Abstract
This study aimed to explore the predictive role of peripheral blood indicators in the prognosis of cutaneous squamous cell carcinoma (cSCC) patients treated with immune checkpoint inhibitors (ICIs). Clinical data of 139 cSCC patients receiving ICIs treatment were retrospectively collected. Peripheral blood indicators, including blood cell counts, neutrophil-to-lymphocyte ratio (NLR), liver and kidney function markers, and inflammation markers, were examined. A binary logistic regression model was used to identify risk factors for non-response to ICIs, and a predictive model was constructed. Additionally, multiple linear regression and Pearson correlation analysis were employed to assess relevant influences and relationships. Results showed that immunotherapy timing, lymphocyte count, NLR, and C-reactive protein (CRP) were influencing factors for non-response to ICIs (all P<0.05). The area under the curve (AUC) for these indicators in predicting non-response risk was 0.651 (95% CI: 0.529-0.773), 0.671 (95% CI: 0.542-0.801), 0.775 (95% CI: 0.682-0.868), and 0.717 (95% CI: 0.573-0.861), respectively. The combined AUC of these four factors was 0.878 (95% CI: 0.790-0.966), with sensitivity and specificity of 76.0% and 93.0%, respectively. After internal verification, the constructed model exhibited predicted sensitivity and specificity of 80.00% and 94.29% respectively. Multiple linear regression analysis indicated that these four factors were independent predictors of progression-free survival (PFS) in cSCC patients. Immunotherapy timing, NLR, and CRP were negatively correlated with PFS (r = -0.235, -0.330, -0.494), while lymphocyte count was positively correlated with PFS (r = 0.326). In conclusion, peripheral blood indicators are valuable for predicting the response to ICIs in cSCC and can influence patients' PFS.