# Relationship of nutritional or systemic inflammatory markers with efficacy of gemcitabine and cisplatin with or without durvalumab therapy for patients with unresectable or metastatic biliary tract cancer: a retrospective study

## TL;DR

This study shows that certain blood markers, like mGPS and PLR, can predict survival outcomes in patients with advanced biliary tract cancer undergoing specific chemotherapy treatments.

## Contribution

The study identifies mGPS and PLR as significant predictors of survival in biliary tract cancer patients treated with gemcitabine, cisplatin, and durvalumab.

## Key findings

- Patients with mGPS 2 had significantly shorter median OS (7.9 months) compared to those with mGPS 0–1 (16.9 months).
- PLR ≥ 148 was associated with a median survival of 13.7 months, while PLR < 148 had not been estimated, showing a significant difference.
- Multivariate analysis confirmed mGPS 2 as an independent risk factor for shorter OS.

## Abstract

Nutritional or systemic inflammatory markers like modified Glasgow Prognostic Score (mGPS), prognostic nutritional index (PNI), neutrophil-lymphocyte ratio (NLR) and platelet-lymphocyte ratio (PLR) are useful prognostic indicators, but their role in patients with unresectable or metastatic biliary tract cancer receiving gemcitabine plus cisplatin (GC) or GC plus durvalumab is unclear. This study investigates the relationship of these markers with treatment outcomes and survival.

Chemotherapy consisted of protracted infusion of gemcitabine (1000 mg/m2/day) and cisplatin (25 mg/m2/day) on day 1 and 8, or gemcitabine (1000 mg/m2/day), cisplatin (25 mg/m2/day) on day 1 and 8 and durvalumab (1500 mg/day 1) given once every 21 days. After six cycles of GC plus durvalumab therapy, durvalumab monotherapy was administered at 1500 mg every 4 weeks. PNI, NLR, PLR and mGPS were assessed before treatment using established cutoffs. Overall survival (OS) and time to treatment failure (TTF) were analyzed with Kaplan–Meier and Cox regression models. The primary endpoint was OS, defined as the time from the start of chemotherapy to death.

This study analyzed 103 patients with unresectable or metastatic biliary tract cancer treated with GC or GC plus durvalumab. For patients with PLR ≥ 148 the median survival was 13.7 months and for those with PLR < 148 the median OS was not estimated; this difference was significant (p = 0.016). For patients with mGPS 0–1 and 2 the median OS was 16.9 months and 7.9 months, respectively, and this difference was also significant (p < 0.001). Univariate analysis revealed that, among the nutritional or systemic inflammatory markers tested, mGPS 2 and PLR ≥ 148 significantly predicted shorter OS. Multivariate analysis further confirmed that mGPS 2 is an independent risk factor for shorter OS.

A mGPS of 2 was a predictor of shorter OS for patients with unresectable or metastatic biliary tract cancer receiving GC or GC plus durvalumab. Evaluating mGPS prior to treatment initiation may support planning of individualized treatments that can enable clinicians to identify patients who are likely to benefit from chemotherapy and to exercise caution when considering systemic therapy for those with mGPS 2.

The online version contains supplementary material available at 10.1186/s40780-026-00555-5.