Low Skeletal Muscle Index Predicts Neoadjuvant Immunochemotherapy Tolerance and Survival in Oesophageal Squamous Cell Carcinoma
Yizhou Huang, Maohui Chen, Zhenyuan Yang, Bingqiang Cai, Yongcong Zhang, Hongjin Wang, Shuliang Zhang, Taidui Zeng, Chun Chen, Bin ZhengABSTRACT
Background
Low skeletal muscle index (SMI) may identify host vulnerability in patients receiving intensive cancer therapy. However, its association with treatment tolerance and outcomes during neoadjuvant immunochemotherapy (nICT) for oesophageal squamous cell carcinoma (OSCC) remains unclear.
Methods
This multicentre retrospective cohort included 461 patients with locally advanced OSCC who initiated nICT between 2020 and 2023; 404 underwent oesophagectomy. Pretreatment SMI at the third lumbar vertebra was measured by two blinded radiologists. Low SMI was defined using sex‐specific outcome‐derived cutoffs, with sensitivity analyses using published thresholds. Neoadjuvant treatment tolerance grade (NAT‐TG) was used as a composite measure of treatment tolerance and feasibility, integrating treatment modification and failure to undergo curative surgery.
Results
Among 461 patients (mean age, 61.2 ± 7.3 years; 21.3% female), low SMI was present in 121 (26.2%) and was independently associated with an unfavourable NAT‐TG composite (NAT‐TG2–3 vs. NAT‐TG1; adjusted OR, 2.48; 95% CI, 1.62–3.82; p < 0.001). The association persisted after excluding six progression‐related surgical omissions (adjusted OR, 2.51; 95% CI, 1.59–3.95; p < 0.001). The overall distribution of treatment‐related adverse event severity differed between groups ( p = 0.012), with grade ≥ 3 events occurring in 36.4% of patients with low SMI and 22.9% of those without low SMI. In the surgical population, low SMI was independently associated with severe postoperative complications (adjusted OR, 2.92; 95% CI, 1.51–5.63; p = 0.001). During a median follow‐up of 43 months, 165 deaths occurred. The 1‐year mortality rate in the intention‐to‐treat population was 8.5% (95% CI, 5.9%–11.0%). In the primary intention‐to‐treat model, low SMI was associated with poorer overall survival (adjusted HR, 2.36; 95% CI, 1.67–3.33; p < 0.001). Associations were also observed for surgical overall survival (adjusted HR, 1.91; 95% CI, 1.31–2.80) and disease‐free survival (adjusted HR, 1.57; 95% CI, 1.11–2.22). Results were generally consistent when published SMI thresholds were applied.
Conclusions
Pretreatment low SMI was associated with less favourable nICT tolerance and feasibility, a higher incidence of severe postoperative complications and poorer survival. CT‐derived skeletal muscle assessment may complement baseline risk stratification, while intervention strategies require prospective validation.