Matching-adjusted indirect comparison between asciminib and flumatinib as first-line treatment for chronic myeloid leukemia in China

Introduction

Chronic myeloid leukemia (CML) is a malignant clonal disorder of hematopoietic stem cells characterized by the BCR::ABL1 fusion gene (1). For patients with CML in the chronic phase (CML-CP), tyrosine kinase inhibitors (TKIs) are commonly used in clinical practice, as guidelines recommend their use as the standard of care in first-line therapy (2,3). Imatinib was the first TKI to be approved in China, followed by the second-generation TKIs nilotinib, dasatinib, and flumatinib (4-6).

Despite the wide availability of multiple TKIs, a substantial unmet need persists in the first-line setting. All currently approved TKIs are adenosine triphosphate (ATP)-competitive agents and involve a mechanism of action inherently associated with off-target kinase inhibition reported to contribute to a broad spectrum of treatment-related adverse events (AEs) and impaired tolerability (6). Treatment intolerance is frequently associated with dose modifications, treatment interruptions, or discontinuations, which can compromise the achievement of response milestones and/or long-term disease control (6-8). Given the challenges in achieving an optimal balance between efficacy and safety, real-world rates of treatment-free remission (TFR) remain suboptimal (9,10).

In 2025, asciminib, the first BCR::ABL1 inhibitor that functions by specifically targeting the myristoyl pocket, received approval in China for first-line treatment of newly diagnosed CML-CP (4). Asciminib offers a distinctive advantage over ATP-competitive TKIs due to its novel mechanism of action in binding to the myristoyl pocket. This provides specificity in targeting ABL over other kinases, a reduction in off-target effects with improved safety and tolerability, and desirable efficacy outcomes (11). Results from the global ASC4FIRST trial demonstrated the superior efficacy and safety of asciminib over first- and second-generation TKIs in patients with newly diagnosed Philadelphia chromosome-positive (Ph+) CML-CP (12). As treatment goals increasingly emphasize quality of life, cost savings, and TFR, asciminib is being increasingly considered a favorable option in the first-line treatment of CML-CP, and this has been supported by clinical data on TFR from extension studies (13).

Currently, flumatinib is only approved in China and was not included as one of the comparators in the phase III clinical trials of asciminib (12). In the absence of direct comparative evidence from head-to-head trials, indirect comparison methods such as matching-adjusted indirect comparison (MAIC) are necessary to address this deficiency in data and generate comparative evidence for first-line settings.

This study aimed to bridge this gap by conducting an anchored MAIC based on individual patient-level data (IPD) from the ASC4FIRST trial and aggregate data from the FESTnd trial to assess the relative efficacy and safety of asciminib and flumatinib in patients with newly diagnosed CML-CP in China. We present this article in accordance with the STROBE reporting checklist (available at https://tcr.amegroups.com/article/view/10.21037/tcr-2026-1659/rc).

Methods

Design

An anchored MAIC was conducted in which imatinib was the common comparator between asciminib and flumatinib (Figure 1). In this context, IPD from the ASC4FIRST trial were used for the asciminib and imatinib arms, whereas published aggregate data from the FESTnd trial provided information on flumatinib and imatinib. According to the established MAIC methodology, no patients were excluded from the IPD in analysis, instead differential weights were assigned to patients to overcome any bias related to heterogeneity in the two trials. The IPD were reweighted to match the baseline characteristics reported in the aggregate data of Chinese CML-CP patients from the FESTnd trial to account for imbalances in key prognostic factors and effect modifiers across the trials.

Figure 1 Evidence network of studies included in the MAIC. The solid line represents a direct comparison. The dotted line represents an indirect comparison. IPD, individual patient-level data; MAIC, matching-adjusted indirect comparison.

Data sources

Efficacy and safety data for newly diagnosed CML-CP were obtained from two sources: the ASC4FIRST (NCT04971226) and FESTnd (NCT02204644) trials. IPD were obtained from ASC4FIRST, a multicenter, open-label, randomized phase III trial that compared asciminib with investigator-selected TKIs, including imatinib, in the treatment of adults with newly diagnosed Ph+ CML-CP. Published aggregate data were obtained from FESTnd, a multicenter, open-label, randomized phase III trial that compared flumatinib with imatinib in the treatment of adults with newly diagnosed CML-CP at a 1:1 ratio. Key design characteristics of the ASC4FIRST and FESTnd trials are summarized in Table S1. This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments.

