Survival Outcomes for US and Canadian Patients Diagnosed with Hodgkin Lymphoma before and after Brentuximab Vedotin Approval for Relapsed/Refractory Disease: A Retrospective Cohort Study

Response to Reviewer 1 Comments

1. Summary

Thank you very much for taking the time to review this manuscript. Please find the detailed responses below and the corresponding revisions/corrections highlighted/in track changes in the re-submitted files.

2. Questions for General Evaluation

Reviewer’s Evaluation

Response and Revisions

Does the introduction provide sufficient background and include all relevant references?

Must be improved

Please see below for full details.

Is the research design appropriate?

Must be improved

Are the methods adequately described?

Yes

Are the results clearly presented?

Yes

Are the conclusions supported by the results?

Yes

 3. Point-by-point response to Comments and Suggestions for Authors

Comments:

This is a registry-based epidemiological study of outcomes of HL patients diagnosed during defined time periods (2007-2010 or 2011-2014) in USA and Canada and included in the SEER registry (for USA) and the National Cancer Registry (for Canada). The data show that the outcome of both cohorts improved, but Canadian patients fared better in both time periods analyzed. This was due to poor outcomes of uninsured and Medicaid insured US patients. The latter group improved in the second period, but still fared worse than Canadian or privately insured US patients. 

As the authors themselves state in the discussion, none of these results can be traced back to the approval and use of brentuximab vedotin (BV), which anyway would be administered only to a small fraction of HL patients. Also, US patients diagnosed in 2010 could still have received the drug if they relapsed in 2011 or later and the same is true for Canadian patients diagnosed prior to, but relapsing after 2014, which makes it impossible to reach any conclusions on the merit of this drug using such study design. 

Therefore, the title of the paper is completely misleading and should be changed. Also, the abstract should be rewritten reducing the emphasis on BV approval and use.  

Response 1: Thank you for sharing these suggestions, and we appreciate the opportunity to improve our manuscript. We agree with the challenges you have identified in your review, however given there are no registries that capture date of relapse or relapse therapy (even the National Cancer Data Base, the most robust U.S. cancer database capturing almost all cancer cases), this led to the development of our surrogate endpoint. As described in our methods section, 72% of relapses occur within 2 years of diagnosis, thus we would expect the majority of patients who relapsed within the U.S. during period 1 to not have access, and to likely have access to Brentuximab Vedotin in the latter time period if privately insured (variable for Medicaid and unlikely for uninsured patients). This would not be the case at all for Canadian patients, appreciating your comments of the patients diagnosed in the latter half of period 2 who may have relapsed past 2014 and been eligible to receive BV. Fortunately, care for patients with Hodgkin Lymphoma was otherwise very stable (no other therapies were approved for decades before, just perhaps slight improvements in supportive care), so as far as holding all other variables stable, we can be relatively confident. Canadian private insurance would be very unlikely to cover an expensive therapy such as Brentuximab Vedotin, and the number of individuals who would have been eligible and able to enroll on a clinical trial at an academic centre would have been quite small owing to Canada’s research infrastructure.

Respectfully, regarding our title, we have discussed and feel this is the most informative title given the research question, without overstating the results given the limitations, as you have expressed concern with. Given your concerns we have adapted our abstract to include the comment that approval date was used as a surrogate as therapy was unavailable in registries to prevent the misunderstanding that BV usage was directly recorded and compared (page 1, paragraph 1, line 15-29).

Additionally, the other Reviewers and yourself have identified further limitations in interpretation of our results, and thus we have modified our limitations section to further clarify and expand upon the variables that are absent (page 9, paragraph 3, lines 267-270).

Response to Reviewer 2 Comments

1. Summary

Thank you very much for taking the time to review this manuscript. Please find the detailed responses below and the corresponding revisions/corrections highlighted/in track changes in the re-submitted files.

2. Questions for General Evaluation

Reviewer’s Evaluation

Response and Revisions

Does the introduction provide sufficient background and include all relevant references?

Yes

See below.

Is the research design appropriate?

Yes

Are the methods adequately described?

Yes

Are the results clearly presented?

Yes

Are the conclusions supported by the results?

Yes

 3. Point-by-point response to Comments and Suggestions for Authors

Comments:

This is a retrospective study based on 16,213 patients' data derived from 2 cancer registries, the US SEER and the Canadian cancer registry. The study compares survival between two periods, before (period 1) and after (period 2) the FDA approval of brentuximab vedotin (BV), and by insurance status. The study is well written and no methodological error in statistical analyzes is evident.  The most important finding is the worse survival in uninsured patients in period 2. This observation confirms the importance of ensuring equity in access to care in order to improve overall survival in HL patients. The second important finding is the improvement in survival in period 2 in the Canadian patient population despite the delay in approval and funding of BV. Unfortunately, as already recognised by the authors, due to limited data captured on the registries, no association could be evaluated according to stage, risk factors, comorbidities, chemotherapy oval regimen, interim (PET-2) and end of treatment response as well as cause of death. Therefore, further studies are needed to conduct cost-benefit analyses.  

Response 1: Thank you for your comments which we agree with. Certainly it was challenging to decide on an appropriate surrogate given a lack of data for treatment at relapse. Even the National Cancer Database does not capture therapy at relapse, and databases collecting this data are sorely needed and potentially possible given the overall improvement in EMRs nationwide. You raise several other good limitations, and as such we have included these in our discussion (page 8, paragraph 3, line 230-241) and limitations section (page 9, paragraph 3, line 268-271).

Response to Reviewer 3 Comments

1. Summary

Thank you very much for taking the time to review this manuscript. Please find the detailed responses below and the corresponding revisions/corrections highlighted/in track changes in the re-submitted files.

2. Questions for General Evaluation

Reviewer’s Evaluation

Response and Revisions

Does the introduction provide sufficient background and include all relevant references?

Yes

See below.

Is the research design appropriate?

Yes

Are the methods adequately described?

Yes

Are the results clearly presented?

Yes

Are the conclusions supported by the results?

Yes

 3. Point-by-point response to Comments and Suggestions for Authors

Comments:

This article deals with survival outcome for Hodgkin lymphoma in US and Canada, by comparing different group of years according to Brentuximab Vedotin access.

The paper is well written, is based on registry data and on a large population of patients, which are positive elements. The conclusions are supported by described data and support the hypothesis of a worst outcome for patients without an assurance, compared to medicaid and those with a private insurance.

However, we miss the data on Brentuximab Vedotin employment in both cohorts, so we don't know exactly wether survival differences are related to the availability of the drug. It is remarkable that in the group of patients with a private insurance (those who would probably obtain easily BV if needed), there's essentially no modification in survival (92% vs 93%). 

I have no specific corrections to suggest, seeing that the two main limitations according to me (almost no data about disease stage for the Canadian cohort and no data about BV utilisation) could probably not be solved due to lack of data.