Yao Yang, Wanli Zang, Yanling Zhou, Yuan Chen, Ben Waterson, Junchen Zhu, Chen Zhu, Xianzuo Zhang
The LPA of preoperative CT-derived quadriceps muscle volumes identified three distinct KOA muscular phenotypes associated with different symptom burdens and recovery trajectories after TKA. Patients who have preserved quadriceps muscles achieved the best absolute outcomes, whereas those who had marked atrophy showed greater relative improvement. These findings support CT-based phenotype-guided perioperative rehabilitation.
BACKGROUND: Knee osteoarthritis (KOA) is characterized by complex structural and muscular alterations. Quadriceps muscle atrophy is a critical, but heterogeneous feature influencing pain, function, and postoperative recovery. Understanding muscle phenotypes may provide novel insight into disease heterogeneity and guide personalized rehabilitation.
METHODS: This retrospective study included 191 consecutive patients undergoing primary robot-assisted total knee arthroplasty (TKA; 76.4% women; mean age 67 years (range 46 to 86)), with preoperative computed tomography (CT) imaging and six-month follow-up. There were three CT-derived quadriceps muscle volumetric ratios (rectus femoris/intermedius, vastus medialis, and vastus lateralis normalized to femoral volume) entered into latent profile analyses (LPA) to identify distinct muscle phenotypes. Associations between latent classes and functional outcomes were evaluated using linear mixed-effects models.
RESULTS: The LPA identified three subgroups: Class 1 (n = 125) with moderate muscle ratios and intermediate function, Class 2 (n = 27) with higher muscle ratios and the best outcomes, and Class 3 (n = 39) with reduced muscle ratios and the worst baseline status. At six months, Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC, the improvements of WOMAC in classes 1 to 3 are as follows: 28.6, 41.8, 37.8) and visual analog scale (VAS, the improvements of VAS in classes 1 to 3 are as follows: 3.5, 4.3, 2.8) improved significantly in all classes (all P < 0.05), with significant class × time interactions (all P < 0.001). The range of motion also improved over time (P < 0.001), without a significant class × time interaction (P = 0.562). Overall, Class 2 achieved the best postoperative outcomes.
CONCLUSION: The LPA of preoperative CT-derived quadriceps muscle volumes identified three distinct KOA muscular phenotypes associated with different symptom burdens and recovery trajectories after TKA. Patients who have preserved quadriceps muscles achieved the best absolute outcomes, whereas those who had marked atrophy showed greater relative improvement. These findings support CT-based phenotype-guided perioperative rehabilitation.