Endpoints

The selection of clinical outcomes for comparison was challenging due to the variability in the availability of relevant data at the same time points in both trials.

Regarding, efficacy outcomes, given the reported prognostic significance of key response outcomes, early molecular response (EMR) at 12 weeks and major molecular response (MMR) at 48 weeks and 96 weeks were analyzed in this MAIC (14,15). Published data for the FESTnd trial reported outcomes at “1 year” which was assumed to be equivalent to data at 48 weeks available from the ASC4FIRST trial.

Additionally, as the FESTnd trial did not report outcomes at 96 weeks (~equivalent to 2 years) and as anomalies were identified in the published time-to-event curve, the 96-week (or equivalent) results for flumatinib were estimated by extrapolating from the 48-week data under the assumption of a constant odds ratio (OR) over time. The assumption of a constant OR for asciminib and flumatinib was validated by empirically testing the ORs at two different time points.

For the outcome of safety, discontinuations due to AEs reported by 48 weeks and/or equivalent time points in both trials were used for the MAIC.

Prognostic variables and effect modifiers

In line with standard practice of indirect comparisons, effect modifiers and baseline prognostic variables for the MAIC were identified based on global standards, clinical expert input, and evidence from the published literature (16). Age, sex, platelet count, and white blood cell (WBC) count were selected as the key matching variables for reweighting IPD from the ASC4FIRST trial to align with the FESTnd population. These parameters are core metrics associated with the European Treatment and Outcome Study Long-Term Survival (ELTS) and Sokal scoring systems, which were used in the two trials, respectively (2,17) and are prognostic scoring systems for predicting survival outcomes in patients with CML (16).

Statistical analysis

The analysis used balancing entropy to reweight the ASC4FIRST IPD to match the baseline characteristics reported in the FESTnd trial and monitored the effective sample size (ESS) to ensure weight stability. The reduced ESS does not reflect patient removal, instead it supports the credibility of the indirect comparison as the heavily down-weighted patients contribute little, making the weighted dataset statistically behave like a smaller, but more comparable sample. Weight-distribution histograms were used to confirm balanced distribution of weights assigned to patients. Individual patient weights were adjusted via R software (The R Foundation for Statistical Computing, Vienna, Austria) to account for the baseline differences between trials. ORs and their 95% confidence intervals (CIs) and risk ratios (RRs) and their 95% CIs were calculated after matching was performed for efficacy and safety outcomes, respectively.

Results

Baseline characteristics

Both the ASC4FIRST and FESTnd trials enrolled adults with newly diagnosed Ph+ CML-CP and with an Eastern Cooperative Oncology Group (ECOG) performance status of 0–1, applying largely similar inclusion and exclusion criteria (detailed in Table S1). Baseline characteristics for the overall study populations were summarized as descriptive statistics, and key baseline variables (age, sex, platelet count, and WBC) were matched after reweighting the ASC4FIRST population to match with the aggregated data of FESTnd (Tables 1,2). The ESS for asciminib after matching was 184.78, representing a 7.6% reduction from the original 200 patients in ASC4FIRST, compared with 196 patients in FESTnd. There was a balanced distribution of assigned weights noted in the weight-distribution histograms (Figure 2).

Figure 2 Weight-distribution histograms. ESS, effective sample size.

Efficacy outcomes of asciminib and flumatinib at different time points

EMR

The adjusted EMR rate at 12 weeks for asciminib and flumatinib was 90.0% and 82.1%, respectively, with asciminib-treated patients demonstrating significantly higher odds (1.95 times) of achieving EMR than those receiving flumatinib (95% CI: 1.05–3.73; P=0.03) (Figures 3,4).

Figure 3 Comparative analyses of efficacy and safety outcomes (%). EMR, early molecular response; MMR, major molecular response.

Figure 4 Summary of the OR and RR for the efficacy and safety outcomes. For efficacy (OR), values >1 favor asciminib, while values <1 favor flumatinib. For safety (RR), values <1 favor asciminib, while values >1 favor flumatinib. AE, adverse event; CI, confidence interval; EMR, early molecular response; MMR, major molecular response; OR, odds ratio; RR, risk ratio.

