The Epidemiology of Myeloproliferative Neoplasms in New Zealand between 2010 and 2017: Insights from the New Zealand Cancer Registry

Round 1

Reviewer 1 Report

This paper describes a sample of MPN patients included in the New Zealand Cancer Registry, 2010 – 2017. The study addresses an important and potentially novel issue, which has clinical implications that may extend beyond New Zealand to other Polynesian countries. Overall, the study has value but there are several areas that need to be addressed to improve the clarity and strength of the paper.

Specific suggestions are:

 

Introduction: The scientific background and rationale for the study could be strengthened by more providing more detail. That is, more detailed reporting of findings from previous studies would add strength to the study rationale and help to emphasise the importance of the study.

 

Introduction, page 2, line 48: In the sentence “incidence and mortality risk increases with age.”, it is recommended that the authors support this sentence by providing estimates and the relevant references, which in turn would support the study rationale.

 

Introduction, page 2, lines 51-57, and lines 59 - 63: As with Line 48, lines 51 – 57 and 59 - 63 need to provide estimates to support statements of ethnicity-based differences in incidence rates.

While a research gap has been identified, the rationale for the importance of analysing this data could be emphasised with regards to highlighting the implications of study findings for improving patient/clinical and public health outcomes. The results may also have implications for recommending increased screening of extended families of PV patients perhaps for instance.

Suggest discussing the implications of the study findings in the last sentence of the introduction.

 

Introduction, page 2, lines 65 - 68: Suggest rephrasing the final paragraph of the introduction to clearly state the aims of the study and a brief outline of the study design, including describing where the data will come from (NZCR).  It would also be useful to include here any hypotheses and the implications of the study findings, as discussed above.

 

Results:

Would recommend changing the order of results to start with a description of the study participants and demographic data, with a comparison of data from different regions (Table 1) before discussing capture rates (Figure 1, Table S1).

 

This would include describing the number of participants in the cohort, characteristic and a summary of follow-up time to mortality. Also suggest discussing the data presented in Table 1 as a whole rather than interspersed. For instance,  for lines 122 – 128, would recommended making that paragraph all about results found in Table 2, and move description of crude data presented in Table 1 to a separate paragraph. This will aid clarity and assist the reader with flow of data and text.

 

Table 1: Recommend amendment of title to" Table 1: Characteristics of three Ph-negative myeloproliferative neoplasm patients captured in the New Zealand Cancer Registry, 2010 - 2017 (n = 275).

 

Table 2:

Suggest amendment to the title (omit adjusting for age, gender and ethnicity – that can be mentioned as a footnote and perhaps mention why these variables were adjusted for – eg because they are well-known confounders or significant in univariate Cox modelling?).

 

Suggest adding univariate analysis as part of main text (not supplementary)

Were any variates time varying? Perhaps mention.

 

Methods: Please include a statement related to the study design and description of the study setting – eg. retrospective cohort / population-based? Were there any sources of bias? And if so, where any efforts to address potential sources of bias made? Were any missing data identified and how were these handled?

 

Methods, lines 268 - 275: The details provided in this section are very useful and clear. Would recommend mentioning if all ages or >18 years are included in the current study.

 

Methods, lines 298: Useful description of variables included here, and important acknowledgement  that ethnicity was self-reported.

 

For statistical analysis, suggest subheadings.

 

Suggest outlining “Exposures” and “primary outcome” and “secondary outcome” in subheadings and describing in detail what they were.

 

Also subheadings for univariate and multivariate analyses.

 

Methods, page 10, lines 312:

 

Recommend the authors include a relevant reference for Cox PH regression and a reference for schoenfeld residuals to test the Cox PH assumption. Also recommending more details of variables:

- include whether age was included as a categorial or continuous variable in the adjusted model, and if continuous, was this by single year increments.

- explain why certain covariates (age, gender and ethnicity) were adjusted for? Were they significant in univariate analyses (eg p=0.15)

 

Discussion, page 7, lines 159 – 167: The authors have given a clear summary of aims and findings. However, it may be worth mentioning that the study identified no risk factors for thrombocythaemia.

Also suggest the authors emphasise that the research findings represent novel findings in the New Zealand population.

 

Discussion, page 8 page line 215: Suggest authors not just report the results already presented in the results section but perhaps reflect on why there would be differences in age of diagnosis and mortality between different Pacific Islands and Europeans or different populations within the Pacific. Could this be due to diet, lifestyle or other differences?

 

It may be beneficial to add a DAG diagram (http://www.dagitty.net/dags.html) to help illustrate which confounders may be relevant to mortality outcomes and which confounder data was available and could be adjusted for, and which variables there was no data for. In turn, this may help identify variables for future research to collect data on and identify potential areas of study bias.

 

Discussion, page 8 page line 236: Recommend modifying this sentence to reflect the importance of evidence to improve outcomes. Would suggest something similar to “Further research into genomic associations between ethnicity and MPN incidence and outcomes in the New Zealand population may be warranted as this may lead to improved clinical care for patients with MPN.” Is there also a similar genetic disorder or example that can be discussed where research into genomic associations in a certain ethnic population was found which led to novel preventative/diagnostic/therapeutic measures?

 

Discussion, page 8, line 239-240: Suggest re-wording this sentence as it is unclear what the authors mean by research being under-valued. Perhaps it would be better to say that increased focus in this area is needed/warranted and provide evidence (eg. not many prior studies, no funding in this area???) Also suggest emphasising what the potential clinical and public health benefits of this research would bring. Eg."Further research into the drivers of the ethnicity-based differences in haemotological cancer outcomes in NZ may aid ….." (this may have been address in above comment)

 

Discussion, line 242 - 256:

Suggest separating this paragraph into limitations in one paragraph and strengths in another.

With regards to study limitations, the authors should be commended for mentioning they did not have available data on confounders, but which confounders (as mentioned previously). Suggest discussing confounders adjusted for and not adjusted for by other studies.

Suggest adding short follow-up time as a limitation (and perhaps comparing follow-up times with the US study)

 

 

Minor comments

Table 1 - Otherwise, suggest only using median unless a logical reason can be given for providing both mean and median.

Discussion – Suggest adding a brief discussion for the reasons why risk factors for thrombocythaemia were not identified.

 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

An interesting and important reporting of MPN incidence in New Zealand highlighting increased incidence and death rates by ethnicity. A few points to consider:

• In the methods section there is information on the guidelines produced in 2016 but no information as to when practices in New Zealand would have adopted these guidelines. The low incidence of ET is reflective of inadequate capture of cases. A more current pulling of data is warranted to see the change in case numbers before and after implementation of reporting guidelines.
• Results section needs to highlight that there is no incidence data for PMF prior to 2014 at the start. 
• In the results section it states "Most MPN patients were European". This is inaccurate and should state European/European decent. 
• Given the age difference in MPN patients by ethnicity the statement "There was no clear difference in the survival of
  Polynesian and European patients for all cases combined" needs to be age and sex adjusted. 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf