- TX-004HR’s 4 mcg dose may not meet the FDA’s efficacy criteria.
- The burden of proof for the long term safety of the 10 mcg dose rests with TXMD.
- Based on available information, it appears the NDA has a good chance of approval.
May 2017: TherapeuticsMD (TXMD) announced that the FDA had issued a Complete Response Letter (CRL) for TX-004HR (brand name: Yuvvexy; later switched to Imvexxy).
TX-004HR is a local (vaginal) softgel capsule that contains 17β-estradiol (E2). It’s indicated for the treatment of vulvar and vaginal atrophy (VVA) in postmenopausal women.
Shortly after the CRL, TXMD issued a press release and an 8-K to affirm that the only approvability-related concern in the CRL was a lack of long-term endometrial safety data beyond the pivotal 12 week trial; likely because TX-004HR is a novel dosage form.
In the CRL, the only approvability concern raised by the FDA was the lack of long-term endometrial safety data for TX-004HR beyond the 12 weeks studied in the pivotal phase 3 Rejoice Trial. No cases of endometrial hyperplasia were observed in the Rejoice Trial at the end of week 12 for all the doses studied and included in the NDA.
However their 10-K statement on the point was broader, and mentioned the need for additional (unqualified) long-term safety data.
In the CRL, the only approvability concern raised by the FDA was the lack of long-term safety data for TX-004HR beyond the 12 weeks studied in the phase 3 REJOICE Trial.
This could’ve been because the press release and 8-K were put forth shortly after the CRL, while the 10-K was issued after numerous discussions with the FDA post-CRL. Discussions that appear to have raised additional long-term safety concerns.
DISCUSSIONS WITH THE FDA
As a consequence of those discussions, TXMD submitted data to make the case that:
- TX-004HR was safe in the long-run.
- They didn’t need to conduct an additional pre-approval safety study.
- Any long-term safety concerns could be allayed with a phase IV, observational study instead.
Upon initial review of TXMD’s supporting evidence for this, the FDA agreed and asked TXMD to resubmit the NDA.
TXMD resubmitted the NDA alongside an observational study it (in essence) co-authored about the long term safety of local VVA treatments. A systematic review of the known literature on the point was put forth as well.
Partially as a result of the volume of data they submitted on the point, this was a class 2 resubmission with a PDUFA target action date of 05/29/2018.
WHAT IS THE FDA CONCERNED ABOUT?
What safety issues is the FDA potentially worried about?
For estrogen-containing products indicated for the treatment of VVA, the long-term effects the FDA is most concerned about include:
- Endometrial—especially atypical—hyperplasia
- Endometrial polyps
- Uterine neoplasia
- Breast cancer
- Cardiovascular events, including: myocardial infarction, stroke, DVT, and pulmonary embolism.
WHAT IS TXMD CONCERNED ABOUT?
Like the FDA, TXMD has worries of its own with respect to its NDA.
The original NDA was for three different strengths: 4 mcg, 10 mcg, and 25 mcg.
The resubmission did not include the 25 mcg dose, limiting sales of the product. There is likely more than one reason for why the sponsor got rid of this dose:
- Some of the PK data on the dose pointed to the possibility of an increased risk of long-term safety issues. This means this dose would’ve faced an even bigger hurdle for approval.
- The FDA always prefers the lowest effective dose. There is no good justification for the 25 mcg dose when 10 mcg appears to work just as well.
- As a result of the latter (and safety/marketing incentives in general), Vagifem moved to get rid of its 25 mcg dose and switch to 10 mcg in mid-2010. Vagifem (a vaginal insert) is TX-004HR’s closest direct competitor/comparison in terms of active ingredient, strength, indication, route of administration, and treatment regimen.
Thus, it appears TXMD made the right decision to move away from the 25 mcg dose based on industry best practices and their own data.
THE 4 MCG DOSE
Among the remaining strengths up for approval, it’s the 4 mcg dose that actually faces the highest risk of rejection.
This is because the FDA now prefers to use an alpha value of 0.01 when judging the data from a single pivotal 12-week trial for this class of drugs.
TXMD’s phase III study used an alpha value of 0.05; and almost all the p-values on all mandated co-primary endpoints were much lower than 0.01.
The one caveat? The 4 mcg dose’s p-value for one co-primary endpoint was 0.0149. So it “passed” TXMD’s criteria for statistical significance but not the FDA’s.
