CT-derived abdominal organ volumetrics for predicting recurrence-free and disease-free survival in resected non-small cell lung cancer: a multicenter retrospective cohort study

Background: Postoperative recurrence remains a major challenge after curative-intent resection for non-small cell lung cancer (NSCLC), and conventional clinicopathologic factors do not fully capture recurrence risk heterogeneity. Although computed tomography (CT)-derived body composition biomarkers have been associated with outcomes in NSCLC, the prognostic relevance of abdominal solid-organ volumetrics remains insufficiently characterized. We investigated whether preoperative CT-derived abdominal organ volumes and body surface area (BSA)-normalized organ indices were associated with recurrence-related outcomes in resected NSCLC.

Methods: This multicenter retrospective cohort study included 1,762 patients with clinically and pathologically confirmed NSCLC who underwent curative-intent resection at five centers and had preoperative CT within 1 month before surgery. A healthy control cohort of 3,263 individuals was included for baseline organ-volumetric comparison. Clinical and laboratory variables were extracted from preoperative records, with laboratory tests generally obtained within 7 days before surgery. A deep learning-based automated segmentation pipeline quantified the spleen, liver, pancreas, bilateral kidneys, and bilateral adrenal glands, and organ indices were normalized to BSA. Propensity score matching (PSM) was restricted to patient-control organ-volumetric comparisons. Recurrence was ascertained from follow-up clinical and imaging records. Recurrence-free survival (RFS) and disease-free survival (DFS) were evaluated as recurrence-related outcomes. Multivariable Cox models were performed in the full NSCLC cohort and adjusted for age, sex, T stage, pathological types, diabetes, hypertension, smoking, drinking, body mass index (BMI), albumin, neutrophil-to-lymphocyte ratio (NLR), and platelet-to-lymphocyte ratio (PLR).

Results: The NSCLC cohort had a median age of 60 years, and 934 patients (53.0%) were women. During median follow-up periods of 39.7 months for RFS and 39.4 months for DFS, 198 RFS events and 219 DFS events occurred. Compared with healthy controls after PSM, patients with NSCLC showed larger spleen, liver, kidney, and adrenal volumes and smaller pancreatic volume. In multivariable Cox analyses, lower pancreatic volume and pancreatic index were independently associated with shorter RFS [hazard ratio (HR) =0.735, P<0.001; HR =0.729, P<0.001, respectively] and DFS (HR =0.791, P=0.002; HR =0.787, P=0.001, respectively). Lower left adrenal gland index (LAGI) was also associated with shorter RFS (HR =0.845, P=0.030) and DFS (HR =0.833, P=0.01). Lower liver volume (LVV) index was associated with RFS only (HR =0.834, P=0.02), whereas lower left adrenal gland volume (LAGV) was associated with DFS only (HR =0.843, P=0.03).

Conclusions: Preoperative CT-derived abdominal organ volumetrics, particularly pancreatic and left adrenal gland metrics, were independently associated with recurrence-related outcomes after curative-intent resection of NSCLC. These imaging biomarkers may provide complementary information for postoperative recurrence risk stratification. Further prospective validation is warranted.