Abstract
We describe a novel surgical technique used to correct feline patellar luxation (PL) where abnormal patellar tracking persists despite conventional corrective surgery. An anatomical difference between feline and canine stifles is that the feline patella is wider relative to the trochlear sulcus. This results in less constrained patellar tracking. Therefore, patellar subluxation is common in normal cats. It was noticed that in some feline cases with clinically significant PL, PL persisted intraoperatively despite performing the standard corrective procedures. We report a novel surgical technique – partial parasagittal patellectomy – to address the wide shape of the feline patella relative to the sulcus. This technique has been successfully performed in four cats with good outcomes. However, the immediate risks and long-term effects of partial parasaggital patellectomy are not known. We reserve this technique for surgical cases where PL cannot be controlled by conventional means.
Patellar luxation (PL), its surgical management, outcome and complications are well described in the dog.1–5 However, there is little published about PL in the cat.6–8 Knowledge of feline orthopaedic disease and treatment are frequently assumed to be equivalent to that required for small dogs. Management of feline orthopedic disease is frequently adapted from techniques for the canine patient without consideration of anatomical features unique to the cat. 9 One feature of the feline stifle is that the patella is wider in a medio-lateral direction relative to the femoral trochlear sulcus, and flat in the cranio-caudal plane. 9 Unlike dogs, many clinically normal cats have mediolateral laxity of the patella with respect to the sulcus. It is recognised that patellar subluxation is a common finding in the normal cat.7–9
Cats with patellar subluxation or low-grade PL may have no clinical signs. A cross-sectional prevalence study found that 45/78 cats (58%) had PL and 78% of the cats with PL had a grade 1 luxation, but only 11 cats (24%) had pelvic limb lameness.7,10 Medial patellar subluxation was seen in 31/33 cats (94%) with otherwise normal stifle joints and no lameness. 7 Although patellar subluxation is a common finding in cats, PL is an infrequent clinical problem, but, when present, it can cause pelvic limb dysfunction presenting as pelvic limb lameness, intermittent locking of the stifle or a crouched stance. 6 Surgical and non-surgical management of feline PL can result in excellent clinical outcomes.8,11 Loughin et al 8 recommend surgical management for cats with persistent lameness refractory to conservative management.
Surgical techniques commonly employed for correction of PL in dogs include femoral trochleoplasty, tibial tuberosity transposition (TTT), soft tissue imbrication and/or release, and femoral corrective osteotomies.1,4,12 Combinations of these techniques are performed on a single stifle. The same techniques are utilised in cats, but the literature does not address how to apply them to cats nor whether their application should be different from dogs. 8
We have noticed that, compared to dogs, applying these techniques does not always achieve intraoperative patellar stability in cats. Our intraoperative assessment in such cases has been that the wide and flat shape of the feline patella relative to the femoral trochlear sulcus hinders stable tracking of the patella. To address this and prevent persistent PL, we have developed the novel surgical technique of partial parasagittal patellectomy. The purpose of this report is to describe the surgical technique and to report a series of cases.
Case descriptions
Four cats had partial patellectomy for persistent PL (Table 1). The same lead surgeon (GA) performed all the surgeries in the same procedural order as described in Table 1. Pre- and postoperative radiographs were taken for all cases (Figure 1a,b, Figure 2a, b). All of the cats were hospitalised for 2–4 days postoperatively. Peri- and postoperative analgesia consisted of injectable opiates (methadone [Physeptone; Glaxosmithkline] or buprenorphine [Vetergesic; Alstoe Animal Health]) and oral meloxicam (Metacam; Boehringer Ingelheim). Gentle physiotherapy and cold compresses were applied several times daily to those cats that would allow it. The cats were discharged with instructions that included exercise restriction in the form of cage rest for 6 weeks, with 5–10 minutes of supervised gentle, room-confined exercise and passive range-of-motion physiotherapy.