THE 10 MCG DOSE
Theoretically this means the FDA can reject the 4 mcg dose, leaving TXMD with only the 10 mcg dose. This might seriously limit the drug’s revenue potential, namely because it won’t be able to claim to have a very low, yet effective, dose out on the market.
Even then, the 10 mcg’s chances of approval rest on the literature/data TXMD submitted to make the case for its long-term safety.
Consequently, we address the following points below in order:
- Is the 4 mcg dose really going to be tabled?
- Does the published literature and other data support the long-term safety and approval of the 10 mcg dose?
IS THE 4 MCG DOSE REALLY GOING TO BE REJECTED?
To fully understand the issues with respect to the 4 mcg dose, we need to take a step back.
It used to be that an alpha value of 0.05 was perfectly reasonable for a single phase III study for this class of drugs.
The FDA appears to have gotten stricter in the recent past. Perhaps that’s as a result of the fact that TX-004HR uses a novel dosage form even though the active ingredient, E2, is exactly the same as numerous other approved medications for the same indication.
But does this really mean the 4 mcg dose is going to be rejected under the new p-value requirements?
A LOOK AT SIMILAR NDAs
In fact, we should take a look at all of the estrogen-only medications—subject to the 2003 Guidance for VVA and vasomotor symptoms (VMS)—which have been approved thus far to get a better sense of why that is.
As a preface: we have to admit that this entire discussion may be a case of pure survivorship bias and nothing else. However, it appears that there were few (if any) estrogen-only NDAs submitted for VVA/VMS in the time since the 2003 Guidance that were not eventually approved.
With that in mind, it appears plausible that these medications may constitute a good look at the FDA’s thinking on the matter.
As such, there were six drugs that fell under this criteria: three gels, one oral tablet, one spray, and one patch.
And while only one was approved for VVA/VMS—the others were technically approved solely for VMS—they were all subject to the same Guidance and FDA review process nonetheless.
WHAT DO THE OTHER NDAs REVEAL?
After taking a close look at all of the documents pertaining to their NDAs, several reasonable and over-arching conclusions could be made:
- The FDA isn’t as strict as it appears to be with these medications. In numerous cases, the FDA bent the rules in favor of the sponsor even when the drug did not meet seemingly important transitional criteria set out in the Guidance.
So long as a statistically significant difference was found in the very end, it was good enough despite poor intermediate results.
- The FDA’s reviewers may conduct their own statistical analyses of the data, which can actually work in the sponsor’s favor in terms of finding adequate statistical significance where the sponsor found none in their original analysis.
- Only a minority of the sponsors had to conduct anything other than one, 12-week, pivotal trial to prove long-term safety. The exceptions were: a submission of long-term, right to reference, data in one NDA and a recommendation for a phase IV study in another.
And what’s happening in the case of TXMD? They’ve committed to a phase IV study as well.
As a result of all of this, it appears that the 4 mcg dose has a reasonable chance of approval.
THE FDA’s CONCLUSION
What may happen is the FDA will conclude that the p-value is significant enough at an alpha of 0.05 anyways. As a result, a lot of women may benefit from this lower, potentially safer, and technically effective dose.
Yes, it may take a longer amount of time to see the full benefits but this possibility can be reflected in product labeling.
Remember: the FDA wants to use as low an effective dose as possible.
DOES THE PUBLISHED LITERATURE SUPPORT THE APPROVAL OF THE 10 MCG DOSE?
Although this discussion will apply to the 4 mcg dose in a way, we focus on the 10 mcg dose as it probably poses a higher safety risk over the long run and, because of the last section’s points, might indeed be the only dose being considered for approval anyways.
TXMD submitted a systematic review of the published literature on the long-term safety front. They submitted a co-authored observational study as well. The FDA will review this and the FDA Adverse Event Reporting System (FAERS) in order to reach a conclusion about the likely long-term safety of this drug.
TRIAL DATA: THE GENERAL CONCLUSION
When we take a look at the published trial literature, it’s clear that low-to-moderate quality evidence supports the fact that local, estrogen-only, treatments for VVA/VMS are indeed safe with respect to all of the safety issues the FDA is most concerned about.
There’s one caveat. Most of this data includes studies conducted on a wide variety of dosage forms. Furthermore, the majority of the trials were relatively short-term studies (3 months) while the FDA is concerned about long-term safety.
