January 2006


James M. Fingeroth, DVM, DACVS
Veterinary Specialists of Rochester, NY
Consultant - Veterinary Information Network

Understanding and Explaining Current Concepts
in Canine Cruciate Ligament Disease

Considering that cruciate ligament disease is probably the most common orthopedic affliction we contend with in our canine patients, and that the disease has been recognized for decades, it is surprising how much we still don't understand, and how much our concepts have shifted, especially over the last 15-20 years. It is a subject of some considerable confusion and controversy for specialists, so it is hardly a shock to find veterinarians in general practice that are even more at a loss to cogently understand - and explain to their clients - the intricacies of this problem. My goal in this presentation is to help review some of the concepts and also hopefully assist you in being able to relate relevant aspects to your clients.


In anatomy we learned that the function of the cruciate ligaments was to resist certain forces acting in the knee. The cranial (anterior) cruciate ligament (CCL/ACL) was credited with resisting cranial drawer motion, internal rotation, and hyperextension. It was assumed, based largely on experience with human ACL injuries, that ligament failure occurred as a result of trauma that produced excessive amounts of one or more of these forces. This was classically explained as "the hunting dog that steps into a hole and hyperextends its knee". However, even 30 years ago there were situations that seemed to be at odds with this concept of canine knee injury. These included the large number of "middle-aged, overweight, sedentary" dogs that ruptured their ligaments with minimal trauma while in the house, and the usual absence of clients observing their "hunting" dog twisting or hyperextending its knee in accordance with the theory. In the former case, theories were proposed to invoke a concept of these poorly conditioned dogs having some ill-defined "degenerative" process causing weakening of the CCL. The latter group was conveniently explained away by suggesting that the dogs were running out of sight of the client and so we simply did not witness the actual injury. To some degree there seemed to be support for the theory of a degenerative predisposition based on histopathology studies in the 1960's that looked at supposedly acutely torn CCL's. Pathologists found that there were, in fact, chronic degenerative changes in the ligaments that could not be explained by acute trauma of recent vintage.

Chinks in the armor of the traditional view of canine knee and CCL biomechanics began to appear in the 1970's and '80's, with the increasing recognition of apparent increased incidence of CCL injuries in young, otherwise fit and athletic large breed dogs, especially Labrador retrievers. Many of these dogs were pets, not hunting dogs, and while some injured their knee while romping in a park or field, a good number of them started limping despite having no trauma beyond normal, everyday activity. During this same time reports in the literature appeared that began documenting the occurrence of partially torn cruciate ligaments. We also recognized an apparent increase in the number of dogs with bilateral cruciate ligament injuries.

Seminal studies by Marshall, Arnoczky and others helped to better define the substructure of the CCL/ACL, including the identification of the various bands that comprised the overall ligament. They showed how damage isolated to one band could lead to cranial drawer only being elicited with the knee in a flexed, but not extended position. This helped clinicians understand that the standard cranial drawer test was not always reliable as a determinant of CCL injury if only a partial tear was present. But these studies did not really elucidate the mechanism by which only part of the CCL would fail initially.

Because of these apparent inconsistencies a torrent of theories were advanced to try to explain the findings of CCL injuries without trauma, bilateral CCL injuries, and partial CCL injuries. For example, synovial biopsies in this "new" population of dogs with CCL injury demonstrated lymphocytic-plasmacytic synovitis. This suggested a possible immune-mediated basis for CCL degeneration, ultimately resulting in spontaneous failure - partially at first - then progressing to complete tear, without major trauma. Studies were launched looking for the triggering mechanism. But what these theories failed to take into account were: (1) Could the synovitis seen in dogs with CCL injuries be the consequence of the injury, rather than pre-existing (…which was the chicken, and which was the egg?)? (2) How could we explain that a degenerative process only affects the cranial cruciate ligament, but not the caudal ligament, even though the two share the same anatomic and biochemical milieu? Furthermore, in invoking an immune-mediated etiology, how can we account for a rheumatological disease that only affects the knee joints of dogs, but no other joints? So, even as some of these theories looked promising, a teleologic scratch below the surface certainly seemed to reveal large gaps.


One of the most fertile grounds for human-animal comparative research over the years has been in the study of knee injuries and their treatment. The recognition by physicians that there seemed to be a spontaneous animal model for a common and costly affliction of people stimulated great interest in using dogs as a proving ground for developing treatments. Similarly, veterinarians were able to draw from a vast literature on human ACL injuries to help adapt new concepts to managing the corresponding problem in dogs. This was and remains a rich intercourse between the professions, and many of the operations used in people originated in veterinary surgery, while in likewise fashion, many of the advancements in veterinary knee surgery owe their roots to operations devised for human use.

But it may all be wrong.

Because of the common anatomy of knees across species (and across evolutionary time; just visit the natural history museum and look at a prehistoric animal's knee joints) and the occurrence of clinically significant ACL/CCL injuries affecting humans and dogs, it was long and logically assumed that dogs and people injured their ligaments in the same way. This led, ipso facto, to the conclusion that treatments that work for one species should work well in the other. This has been the prevailing view for decades, but may have blinded us to what may be the truth.

One "blind spot" already alluded to is the absence of clearly defined trauma in most of our canine patients with knee injuries. In people, ACL tears are largely a sports-related injury. And it is usually acute and catastrophic, associated with documented hyperextension or internal rotation, such as so-called "football knee", or tumbling down a ski slope. People with ACL tears usually know the moment it has happened, with no preceding history of problems. Contrast that with most of our canine patients. Not only do most not have an acute trauma history, but most dogs, even those with a seemingly acute onset of signs, have some earlier history of on and off limping. We also have the aforementioned long-recognized finding of chronic degenerative changes in "acute" dog CCL injuries. Dogs (and cats) with truly traumatic knee injuries often have deranged knees, with damage to both cruciates and the medial collateral ligament, similar to some of the more violent knee injuries suffered by professional human athletes. But deranged, three-ligament injuries are rare exceptions in dogs, and most dogs suffer only from cranial cruciate ligament pathology and associated meniscal damage.

Another variance between humans and dogs is the utter lack of "partial cruciates" in people, and, excepting athletes who return to competition, the occurrence of bilateral cruciate problems in people. Moreover, when dogs have bilateral cruciates they are likely to have both fail acutely at the same time or within a very short temporal span, a situation virtually unheard of in humans.

Perhaps the biggest variance between people and dogs is in results after surgical reconstruction. Armed with the concept that the cruciate ligament failed because of an acute, violent over-stressing in extension or internal rotation, the treatment seems obvious: Repair or replace the broken rope. This fundamentally underlies all of the various procedures that have been devised and refined over dozens of years in both human and veterinary orthopedic surgery. We shall term these surgeries as "conventional" or "traditional". As we know, there are a multitude of different techniques that vary in the type of material used as a cruciate substitute (since the ligament is almost always shredded beyond primary repair), where and how it is placed in or on the joint, and how it is anchored. But all these conventional/traditional procedures share the goal of reproducing the function provided by the ACL/CCL, particularly the elimination of cranial (anterior) drawer. This includes all the various intra-articular reconstructions (Paatsama; Over-the-Top; Vascularized Patellar Tendon Graft; etc.) and extra-articular repairs (DeAngelis-Flo lateral suture; Fibular Head Advancement; etc.).

In people with conventional reconstruction and appropriate physiotherapy it is possible to restore most patients to normal function without limitation. It has become commonplace to see star athletes suffer devastating injuries and yet return to their previous level of competition after knee reconstruction. The same cannot be said for our results in dogs. Study after study evaluating the various procedures used in dogs has consistently identified the same ceiling of results: About 80-85% of dogs will have good to excellent function after surgery, almost regardless of which type. Intra-articular repairs, touted so long for being more physiologic and closer to restoring the natural biomechanical functions of the CCL, actually fare poorer in such studies. Yet with all the attempted refinements in material choice, placement, and anchorage the results never seem to improve. We still see a consistent 15-20% of knees that have less than good long-term results. Moreover, even as a given surgeon uses the same technique repeatedly in hundreds or even thousands of knees, the results do not improve. More consternating, that same surgeon is unable to predict which knees will do well and which won't. There is no correlation between time from onset of signs until surgery, the degree of arthrosis present at the time of surgery, the size of the dog, or any of a myriad of other factors that one might presume would influence outcome. Even an individual dog with bilateral CCL injuries and identical reconstructions might have very different results from one side to the other. Furthermore, how we define "good" and "excellent" may not be demanding enough, so overall results with traditional surgeries might even be less good than we give ourselves credit for. For example, should a dog that doesn't limp, but who can't sit squarely, be considered to have had an "excellent" outcome? If a dog runs and plays after CCL reconstructive surgery, but doesn't have normal stamina, or limps a bit after heavy exertion, would that be considered a "good" result?

"Anatomy is not the whole answer"

The similarity between knees in humans and dogs (and other species), and the shared problem of cruciate ligament injury has long led us to presume that the mechanism of injury and treatment should be the same. Yet our results with surgery in dogs have generally been inferior to those in people. Even dogs with good clinical results will have recurrent laxity - drawer - within a few months after surgery, regardless of material used for repair, intra- or extra-articular placement, and despite being stable and tight at the end of the operation. And the knees that are stable long-term often achieve that stability not by virtue of the cruciate ligament substitute, but by a process of generalized fibrosis around the knee. This scarring may make the knee stable, but does so at the expense of normal range of motion.

What has been occurring over the many years we have been treating knee injuries in dogs is a classic case in science of missing the obvious until looking with "new eyes." While canine and human knee anatomy is very similar, our biomechanics differ greatly. Start by looking at the differences in posture: Humans stand plantigrade (heels on the ground) and our knees are essentially straight during weight bearing. Contrast that with dogs who stand digitigrade (hocks elevated), with their knees bent during weight bearing. The significance of this difference is that there are different force vectors in the canine knee than in the human knee. Because of the complexity of movement in knees and the number of structures that contribute both to motion as well as restraint of motion, engineers are unable to identify and quantify all these forces accurately, especially for different angles of flexion and extension. However, bioengineers do agree that there are shearing forces in dog knees that tend to produce a vector of force that moves the tibia cranially relative to the femur. The engineers may disagree over the exact cause for or magnitude of this force vector, but there is concurrence that it exists. For the purpose of our discussion we will refer to this forced vector by its most commonly known moniker, cranial tibial thrust. In a normal canine knee this shearing force is counteracted by the large muscles of the leg, notably the quadriceps group and hamstrings. However, in situations where the leg muscles are unable to resist this shear the last line of defense, Mother Nature's "emergency brake" if you will, is the cranial cruciate ligament.

The concept of the CCL as a final restraint on cranial tibial thrust casts the ligament in a whole new light. Remember that cranial drawer is a passive laxity. It occurs when the leg is non-weight bearing, and usually only in the hands of an examiner. Cranial tibial thrust on the other hand occurs during active weight bearing or when weight bearing is simulated via the tibial compression test. So the question arises: What happens if a dog consistently has more shear (tibial thrust) than its muscles can counteract? Put another way: What happens if the CCL is called upon to counteract thrust every time the dog bears weight? One can imagine the CCL sustaining a slight "jerk" every time the dog steps on its leg in this situation. If the stretching and microdamage that ensues is greater than the material strength of the ligament, or exceeds the body's ability to self-repair the internal microdamage, the ligament will start to suffer material failure. Engineers refer to this as cyclical loading ("cycling"), fatigue failure, and plastic deformation. The clinical term we most often use is "repetitive stress injury". So, could CCL injuries in dogs be nothing more than repetitive stress? It seems plausible, and would explain many of the observations we noted above (failure with normal activity, presence of chronic changes in supposedly acute injuries, etc.). The obvious question to ask is, "What distinguishes a normal dog's knee from one that is predisposed to failure from repetitive stress?" There are no clear answers to this. I have two theories. One is that there is a genetic predisposition to CCL failure in some dogs as a consequence of muscles being inadequate to limit cranial tibial thrust. If you will, their muscles are "too small for their knees". This certainly makes sense when one considers the abnormally high incidence of CCL injuries, and bilateral CCL injuries, in certain breeds such as Labrador retrievers and Rottweilers. The second pertains to the classically recognized "overweight, sedentary, middle-aged" dog. It is not hard to imagine that these dogs have poorer muscle tone and function, and combined with obesity, they constantly yank on their cranial cruciate ligament. In either case the primary problem is not in the ligament, but in the muscles that are failing to "protect" the CCL from repetitive over-stress.

The concept of CCL injury as an accumulation of damage from normal wear and tear, as opposed to the result of catastrophic trauma, has dramatic implications with respect to treatment. As discussed above, all of our classical, conventional surgical procedures are focused on eliminating drawer sign - passive laxity - that occurs after the ligament has failed. But if the new paradigm is correct, the unchecked continuation of tibial thrust/shear should be expected to lead to - what? It will lead to gradual loosening of such reconstructions, which is exactly the clinical finding we have long been plagued by! This is the beginning of an understanding of some of the inherent deficiencies in our traditional surgical methods. If we replace the old CCL with a new "rope" (and regardless of the substitute, most would acknowledge that nothing is materially as good as he original CCL in the first place) but subject it to all the same (shearing) forces the old ligament was affected by, it is no wonder that our repairs seem not to hold up well over time.


The understanding of cruciate ligament deficiency with an active force model as opposed to a passive laxity model also helps suggest new treatment strategies. If the fundamental problem is tibial thrust (or some equivalent shear force), it should be possible to make the knee functionally stable during weight bearing by elimination of this force. A review of physics and trigonometry (which we will not undertake here) shows that force vectors can be altered by changes in geometric angles. These new concepts share the idea that if shear forces are eliminated there is no longer need for a cranial cruciate ligament. It becomes a moot structure.

There are two angles that are currently the subject of alteration: One is the angle known as tibial slope, the angle between the tibial plateau and the long axis of the tibia; The other is the angle between the tibial plateau and the straight patellar tendon in its course between the patella and the insertion point on the tibial tuberosity.

The pioneer in recognizing this new paradigm was the late Dr. Barclay Slocum. He was inspired by reports describing cranial tibial thrust. In the description of thrust the authors focused on a new test - the tibial compression test - as an alternative to the drawer test in diagnosing cruciate ligament deficiency. But what caught Dr. Slocum's attention was the thrust itself, and its relation to active weight bearing. In his first attempt to address this phenomenon Dr. Slocum devised the (closing) tibial wedge osteotomy (CTWO) to reduce the angle of tibial slope. Later on he revised his procedure with the development of the curved tibial plateau leveling osteotomy (TPLO) and associated instrumentation. He metaphorically compared the relationship of the femur to the tibia as a "wagon on a hill", where the cranial cruciate ligament holds the femur (wagon), and the caudal aspect of the medial meniscus acts as a backstop. With disruption of the ligament, the wagon can freely roll down the hill as the tibia thrusts forward, and he accounted for meniscal damage, so common with CCL tears, as the result of crushing by the now unrestrained femur. By leveling the tibial plateau (elimination of the downward angle), the femur now has no tendency to roll backwards as the dog bears weight, and the tibia no longer slides forward since thrust is eliminated.

A key point to bear in mind with this theory is that the tibial slope itself is not necessarily pathologic. Dog breeds with normal knees (ie, no tendency to have cruciate injuries) have tibial slope that is similar to the slope measured in predisposed breeds. There is also no demonstrable difference between slope in affected knees and unaffected knees. Thus, one cannot use measurement of tibial slope to predict cruciate ligament injury, and hence to consider prophylactic treatment. On the other hand, some dogs do seem to have excessively high slope angles, but it is not yet known if this alone predisposes these knees to CCL failure. But once the ligament has failed, even "normal" slope will satisfy the Slocum model of the "wagon on the hill."

An alternative concept in "geometry modifying" surgery, developed by Pierre Montavon and Slobodan Tepic in Switzerland, is tibial tuberosity advancement (TTA). Their approach to eliminating shear is to osteotomize the cranial aspect of the tibial tuberosity (and insertion site of the straight patellar tendon), and use a titanium "spacer" to advance the bone cranially. This results in creation of an idealized 900 angle between the tendon and the tibial plateau, and elimination of shear.

With the concept of inadequately restrained cranial tibial thrust being the progenitor of CCL injuries in dogs gaining increased acceptance, new innovations for modifying anatomy to reduce this force vector have been appearing in the literature. One presumes that as more questions are raised, and more knowledge is gained, more new procedures and modifications of these existing procedures will yet appear.

One area of confusion with these newer procedures is that none are designed to eliminate the drawer sign that all of us have long relied on as the pathognomonic sign of an unstable knee. This is not an oversight. When dogs are weight bearing they do not have drawer; they have tibial thrust. Geometry modifying operations are deemed successful if tibial thrust - as measured by the tibial compression test - is eliminated. The knee is then considered stable in the active force model. None of these operations currently incorporates a substitution for the CCL to reduce drawer. It is thought that this is unnecessary since drawer only occurs in a non-weight bearing situation. Whether this thinking is correct is not proven. Might there be an advantage to combining geometry modification with passive restraint procedures? This might be an area for research, but as yet has neither been described nor studied. We do know that meniscal injuries can still occur after any knee reconstruction (conventional or geometry modifying), and can account for a recurrence of lameness and need for revision surgery. The potential for drawer, which is the equivalent of pushing the wagon along the tibial plateau (whether sloped or level), suggests how the caudal pole of the medial meniscus remains at risk even after TPLO or TTA. How to prevent this late sequela remains undetermined and controversial.


Probably the most important statement to make at the outset, and to convey to clients, is that we are not in a situation where we have a "good" operation and a "bad" operation. Both traditional passive laxity type surgeries and geometry modifying surgeries have significant track records of success, and it can be stated unequivocally that any surgery is better than no surgery when it comes to CCL injuries in dogs. The question is whether one is better than the other in getting more of our patients closer to the ideal of 100% restoration of normal function, and whether the results of newer procedures are more predictably good than the conventional procedures. This is especially important since in most institutions surgeries like TPLO and TTA can cost about 50% more than a conventional reconstruction.

The answer of course lies in "evidence based medicine", and the gold standard for EBM is the prospective, double-blind, controlled clinical study. The required study would take a population of dogs presenting with CCL injury and randomly assign them to treatment groups, either conventional surgery or geometry modifying, and the follow these dogs for a specified number of years. Costs would be kept identical to remove any client bias in measuring satisfaction. Ideally those who evaluated function post-operatively would be "blinded"(which could be done for the physical exam by producing identical skin incisions in both groups, but would be difficult for radiographic exams for obvious reasons). Assuming analysis of variance proved that no biases were incorporated in assignment to groups, and that each group had a similar composition with respect to breed, age, gender, duration and type of pathology, and follow-up time, it should be easy to answer the question of whether, theoretical considerations aside, TPLO or TTA is clinically superior to conventional reconstruction.

To date no such study has been reported, and given the nature of veterinary clinical trials and funding, it is doubtful that such a large scale study will ever be completed.

One comparative study out of Iowa that was published in 2005 looked at a cohort of dogs receiving either a lateral suture-type conventional repair with dogs having TPLO. This study purported to show no advantages for the TPLO. However, this study had some inherent design flaws that make its conclusions suspect. Chief among these was the failure to randomly assign dogs to one treatment group or the other. They also had a very limited (6 month) follow-up period. Finally, their data were based to a large degree on the use of force plate analysis of gait. This theoretically objective means of measuring how much weight a dog bears on its limb has come under some scrutiny - and criticism - as a result of the methodology itself, and similar studies where force plate-derived data were at odds with other observations.

While we wait for definitive, objective studies to appear, the best evidence we have, albeit one of the weakest forms of EBM, is the clinical impression of surgeons and clients. There seems to be a consensus that properly performed TPLO and TTA does result in better outcomes, with movement of numbers from the 80-85% "good to excellent" range up to the 90-95% range. Moreover, after observing dogs that have had TPLO and TTA the very definition of what constitutes "good" or "excellent" seems to be shifting. Among the advantages claimed for these geometry modifying procedures over conventional surgery are retention of more normal range of motion, dogs sitting more squarely, reduced progression of secondary degenerative joint disease, better recovery of full athletic function, faster return to weight bearing after surgery, and lower incidence of recurrent problems necessitating reoperation due to loss of stability or late meniscal tear.

There are a myriad of other, mostly technical controversies with regard to some of the specific elements of these geometry modifying surgeries, akin to debates with conventional procedures (like the lateral suture technique with respect to ideal orientation and anchorage of the suture substitute). We will not dwell on these here. However, there are a few controversial areas that are more than technical so we'll briefly address those:

What is the Role of Arthrotomy?

