Systemic inflammation and patient-reported quality of life as combined prognosticators in patients with solid tumors

The modified Glasgow Prognostic Score (mGPS) and quality of life (QoL) are associated with overall survival (OS) in oncology. Here, we investigated their interplay and combined prognostic potential in patients with solid tumors.

Patients were prospectively enrolled at the Division of Oncology, Medical University of Vienna. Baseline QoL was assessed using the EORTC QLQ-C30. Patients were categorized by mGPS (0-2) and global health status/QL2 score (Low=0-33, intermediate= ≥33-66 and high= ≥67-100). Cox models included mGPS, continuous QL2, age, sex, ECOG, tumor type, metastatic and treatment status. Incremental model fit and discrimination were assessed using likelihood-ratios and five-fold cross-validated time-dependent AUCs.

Overall, 1452 patients were analyzed. Patients with mGPS 1/2 had a significantly lower QL2-score compared to mGPS 0 (p < 0.001). 1171 patients had survival data; adjusted analyses included 1076 patients and 549 deaths. At 6 months, adjusted HRs were 2.23 (95% CI, 1.80-2.77) for mGPS 1 versus 0, 3.22 (2.38-4.37) for mGPS 2 versus 0, and 0.895 (0.860-0.930) per 10-point increase in QL2. Associations attenuated over follow-up. Each marker improved model fit beyond clinical factors and the other marker (both p < 0.001). Cross-validated AUCs increased from 0.734 to 0.803 at 6 months and from 0.718 to 0.789 at 12 months after adding both markers. The additional AUC gains from QL2 beyond mGPS were 0.008 and 0.010.

mGPS and QL2 provide complementary prognostic information beyond measured clinical factors. External validation and evaluation of clinical utility are required before using the combined approach.