Radiographic measurement to predict the risk of periprosthetic tibial plateau fracture after Oxford cementless unicompartmental knee arthroplasty - 23/09/26

Doi : 10.1016/j.otsr.2026.104881

Rapeepat Narkbunnam, Paphon Hirunyachoke, Krittin Kittipibul, Keerati Chareancholvanich, Kit Awirotananon, Chaturong Pornrattanamaneewong ⁎

Department of Orthopaedic Surgery, Faculty of Medicine Siriraj Hospital, Mahidol University, 2 Wanglang Road, Bangkoknoi, Bangkok 10700, Thailand

^⁎ Corresponding author.

Sous presse. Épreuves corrigées par l'auteur. Disponible en ligne depuis le Wednesday 23 September 2026

Abstract

Objectives

Cementless Oxford unicompartmental knee arthroplasty (OUKA) reduces the radiolucent lines associated with cemented fixation but carries an increased risk of early postoperative medial tibial plateau fracture, particularly in Asian populations. This study aimed to identify postoperative radiographic parameters predicting periprosthetic medial tibial plateau fracture after cementless OUKA and to derive a clinically usable cut-off.

Methods

This retrospective study analysed knees undergoing cementless medial OUKA between 2015 and 2022. Knees were classified into a fracture group (14 knees) and a non-fracture group (65 knees). Demographic data and twelve radiographic parameters were compared, including femorotibial angle (FTA), medial eminence line (MEL) type, tibial tray coronal angle (TCA), vertical tibial cut depth (VCD), vertical-cut-to-medial-cortex (VM) distance, V/M ratio, keel-to-medial-cortex (KM) distance, K/M ratio, lateralized tibial tray (LTT) ratio, keel-tip-to-posterior-cortex (KTP) distance, and tibial tray slope (TS). Logistic regression and receiver operating characteristic (ROC) analyses identified predictors and optimal cut-offs; reliability was assessed using intraclass correlation coefficients (ICC).

Results

Baseline characteristics did not differ between groups. The mean time to fracture was 37 days (range, 7–56). The fracture group had a higher TCA (89.0 vs 87.0°, p = 0.005), VCD (10.9 vs 7.4 mm, p < 0.001) and V/M ratio (0.4 vs 0.2, p < 0.001), and a lower VM distance (26.1 vs 30.2 mm, p < 0.001), KM distance (12.0 vs 15.0 mm, p < 0.001), K/M ratio (1.5 vs 1.8, p = 0.002) and KTP distance (7.0 vs 9.0 mm, p = 0.023). VCD showed the strongest discriminative performance (AUC 0.984). The optimal VCD cut-off was 9.4 mm, yielding a sensitivity of 92.9%, specificity of 95.4%, positive predictive value of 81.3%, and negative predictive value of 98.4%.

Conclusion

A postoperative VCD greater than 9.4 mm represents a postoperative, rather than a preoperative, radiographic marker of periprosthetic medial tibial plateau fracture after cementless OUKA. This measurement is simple, reproducible and reliable, and because it reflects an intraoperative surgical action, it can additionally be assessed intraoperatively to guide surgical technique. Prospective validation is required before clinical adoption.