Association of malignant ascites with systemic inflammation and muscle loss after treatment in advanced-stage ovarian cancer

Abstract

Background: Malignant ascites is prevalent in advanced-stage ovarian cancer and may facilitate identification of the drivers of muscle loss. This study aimed to evaluate the association of ascites with changes in systemic inflammation and muscle after treatment of advanced-stage ovarian cancer.

Methods: We evaluated 307 patients with advanced-stage (III/IVA) ovarian cancer who underwent primary debulking surgery and adjuvant platinum-based chemotherapy between 2010 and 2019. The changes in skeletal muscle index (SMI) and radiodensity (SMD) were measured using pre-surgery and post-chemotherapy portal-venous phase contrast-enhanced computed tomography scans at L3. Systemic inflammation was measured using albumin levels, prognostic nutritional index (PNI), neutrophil-lymphocyte ratio (NLR), and platelet-lymphocyte ratio (PLR). Primary endpoint was the changes in SMI and SMD after treatment. Linear regression analysis was used to test associations between muscle change and other covariates. Mediation analysis was used to determine the mediator.

Results: The median (range) age was 53 (23-83) years. The median duration (range) of follow-up was 5.2 (1.1-11.3) years. Overall, 187 (60.9%) patients had ascites. The changes in muscle and systemic inflammatory markers after treatment were significantly different between patients with and without ascites (SMI: -3.9% vs. 2.2%, P < 0.001; SMD: -4.0% vs. -0.4%, P < 0.001; albumin: -4.4% vs. 2.1%, P < 0.001; PNI: -8.4% vs. -0.1%, P < 0.001; NLR: 20.6% vs. -29.4%, P < 0.001; and PLR: 1.7% vs. -19.4%, P < 0.001). The changes in SMI and SMD were correlated with the changes in albumin, PNI, NLR, and PLR (all P < 0.001). In multiple linear regression, ascites and NLR changes were negatively while albumin change was positively correlated with SMI change (ascites: β = -3.19, P < 0.001; NLR change: β = -0.02, P = 0.003; albumin change: β = 0.37, P < 0.001). Ascites and NLR changes were negatively while PNI change was positively correlated with SMD change (ascites: β = -1.28, P = 0.02; NLR change: β = -0.02, P < 0.001; PNI change: β = 0.11, P = 0.04). In mediation analysis, ascites had a direct effect on SMI change (P < 0.001) and an indirect effect mediated by NLR change (indirect effects = -1.61, 95% confidence interval [CI]: -2.22 to -1.08) and albumin change (indirect effects = -2.92, 95% CI: -4.01 to -1.94). Ascites had a direct effect on SMD change (P < 0.001) and an indirect effect mediated by NLR change (indirect effects = -1.76, 95% CI: -2.34 to -1.22) and PNI change (indirect effects = -2.00, 95% CI: -2.79 to -1.36).

Conclusions: Malignant ascites was associated with enhanced systemic inflammation and muscle loss after primary debulking surgery and adjuvant chemotherapy in advanced-stage ovarian cancer. The association between ascites and muscle loss may be mediated by systemic inflammation.

Keywords: Malignant ascites; Ovarian cancer; Skeletal muscle; Survival; Systemic inflammation.

Figure 1

In the hypothetic model, malignant ascites is a key trigger for skeletal muscle loss and enhanced systemic inflammation. The changes in systemic inflammation and skeletal muscle are correlated. The association between malignant ascites and muscle loss may be mediated by systemic inflammation.

Figure 2

Scatter plot depicting the relationship between changes in albumin and PNI and changes in SMI and SMD. The red and blue dots represent patients with and without malignant ascites, respectively. rs, Spearman's rank correlation coefficient; PNI, prognostic nutritional index; SMD, skeletal muscle radiodensity; SMI, skeletal muscle index. *P < 0.001.

Figure 3

Scatter plot depicting the relationship between changes in NLR and PLR and changes in SMI and SMD. The red and blue dots represent patients with and without malignant ascites, respectively. rs, Spearman's rank correlation coefficient; NLR, neutrophil‐lymphocyte ratio; PLR, platelet‐lymphocyte ratio; SMD, skeletal muscle radiodensity; SMI, skeletal muscle index. *P < 0.001.

Figure 4

Mediation analysis. (A, B) Effect of ascites on changes in NLR, albumin level, and SMI. (C, D) Effect of ascites on changes in NLR, PNI, and SMD. The changes in SMI or SMD are used as outcome variable, while the changes in NLR, albumin, or PNI are used as their mediator. NLR, neutrophil‐lymphocyte ratio; PNI, prognostic nutritional index; SMD, skeletal muscle radiodensity; SMI, skeletal muscle index.