Case details of cats that had surgical correction of patellar luxation (PL), including partial parasagittal patellectomy
DSH = domestic shorthair; M(N) = male (neuter); F(N) = female (neuter); y = year; mo = month; L = left; R = right; W = wedge sulcoplasty; T = tibial tuberosity transposition; Rl = release; PP = partial patellectomy; B = block sulcoplasty; I = imbrication; A = abrasion sulcoplasty; MPL = medial patellar luxation; LPL = lateral patellar luxation
Numbers refer to the grade of PL (0–4)
Preoperative craniocaudal (a) and mediolateral (b) radiographs of case 4 demonstrating a grade 3 medial patellar luxation and relatively wide patella compared with the trochlear width
Postoperative craniocaudal (a) and mediolateral (b) radiographs of case 4. The medial and lateral abaxial portions of the patella have been removed. The patella is centrally located within the deepened trochlear sulcus
Case 1
Case 1 was a 1-year-old Maine Coon with grade 2 left medial patellar luxation (MPL). A lateral parasagittal patellar arthrotomy was made. Femoral wedge trochleoplasty was performed after arthrotomy and intra-articular assessment. Quadriceps alignment was assessed and corrected by lateral TTT. Medial and lateral soft tissue tensions were assessed and were normal; therefore, lateral imbrication and medial release were not performed. Prior to arthrotomy closure, patellar stability was assessed through a normal physiological range of movement, including stifle flexion/extension and internal/external rotation.
Despite the corrective surgery, PL persisted. Further detailed inspection showed that the patella was not engaging the recessed femoral trochlear sulcus correctly because the patella was wider than the trochlear sulcus; therefore, it was riding on, and contacting only, the proud medial and lateral trochlear ridges. In other words, although trochleoplasty successfully resulted in recession of the sulcus, this did not influence patellar position; instead, it served only to reduce contact area between the patella and the femoral sulcus to just the trochlear ridges. Consequently, the patella was not constrained by the trochlea, normal tracking did not occur and PL persisted.
As a result of this and in order to match the mediolateral width of the patella to the dimensions of the trochlear sulcus, partial parasagittal patellectomy was performed. The stifle was extended and Adson forceps were applied to the cut surface of the retinaculum and the patella was retroflexed (Figure 3). The width of the recessed femoral trochlear sulcus and the patella were measured and the difference calculated to determine the reduction in patellar width that would be necessary to achieve constrained patellar tracking within the trochlear sulcus. Lateral partial parasagittal patellar ostectomy was performed using an oscillating saw (Colibri; Synthes) with extreme care to avoid damage to the overlying patellar ligament. A Freer elevator and minimal sharp dissection were then used to free and remove the ostectomised segment. Following completion of the partial patellectomy, the patella was returned to the recessed trochlear sulcus. Improved seating and correct tracking of the patella in the trochlear sulcus was confirmed on stifle manipulation, and patellar luxation was no longer possible. The arthrotomy and surgical site were closed routinely.
Intraoperative view of retroflexed patella (case 1) showing the caudal articular patellar surface when retroflexed and the width discrepancy between the wide, flat feline patella and narrow femoral trochlear sulcus. The top of the photograph is proximal. The two black lines show the approximate ostectomy positions for medial and lateral partial parasagittal patellectomy
The cat recovered well, but 5 days postoperatively it developed an erythematous, painful wound with purulent discharge. The signs were consistent with surgical site infection, but bacteriology was not performed. Seven days of antibiotic therapy (amoxicillin-clavulanic acid) were administered; the signs resolved and did not recur. At 6 week postoperative re-examination, the cat had marked left pelvic limb muscle atrophy and was mildly lame. Stifle manipulation was well tolerated and pain free through a full range of movement. Normal patellar tracking was confirmed and luxation was not possible. Radiographs showed progressive union of the tibial tuberosity and no evidence of implant-related or other problems. There was no evidence of cortical remodelling of the patella, nor of osteophytes associated with the ostectomised lateral edge. Four months postoperatively, the cat was not lame and there was mild muscle atrophy. The patella tracked normally and did not luxate. There was no pain on stifle manipulation. Further radiographs were offered, but declined by the owner. The referring veterinary surgeon re-examined the cat on several occasions up to 48 months postoperatively; no problems or lameness were noted, and the owner reported excellent pelvic limb function.
