Help w/ ? please: How do FDA and DRs handle meds which benefit subpopulation
I'm hoping my team of experts can help me think through something. In one of the previous conference calls, Vertex noted that there was a way forward for FDA approval with the 770/809 combo either if there was an overall improvement OR a subpopulation which benefitted from the combo. And Vertex noted that there is precedent with other drugs for that. Add to that the recent Chest (April 2011) article discussing VX770's impact on homozgous f-508d CFers, which noted a subpopulation appeared to benefit from 770 ALONE and that this was consistent with in vitro tests.
So my question is two-fold (with sub-parts! 1) What does/how does the FDA handle a drug where there is a subpopulation which benefits from the drug? What if there is no way to know who benefits without the individual taking the drug and seeing if it works? What are the precedents?
2) Separately, how do doctors handle off-label use where it is established that some individuals might benefit from off label use, but only a small percentage?
One particularly promising (though still expensive) way to measure treatment efficacy is via CT imaging. See the abstract of ‘Quantitative evaluation of hyperpolarized helium-3 magnetic resonance imaging of lung function variability in cystic fibrosis.’ - http://www.ncbi.nlm.nih.gov/pubmed/21536462
In answer to your questions, (1) If Vertex and the FDA can agree on the use of particular surrogate endpoints in their trials and it is determined that Vertex’s treatment is efficacious for one or more patient subpopulations using those endpoints, if/when the drug is approved the label will reflect that subpopulation information and the package insert will describe the surrogates used.
(2) If studies are subsequently performed for this treatment on other patient subpopulations that were not the focus of the clinical trial and those studies tie efficacy to surrogate endpoints used in the clinical trial (or even not used in the trial if the surrogate endpoint appears to have relevance in published studies), a prescribing doc may decide to go ‘off-label’ using one or more of those surrogate endpoints as guides in the early stages of the treatment and functional measures later.
Perhaps the simplest way to deal with the problem of not knowing which CF patient sub-groups might benefit from Kalydeco is for Vertex to provide the drug at cost (basically free) to any CF patient not included on the Kaydeco FDA label for up to 6 months. In return, Vertex could ask that certain clinical endpoints be measured by the patients physicians. If there is strong objective clinical evidence of improvement, then payers would be compelled to pay for off-label use in these patients.
I've seen the references to that Chest paper but don't have access myself. Are there any interesting quotes worth sharing?
As for your questions, here's my impression of how things could work. With the subpopulations, Vertex will have to show efficacy in a prospective clinical trial to get approval. To do that they're going to need to identify the subpopulation at the beginning of the trial. Normally it would be defined ahead of time (by CFTR expression, sweat chloride levels, whatever). But possibly they could take all comers and dose them for 8 weeks. Patients who don't respond (on whatever measure they decide) get dropped. Those who did respond are then divided into drug/placebo groups and the trial goes forward.
As to how doctors handle off-label usage, probably they just use their judgement based on: 1) how likely is it that the drug will work for this patient?, 2) how much would it help if it did work? and 3) what are the risks of harmful side effects from taking the drug off-label?
If the answer to '3' is "minimal" and the answer to '2' is "a whole lot", then '1' probably becomes irrelevant and most doctors will try it on most patients. If '2' and '3' are less obviously favorable, then the risk/benefit calculation gets more complicated. Given the gravity of CF, and the safety profile of Kalydeco (so far), I would think most doctors will be willing to try if the patient understands the situation and wants to try. I had always assumed insurers were going to be the major barrier, but it seems they're mostly going along too.
Here is the relevant language from the Chest article. This is really all there is re VX770 and f508 homozygous that I can find.
F508del-CFTR homozygotes, 15.2% of ivacaftor subjects demonstrated an average sweat chloride reduction 10 mmol/L across all time points through Week 16 compared to no subjects in the placebo group. This may indicate a subpopulation of subjects homozygous for the F508del-CFTR mutation with some F508del-CFTR that reaches the epithelial cell surface, supporting in vitro evidence. One in vitro study tested ivacaftor in human bronchial epithelial (HBE) cells from subjects with CF who were homozygous for the F508del-CFTR mutation10. Among HBE that showed potentiation by ivacaftor, the effect size (up to 16% of normal activity) was considerably less than in HBE from a G551D/F508del donor (nearly 50% normal).
The difficulty I have is my understanding is that they cannot tell what subpopulation benefits.. . . there is no identified modifying gene, pre-drug marker. . . What do you do in that situation? Would the FDA really approve a drug in that situation where the Phase 3 study is really stacked, i.e., you pick only those you know will benefit and then they benefit. . . how do you lable b/c you only know who benefits after? Could Vertex really do a study where they give free of charge to all double deltas to see who benefits? (They wouldn't need 6 months--only one month!) But that would seem illegal under FDA rules. Does anyone know more re the liver sideeffects? I've heard that they are testing those monthly for the first several months. . . I had no idea an issue BUT this might be because of CF liver disease some have and safe for those without the liver disease. I've also heard many doctors are saying absolutely "no" for off-label use--not even getting to insurance issue. That's why I'm wondering what analogous situations there are for other drugs.
Thanks all for input. Please keep coming with thoughts/answers if you can!!