Search for a command to run...
To the Editor: Proximal fibular osteotomy (PFO) has emerged as a minimally invasive, knee‑preserving procedure for mild‑to‑moderate medial‑compartment knee osteoarthritis (KOA), offering faster recovery and fewer complications than high tibial osteotomy (HTO);[1] however, its mid‑ to long‑term survival rates and predictors of conversion to subsequent surgery remain unclear. Consequently, we conducted a large cohort study to assess the incidence and timing of conversion from PFO to subsequent surgery and to identify associated perioperative predictors. This study was approved by the Ethics Committee of the Hebei Medical University Third Hospital (No. 2014-015-1), and all participants signed an informed consent form before the surgery. The inclusion criteria were as follows: (1) radiographic medial-compartment KOA (Kellgren–Lawrence [K–L] grade 2–4), (2) predominant self-reported medial knee pain, and (3) pain unresponsive to conservative treatment of ≥3 months. The exclusion criteria were as follows: (1) comorbidities limiting ambulation more than knee pain (e.g., diabetic neuropathy and inflammatory arthritis); (2) knee surgery or intra-articular injection within 6 months; (3) oral corticosteroid use within 6 weeks; (4) prior knee replacement or tibial osteotomy; and (5) valgus alignment (hip–knee–ankle [HKA] angle >180°) or any other conditions (e.g., prior knee/ankle surgery and cerebral palsy) affecting lower‑limb function. This study is a retrospective secondary analysis of the prospectively maintained Surgical Site Infection in Orthopedic Surgery (SSIOS) registry, which manually captures and updates data on all orthopedic inpatients at the Orthopedics Department of the Hebei Medical University Third Hospital.[2,3] The potential predictors were identified a priori, including demographic characteristics, lifestyle factors, medical comorbidities, clinical characteristics, and perioperative laboratory parameters. The PFO procedures were performed by a surgical team with a minimum of 5 years of orthopedic experience, including two senior surgeons, each of whom had performed over 30 PFO procedures before the study. All operations were conducted under either general or spinal anesthesia. The surgical technique followed previously described protocols. A 3–5 cm lateral incision was made at the proximal third of the fibula. After adequate exposure, a 2-cm segment of fibula was excised 6–10 cm distal to the fibular head using an oscillating or fret saw. Bone wax was applied to seal the proximal and distal fibular ends.[4] Patients with mechanical symptoms and magnetic resonance imaging (MRI)-confirmed intra-articular pathology additionally underwent arthroscopic procedures, including 78 cases of debridement alone (removal of cartilage flaps, loose bodies, or hypertrophic synovium) and 42 cases of debridement combined with partial meniscectomy. Postoperatively, patients were encouraged to mobilize the operated limb and were prescribed analgesics as needed. Ankle pumps, quadriceps stretching, and straight-leg raise exercises were initiated on the day of surgery. Assisted ambulation was permitted 24 hours postoperatively if quadriceps strength was grade IV or higher and the incision site showed satisfactory healing. Patients who underwent PFO between March 11, 2014 and September 1, 2019, were followed until conversion to subsequent surgery (arthroplasty or other osteotomy), loss to follow-up, death, or September 1, 2023 (minimum 4-year potential follow-up). The primary outcome was time from PFO to subsequent surgery, determined from our electronic operative reports and radiographs; procedures performed elsewhere were verified through outside medical records provided by the patients. Baseline characteristics were compared between conversion and nonconversion groups using descriptive statistics, with normally distributed continuous variables presented as mean ± standard deviation (SD), categorical variables reported as n (%), and between-group differences tested via t-test and chi-square test with corresponding P values provided. Preoperative serum biomarkers and age cut-offs were determined through the surv_cutpoint function, and biomarkers with missing values <5% were addressed by multiple imputation. Cumulative incidence of conversion to subsequent surgery was estimated by Kaplan–Meier analysis (with subgroup curves by age, K–L grade, and arthroscopy) and by Fine–Gray competing-risks regression (death as a competing event). Person-years accrued from PFO date to the first occurrence of conversion, death, loss to follow-up, or September 1, 2023. Univariable Cox models screened a priori, selected predictors (age, body mass index [BMI], residence, K–L grade, settlement value, preoperative visual analog scale [VAS] score, HKA angle, use of arthroscopic procedure, and several preoperative serum biomarkers), followed by a multivariable Cox model (robust sandwich estimator for bilateral cases; Schoenfeld residuals tested proportionality). Three sensitivity analyses assessed robustness: Stratification by surgeon to account for experience, comparison of multiple‑imputation vs. complete‑case results for serum biomarkers (red blood cell [RBC] and creatinine [CREA]). Post hoc analyses included separate Cox models adding 6‑month postoperative HKA angle and VAS score as predictors, and a subgroup analysis restricted to patients aged 40–60 years. All tests were two‑sided (α = 0.05) and a P value <0.05 was defined as the threshold for statistically significant differences. All analyses were performed using R software (Version 4.1.3, R Foundation for Statistical Computing, Vienna, Austria. https://www.R-project.org/). We screened the knees of 832 patients for potential inclusion. The final cohort included 546 patients with 764 knees undergoing PFO procedures, including 218 simultaneous bilateral PFO. The cohort (median age 62 years, range 39–89 years) was predominantly female, overweight, with varus alignment, moderate-to-severe knee pain, and medial-compartment KOA (K–L grade 3–4) [Supplementary Table 1, https://links.lww.com/CM9/C826]. Median