Cases 2 to 4 were managed in the same way and in the same sequence as case 1, except for the differences that are described below.
Case 2
Case 2 was a 2-year-old domestic shorthair (DSH) cat that initially had bilateral grade 3 MPL. Left stifle patellar luxation was not addressed surgically. The right stifle had lateral imbrication and medial release performed by the referring veterinary surgeon 11 months prior to presentation, but medial PL had persisted. Surgery consisting of block trochleoplasty, TTT and medial release was performed by colleagues at our institution. The cat was presented to us for 6 week postoperative check, and the complication of right stifle grade 3 lateral patellar luxation was diagnosed. This was addressed surgically as described for case 1, except that medial parapatellar arthrotomy and abrasion sulcoplasty were performed. This did not achieve constrained patellar tracking and stability within the trochlear sulcus for reasons identical to case 1; therefore, partial parasagittal patellectomy was performed.
Initially, only lateral partial patellectomy was performed, but the patella remained wider than the recessed trochlear sulcus and, as result, the patella did not seat correctly; the lateral edge recessed, but the medial edge was tipped up on the medial trochlear ridge. Therefore, medial partial parasagittal patellectomy was also performed, which resulted in correct patellar seating within the recessed femoral sulcus. Repeat intraoperative stifle assessment showed normal patellar tracking without luxation and the stifle and surgical site were closed routinely. TTT was not performed, as it had been performed previously, and quadriceps alignment was good.
Six weeks postoperatively, the cat had mild muscle atrophy, was mildly lame and mildly uncomfortable on stifle manipulation. The patella was tracking normally within the trochlear sulcus and could not be luxated. The owner declined radiographs. Repeat examination 26 months postsurgery showed that the cat was not lame and had no appreciable muscle atrophy. The stifle was mildly thickened and had normal pain-free range of movement with no crepitus. The patella was tracking normally and could not be luxated. The owner reported that the cat was sound except that every few months, mild weightbearing lameness was noted that lasted a short period of time and resolved without medication or restricted exercise. The owner was satisfied with the outcome and declined check radiographs.
Case 3
Case 3 was a 6-year-old DSH with acute-onset lameness and grade 2 left stifle lateral PL of 4 months’ duration that had not improved with conservative management of rest and non-steroidal anti-inflammatories (NSAIDs). This was addressed in a similar way to case 2, including medial parapatellar arthrotomy and block trochleoplasty. This did not achieve constrained patellar tracking and stability within the trochlear sulcus for reasons identical to cases 1 and 2; therefore, medial and lateral partial parasagittal patellectomy was performed. TTT was not performed as quadriceps alignment was normal. Intraoperative assessment prior to soft tissue closure showed normal patellar tracking and no luxation. Six week re-examination showed mild lameness (3/10) and no pain on stifle manipulation. The patella was normal on palpation, tracking normally within the femoral trochlear sulcus and could not be luxated. Radiographs were offered, but declined by the owner. Further follow-up examination at our institution was declined, but the cat was re-examined by the referring veterinary surgeon who reported that the cat was not lame and did not resent stifle palpation.
Eighteen months postoperatively, the owner reported that the cat was exercising normally with mild intermittent lameness.
Case 4
Case 4 was a 3-year-old DSH with acute-onset lameness and grade 3 left stifle MPL that had occurred after a dog attack. The cat was stabilised and life-threatening injuries, including pneumothorax, were treated. Despite rest and NSAIDs for 4 weeks, left pelvic limb lameness and PL persisted. The case was managed surgically, as described above, including lateral parapatellar arthrotomy, block trochleoplasty, medial release and lateral imbrication. This did not achieve constrained patellar tracking and stability within the trochlear sulcus for reasons identical to cases 1 to 3; therefore, medial and lateral partial parasagittal patellectomy was performed. TTT was not performed, as quadriceps alignment was normal. Intra operative assessment prior to soft tissue closure confirmed normal patellar tracking and no luxation. Examination 16 weeks postoperatively by the referring veterinary surgeon showed the stifle was not painful on manipulation, the patella was tracking normally within the trochlear sulcus, the cat was not lame and was able to jump.
