Abstract
BACKGROUND: Despite the clinical success of PD-1/PD-L1 inhibitors, resistance remains a major barrier to durable responses in advanced cervical cancer. Identifying readily available biomarkers predictive of immunotherapy resistance may improve treatment selection and patient outcomes. METHODS: This single-center retrospective cohort study included 140 patients with histologically confirmed advanced or recurrent cervical cancer who received at least two cycles of PD-1 inhibitor therapy. Clinical characteristics, laboratory indices-including neutrophil-to-lymphocyte ratio (NLR) and lactate dehydrogenase (LDH)-and prior treatment history were analyzed. Resistance was defined as confirmed progressive disease (iCPD) according to iRECIST criteria. Logistic regression and receiver operating characteristic (ROC) analyses were performed to identify independent predictors and evaluate model performance. RESULTS: Of 140 patients, 68 (48.6%) developed PD-1 inhibitor resistance. Resistant cases exhibited significantly higher baseline NLR (4.09 ± 1.64 vs. 3.02 ± 1.28, p < 0.001), elevated LDH (264.4 ± 88.9 U/L vs. 216.8 ± 69.4 U/L, p = 0.003), and more frequent prior chemotherapy (67.6% vs. 47.2%, p = 0.018). In multivariable logistic regression analysis, baseline NLR (OR = 1.62, 95% CI 1.21-2.17, p = 0.002), LDH (OR = 1.05 per 10 U/L increase, 95% CI 1.01-1.09, p = 0.012), and prior chemotherapy (OR = 2.08, 95% CI 1.01-4.28, p = 0.047) were independently associated with early PD-1 inhibitor resistance. The combined model incorporating NLR, LDH, PD-L1 expression, and prior chemotherapy achieved an AUC of 0.842 (95% CI 0.773-0.911), outperforming individual parameters (p < 0.05). Exploratory analysis showed that PD-1-based combination regimens achieved a disease control rate of 72.2% in previously resistant patients. CONCLUSION: Systemic inflammatory and metabolic markers, together with clinical treatment history, can effectively predict PD-1 inhibitor resistance in advanced cervical cancer.