A nomogram based on plasma HSP90α and NLR for predicting prognosis in advanced gastric cancer patients treated with PD-1 inhibitors

Purpose: To investigate the use of heat shock protein 90alpha (HSP90α) as a marker for prognostic evaluation and efficacy monitoring in patients receiving PD-1 inhibitors treatment for advanced gastric cancer (AGC). Methods: We investigated the value of HSP90α in AGC patients treated with PD-1 inhibitors from a clinical perspective using human plasma samples. Results: In summary, plasma HSP90α was significantly associated with neutrophil-to-lymphocyte (NLR) in AGC patients at baseline. Regarding short-term efficacy, HSP90α levels decreased considerably after PD-1 inhibitor treatment in the partial response (PR) group (P=0.016). Furthermore, there was no significant difference between HSP90α levels in stable disease (SD) group before and after immunotherapy (P=0.659). However, HSP90α levels were considerably greater in AGC patients at disease progression and eventual PD-1 inhibitor therapy failure compared to baseline (P=0.041, P=0.005). Notably, plasma HSP90α, treatment lines, metastatic sites, and NLR level were independent predictive variables for overall survival (OS) in AGC patients receiving PD-1 inhibitors treatment before and after propensity score matching. Additionally, we constructed the nomogram model depending on the above independent prognostic variables, which can well differentiate the clinical prognosis of patients (P<0.001). And ROC curves, calibration curves, and decision curve analysis curves revealed promising discrimination and accuracy of the nomogram. Finally, plasma HSP90α showed specific prognostic value in different subgroups of clinical characteristics. Conclusion: Plasma HSP90α can be used as a marker for efficacy monitoring and prognostic assessment in AGC patients receiving PD-1 inhibitors treatment. We combined plasma HSP90α, NLR, and clinical characteristics to construct a nomogram for predicting the prognosis of gastric cancer immunotherapy, providing a powerful tool for clinical decision-making.