At the 15 month postoperative examination, performed by us, the cat was not lame. Muscle atrophy was not apparent. There was mild stifle thickening, normal range of movement, and no crepitus or pain. The patella was palpably normal in the trochlear sulcus, was tracking normally and could not be luxated. The owners reported mild intermittent weightbearing lameness of a few hours’ duration that occurred every few months and resolved without intervention. Radiographs were declined.
Discussion
This case series describes partial parasagittal patellectomy as a novel surgical technique that has been successfully employed to address PL in cats. It describes how to perform this surgical procedure for which cases it may be appropriate, and how it can be used successfully as an adjunctive surgical technique.
The patella of cats is comparatively wider than that of dogs, and is relatively wider than the feline femoral trochlear sulcus. 9 The normal feline patella is flat in a cranial to caudal plane and wide in a medial to lateral plane, and the patella is relatively wide compared with the femoral trochlear sulcus, but, to our knowledge, the relevance of this conformation to normal patellar tracking in cats and the occurrence of patellar subluxation or luxation has not been previously investigated. The common feature in the four cats reported here was that conventional PL surgical techniques were performed, but these did not result in correction of patellar tracking or luxation. Specifically for each case, and despite trochlear recession, the problem was identical, that is, intraoperative assessment of the stifle showed that the patella was wider than the sulcus and, as a result, was riding on just the medial and trochlear ridges, that is, trochlear recession did not result in patellar recession. In other words, the effect of trochlear sulcoplasty without partial patellecomy was detrimental in all four cases as patello-femoral contact area was reduced to a minimum at just the medial and lateral trochlear ridges. To correct this mismatch, abaxial parasagittal partial patellectomy was performed by removing the medial and/or lateral edges of the patella. The width to be removed was determined by measuring the difference between patellar and trochlear sulcus widths. In the first case (case 1), only lateral parasagittal partial patellectomy was performed, but in the subsequent three cases, medial and lateral partial parasagittal patellectomies were performed to minimise the total volume of bone ostectomised (as discussed below), and to avoid abnormal patellar position and orientation within the sulcus. Patellar tipping within the recessed sulcus can occur if only one edge of the patella is ostectomised (as in case 2).
An oscillating saw was used for increased accuracy and control to resect a narrow 2–3 mm sliver of abaxial patellar bone with less likelihood of slippage than using a small hand saw. The use of an oscillating saw was a reliable and safe technique in these cases; significant observable iatrogenic damage did not occur. Nonetheless, theoretical risk factors are damage to the patellar ligament or mechanical weakening of the patella that could precipitate patellar fracture. It is paramount that the surgeon avoids iatrogenic damage to the patellar ligament. The safe limit for partial patellar ostectomy is unknown; cadaveric and biomechanical studies would be useful to better understand the associated risks and produce guidelines. In the absence of such studies, it seems prudent to remove the minimum amount of patella necessary for adequate patellar recession. All the cases reported had a good to excellent clinical outcome with no major complications.
The partial parasagittal patellectomies were performed either laterally (1 case) or medially and laterally (3 cases). The intended width of the patella was determined from the width of the sulcus. Typically, this required removal of patellar bone (2–3 mm width) medially and laterally. We hypothesised that by removing the most medial and lateral borders of the patella, this would achieve width reduction, while minimising the likelihood of patellar ligament injury. The shape of the feline patella suggests that the lowest volume of bone per unit length is at the abaxial edges, but this is an assumption because, to our knowledge, there are no published studies that document this. If this assumption is true, performing medial and lateral abaxial patellectomies as we have reported would achieve the necessary reduction in width, while removing the least bone volume.
