PurposeThrombocytopenia is among the most common chemotherapy-related hematologic toxicities. We aim to determine the predictors of oxaliplatin chemotherapy-induced thrombocytopenia in patients with gastrointestinal tumors to guide the clinic.MethodsClinical data of 750 patients with a malignant gastrointestinal tumor were included as the primary cohort. Basic clinical data, serological indices, and anthropometric indices of these patients were collected. According to the presence or absence of CIT, univariate analysis was performed to identify significant factors for multivariate analysis. In R language software, nomogram was constructed based on the results of multi-factor analysis, and the calibration curve and ROC curve were drawn.ResultsUnivariate analysis identified 17 factors as closely related to CIT occurrence, namely age, lymph node metastasis (N) stage, metastasis (M) stage, lung metastasis, other site metastasis, chemotherapy regimen, course of treatment, total dose of oxaliplatin, AST, albumin, neutrophils, monocytes, baseline platelets, transferrin, natural killer (NK) cell, phase angle, and SMI (P < 0.10). The binary logistic multivariate regression analysis revealed five independent risk factors for developing CIT (P < 0.05), including the M stage, total dose of oxaliplatin, albumin, baseline thrombocyte count, and NK cell. Based on the results of multivariate logistic regression analysis, R software was used to establish a nomogram model. The calibration curve shows that the combined predictor has good consistency. The area under the ROC curve was 0.877 and the best cut-off value was 0.3579613 (sensitivity, 78.9%; specificity, 81.8%), which showed the better prediction efficiency.ConclusionThe total dose of oxaliplatin, M stage, albumin, baseline platelet count, and NK cell was independent risk factors for CIT. The sequentially constructed histogram model had a good predictive effect on the risk of thrombocytopenia caused by oxaliplatin chemotherapy in patients with gastrointestinal malignancies.
Read full abstract