Just as the advent of TPLO and TTA has made it apparent that dogs don't require a functioning cruciate ligament so long as active instability has been eliminated, the question arises as to whether the remnants of the torn native CCL require debridement? Most of us learned, at least in some vague way, that one of the goals of conventional surgery was to "clean up" the knee. Removal of the torn CCL was seen as integral to the success of the operation, and there was an implied (or sometimes explicit) suggestion that the shreds of torn ligament served as incitement for ongoing inflammation within the joint. Like many of you I have spent many hours of my professional life diligently carving out every last shred of CCL because of this assumption, and it was considered near gross negligence not to do this when treating canine knee injuries. But it is likely that this is untrue. The strongest evidence for the unimportance of CCL debridement is derived from our experience in operating knees with very chronic CCL deficiency. In many of these dogs there is hardly any remnant of the torn CCL left at the time of surgery, suggesting that there is a spontaneous resorption process that takes place. The corollary to this is that dogs that have stabilizing surgery without CCL debridement have outcomes no different from those in which the remnants have been painstakingly removed.

However, a less controversial aspect centers on the concern about meniscal pathology. There is little debate that tears of the medial meniscus constitute a major contributor to clinical lameness, that dogs with untreated meniscal injuries have more rapid and dramatic degenerative joint disease, and that failure to address meniscal pathology, or the later occurrence of meniscal injury, is a major cause for knee surgery failure and need for further operation. So, whether one chooses to remove remnants of CCL or not, it is imperative that the joint be explored for meniscal lesions at the time of stabilizing surgery (regardless of method), and any meniscal trauma be addressed. There is some controversy about whether it is better to perform partial meniscectomy vs. total meniscectomy, although most data support preserving as much normal meniscus as possible.

There is also some debate as to the best method for examining the interior of the joint to assess the meniscus. Probably the best method is arthroscopy, since it is minimally invasive, allows more physiologic appreciation of joint structures in their natural liquid environment, and provides magnified views of structures. Arthroscopy has proven to be the most sensitive test for finding pathology that might otherwise be missed during standard open arthrotomy. Unfortunately, arthroscopy is not available in every institution, and it takes a good deal of experience to become a good knee arthroscopist.

In conjunction with his promulgation of the TPLO surgery Dr. Slocum also described a mini-arthrotomy made just caudal to the medial collateral ligament. This approach was felt to provide adequate visualization of the medial meniscus for identifying trauma, and performing any needed meniscectomy, and shared with arthroscopy a minimally invasive nature. Some surgeons are more adept and comfortable with this approach than others. Many who have tried it have since abandoned it for a return to a more traditional parapatellar arthrotomy. Ultimately, the choice between arthroscopy, mini-arthrotomy, or standard arthrotomy is more a matter of surgeon's preference and is not correlated with quality of outcome.

What Should be Done With Partial Cruciate Ligament Injuries?

This is a question that can be asked at two times: Before surgery and during surgery. As we have become increasingly aware of partial CCL injuries being a cause for waxing/waning lameness in dogs, and as we have become more adept at making this diagnosis (cranial drawer sign only with the knee in flexion; subtle radiographic changes; positive tibial compression test, positive joint tap, clinical suspicion even in the absence of any objective evidence), we must decide what the appropriate treatment is. It is quite common for such patients to be treated with rest, NSAIDs, and neutraceuticals, often for prolonged periods. Since there is no reporting in veterinary medicine there are no data to show whether, and how many dogs so managed eventually recover and have resolution of their lameness. My impression is that most dogs do not recover, and in fact progress over time until they are either markedly affected or suffer an acute lameness in conjunction with either rupture of the last remnant of CCL and/or tearing of the meniscus. Once the process of CCL degeneration and structural failure has been instigated it is difficult to reverse.

I think there may be some dogs that can be managed non-surgically if they are caught early (at least before there is measurable instability or arthrosis), and can undergo a rigorous program of physical therapy to strengthen their quadriceps and hamstring muscles. If the dog's muscles can be recruited to restrain cranial tibial thrust it seems plausible that the injured CCL might heal. This approach is also used sometimes in humans with ACL tears; the patient may have demonstrable drawer signs when sitting, but have a stable knee when weight bearing, and thus not require an operation. However, the majority of dogs will either not be good candidates for non-surgical management, or will fail with such management. Therefore, a diagnosis of partial CCL tear should prompt consideration of surgery, even if there is no dramatic instability.

If a dog is undergoing surgery, be it conventional or geometry modifying, what should one do with that part of the CCL that is still grossly intact? The short answer is to recognize that "grossly" intact does not necessarily mean functionally or histologically intact. The surviving portion may be quite damaged and incompetent, a fact sometimes easily demonstrated by forcing the knee into cranial drawer with an instrument levered caudal to the tibia, and observing the "intact" portion of the CCL pop apart or at least stretch unnaturally. There is probably no specific need to excise it (see above), but simultaneously one should not assume it is going to be of much use to the dog, or to in any way "skimp" on the reconstruction you were already intending on doing. With parapatellar arthrotomies the CCL acts as an impediment to visualizing the medial meniscus, and for that reason should be removed as a matter of course, in my opinion.

What is "Meniscal Release" and Should it be Part of Treatment?

The recognition that meniscal pathology is a major player in producing more severe clinical signs and worsened arthritis, combined with the recognition that meniscal injuires sometimes occur after knee reconstruction surgery, leads surgeons to speculate whether anything can be done prophylactically to reduce the incidence of such late meniscal tears and the subsequent need to reoperate a knee.

We earlier described Dr. Slocum's "wagon on a hill" metaphor and his identification of the caudal pole of the medial meniscus as the "backstop" that gets crushed when the femur rolls over it. With less florid imagery other physicians and veterinarians have long recognized the association of meniscal injuries with CCL injuries, and have looked for anatomic explanations. It is well known that medial meniscal injuries are far more prevalent than lateral meniscal injuries. The latter had hardly been recognized at all until the advent of arthroscopic surgery in dogs. The explanation for this discrepancy has been linked to the differences in the attachment of the caudal poles of the two menisci. The lateral meniscus has a caudal meniscofemoral ligament such that the caudal pole actually sits slightly off the tibial plateau, and moves cranially and caudally with the femur during any tibial thrust or drawer. The medial meniscus, on the other hand, is bound down by a caudal meniscotibial ligament, as well as a strong attachment to the medial collateral ligament. When the femur moves caudally relative to the tibia the caudal pole of the medial meniscus is fixed, and thus can get trapped by the femoral condyle in a shearing and crushing fashion.

Meniscal release was Dr. Slocum's proposed solution to eliminating the vulnerability of the medial meniscus. He recognized that even with TPLO one could not ensure the elimination of all translational motion between the femur and tibia in all joint angles and in all phases of stride and weight bearing. Joint biomechanics are far too complex for any single alteration to account for every facet of motion. His concept was to either transect the caudal meniscotibial ligament (allowing the caudal pole to retract caudally "out of the way" when the femoral condyle rolled past), or alternatively to transect the caudal horn of the medial meniscus itself allowing similar mobility to mimic the situation with the lateral meniscus. The choice of where to perform the release (ligament or meniscal horn) is dependent on the type of arthrotomy used. Meniscal release can be incorporated into either a conventional reconstructive procedure or a geometry modifying procedure. However, recent studies done in Germany have suggested that meniscal release may be ineffective at reducing the incidence of late meniscal tears, and moreover, may, like a spontaneous meniscal injury, actually result in increased arthrosis in the knee. This remains a controversial area and some surgeons are now abandoning the practice, while others continue to employ it.


A recent study concluded that CCL injuries in dogs accounted for over $1 billion of cost each year in the U.S. (Wilke VL, Robinson DA, Evans RB, et. al.: Estimate of the annual economic impact of treatment of cranial cruciate ligament injury in dogs in the United States. J Am Vet Med Assoc 227:1604-1607, 2005). This is a boon to veterinarians, but also places upon us an ethical burden to do all we can to learn about this disease and its management, continue to make strides in prevention, and to try to objectively determine both the most efficacious and cost-effective means for treatment.

The revolution that was touched off in 1983 when Slocum first described CWTO for treatment of cranial tibial thrust, through the late '90's and first part of the 21st century when TPLO became the new gold standard for treating CCL injuries (spawning vast amounts of research into the particulars of the TPLO itself, as well as all the new geometry modifying procedures that have and continue to follow), should remind us just how tenuous any theory of disease is, and prompt us to always mix skepticism with our enthusiasm, whether we are evaluating established dogma or appraising the newest mousetrap. Our particular challenge is to sift though all the shades of grey that objective data provide and yet somehow distill it down to black and white answers for how we will practice, and how we will educate and inform our clients. This is no easy task, but I hope this synopsis assists you in the process.

To quote Andre Gide (1869-1951):
Believe those who are seeking the truth.
Doubt those who find it.

Practitioner's Perspective on Oncologic Surgery

We are frequently presented with lumps, masses and syndromes related to tumors on a patient's surface or internally. Since most cancers are still best managed by either surgery alone, or by surgery as the most important arm of treatment, it behooves all of us to have a good idea of how to best manage patients with tumors with which we are presented.

To be an effective surgeon for patients with tumors it is imperative that one have a sound understanding of tumor biology for various classes of tumor. Knowing the expected behavior of a tumor greatly assists in guiding therapy, including the role for surgery, its timing, and the specific requirements of the operation.

As a general rule we are going to find ourselves in one of two situations: Either we will know in advance what type of tumor we are dealing with, or we will only know that there is a tumor, but not know what type. Examples of the first situation might include such things as functional endocrine tumors with very specific syndromes (egs, primary hyperparathyroidism, insulinoma, adrenocortical tumors with Cushing's syndrome), or masses for which we have a cytologic or histologic diagnosis already established. When in this situation plans can be made immediately for comprehensive treatment based on an appropriate staging work-up.

The second situation presents a little bit more of a conundrum. For example, a skin mass may be innocent/benign or something quite aggressive, and its gross appearance may be misleading. How should we proceed? The most basic answer is to aspirate or biopsy, or both. It is outdated to think that the act of aspirating or biopsy alone will induce more aggressive behavior by the tumor. The biggest risk to doing an aspiration or biopsy comes from the potential to injure vital structures (eg, an intrathoracic mass) or to cause hemorrhage (eg, suspected thyroid tumors). The other consideration is whether to do an incisional biopsy or excisional biopsy. The latter has the potential advantage of being both diagnostic and therapeutic, and hopefully obviates the need for further surgical treatment. However, excisional biopsy may be hampered because of proximity to other vital structures, inability to close the resultant wound in primary fashion, and uncertainty of how much grossly normal margin is required to achieve the local cure. Incisional biopsies can be performed by actually removing a wedge of tissue, or via such instruments as biopsy needles and dermal punches.

Besides having an understanding of tumor biology and expected behavior to help guide surgical planning, it is always important to take a "giant step back", and consider the issues pertaining to the client and the patient as a whole. Failing to do this appropriately can lead us to either over-treat or under-treat. Perhaps our most important role is in making judgments about the patient's and client's quality of life - both as they are at the outset - and as we expect they might be with various treatment options (including NO treatment). Yet even as we make such judgments we have to guard against making overly biased choices based on our perceptions. This is especially apropos when dealing with aged animals, aged clients, or clients who we think may have financial limitations.


When we suspect a patient has a tumor we need to consider getting the following information:
  1. Tumor type
  2. Tumor grade (if appropriate)
  3. Clinical stage of disease
  4. Presence of related or unrelated medical problems that might impinge on decision-making
As mentioned above, tumor type will be determined based on cytology, biopsy or clinical syndrome, or some combination of these. Cytology is useful because samples can be obtained quickly and easily, without much preparation or anesthesia, and results are generally quicker than with histopathology. But we need to bear in mind that the quality of results are dependent in such factors as sample acquisition, slide preparation, the abilities of the cytologist, and the nature of the underlying tumor to "reveal" itself. For some diseases cytology can be diagnostic (many mast cell tumors and lymphomas fit into this category), but even in these instances there can be misinterpretation. Cytology is also not generally useful for providing information about tumor grade, except for certain special staining techniques (such as AgNOR staining of mast cell tumors). For most cytological preps the examiner will be doing well to distinguish whether cells have criteria for malignancy or not, but not necessarily establish a diagnosis. So cytology is a good first step, but formal biopsy should be done concurrently or should follow if the cytologic diagnosis is neoplasia. To quote a famous observation: "Cytology is like looking at a few bricks and trying to tell what a whole wall looked like, whereas biopsy is looking at a representative section of the wall".

One of the considerations in trying to decide between incisional vs. excisional biopsy is what options exist if I don't remove the entire mass at the outset? This frequently comes up when dealing with suspected bone tumors. I will reiterate that it is never wrong to do a limited, incisional biopsy, but there are instances where it may either cause no more morbidity to remove the entire mass (eg. certain skin masses), or, regardless of the diagnosis, there are no good alternatives to more radical surgery (eg, an osteodestructive proximal humeral mass with pathologic fracture where nothing short of amputation will palliate the patient irrespective of whether it's cancer or not).

It is not uncommon for excisional biopsies to have "dirty" margins. This is essentially always the case with the various soft tissue sarcomas (STS), and is a frequent finding for mast cell tumors (MCT). The key is having some clinical suspicion, based on appearance or cytology. If one suspects a STS (such as hemangiopericytoma, nerve sheath tumor, fibrosarcoma, etc.), it might be best to first do an incisional biopsy to establish the diagnosis and grade, and then plan for more comprehensive treatment with wide excision and adjuvant radiation. If a MCT is suspected, and not too large or inflamed, an excisional biopsy can be curative if sufficient margins are obtained. For many years the standard of care was to achieve 3cm margins horizontally and one fascial plane deeply. Newer research suggests that for most tumors 2cm margins are sufficient. This translates into many more of these lesions being amenable to excisional biopsy. In any case, excisional biopsy results that indicate dirty margins create a responsibility to follow-up with either more surgery, or to use adjuvant therapy, depending on the specific diagnosis.

Clinical staging is important since it helps us and the client decide about prognosis as well as how best to treat. Staging work-ups will vary to some degree based on the type of tumor and its expected biologic behavior. Since many tumors have a propensity to metastasize to the lungs, thoracic imaging is frequently part of this evaluation. However, certain tumors such as MCT's have a very low pulmonary metastatic rate, so the value of thoracic imaging for patients with this disease may be more limited. Similarly, knowing that certain tumors may produce paraneoplastic syndromes or metastasize to locations other than lungs will guide us in which tests to perform. So, in addition to thoracic imaging, we may find ourselves looking at certain specific blood or urine tests, bone marrow, and so forth.

Beyond the specifics of clinical staging for the tumor we also need to evaluate the rest of the patient with respect to other potentially complicating medical problems. Many of the tests done in this regard will overlap with those done as part of the staging work-up. But it is important we determine whether the animal has other issues that may impinge on our management of the tumor. It is here that we need to remind ourselves and our clients that "age is not a disease". If the patient happens to be older, but is otherwise in good health besides its tumor problem, it is entirely appropriate to offer treatment. Remember our fundamental role for the client is not necessarily to cure or even treat their pet; rather, it is to be a source of reliable information upon which they can base their decision about whether to proceed or not, and how.


Armed with the above information we are in a position to discuss whether surgery can or should play a role in treating the patient. As with all interventions (both diagnostic and therapeutic) we must analyze and weigh the potential good we seek to do against the harm we might do. These analyses are rarely black and white, but it is important to explain as much as we can to the client and involve them in the decision making process. We neither want to "sell" the client a procedure nor deny a patient a chance. As long as the client is given good information upon which to make a choice, and they cannot later on claim that they were given either too optimistic or too pessimistic an assessment by you, you are less likely to have a disgruntled client regardless of the ultimate outcome. This is one place where referral to an oncologist and/or surgeon can be very valuable. You may intend to do the requisite treatment yourself, but by suggesting referral for a second opinion, not necessarily for the specialist to take the case over, you ensure that the client and you have properly assessed all the factors before a decision has been made.

Similarly, whether surgery should be done, how it is done, and by whom is predicated on such factors as level of training, experience, equipment, willingness and ability to deal with complications, and ability to monitor the patient appropriately. This is a difficult decision for you and for your client. On one hand, you are motivated by the personal reward factor. Surgery for many veterinarians is "fun" to do, even if not particularly profitable. It is also a place where we tend to measure our egos, and feel that we can handle things that may, in fact, be beyond our expertise. On the other hand is the client who has faith in their primary care doctor, and may neither be aware of specialists nor want to have to go to a "stranger" for their pet's care. The combination of these two factors is a powerful motivation for the generalist to do oncologic surgery that might get them in over their head, or at least have a suboptimal result. We all should be willing to recognize these inherent biases, and actively ask ourselves, as objectively as we can, if we are really the best choice for surgeon? In some cases the answer will be a resounding yes, in other cases a clear no, and for some it may be uncertain. In all cases, again, it can be valuable to solicit the input from a specialist to discuss the matter. That discussion, whether via calling the specialist and talking about the case, or via formal referral for a second opinion, will raise questions and issues that will usually make it clear to everyone, you, the specialist, and your client, what the best answer is. In many cases such input from the specialist will serve to bolster your confidence and you can proceed with the operation. In other instances it may make clear a whole list of issues and concerns that you might not have considered before, and consequently convince you that you ought not be doing the procedure yourself. In either instance the needs of the client and patient are put first and foremost, as they should be.

Among the factors that have to be considered when planning an operation are: Is the tumor operable (ie, can it be removed entirely or debulked substantially without causing undo harm to the patient)? What is the best approach? What will the requirements for closure be? Physical diagnosis and routine radiographs may be insufficient for determining the operability of certain tumors. Examples include thoracic, abdominal, pelvic, head and neck, and certain appendicular tumors. Other imaging modalities can be very helpful in these situations. Ultrasound has become commonplace and is of great value for abdominal evaluation especially. Yet even in the abdomen ultrasound can be deficient in providing the surgeon with a clear, three-dimensional understanding of the tumor and its relationship to other viscera. Of increasingly common use in veterinary surgery are the cross-sectional imaging modalities, CT (computed tomography) and MRI (magnetic resonance imaging). I personally find these of great value for treatment planning, and in not a few instances in telling me (and the client) before doing any "exploratory" surgery that surgery will not be helpful. The biggest downside to CT and MR is their cost, which can represent a substantial percentage of the cost of an exploratory surgery. But here again it is a matter for the oncologic surgeon to first consider whether such imaging will be particularly useful (in some instances there is enough information to make CT or MR moot), and if it is, to discuss with the client the cost/risk of such imaging relative to the benefit derived.

The challenge for many thoracic and abdominal tumors relates to anesthetic concerns, availability of assistance for retracting, etc, having proper instruments for handling the intended operation and any "contingencies" that might arise, and managing the patient's post-operative needs. But it is critical that these factors be considered. Although one may be very familiar, say, with doing celiotomies because of experience with operations such as overiohysterectomies, cystotomies, or intestinal foreign body removals, it can be a very different story going into an abdomen with a tumor. And it is an exquisitely uncomfortable situation to be in the middle of an operation and feel like you're in over your head. Moreover, lack of familiarity with tumor management or having some specialized instruments might lead you to think the situation is untenable, and have you counsel euthanasia, when it might in fact be a more treatable condition.

For tumors on the body surface, the biggest challenge most face is creation of a large defect that cannot be closed primarily by suturing the edges together. There are a couple of key things to keep in mind: First, one needs to consider how much resection is required based on the tumor, and not based on "not wanting to make a hole so big I can't close it". Pre-planning appropriate plastic surgery techniques such as axial pattern flaps, random flaps, and grafts allows one to proceed with resection more liberally. Paying attention to lines of skin tension when planning incisions is also very helpful in avoiding creation of non-closeable wounds or wounds closed under tension (and doomed to dehisce). Dead space and flaps almost always demand incorporation of drains, preferable closed suction type devices. If one has determined before incising that there is enough skin mobility to close the wound primarily it is helpful to use a sterile marking pen to outline the cut before starting. Once the skin is incised it will tend to retract, and it can be easy to get off your proposed incision line. Moreover, once the skin retracts it will seem that you created too large a wound to close. But if your planning was correct, you should be able to draw the edges together progressively without undo tension,

One of the biggest problems I see in referral practice is with subcutaneous masses, and also masses in the deep cervical region. The tendency too often is to make an incision directly over the mass, without regard to underlying anatomy or lines of tension. For cervical masses, even those that seem very lateralized, it is almost always more appropriate to approach these via a standard midline approach rather than directly over where you palpate the mass, unless all you are doing is an aspirate or needle biopsy. By not going in on midline you almost invariably find that you can't remove the mass, or that there is a lot of important stuff with names in the way, and the patient ultimately comes to a specialist for a second operation. This second operation becomes more difficult because now there is fibrosis and adhesion of tumor to the first, non-midline scar that requires more extensive and difficult dissection. By going in on midline the first time you have the advantage of more familiar anatomy, easier retraction of vital structures, and a far better chance for complete tumor removal.

