Troy Brown, RN
October 31, 2019
The US Food and Drug Administration's (FDA's) Bone, Reproductive and Urologic Drugs Advisory Committee voted overwhelmingly on Wednesday to recommend levonorgestrel and ethinyl estradiol transdermal system (AG200-15; Twirla, Agile Therapeutics) for the prevention of pregnancy — despite significant concerns noted by the FDA in its pre-meeting briefing document.
The panel overwhelmingly decided (14 yes and 1 no, with 1 abstention) that the product's efficacy outweighs the risks for arterial thrombotic events and venous thromboembolism (VTE).
AG200-15 is a combined hormonal contraceptive (CHC) in the form of a transdermal patch that is applied to the abdomen, upper torso, or buttock once weekly for 3 weeks with 1 week off. It delivers 120 mcg of levonorgestrel (LNG) and 30 mcg of ethinyl estradiol (EE) per day.
Third Time a Charm?
This is the third new drug application review cycle by the FDA for AG200-15. The same EE/LNG formula was used in each of the three phase 3 clinical trials (ATI-CL12 and ATI-CL13 from the first review cycle; ATI-CL23 from the second review cycle) and the in-vivo adhesion trial included in the current review cycle.
The FDA voiced significant concerns in its briefing document about the balance of efficacy against safety and the company's proposed "limitation of use" (LOU) in the label regarding the potential for decreased efficacy among women who are overweight or obese.
"If the drug is approved ... I hope [future discussions between the sponsor and the FDA] include some discussion that in fact from an efficacy standpoint, it's not just women with BMIs over 50 where there's some question [but rather] it's all overweight women," voting committee member Douglas C. Bauer, MD, professor of medicine and epidemiology & biostatistics, University of California, San Francisco, said about his "yes" vote.
He added that he hoped "there could be some nuance on the label" that says efficacy may be reduced in overweight women, particularly those with a BMI higher than 30 km/m2."
More than half of women in one study reported unscheduled vaginal bleeding and almost half said they had difficulty applying the patch. Several panel members mentioned the fact that the patches come in boxes of three and provide enough patches for one cycle only, with no replacement patches if one comes off or is damaged in some way.
Although pharmaceutical companies have lowered the hormonal amounts and developed more convenient dosing regimens over time, CHCs continue to be associated with an increased risk for "rare but serious side effects such as arterial thrombotic events (eg, myocardial infarction and stroke) and venous thromboembolism (eg, deep vein thrombosis and pulmonary embolism)," according to the FDA's briefing document.
The safety profile of AG200-15 appears to be generally comparable to other approved CHCs; however, there was a safety signal for potentially fatal VTE, according to FDA documents. Four women (out of over 1700) in Study 23 developed five drug-related VTEs deemed to be treatment-related, three of which were pulmonary emboli.
The company asserts that AG200-15 is a low-dose estrogen-containing CHC, but the FDA does not consider it to be low-dose and believes the LNG component conveys no specific advantage compared with other approved CHC products. Whereas the company says AG200-15 represents an unmet need in contraceptives, the FDA in its briefing document states that it does not believe it does and says that adequate therapy alternatives are available.
Several panel members said they want patients to have more choices when it comes to contraception. Many women have trouble with at least one contraceptive method, including long-acting reversible contraceptives, because of side effects such as bleeding or because their lifestyles make it difficult to take a daily pill. Other women may be reluctant to use a method that requires vaginal insertion.
Of the women who provided electronic diary information about vaginal bleeding and product application during cycles 1 and 13, 60% and 41%, respectively, reported unscheduled vaginal bleeding or spotting that ranged from inconvenient to severe enough that they stopped using the patch.
"One of the most bothersome symptoms to women is lack of predictability of their cycle ... This event is concerning to the [FDA's] division in the context of benefit-risk," Nneka McNeal-Jackson, MD, an obstetrician-gynecologist and clinical reviewer with the FDA's Center for Drug Evaluation and Research, said during a presentation at the meeting.
One speaker at the open public hearing who worked in a clinical trial for the product said some women preferred it over other patches on the market because of its round shape.
Concerns About Efficacy
Committee members at the meeting were primarily concerned with data from Study ATI-CL23 (Study 23) — a United States-only, multicenter, phase 3, open-label, single-arm clinical trial examining the product's effectiveness, safety, cycle control, treatment compliance, and skin adhesion. The study did not exclude women on the basis of body mass index (BMI).
Study 23 enrolled 2032 women at 102 investigational sites, most of whom were aged 18 to 35 years. A total of 1736 women contributed 15,165 cycles that provided enough information to allow for analysis.
The prespecified primary efficacy outcome was pregnancy rate according to the Pearl Index (PI), which describes the pregnancy rate per 100 women-years of exposure to the drug, without respect to BMI. PIs were higher than the FDA's acceptable value of 5 in all race subgroups, even the non-obese participants. PIs were twice as high in women with higher BMI or weight.
Pre-specified secondary outcomes were PIs according to BMI (< 30 kg/m2 vs ≥ 30 kg/m2), self-identified race (white vs black vs other), and ethnicity (Hispanic or Latino vs Not Hispanic or Latino).
More than half (51.3%) of the women discontinued the study; the most common reasons were subject's decision (310 women; 15.3%), loss to follow-up (229 women; 11.3%), and adverse event (222 women; 10.9%).
Study 23 was designed to demonstrate a PI no larger than 3.5 with an upper bound of the corresponding two-sided 95% confidence interval (CI) not exceeding 5. In addition, the 10,000 cycles of exposure required by FDA is usually adequate to rule out an upper bound (95% CI) of 5 to evaluate a PI of 3.5 or less in a clinical trial.
For the primary endpoint, the PI for women aged 35 years or younger was 5.83 (95% CI, 4.45 - 7.21). The lowest PIs were seen among women whose BMI was < 30 kg/m2 (4.34; 95% CI, 2.86 - 5.82), whose race was "other" (4.55; 95% CI, 0.57 - 8.52), and whose weight was below 92 kg (202 lb) (4.87; 95% CI, 3.47 - 6.28).
PIs doubled for women whose BMI was ≥ 30 kg/m2 (8.64; 95% CI, 5.79 - 11.50) and whose weight was ≥ 92 kg (9.90; 5.78 - 14.02).
PIs exceeded 5 for women who were white (5.82; 95% CI, 4.14 - 7.49), black (6.40; 95% CI, 3.36 - 9.43), Hispanic or Latino (5.47; 95% CI, 2.38 - 8.56).
"It is concerning to the [FDA's] division what's noted here with the upper bound in obese women of 11.4. It is concerning to the division in the non-obese population that the upper bound is 5.8. This, in the context of a known VTE safety signal for this product, is concerning," McNeal-Jackson explained.
The FDA noted in its briefing document that although it has approved hormonal contraceptives and CHC products with overall PIs above 2.0 during the last 10 years, the upper bound of those 95% CIs has not exceeded 5.
"I voted 'yes' but it was a conditional 'yes', because if the LOU [limitation of use] was not in place I would not [have]," voting committee member Pamela Shaw, PhD, associate professor with the department of biostatistics and epidemiology at the University of Pennsylvania School of Medicine, Philadelphia, said.
"There [are] more questions about the over-30 BMI group in terms of whether or not there is increased risk and how much, and the efficacy is underwhelming," she added.
The FDA is expected to make a decision on the new drug application by November 16.