follow-up was 7.5 years; 60 patients were lost to follow-up and 23 died [Supplementary Figure 1, https://links.lww.com/CM9/C826]. At the end of follow-up, 91 conversions (88 total knee arthroplasties, 3 HTOs) occurred—equivalent to 17.4 events per 1000 person-years. The cumulative incidence of conversion was 3.2% (95% confidence interval [ CI]: 2.1–4.7%) at 3 years, 7.6% (5.9–9.8%) at 5 years, and 13.0% (10.6–15.8%) at 8 years [Figure 1]. Subgroup analyses showed higher 8-year cumulative incidence of conversion in patients ≤63 years (15.4%; 95%CI 11.7–19.0%), those with K–L grade 4 KOA (19.9%; 95%CI 14.9–24.6%), and those who did not undergo arthroscopy (14.7%; 95%CI 11.6–17.7%) [Supplementary Figures 2–4, https://links.lww.com/CM9/C826].Figure 1: Cumulative incidence curve (with 95% confidence intervals [CIs] as shaded areas) for patients undergoing one or more proximal fibular osteotomies (n = 764) with conversion to subsequent surgery defined as the event of interest. The number of patient knees at risk is also presented for selected timepoints.In multivariable-adjusted Cox models, K–L grade 4 (adjusted hazard ratio [aHR] 3.19; 95%CI 1.50–6.81) and greater settlement value (aHR 1.08 per unit; 1.07–1.09) were associated with increased conversion risk, whereas age >63 years (aHR 0.36; 0.22–0.59) and use of arthroscopy (aHR 0.19; 0.08–0.43) were protective. Preoperative biomarkers—RBC <4.46 × 1012/L (aHR 2.67; 1.08–8.01) and CREA <60 µmol/L (aHR 2.71; 1.43–5.13)—also predicted higher conversion. Multicollinearity was evaluated using variance inflation factors (all VIFs <2), confirming no significant collinearity among covariates [Supplementary Tables 2 and 3, https://links.lww.com/CM9/C826]. Results were consistent in a multivariable Fine–Gray competing-risks model. Sensitivity analyses (surgeon stratification; multiple imputation vs. complete case) did not substantially alter findings. Post hoc analyses suggested that higher VAS score and HKA ≤174° at 6 months postoperatively were linked to increased conversion; higher age was associated with an increased rate of conversion in patients aged 40–60 years, with 52 years identified as the cutoff age in this age group [Supplementary Tables 4–13, https://links.lww.com/CM9/C826]. This study provided medium‑term (5–8 years) survival data for PFO in KOA, the comparable survival rates to HTO, and even unicompartmental knee arthroplasty emphasize its potential in knee-preserving treatment. Notably, patients under 52 years—particularly those receiving adjunctive arthroscopic procedures—may derive greater benefit, whereas selected patients over 63 years of age also achieved favorable outcomes when preoperative parameters such as K–L grade and settlement value fell within appropriate thresholds. Besides, beyond conventional preoperative assessments such as K–L grade, HKA angle, and other physiological leg axes, settlement value (aHR 1.08 per mm) should be incorporated into routine evaluation for medial-compartment KOA due to its potential to reflect nonuniform tibial plateau settlement and predict surgical conversion. To enhance clinical applicability, a Youden index-based cutoff of 7.56 mm was identified; patients with settlement >7.56 mm had more than twice the risk of conversion (aHR 2.06; 95% CI: 1.31–3.25) [Supplementary Figure 5 and Supplementary Table 14, https://links.lww.com/CM9/C826]. Furthermore, although we noted borderline associations of lower RBC and CREA with conversion, the evidence is not yet strong enough to guide specific interventions. These exploratory findings are hypothesis-generating and warrant further study. Post hoc analyses highlight the value of early postoperative assessment in identifying patients at risk for conversion. At 6 months, those with persistently high VAS scores despite PFO likely experienced insufficient medial unloading, and patients whose post-correction HKA remained ≤174° may have retained a deleterious varus vector that predisposes to further degeneration. These exploratory results underscore the need for standardized 6-month functional and radiographic milestones to guide timely decision-making, while acknowledging that confirmatory prospective studies are warranted before routine clinical adoption.[5] Limitations of our study should also be acknowledged. First, this single-center study enrolled patients from a tertiary university-affiliated specialty orthopedic hospital. Although PFO is technically straightforward—minimized by our standardized protocol—our findings may not generalize to centers without equivalent infrastructure. Second, as an innovative, research-funded procedure, our cohort skewed older (mean age 63 years) with more severe KOA (K–L grade 3–4 in 82.5% participants), which may not reflect ideal indications. Third, the SSIOS registry, while prospectively maintained and incorporating a priori follow-ups for subpopulations, does not capture preoperative functional knee scores; their prognostic value therefore requires further validation. Finally, using conversion to subsequent surgery as our primary outcome may not fully represent PFO “survival”, since reoperation decisions can be influenced by socioeconomic status, pain tolerance, and activity demands that we did not assess. In conclusion, in a large cohort with a median follow-up of 7.5 years, we observed a notably lower conversion rate from PFO for KOA to subsequent surgery. Among the established predictors, higher age exerted an inverse protective effect, with elderly patients demonstrating a significantly lower probability of conversion to subsequent surgery compared with younger counterparts. Besides, settlement value, combined use of arthroscopic procedure, and several preoperative serum biomarkers were also identified and should be considered when evaluating surgical indications and providing preoperative and postoperative counseling to patients. Acknowledgments We sincerely thank all the patients in this study. Funding This research was supported by the Hebei Department of Science and Technology High-Level Talent Team Construction Project (No. 225A7703D). Conflicts of interest None.