Oral tumors often have more extensive bone involvement than is apparent from looking only at the gingival component or even radiographs. Few of these tumors are cured by local resection of soft tissues alone, or even extractions of involved dentition. Most will require some degree of mandibuectomy or maxillectomy for effective control, but fortunately most dogs and cats tolerate such surgery amazingly well. CT is an excellent modality for determining the full extent of the tumor and for achieving clean surgical margins.

Tumors of the nose, skull and orbits can rarely be appreciated in their full extent without cross-sectional imaging. Very typically any visible tumor or bony deformity is the tip of the iceberg, so if surgery is contemplated, either CT or MRI should be part of treatment planning.

Tumors of the brain and spinal cord may in some instances be amenable to surgical treatment. The greatest success with brain tumor surgery has generally been with convexity tumors (ie, those on the surface of the cerebral hemispheres). Most of these are meningiomas, and in cats especially they tend to be benign (in dogs they have a higher propensity for being aggressive and invasive). Success has also been seen with tumors of the olfactory lobes. The current results of surgery for more deeply seated brain tumors, tumors in the cerebellopontine angle, and pituitary have been more mixed. Similarly intramedullary spinal cord tumors are difficult to remove without leaving the dog paralyzed distal to the lesion, although sporadic successes have been achieved. Intradural but extramedullary tumors have a higher success rate. Extradural tumors of the spinal cord may be amenable to surgery, but much depends on involvement of surrounding bone and muscles.

One of the "luxuries" enjoyed by our physician counterparts is the frequent and deliberate division of labor between a team of surgeons charged with removing a tumor and another team responsible for reconstruction. Not having to be the surgeon who also has to close the wound allows for more boldness in resection based on the requirements for good tumor control. But with due consideration of the needs for the former and the latter, veterinarians can still be effective oncologic surgeons.


One of the first things that must be borne in mind is that a tumor that seems to us "inoperable" might be, in fact, successfully managed by someone who possesses greater skills, more experience, more technical support, or a combination of these. For the general practitioner this may mean the local specialty hospital and board-certified surgeon. But even a board-certified surgeon may, at times, recognize that a given case might best be treated by someone more specifically engaged in oncologic/reconstructive surgical practice than he or she. One of the seminal pioneers in veterinary oncologic surgery, Steve Withrow at CSU, is famous for asking new residents and fellows at the start of their rotation on his service to make a list of "things that cannot be removed from a dog or cat [without causing intolerable morbidity or death]". At the end of their rotation the list is reviewed, and invariably it is reduced to zero or near zero. On the other hand, that we sometimes can do an operation to prolong life or save some body part does not always mean we should. Here again it is incumbent for the surgeon and client to work together as a team and have frank discussions about what each desires, what each is capable of, and ultimately what is in the best interests of the patient.

For most clients any major surgery will likely be the only surgery they will accede to. Put another way, if you diagnose a dog, say with a large abdominal mass, and elect to explore it yourself, it is not likely that the client will agree to you closing the dog up and then referring it for someone else to operate on because what you encountered was beyond your skills or equipment. Such "peek and shriek" operations almost inevitably end up with euthanasia during or shortly after surgery. So you must really think hard about what you might find, what you think you can handle, and what you cannot. In some cases this will result in you referring away a patient that you may well have been perfectly capable of handling yourself. So be it, since the outcome is still the best for the patient, and it is certainly preferable to putting to sleep a dog or cat that in your heart of hearts you know may have been helped more elsewhere.

For abdominal tumors it is important to be disciplined and methodical in your approach. First off, be sure to prep an adequate surgical field. If a liver or other cranial abdominal mass is involved there can be insufficient room to visualize or operate because of the constraints of the rib cage, especially in narrow/deep-chested dogs. You should be prepared to extend your celiotomy incision into a caudal median sternotomy for exposure, which also means you and your staff need to be prepared to manage a thoracotomy intra-operatively and postoperatively. In some cases a paracostal extension from the xiphoid must be made to flap down the cranial abdominal wall. Therefore, your shaving, skin preparation, and draping need to be long and wide. Although we repeat the mantra to our clients that "incisions heal side-to-side, and not end-to-end", we too often confine ourselves to overly short incisions. Although you may have a specific focus for your surgery based on pre-operative evaluations that only requires a limited incision to expose, you should treat every abdominal surgery case as an exploratory, and to accomplish a thorough evaluation of the abdomen you must make a sufficiently long incision (at least xiphoid to pubis).

Whenever possible you should try to use a standardized, rote approach to your exploration, and do this first, before you address the primary tumor. Much like reading a radiograph and disciplining yourself to look first at the periphery before focusing in on the most obvious lesion, it is too easy to become excited by the primary lesion during surgery and overlook other significant findings. Sometimes tumors are so large (eg, splenic masses) and/or enveloped in omentum that they must be attended to first. Likewise, bleeding abdominal masses need to be addressed early on to stop the blood loss. But if this is the case, be sure to go back and do the rest of the exploratory at the conclusion. Also, in the instance of hemoabdomen or other effusion, it is essential that you have the ability to aspirate the fluid with suction. Large dogs may have multiple liters of free fluid that needs to be removed before you can visualize much of anything. I like to make an initial small incision in these cases with enough room to insert a Poole suction tip. I then wait patiently to aspirate the effusion before commencing to extend my incision cranally and caudally. This helps you document the actual volume of free fluid, and if nothing else, keeps your socks cleaner and drier. Whether you have assistance or not, it is helpful to have self-retaining retractors such as Balfours available to improve your visualization. I also like to use laparotomy pads as opposed to 4X4 sponges once I've entered the peritoneal cavity since I am far less likely to inadvertently leave a large lap pad behind than a small sponge. Of course the ideal practice is to count sponges of every type (along with hemostats, scissors, etc.) to avoid the embarrassment and potential morbidity of leaving something in the patient.

Whether you see gross pathology beyond the primary tumor or not, you must bear in mind the biologic behavior of the tumor, and consider biopsies of lymph nodes and liver if these are common metastatic sites. If lesions are seen these are specifically excised. If no lesions are seen random samples should be harvested to look for microscopic disease.

A good practice for tumor removal anywhere is to change gloves and instruments after manipulating neoplastic tissues. Some tumors are quite adept at exfoliation and implantation, so using uncontaminated gloves and instruments for closure is very sensible. However, if a tumor is already widely disseminated (eg, ruptured splenic hemagiosarcoma) this precaution is made moot. Another principle, where applicable, is to do any "clean" part of the surgery first, and save handling neoplastic tissues for last. Again, in the case of abdominal or thoracic cavity masses, the exploratory should ideally precede tumor resection. Another example is when performing a combined mastectomy and overiohysterectomy: You should first make your midline celiotomy incision and complete the spay, then close the linea, and only then proceed to remove breast tissues.

In cases where there is doubt about the completeness of tumor resection it is helpful to have the pathologist examine the margins of any excised tissues. If the margins are not obvious or will be obscured by tissue folding, etc., then it is useful to use some method (such as India ink) to stain the marginal sample before placing the specimen into formalin.

For major subcutaneous or body wall tumor excisions there will generally be creation of a good deal of dead space. This may be complicated by incorporation of foreign material such a mesh, or movement of a large flap or graft to close the defect. This dead space cannot and should not be closed with "tack down" suture techniques. The preferred approach is to incorporate one or more drains in any dead space and exit these away from any primary incision lines. The best drains are closed-suction, grenade-type drains that are now widely available and affordable. These drains are usually made of a silastic material that is non-irritating, and come in a variety of sizes. Penrose and other passive, open drains are frowned upon since they are less effective at keeping wounds decompressed, and they permit microorganisms to travel retrograde from the environment.

Practical Approach to Fracture Management and Decision-Making:
When and How Should I Refer?

Fractures will invariably continue to be a common disease entity that veterinary clinicians in general practice will be confronted with. They can be disconcerting for several reasons. First, by their nature they are unexpected, and so trauma patients tend to be disruptive of normal schedules. Second, many fractures are but part of a larger patient trauma situation, such as motor vehicle accidents. This only adds to the anxiety and stress felt by both the veterinarian and client. Finally, the veterinarian has to make judgments about how to manage these patients and their injuries, including decisions about whether to treat the fracture themselves or to refer, and what to do for the patient between admission and definitive treatment. My hope in this presentation is to help put some of these things into perspective, and provide veterinarians with some tools for answering the above, thereby reducing some of the stress and mystery associated with fracture management.

We have to first remember that we are not confronted with "the broken leg in exam room 2". What we have is a patient for whom we must generate a problem list that may or may not include a fracture. In the book, "The House of God" by Shem, the first "law" is: "At the scene of a cardiac arrest the first thing to do is…take your own pulse." This philosophy of taking a deep breath before plunging into a crisis situation is obviously more universally applicable, and is very apropos in the trauma setting. This can be made difficult by clients for whom the broken leg is the most obvious and distressing injury, and they want you to address it immediately. Reassurance with a calm demeanor, or if necessary, separating the patient (and yourself) from the client can be helpful to both the client and to you, so as to allow you to do your patient survey without distraction.

Your initial survey should be to make a judgment as to whether the animal is in shock or not. If so, attention is focused on general patient resuscitation using such things as flow by oxygen, intravenous fluids, and analgesics. A methodical exam should be conducted very much like your standard wellness exam. This ensures that all body systems are evaluated. I usually reserve examination of any obvious injuries - fractures and luxations - for last. It is too easy to become focused on these "big" problems and miss more subtle, but no less significant injuries elsewhere. Have in your mind the concept that if enough energy was delivered to the animal to break bone, there may well be significant soft tissue/internal trauma sustained in addition. Concentrate your initial efforts here before being sidetracked by a fracture. If the patient has an open wound it is certainly reasonable and sensible to cover or otherwise protect it if there needs to be a delay before you can address it more specifically. If it is going to be a long delay (> 1hr - the so-called "golden period") you should do some preliminary clipping of hair and cleansing of the wound before simply encasing it in a bandage. This wound toiletry will help extend the golden period.

Radiographs may play a role early on in the assessment of the patient, but here again the chief concern should be with non-orthopedic injuries. Trauma victims deserve thoracic radiographs and probably abdominal radiographs as well. These films will allow assessment for grotesque spinal column injuries, lung and diaphragm injuries, bladder integrity, and loss of detail that might signal internal hemorrhage or other life-threatening effusion. If you already know there is an appendicular or pelvic fracture there is probably little to be gained by early radiography of such injuries. Chances are you are not anesthetizing the animal or being too attentive to positioning, technique, or the acquisition of orthogonal and oblique views. So the films you have (and must charge the client for) contribute little to better understanding the problems or deciding about management at this stage. In most cases, knowing right off the bat if the fracture looks "nasty" or not will not be a factor in your ultimate decision about its management.

To quote from "The Age of Miracles: Medicine and Surgery in the Nineteenth Century" (Guy Williams, Academy Chicago Publishers, 1987):

"[In 1895] Robert Jones heard of the momentous discovery that William Conrad Roentgen had made…Having a lively and forward-looking mind, Robert Jones set off at once for Prussia to find out a bit more about Roentgen's researches. Having heard Roentgen's story and having seen a demonstration of his extraordinary flesh-penetrating process, the young Welshman ordered and paid for an X-ray tube. Then he returned to England, and there he and another young man called Thurston Holland, later to become one of Britain's most eminent radiologists, took an X-ray picture [that] showed plainly a small bullet that was embedded deeply in a boy's wrist. Looking at it, Robert Jones realized at once how immensely valuable X-ray photography would be to all future orthopedic surgeons…"
"But he did not allow his judgment to be clouded by the excitements roused by Roentgen's new technique. Jones knew only too well how dangerous were all modern short cuts to genuine scientific knowledge. 'While Roentgen's discovery has been to us of immense value, chiefly in the classification of our injuries', he wrote, 'it has done little if anything to perfect or alter our treatment of fractures'. All medical students should acquire proper clinical skills in diagnosis, he went on. They should not rely exclusively on the interpretation of a fallible photograph."


Once the time comes to take radiographs of any fractures or luxations, there are some principles to bear in mind. Ideally animals should be positioned using sandbags, ropes, tape, and the like, and not hand-held. Depending on the animal's mental status and cooperativeness this may require no sedation, but usually demands some form of chemical restraint. The drug(s) used will be determined by the patient's overall health status at that point. Attention should be paid to exposure technique and collimation; try to focus the image on the target area without including extraneous anatomy (but DO include all the soft tissues and skin around the target bone). At the same time, try always to include the entire bone from at least the entire joint above through the entire joint below. Try always to obtain two orthogonal views for appendicular and spinal radiographs. For pelvic/sacral fractures it can be useful to obtain oblique views as well. When luxations are suspected it will be important to take appropriate stress views that demonstrate any loss of soft tissue stability to the joint (and in the case of the carpus or tarsus, which joint level is affected). Although currently rare in general practice and nascent even in specialty hospitals, expect to see increased use of computed tomography (CT) for fracture assessment in the next few years. This modality is particularly helpful for spinal, pelvic, and sacral injuries, allows three-dimensional pre-operative planning, and is seeing increased use for long bone fracture assessment as well.

For your own medical records, but especially if you contemplate referral, you should strive to interpret the radiographs and classify the injury using some standardized system,
Standard terminology includes noting whether the fracture is open or closed, simple or comminuted, orientation of fracture line, and type of displacement.

Open fractures (old synonym = "compound") can be further classified by degree. Grade I open fractures are caused by penetration of the skin by the bone from within; there is typically minimal soft tissue disruption, and most fractures are simple or minimally comminuted. With a modicum of wound care (clipping, cleansing, protecting with a bandage, possibly systemic antibiotics) they can be managed just as a closed fracture would. Grade II open fractures are more often caused by penetration of the skin from without, and also induce more soft tissue damage and may have more comminution. Some femoral and other fractures that have skin perforation from within also have extensive muscle and bone trauma and better fit this grade II scheme. Grade III open fractures have more significant bone and soft tissue trauma, including quite often neurovascular bundle damage, and actual loss of bone and soft tissues. They are almost always caused by external wounding of the skin, and have more wound contamination than grade II injuries, The most common examples of grade III open fractures we see are those caused by gunshot wounds and shearing wounds to the medial aspect of the tarsus, metatarsus, and digits caused by collision with an automobile (so-called "street pizza"). In human medicine there are many sub-classification schemes, often tied to a particular region of anatomy (eg, lower leg injuries), and these are clinically useful because of statistically-derived predictive value in prognostication and treatment. For us, it would be good to at least classify within the three-grade format, understanding that these grades are on a continuum and are not discrete. Grade II and particularly grade III open fractures demand much more intensive wound management. Grade III open fractures represent a true orthopedic emergency and require appropriate care within the first minutes to hours.

Simple fractures are two-piece fractures. The term "greenstick fracture" is reserved for those injuries that only disrupt one cortex, and these are relatively uncommon. Fractures with three or more fragments are termed comminuted. These can be further described by degree, such as "mildly comminuted with a single large 'butterfly' fragment", or "highly comminuted with multiple fragments of varying size." Note that the terms "comminuted" and "compound" are NOT synonymous; You can have one, both, or neither in a fracture. Sometimes we may have a "relatively" simple fracture where there are a few, very tiny "crumb-like" pieces. It is a judgment call whether to term these simple or comminuted, but most of the time these can be considered as "simple." A special case of the comminuted fracture is the segmental fracture in which an entire cylinder of bone is broken free in the diaphysis.

Simple fractures can also be described by the orientation of the fracture line. The common terms used are transverse (perpendicular to the long axis of the bone), oblique (some angle between perpendicular and parallel), compression (mostly applicable to physeal fractures [Salter-Harris type V], and vertebral fractures), and spiral fractures (those that wind around the bone with a 450 angle. For oblique fractures it is also helpful to describe if "short" oblique (< 3-4 X the bone diameter in length) or "long" oblique. The orientation can reveal a great deal about the biomechanical forces that caused the fracture, and in turn help inform decisions about appropriate fixation methods.

With simple or mildly comminuted fractures you can also describe displacement. By standard phraseology displacement direction is determined by the distal fragment. In addition to the standard anatomical nomenclature (cranial, caudal, medial and lateral), displacement may also be overriding.

Disciplining yourself to use of these standardized terms will greatly facilitate discussion with a surgeon over the telephone, and in turn improve the information you can provide your client with respect to timing of referral, potential fixation methods, and costs (by allowing the surgeon on the receiving end to have a more precise mental image of the fracture beyond, "Gee have I got a nasty one for you!").


With rare exceptions, most fractures in veterinary medicine, even grade III open fractures, do not go to the operating room on an emergency basis. This is largely because of shock and other serious injuries that need to be stabilized (and the absence of teams of anesthesiologists and criticalists to keep the patient alive while the orthopedist works away), requirements for clients to absorb what is happening and make financial decisions/arrangements, and not always having a surgeon immediately available. Our focus then should be on preventing worsening of the orthopedic injury, wound management as needed, and pain control.

Fractures below the knee or elbow are usually amenable to coaptation. Fractures of the distal humerus and distal femur may also be coapted if the device can get high enough into the axilla or groin, respectively, to effectively immobilize the fracture. The optimum device in this acute setting is the Robert Jones bandage. These are very effective at controlling soft tissue swelling without compromising vascular function. Modifications of the Robert Jones bandage include a less bulky bandage into which a rigid splint is incorporated, and use of rolls of newspaper, towels, and duct tape as advocated by Tim Crowe.

If there is no open wound there is no orthopedic indication for starting systemic antibiotics. An open fracture however probably warrants initiation of broad-spectrum antimicrobials. Limbs with suspected open wounds should be thoroughly clipped (ranging from a local hairclip around a small, grade I puncture wound, to a complete shaving of the limb for more extensive injuries) and cleaned prior to any coaptation being applied. A water-based sterile lubricant can be placed into the open wound during shaving to prevent migration of hair deeper into the wound. It is less critical to use antiseptics in this cleansing solution than to use large enough volumes to mechanically remove and dilute any contaminants. IV crystalloid fluids or sterile physiologic saline for irrigation make fine, readily accessible lavage solutions. Open wounds can be covered by either a non-adherent dressing (puncture wound) or more elaborate dressings (eg, wet-to-dry, sugar, etc.). In addition to lavage, gross contamination should be removed with sterile surgical instruments using aseptic technique. The goal should be to convert the wound, as much as possible, from one that is contaminated (and therefore more likely to become infected) to one that is clean, However, any debridement should spare removal of large bone fragments, tendons, ligaments, nerves and vessels so these can be better evaluated during definitive treatment later on.

Analgesia can be provided by several routes. Coaptation alone and prevention of motion at the fracture site will do much to alleviate pain. Beyond standard parenteral techniques for analgesia one can also consider epidural or regional routes, and for distal extremity fractures even consider injecting local anesthetic directly into the fracture hematoma.


The technical term for fracture repair using devices is "osteosynthesis". It is a cooperative effort between the bone and any hardware applied by the veterinarian. It is important to recognize that we have many systems for fracture management at our disposal. These include casts/splints, intramedullary pins and cerclage/hemicerclage, plates and screws, interlocking nails, and rigid external skeletal fixators (linear, ring, and hybrids of the two). Systems of osteosynthesis should be separated from the specific hardware or devices, since the latter may be applicable in more than one system (eg, pins and wires may play an adjunctive role in an external skeletal fixator repair), or there may be alternative devices within a system (eg, Kirschner-Ehmer, Securos, and Imex all have devices for external skeletal fixation). It is equally important to recognize that no one system is applicable to every fracture. This is why those who do orthopedic surgery for a living maintain a large inventory of systems and devices that permit customization for every case. This can be one source of problems for those who do orthopedic surgery less often. They may have access to fewer osteosynthesis systems, and may try to apply the devices they have accessible to fractures where they are really contraindicated, or at least where other systems or devices might be far superior.

One of the real challenges of deciding how to manage and repair a fracture is in understanding all the factors that impinge on this decision, as well as a consideration of the technical skill and experience of the veterinarian. It should also be remembered that none of us really ever "repairs" a fracture. Bones heal because bodies are self-reparative. Bones have been healing for eons before there were humans to finagle them. If anything, our interventions and manipulations may serve in many cases to delay or impede the healing process. So our goal and raison d'etre as orthopedists is to assist Mother Nature, and try to provide the optimum conditions for functional bone healing. For us, it is not sufficient just for the bone to heal, but we want it to heal with restoration of normal anatomic alignment and relationships, and with full and hopefully early return to normal function.