MMR

After MAIC, the MMR rate at 48 weeks was significantly higher for asciminib than for flumatinib (66.5% versus 52.6%), with asciminib-treated patients being 1.79 times more likely to achieve MMR than those treated with flumatinib (95% CI: 1.17–2.75; P=0.006).

Under the assumption of constant a OR estimated at 48 weeks, the estimated MMR rate for flumatinib at 96 weeks was 61.25%, while that for asciminib was 74.0%, indicating a significantly greater likelihood of achieving MMR with asciminib than with flumatinib (OR: 1.79; 95% CI: 1.17–2.75; P=0.006).

Safety outcome: discontinuation due to AEs

Asciminib demonstrated a favorable safety profile, with a lower proportion of AE-related discontinuations by 48 weeks compared with flumatinib in both the unadjusted and adjusted analyses (5.5% versus 10.2% for both comparisons). The risk of AE-related discontinuations was 71% lower with asciminib than with flumatinib, representing a significantly superior safety profile (RR: 0.29; 95% CI: 0.10–0.84; P=0.02).

Discussion

In the absence of a direct head-to-head comparison between asciminib and flumatinib in China, we conducted an anchored MAIC to strengthen the comparative evidence for first-line CML-CP treatments. To our knowledge, this is the first comparative analysis of asciminib and flumatinib as initial therapy in this setting and included structured reweighting of IPD from the ASC4FIRST trial to improve the validity of the efficacy and safety estimates.

The approach used in our analysis provides a more robust estimate of relative treatment effects as compared to naïve indirect comparisons or standard network meta-analyses, particularly given the key differences present in patient populations across studies. A MAIC allows for the matching of patients for whom IPD are available with patients from a comparator study by excluding those patients which would be deemed ineligible based on the reported aggregate baseline characteristics of the comparator. Additionally, the reliability of the comparative results can be enhanced via the adjustment of the remaining “matched” patients to align with the observed marginal distribution of key effect modifiers and prognostic factors in the comparator study (i.e., means and standard deviations of continuous variables and proportions of categorical variables) (18).

The findings of this MAIC are consistent with results reported from global clinical studies demonstrating a favorable benefit-risk profile for asciminib compared with other ATP-competitive TKIs (19,20). Our anchored MAIC, which included adjustments for age, sex, WBC count, and platelet count, offers strong internal validity despite a limited but clinically appropriate set of covariates for patients with newly diagnosed CML-CP. Since spleen size was not available for the asciminib-treated patients and blast count was not available for flumatinib-treated patients, these were not included in the adjustment. Additionally, the risk scores from the two trials were not included as one of the effect modifiers since the scoring systems were different between the trials.

In line with other indirect comparative studies, our analysis evaluated molecular response, as hematologic and cytogenetic response provide limited information on the residual disease kinetics, and many patients achieving complete hematologic response and complete cytogenetic response may experience relapse (21,22). Molecular response monitoring is an optimal choice for monitoring patients with CML-CP, as it offers the highest degree of sensitivity and an ease of evaluation that does not require bone marrow puncture (21-24). Furthermore, it is well established that the clinical outcomes of patients with CML are closely associated with molecular response; specifically, EMR at 12 weeks is a validated predictor of long-term survival and attainment of molecular response 4, whereas MMR functions as a robust prognostic indicator for disease progression (2,25,26). The clinical significance of these milestones is corroborated by the recommendations of European LeukemiaNet, which states that MMR confers a near-total probability of CML-specific survival, as disease progression is uncommon once this level of cytoreduction has been achieved (2,17,27,28). Thus, the outcomes generated in our MAIC align with the accepted importance and relevance of key outcomes in similar comparative studies.