There is a precedent for performing partial patellectomy: partial edge patellectomy and proximal hemi-patellectomy have been described as treatments for patellar fractures in cats.13,14 In a survey of 52 patellar fractures in cats, six were treated with partial patellectomy and one with complete patellectomy. The outcome was good with mild lameness for the cases of partial patellectomy, but the outcome for the complete patellectomy case was not clear. 15 Complete patellectomy is reported as a salvage procedure because persistent lameness, marked quadriceps muscle atrophy and reduced pelvic limb function occurs. 16 Partial patellectomy of the distal pole has been described for the treatment of patellar fracture and for patellar osteochondrosis.17,18 Arguably, the biomechanical effect of distal pole patellectomy would be much greater than abaxial parasagittal patellectomy as the former disrupts the patellar-to-patellar ligament interface, whereas abaxial patellectomy does not. However, the effect of force re-distribution and abnormal stresses after abaxial patellectomy is unknown.
Three different femoral trochleoplasty techniques (abrasion, wedge and block recession) were used for the four cases we report. The choice of technique was at the surgeon’s discretion and decisions were made according to previous experience, surface area of trochlear hyaline cartilage that could be salvaged (case 2) and perceived risk of complications, including femoral sulcus, osteochondral graft or trochlear ridge fracture. 12 Regardless, the common feature for all cases, irrespective of trochleoplasty technique, was that adequate trochleoplasty did not result in corresponding recession of the patella. Indeed, the effect of trochleoplasty alone was actually detrimental in that it resulted in reduced contact between the patella and trochlear sulcus. This occurred because although the sulcus was recessed, the patellar position was unchanged as the patella was wider and over-sized with respect to the trochlea and therefore could not recess into it as it was ‘suspended’ on the medial and lateral trochlear ridges. Trochleoplasty techniques had no impact on this.
It is important to note that the partial patellectomy technique was used in only a minority of our feline surgical PL cases. It might be argued that performing standard techniques such as trochlear sulcoplasty and TTT more aggressively could have resulted in better patellar tracking without the need for partial patellectomy. However, there is a limit to what each of the standard surgical techniques can achieve, and there is no absolute formula as to the correct combination of techniques. Failing to achieve this delicate balance can result in further complications, for example, converting medial patellar luxation to lateral patellar luxation. The decision as to which techniques to perform and how aggressively is subjective, and has to be made by the surgeon intraoperatively.
One limitation of this case series is the lack of long-term follow-up radiographs, and the loss of three-dimensional data associated with radiography. Computed tomography imaging would have been advantageous to assess limb alignment and patellar seating, but would have incurred additional costs for the owner and was not readily available for all cases. In all but one of the cases, follow-up radiographs were declined by the owners as the cats were doing very well, and they wished to avoid the expense and potential risk associated with sedation. We are therefore unable to document the radiographic effects of partial patellectomy, in particular with respect to the potential development of degenerative joint disease such as osteophytes associated with the patella or patellar fracture. It is likely that degenerative joint disease developed as a result of the stifle surgery, but this could be as a result of any part of the stifle surgery, including the arthrotomy, trochleoplasty or partial patellectomy. Regardless, the effect of this is arguably minimal because cats were documented to be doing clinically well at re-examination between 15 and 48 months post-operatively with no or mild intermittent lameness, and no palpable stifle dysfunction or pain. For the same reasons, it seems very unlikely that postoperative patellar fracture occurred, particularly as studies of cats with patellar fractures have documented moderate to severe lameness, and this was not a feature for any of our cases.15,19
Conclusions
We report a novel technique of partial parasagittal patellectomy for augmentation of surgical correction of PL in cats. The procedure has been successfully performed in four cats with good outcomes and without complications requiring revision surgery. However, the technique should be utilised cautiously as the risks and long-term effects of partial patellectomy are not fully understood. Nonetheless, the technique can be used successfully in cases that are judged to be refractory to conventional surgical correction for PL.
Footnotes
Funding
The authors received no specific grant from any funding agency in the public, commercial or not-for-profit sectors for the preparation of this case series.
Conflict of interest
The authors do not have any potential conflicts of interest to declare.