As we have alluded to before, the House and Senate have been considering another tweaking of regulations related to pharma as part of the broad 21st Century Cures Act (Act). The House has taken a big step forward by approving on a bi-partisan basis the Act. Most of the headlines are focused on the increase in NIH funding, but for pharma people there is a lot of interesting suggested changes in the regulatory environment. NOTE: this is just the House version, the Senate will not take up the bill to later this year so we don’t know if any of these changes will come to pass – so far at least the White House is not threatening a veto.
You can find the whole bill here.
Medical Affairs: All about off label communication
Subsection 2102 explicitly requires the regulatory clarification about acceptable dissemination of off-label information:
Not later than 18 months after the date of enactment of this Act, the Secretary of Health and Human Services shall issue draft guidance on facilitating the responsible dissemination of truthful and non-misleading scientific and medical information not included in the approved labeling of drugs and devices.
Implication for MA:
While conceivably once this guidance is issued anyone (commercial and medical) from the company can share off-label information, if the guidelines are anything like the guidance provided by the FDA in their response to the Amarin lawsuit which we discussed here that guidance will explicitly require that the information be discussed by someone from the company with “… the appropriate background or training to accurately communicate scientific information.”
So, conservatively interpreting that definition, if this act passes MA should finally have clear guidelines on how to proactively discuss off-label information. A BIG win for MA.
Clinical Development and Medical Affairs: Support for innovation in clinical trials
There are a number of requirements that could be clarified by the Act. I will highlight a few I think are most impactful.
The first is section 2021 which discusses the use of biomarkers. It establishes a process whereby biomedical research consortia can identify and agree upon biomarkers, then submit those biomarkers to the FDA for their review and approval, and once approved can be used by industry as surrogate endpoint. Draft guidance on this process is due no later than 24 months after enactment.
The second is subsection 2061, focused on broader use of adaptive trial design and Bayesian statistics. The Act requires the FDA to update and finalize the guidance on adaptive trial design within 18 months and issue draft guidance on the use of Bayesian statistical models within 48 months. So not exactly pushing the envelope on the Bayesian timeline for draft guidance.
The next is subsection 2062 deals with the use of “real world” observational / registry / safety data in the application for additional indications or to meet post-approval requirements. It requires that a program for the use of this type of data in these ways be established within 24 months including delineating when this type of data will be acceptable to the FDA, the standards and methods needed to be followed when collecting this data. One year after the program is up and running draft guidance is due and then a year after that final guidance is due.
Implications for CD MA
The implications of the acceptance of biomarkers as acceptable surrogate endpoints should have huge implications. Currently these endpoints are gathered but are not relied upon for fear that they will not be acceptable. Knowing in advance that they will be acceptable to the FDA will allow for much more streamlined data collection efforts.
Further clarity on adaptive trial design will be helpful but not earth-shattering. Adaptive trial design draft guidance already exists so all the act is really doing to pushing to get it out of the draft phase and finalized which should reduce risk and make adaptive trial designs more acceptable to less risk-tolerant organizations. This will hopefully push into becoming a standard in P2 trials, reducing the need for multiple trials. I’m not really qualified to discuss how the Bayesian statistical model impact development – if you understand this better than me please leave a note in the comments.
The “real world” data requirement, assuming it produces clear definitions from the FDA about the qualities that the data has to have to be acceptable, could be a dual edged sword. On the one hand, if they have very high expectations about the quality and cleanliness of this data in some cases, especially registries, it could raise the costs. On the other hand, clear definition of what data will be acceptable for additional indication applications will allow for the use of this data more broadly and thus more of these studies to be conducted, which is great overall for pharma.
I am curious to know what you think. Please leave a comment here.