Selecting the optimal first-line treatment for patients with newly diagnosed CML-CP is critical, particularly given the wide array of TKIs available (2,29,30). Asciminib, with its novel mechanism of action that specifically targets ABL myristoyl pocket, increases selectivity and reduces off-target toxicities as compared to ATP-site inhibitors (31). Clinical trials have proven the superiority of asciminib’s favorable safety and tolerability profile over that of investigator-selected TKIs, suggesting it may transform the first-line treatment paradigm (20,21). Meanwhile, flumatinib is another second-generation ATP-competitive BCR::ABL1 TKI that was developed as an imatinib derivative to enhance potency (16). In another MAIC that anchored flumatinib to imatinib and indirectly compared outcomes across approved second-generation TKIs, no significant superiority in MMR rates was found for flumatinib over other approved second-generation TKIs once baseline characteristics were adjusted for (30). Similarly, a recent real-world comparative analysis by Zhang et al. evaluating first-line TKIs in patients with CML reported comparable but numerically lower MMR rates with flumatinib than with other second-generation TKIs (32). These findings provide a strong basis for using statistical methods to correlate the ASC4FISRT results more generally across TKIs that were not part of the trial. In our MAIC, asciminib demonstrated a significantly lower risk of AE-related discontinuations compared with flumatinib, suggesting that asciminib may offer a meaningful incremental benefit as a first-line therapy for CML-CP in China, providing superior molecular response and improved tolerability.

As with any analysis that synthesizes indirect evidence, there were several limitations to this study. The MAIC was based on some key underlying assumptions—such as the comparability of patient populations and the consistency of treatment effects across studies—which, if violated, could have resulted in biased estimates. Consistent with the approach used by Bucher et al. and Phillippo et al., any inherent biasness was addressed by reweighting the IPD from ASC4FIRST with the aggregate patient characteristics in FESTnd trial (33,34). The variables that were selected for matching are established in CML prognostic systems and were supported by inputs from clinical experts. Additional constraints in our analysis include the dependence on the baseline characteristics reported in the Chinese flumatinib study and the need to extrapolate the 96-week MMR for flumatinib via a constant OR derived at 48 weeks. The extrapolation approach used in this MAIC avoided a reliance on unstable digitized curves, such as the discrepancies in the published flumatinib MMR KM curve (14) and maintained consistency with the anchored relative effect. However, it relied on strong assumptions—that the 48-week OR was unbiased and remained constant up to week 96 and that cross-trial differences in response definitions and follow-up durations did not materially distort relative effects. In line with MAIC guidance from the National Institute for Health and Care Excellence Decision Support Unit, these extrapolated estimates should therefore be interpreted with caution and considered supportive rather than definitive evidence of long-term comparative efficacy (33). Nonetheless, the anchored design, the alignment between adjusted and unadjusted findings, and the consistency with the broader asciminib evidence base—including a similar analysis by Atallah et al. and other global comparative studies—support the robustness and relevance of our results. Finally, as flumatinib is not widely available globally, there are no head-to-head randomized trials and no robust indirect comparisons of flumatinib with asciminib, with only data from a few Asian countries being reported in the literature. Our MAIC addresses an important deficiency in the available data, providing meaningful comparative insights to strengthen the evidence supporting the use of asciminib in the first-line treatment of CML-CP.

Conclusions

Asciminib consistently outperformed flumatinib across all key comparative endpoints in this anchored MAIC of first-line treatment for patients with newly diagnosed CML-CP in China. This higher efficacy was complemented by a clinically meaningful safety advantage, with significantly fewer treatment discontinuations due to AEs. These findings further support asciminib’s clinical value and its increasing use as a standard of care in the first-line setting, representing a meaningful shift from current ATP-competitive options.

Acknowledgments

Medical writing support was provided by Lovneet Saini and Aditi Kataria (both of Novartis Healthcare Pvt. Ltd., India) in accordance with the Good Publication Practice guidelines.

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://tcr.amegroups.com/article/view/10.21037/tcr-2026-1659/rc

Peer Review File: Available at https://tcr.amegroups.com/article/view/10.21037/tcr-2026-1659/prf

Funding: This study was funded by Novartis Pharma.

