Chrome Extension
WeChat Mini Program
Use on ChatGLM

A Random Forest Model to Predict the Efficacy of KL-A167 in Previously Treated Recurrent or Metastatic Nasopharyngeal Carcinoma Based on the Data from the KL167-II-05-CTP Study.

JOURNAL OF CLINICAL ONCOLOGY(2024)

Cited 0|Views10
No score
Abstract
e18012 Background: KL-A167 is a fully humanized monoclonal antibody targeting programmed cell death-ligand 1. KL167-Ⅱ-05-CTP study had evaluated the efficacy and safety of KL-A167 in Chinese patients with previously treated recurrent or metastatic (R/M) nasopharyngeal carcinoma (NPC). The results of KL167-Ⅱ-05-CTP study showed that not all patients with R/M NPC could benefit from KL-A167 therapy, and an effective model to predict the efficacy of KL-A167 is necessary. Methods: The baseline data of clinical characteristics, laboratory examination and imaging of patients with R/M NPC in KL167-Ⅱ-05-CTP study were collected, who received KL-A167 therapy. The progression-free survival (PFS) was selected as the predicted variable. Patients were randomly divided into training cohort (70%) and test cohort (30%). In the training cohort, uni-Cox regression was used to estimate the value in predicting PFS, and least absolute shrinkage and selection operator (LASSO)-Cox regression was performed to screen clinical, laboratory and imaging indexes to build random forest model. The accuracy of this random forest model was validated in the test cohort. Results: In this study, 137 patients with R/M NPC were enrolled, and the median PFS of them was 2.689 (95% confidence interval [ CI]: 1.475-4.098) months. A total of 14 indexes with P < 0.05 were selected to perform LASSO-Cox regression, and 10 indexes (blood lactate dehydrogenase, blood creatinine, left ventricular ejection fraction, body temperature, weight, blood lymphocyte count, blood chloride ion concentration, Eastern Cooperative Oncology Group performance status, blood triglyceride and blood alanine aminotransferase) were used to build prediction model. The C-index of risk score from this model was 0.820 and 0.790 in the training and test cohort, respectively. Patients were divided into the high and low risk groups according to the median risk score, and the median PFS of the high risk group was shorter than that of the low risk group (training cohort: 1.377 vs. 4.197 months, P < 0.001, hazard ratio [ HR] = 1.142, 95% CI: 1.111-1.175; test cohort: 3.475 vs. 10.984 months, P < 0.001, HR = 1.100, 95% CI: 1.060-1.142). Conclusions: In this study, a random forest model based on 10 clinical, laboratory and imaging indexes was constructed and showed a potential value in predicting the PFS of KL-A167 therapy in Chinese patients with previously treated R/M NPC, which could help to guide clinical trials and practices.
More
Translated text
AI Read Science
Must-Reading Tree
Example
Generate MRT to find the research sequence of this paper
Chat Paper
Summary is being generated by the instructions you defined