The results of our study support our first hypothesis that a systematic decrease in patellar height occurs following TPLO and AMA-based CCWO even when these procedures are performed by surgeons who are experienced in the relevant techniques. For both procedures, postoperative patellar height was found to be distributed into either the normal height class or the low height class with no TPB or PPB, but the two procedures had different effects on the patellar ligament at the two-month follow-up, with patellar height decrease and PLT occurring regardless of the osteotomy position in the TPLO group.
Assessment of decreases in patellar height following TPS alteration procedures depends on which index is most appropriate for quantifying a change . During CCWO, the tibial crest moves distally, but the patellar ligament length (PLL) and the patellar length (PL) remain unchanged, resulting in similar pre- and postoperative ISIs; however, the distance between the distal pole of the patella and the tibial plateau (TP) decreases, and the GVI has been shown to be reliable and accurate for calculating changes in patellar height after this procedure and for predicting subsequent PPB [16, 21]. Following TPLO, the relationship between the patella and the TP is unchanged; thus, the GVI remains within the normal range, whereas the ISI, which reflects PLL, has been shown in previous studies to be a suitable index for assessing TPB [14, 16].
The aim of this study was not to statistically compare the magnitude of patellar height between groups because the indices used in the two groups were different but to quantify the changes in patellar height in the two groups to provide objective data that can be used by TPLO surgeons and evidence that a decrease in the height of the patella also occurs following TPLO.
PPB and TPB are well-known undesirable results following high tibial osteotomy (HTO) in human patients, and they have been correlated with poor functional outcome [14, 15]. It has been stated that in dogs that undergo CCWO, the decreased height of the patella may lead to PB [11, 13], and this is thought to be avoided with TPLO [11, 16].
PB could be a result of tibial shortening following wedge reduction with standard CCWO and has been reported to be a critical disadvantage of CCWO compared to TPLO [11–13]. Such shortening could increase the strain on the patellar ligament and increase the likelihood of secondary inflammation, but these are theoretical suggestions that have not been proven or documented in practice [6, 7, 9, 10]. In our study, precise planning of the osteotomy allowed us to superimpose the AA and MA axes postoperatively, thus reducing the median TLAS to 0°; this resulted in 82.5% of the dogs having strictly aligned axes with consecutive, predictable postoperative TPAs. Furthermore, the reduced angle of correction compared to previous recommendations for standard CCWO [4, 9, 10, 12, 23] and the positioning of the tip of the wedge at the caudodistal insertion of the medial collateral ligament rather than at the caudal cortex of the tibia, as described in previous studies [6–9], reduced the size of the wedge and limited lowering of the patella. As a result of the reduced size and precise positioning of the wedge, the median tibial shortening, even in this cohort of large dogs with large TPAs, was 3.5 mm, comparable to the mean difference in tibial length of 2.5–3 mm for CCWO documented in previous publications [6–9, 16].
In the present study, neither PLT nor patellar tendonitis (PLT=0) was recorded at the two-month follow-up after CCWO, confirming what has been reported in previous studies concerning stifles subjected to AMA-based CCWO [7, 16]. In contrast, PLT and patellar tendonitis have been reported to occur in 80–100% of patients after TPLO [19, 20], although the position of the tibial crest remains unchanged. In our study, only 40% of the dogs in the TPLO group had PLT grade 0–2 at the two-month follow-up, with 40% having a grade higher than 4.
It has previously been suggested that PLT following TPLO procedures could be a consequence of arthrotomy and abnormal stresses on the patellar ligament following rotation of the proximal tibial segment [19, 20] and that a TPA of less than 6° may contribute to increased PLT, but this was not recorded in our case series; regardless of group, dogs with TPA >30° had almost the same amount of rotation, the same TPA postoperatively and the same mini-arthrotomy, but they had different PLT grades at 2 months postoperatively .