Less experienced veterinarians are often puzzled by the ability of a more experienced orthopedic surgeon to silently contemplate a set of radiographs, examine a patient, and then declare how they propose to repair a given fracture or set of fractures, or sometimes elect not to treat certain injuries. The novice may be even further perplexed by the presentation of what seems like a very similar fracture case, and observing the same experienced orthopedist choose an entirely different pathway for management. The mind of the experienced doctor seems to be a "black box". Also consternating to less experienced surgeons is the failure for "textbook" solutions to have successful results when applied to their real patients. The problem with textbooks - even the very best of them - is that they usually only describe the fractured bone, and say nothing about other patient factors that might influence decision-making and outcome. It is risky to thumb through a textbook looking for a fracture that is "similar" to the one you are presented with, and then assume the textbook now becomes a cookbook, allowing you to simply transcribe the published osteosynthesis to your patient.

In an effort to unlock the mysteries of the "black box" that is the mind of the more experienced orthopedist, Ross Palmer and Don Hulse several years ago developed and published the Fracture-Patient Assessment Score (FPAS), a methodical system for helping veterinarians account for many of the factors that the experienced orthopedist uses in his or her assessment, almost unconsciously, in designing a treatment plan. I will review the basics of this strategy shortly and encourage each of you to employ it with your fracture cases.

The development of newer and better systems for fracture management in the twentieth century led increasingly to a concept of "rigid anatomic reduction" as the sine qua non of being a successful orthopedist. The top surgeons were revered for their ability to take a jigsaw puzzle of a fracture and put it all back together with a combination of pins, wires, screws, plates, and the like. My own final exam in orthopedic surgery as a 3rd year veterinary student was to illustrate, on a diagram of a comminuted fracture, how I would reconstruct it. It was all about the bone and the hardware. This devotion to recreating the intact column of bone has been likened to being a carpenter. Lip service was always given to "gentle handling of soft tissues", but in actual practice, the demands of approaching a fracture from the carpenter's perspective required rather brutal treatment of the soft tissues in order to achieve the primary goal of anatomic reconstruction.

In recent years the new buzzword has been "biological" osteosynthesis. The concept is based on an increased appreciation of how bone cannot be isolated from its soft tissue milieu, especially since its blood supply and therefore ability to heal is dependent on the integrity of those soft tissues. At its extreme it has led to increased use of either closed reduction methods or so-called "open (or "look") but don't touch" methods of fracture manipulation. The advent of these less aggressive attempts to reconstruct fractures anatomically has been made feasible by the concomitant development or refinement of systems for osteosynthesis that permit rigid fixation without anatomic reduction. This attention to the soft tissues, even sometimes at the expense of actually repairing the broken bone, has been likened to that of a gardener.

The truly expert orthopedic surgeon of course wears both hats; the carpenter's and the gardener's. It is in the understanding of how to exploit the best of both, and knowing when to tip the balance one way or the other, that makes the difference between successful osteosynthesis and failure. Even the best "carpenter" must acknowledge that no implants or fixation system will hold up forever, and that ultimately the bone must be able to take over weight bearing loads. The difference between bone and metal or plaster is that the former has some inherent plasticity, and also can remodel/repair when subjected to repetitive stress, strain and mechanical loading. Metal and plaster, no matter how stiff and strong, undergo fatigue from cyclical loading. So every fracture repair is a race between the time the bone heals (making the hardware superfluous) and the time the metal or plaster undergoes structural failure because the bone has not healed sufficiently. Our job of course is to do all we can in surgery, and after surgery, to promote the bone winning the race. Conversely, bone requires both blood supply and stability to heal. Over-emphasis on nurturing the soft tissues as a "gardener" may lead to delayed unions or non-unions as a consequence of inadequate control of motion at the fracture site. The ability to wear both hats to the right degree has been termed "balanced fracture treatment"

Although much of this balanced approach is seen as a "modern" development, it really represents the recurring swing of a pendulum over time. For example, the derivation of external skeletal fixation systems dates back to the 1940's with the initial introduction of the Kirshner-Ehmer apparatus. "KE's" fell out of favor in the 1960s-1980's because their failure and complication rates were too high, and because surgeons had become infatuated with modern systems of bone plates and screws. Over time this became a general discrediting of external skeletal fixation. However, a resurgence of interest since the 1980's, with the application of more basic scientific standards, established that the principles of ESF were in fact sound, but that the devices available at the time were inadequate. Since then we have seen ESF become rightfully reincorporated into the surgeon's armamentarium with the advent of devices that incorporate the requirements for successful application. Even more basically, the concept of "gardening" fractures is hardly new. Quoting from the same book, The Age of Miracles:

"Hugh Owen Thomas, in 1859, set up practice on his own. [He] managed to develop, on his own initiative, the modern treatment of fractures. He was completely opposed to the growing practice of immobilizing limbs with plaster since this, in his view, 'did not allow frequent inspection, was much labor at first, and little afterwards, and provided no opportunity for the display of skill'. He despised the rudimentary wooden splints that were used by most practitioners of the time for supporting limbs with broken bones - if, indeed, those limbs were supported at all."
"Needing relatively sophisticated splints of every size, that would suit any type of fracture…he decided to use, principally, iron for strength and rigidity and leather for comfortable padding. On Hugh Owen Thomas's home-made splints, many of our most necessary modern orthopedic appliances are based. There is no doubt that Hugh Owen Thomas was years ahead of his time…"
"It is ironic that on one of the few occasions Hugh Owen Thomas left his practice, in the 1870's, he went to offer the use of the Thomas Splint to the French Army. His offer was refused. Nearly half a century later, Thomas's nephew Robert Jones was to record:
'The Great War afforded the most convincing proof of the mishandling of complicated, and even simple fractures. Fractures of the femur were a notable example. The splint with which we are all so familiar, invented by Thomas, was barely known, and yet it was the type of splint which ultimately saved the situation. In 1916 the mortality from these fractures amounted to 80 per cent, a large proportion of the deaths occurring on their way to or at the Casualty Clearing Stations. Later, when the Thomas Splint was applied almost exclusively, and as near to the firing line as possible, the mortality in 1918 was reduced to 20 per cent.'

The Palmer-Hulse Fracture-Patient Assessment Score

Ross Palmer and Don Hulse identified three main areas that the experienced clinician considered when making a fracture treatment plan: Mechanical factors; biological factors; and clinical factors. In their FPAS they suggested use of a subjective scoring system from 1-10 to assign numbers to each of these factors and their subfactors, ultimately to derive an average score for the particular patient and particular fracture. Lower scores indicate the worst case scenarios and the most difficult to manage successfully while higher scores are associated with more optimum outcomes and lower complexity. The score could then be used to suggest the osteosynthesis requirements (More carpenter? More gardener? Which system?), as well as the demands on the surgeon (Reasonable for a generalist to tackle? Better to be referred to someone more experienced?).

Mechanical Factors

How stable a fixation device needs to be is determined by several subfactors, including the patient's size and weight, the presence or absence of other limb injuries (ie, whether the animal will be forced to bear more weight earlier on this limb), and the ability of the fractured bone itself to share weight bearing load.

Load sharing between the bone and the fixation device is determined in large measure by the configuration of the fracture as we described in the radiographic assessment. Transverse fractures are capable of immediate and substantial load sharing whereas comminuted fractures with more than just a single, large, reconstructable butterfly fragment generally cannot share any load at the outset. In the latter situation the fixation system chosen and devices used within that system must be constructed to retain their strength for a protracted period of time without succumbing to structural failure from cyclical loading. Oblique fractures represent a middle ground between ideal load sharing (transverse fracture) and non-load sharing (comminuted fracture). They may be partial load sharing by virtue of anatomic reconstruction with cerclage wires, olive wires or interfragmentary lag screws to achieve interfragmentary compression, but the shear forces acting on these interfragmentary devices must be neutralized to prevent their failure. This can be accomplished by such systems as interlocking nails, external skeletal fixators, or application of bone plates in neutralization mode.

The mechanical FPAS approaches the high end of the scale when the patient is small, lean, has only one injured limb, and the fracture is capable of immediate load sharing. Deviations from these ideal circumstances shift the score lower.

Biological Factors

Some of the issues to consider here include whether the fracture is open or closed, the degree of soft tissue injury (which can yet be substantial even if the fracture is closed, as is typical when the injury involves a large shock wave of contusive energy, such as with automobile trauma), and the overall health status of the patient (Young? Old? Well? Sick?). Obviously, young, healthy patients with low energy closed injuries (eg, jumped from a modest height) will have a higher FPAS than older, cachectic, systemically ill animals or those with more trauma to the bone and soft tissues.

Clinical Factors

The major concerns here are the ability of the client to appropriately care for any fixation device (especially applicable to the intended use of either coaptation - splints and casts, or external skeletal fixation), ability of the client to appropriately restrict the dog from over-activity, the cooperativeness of the patient in respecting the device (ie, will the dog leave a cast alone or will he chew at it?), the consequences of having to immobilize any joints (as with the use of casts and splints), the availability of physical therapy to promote return to function, and putting all of this into the context of the anticipated time needed for good client and patient compliance with restrictions (ie, is this a fracture we anticipate to heal within 3 weeks or is it one that will likely take 3 months?).

Note that it is pro forma for us to instruct our clients about how to properly restrict the patient post-operatively, and it is just as certain that even for the best intentioned client or most cooperative dog there will likely be some departures from the ideal. I have yet to have a dog use its crutches as I have told them to do, and as someone who has also been on the client side of the equation, I can attest to how difficult it can be to police a dog 100% of the time. So even as we make judgments about our clients and patients, we should bear in mind that we should design our osteosynthesis to withstand these unintentional overstresses (or in the parlance of the engineers, to "overbuild" a bit) rather than blaming failures on the dog or the client. Denis Aron at the University of Georgia is explicit in saying that any osteosynthesis failure is the fault of the surgeon, either in the original construction, or in the surgeon's failure to adequately educate the client in how to care for the patient.

Nonetheless, we can certainly anticipate some potential problems with compliance with postoperative care instructions, and the degree to which this is expected will determine the score assigned: Closer to 10 for the ideal patient and client, and closer to 1 for those patients and/or clients where we expect poor compliance. This realm of "clinical factors" is one of the major determinants for choosing between internal fixation vs. external skeletal fixation where the two systems might otherwise be equally effective. It is sometimes preferable to open a fracture and apply, say, a bone plate, even though the fracture may be quite amenable to a closed or limited open approach and application of ESF, if in our judgment (or that of the client's) care for the ESF may be problematic.

The overall FPAS can then be derived as the arithmetical average of all these scores (Palmer and Hulse do not describe any weighting scheme by which some factors are held to be more consequential than others), with the following generalizations:

FPAS 1-3

The lowest scores are more complex and problematic cases that require greater expertise, stronger and more durable osteosynthesis, minimal further disruption of soft tissues, and are intolerant of judgment or technical errors. For fractures that are deemed non-reconstructable or where reconstruction would be time consuming and require extensive soft tissue dissection in the fracture zone, the gardener's hat is worn and the focus is on maintaining spatial orientation of the bone fragments (bringing the fracture out to length and keeping joints properly aligned) while using devices that permit rigid fixation of the bone on either side of the fracture zone. Such cases should not be undertaken by novices, or those who are not in possession of the fixation system required to achieve the stated goals. Even if a client insists that referral is not an option and states "just do your best, doc", I would recommend against feeling backed into a corner and doing what the client requests. Almost invariably there will be complications (often outright failure), you will feel terrible, you will lose money in your effort to help, the client and patient will suffer, and most depressing of all, at the end of all your efforts to be "the good guy", YOU will be blamed by the same client that asked you to "just do your best" for all their subsequent unhappiness. And in this day and age of increased litigation for veterinary malpractice, you can bet that the client and their lawyer will not hold you to some lower standard even though it was at the client's behest that you tackled something YOU knew you were ill-prepared for. A client's refusal for referral is not, ipso facto, a requirement for you to assume a burden you already judged was not appropriate for you to take on.

FPAS 4-6

These cases are only slightly less tolerant of technical error or lapses in judgment, and still require a surgeon with considerable experience. However, a less experienced surgeon or general practitioner who does a fair bit of orthopedic trauma surgery, who understands his or her limits (both in expertise and equipment), and who feels he or she can adequately manage the case is in a position to proceed if referral to a specialist is not possible or is declined by the client.

FPAS 7-9

Somewhat more leeway is available here for minor technical errors or mistakes in judgment; these cases are more "forgiving". Nonetheless, some firm grounding (such as attendance at a post-graduate continuing education course and laboratory that teaches orthopedic principles and basic techniques), and some clinical experience under the tutelage of a more experienced orthopedic surgeon permits the general practitioner to handle many of these cases with fewer concerns for severe complications.


For the novice who has an interest in orthopedics and desires some clinical case examples to "test the waters", these cases are good for building experience. There is less critical need to properly balance the carpenter and gardener hats, these cases are tolerant of miscues in judgment and technique, and one generally has to work hard to yield a failure. Still, some background (academic courses and laboratories, prior work with a mentor, having someone with more experience scrub in to provide some advice) is beneficial. These cases are also ideal for the more experienced surgeon to begin using a new system or device to gain familiarity and experience after an initial exposure at a CE course and lab.


The increasing variety and sophistication of osteosynthesis systems has made it possible to effectively treat almost every type or combination of fractures. The corollary is that it is no longer feasible for every general practice to maintain the necessary inventory, not to mention having the training and experience to exploit all these systems. Use of the FPAS will help the generalist decide whether he or she has the proper personal background, and is properly equipped to handle a given case, or whether it would be better to refer. It is also entirely appropriate, if the general practitioner is not sure whether to refer or not, to call the more experienced person to whom he or she intends to refer and discuss the case over the phone. Using the systematic radiographic description system above allows clarity in communication, and permits the more experienced/better equipped surgeon to advise the potential referring doctor more precisely.

One trap to avoid is in being too specific with the client with respect to how the case will be definitively handled by the orthopedic surgeon to whom you are referring the patient. It is disconcerting for a client to come into the referral hospital expecting their dog's fracture to be "plated", or have some other specific management, only to have the surgeon tell them that he or she intends to pursue some other method. Moreover, most orthopedic surgeons tend to have contingency plans in mind if further radiographs or the situation encountered in the operating room suggests that their "plan A" will not be tenable.

Another trap is to suggest a financial estimate to the client without having first discussed this with the surgeon to whom you are referring. An over-estimate might discourage the client from pursuing the referral in the first place, where estimating too low will frustrate the client, the surgeon, and reflect poorly on you. If you cannot for some reason discuss this with the surgeon prior to making the referral, explain to the client that their only commitment if they go for the referral is for the initial examination/consultation fee. The surgeon will prepare an assessment of the case, make a recommendation for optimum treatment, possible alternatives, and provide an accurate cost estimate. At the worst they will be better educated about the nature of the problem and the prognosis associated with each of the management options, even if they ultimately decide not to pursue treatment by the referral surgeon.

Another thing to bear in mind is that the receptionist who takes the call from you or your client at the referral surgeon's hospital may or may not be well-trained to triage cases. That person may not be able to distinguish an urgent or emergency case from something more "routine", and consequently may schedule a fracture case days or longer down the road. Your client may similarly not fully understand the urgency of the situation, and assume that the appointment wouldn't be delayed if it wasn't OK to do so, compounding the error. My suggestion, for this or any case that you perceive to be urgent or emergent, is for you to call directly (and if need be, insist on speaking with the surgeon), or to tell your client to advise the receptionist that the patient needs to be seen right away. No matter how busy a referral surgeon may be, he or she will be much happier squeezing in your case, or otherwise being sure that the case is handled appropriately, than to see a patient with a fracture a week or more after the injury simply because no one insisted on an earlier appointment time.


Since the days of the psalmist, who wrote, "…Thou shalt make me hear of joy and gladness: that the bones which thou hast broken may rejoice…" (Psalm 49), there has been a progressive improvement in our understanding of fracture biology and in our ability to work better with Mother Nature in restoring function after bone injury. I hope this presentation helps you better cope with fracture cases yourself in the future.

Decision-Making, Management, and Referring the "Disc" Case

Primary care veterinarians and emergency clinicians are frequently presented with dogs experiencing varying degrees of weakness, paralysis, and pain. They are challenged in making such decisions as: Should the patient be referred for surgery? If so, at what point should such referral be made? Should I treat the dog medically? If so, what drugs should I use? Not use? Dose? Should I take radiographs? What kind of prognosis should I offer the client?

My goal is to try and provide some guidelines to help answer the above questions, and to identify areas of controversy as well as areas where we lack sufficient data to have clear resolution to these and other dilemmas.


It is obvious to each of us, but often forgotten in the heat of an emergency, that not every dog that presents with spinal pain and/or paralysis has a disc-induced injury as the cause. It may be quick shorthand to list the problem as "IVDD", but this is a trap that could potentially blind you to other differential possibilities, and have you insufficiently prepare the client for possible options and outcomes.

The first course of action is to do an initial, basic neurologic exam. A more detailed and comprehensive exam can follow, but within a few minutes you can do much to establish a more accurate rendering of the problem. One of the most important aspects of the "exam" often overlooked or given short shrift, is the anamnesis. Much can sometimes be gleaned from the history provided by the client. Of particular interest are the following:
  • Onset
  • Progression
  • Lateralization
  • Prior history
Onset refers to how quickly the signs appeared. Has the dog been acting uncomfortably or limping for a time before gross weakness or paralysis appeared? Did the signs come on at all gradually or did the dog go from normal to the current state seemingly all at once? The clinician should try to establish from this information whether the onset was chronic, subacute, acute, or peracute.

Progression refers to what has transpired since the client became aware of signs. Are the signs waxing and waning? Has the dog improved? Worsened? How rapidly? If the dog is painful as well as paralyzed, did the two come on and progress at the same rate, or did one precede the other?

A very important part of the history, when available, is any notation by the client of lateralization. By the time you see the dog it may be symmetrically paralyzed, but the client may be able to report that signs began or seemed worse initially on one side or the other. This information can be extremely helpful to a surgeon since imaging studies (particularly myelography) may fail to disclose which side the lesion is on. Even with syndromes where pain is the only sign the client may be able to relate that the dog seemed to limp on or elevate a particular limb consistently.

How you manage the patient will also be influenced to some degree by any prior history of similar signs. Much like a cat with urethral obstruction, dogs with recurrent signs may be referred for surgery as much on the basis of recurrent history as on the issue of severity of current signs. For example, dogs that only are painful (no deficits), or those who have mild deficits, but are repeatedly or relentlessly affected, might be in as much need for referral to a surgeon as the dog who is having a first episode of more profound paraparesis that is not improving with medical treatment. On this same track, dogs that have recurring signs of back pain or weakness occasionally have multiple disc displacements responsible, but the vast majority in fact are having recurrence of signs attributable to the same, single disc. This "ticking time bomb" underscores the need to sometimes consider surgery even when the signs are not dramatic.

While you are taking the history it can be helpful to have the patient on the floor where you can observe it. This may give you an insight into whether the dog still has voluntary motor function or seems uncomfortable in a way that you may not appreciate as well once you start handling the dog.

Your objectives in doing the neurologic exam proper are to start to get an idea about localization and severity of signs.

Localization can be divided into zones based on spinal cord segments:
  • Cervical: C1-C6
  • Cervical: C7-T2 (intumescence)
  • Thoracolumbar: T3-L3
  • Lumbar: L4-L6
  • Lumbosacral: L7-S1
  • Lumbosacral: S2-S3-Coccygeal (caudal)
Remember that spinal cord segments do not usually lie within the vertebral canal at the correspondingly numbered vertebrae. Our neurologic exam establishes segmental localization, and we must then refer to anatomy books to see where those segments lie within the vertebral canal.

Cervical lesions usually induce signs in both the thoracic and pelvic limbs, although the latter may be more dramatically affected. Lesions below T2-T3 should leave thoracic limb function normal. The keys to determining the localization include observation of gait, proprioception, postural reactions and segmental reflexes.

A dog that is dragging its pelvic limbs but seeming to walk normally with the thoracic limbs is very unlikely to have a lesion cranial to T3. While pelvic limb signs can be worse than thoracic limb signs with cervical disease, they will not have this degree of disparity. Dogs with less dramatic signs may exhibit deficiency in proprioception or postural reactions (hopping, extensor postural thrust, hemistanding, etc.) that can alert the clinician to involvement of the thoracic limbs. Finally, testing of segmental reflexes is critical. Unfortunately, the ability to reliably test thoracic limb reflexes (biceps/triceps) is problematic. They are much more difficult to test than corresponding pelvic limb reflexes. The easiest forelimb reflexes to test are the flexor (withdrawal) reflexes (always check with separate stimulation to the medial and lateral digits) and general tone in the limb. Subtle muscle atrophy of the scapular muscles might be an indicator of lower motor neuron deficits, but neurogenic atrophy of this type takes about a week to become evident.

