The Camargo team got its start in the 505(b)(2) process at Duramed Pharmaceuticals with the 1990’s development and approval of Cenestin (synthetic conjugated estrogens, A). The product was approved based on a Phase 3 clinical study demonstrating the treatment of moderate to severe vasomotor symptoms (VMS) due to menopause – (known as ‘hot flashes’). For decades, estrogens have been the standard for treating VMS, but the potential side effects have concerned both doctors and patients – a non-hormonal product is needed.
The FDA’s Reproductive Health Drugs Advisory Committee met on March 4, 2013 to discuss two proposed non-steroid drugs indicated for the treatment of moderate to severe vasomotor symptoms. In the morning session the Committee discussed Depomed’s Gabapentin 600 Mg Tablets (FDA Brief, Depomed Brief and Errata). In the afternoon session, the same Committee discussed Noven’s Paroxetine Mesylate 7.5 Mg Capsules (FDA Brief and Errata, Noven Brief).
The FDA has a guidance with well-defined criteria for the study design and the required acceptance criteria. Basically, postmenopausal women presenting with moderate to severe hot flashes (>7/day or >50/week) are treated for 12 weeks with placebo or the test article. Reduction frequency and severity in hot flashes at 4 and 12 weeks versus placebo is required.
Both of the products presented to the Advisory Committee failed to meet one or more of the efficacy endpoints; the drugs were not significantly better than placebo. Yet, both drugs are currently used off-label for the treatment of VMS, an alternate to estrogen treatment is compelling and soy and other non-drug treatments have no supporting evidence of efficacy, so the FDA brought the products before the Advisory Committee to determine the risk/benefit.
For both drugs, the sponsors cited the issue of high placebo response as a reason for failure to meet endpoints. Our team at Duramed, in conjunction with the Cenestin investigators, was also concerned about the placebo response when designing our study. In a post-study evaluation, we examined the frequency and severity response to treatment and placebo and showed that the placebo lowered the frequency but not the severity of hot flashes, concluding that the placebo had no treatment effect.
For the gabapentin studies, the pooled age at menopause (see graph below) indicated a relatively high number of patients who were younger than the usual age of postmenopause. We had previously published that perimenopausal women give a higher placebo response than confirmed postmenpausal women.
The placebo response in the clinical studies for the two subject drug products is similar to the placebo response seen in the Cenestin study. Thus, for gabapentin’s studies, the placebo effect ranged from 47-61%, while for the placebo effect in the Cenestin study, it ranged from 48-56% (see figures below).
So, perhaps it’s not so much the placebo response, but the low drug response. Clearly, the conjugated estrogens had a large difference between drug and placebo response. In contrast, both gabapentin and paroxetine had less than a 2 hot flash/day reduction compared to placebo – a hurdle of clinical signficance.
Hence, the Advisory Committee voted 13-1 that there wasn’t sufficient evidence of effectiveness for gabapentin and split 7-7 for paroxetine. On the question of benefit/risk, members voted 12-2 and 10-4 respectively that gabapentin and paroxetine should not be approved.