In their comprehensive systematic review and network meta-analysis, Fujisaki et al1 compared the efficacy and safety of direct oral anticoagulants (DOACs) for managing venous thromboembolism (VTE) in patients with cancer. We appreciate the authors’ efforts in synthesizing the evidence to guide clinicians on the selection of DOACs in this patient population. However, we noted that assessments of the certainty of evidence were not provided. The Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines recommend presenting assessments of certainty in the body of evidence for all outcomes analyzed.2 This information is crucial for readers to understand the reliability of treatment effect estimates, aiding clinicians in accurately interpreting and applying the findings in clinical practice.
In the authors’ analysis, apixaban was identified as the best anticoagulant for both recurrent VTE and major bleeding outcomes. However, relying solely on treatment rankings can lead to potentially spurious conclusions for several reasons.3 First, differences in rankings may be caused by chance. Second, even if differences are real, they may not be clinically significant. Lastly, and perhaps most importantly, treatment rankings do not account for the certainty of evidence; although apixaban is ranked highest, the evidence supporting this may be of low or very low certainty. Patients and clinicians may prefer a lower-ranked treatment with more certain evidence over a higher-ranked treatment with uncertain evidence. Therefore, it is essential to consider the certainty of evidence when determining treatment rankings to ensure that patients and clinicians make informed decisions on the implementation of findings from a network meta-analysis.
We used the Grading of Recommendations, Assessment, Development, and Evaluations3 approach to assess the certainty of evidence from the authors’ network meta-analysis for comparisons involving apixaban and outcomes of recurrent VTE and major bleeding. With respect to recurrent VTE, we found high certainty evidence for parenteral anticoagulation; moderate certainty for vitamin K antagonists; and very low certainty for rivaroxaban, edoxaban, and dabigatran compared to apixaban (Table 1). For major bleeding, we found very low certainty evidence for all anticoagulants compared to apixaban (Table 1). Therefore, the authors’ conclusion that DOACs seem to have similar efficacy, whereas apixaban may offer a more favorable safety profile is only supported by very low certainty evidence; this level of uncertainty, as per Grading of Recommendations, Assessment, Development, and Evaluations interpretation, indicates “an uncertain effect.” Clinicians should take this certainty of evidence into consideration when assessing the risk-benefit ratios of different DOACs for managing VTE in cancer patients.
Table 1.
GRADE Assessments of Certainty of Evidence From Fujisaki et al’s Network Meta-Analysis1
| Comparison | Direct Estimate HR (95% CI) | GRADE Certaintya | Indirect Estimate HR (95% CI) | GRADE Certaintya | NMA Estimate HR (95% CI) | GRADE Certainty | Narrative Summary |
|---|---|---|---|---|---|---|---|
| Recurrent VTE | |||||||
| VKA vs apixaban | 1.80 (0.42-7.69) | High | 2.85 (1.66-4.89) | Moderateb | 2.69 (1.62-4.47) | Moderatec | VKA probably increases the risk of recurrent VTE compared to apixaban. |
| Rivaroxaban vs apixaban | Not applicable | Not applicable | 1.27 (0.66-2.46) | High | 1.27 (0.66-2.46) | Very lowd | There is an uncertain effect of rivaroxaban compared to apixaban on recurrent VTE. |
| Parenteral vs apixaban | 1.75 (1.10-2.77) | High | 1.11 (0.25-4.85) | Moderateb | 1.68 (1.08-2.61) | High | Parenteral anticoagulants increase the risk of recurrent VTE compared to apixaban. |
| Edoxaban vs apixaban | Not applicable | Not applicable | 1.22 (0.68-2.18) | High | 1.22 (0.68-2.18) | Very lowd | There is an uncertain effect of edoxaban compared to apixaban on VTE. |
| Dabigatran vs apixaban | Not applicable | Not applicable | 1.70 (0.48-5.99) | Moderateb | 1.70 (0.48-5.99) | Very lowd | There is an uncertain effect of dabigatran compared to apixaban on recurrent VTE. |
| Major bleeding | |||||||
| VKA vs apixaban | 2.25 (0.39-13.09) | High | 1.68 (0.73-3.85) | High | 1.77 (0.84-3.75) | Very lowd | There is an uncertain effect of VKA compared to apixaban on major bleeding. |
| Rivaroxaban vs apixaban | Not applicable | Not applicable | 1.22 (0.47-3.19) | High | 1.22 (0.47-3.19) | Very lowd | There is an uncertain effect of rivaroxaban compared to apixaban on major bleeding. |
| Parenteral vs apixaban | 1.44 (0.70-2.96) | High | 1.93 (0.32-11.73) | High | 1.50 (0.77-2.92) | Very lowd | There is an uncertain effect of parenteral anticoagulants compared to apixaban on major bleeding. |
| Edoxaban vs apixaban | Not applicable | Not applicable | 2.66 (1.07-6.58) | High | 2.66 (1.07-6.58) | Very lowd | There is an uncertain effect of edoxaban compared to apixaban on major bleeding. |
| Dabigatran vs apixaban | Not applicable | Not applicable | 0.76 (0.12-4.99) | High | 0.76 (0.12-4.99) | Very lowd | There is an uncertain effect of dabigatran compared to apixaban on major bleeding. |
GRADE = Grading of Recommendations, Assessment, Development, and Evaluations; NMA = network meta-analysis; VKA = vitamin K antagonist; VTE = venous thromboembolism.
Does not incorporate imprecision (only assessed at the network meta-analysis level).
Certainty downrated for inconsistency in comparisons making up the most informative first-order loop.
The vast majority of evidence coming from an indirect estimate, which is of moderate certainty.
Certainty downrated multiple levels because of extremely serious imprecision with very wide CIs that do not exclude the possibility of important benefit or harm.
Footnotes
Dr Khan has received financial support through the Banting Postdoctoral Fellowship from the Canadian Institutes of Health Research. Dr Carrier has received grants from BMS, Leo Pharma, and Pfizer; and has received personal fees from BMS, Leo Pharma, Bayer, Pfizer, Servier, and Sanofi. Dr Rochwerg has reported that he has no relationships relevant to the contents of this paper to disclose.
The authors attest they are in compliance with human studies committees and animal welfare regulations of the authors’ institutions and Food and Drug Administration guidelines, including patient consent where appropriate. For more information, visit the Author Center.
References
- 1.Fujisaki T., Sueta D., Yamamoto E., et al. Comparing anticoagulation strategies for venous thromboembolism associated with active cancer: a systematic review and meta-analysis. J Am Coll Cardiol CardioOnc. 2024;6(1):99–113. doi: 10.1016/j.jaccao.2023.10.009. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Page M.J., McKenzie J.E., Bossuyt P.M., et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ. 2021;372 doi: 10.1136/bmj.n71. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Puhan M.A., Schünemann H.J., Murad M.H., et al. A GRADE Working Group approach for rating the quality of treatment effect estimates from network meta-analysis. BMJ. 2014;349 doi: 10.1136/bmj.g5630. [DOI] [PubMed] [Google Scholar]