Abstract
We sincerely thank the writers of this Letter 1 for the insightful and constructive comments on our study. 2 We appreciate the recognition of the potential contribution of our work and have carefully considered each point. Our responses are provided below.
Heterogeneity in Disease Staging
We acknowledge the reviewer’s concern regarding heterogeneity in disease staging among the included studies. In our analysis, the target population was defined as patients with “unresectable advanced NSCLC,” which broadly corresponds to stage III to IV disease in clinical practice. However, some included trials did not clearly report staging information or may have enrolled mixed-stage populations. This broader inclusion reflects the variability in reporting standards across trials and was intended to capture real-world clinical settings where strict staging classification is not always consistently applied. In cases where disease stage was not explicitly reported, eligibility was determined based on a comprehensive review of the full text, including the applied inclusion criteria and the nature of comparator treatments, to ensure consistency with the definition of unresectable advanced disease. Nevertheless, this approach may introduce clinical heterogeneity, which should be considered when interpreting the pooled results.
Variability in Comparator Definitions
We agree that differences in comparator regimens across studies may contribute to heterogeneity. Although the analysis was framed as a comparison between herbal medicine combined with palliative chemotherapy and chemotherapy alone, some included trials incorporated additional background treatments such as radiotherapy or supportive care. In our study, these trials were included provided that chemotherapy remained the primary treatment backbone in both arms, reflecting the complexity of real-world oncology practice. To further explore this issue, subgroup analyses according to chemotherapy regimens were conducted and presented in the forest plots. The results showed that the overall heterogeneity did not substantially differ across subgroups, and the magnitude and direction of effect estimates were generally consistent. These findings suggest that variability in background treatment regimens may not have been a major contributor to the observed heterogeneity, although residual clinical heterogeneity cannot be entirely excluded.
Methodological Limitations and Subjective Outcomes
We acknowledge that blinding was not consistently implemented across the included trials, and allocation concealment was often unclear. This reflects a common methodological challenge in studies involving herbal medicine interventions, where the nature of the treatment may limit the feasibility of double-blinding. As a result, subjective outcomes such as Karnofsky Performance Status and quality-of-life measures may be particularly susceptible to performance and assessment bias.
These limitations were systematically evaluated through the risk-of-bias assessment and should be taken into account when interpreting the findings. In contrast, more objective outcomes, such as tumor response, are less likely to be substantially influenced by the absence of blinding. Overall, while the observed improvements in patient-centered outcomes are encouraging, cautious interpretation is warranted given these methodological considerations.
We appreciate these insightful comments, which highlight important considerations in the interpretation of evidence in this field. We hope that this clarification helps to further contextualize our findings and supports ongoing efforts to improve methodological rigor in future studies of integrative oncology.
Footnotes
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Korea Health Industry Development Institute (KHIDI) (Grant No. RS-2023-KH139460).