Description

The sample size was calculated according to the One-Sample Log-rank Test.During the treatment of amlotinib, metastatic site imaging was performed every 2-3 cycles, and the efficacy assessment and safety information were collected according to the RECIST version 1.1 criteria. After the completion of amlotinib treatment, survival follow-up was continued to clarify the survival status of the patients. When descriptive statistical analyses of clinicopathological characteristics were performed, categorical information was described by frequency counts and constitutive ratios, and quantitative information, such as conforming to normal distribution, was described by mean ± standard deviation, and not conforming to normal distribution, was described by median and interquartile spacing. Progression-free survival and overall survival at each time point were calculated by the Kaplan-Meier method and plotted as survival curves, Log-rank test was used to compare the prognostic differences between groups of latent variables, and the prognostic impact of latent variables on the anilotinib-treated patients was analysed using the COX regression risk model for exploratory analysis, and the risk ratios (hazard ratios (HRs) for quantitative evaluation. Variables with p-values less than 0.1 in the univariate COX regression analyses were screened by stepwise regression into the multivariate COX regression analyses. If the lapse rate exceeded 20%, sensitivity analyses were used to assess the impact of the lapse on the endpoint outcome event, which was performed by assuming a death outcome for the study subjects who were lost and analysing the endpoint event, and then treating the lost patients as normal and analysing the endpoint event and assessing the impact of the lost patients on the endpoint event. spss version 21.0 was used for the statistics, and a p<0.05 was considered to be clinically significant.