In a previous posting, I provided background on KV’s Makena (17a-hydroprogesterone caproate injection aka 17P). The development and regulatory history contains many lessons. In this posting I’d like to examine the accelerated approval process.
Makena was approved under 505(b)(2) as seen from the approval letter (at this writing the approval documents are not posted at Drugs@FDA). Furthermore, it was granted Accelerated Approval under the provisions of 21 CFR 314.510, Subpart H. The essence of this regulation is that it allows FDA to approve a new drug for an unmet, life-threatening condition based on surrogate endpoints that can be reasonably expected to show clinical endpoints in required post-approval clinical studies. The regulation was originally developed as a way to get approval for HIV-related drugs where the time between surrogate endpoints and clinical endpoints was years and the drugs were widely believed to work.
In the case of Makena, the sponsor had to make a case for the unmet need and convince the FDA that the use of the surrogate endpoint was likely to be confirmed as a clinical endpoint. Makena was developed to reduce the risk of preterm birth in women with a singleton pregnancy who have a history of singleton spontaneous preterm birth. The surrogate endpoint is that treatment with Makena reduces the likelihood of the patient from giving birth before the natural term; reduction in preterm births. The clinical endpoint is neonatal morbidity and mortality (survival). The surrogate endpoint was the focus of the single clinical study submitted in the original NDA by the then-sponsor, Adeza Biomedical.
It is unclear if Adeza provided a rationale for the unmet need since the FDA Medical Reviewer doesn’t refer to any sponsor comments. Instead, the Medical Officer (MO) provided the rationale. Until the approval of Makena, there was no approved drug product to reduce the risk of preterm labor. 17P had been used for many years via pharmacist compounding, but FDA has historically believed that compounding sterile IV products (such as Makena) is a high risk. In her final review , the Medical Officer makes a strong case (see page 30-31) for the need for 17P “(published) studies demonstrate that there is greater morbidity than previously thought in these offspring (from preterm births). Increased neonatal morbidities include respiratory, metabolic, infectious, and neurologic disease. Childhood morbidities include neurologic, cognitive and behavioral delays.”
From FDA’s perspective, the unmet need was clear and caused little internal debate. The real challenge was the single trial and the use of a surrogate endpoint. In the MO’s review, she makes it clear there was a strong belief that the surrogate endpoint was related to clinical efficacy, but the issue is that “There was no statistically significant improvement in neonatal mortality or morbidity.” (page 24).
The statistical reviewer was adamant that the original NDA didn’t pass the test for subpart H concluding in her review (page 9) that the “single study was not sufficient to support the effectiveness of 17P in preventing preterm deliveries“. The statistical review is quite instructive on the issues surrounding subpart H but weighs statistical weight more heavily than clinical. This issue was such that the Division decided to present it at an Advisory Committee for Reproductive Health Drugs.
The Advisory Committee was asked to answer six questions (Committee vote in brackets, 17P is referred to as 17OHP-C):
- Is the primary endpoint of Study 17P-CT-002 – prevention of preterm birth prior to 37 weeks gestation – an adequate surrogate for a reduction in fetal and neonatal mortality or morbidity? [Yes=5, No=16]
a. If not, would prevention of preterm birth prior to prior to 35 weeks gestation be an adequate surrogate? [Yes=13, No=8]
- Do the differences in the incidence of preterm birth in Study 17P-CT-002 prior to 37 weeks in the vehicle (control) group (55%) compared to those in the control arms of (a) another Maternal Fetal Medicine Units Network trial (approximately 37%) and (b) Study 17P-IF-001 (36%) evaluating similar high risk populations indicate the need to replicate the findings of Study 17P-CT-002 in a confirmatory trial? [ Yes=9, No=12]
- Do the data reviewed by the Committee provide substantial evidence that 17OHP-C prevents preterm birth prior to 35 weeks gestational age? [Yes=12, No=9]
- a. Do the data reviewed by the Committee provide substantial evidence that 17OHP-C prevents preterm birth prior to 32 weeks gestational age? [Yes=7, No=14]
b. Do the data reviewed by the Committee provide substantial evidence that 17OHP-C reduces fetal and neonatal mortality or morbidity? [Yes=2, No=19]
- Is further study needed to evaluate the potential association of 17OHP-C with increased risk of second trimester miscarriage and stillbirth? [Yes=21, No=0]
- Are the overall safety data obtained in Studies 17P-CT-002 and 17P-IF-001 and Study 17P-FU (long term follow-up) adequate and sufficiently reassuring to support marketing approval of 17OHP-C without the need for additional preapproval safety data? [Yes=13, No=8]
- If 17-hydroxyprogesterone caproate were to be approved for marketing without additional preapproval clinical studies, would you recommend that the Applicant conduct a post approval clinical trial(s) to investigate further safety or effectiveness? [Yes=21, No=0]
b. If not, would prevention of preterm birth prior to 32 weeks gestation be an adequate surrogate? [Yes=20, No=1]
a. If so, should this information be obtained prior to approval for marketing or post-approval? [Pre-Approval=8, Post-Approval=13]
a. If so, what would be the primary objective of the trial(s) (i.e., what unanswered question(s) would the study investigate)? [Numerous suggestions were made with no consensus]
The Medical Officer (review, pages 29-30) found that the Committee recommendation was in line with the Agency thinking. Adeza submitted the results of the follow-up study to the original study which addressed the outcomes of the children born. Unfortunately, the statistical power was too low to arrive at a firm conclusion. Thus, the reviewers focused on the design of a post-approval confirmatory trial – one that would be a condition of approval of the drug. With that behind them, the FDA approved the product under subpart H as a 505(b)(2) NDA.
An interesting issue arises in a subpart H development. Subpart H requires that the surrogate endpoint be confirmed by a post-approval clinical study. But this can give rise to an ethical question: can a phase 4 clinical study using a placebo-controlled design be ethically conducted if the study drug has already been approved? For Makena, this was a very real concern. After the NDA for Makena was submitted and after the Advisory Committee concluded that that FDA should approve the drug, the American College of Obstetrics and Gynecology (ACOG) issued a guidance stating, in part, “Progesterone supplementation for the prevention of recurrent preterm birth should be offered to women with a singleton pregnancy and a prior spontaneous preterm birth due to spontaneous preterm labor or premature rupture of membranes.” The MO (see page 17 of her review) raised her “concern that successful completion of the placebo-controlled study that was proposed was not likely to be feasible if the trial is conducted primarily in the U.S. I believed that the ACOG opinion virtually established offering treatment with progesterone to such high-risk patients as a de facto standard of care. Institutional Review Boards (IRBs) and patients might interpret the ACOG committee opinion as indicating that any remaining questions regarding the efficacy and safety of hydroxyprogesterone caproate are not sufficient to justify conducting a placebo-controlled study.” It turned out that the study was able to be conducted.