Kowaleski et al.  postulated that in the TPLO procedure, the osteotomy must be on the long axis of the tibia and centred on the point that divides the intercondylar tubercles (Position E) to avoid TLAS and a subsequent patellar height decrease. The present study shows that regardless of the location in which the radial osteotomy was performed, even if it was in the correct position, a similar decrease in patellar height occurred. The patellar height decrease following TPLO is thus a consequence of the radial osteotomy followed by caudal rotation of the proximal segment of the tibia and not a result of misplacement of the osteotomy. However, the decrease in patellar height is exacerbated by osteotomy in the caudodistal position. In this study, we found that dogs with grades >4 had greater magnitudes of patellar height decrease and TLAS.
The consequence of a decrease in postoperative ISI with TPLO is a decrease in the distance between the distal pole of the patella and the tibial crest. Thus, postoperatively, the patellar ligament and collateral ligaments are under less tension, as reported in a recent study , and with active quadriceps muscular contraction, fibre tearing in the patellar ligament and its paratendon could occur due to abrupt loading; this could explain the higher frequency of PLT at the level of the tibial crest observed in this study, and it confirms what has been reported previously [19, 20].
Our study also confirmed our second hypothesis, which proposes that the craniocaudal position of the patella following TPLO is different from that in a normal stifle and that the patella shifts from caudal to the AA to cranial to the AA with a reduced AMA angle with AMA-based CCWO, a position more closely related to the position found in the control animals.
As postulated by Mazdarini et al. , despite the fact that an increased magnitude of the AMA angle has been associated with an increased risk of CCLR based on comparisons of dogs at low risk of developing CCL disease with dogs predisposed to CCLR and on comparisons of predisposed dogs with and without CCLR [21, 22], there are insufficient data to indicate that decreasing the AMA angle by itself is an aid to joint stability . However, a similar approach to AMA-based CCWO has been described recently. CORA-based levelling osteotomy (CBLO) achieves a similar rotation of the entire proximal tibial metaphysis relative to the AA, reducing the TPA by the same ratio as does AMA-based CCWO (65% of the TPA), which is less than the desired rotation needed to obtain joint stability with the TPLO procedure ; this induces alignment of the proximal and distal anatomic axes of the tibia and thus reduces the AMA angle. Good to excellent outcomes after CBLO, even in dogs with TPA >30°, have been recorded in several recent studies [27–30]. Accordingly, it could be argued that although the desired planned postoperative TPA is higher with CBLO and AMA-based CCWO than with TPLO, alignment of these two axes during the change in TPA limits the secondary translation of the tibia and the caudal tibial thrust during weight bearing and thereby contributes to the stability of the stifle joint [7, 16, 22, 27, 28, 30].
The AMA angle should be taken into consideration when a TPS alteration procedure must be chosen, especially since the AMA angle and TPA are strongly correlated with TPA >30°, with almost 80% of dogs having AMA angles >3° [21, 22].
In this subset of dogs, following TPLO, the malalignment between the AA and the MA persists postoperatively, and the AMA angle remains greater than 3°, resulting in increased caudal displacement of the weight-bearing axis that has been recognized to cause a focal increase in joint forces at the caudal aspect of the tibial plateau, with resulting loss of compliance of supporting structures such as the joint capsule, leading to cartilage erosion [27, 31, 32].
Furthermore, as reported by Kim et al. , the clinical consequences of the persistence of these abnormal joint contact mechanics could explain the progression of osteoarthritis that is frequently observed in stifles treated with TPLO, particularly those with a greater magnitude of TPA [31, 33–35].
In contrast, as described for the CBLO, CCWO allows modification of the TPA while the AMA angle is reduced, resulting in alignment of the proximal tibial epiphysis on the tibial shaft with the weight-bearing forces that are transmitted through the tibial diaphysis, a situation recorded in unaffected dogs of the control group. Therefore, there is no secondary translation or “balcony” effect, as seen in dogs with TPA >30° treated with TPLO [27–29].
CCWO should therefore be considered as a treatment for CCL-deficient canine stifles with TPA >30° and AMA angle >3° and may be more suitable than TPLO in this subset of dogs.
Additional studies that assess long-term clinical outcomes in dogs with CCL injury with TPA >30° treated with either TPLO or AMA-based CCWO are warranted, as are long-term radiographic studies to determine whether osteoarthritis progresses more severely in one group and to assess a possible relationship to the magnitude of patellar height decrease, PLT and the postoperative persistence of an AMA angle >3°.