Pretreatment prognostic nutritional index and immune status for response to preoperative neoadjuvant chemotherapy in...

Pretreatment nutritional and immune status may be associated with response to preoperative neoadjuvant chemotherapy (NACT) in cervical squamous cell carcinoma (CSCC), but its clinical utility remains uncertain.

This exploratory retrospective study evaluated the association of pretreatment immunonutritional and tumor-related factors with short-term radiologic response to NACT. Ninety-two patients with histopathologically confirmed CSCC who received paclitaxel plus platinum-based NACT followed by surgery were included. Response was defined as a ≥30% reduction in the sum of target-lesion diameters on post-NACT magnetic resonance imaging relative to pretreatment imaging.

Sixty patients were responders and 32 were non-responders. Responders had higher pretreatment prognostic nutritional index (PNI) values, and FIGO stage distribution differed between groups. Higher PNI remained associated with response after adjustment for FIGO stage and pretreatment maximum tumor diameter (odds ratio [OR] = 1.154, 95% confidence interval [CI] 1.030-1.295; P = 0.014) and after adjustment for pretreatment MRI-assessed lymph-node metastasis and maximum tumor diameter (OR = 1.136, 95% CI 1.023-1.262; P = 0.017), with consistent findings in the weighted sensitivity analysis. Repeated nested cross-validation yielded areas under the receiver operating characteristic curve of 0.630 for PNI alone, 0.684 for five-variable ridge logistic regression, and 0.737 for random forest. Neither multivariable model showed a statistically reliable improvement over PNI alone. The random forest showed a train-validation AUC gap of 0.163 and was considered unstable and exploratory.

Pretreatment PNI may provide a simple marker for preliminary response risk stratification, but prospective external validation is required before clinical application.