Changes in the Use of Hydroxyprogesterone Caproate Injection After Confirmatory Trial Failure

Abstract

This cross-sectional study examines changes in hydroxyprogesterone caproate fills from 2010 through 2020.

The US Food and Drug Administration (FDA) accelerated approval pathway allows new drugs with uncertain clinical benefits to be approved on the basis of clinical trials involving surrogate end points. For state Medicaid programs, which must cover nearly all drugs, including those receiving accelerated approval, stakeholders are concerned about spending on accelerated approval products that remain on the market despite failing to complete FDA-required follow-on clinical trials or where those trials fail to confirm clinical benefits.

Hydroxyprogesterone caproate injection (Makena) illustrates the challenges that accelerated approval may pose for Medicaid. Makena was approved by the FDA in 2011 based on its ability to reduce recurrent preterm births and required postapproval trials confirming clinical benefits for neonatal outcomes. Yet a decade later, the use of and appropriate patient selection for progesterone supplementation, including Makena, remains controversial.¹ How Makena’s use in Medicaid—the largest payer nationwide for preterm births²—has changed with the evolving evidence base is unknown.

Methods

Using national Medicaid drug utilization and spending data, we examined changes in hydroxyprogesterone caproate fills from January 1, 2010, through September 30, 2020. We examined fills for both generic and brand-name hydroxyprogesterone caproate, including powder, injectable, and autoinjector forms. Our primary analysis includes quarterly fill counts. We also assessed fills per 100 000 Medicaid enrollees, which were similar to the primary analysis (not shown).

We mapped these data to important events in Makena’s history (Table). The FDA granted accelerated approval to Makena in 2011. In February 2018, Makena’s manufacturer introduced a new branded autoinjector before generic entry in 2018. Makena’s confirmatory trial, reported in 2019, failed to replicate the initial trial’s reduction in preterm births and did not confer neonatal benefits. In October 2019, a closely divided FDA advisory committee recommended that the FDA withdraw Makena’s approval: 9 members voted for withdrawal, but 7 voted against, citing differences between the trial that supported initial approval and the confirmatory trial. The FDA formally proposed withdrawing Makena in October 2020, but the manufacturer disagreed and requested a hearing.³ The FDA granted the hearing request in August 2021, but no verdict has been reached to date.⁴

Table. Key Events in Makena’s Regulatory History.

Date	Key Event
February 2011	Makena receives accelerated approval on the basis of a clinical trial demonstrating reductions in the recurrence of preterm births. The FDA requires postapproval trials confirming clinical benefits for neonatal outcomes, requesting final submission by October 2018.
February 2018	The FDA approves an autoinjector form of Makena.
June 2018	First generic version of Makena (injectable form) receives FDA approval. Additional generic approvals follow in 2018 and 2019.
March 2019	Makena’s manufacturer announces that the drug’s confirmatory trial failed to replicate the initial trial’s reduction in recurrent preterm births and did not confirm clinical benefits for neonatal outcomes.
October 2019	The FDA advisory committee votes narrowly (9-7) to withdraw Makena’s approval, on the basis of its failed confirmatory trial. Confirmatory clinical trial results are published online.
October 2020	The FDA Center for Drug Evaluation and Research proposes to withdraw Makena’s approval. In response, Makena’s manufacturer requests a hearing.
August 2021	The FDA grants the request for a hearing. Proceedings remain ongoing.

Abbreviation: FDA, US Food and Drug Administration.

This cross-sectional study used deidentified, publicly available data and was exempt from institutional review board review. It conforms to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guidelines.

Results

Hydroxyprogesterone caproate use increased from less than 11 000 fills/quarter between 2010 and 2014 to more than 30 000 fills/quarter in quarter 1 of 2019 (Figure). In quarter 2 of 2019, following the initial announcement of Makena’s failed postapproval trial, we observed no declines in use. Between quarter 3 and quarter 4 of 2019, use declined more than 8000 fills/quarter, coincident with trial publication and the advisory committee’s recommendation to withdraw Makena from the market. A subsequent use decline occurred in quarter 3 of 2020, but use remained high at 19 554 fills/quarter (54.9% of the maximum volume reached in quarter 2 of 2017).

Figure. Use of Brand-name and Generic Hydroxyprogesterone Caproate by Formulation.

Total fills of brand-name and generic hydroxyprogesterone caproate by formulation in each quarter between quarter 1 of 2010 and quarter 3 of 2020. Any hydroxyprogesterone caproate represents total use across all brand-formulation and generic-formulation combinations. Data were not available for quarter 4 of 2020 at the time of analysis. AA indicates accelerated approval; adcomm, advisory committee; FDA CDER, US Food and Drug Administration Center for Drug Evaluation and Research.

Since quarter 2 of 2019, brand-name Makena has maintained more than 50% market share, despite availability of generic versions since 2018. Low generic adoption may be driven by Makena’s newer branded autoinjector formulation, a form of “product hopping,” to extend Makena’s functional exclusivity period.⁵ The autoinjector has resulted in continued high spending on hydroxyprogesterone caproate, despite generic availability. In quarter 3 of 2020 alone, state Medicaid programs reimbursed (before rebates) $41 872 080 for all forms of hydroxyprogesterone caproate, 73.2% of which was for the Makena autoinjector.

Discussion

Makena provides evidence for stakeholders’ concerns about use of the accelerated approval pathway and state Medicaid program coverage requirements, though our analysis has limitations: we may be underestimating use, as our data sources do not include drugs when used in the inpatient setting, in the 340B drug pricing program, or when billed through the medical rather than the pharmacy benefit. Despite the negative trial results and FDA efforts to remove the drug from the market, use and spending continue, demonstrating a level of support for the drug among physicians, likely owing to conflicting evidence supporting the drug’s use.⁶ However, states remain concerned about spending their limited resources on products with high prices and limited clinical benefit for patients.