Determinants of the Cancer Drug Funding Process in Canada

Round 1

Reviewer 1 Report

 

The authors have done an analysis of determinant of the Cancer Drug Funding Process and concluded that cancer type, drug class and pCODR recommendations more than cost/list price.

 

The authors start with that pCODR decisions are not binding; however, the data suggests otherwise as per their conclusion.  It would be interesting to look at the negative pCODR decisions to see the basis of those decisions and if clinicians felt there is value – or if there is discordance with other regions.  For example INESS funds Lonsurf and regorfenib for metastatic colon cancer – both received a negative pCODR recommendation.  If the pCODR decision is non-binding – why did INESS choose to fund based on the same data?  And if non binding – why didn’t provinces choose to try and fund these indications?

 

While cancer type seemed to be determinant I suspect that this is really due to external factors.  It would be interesting to look at the clinical data itself – what is the OS,PFS, toxicity, QOL.  IS there cutoff that seems to be a threshold for funding or TTF?  These tend to be the main factors for the HTA process.  Another example with the colon area is the use of EGFR in left sided tumors – it is a used world wide – yet pCODR has declined this as well.

 

I agree that the list price likely does not drive this as the HTA dictates what the price ceiling is for pCPA and hence the negotiations.  Also hence the fund with conditional recommendations – it is dependent of the future price negotiation – so unless a list price comes in at the $50-60K QALY that is now quoted – it will always go to pCPA.  So the study shows that the list price really is meaningless.   Although I admit it is surprising that the list price did not correlate to longer TTF thinking it would take longer in pCPA.   It is curious to see some of the times – it would be helpful in Table 1 to have the TTF and perhaps even the primary endpoint data for those studies.  If there are studies with long TTF what are those issues?  It is also curious that NET had a TTF of 0 and listed so quickly.   

 

The authors also state there should be a focus on expediating funding approval process post HTA completion.  However, the TTF is from Health Canada approval to first provincial funding.  There are several components – pCODR/HTA then pCPA and then provincial process.  The authors have not broken down these timelines to show that it is the pCPA or provincial process after HTA that is the barrier the access (not saying that it isn’t the barrier but the data is not in the paper to back up the statement)

 

This paper has a great start presently.  The data from the trials should be presented to look at if that is the determinant at what is the threshold given that these should be the factors that the HTA is using.  The tumor type and drug type in the end are likely more factors that reflect this.  Otherwise from the HTA process it would be incorrect to say that the tumor type is the determinant.  The authors do comment that this is not causal and it is an observation though.

 

In order to have this research published – I would recommend including more of the clinical data endpoints from the study, greater examination of why some drugs had such a long TTF (they explained the short TTF – or for the NET take out BC and look at the next province that funded it since it seemed BC jumped the gun) and looking into the factors of the studies that were not approved to examine the non-binding ruling which in reality seems to be binding.  

Thank you for your hard work and efforts.  There are some great thoughts that are in the works here.

Author Response

Please see attached file.

Author Response File: Author Response.pdf

Reviewer 2 Report

The data collection process should include where the date of the first public reimbursement came from. CADTH reports list when provinces decide to fund, not when the first reimbursement occurred. 

The authors state that the analysis is descriptive. Why did they not perform an additional multivariate analysis to assess which of the factors they measured were the most important? This would add much to the research. They also appear to have used simple linear correlation analysis (for example in Figure 1). They should have re-analyzed omitting the obvious outliers and also considered non-linear association.

The analysis assesses listing of drugs but makes no attempt to evaluate access criteria. A drug may be listed but the criteria to access can be so restrictive that few patients qualify for access. 

Finally, the Discussion is extensive and much of it speculates on factors not identified in the research. The authors identify no impact of price but spend much discussion suggesting it is a factor. A more comprehensive analysis may have enabled them to say more on the issue based on their results rather those of other authors. Similarly, their discussion of the impact of tumor type is based more on external items than their work. 

Author Response

Please see attached file.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Thank you for efforts to address my comments – I really have this one area to discuss:

 

The authors start with that pCODR decisions are not binding; however, the data suggests otherwise as per their conclusion. It would be interesting to look at the negative pCODR decisions to see the basis of those decisions and if clinicians felt there is value – or if there is discordance with other regions. For example INESS funds Lonsurf and regorfenib for metastatic colon cancer – both received a negative pCODR recommendation. If the pCODR decision is non-binding – why did INESS choose to fund based on the same data? And if non binding – why didn’t provinces choose to try and fund these indications?

 

Thank you for this comment. While the pCODR recommendation is not binding, certainly it seems that the provinces usually do what pCODR has suggested, even though they are not obligated to (as our results show). Determining the province’s reasoning for this is beyond the scope of our analysis, and unfortunately there is no reliable data available regarding the nuances/motivations of the decision process of the provinces themselves. It does seem quite clear that the provinces generally follow the pCODR recommendation.

 

With respect to the revised paper:

 

Impact of pCODR Recommendation:

 

Line 49 – I suggest legally nonbinding – I argue that that pCODR recommendations end up being binding due to the fact that provinces can not negotiate a price for a drug without pCPA. At least this is my understanding – see below.

 

On line 251-252 – the authors state

 

Clearly the pCODR recommendation influences provincial funding decisions, particularly when negative, as no drugs in our study with a negative pCODR recommendation were funded in any province.

 

I would suggest the following be considered after this statement.

 

Currently price negotiations can not occur with pCPA with a negative pCODR recommendation.  While there may be cases where clinicians disagree with the recommendation, provinces can not pursue funding of a therapy with a negative pCODR recommendation given that there are no pCPA negotiations. 

 

There are examples of negative pCODR recommendations that clinicians would want access – immunotherapy in small cell lung cancer or first line EGFR therapy in left sided colon cancers – these have received negative pCODR recommendations and in the case of colon cancer – have written guidelines stating the benefits.

 

 

However on line 320:

 

In fact, political pressure, in the form of advertisements, or from pharmaceutical companies, has been cited as a top reason for funding drugs that received a negative pCODR recommendation [13], although this would seem to occur rarely.

 

There is some discordance with these statements as it does seem like the time periods of the studies overlap.  There are not any examples listed of this in your study and you state that negative pCODR recommendations are not funded. 

 

The other option is to remove the sentence in line 251 – 252.  I think the negative recommendation and non funding are due to many other factors that are not addressed in the study.

 

Great job – this is a timely topic.  Thank you for your hard work and analysis.  Sorry for being a bit picky about this point – I think you would receive comments about the negative pCODR recommendations – and this is not addressed in this study.  I think you could do a whole paper on negative pCODR recommendations actually.  I agree with the observation but there are a lot of factors that go into the lack of funding and the question is whether clinicians truly agree with those decisions. 

 

 

 

 

Author Response

Please see attached file.

Author Response File: Author Response.docx

Reviewer 2 Report

Good to see the multivariate results and discussion now much better.

Author Response

Please see attached file.

Author Response File: Author Response.docx