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://tcr.amegroups.com/article/view/10.21037/tcr-2026-1659/coif). All authors report that this study was funded by Novartis Pharma. M.T., X.G., and B.Z. report an employment relationship with Beijing Novartis Pharma Co. Ltd. A.C. and S.G. reports an employment relationship with Novartis Healthcare Private Limited in Hyderabad. The authors have no other conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Expert Consensus Working Group on Individualized Whole-Course Management of Chronic Myeloid Leukemia in China. Chinese expert consensus of chronic myeloid leukemia individualized whole process management. J Clin Hematol 2026;39:1-20.
2. Apperley JF, Milojkovic D, Cross NCP, et al. 2025 European LeukemiaNet recommendations for the management of chronic myeloid leukemia. Leukemia 2025;39:1797-813. [Crossref] [PubMed]
3. Hochhaus A, Saussele S, Rosti G, et al. Chronic myeloid leukaemia: ESMO Clinical Practice Guidelines for diagnosis, treatment and follow-up. Ann Oncol 2017;28:iv41-51. [Crossref] [PubMed]
4. Chinese Society of Hematology, Chinese Medical Association. Guideline for the diagnosis and treatment of chronic myeloid leukemia in China (2025). Zhonghua Xue Ye Xue Za Zhi 2025;46:1081-93.
5. Jiang L, Yang M. Clinical Efficacy and Safety of Flumatinib in Newly Diagnosed Chronic Myelogenous Leukemia. Pharmazie 2023;78:13-6. [Crossref] [PubMed]
6. García-Gutiérrez V, Hernández-Boluda JC. Tyrosine Kinase Inhibitors Available for Chronic Myeloid Leukemia: Efficacy and Safety. Front Oncol 2019;9:603. [Crossref] [PubMed]
7. Hijiya N, Mauro MJ. Asciminib in the Treatment of Philadelphia Chromosome-Positive Chronic Myeloid Leukemia: Focus on Patient Selection and Outcomes. Cancer Manag Res 2023;15:873-91. [Crossref] [PubMed]
8. Jabbour E, Kantarjian H. Chronic myeloid leukemia: 2025 update on diagnosis, therapy, and monitoring. Am J Hematol 2024;99:2191-212. [Crossref] [PubMed]
9. Bruzzese A, Vigna E, Martino EA, et al. Treatment‐Free Remission in Chronic Myeloid Leukemia: Revisiting the “W” Questions. European Journal of Haematology 2025;115:218-31.
10. Shanmuganathan N, Osborn M, Hughes TP. Which Is the Best Tyrosine Kinase Inhibitor for Newly Diagnosed Chronic Myelogenous Leukemia? Am Soc Clin Oncol Educ Book 2025;45:e473082. [Crossref] [PubMed]
11. Hochhaus A, Kim DW, Cortes JE, et al. Asciminib monotherapy in patients with chronic myeloid leukemia in chronic phase without BCR::ABL1(T315I) treated with at least 2 prior TKIs: Phase 1 final results. Leukemia 2025;39:1114-23. [Crossref] [PubMed]
12. Hochhaus A, Wang J, Kim DW, et al. Asciminib in Newly Diagnosed Chronic Myeloid Leukemia. N Engl J Med 2024;391:885-98. [Crossref] [PubMed]
13. Yoshida C, Takaku T. Asciminib: the next-generation bullet for first-line treatment of chronic myeloid leukemia. Med 2024;5:856-8. [Crossref] [PubMed]
14. Held N, Atallah EL. Real-world Management of CML: Outcomes and Treatment Patterns. Curr Hematol Malig Rep 2023;18:167-75. [Crossref] [PubMed]
15. Zhang Y, Chen W, Pathak A, et al. Burden of tyrosine kinase inhibitor failure in Chinese chronic myeloid leukemia patients: a systematic literature review. J Comp Eff Res 2022;11:621-37. [Crossref] [PubMed]
16. Lauseker M, Hoffmann VS, Pfirrmann M. Prognostic scoring systems in chronic myeloid leukaemia. Leukemia 2025;39:1046-52. [Crossref] [PubMed]
17. Zhang L, Meng L, Liu B, et al. Flumatinib versus Imatinib for Newly Diagnosed Chronic Phase Chronic Myeloid Leukemia: A Phase III, Randomized, Open-label, Multi-center FESTnd Study. Clin Cancer Res 2021;27:70-7. [Crossref] [PubMed]
18. Hochhaus A, Baccarani M, Silver RT, et al. European LeukemiaNet 2020 recommendations for treating chronic myeloid leukemia. Leukemia 2020;34:966-84. [Crossref] [PubMed]
