I recorded my thoughts as I worked through your paper, so what you'll hear reflects my genuine reactions as a reader. I hope it gives you something useful to take forward.
The manuscript demonstrates a sophisticated methodological framework by integrating data from multiple sources: the global NAVIGATOR trial, ethnicity-specific subgroup analyses, and the dedicated Chinese phase I study. This totality-of-evidence approach is particularly valuable given the preliminary nature of the Chinese data and provides readers with a more complete picture of avapritinib's efficacy profile. The authors appropriately acknowledge the limitations of this approach while leveraging its strengths to provide clinically meaningful estimates.
"This review synthesises available data from published studies and from an ongoing phase I trial... with the understanding that final Chinese-specific results may refine the estimates."
The introduction effectively establishes the clinical context by documenting the dismal prognosis of 4L+ GIST patients (median PFS of 1–2 months, OS less than 6 months). This stark presentation of the therapeutic landscape provides compelling justification for the review and helps readers immediately appreciate the clinical significance of even modest efficacy improvements. The progression from imatinib through sunitinib and regorafenib to the current unmet need is clearly articulated.
The safety analysis is comprehensive and well-structured, presenting adverse event data in multiple complementary ways: any-grade TEAEs, grade ≥3 events, treatment-related AEs, serious AEs, discontinuation rates, and adverse events of special interest. The side-by-side comparison with global data allows readers to assess consistency, and the authors appropriately note that no unexpected safety signals emerged in the Chinese population. The inclusion of laboratory abnormalities adds depth to the safety characterization.
The inclusion of pharmacokinetic data demonstrating comparable exposure between Chinese and non-Asian patients at the same 300 mg dose is a critical strength. This bridging analysis provides pharmacological justification for the dose selection and addresses a key regulatory and clinical question about the need for ethnicity-specific dosing. The detailed PK parameters (AUC, Cmax, Ctrough, Tmax, t½) with coefficient of variation data allow readers to assess inter-patient variability.
"These values are similar to those observed in non-Asian patients, supporting the same dosing regimen."
While the manuscript reports ORR across various secondary KIT mutations, the data are presented in aggregate rather than by specific mutation type. Given that avapritinib's mechanism involves targeting the active conformation of KIT, understanding whether responses differ between ATP-binding pocket mutations (exon 13/14) and activation loop mutations (exon 17/18) would be clinically valuable for patient selection. The authors mention that "responses were observed across various secondary KIT mutations" but do not provide mutation-specific response rates.
"The Chinese trial did not mandate central mutation testing for all patients, so the relationship between specific secondary KIT mutations and efficacy could not be fully explored."
The manuscript acknowledges that duration of response data in Chinese patients is immature (median DoR "not reached" due to short follow-up). While the preliminary PFS and OS data appear encouraging, the durability of responses is a critical question for clinical decision-making. Longer follow-up will be essential to determine whether the 7.6-month median DoR observed globally is replicated in Chinese patients, and whether late resistance patterns differ.
While the manuscript reports cognitive effects in 33% of Chinese patients (mostly grade 1–2), there is limited discussion of how these events affect patients' daily functioning, quality of life, and willingness to continue treatment. Given that cognitive effects are a recognized class effect of type I TKIs that cross the blood-brain barrier, and that GIST is often a chronic disease requiring prolonged treatment, understanding the patient experience with these neurological adverse events is important for shared decision-making.
The manuscript appropriately acknowledges this limitation, but it remains a significant challenge. The Chinese cohort (n≈30) is substantially smaller than the global NAVIGATOR 4L+ population (n=111), resulting in wider confidence intervals for efficacy estimates (e.g., ORR 95% CI: 8–38% vs 15–31%). The preliminary nature of the data means that point estimates may shift with longer follow-up, and the current conclusions should be interpreted with appropriate caution.
"The main limitation of this review is the preliminary nature of the Chinese data. Final analysis with longer follow-up and larger sample size is needed to confirm the point estimates."
While the comparison with historical controls (median PFS 1–2 months in 4L+ GIST) provides context, the absence of a randomized comparator means that the absolute treatment effect cannot be quantified without bias. The manuscript acknowledges this limitation but does not fully explore how differences in supportive care, patient selection, or assessment schedules between the historical cohort and the current trials might influence the comparison.
The comparison between avapritinib (ORR ~20-22%) and ripretinib (ORR 11.8%) from the INVICTUS trial is presented without sufficient methodological caveats. The authors briefly note that "cross-trial comparisons are confounded by differences in patient populations and mutation profiles," but this critical point deserves more emphasis. Differences in mutation testing, response assessment criteria, and patient selection between NAVIGATOR and INVICTUS make direct comparison problematic.
The manuscript touches on PDGFRA D842V-mutant GIST but focuses primarily on the broader 4L+ KIT-mutant population. There is an opportunity to more explicitly develop a biomarker-guided framework for treatment positioning: when to use avapritinib based on mutation status, how to integrate molecular profiling into clinical decision-making, and where avapritinib fits relative to other emerging therapies in the treatment algorithm. This would enhance the clinical utility of the review.
"For patients with PDGFRA D842V mutations, avapritinib should be considered in earlier lines (even first line), but that is beyond the scope of this review."
The current review focuses on traditional efficacy endpoints (ORR, PFS, OS) and safety (AEs, laboratory abnormalities) but does not address patient-reported outcomes (PROs) or quality of life (QoL). Given the chronic nature of GIST treatment and the cognitive adverse events associated with avapritinib, incorporating PRO/QoL data would provide a more holistic assessment of treatment value. If such data exist from the Chinese trial, including them would strengthen the manuscript.
China's healthcare system faces unique challenges including regional disparities in access, variable insurance coverage, and competition from traditional Chinese medicines (mentioned briefly in the manuscript). There is an opportunity to discuss the practical implications of avapritinib's cost-effectiveness in the Chinese context, potential barriers to access, and strategies for implementation within the evolving Chinese oncology care landscape. Health economic considerations would make the review more relevant to Chinese healthcare policymakers and clinicians.
Dr. Morben has six years of clinical experience in veterinary medicine, having worked as a general practitioner in the United States from 2004 to 2010. She has extensive experience editing manuscripts in veterinary medicine, clinical human medicine, surgery, dentistry, genetics, and related fields. She joined Edanz in 2010 and became a board-certified editor in the life sciences in 2014, and is a current member of the Council of Science Editors.
Items 1 to 5 are addressable with targeted editorial support. Items 6 and 7 require more substantial engagement. Items 8 and 9 involve returning to the data or planning follow-up work. To discuss next steps, please contact your Edanz representative or visit edanz.com.
This Manuscript Assessment Report was prepared by Edanz Editorial using our AI-assisted assessment system. It is intended to support your understanding of your manuscript's current position and to inform your next steps; it is not a guarantee of any editorial outcome.
This report is generated using Anthropic's Claude. Edanz operates under Anthropic's commercial API terms, which include a configuration under which submitted content is not used to train AI models.
Information submitted to this system is transmitted via API to Anthropic's Claude for processing. Edanz applies reasonable security management measures, including training opt-out configuration. Edanz is not liable for information disclosure caused by third-party providers or causes not attributable to Edanz.
Edanz makes no warranty as to the accuracy, completeness, or fitness for any particular purpose of the assessments in this report. AI-generated content may include errors; users must verify accuracy before acting on any information. Edanz is not liable for damages arising from use of this report, except in cases of intent or gross negligence.
Edanz makes no warranty that AI-generated outputs do not infringe third-party intellectual property rights. Users are responsible for verifying outputs before use.