Response to Kong-Ying Lin, Shi-Chuan Tang, Cheung-Wu Zhang, Yong-Yi Zeng, Tian Yang

Abstract

We thank Lin KY et al for their comments on our manuscript entitled “Use of antibiotics during immune checkpoint inhibitors associates with lower survival in hepatocellular carcinoma”.1 Several issues were raised in the letter including the start time for survival analysis, the details of survival time with Kaplan-Meier curve and definition of cancer-related death. In the “Outcome definition” paragraph of our manuscript, it was mentioned that patients were observed from the start date of first immune checkpoint inhibitor as the index date till the occurrence of the outcome of interest or end of study. In the same paragraph, we have defined the primary outcome of cancer-related mortality as “liver cancer specified as the cause of death in the database” by the diagnosis codes. In the whole cohort, the 3-month, 6-month, 1-year and 2-year probability of cancer-related survival was 70.7% (95% CI: 66.3 – 75.4%), 57.4% (95% CI: 52.6 – 62.6%), 45.9% (95% CI: 41.1 – 51.4%) and 37.4% (32.4 – 43.3%), respectively. Figure 1 shows the Kaplan Meier survival plot of the whole cohort. Among the antibiotic users, the 3-month, 6-month, 1-year and 2-year probability of cancer-related survival was 50.9% (95% CI: 42.0 – 61.7%), 35.3% (95% CI: 26.9 – 46.4%), 26.8% (95% CI: 19.1 – 37.7%) and 19.3% (12.0 – 31.1%), respectively. For antibiotic non-users, the 3-month, 6-month, 1-year and 2-year probability of cancer-related survival was 77.8% (95% CI: 73.1 – 82.8%), 64.7% (95% CI: 59.4 – 70.6%), 52.2% (95% CI: 46.5 – 58.5%) and 43.6% (37.7– 50.5%), respectively. There was significant difference in cancer-related survival between the two groups (log-rank p < 0.001). Figure 2 shows the Kaplan Meier survival plot according to antibiotic use.