Caudally, the reflexes that need to be tested include the patellar tendon reflexes, cranial tibial reflexes, withdrawal reflexes (again, separately check medial and lateral digit stimulation), and perineal reflexes. A digital rectal examination should be performed to check anal tone.

Crossed extensor reflexes can be checked in addition to segmental reflexes The finding of a positive crossed extensor reflex (which is a pathological reflex) indicates that there is an upper motor neuron deficit to the limb that extends in response to forced flexion of the contralateral limb while the dog is in lateral recumbency. It is of no prognostic value however. It tells us nothing about the severity of the lesion or the likelihood for recovery.

The goal of these examinations is to define whether weakness (paresis) or paralysis (plegia) can be characterized as upper motor neuron (UMN) or lower motor neuron (LMN) with reference to particular nerve distributions. Knowledge of which peripheral nerves subserve particular reflexes, and in turn, which nerve roots and spinal cord segments contribute to those peripheral nerves, allows localization to particular zones of the spinal cord as listed above. As one is trying to deduce this localization based on one's exam findings, it can be helpful sometimes to turn the process around into an inductive problem. That is, ask yourself, "If I put a lesion in a particular spinal cord segment, what deficits and reflex changes should I see?" Does this match your actual findings?

Another reflex that can be helpful with localization is the cutaneous trunci (nee: panniculus) reflex. Stimulation of the skin along the truck induces a signal that enters the spinal cord, travels cranially to the C7-T2 segments, and then synapses on the origin of the lateral thoracic nerve which emerges in the brachial plexus and provides motor innervation to the cutaneous trunci muscle, producing a twitch in response. A lesion in the cervical intumescence would likely cause a complete areflexia of the cutaneous trunci reflex, while lesions more caudal might produce what is termed a cutaneous trunci "cut-off". In that instance, stimulation of the skin caudal to the level of the spinal cord lesion (taking into account the tendency for nerves to branch off the spinal cord at a caudal angle, rather than perpendicularly), will induce no twitch, whereas stimulation cranial to the lesion will elicit a normal reflex.

The final part of the examination that must be performed is evaluation of nociception - pain perception. Noxious stimulation to the digits and tail is applied to observe if the animal has a conscious reaction. Such a reaction is inferred to indicate an intact neural pathway between the sensory receptor and the brain.

In dogs with tetraparesis the status of thoracic limb tone and reflexes (and possibly the loss of cutaneous trunci reflex) will help determine if the thoracic limb signs are UMN (ie, C1-C6) or LMN (C7-T2).

Putting It All Together...And Some Key Caveats/Reminders

At the end of the above, you want to characterize the dog's neurologic deficits in a descriptive way. First off, is the dog ambulatory or not? Does it have proprioceptive deficits? Ambulatory with support or without? If not ambulatory, the question to ask is "Does the dog still have voluntary motor or not?" Finally, if no voluntary motor the key question becomes "Does the dog still have intact nociception (nee "deep pain")?" Note that there is a sequential nature to this. For example, if a dog is still walking to some degree there is no reason to ask the subsequent questions about voluntary motor function or about sensation. So, when you are doing your neurologic exam, you need not worry about verifying the presence of nociception unless the dog has lost all voluntary motor function, since this can be presumed to be intact if the dog is still moving the limb. This information is combined with your other findings relative to localization. So, for example, you might describe a dog as "non-ambulatory, left lateralizing, UMN paraparesis with preserved voluntary motor" (note no need to comment on sensory status). Or you might substitute "Left lateralizing T3-L3 myelopathy" for "UMN paraparesis". The description can be refined further with application of adjectives pertaining to onset and progression (eg, "acute, non-progressive, non-lateralizing, non-ambulatory paraplegia with loss of voluntary motor but intact nociception").

Many like to use a grading system for paresis scaled from 0-5. The problem with this is that there are two such systems, and they are diametrically opposed. In one system, grade 5 is considered normal and grade 0 implies complete paraplegia with loss of nociception. In the other system the numbers are reversed. So without knowing which system someone is using, being told that a dog has "grade 4 paraparesis" can be very confusing, and ultimately conveys no useful information. The descriptive method is much more helpful.

Once the problem has been accurately described the veterinarian can proceed to differential diagnosis and begin considering diagnostic options, therapeutic options, prognosis, and need for and urgency of referral. Signalment, along with onset, progression and localization may suggest etiology. Certainly a dachshund with signs of back pain or weakness is likely to have intervertebral disc disease, but sometimes we do see other lesions in this breed. A 14-year-old German shepherd that presents with acute back pain and paralysis, on the other hand, could have a disc lesion, but one would also be worried about such entities as vertebral tumor with pathologic fracture. Remember that the process of differential diagnosis is a process of formulating, and ranking, educated guesses as to cause. This in turn guides diagnostic testing to confirm or rule out specific diagnoses.

A few things to remember…

If a dog has LMN deficits, then one can be even more localizing/descriptive. For example, if a dog has normal thoracic limbs, is non-ambulatory in the pelvic limbs, has no patellar reflexes, but has exaggerated cranial tibial reflexes, anal tone, and perineal reflexes, the lesion must be in the segments that subserve the femoral nerve, ie. L4-L5. An important thing to keep in mind with LMN deficits is that the lower motor neuron, by definition, incorporates the spinal cord segments (grey matter/cell bodies for the motor nerve), the nerve roots, peripheral nerve, neuromuscular junction, and muscle itself. So a lesion anywhere in this motor unit can produce identical signs. One must be careful to distinguish spinal cord disease from lesions peripheral to the spinal cord proper in such patients.

When looking at segmental reflexes some veterinarians like to do a "sciatic" test where they use their plexor hammer to tap the area caudal to the greater trochanter and observe for a leg jerk. Remember that this is not a true myotatic reflex; it does not involve the spinal cord. All it verifies is that the sciatic nerve distal to the hip is intact. The true tests for sciatic nerve and segments L6-S1 integrity are the cranial tibial reflex and flexor (withdrawal) reflex with stimulation of the lateral digits.

The most critical error that can be made is to confuse an intact withdrawal reflex with intact nociception. When one stimulates the digit the impulses travel to two key places: They synapse within the segment directly on the alpha-motor neuron resulting in a reflex withdrawal or flexion of the limb, and they ascend the cord to the somatosensory cortex of the cerebral hemispheres, producing a perception of discomfort (pain). This should provoke a conscious reaction by the dog beyond the withdrawal of the limb. It is possible to have one without the other. For example, the dog could have a lower motor neuron paralysis due to a peripheral lesion yet have an intact sensory pathway and spinal cord. In this case the dog would react consciously to having its toe pinched with a hemostat and yet might have a diminished or absent withdrawal response. Conversely, and most critical for our discussion here of spinal cord disease, a dog could have preservation of its withdrawal reflex despite complete transaction of the spinal cord cranial to the segment. It is not uncommon, unfortunately, for some dogs to be judged as having intact nociception based on observation of a withdrawal reflex only, even though in reality they have lost "deep pain". Don't let yourself fall into this trap.

The corollary to this is that it can sometimes be difficult to ascertain whether the dog has conscious perception of noxious stimulation. This is particularly true when the dog is first presented to you. Dogs that are brought in for acute paraplegia (no voluntary motor), the very situation where the issue of preservation or loss of deep pain is most critical to determine, are often painful, fearful, and distracted by suddenly being in a strange place and subjected to scary manipulations. If the dog reacts consciously to noxious stimulation you can conclude that it still has "deep pain". But what can you conclude if the dog doesn't turn around and try to bite your hemostat, or otherwise show unequivocal conscious recognition of the stimulus? It can mean one of two things: Either the dog has truly lost sensation (with all the implications thereof) or it has deep pain but is not choosing to communicate this to us (would that we could just ask the question and get a verbal reply!). In this situation it may be useful to hospitalize the dog and check it later on after it has calmed down and become a bit more acclimated. When doing these subsequent checks I recommend trying to "sneak up" on the dog and apply the noxious stimulation before otherwise handling the dog (or even opening the cage door if possible), since this will be the best chance you have to get an accurate response. A consistent absence of nociception on several exams is more suggestive of the dog truly being "deep pain negative".

Many veterinarians get hung up on trying to determine if a reflex is exaggerated (hyper) or not. Certainly our expectation with UMN disease is that reflexes distal to the spinal cord lesion should be increased. It is certainly the ideal to characterize reflexes on a scale from 0-5, where 0 implies complete areflexia , 3 is normal, and 5 indicates clonus (severe hyperreflexia with post-stimulus discharge). By definition grades 0-2 indicate LMN disease while grades 4-5 indicate UMN disease. However, in the heat of battle it can be sufficient to note simply the presence or absence of particular reflexes. If a reflex is present, unless grossly diminished, one can generally assume the lesion is cranial to the associated spinal cord segment (ie, UMN).

One place where reflexes may be misleading is the situation that has been termed "pseudohyperreflexia". This applies to lesions in the L6-S1 segments. The result of a lesion in these segments is depression or loss of sciatic reflexes (cranial tibial and withdrawal), with associated hypotonia in the muscles subserved by the sciatic nerve (chiefly the hamstrings). As a consequence of hamstring muscle hypotonia there is loss of normal balancing antagonism to excursions of the knee in extension when the patellar reflex is stimulated. Hence, tapping the patellar tendon will produce an exaggerated response, and this may be interpreted as patellar hyperreflexia. Without incorporation of the knowledge about the sciatic reflexes, the examiner might conclude that the lesion is cranial to L3, rather than caudal to L5.

Control of micturition is a complex interaction of multiple neuronal pathways. On the "front lines" however, we can reduce it to some generally true and simple concepts. The bladder and sphincter are innervated by the pudendal nerves arising from spinal cord segments S2 and S3. A LMN lesion in these segments will result in so-called LMN bladder/incontinence; the bladder will be hypotonic/flaccid, the sphincter will be incompetent, and the dog will continuously leak urine or whenever its position changes. Palpation of a LMN bladder reveals a hard to distinguish organ that is difficult to get a good grip on. Lesions above S2 may produce so-called UMN bladder/incontinence. In this situation the sphincter is overly tight and makes emptying difficult. The bladder retains tone and is easily palpated. It is difficult to express even though one has a good grip on it. As the bladder distends with urine it will eventually stimulate stretch receptors in the detrussor muscle that will induce a reflexive emptying, resulting in sporadic voidance of urine. However, unless the bladder is grotesquely over distended it will not leak continuously or based on gravity (like a LMN bladder), but it also will not empty on its own since the sphincter will close down again as soon as the detrussor stretch is relieved and the reflex relaxation of the sphincter via the pudendal nerves is shut off.

As a general rule, bladder function and bowel (anal) function are linked. You will not usually see a dog with preserved bladder function but loss of anal tone. So, anal tone can be used as a determinant of bladder function. If an animal has diminished or absent anal tone you can assume that it also has a LMN bladder, and you can localize the lesion to the S2-S3 segments.

In most dogs the onset of UMN incontinence occurs in conjunction with loss of voluntary motor function in the pelvic limbs. Thus, if a dog still can move its legs (even if non-ambulatory) it will most likely still be able to consciously control micturition. This also tends to work in reverse as dogs recover from paralysis: they will regain control over urination and no longer require manual expression or catheterization about the time they recover voluntary motor function in their legs.


The first consideration that applies is making a decision about prognosis and whether it is more appropriate to pursue strictly medical management, or whether surgery is required.

Prognosis depends on the severity, progression, and etiology of the spinal cord lesion. These same factors govern the decisions about how to intervene. Severity is determined by defining the neurologic deficits as described above. Clearly a non-ambulatory dog is worse off than one that is still ambulatory with some support, who in turn is worse off than a dog that only has signs of pain or pain with mild conscious propriocptive (CP) deficits. However, a dog with yet modest deficits but who is worsening hour to hour (progression) may represent more of an emergency that another dog with more severe but stable deficits. If the etiology is something like an unstable vertebral column fracture/luxation, the patient may require more urgent care even if its current deficits are still mild or don't seem to be worsening immediately.

The attending clinician can help sort through these multiple variables by construction of a time-sign graph. The signs on the y-axis can be defined as anything: weakness, discomfort, nociception, etc. The line is determined by the known past history (was the onset gradual or peracute? Has the sign in question worsened, and if so, how rapidly?). The clinician then uses the existing line to estimate what will likely occur over the ensuing hours or days with no further intervention. Clearly a combination that includes either very severe signs (now or projected in the near term) and/or rapid deterioration would warrant more aggressive treatment than situations that are of low severity, non-progressive, or already showing evidence of spontaneous improvement (or improvement from treatment already initiated).

Probably the most urgent situation is one in which the dog has, or is projected to have lost nociception (deep pain). The reason for this concern is that the deep pain fibers are small, unmyelinated nerves in the spinal cord. As such, they are the least vulnerable to the effects of spinal cord compression and injury. Loss of deep pain is interpreted as a more severe spinal cord lesion, and can be a harbinger of irreversible spinal cord injury and permanent paralysis. There are conflicting data in the veterinary literature regarding the prognosis for dogs once deep pain has been lost. Suffice to say that such dogs will rarely recover without surgical intervention (where such intervention can achieve decompression - see below), and that the best data we have suggest that 50% of deep pain negative dogs will recover if they receive surgery within 12 hours of the loss of nociception. Beyond 12 hours the numbers decline toward zero. Recent and very preliminary studies of the use of polyethylene glycol (PEG) in these dogs suggest that this drug may improve the odds for recovery, but this drug is not a substitute for surgery, may not pan out in clinical trials, and has not yet received FDA approval for this use. The prognosis for recovery of paraplegic dogs that have retained sensation prior to surgery is generally felt to be around 80-90%.

Dogs with suspected loss of deep pain (recall from above that we may not always be absolutely sure of this since dogs cannot verbalize their conscious sensation) should be considered hyperacute surgical emergencies. This is especially true since in many cases we do not know the exact time the dog lost nociception. Often a client will come home to find their dog paralyzed, but only know that the dog was fine 8 hours earlier. We have no way of knowing in these cases whether the paralysis - or loss of sensation - occurred 5 minutes after the client left the house earlier or 5 minutes before they arrived back home.

When faced with this hyperacute emergency the appropriate action is to recommend immediate referral for imaging and decompressive surgery, since this offers the best chance for the dog to recover. If such referral is accepted and can be accomplished in a timely fashion there is probably no need for you to administer medication. If the dog is extremely painful it would certainly be appropriate and acceptable to administer an analgesic such as a potent opioid. If there is to be any kind of delay, or long travel time involved, it may also be appropriate to administer intravenous fluids to help preserve spinal cord blood flow, and corticosteroids. Steroids may also be appropriate if the client declines referral. We will address the use of steroids more below.

The question of whether radiographs should be taken is predicated entirely on the differential diagnosis, and whether the information derived will alter your recommendations for treatment, referral , etc. The major indication for taking spinal radiographs is when the differential diagnosis includes fracture/luxation, discospondylitis or vertebral neoplasia. Bear in mind that false negatives are possible with all these entities. Elastic recoil after vertebral trauma may result in minimal and subtle displacement of affected segments, and may not be apparent on a single lateral view. Many vertebral neoplasms (most common is osteosarcoma) are purely lytic, and do not produce the type of proliferative reaction typically seen with appendicular bone tumors. There must be at least 50% loss of bone density usually to see the lesion with plain radiographs, and the presence of overlying structures (ribs, lungs, other viscera) may further hamper the ability to discern a lesion. Early discospondyltis can be very subtle and may be missed even by experienced radiologists, neurologists, and surgeons.

Radiographs are rarely of benefit in dogs with suspected disc disease, unless part of an imaging work-up being conducted for the direct benefit of the surgeon who intends to operate on the dog. In the case of chondrodystrophic breeds where a disc lesion is suspected there is a high probability that lesions will be seen. About 90% of dogs in these breeds have demonstrable degenerative disc lesions that may be evident on radiographs. The finding of disc calcification or even a collapsed disc space may or may not correlate with the actual site of the current compressive lesion. And even if it does, how would it benefit the attending veterinarian or client to "know", for example, that the T13-L1 disc looks "suspicious" as opposed to, say, L1-L2? There was already the suspicion for disc disease, and a high likelihood of finding a lesion, so nothing really has been gained. Moreover, obtaining good spinal radiographs that are not burdened with positioning artifacts requires anesthesia and careful technique. So, if unanesthetized radiography is used one may make the animal more uncomfortable by having to hold it in place, and the value of such films, already dubious by definition, is further degraded. And if it makes little sense to take x-rays at the outset, does it make sense to add an anesthetic (and the cost to the client) in order to pursue this form of imaging? Besides the potential for false positives (seeing lesions that are in fact red herrings), plain radiography can produce false negatives. Not every disc lesion results in obvious collapse of the disc space, or observation of mineralized material within the vertebral canal or intervertebral foramen.

Drug Therapy

Pharmacologic intervention for spinal lesions is an ever-changing and perennially controversial area. For most of us the traditional mainstay has been the use of corticosteroids. There are some good theoretical underpinnings to the use of these drugs. However, there are no published studies in dogs that demonstrate efficacy or safety of steroids for spinal cord disease, in particular disc disease, and also no studies that indicate which drug or dosage is most appropriate. What we are left with is tradition, supposition, and extrapolation from human studies. We do know that steroids can be associated with some serious side effects in both humans and dogs. Life-threatening and fatal gastrointestinal complications are fairly common in dogs. Unfortunately, gastrointestinal protective agents, such as antacids, coating agents (eg. sucralfate), and prostaglandin analogs (misoprostil) have not been shown to be helpful in preventing or treating such complications, especially colonic lesions where these protective agents have no major action. So it is little more than wishful thinking and a bit of defensive medicine to "cover" dogs with such drugs when we choose to treat with steroids.

For many years there was a consensus that dexamethasone was the "steroid of choice" for neurologic disease in general. There is no supportive evidence for this. Initially it was hoped that the use of prednisone instead of dexamethasone would reduce the incidence of GI complications, but this has not been the case. The major differences between various forms of corticosteroids have to do with their relative potency and half-lives. Dexamethasone for example has about 10 times the glucocorticoid activity as prednisone, and also has a longer half-life. The longer half-life makes tapering to alternate day treatment with dexamethasone pointless since the drug will still suppress the adrenals on "off" days.

In recent years the focus has been on the use of methylprednsisolone sodium succinate (Solu-Medrol; UpJohn) for parenteral use in acute spinal cord disease. This focus was generated by meta-studies in human spinal cord injury that demonstrated improved outcomes in people with spinal cord injury that received the drug compared with those who did not. The drug is administered in suprapharmacologic doses where it has actions well beyond the familiar glucocorticoid effects. Various protocols have been advocated ranging from initial boluses of 30mg/kg IV to constant rate infusions. There are two major problems with such use in dogs. First, in extrapolating results from humans we need to look more closely at what constituted "improved outcomes". For very few patients did the use of methylprednisolone (MPSS) mean the difference between walking again vs. permanent paralysis. The major benefit seen in MPSS-treated patients was in quadriplegics, where those in the treatment group had higher recovery rates of such things as finger movement. For a person, this can be a huge advantage; the difference perhaps in being able to grasp a fork and feed oneself vs. not. A similar benefit would be of little use to a dog. Our demands are actually much higher, where we measure success of spinal cord injury management in terms of restoring the ability to walk. It is doubtful that any medication will ever have this ability. The second issue is that there have been no prospective studies in dogs to evaluate the therapeutic index for MPSS (efficacy vs. toxicity).

Another issue relative to the use of MPSS is the timing. It is known in humans that to be effective the drug needs to be administered as soon as possible after the injury, and that administration more than 8 hours after injury is actually associated with poorer outcomes. Assuming the same would be true in dogs, it means that for many of our patients, where they may not be presented for more that 8 hours after the onset of signs, we are on questionable ground using this drug. However, in those instances of acute injury of less than 8 hours' duration it is probably reasonable to employ one of the MPSS protocols. Personally, I find my greatest use of this drug is as a bolus or CRI in an adjunctive mode to surgery (both emergency and elective), since it is well known for steroids and many other experimental drug treatments that they are most effective when given before the spinal cord is injured.

As the pendulum has swung somewhat away from the use of steroids for dogs with suspected disc disease there has been an increasing use of NSAIDs. These drugs certainly can be useful for pain control, and in that setting may be as or more effective, and probably safer than steroids. The downside is that in those dogs that progress to more severe deficits, where the use of steroids may be indicated (for example, a dog that initially presents with ambulatory paraparesis but who progresses rapidly to paraplegia), the prior use of NSAIDs precludes the concomitant administration of steroids without greatly increasing the risk for GI complications.

