Paper 35: Failure following medial patellofemoral ligament reconstruction: a propensity-matched comparative analysis

Objectives: Medial patellofemoral ligament reconstruction (MPFLR) has emerged as a foundation of patellofemoral instability (PFI) treatment, with numerous studies demonstrating its efficacy in preventing recurrent instability. While generalized ligament laxity and bony risk factors have been shown to increase risk of reconstruction failure, optimal graft selection, fixation technique, and concomitant procedures in MPFLR remain an area of uncertainty. There is a paucity of research on differences in demographic, radiographic, and operative variables between matched failed and non-failed MPFLR cohorts. The aims of this study were to evaluate such differences between failed and non-failed reconstructions in order to identify potential risk factors for MPFLR failure and to inform surgical decision making to prevent recurrence of PFI or the need for revision.
Methods: To identify the failed MPFLR cohort, we retrospectively identified patients who underwent primary MPFLR between 2004 and 2022 and underwent subsequent revision MPFLR for recurrent instability at a single institution. Patients presenting with repeat revision or graft over-tensioning were excluded. A control group of patients diagnosed with PFI who underwent MPFLR without documentation of graft failure or recurrent patellar dislocation was propensity matched 2:1 for age and gender. Demographic measures, graft characteristics, fixation techniques, and concomitant procedures were recorded. Preoperative radiographic measures including patellar tendon to lateral trochlear ridge (PT-LTR), tibial tubercle to trochlear groove (TT-TG), tibial tubercle to posterior cruciate ligament (TT-PCL), and Caton-Deschamps Index (CDI) were collected. Statistical analysis was conducted utilizing RStudio (Rv4.4.1) to compare failed and non-failed reconstructions.
Results: A total of 102 patients were included (34 failed MPFLR, 68 non-failed MPFLR). The mean age at surgery was similar between groups (19.9 +- 6.9 vs 19.3 +- 6.3 years, p > 0.9). Patients in the failed cohort had a significantly higher BMI compared to controls (28.4 +- 7.6 vs 25.4 +- 7.6, p = 0.022). Hypermobility was more frequent among the failed cohort, though not statistically significant (41% vs 24%, p = 0.065), and mean time from surgery to revision for the failed cohort was similar to follow-up time for non-failed reconstructions (2.4 vs 2.8 years, p = 0.33) (Table 1).Radiographic parameters were comparable between groups. Mean PT-LTR was 9.8 +- 4.5 mm in the failure cohort and 11.7 +- 9.2 mm in controls (p = 0.4). TT-TG distance was similar (15.8 +- 4.5 mm vs 15.1 +- 5.3 mm, p = 0.7), as was TT-PCL (21.4 +- 5.3 mm vs 22.5 +- 5.5 mm, p = 0.3). CDI did not differ between groups (1.2 +- 0.2 for both, p = 0.5). Dejour classification distribution showed no significant differences (p = 0.3), with most failed reconstructions classified as type A (67% vs 41% in controls) (Table 2).Concomitant procedures performed with MPFLR did not differ significantly between groups (p = 0.502). Isolated MPFLR was the most common in both cohorts (61.8% failed vs 57.4% control) (Figure 1). Graft tendon type (semitendinosus vs gracilis) was similar between groups (p = 0.7). However, use of autograft versus allograft differed significantly with 100% of control patients receiving allograft compared to 61% among the failed reconstructions (p < 0.001). Graft diameter trended larger in the failure group, though not statistically significant (p = 0.088). Patellar fixation technique differed significantly with failed reconstructions more likely to use inlay fixation compared to controls (48% vs 18%, p = 0.003). Femoral fixation method also differed significantly with failed reconstructions less likely to use inlay fixation compared to controls (76% vs 94%, p < 0.007) (Table 3).
Conclusions: In this propensity-matched cohort, failed MPFL reconstructions were characterized by significantly higher BMI and greater utilization of autograft tissue. While demographic, radiographic, and concomitant procedures were otherwise comparable, fixation technique differed where failed cases more frequently used patellar inlay fixation and femoral onlay or loop constructs, whereas controls more commonly underwent patellar onlay and femoral inlay fixation. Collectively, our results demonstrate the potential importance of graft selection and fixation approach in reducing the risk of MPFLR failure and may help guide surgical decision making to optimize the success of this procedure in treating PFI.