When looking over the nine approved VVA/VMS products that were subject to the 2003 Guidance, including three non-estrogen or non-estrogen/progestin products, something becomes immediately clear.
With rare exception, the FDA’s definition of long-term safety in these cases referred to, at minimum, a 52-week study.
TRIAL DATA: THE SPECIFIC CONCLUSION
And when we go over this subset of published literature most relevant to TX-004HR (including similar indications, dose, dosage, etc.) with respect to safety, the conclusion is mixed.
On the one hand, there isn’t much data that fits this criteria. But in the handful of studies that do fit the bill, it appears that there are generally no serious long-term safety concerns.
However, while the methods (with respect to blinding, randomization, and use of controls) are overall moderate to high in quality, there is a problem.
The problem is that most of the studies did not follow the Guidance’s instructions towards adequately assessing the potential endometrial issues that might arise. For instance, ultrasound was used to assess for endometrial changes as opposed to histopathology; or baseline biopsies were never conducted.
With respect to this point, the studies were of a generally low to moderate quality.
This doesn’t help TXMD but we must consider the fact that the FDA is not a black and white organization. It won’t reject the NDA solely because these studies didn’t follow the Guidance to a T.
The reviewers will consider all of the studies and data on the point. And there is hope yet for TXMD as a result.
First, most published long-term data (including observational data) generally concludes there is nothing significant to worry about.
The one major exception is a sizable observational Danish study that found a strong likelihood of endometrial cancer due to the use of vaginal estrogen—even when accounting for prior systemic estrogen therapy.
It’s unclear why there’s such a strong difference between this study and most others given the published data but true comparisons between such studies could not be made due to a lack of important information.
Second, if we compare the safety data between TX-004HR and the nine other NDAs in question, something else comes to light. The FDA has approved drugs that demonstrated, comparatively speaking, more serious endometrial concerns (even at 12 weeks) than TX-004HR ever did.
THE FDA ADVERSE EVENT REPORTING SYSTEM (FAERS) DATA
Third, the FDA is going to search FAERS for clues to long-term safety issues. And so, there’s an important lesson we can learn from FAERS.
In December 2009, a 10 mcg version of Vagifem was approved. In mid-2010 Vagifem discontinued its 25 mcg version.
We ignore 2010 in the analysis as it was the year when both doses were available. Thus, we can split the reported safety data in to two equal periods, the 25 mcg period of 2003-2009 and the 10 mcg period of 2011-2017.
As you can see below, the number of serious adverse events dropped by about 15%.
A 15% drop might be conservative.
That’s because it’s likely that the number of women using prescription products like Vagifem for VVA has increased over time so, all else equal, we should expect an increase in serious adverse events being reported if the 10 mcg version wasn’t a safer alternative.
Even if the number of women using Vagifem has remained steady or declined a bit, the overall number of adverse events input into FAERS (for any drug) between 2011-2017 has more than doubled since 2003-2009. This is likely due to increased access to technology and public awareness of the system as well as an increase in the number of drugs out on the market overall.
And let’s not forget one thing: many adverse effects may arise as a result of a cumulative effect. In essence, some of the adverse effects in 2011-2017 may be associated with the use of the 25 mcg dose in the prior years. A carryover effect, of sorts.
Despite all of the above, the number of adverse events specifically for Vagifem has dropped nonetheless.
REFERENCE LISTED DRUG DATA
Finally, while TXMD is depending on published literature to support TX-004HR’s approval, it’s also possible that it’s referencing a listed drug with an expired patent to support their case as well.
While referencing Vagifem would be ideal to support approval, the patent/legal issues involved make this highly unlikely.
Yet there are several other medications it could reference with expired patents that would be beneficial for approval; so long as TXMD submitted a quality scientific bridge between TX-004HR and the listed drug.
To recap the safety evidence discussion:
- Low-moderate quality literature generally supports the long-term safety of products very similar to TX-004HR.
- FAERS data, albeit confounded by numerous factors, appears to support the long term safety potential of TX-004HR at the 10 mcg dose.
- TXMD has committed to a phase IV study; there’s only a low risk a CRL will be issued due to a poorly proposed protocol.
Overall, given the newly submitted evidence not available (or not submitted) for the original NDA, it’s quite likely the 10 mcg dose will be approved from a purely biomedical perspective.
Last Updated: 5/13/2018 for elements of style. 5/30/2018 – added Imvexxy as the new brand name.
No one associated with writing this article has any financial stake in, or ties to, TXMD.