Another good way to manage the sometimes intense pain associated with spinal cord injury (the pain is usually radicular pain from compression and ischemia in the affected nerve roots, possibly combined with meningeal inflammation) is with the use of opioids. For hospitalized patients this can be accomplished with injectable forms such as CRI's of fentanyl. For outpatients transdermal patches may be helpful.

Some of the pain associated with nerve root compression is associated with muscle spasm as well. In dogs with cervical disc disease this may be the most predominant clinical sign. It is therefore useful to employ muscle relaxants in the treatment regime. Diazepam is an excellent muscle relaxant, but in dogs reaches inconsistent and generally poor blood levels after oral administration. More effective is injectable or rectal administration. The former, if given intramuscularly, can cause pain at the injection site. More frequently used are drugs such as methocarbamol (orally or parenterally). The exact mechanism of action of methocarbamol is unknown.

One of the most important treatments for dogs with spinal cord injury is preservation of spinal cord blood flow. One of the benefits of suprapharmacologic doses of MPSS rests in the drug's ability to do just this, even as it scavenges free radicals released during reperfusion. More basic, the use of intravenous fluids may be beneficial in this regard. Because of pain and fear dogs may not eat or drink adequately, and some treatments may be dehydrating (for example, steroids have a diuretic effect). Dogs that are hospitalized because of spinal cord disease would best be managed by having them on maintenance fluids.

Serial Exams

With disc disease in particular we must be sensitized to the possibility of progressively worsening signs. Discs tend to extrude in stages, and so a dog may initially present with vague signs where disc disease is not even on the initial differential list, or very mild signs of CP deficits and ataxia. Whether treated as an inpatient or outpatient (that decision being based on the time-sign graph discussed earlier) it is important for the client or veterinarian to reassess the signs as time goes by. The more severe/acute the initial presentation the stronger the requirement for in hospital observation, and likewise for frequent rechecks. This may mean serial exams as often as hourly. In turn, that might suggest referral, if needed, to a 24-hour care facility. The most critical observation of course would be any loss of nociception. By spotting this development immediately, the opportunity for prompt intervention early in the 12 hour "window" is increased, and the prognosis for recovery is improved.

Whether a patient is managed as an inpatient or outpatient, and regardless of particular medical therapy, it is also thought to be beneficial to keep dogs cage rested while they are being treated. Although no one knows if disc displacement is induced by heavy exertion, it is probable that activity could promote exacerbation by evolving more disc material from the interspace.

Bladder Management

One of the most frequently overlooked aspects of patient needs when dogs are paralyzed is the requirement for emptying the bladder. Most of our patients have UMN bladders since most disc lesions occur well cranial to the S2 segment. These dogs tend to retain urine and only void reflexively when the detrussor is overstretched. Failure to assist in emptying these bladders results in discomfort, damage to the detrussor, and promotion of urinary tract infections. If the detrussor is chronically stretched there may be detrussor atony even after the pudendal nerves have recovered their connection to the brain. Urinary tract infections have a negative impact on overall patient health, and are one of the leading causes of death in human paraplegics and paraplegic dogs. It is far too common to have dogs seen by a primary care veterinarian for paralysis that go hours to days without attention to this issue. Too many veterinarians misinterpret the reflexive voiding of urine as voluntary control, and thus institute no steps for active bladder emptying. My recommendation is to assume that at any dog with absent voluntary motor function has concomitant bladder paralysis, and to initiate management immediately.

The bladders of female dogs are generally easier to manually express than males', although obesity can make this difficult. Likewise, smaller dogs are easier to manage with manual expression than larger dogs. For any dog where manual expression is difficult it should be a consideration to use a catheter. From the urologist's standpoint it is preferable to use intermittent catheterization to indwelling catheterization. However, from a practical standpoint it is much easier on the patient and nursing staff to place an indwelling catheter, even in females. Ideally, Foley-type catheters should be used, placed using proper aseptic technique, connected to a sterilized collection bag. Bags can be weighed daily or twice daily to ensure and record proper urine output, and the bags emptied or replaced using clean techniques.

If dogs sustain any soilage or scalding from urine or feces they should be bathed as needed and appropriate topical agents employed to protect the skin.


There are several different surgical procedures described in the literature for disc disease. They can broadly be divided into two classes: Operations that effect decompression of the spinal cord and nerve roots, and operations that remove disc material from the interspace. The latter category includes disc fenestration, laser ablation, and chemonucleolysis. The greatest potential benefit to these latter operations is for prophylaxis - prevention of other discs from displacing - and/or prevention of more disc extrusion from a site where rupture has already occurred. They are of dubious therapeutic value once disc material has escaped into the vertebral canal or intervertebral foramen.

Once there is actual displacement of material in the vertebral canal decompressive surgery is needed. The different operations, laminectomy, hemilaminectomy, pediculectomy, partial corporectomy, foramenotomy, and ventral slot all share one common feature: They are simply methods for gaining access to the vertebral canal. They are bone removing procedures. None of them by themselves are inherently decompressive. The spinal cord is tethered within the vertebral canal by the nerve roots. Removal of bone does not allow the spinal cord to "move away" from the compressive mass. To achieve true decompression the offending mass must be removed. This applies as well to fractures and luxations. In many of these cases decompression can be accomplished just by restoring the normal anatomic alignment of the vertebral canal without the need for any bone removal at all, since the "mass" in this instance may be the malpositioned vertebral bone.

There are some instances with disc disease where surgery will be of little to no benefit. Besides the previously noted guarded prognosis with loss of nociception, a distinction must be made between spinal cord injury and spinal cord compression. The former refers to all the many microscopic and biochemical alterations that occur after trauma, some of which tend to be self-propagating. One of the major determinants of how much "injury" occurs is the amount of kinetic energy that is delivered as a bolus to the cord. Compression is a measure of spinal cord deformation, which is directly related to the volume of space-occupying mass in the canal. The relationship between all these variables is described by the equation KE = ½ mv2. Analysis of this equation makes it apparent that the amount of energetic contusion, which is proportional to spinal cord injury, is much more strongly influenced by the velocity component than by the mass component. From the clinical standpoint this has several implications. It tells us that we can sometimes have very large compressive masses, but if the compression is slowly applied (velocity approaches zero), there will be essentially no "injury", and in fact signs may be very mild despite marked deformation of the spinal cord. This has been demonstrated experimentally and occurs clinically with slow growing spinal cord tumors, and even more frequently with cervical disc disease. Conversely, a small mass, such as a tiny fragment of disc, may be ejected at very high velocity, with the result being a devastating spinal cord injury with no real "mass" to remove. Surgery in these cases is usually unrewarding as the lesion is limited to spinal cord swelling and perhaps some epidural hemorrhage if the sinus was lacerated.

When dogs are presented for presumed acute disc displacement there is usually some combination of injury and compression, the degree of each being unknowable. This is probably the major reason we see such variance in outcome, especially the time required to improve from paraplegia to recovery of voluntary motor function and recovery of continence.

The knowledge that dogs can have massive compression with yet modest deficits, or no deficits at all and just discomfort is very important to bear in mind. We frequently see dogs that are presented simply because of pain, most commonly cervical, but sometimes thoracolumbar. There is a mistaken assumption that the absence of neurologic deficits implies a non-surgical lesion. Put another way, too many clinicians assume that there is a correlation between amount of compression (mass) and severity of signs, such that in dogs where the only sign is discomfort they must only have minimal to no compression, while dogs with peracute paraplegia must have substantial compression and therefore would benefit from decompressive surgery. As we have seen from the foregoing discussion of the role of velocity, it should be borne in mind that a dog that has discomfort alone, but is unresponsive to non-surgical therapy, is a candidate for imaging and decompressive surgery if a mass lesion can be identified.

Post-operative management follows similar guidelines as for pre-operative care. In most cases steroids will be discontinued, although a more gradual weaning process would be in order if the dog had been on chronic therapy prior to surgery. An important addition to post-operative management is the initiation of physical therapy for rehabilitation. It is beyond the intent and scope of this presentation to discuss all the methodologies employed, but suffice it to say that the introduction of more intensive physical therapy in veterinary medicine shows great promise for speeding and improving recoveries from spinal cord disease.

De-Confusing Hip Dysplasia
and Explaining the Disease to Clients and Breeders

For an old disease, the questions I am asked by clients and referring veterinarians, and the timing and reasons for referral (or delayed referral), all suggest that hip dysplasia is still a poorly understood and confusing disease. The proliferation of new data, new methods of diagnosis, and new treatments over the years have only served to increase the confusion and frequent misinterpretation by the lay public, as well as to make it difficult for the general practitioner to keep up with the latest details. While it is impossible to incorporate, or even summarize all the knowledge gained about this common disease in anything short of a textbook, I will attempt to touch on those practical aspects that will hopefully assist you in your understanding and discussions with clients.


The term "dysplasia" literally means "abnormal growth and development". It tells us nothing about radiographic signs, clinical signs, causation, or treatment. Some important concepts to always keep in mind:
  1. dogs can be dysplastic and have no appreciable lameness or discomfort
  2. dogs can be dysplastic, and depending on the positioning technique employed, be interpreted as having normal hips radiographically
  3. radiographic signs are evident (but as in (2) above, may be missed)
  4. the presence (or absence) of clinical signs is poorly correlated in many cases with the radiographic findings. Dogs can be crippled in the face of fairly mild radiographic changes, while others may have advanced osteoarthritis and yet function near normally
  5. there are frequently other causes of lameness or soreness in dogs that have concomitant hip dysplasia or hip arthritis.
Putting (4) and (5) together, it is critical to avoid jumping on "bad hip dysplasia" as a diagnosis when you evaluate a lame dog, forgetting that the awful looking hips may be a "red herring"; the problem may well lie elsewhere.

In humans, hip dysplasia is typically a congenital lesion. Heritability is poorly described. It is usually diagnosed by palpation at or near birth, confirmed with ultrasonography of the joints, and in most instances successfully resolved by special diapering techniques. If properly recognized by the obstetrician or pediatrician it rarely progresses to the degree where operative treatment is required.

This contrasts sharply with the situation we see in dogs. Canine hip dysplasia (CHD) is a developmental disease that usually has its onset at 3-6 months of age. There have been anecdotal claims in the literature by veterinarians who say they can diagnosis the disease in puppies as young as 6 weeks, and moreover, can eliminate any progression of disease (i.e. "cure the dysplasia) with manipulation techniques similar to those used in human infants. These claims have never been substantiated or scientifically reproduced or tested. As best as we can tell, dogs with CHD are born with normal hips, and the disease ensues during adolescence.

Hip dysplasia can occur and is seen in any breed or size of dog, is not uncommon in cats, and has also been documented in many other domestic animal species. However, the clinically significant form of hip dysplasia is that which typically affects large breed dogs. Pedigree analyses clearly demonstrate that while not congenital, CHD is heritable and tends to be familial. Unfortunately, it is polygenic and so defies any simple diagnostic test for genetic markers at this time. This is critical, since it relegates our ability to diagnose the presence of this genetically based disease to evaluation of phenotypic characteristics. Wouldn't our lives, and those of breeders, be simplified if we could diagnose hip dysplasia with a blood test, before the patient ever developed the disease or was used for breeding?! Work being spearheaded by Rory Todhunter at Cornell and facilitated by the completion of the canine genome project holds promise for the development of such tests in the future.

It is unlikely that the "hip dysplasia genes" directly cause hip dysplasia. More likely, they cause other, predisposing conditions that combine to create the disease entity we recognize as CHD. For example, factors such as rate of growth, ultimate size at full growth, composition and volume of synovial fluid, composition and structural characteristics of muscles, ligaments, tendons and bones, and cartilage physiology are all under the influence of genetic factors. As scientists search for the "holy grail" - the "ultimate" fundamental "event" that precipitates CHD - they instead find that all of these factors, and more, play a role in pathogenesis. It is analogous to the searches for the fundamental elements of diseases such as shock or spinal cord injury, and just as frustrating.

Basic genetics also teaches us that the phenotype is the result of a genetic predisposition combined with "environmental" factors (P=G + E). We have a poor understanding of what the "environmental" factors are in CHD. Does uterine positioning play a role? Birth order or puppy size at birth? What about the two "biggies", namely diet and exercise during growth? Despite years of research, we still have inconclusive evidence of the importance of each of these, and other, factors.

There are some things which do seem evident, and which we can use to try and help reduce the incidence or severity of CHD. On the "G" side, we know that breeding affected dogs together greatly increases the risk of subsequent puppies having the disease. We know equally well that breeding together unaffected dogs lowers the incidence of CHD in subsequent generations. Therefore, the use of basic population genetics will allow what is termed a "genetic drift" towards or away from a particular phenotypic characteristic. The difficulty here stems from two major limitations: The first is that CHD is polygenic, the genes are widely disseminated in the affected breeds, and the involved genes are probably useful to the species for other traits. We cannot hope to "eliminate" CHD genes completely because there probably are no such exclusive things. The second, as touched on above, is that we lack any direct genetic testing. Thus, our detection of CHD, for the purpose of inclusion of a particular dog in the breeding population, is predicated on the phenotypic findings. This may cause us to fail for several reasons. First, the particular test we are doing to detect CHD, such as OFA-type radiographs, may lack enough sensitivity (i.e., it produces "false negatives"). This allows affected dogs to "slip through" and contribute to the genetic pool, all the while we are thinking they are "normal" dogs. Second, even if a given dog is truly phenotypically normal, has not the slightest whiff of dysplasia, what does this tell us about its genotype? The answer is absolutely nothing! We cannot assume there is a correlation between "dose" of CHD genes and presence or severity of clinically measurable disease for a given individual. As with many genetically based disorders, there can be "silent carriers". Such patients, though normal themselves, may be just as likely to pass along "bad" genes to the pool as dogs that are severely affected. In these instances it may be as valuable to find out if the "normal" puppy has any affected siblings, cousins, or offspring from those relations. Such a finding would strongly suggest that this normal puppy is a carrier, and should be excluded from breeding with the same vigor as one would the puppy with gross hip dysplasia.

Of course the biggest problem with all the above is trying to remember these principles from Genetics 101 ourselves, and then having to explain them to clients and breeders. The latter, in particular, are often worse than ignorant, because they have just enough knowledge and misinformation (or misinterpretation) to be dangerous (plus, in an unfortunate number of them, enough arrogance to feel like they "know it all").

On the "E" side of the equation, we recognize that one of the apparent risk factors for CHD, as it is for a variety of pediatric musculoskeletal disorders (panosteitis, osteochondrosis, elbow dysplasia, etc.) is rapid growth during adolescence (4 months to about 1 year of age). We know, and can inform our clients that genes largely determine the ultimate size of the patient. This is important, because clients do not generally purchase a shepherd or a Rottweiler so they can end up with the smallest dog possible; they want a giant, and mistakenly think that feeding the dog a rich diet, and more of it, will achieve this result. So too, the rate of growth is to some degree genetically influenced, but at least this can be altered by diet. It seems prudent, therefore, to try and get dogs from breeds at risk for CHD to grow more slowly to whatever degree this eliminates or reduces one of the risk factors for development or progression of CHD, it should be useful. What is unknown is whether it is the amount of food, the caloric density, or the composition of the food that is most critical to limit. Is excessive protein a factor, or is it simply that higher protein foods also tend to be more calorically dense? What about minerals? We know that calcium, magnesium, zinc and copper levels can influence the development of osteochondrosis in different species. Do they affect the development of CHD? How about vitamins? Have we not all heard stories of Vitamin C being a preventive or cure for hip dysplasia?

Many questions are raised, but so far few have been thoroughly answered by the nutritionists. Even as commercial diet manufacturers produce "large breed growth formulas", it is still rough guess. My own recommendation to clients is to feed puppies from breeds at risk an adult dog food (or one of the large breed growth diets) from weaning, and to especially avoid calorically dense or high (>22-23%) protein diets. I explain that this may not prevent hip dysplasia (or other pediatric disorders), and may not lessen the severity of any problems that develop, but it seems to be the best we can do for the present time. There are currently no valid data to support the use of megadose therapy with vitamin C, or other vitamins or trace minerals, as a preventive or treatment for CHD. We will touch later on the controversy surrounding the use of "chondroprotectants" and other dietary supplements.

Another poorly understood factor is the role of exercise on the development or progression of CHD. We have all seen recommendations at either end of the spectrum, from those that suggest cage rest to those that champion vigorous activity. There are few reliable data to help settle this debate. My own bias is to allow puppies to run and play as vigorously as they desire. First, this allows better development and conditioning of the muscles and other periarticular soft tissues. To what ever degree these soft tissues can then stabilize the hip, it may help lessen the effects of any underlying laxity. Second, the joy of being a puppy is to be able to run and play. If the puppy is unable to be normally athletic (which usually suggests that CHD is affecting its quality of life), then it should be evaluated, and if CHD is the cause, treated. From a teleological standpoint it seems unlikely that exercise would cause CHD, or exacerbate it, in a puppy that was otherwise going to be unaffected or minimally affected. That it might exacerbate the clinical signs of CHD in an affected dog is certainly possible, or even probable, but what is the point of having a puppy on continued confinement just so it won't be lame doing the things that make it a puppy? And even if we were successful in "hiding" the early signs of CHD, would even a lifetime of such restrictions necessarily limit the subsequent development of secondary degenerative joint disease, and accompanying discomfort? Let 'em run, I say, and if the puppy can't perform as suspected, find out why and correct it.


While it I useful to investigate and understand the histologic, ultrastructural and biochemical change that occur in the cartilage, bone, joint capsule, tendons, ligaments and muscles in dogs with hip dysplasia, as clinicians we need to summarize this knowledge into a form that helps us and our clients comprehend the progression from a normal to dysplastic to arthritic hip. We should keep in mind, however, that as we simplify explanations we may oversimplify, and thus be left with exceptional cases that prove the rule [remember that to "prove" in this adage is to "test", as in "proving grounds", not to "confirm"].

The fundamental event or problem with early hip dysplasia is the development of laxity between the femoral head and the acetabulum. This may or may not be accompanied by incongruency - poor fit - between the ball and socket. The hip was evolutionarily designed to function with rotatory motion only; it was not designated for more than miniscule translational forces and strains. If the ball and socket are congruent, the primary lesion then is in the synovial fluid and soft tissues that surround the articulating bones. Whether there is a primary synovial effusion, or whether this occurs secondary to soft tissue laxity is not well defined. Regardless, the problem from the dog's standpoint is that its hips are starting to do things for which they were not designed.

In its earliest stages, this laxity is manifested as what I term "mechanical lameness." There may yet be no damage to the articular cartilage, no changes in the subchondral bone, no changes suggestive of arthritis, and yet the puppies are lame. They struggle to get up from a recumbent position. When they walk they have a "sashaying" or wiggling gait. When they run, they tend to bunny hop. They sit funny. In fact, it is often the trainer at obedience class that first recognizes a problem, when they note that a puppy always sits with its legs out to one side or the other, but not squarely on its haunches. With time other signs ensue, including poor exercise tolerance. The puppy has to stop and sit during long walks. It "hangs back" when playing with cohorts. It is "quieter" than other puppies. Yet, when you question the client, or examine the puppy, clinical signs of pain are conspicuously absent. The puppy seems happy, never cries, and if treated with an analgesic/anti-inflammatory medication, seems about the same. These signs typically begin at about 4-5 months of age.

As the underlying disease progresses, do so the clinical signs. However, the rate of this progression, and the correlation between the changes in the hip and the clinical degree of dysfunction, are not uniform. Not only do different dogs have progression at different rates, individual dogs often have discrepancy between the severity of changes, and signs, between their two hips. Furthermore, the rate of progression is not linear, and may wax and wane much as overall growth tends to occur in "spurts."

In response to the translational instability, progressive changes occur in the soft tissues, cartilage and bone. This is the beginning of the process we recognize as osteoarthritis. The soft tissues thicken, probably a physiologic hypertrophic response (the soft tissue equivalent to Wolff's law). Stresses on the articular cartilage and subchondral bone lead to a loss of homeostatic balance between production, repair and resorption, the net result is increased "wear and tear" on all components of the joint. At this point discomfort may start to become an increasingly apparent clinical sign. Sprain-type injuries to the joint capsule and supporting tendons/ligaments, coupled with microfractures in the bone and cartilage can stimulate nociceptors in all of these structures, with the result of conscious pain/soreness.

