We appreciate the insightful comments relating to our article. Regarding the first point, we report a landmark analysis for surgery status to overcome the possibility of immortal time bias.
In the landmark approach, a fixed time after the start of therapy is chosen as a landmark for analyzing survival. Only patients alive at the landmark are then included in the analysis and separated into two categories, distinguished by whether they received conversion surgery up to that time. Survival probabilities in the two groups at 4.2 months were illustrated using Kaplan–Meier (KM) curves, and median overall survival (OS) was calculated in the two groups.
Landmark analysis for OS was analyzed at 4.2 months until death or end of follow-up. 36 patients were included at 4.2 months; 6 patients with conversion surgery and 30 patients without conversion surgery prior to 4.2 months. One patient died, and 1 patient loss to follow-up before 4.2 months; these patients were not included at 4.2 months landmark analysis. The KM curve for OS by surgery status at the 4.2 months landmark is shown in Fig. 1. Compared with the median OS of 17.6 months in the group without conversion surgery, the median OS in the group with conversion surgery was not reached [landmark analysis]. The landmark analysis for surgery status was conducted in R version 4.1.3 with survival package (version 3.8.3).
Fig. 1.
Kaplan–Meier curve of overall survival (landmark analysis); with conversion surgery (blue) or without conversion surgery (red).
We will continue to follow up with this cohort to obtain more mature survival data and further clarify the long-term effect of the treatment and the contribution of conversion surgery.
Regarding the second point, the dichotomization of baseline haploid genome equivalents (hGE) using a data-driven cut-point, while optimizing performance across thresholds, raises valid concerns about potential overestimation of prognostic separation and reduced reproducibility. We appreciate this insightful critique. To strengthen the clinical relevance of hGE as a biomarker, reporting continuous associations and validating the selected threshold in an independent cohort would enhance robustness. Such steps could improve confidence in its actionable utility while addressing limitations inherent to dichotomization.
Our research team has been dedicated to investigating the clinical utility of circulating tumor DNA (ctDNA) in the detection and management of peritoneal metastasis in gastric cancer.1 We are conducting dynamic ctDNA monitoring clinical research and preparing the manuscript for submission. We look forward to presenting our findings soon.
Contributors
Collection and assembly of data: All authors.
Data analysis and interpretation: All authors.
Manuscript writing: Hong Yuan.
Final approval of manuscript: All authors.
Accountable for all aspects of the work: All authors.
Min Shi had accessed and verified the data.
Declaration of interests
No potential conflicts of interest were reported.
Reference
- 1.Yuan H., Xu F., Wang S., et al. Analysis of circulating tumor DNA identifies distinct therapeutic response to intraperitoneal and intravenous paclitaxel plus S-1 in gastric cancer patients with peritoneal metastasis. Ther Adv Med Oncol. 2024;16 doi: 10.1177/17588359231225038. [DOI] [PMC free article] [PubMed] [Google Scholar]