19. Signorovitch JE, Sikirica V, Erder MH, et al. Matching-adjusted indirect comparisons: a new tool for timely comparative effectiveness research. Value Health 2012;15:940-7. [Crossref] [PubMed]
20. Réa D, Boquimpani C, Mauro MJ, et al. Health-related quality of life of patients with resistant/intolerant chronic phase chronic myeloid leukemia treated with asciminib or bosutinib in the phase 3 ASCEMBL trial. Leukemia 2023;37:1060-7. [Crossref] [PubMed]
21. Cortes JE, Hughes TP, Wang J, et al. Asciminib demonstrates superior efficacy and safety in newly diagnosed chronic myeloid leukemia in the ASC4FIRST trial. Blood 2026;147:1433-46. [Crossref] [PubMed]
22. Press RD. Major molecular response in CML patients treated with tyrosine kinase inhibitors: the paradigm for monitoring targeted cancer therapy. Oncologist 2010;15:744-9. [Crossref] [PubMed]
23. Fachi MM, Tonin FS, Leonart LP, et al. Comparative efficacy and safety of tyrosine kinase inhibitors for chronic myeloid leukaemia: A systematic review and network meta-analysis. Eur J Cancer 2018;104:9-20. [Crossref] [PubMed]
24. Baccarani M, Soverini S, De Benedittis C. Molecular monitoring and mutations in chronic myeloid leukemia: how to get the most out of your tyrosine kinase inhibitor. Am Soc Clin Oncol Educ Book 2014;167-75.
25. Hughes T, Deininger M, Hochhaus A, et al. Monitoring CML patients responding to treatment with tyrosine kinase inhibitors: review and recommendations for harmonizing current methodology for detecting BCR-ABL transcripts and kinase domain mutations and for expressing results. Blood 2006;108:28-37. [Crossref] [PubMed]
26. Hughes TP, Saglio G, Kantarjian HM, et al. Early molecular response predicts outcomes in patients with chronic myeloid leukemia in chronic phase treated with frontline nilotinib or imatinib. Blood 2014;123:1353-60. [Crossref] [PubMed]
27. Breccia M, Alimena G. The significance of early, major and stable molecular responses in chronic myeloid leukemia in the imatinib era. Crit Rev Oncol Hematol 2011;79:135-43. [Crossref] [PubMed]
28. Pennisi MS, Stella S, Vitale SR, et al. BCR-ABL1 Doubling-Times and Halving-Times May Predict CML Response to Tyrosine Kinase Inhibitors. Front Oncol 2019;9:764. [Crossref] [PubMed]
29. Wang R, Cong Y, Li C, et al. Predictive value of early molecular response for deep molecular response in chronic phase of chronic myeloid leukemia. Medicine (Baltimore) 2019;98:e15222. [Crossref] [PubMed]
30. Zhang JJ, Qian YL, Wu ZY, et al. Comparative efficacy and safety of first-line tyrosine kinase inhibitors in chronic myeloid leukemia: a systematic review and network meta-analysis. Transl Cancer Res 2024;13:3783-97. [Crossref] [PubMed]
31. Cortes JE, Hochhaus A, Hughes TP, et al. Asciminib (ASC) demonstrates favorable safety and tolerability compared with each investigator-selected tyrosine kinase inhibitor (IS TKI) in newly diagnosed chronic myeloid leukemia in chronic phase (CML-CP) in the pivotal phase 3 ASC4FIRST study. Blood 2024;5;144:475.
32. Zhang X, Xu N, Yang Y, et al. Comparison of the Efficacy Among Nilotinib, Dasatinib, Flumatinib and Imatinib in Newly Diagnosed Chronic-Phase Chronic Myeloid Leukemia Patients: A Real-World Multi-Center Retrospective Study. Clin Lymphoma Myeloma Leuk 2024;24:e257-66. [Crossref] [PubMed]
33. Phillippo D. Population-adjusted treatment comparisons: Overview of approaches and recommendations from the NICE DSU. 64th Biometrisches Kolloquium; 2018.
34. Bucher HC, Guyatt GH, Griffith LE, et al. The results of direct and indirect treatment comparisons in meta-analysis of randomized controlled trials. J Clin Epidemiol 1997;50:683-91. [Crossref] [PubMed]

(English Language Editor: J. Gray)