These changes may occur quickly and puppies may start evidencing a pain reaction by the time they are 6-7 months old, or they can progress more modestly, with no signs of pain until 8-9 years of age (or never). Further complicating our clinical awareness, individual patients (human or animal) have differing thresholds or tolerances for painful stimuli. As we all know intuitively, but require continued reminders from the neuroscientists, the phenomenon of pain is not just mediated by stimuli and receptors, but by the processing of data in our central nervous systems. Some people can tolerate what should be excruciating stimulation of their nociceptors, while other people suffer horribly from parts of their body that they may no longer even have (phantom limb syndrome). With time, the instability is reduced as the soft tissues thicken, but the price paid is the development of periarticular osteophytes, cartilage fibrillation and degradation, eburnation of bone, and all the other hallmarks we use to define degenerative joint disease (osteoarthritis). Once again, the simple presence of DJD, as seen on radiographs or palpated on examination, does not uniformly correlate with clinical disability of signs of discomfort.

The transition from mechanical lameness to eventual painful osteoarthritis is preceded, as we have seen, by a shift from the effects of instability to the effects of wear and tear. This leads to a natural bimodal clinical course in many dogs with CHD. Signs of mechanical lameness develop and predominate for the first year of life. As the hip becomes less unstable, but before DJD has worsened substantially, the dog may experience a period of apparent clinical remission. The dog is limping less, functioning better, and has less need for any analgesic agents. This may lead the owner to erroneously conclude that the dog has "outgrown" its hip dysplasia. If the disease, and this natural biologic phase are not properly explained to the client, they may resent you for "incorrectly" telling them their dog had hip dysplasia in the first place, or worse, having recommended corrective surgery that "clearly" wasn't needed. As we shall see when we discuss diagnostic procedures in more depth, this phase can also result in a negative Ortolani test, leading you to incorrectly inform a client that their dog does not have hip dysplasia. This phase of apparent remission may last anywhere from a few months to generally 3-4 years, before clinical signs recur with a vengeance.

At the other end of the spectrum are those dogs who have been minimally to completely unaffected clinically (but who are dysplastic), who then present with acute pain or lameness. Your examination reveals no other orthopedic or neurologic problems, and they are painful in their hip. Your radiographs reveal no DJD. What is the explanation? For myself, I attribute these situations to an acute sprain of the soft tissues of the hip, recognizing that these dysplastic dogs are at increased vulnerability, even as they are otherwise seemingly unaffected by their primary disease. In these cases I manage with rest, time and analgesics, and most recover to their "pre-sprain" status without the need for radical surgery. They certainly bear monitoring, but may not need any further treatment, as we shall see below.

In those dogs that go on to develop severe osteoarthritis there may eventually be a gradual decline in their functional and comfort level. They may be less active, or, more commonly, be just as active but paying a terrible price afterwards. Stiffness, soreness, trouble rising, difficulty or reluctance to ascend stairs, trouble or refusal to get into vehicles, are all hallmarks of disabling painful arthritis. In some patients their personalities will be affected, and they may become less patient with people, snarl when approached or handled, or start "acting old", irrespective of their chronological age.


Since, as we have seen above, we cannot always uniformly agree on what actually defines "hip dysplasia", it should not be surprising that there is some controversy and confusion about how to diagnose the disease. The safest course of action for us is to presume that all dogs of predisposed breeds be considered potentially dysplastic until proven otherwise. This means that even in the absence of signs, any dog intended for breeding should be appropriately evaluated. Moreover, because of the consequences of progression to DJD, it is a consideration to screen all dogs for CHD. However, this latter concept needs to be tempered by the recollection that dysplastic dogs may be clinically fine.

My personal recommendation is to examine the hips of any dog at risk whenever the opportunity presents itself (e.g., concomitant with anesthesia for surgical sterilization), or if any of the early stages of possible hip dysplasia are seen or reported. The examination has two components, and both are critical to accurately assess the patient. The first is some form of radiography, and the other is palpation. Either alone is often insufficient, or in a number of cases, misleading.

A critically important point here: The most common mistake made by veterinarians and breeders is to forestall the assessment for hip dysplasia until the dog is two years of age. This stems from a misunderstanding of how the OFA examination plays a role in the management of the disease. When we think about CHD we need to think on two levels. The first is whether a given dog is likely to be good or bad for the gene pool. The second is what effect any dysplasia is having or likely will have on this individual animal. It is the latter that many breeders and too many veterinarians lose sight of. Most dogs that will develop hip dysplasia in their lifetimes will have detectable signs before a year of age, and usually by 4-6 months of age. A very few "stragglers" may not develop detectable CHD until 2 years of age or beyond. The OFA certification process is a statistically-based effort to weed affected dogs out from the breeding pool. By two years of age, 95% of dysplastic dogs will be detectable. This accounts for the OFA being unwilling to certify an animal as being FREE of hip dysplasia until it is at least 2 years old. However, probably 90% of dogs with hip dysplasia will be detectable well before this age. Thus, the problem for breeders is the potential for false negatives, i.e., concluding an animal is free of CHD when, in fact, it is affected. The problem is not false positives. Since animals do not "outgrow" hip dysplasia, the detection of the disease at a young age is conclusive. Specificity is high. If the animal has palpable and/or radiographic evidence of CHD, it is dysplastic, no ifs, ands, or buts. Therefore, what is the advantage to a breeder investing two years into a dog intended as a breeding animal, only to have those hopes dashed when it is rejected by the OFA, as opposed to potentially getting this information at 4-6 months of age? What the breeder needs to know is that if the hips look and feel good on the exam(s) prior to 2 years of age, there is still the slight possibility that it may yet fail (if the dog is one of those "stragglers") when it goes for certification at 2 years of age or after.

From the viewpoint of the pet owner, this possibly earlier information is even more critical, since early detection increases the options available for management. It makes no sense to tell a client, who brings their dog to you with a complaint of signs consistent with CHD, that "we cannot x-ray the dog until it's 2 years of age because hip dysplasia can't be diagnosed before then." I hear this all the time, so I again urge you to distinguish between the use of x-rays to diagnose hip dysplasia and x-rays to certify an animal as being free of hip dysplasia. If you send films to the OFA of a patient under 2 years of age they will happily write back to tell you the dog has mild, moderate or severe dysplasia (and will never be certified). If the dog is borderline or better, they will issue a preliminary report ("things are looking OK!), but withhold certification until they see those 2-year films.

Because we have equated laxity or instability with the fundamental lesion of CHD, it is detection of this laxity that determines the diagnosis. If our presumption is incorrect then the detection of instability, in the absence of any supportive clinical or radiographic findings, may cause us to inappropriately diagnose some dogs as having hip dysplasia when all they have is an unstable hip. The implications of calling a dog dysplastic are that it should be eliminated from the breeding population (a minor blow to the pet owner, but a potentially devastating blow to a breeder), and that we "expect" it to progress to arthritis and lameness. It might even inspire some surgeons to recommended prophylactic surgery. Thus, the potential of a false positive diagnosis does exist and may have serious consequences.

The majority of veterinary orthopedic surgeons believe that specificity is high, and that our paradigm of associating laxity with dysplasia is correct. If we accept this, then the challenge is how do we demonstrate this laxity? In part this is done by eliciting a good history and by observing the animal awake as it rises, sits, and ambulates. The sine qua non of detecting laxity is actual palpation of the abnormal translational movement between the femoral head and the acetabulum. The presumption is that laxity between the ball and socket, even in an immature dog, is pathologic. This appears to be true, but is not absolutely proven. Should the effects of sedation, muscle relaxation, or in intact females, estrus, be completely discounted? There are no definite answers. I do know that in a young dog, in whom I would not tolerate hip laxity, I will quickly write off a mild cranial drawer sign in the knee as being "physiologic drawer." I do this based on experience, but it opens some questions that would be worthy of investigation.

The detection of translational movement is best demonstrated via the Ortolani test. The test consists of two parts. The first is detection of movement of the femoral head out of the acetabulum. It is inconsistently detected, even in dysplastic dogs, but when felt is termed "Barlow's sign." The second, more commonly observed, is palpation of the femoral head dropping back into the acetabulum. This is termed "Ortolani's sign." An alternative maneuver is to lift the femoral head directly out of the acetabulum (no abduction or adduction), and this is called "Barden's sign." They are all variations on the same thing! There is a certain amount of skill required to do these examinations with confidence, and the most common problem is a false negative exam. Some cases are easy (the client will tell you they actually hear an audible click or pop as the dog moves, and/or you can feel the laxity with the dog standing in the exam room), while others require an experienced examiner working with the dog under heavy sedation or anesthesia. If you are going to screen dogs for hip dysplasia in your practice you should develop and refine this skill, or else you will be missing some cases!

It is difficult, and can be misleading, to quantify the laxity based on palpation. Does a "small" Ortolani sign imply more mild CHD than a bit "clunker"? Does one require more or earlier treatment than the other? As we mentioned above, laxity is variable in the early stages of the disease, and may resolve as the disease transitions from "hip dysplasia" to hip arthritis. While the presence of a positive Ortolani sign always implies a lax, presumably dysplastic hip, what can we infer from the absence of an Ortolani sign, or just a mild sign? There are three possibilities: The first is that it is a false negative; i.e., the dog really has lax hips and you simply missed it. The second is that the dog has really good (i.e. normal) hips. The third possibility is that the hips are really bad, i.e., they have started the degenerative process, and, as a result of round ligament hypertrophy, bony remodeling of the acetabulum, periarticular fibrosis, or severe luxation, they are no longer unstable and/or there is no longer enough of a socket for the ball to drop into. So, a stable hip is really either very very good or very very bad. The distinction between these is made using radiographs. It will quickly be evident whether the hips appear well seated and congruent, or whether they are dysplastic.

A number or radiographic techniques have been described for hip dysplasia evaluation. The most common is the OFA-style, ventro-dorsal extended leg view, with inward rotation of the hips to align the patellae over the distal femora. This view is extremely useful in detecting degenerative changes in the hip. Radiologists love this view because they can conjure and measure a variety of "angles" as an aid in assessing the "fit" of the hip joints. This view does have drawbacks, leading to either false positive or false negative findings. The common false positives are "flattening" of the femoral head and increased angle of inclination. In most instances where the femoral head appears flattened within the acetabulum, we are really just seeing the fovea capitus (insertion site of the round ligament) accentuated. Coxa valga, an increase in the angle of inclination (the angle with which the femoral neck joins the femoral shaft viewed from the AP plane), used to be considered a "primary" lesion of hip dysplasia. It turns out that when measured directly on the bone, most dogs with CHD have normal angles of inclination. The apparent coxa valga seen on VD films is a radiographic positioning artifact. Trigonometric resolution of the angle, using the VD and lateral views in conjunction, yields the true, usually normal angle.

The most common false negative - and it's a whopper - is that this view will sometimes (often) suggest a normal hip when the hip is palpably unstable! The extended leg VD radiograph will always tend to underestimate the presence or severity of laxity, but never (unless grossly malpositioned with extreme rotation) falsely make a good hip look bad. The implications of this fact are dramatic. Because the certification of "dysplasia-free" by the OFA is based solely on this film, and the radiologists have no access to palpation findings or clinical history, clinically affected dogs with hip dysplasia, and not just "silent carriers" slip through the screening and registry process and are entered into the breeding population! This is why I stress the importance of palpation as part of the hip dysplasia evaluation. If the goal to reduce the presence of affected dogs in the gene pool, and if our method of detection is based on phenotype, we will never make sufficient progress if our detection method is insensitive. The reason for this effect of OFA-type films improving the appearance of a lax hip is explained by torsional mechanics of the joint capsule. In order to rotate within the joint capsule envelope even a normal, non-dysplastic hip has to have a small degree of translational laxity (usually non-detectable). But much of this laxity is reduced when the hip is twisted from a neutral, standing posture to one where the hips are maximally extended, since the joint capsule "winds up" during the maneuver.

Because detection of the Ortolani sign (or equivalent measures of hip laxity) is not quantifiable (and varies based on the skill of the examiner), and because of the critical nature of this information in determining which dogs should not be bred, a method for radiographically documenting the Ortolani sign is desirable. This forms the basis for the PennHip examination introduced by Gail Smith. Using positioning of the femurs that is more physiologic (dogs do not walk with their hips in full extension), and devices to either compress the hips inward (Ortolani-like maneuver) or distract them outward (Barden-like maneuver), the PennHip exam will result in films where the movement of the hip from one position in the acetabulum relative to another can be seen and measured. This measurement is referred to as the distraction index, or DI. Because it is a number, it is by definition quantifiable, and this allows comparison between dogs, between groups of dogs, or over time for a given dog. The complete PennHip exam also includes the standard OFA-type film so other relevant changes can be seen and documented for a more complete evaluation of hip status.

There is a bit of a battle going on between PennHip and OFA. In part it is ideological. As we have mentioned, it may be that laxity is not always a harbinger of or pathognomonic for hip dysplasia. If this is true, then high distraction indices may falsely implicate some dogs as being dysplastic when they may not be. This is the OFA stance. On the other side, there are many documented cases of OFA-certified dogs going on to develop clinical hip dysplasia, and producing affected offspring. Which is worse, a false positive or a false negative? For the individual, it is a wash. If dysplasia develops, even if caught late, it can still be treated with success. If diagnosed with dysplasia but never clinically affected, it will probably never be subjected to any treatment. However, for the breeder, a misdiagnosis either way can be devastating. A false positive will cause the breeder to refrain from breeding a dog that may be normal, with all the associated loss of time and investment in that dog. Worse, though, is the false negative (likelier with the OFA exam), since this will allow an affected dog into the breeding pool, and potentially harm the breeder's reputation as dysplastic puppies or grandpuppies start appearing in the population.

PennHip is moving beyond individual evaluation of hips to looking at DI's by breed. In one seminal paper they have already demonstrated that given equal DI's, German shepherds are far more likely to develop subsequent radiographic and clinical signs of CHD than are Rottweilers. This means that over time a veterinarian and breeder will be able to not only look at a patient's individual DI, but determine where they are on the normal distribution curve for that particular breed. They will have some predictive value for that dog's likelihood for developing other signs of hip dysplasia. It will also be possible to establish guidelines for what amount of distraction (nee, laxity) is "acceptable" for a given breed, and breed only those dogs below this level. By then following the progeny, it will become clear whether this phenotypic test is successfully causing the favorable genetic drift we mentioned earlier, and reducing the incidence of CHD.

Another way that PennHip compares favorably to OFA is that there is no selection of which films to submit. With OFA the breeder gets to choose whether to submit the films. This means that some obviously affected dogs are never "seen" by OFA, making any interpretation of their population data (purporting to show reduced incidence in breeds over time) suspect. ALL PennHip films are submitted for evaluation to eliminate this potential bias.

We should keep in mind, too, that the "battle" between the OFA and PennHip is not just intellectual, but is also a battle over dollars. The OFA has been, and PennHip seeks to become the "standard" for hip evaluation and registration in the United States. These are both for profit enterprises, so they are fighting for the hearts, minds and wallets of veterinarians and breeders. I think the science behind PennHip is superior, but because it is more technically demanding, requires certification to do (and that certification is paid for by the veterinarian), is more costly (more films, more personnel, more time), and takes the veterinary radiologists out of the loop (OFA is a money maker for the member radiologists), I suspect it will be a long time, if ever, that PennHip supplants OFA in the marketplace.

Other radiographic views are also used as part of the hip examination. The lateral view of the pelvis is probably the least useful, except in its ability to also evaluate the lower lumbar spine (important as part of the work up before concluding that hip disease is the cause, or sole cause, for the patient's disability), and for post-operative evaluation procedures such as pelvic osteotomy or hip replacement. The lateral view of the femur and hip (with the opposite limb pulled into abduction to limit superimposition) is useful for pre- and post-operative measurements of femurs for prosthesis size. In cases where it is impossible to get good enough extension for the standard OFA-type radiograph, a true AP view of the femur can be approximated using a laterally recumbent dog and a horizontal beam technique. I find frog leg VD views helpful in the evaluation of young dogs (in addition to the standard OFA-type VD), as a way of assessing hip congruency, and the ability to seat the ball into the socket. This is especially helpful when pelvic osteotomy is being contemplated. We used to emphasize views that would permit direct measurement of the angle of anteversion (the angle that the femoral neck makes with the direct lateral-to-medial cross-section of the femur as measured by a line joining the base of the femoral condyles). These were difficult to obtain, and this angle can be calculated more easily, using trigonometry, from well-positioned VD and lateral views of the hip. The importance of increased angles of anteversion, like those of inclination, has probably been overemphasized in the literature. Most dysplastic dogs have normal angles of anteversion. A dorsal acetabular rim (DAR) radiograph has been described. It is sometimes useful, not always easy to obtain, again the information can be cleaned from interpretation of more "standard" views. Another radiographic technique that has been promulgated is the dorsal lateral subluxation (DLS) view.

None of these radiographic examinations, and little palpation are possible without chemical restraint of the patient, if an accurate database is to be collected. Moreover, it makes little sense to expose yourself or your staff to ionizing radiation by handholding patients when we have such good drugs at our disposal nowadays.

Newer imaging techniques are being explored, including attempts to better reproduce the actual forces acting on the hip during weight bearing in a normal posture, along with ultrasonography, and CT evaluation.

Once all the data are collected, including the signalment, history, intended use of the dog, physical findings awake and under sedation, and radiographs, the tough job still lies ahead in formulating an appropriate management plan.


The major factors influencing treatment decisions are the clinical signs (degree of disability and/or discomfort) and the current status of the hips as assessed above. The question of prophylactic intervention (i.e., performing corrective surgery or administering medication) for patients with a diagnosis of hip dysplasia (based on palpation findings, radiographic findings, or both), when they lack clinical dysfunction is hard to answer. I generally do NOT treat hip dysplasia unless the patient or client is indicating a clinical need. The determination whether there are clinical signs is based on good history taking and good observation of the dog. Clients may demand different things from dogs (police work vs. sedentary pet), dogs may vary in their tolerance of mechanical dysfunction or arthritis, and clients may be unaware of the dysfunction in their pet, attributing dysfunction to "being an awkward puppy", or "I just figured he was getting old", rather than recognizing these signs as being due to hip dysplasia or hip arthritis. These all have to be considered as you formulate a treatment (or non-treatment) plan. Remember that for most stages of the hip dysplasia syndrome the signs develop gradually and only occasionally will animals be presented with acute lameness (and in such instances, you should "smell a rat", and look for another cause, such as cruciate ligament injury).

It is also important to keep in mind that hip dysplasia is a dynamic, progressive disease in each of its stages. When you do an evaluation, no matter how intensive, you are obtaining only a "snapshot" view of the hip status at that time. Particularly in the juvenile dog, things may be changing rapidly, and assessments/decisions made today may be invalid a month from today. Thus, it is critical, especially in the young dog, to monitor the trends, clinically, radiographically, and by palpation. You need to explain to Mrs. Jones that her dog has mild, asymptomatic dysplasia today, but it may worsen in a month's time. Even though moderately costly, the ideal management in such cases is to have the dog return for repeat exam and x-rays in 4-6 weeks so that the rate of progression can be monitored. Such repeat evaluations can be repeated as needed up to skeletal maturity to ensure that the dysplasia is not trending toward greater and greater subluxation. I would also counsel Mrs. Jones to return sooner if the dog starts evidencing signs of mechanical lameness. This is also a place where PennHip may be of some benefit. DI's determined as early as 4 months of age can be predictive of the chances for DJD later in life.

If a young dog has clinical signs, or if it is asymptomatic but clearly worsening in a rapid fashion (palpation and radiographs), I recommend intervention. There are no medications currently available that will address this mechanical failure and prevent the wear and tear sequelae. While chondroprotectants may be prescribed, and while some clients will report a modest improvement in clinical signs (Real? Placebo effect?), such treatment will neither halt progression nor prevent degenerative joint disease.

There has been some investigation of PSGAG (Adequan), given parenterally in Labrador puppies prone to hip dysplasia. The results were encouraging in that treated dogs had somewhat lower incidence and severity of dysplasia than cohort controls. These injections though were started at a very young age, and the long-term effects are not known. From a practical, clinical standpoint, these are treatments that few of us will be in a position to offer, and few clients will avail themselves of. There are few data regarding other so-called chondroprotectants to suggest that they will markedly alter the onset, progression or clinical disability of hip dysplasia. These agents are purported to provide the body "basic substrates" of articular cartilage (chondroitin sulfates, glucosamine, etc.), and the proponents would have us believe that these molecules are absorbed intact, transported to the "diseased" cartilage, and then incorporated into the cartilage to undo damage. It this true? We simply do not know, and the scientific evidence needed to support these claims is lacking. I myself am skeptical that any of these agents work this way. Might they be clinically effective for other reasons? Quite possibly, yes. Just as Adequan works as well (or better) when given intramuscularly as it does given intra-articularly, perhaps the mechanism of any positive action by the other "chondroprotectants' will be due to effects on various enzymatic pathways, such as metalloproteinases, that mediate some of the degradative changes in articular cartilage. However, the burden is on the manufacturers and distributors of these products to prove their efficacy, their safety, and their mechanism of action. Slick marketing, word of mouth and anecdotes are not a substitute for easily designed, double- blind, random, controlled, prospective clinical trials. Period.

