Purpose
The incidence, progression, and prognosis of cancer could be affected by inflammation and nutrition. Female patients have different inflammatory and nutritional states depending on their age and tumor types. It is important to screen for suitable prognostic indicators in female patients with cancer of different ages and tumor types.
Patients and Methods
Baseline clinicopathologic and laboratory characteristics of 1502 female patients with cancer were obtained from a multicenter cohort study. Concordance indices (C-indices) were used to evaluate the prediction accuracy of following inflammation- and nutrition-based indicators: advanced lung cancer inflammation index (ALI), systemic immune inflammation index (SII), modified geriatric nutritional risk index (mGNRI), albumin-to-globulin ratio (AGR), prognostic nutritional index (PNI), lymphocyte-to-C-reactive protein ratio (LCR), controlling nutritional status score (CONUT), modified Glasgow prognostic score (mGPS), and lymphocyte-to-C-reactive protein score (LCS).
Results
The most suitable indicators in different female populations with cancer had C-indices as follows: LCR (0.668; 95% CI, 0.644–0.693) for all females; AGR (0.681; 95% CI, 0.619–0.743) for young females; LCR (0.667; 95% CI, 0.628–0.706) for middle-aged females; ALI (0.597; 95% CI, 0.574–0.620) for elderly females; LCR (0.684; 95% CI, 0.621–0.747) for females with reproductive system cancer; and ALI (0.652; 95% CI, 0.624–0.680) for females with non-reproductive system cancer.
Conclusion
The most suitable indicators for the different female populations with cancer are summarized as follows: LCR for all females, AGR for young females, LCR for middle-aged females, ALI for elderly females, LCR for females with reproductive system cancer, and ALI for females with non-reproductive system cancer.