Impact of Perioperative Body Composition Changes in Patients Undergoing Surgery for Gastrointestinal and...

The survival impacts of longitudinal changes in skeletal muscle and adipose tissue in patients undergoing curative surgery are not as yet fully understood. We investigated the prognostic impact of perioperative changes in CT-derived body composition after curative surgery for gastrointestinal and genitourinary malignancies.

In total, 1868 patients who underwent curative-intent surgery for gastrointestinal (n=1273) or genitourinary (n=595) malignancies were analyzed. Body composition was measured on preoperative CT and on a postoperative CT obtained a median of 6.0 months after surgery (range 3.0-14.9). Perioperative changes in skeletal muscle area (SMA), skeletal muscle density (SMD), visceral adipose tissue area (VATA) and subcutaneous adipose tissue area (SATA) were expressed as post/pre ratios. Overall survival (OS) and cancer-specific survival (CSS) were compared between patients with low and normal ratios, and multivariable Cox models were fitted. Sensitivity analyses addressed multiple testing, immortal time bias, additional confounders, the timing of the postoperative CT, and cancer type.

Gastrointestinal cancers and advanced pathological stage were significantly associated with low SMA, SMD, VATA, and SATA post/pre ratios, indicating unfavorable perioperative changes in skeletal muscle and adipose tissue. Low SMA, SMD, VATA, and SATA post/pre ratios were each significantly associated with poorer OS. Low SMA, VATA, and SATA post/pre ratios were also significantly associated with poorer CSS. On multivariable analysis, low SATA (HR 1.60, 95% CI 1.28-2.00, P < .001) and SMD (HR 1.49, 95% CI 1.18-1.88, P < .001) post/pre ratio remained independently associated with worse OS, together with age ≥70 years, male sex, gastrointestinal malignancy, and pathological stage IV disease. In contrast, low VATA and low SMA post/pre ratios were not independently associated with OS. Furthermore, a low SATA post/pre ratio remained independently associated with poor CSS (HR 1.56, 95% CI 1.19-2.06, P = 0.001), whereas the association of a low SMD post/pre ratio with CSS (HR 1.33, 95% CI 1.00-1.76, P = 0.049) did not remain significant after adjustment for multiple testing. The association of SATA loss with survival persisted after adjustment for multiple testing and additional confounders and in analyses addressing immortal time bias.

Postoperative depletion of body composition parameters, especially subcutaneous adipose tissue, was associated with poor long-term outcomes following curative cancer surgery.