Even if it is shown that "chondroprotectants" have some positive effect, will they be sufficient to protect the cartilage from all the mechanical trauma induced by abnormal translational movement? I rather doubt it. Even if they did, what about the changes in the soft tissues and subchondral bone? Are these unimportant bystanders? I believe a mechanical problem demands a mechanical solution, and this suggests surgery. In this same vein, I do not believe there is much of a role for traditional, non-steroidal anti-inflammatory drugs (NSAID's) in young dogs with the early (non-arthritic) stages of hip dysplasia, except where discomfort from soft tissue sprain is a factor. Most clients with these young, lame dogs will report few clinical signs of pain, and usually report minimal to no change in lameness with these medications.

For very young dogs (<16-20 weeks of age) that are diagnosed with dysplasia, or where dysplasia is a concern based on breed or family history, there is an option for intervention to try and prevent development of instability/incongruency. This is the juvenile pubic symphysiodesis (JPS) procedure initially described by Deuland and co-workers. JPS involves the use of electrocautery to destroy the germinal cells in the pubic symphysis growth plate. In turn this leads to premature closure of this physis, and results in the acetabulums growing more laterally - with more femoral head coverage - than they would naturally. Successful JPS may prevent development of CHD, and obviate the need for further treatment. Unfortunately, JPS does not seem to be of much help beyond 16-20 weeks, and so many dogs who might benefit from the operation are diagnosed too late to employ it.

Surgical options for young dogs (but older than 16-20 weeks) with mechanical lameness include corrective osteotomy (either pelvic or intratrochanteric), temporary symptomatic relief, or salvage. Corrective osteotomies are designed to improve the "fit" (congruency) and stability of the affected hip joint, thus, in effect, "undoing" the dysplasia. One can either move the ball toward the socket or the socket towards the ball, or both.

Intratrochanteric osteotomies (moves the ball) are done to decrease the angles of inclination and anteversion. Even though these are now understood to be normal in most dogs with CHD, the operation has been used with good clinical success for many years. It is technically demanding, and should only be undertaken by a surgeon who is well trained, appropriately equipped, and experienced in its use. It entails a wedge osteotomy of the proximal femur so that the head and neck can be rotated (retroverted) and titled (inclination). Stabilization is usually accomplished with a hook plate. An alternative method for "moving the ball" is an operation termed "femoral neck lengthening". This was touted several years ago but has largely fallen out of favor. The surgery entails longitudinally splitting the femoral diaphysis from the trochanteric fossa to just above the knee on the cranial and caudal cortices. "Spacers" are then placed to wedge the "bivalved" femur apart proximally, this displacing the femoral head and neck toward the acetabulum.

The most common corrective surgery for young dogs with hip dysplasia is pelvic osteotomy. In its various forms (double, triple, acetabuloplasty, etc.) the goal is to reposition the acetabulum and improve the coverage of and contact with the femoral head. Most surgeons today employ a triple pelvic osteotomy (TPO), which isolates the acetabular segment from the ilium, ischium and pubis. Different surgical approaches have been described, as have different locations and orientations for the osteotomies. There are also a variety of stabilization techniques, ranging from lag screw and cerclage wire to several types of bone plates. TPO appears to be helpful for two reasons. First, by in effect deepening the socket, it provides a larger surface area for weight bearing by the femoral head. This translates to improved congruency. Second, it causes some degree of bone and soft tissue remodeling, with the resultant effect of improved stability. It is thus the "ideal" operation for mechanically lame dogs with CHD, or those dogs that are trending to greater degrees of laxity based on serial examinations.

The criteria for successful TPO (success defined as elimination of subluxation, elimination of the Ortolani sign, prevention of secondary degenerative joint disease, and absence of clinical lameness) must be understood before the operation is recommended. First, the hip must be "reducible", that is, the Ortolani test should indicate that the femoral head will reseat into the acetabulum. The frog leg VD radiograph is also useful for making this assessment, since a hip that appears subluxated on the extended leg view should "seat" in this posture. This view also allows an assessment of congruency between the femoral head and the acetabulum. Remember that as CHD progresses the Ortolani sign may be lost as the acetabulum fills with redundant ligament and bone, and/or the femoral head becomes increasingly subluxated (or frankly luxated). If you cannot reduce the hip pre-operatively, TPO will usually fail to "capture" the hip and prevent arthritis later on. Second, even if the hip is reducible, one needs to consider how much rotation of the acetabulum will be required to prevent the hip from subluxating again post-op. Beyond about 30 degrees of rotation there is enough distortion of normal biomechanics to result in an abnormal postop gait. These dogs are not lame or sore, but they walk funny. Beyond 45 degrees this distorted gait may be profound. TPO's have been described ranging from 20 degrees to 90 degrees of rotation. They all may succeed in preventing osteoarthritis, but at some point the trade off between lameness due to hip dysplasia or pain or lameness (altered gait) due to TPO is not worthwhile. It is also known that the results of TPO are dynamic; i.e., some of the correction ("capture") and stability occurs weeks to months after the actual operation. This has taught us not to be overly aggressive in rotating the acetabulum at the time of surgery, but instead to rotate it "enough", and then depend on Mother Nature to complete the correction over time. How much is "enough?" This is controversial. Many determine this by calculating an "angle of luxation" and "angle of reduction", based on the Ortolani test, and using these values in some fashion (e.g., the average of the two) for the amount of acetabular rotation. While logical, there are no data in the literature that prove this is appropriate. Others use fixed-angle plates and many believe the angle of rotation should not exceed 20 or 25 degrees in any patient.

The next criterion is the absence of degenerative joint disease. In part this is determined by palpation, and looking for evidence of crepitus as the hip is flexed, extended, circumducted, abducted and adducted. Sometimes grinding will be detected associated with wearing away of the cartilaginous labrum along the dorsal acetabular rim, or the rim itself. If there are no other signs of DJD, and the hip is reducible as described above, such a dog is still a candidate for TPO, in my opinion. Radiographs are essential for determining whether there are osteophytes or other changes that herald the presence of secondary DJD. This, too, is a controversial area. Some surgeons believe that "a little" DJD is not a contraindication to TPO, feeling that (a) the operation will put healthy areas of cartilage into contact, and (b) the surgery will slow or prevent progression of arthritis. The long-term data do not support this contention particularly well, and the literature suggests that once the osteoarthritis ball is started rolling, it continues despite TPO. Remember too that palpation and radiographs may be limited and insensitive in detecting very early cartilage wear. Arthroscopy of the hip has been shown to be useful in revealing such lesions, and should be considered prior to actually making incisions to perform TPO.

The final criterion, which I consider relative, is the age of the patient. Most surgeons recommend operating before a year of age, and as early as possible (as young as 5 months). I agree with this. However, the main reason for early intervention is fulfillment of the other criteria outlined above. As these puppies age they are more likely to move from reducible, arthritis-free hips to severely subluxated, non-reducible, degenerate hips. In many dogs this transition occurs before a year of age, and in quite a few the hips become non-candidates for TPO by 6-7 months of age! However, there are occasionally dogs whose disease progresses more modestly, and who are otherwise still good candidates at a year-and-a-half or older. TPO may still be considered for these dogs. The chief difference between operating in a 6-month-old dog vs. an 18-month-old dog is the loss of substantial capacity for remodeling. Young (< 1 year) dogs have very plastic bones and soft tissues, so are better able to remodel these tissues in response to the new biomechanics dictated by the TPO. As dogs reach and then pass skeletal maturity, this plasticity is diminished, and so Mother Nature may be less capable of providing the continuing, post-op correction we typically see after TPO. For this reason, when I operate on older dogs, I tend to be more aggressive with my rotation (but still not to exceed 45 degrees) so that the hip is "captured" now.

Since CHD is usually a bilateral disease, most affected dogs will require bilateral treatment. However, it is important that each hip be evaluated independently, since one may be further along clinically or biologically than the other. Most dogs receiving TPO will have both hips done, but some may need no treatment for the second hip, while others' second hips may be poor candidates for TPO. While a few surgeons have described concurrent bilateral TPO's, most surgeons stage these operations. In my practice I typically do the second side as early as 2 weeks after the first (trying to operate before the CHD progresses), giving the patient just enough time to recover and get "sticky" at its osteotomy sites (remember, these are usually young dogs who heal quickly). Waiting 6 weeks between operations, as recommended by some surgeons, may allow a hip that was a good candidate to progress to being a less good candidate. The other disadvantage of widely staged operations is the stress placed on the client. These are young, active, large breed dogs that we are asking be kept crated and leash walked for 6-8 weeks after surgery. No easy task! With a 2-week interval between operations we at least keep the total time of confinement and limited activity down to 2 months, rather than stretched out to 3 or 4 months.

If a young dog is considered to be a suboptimal candidate for TPO (or other corrective procedure), what are the alternatives? First, medical therapy (to control any discomfort) can be instituted. Remember that some dogs will naturally go into a period of clinical remission once they are done growing. For those dogs that continue to have clinical disability or pain, other operations can be considered. Symptomatic surgery has been around for decades, and is based on performing pectineal myotomies or myectomies (tenotomy/tenectomy). The theory is that the tight pectineus is strained as the hip subluxates, and discomfort and lameness will be relieved if the muscle is released. Clinical experience bears this out to some extent, and some young dogs will be noticeably improved after this operation. The problem is that this procedure does nothing to stabilize the hip or to prevent progressive osteoarthritis. It is a temporary surgical band-aid, but further medical and surgical intervention will probably be needed several months down the road. In some instances such surgery can be used as a "bridge" operation to provide symptomatic relief until the dog is old enough for hip replacement. Another operation that can be used as a surgical "band-aid" or "bridge" is denervation of the joint capsule.

An operation which received a lot of press a few years back, but which has (fortunately) fallen out of favor, is the (Sertl) BOP shelf arthroplasty. This was an ill-conceived surgical procedure that was based on extending the dorsal acetabular rim laterally with a polymer material, in an effort to provide better "coverage" of the femoral head and reduce subluxation. Even in principle, this notion is preposterous since this extended shelf is neither intra-articular nor cartilaginous. It clearly does nothing to improve congruency or stability. Not surprisingly, then, results were poor, with young dogs progressing unabated to DJD, and dogs with established DJD having no measurable benefit. Worse still, this supposedly "biocompatible" polymer was anything but! Large numbers of dogs developed infections and draining tracts, and scientific studies demonstrated that not only did the BOP material not foster bony ingrowth, it actually retarded it!

It is reasonable to ask the question, "why not perform a TPO in suboptimal candidate, or do a "stopgap" procedure such as pectineal myectomy, and if it doesn't work, go back and do a salvage surgery?" My answer is that surgery of any kind involves some morbidity and risk for the patient, and is costly for the clients. Whereas in humans it is frequently the tactic to pursue multiple, successively complicated operations for a problem, it is less viable in animals, whose owners have more limited financial resources to pay for surgery, and who emotionally are often ill-prepared to deal with repeated attempts to resolve a problem. For these reasons I believe it is incumbent on us to recommend only that surgery, if any, which is likely (i.e., has a 90 percent or greater prognosis for good or excellent results) to permanently resolve the problem, without the need to go back and do more. In properly selected TPO candidates the data support this 90 percent favorable prognosis. However, dogs with hips that don't reduce nicely, or that already have some DJD, may drop down to a 50 percent prognosis. Is this good enough? For some clients the answer is "yes", and they accept that they may have to come back and do more in the future. For others, the answer is clearly "no", and we are obligated to apprise them of our reservations about a particular procedure.

For those dogs, young or old, that have hip dysplasia that is inadequately managed by medication, and for whom any of the operations discussed above is expected to or has failed, salvage surgery should be discussed. By definition, salvage surgery is an operation employed for the end-stage of disease, or when no other alternatives are acceptable. The radiographic "end-stage" for hip dysplasia is advanced osteoarthritis with deformation and destruction of the normal joint anatomy, complete loss of articular cartilage, eburnation of bone, severe osteophytosis, and all the other changes of DJD. However, we do not treat radiographs! The decision to employ a salvage surgery, then, as with all other treatments for CHD, is predicated on the clinical signs of disability and discomfort. If a dog has hip dysplasia or arthritis, and is functioning at an acceptable level (keeping in mind that some owners may have a skewed opinion as to what is normal or acceptable), it does not require treatment. If the dog's signs are adequately managed with non-surgical therapies (medication, weight control, acupuncture, etc.), it does not need salvage surgery. If the dog has severe hip dysplasia (at any stage of the disease), and develops an acutely painful lameness, look for other explanations (cruciate ligament injury, spinal disease, etc.) before recommending salvage surgery for the hip(s). The hallmark of CHD is gradual loss of function and moderate, albeit chronic, discomfort. Conversely, you should be sensitive to hearing from your client that "the dog is fine; it's just 7 years old now and is slowing down from old age". Horse feathers! How many 7-year-old dogs do you know, absent hip dysplasia or some other disabling illness, that aren't active, playful and athletic? If you hear this or observe this, perhaps that animal should be evaluated as a potential sufferer from CHD, and possibly be offered a salvage procedure.

There are three options for salvage: amputation, femoral head and neck excision arthroplasty (FHNEA, FHO), and total hip arthroplasty/replacement (THR, THA) using a prosthesis (artificial hip). Amputation is virtually never indicated as treatment for hip dysplasia or arthritis. Femoral head and neck excision arthroplasty (femoral head ostectomy) is a viable choice for many dogs. By removing the ball, we hopefully relieve whatever discomfort is associated with contact between the femoral head and the socket. There are advantages and disadvantages associated with FHNEA:

  1. Relatively "simple" operation
  2. Relatively low cost operation
  3. Few potential complications
  4. No need for prolonged or difficult post-op care
  5. Usually successful at relieving discomfort associated with CHD
  6. Excellent results in dogs < 30-40 lbs. body weight (the smaller, the better)
  1. Loss of normal biomechanical function (no articulating joint)
  2. Decreased range of motion as pseudoarthrosis forms from scar tissue
  3. Continued muscle atrophy signals persistent decreased load bearing by pelvic limb(s)
  4. May not relieve all discomfort (grinding bone is only one source of pain)
  5. Poorer results (persistent discomfort/persistent disability) in larger, active dogs
  6. Slow recovery of weight bearing function
  7. Higher likelihood of needing bilateral surgery than THR (making costs more comparable)
If you do FHO's, a few technical points to keep in mind: (1) Be sure to remove the entire femoral neck. Because of natural anteversion, the neck is longer on the caudal aspect than on the cranial aspect. Take this into account as you angle your cutting instrument. (2) Be aware of the location of the sciatic nerve caudal and dorsal to the hip and take care not to injure it. (3) Leave a smooth surface that does not rub or catch at the operating table. (4) Close the soft tissues (joint capsule, plus or minus transposition of the deep gluteal tendon) between the femur and the acetabulum. There is no need for fancy muscle "slings" as have been described in the literature if you create a proper cut.

The alternative to excision arthroplasty is total hip replacement (THR). This procedure has been widely used in clinical practice since the mid-1970's in many thousands of dogs. It has proven itself to be among the most reliable and successful of operations of any kind, with most experienced surgeons returning 90-96 percent of dogs to good or excellent function. Complications are possible (about 10 percent overall rate) but usually can be prevented or successfully resolved if they occur. Refinements in technique, prostheses and instrumentation are continuously being made and investigated. As with FHO, THR has advantages and disadvantages:

  1. Eliminates pain (as with FHO) and provides sound mechanical joint
  2. Restoration of normal biomechanics allows recovery of soft tissues, too
    1. improved muscle mass
    2. improved range of motion
    3. return to full athletic ability
  3. More successful results in large, active dogs when compared with results after FHO
  4. Statistically, 75-80 percent of dogs do well with a single THR, despite bilateral disease
  5. Usually lasts the lifetime of the patient, regardless of age at implantation (barring complications)
  6. Usually immediate weight bearing post-op (disadvantage?)
  1. Complex, costly operation
  2. Cannot be done until skeletal maturity (approximately 9-12 months of age); some young dogs therefore have to wait a few months
  3. Potential complications
    1. implant loosening (infection or aseptic)
    2. dislocation
    3. neurapraxia
  4. Difficult to revise a previous FHO into a THR
  5. Requirement for prolonged (2-3 months) relative inactivity after surgery
  6. If cemented, lifelong risk for hematogenous infection (use prophylactic antibiotic during elective procedures, such as dentals, where bacteremia is likely, and therapeutic antibiotic for skin, urinary tract, dental or other infections that rise)
THR is a wonderful operation when used appropriately. Of all the surgeries I do, the only one that approaches the level of immediate gratification and client satisfaction I get with THR is cervical disc surgery. In both cases, dogs what come in to the hospital crippled and painful go home happy and functional. While it is expensive, when I compare it with other similarly priced veterinary services (radiation therapy, prolonged chemotherapy, thoracolumbar laminectomy for disc displacement), it far and away provides the patient and client with a much better prognosis and longer, better quality of life. In dogs, hip prostheses do not generally "wear out" as occurs in humans, so we do NOT have to hesitate placing a prosthetic hip in a young patient. There are many dogs that have had THR surgery at a year of age ( or slightly younger) that have been followed for the balance of their lives (in some cases, 14-15 years!), and their hips hold up fine. When we do see complications such as loosening, they are as likely to occur a few months postop as they are many years later.

While THR is the preferred salvage surgery for large dogs with hip dysplasia, it does not make FHNEA/FHO "bad" surgery in these same patients. It is only after consideration of the needs of the patient, the financial and emotional capabilities of the client, and the confounding medical or management factors (e.g., does the dog have other systemic, orthopedic or neurologic diseases that might limit the success of THR, or increase the risk for complications? Can the dog be adequately confined/restrained post-op?) that the choice between THR and FHNEA/FHO can be made. Though I might prefer to offer a THR for a patient, if it would still be better off with a femoral head and neck excision than it currently is with its underlying hip dysplasia or arthritis, I would happily do the FHNEA/FHO without regard to the patient's size or age. In my experience, the younger the dog is at the time of FHNE/FHO, the faster and better they adapt and recover. The major caution here is to explain to the client that, for technical reasons, it is usually not feasible to go back and convert from a FHNEA/FHO to a THR, so they need to consider this choice irrevocable. Conversely, if there are major, unresolvable complications with a THR, it can always be converted to a modified FHNEA, albeit a very expensive one! In those (hopefully) rare instances where a TPO has failed to prevent painful, disabling arthritis from developing, THR can be used to salvage the situation. Because the TPO procedure does not involve arthrotomy (in most surgeons' hands), there is no increased risk of infection as there is when THR is used on hips that have previously been opened. Interestingly, pelvic osteotomy has been used, on occasion, to salvage a THR that is plagued by recurrent luxations or acetabular loosenings! If there is insufficient dorsal coverage of the acetabular component it is at risk for loosening. If not properly positioned within the acetabular bed, the head can luxate. Isolating and rotating the acetabular segment containing the prosthetic socket via TPO can rectify both of these situations.


Because of the sophistication and options available nowadays (TPO's, total hip replacement's, etc), it is appropriate to refer dogs with hip dysplasia to a board-certified surgeon in your area for evaluation and treatment. By reviewing the concepts of this lecture, you should be more adept at determining the timing of those referrals, and to better prepare the client for what may be needed. This improved understanding will also save you the potential embarrassment of referring a patient for a specific procedure (e.g., TPO), only to have the specialist tell your client that the procedure is not appropriate for the current stage of the disease. Especially in this day and age of the internet, I see an increasing number of clients who come armed with reams of information about one operation or the other, fully expecting me to simply confirm their understanding and proceed with surgery. How frustrating for them when I have to tell them that all their research was in vain, and that their dog really needs something different! The ultimate frustrating case, which I see routinely, is the dog referred for a hip replacement, accompanied by radiographs that confirm advanced arthritis, but who is found on questioning the client and examination to have some other cause for its current signs (typically a cruciate ligament injury or a myelopathy). Don't let this happen to you. With care, and least at least a little hedging ("Gee, Mr. Smith, your dog has terrible hip dysplasia and arthritis which may be causing its lameness, and if so, he may need a hip replacement, but there may be something else. Let's see what the surgeon says before we make any decisions" ) you will turn the referral experience into a good one for your client, and you will rise in the client's esteem for being knowledgeable enough and caring enough to seek proper assistance at the right time.

© 2006 - James M. Fingeroth, DVM, DACVS - All rights reserved