David Redwine, MD - Leuprolide: The ‘d’ is Silent.
Endometriosis Foundation of America
Medical Conference 2019
From Biomarkers to Precision Surgery
March 8-9, 2019 - Lenox Hill Hospital, NYC
I'm gonna talk about medical therapy of endometriosis with leuprolide and elagolix. Medical therapy of endometriosis follows some immutable principles at this point in time. No medicine eradicates the disease, medicines treat only symptoms. Surgery is the current best treatment of the disease and excision is the only form of surgery that treats both deep and superficial disease. It's the way it's always been and everything new is old again.
Sushruta, an Ayurvedic doctor from ancient India said that surgery has the advantage of having instantaneous effects, so it's the highest of the medical tantras. Hippocrates, the father of endometriosis because his was the first accurate description of the disease and the demographics of patients with the disease said that the diseases which medicines cannot cure, excision cures. These opinions are class C evidence, the opinions of experts.
Lupron was approved by the FDA in 1990 for the treatment of endometriosis. The drug was developed by a combination of Takeda and Abbott pharmaceuticals abbreviated to TAP. And there were several studies that you see here on the right looking at different drugs, different formulations in comparison with different drugs.
And the path to approval of a drug is pretty similar. You get a candidate drug out of your pipeline. You recruit collaborators who collect data. The data is sent back to the mothership, where it's tabulated, a report is written and that report becomes the basis for FDA application and future journal publications.
I had access to these studies at the level of data tabulation from the forms sent in by the researchers and I was interested in whether the broad data supports the summary reports in journal publications. Each of these studies had a similar methodology. They looked at estrogen levels, bone density, adverse effects, etc., but for the purpose of this discussion, I'm going to be looking at estrogen levels in one study and efficacy in several studies and does the raw data support what was said about it?
One of the very first studies of Lupron out of the block was this one. It was comparing Lupron given subcu for a week, followed by about six months of Lupron nasal spray. A comparison drug was danazol. One year after Lupron was stopped, here's what happened to the estrogen levels. 63% of women had not regained baseline. 50% had estrogen levels below 100 pg/mL and 1 out of 8 were menopausal. Yes. Small numbers.
One of the recurring criticisms about the early Lupron studies was only a few dozen typically were studied. These were not due to outliers. Several tests were taken so they were fairly consistent. The conclusion was that hormone profiles during the followup period were similar to baseline.
This was the journal publication that eventuated from that study. It was supported in part by a grant from Abbott Laboratories, which of course is typical. No TAP employee was an author. In this publication, there was no mention of that one-year followup estrogen level.
From the raw data, this is what the estrogen level did before and after the treatment. And this is what it looked like in the publication, in the graph that was published. Basically, the graph that we saw previously, the post-treatment estradiol level was 82.7, but in this graph, they pushed that line up to a little over 150, so it looks like they got almost 100% increase in the post-treatment estradiol level if you look at this graph.
Little things like this here and there. So the raw data does not support the graph that you saw in the literature.
Another study comparing Lupron to Synarel. One of the safety measures was vaginal bleeding after cessation of treatment, the idea being that the ovaries would wake up and people would have menstrual bleeding. The baseline estradiol level for this study was 61 pg/mL, which is very low. The normal range is 100-300 and why was it so low? Did they look at the first study and say, "Wow, recovery of estrogen doesn't seem to go real well. Maybe if we get a lower baseline estradiol level."
I'm being cynical here, but I don't know why they chose to have such a low baseline estradiol level. So that's the estrogen component of what I'm gonna say.
The efficacy component, they measured efficacy with the Biberoglu and Behrman pain scale, which is a simple, common sense three symptoms and two signs, dysmenorrhea, dyspareunia, chronic pelvic pain, pelvic tenderness and nodularity.
We discussed dysmenorrhea before. This is an early bar chart of results from laparoscopic excision of endometriosis in my hands and the red lines, the red bars indicate dysmenorrhea. It's the symptom that responds least well to excision of endometriosis. Which tends to inform me that well, maybe dysmenorrhea isn't always due to endometriosis. And this was before I started doing presacral enterectomies in the 1980s for women with dysmenorrhea.
But this was the reason why I started doing presacral enterectomies because I realized that well, God, I'm doing a pretty good job for the endometriosis symptoms, but for dysmenorrhea, I wasn't. And so presacral enterectomy can help that. So when I said the other day that dysmenorrhea may have a uterine component, I think it definitely does. Whether that's mixed in with the results and whether that matters is another question.
Here is a study where they compared Lupron and danazol and looked at pain medicines. And they found that there was no significant difference in pain medicine use over the six-month study period, either within or between the comparison groups. In other words, during Lupron therapy, there was no reduction in pain medicine requirements just like the previous study that you saw. And does this count as a success because pain medicine use did not increase? I don't know.
While on Lupron, 100% of the patients continued to require some kind of pain medicine. Almost half required narcotics after the initial symptom flare while they were on Lupron. The conclusion was that Lupron was shown to be effective in treating endometriosis pain symptoms, dysmenorrhea, pelvic pain, pelvic tenderness. Keep that in mind, pelvic tenderness. All responded significantly to Lupron compared to placebo.
This is a study that eventuated from that study. One-third of the authors were TAP employees. This article did not talk about the narcotic requirements and continuing pain. And the raw data, it was clear that there were several protocol violations, some willful, they had to be willful from what I saw, that made Lupron look better than it really was. And was the long-term improvement that was seen in some patients due to that long-term ovarian dysfunction that was identified with the first Lupron study?
At the final visit, 46% of successful completers had complete resolution of pelvic pain, but that was increased somehow magically to 55% in the publication that came out in the literature. I have no idea why it was different. Again, 1/3 of the authors were TAPP employees.
Another study, same song. The majority of patients returned to baseline symptoms quickly. Pain medicine and narcotic use is frequent during and after Lupron.
Here is another study comparing four different cocktails of Lupron plus norethindrone and/or Premarin. And what they found was that ... And they were given this for 12 months, these various things were given for 12 months. By three months after completing 12 months of Lupron, these symptoms had returned in about 50% of patients. So the half-life of pain relief after taking Lupron for one year is about two to three months.
This confirms a poor efficacy of Lupron was found in the previous studies by the sponsor.
Elagolix was approved in August 2018. It's a GnRH antagonist. It's given orally. It's got a short half-life and it lowers estrogen production.
I'm going to talk about the publication in the New England Journal of Medicine from July 2017 only. I realize that there were studies on elagolix before that and extension studies and other studies after that, but for this, I'm just going to take a review of this one. Now, 36% of the authors were employees or ex-employees with stock and stock options and others were members of speakers bureaus.
The first draft was written by an [inaudible 00:10:29] employee. The trial design, you're probably aware of it. There were two trials that were essentially identical [inaudible 00:10:36] 1 and 2. They compared placebo with two doses of elagolix. The age range of the patient's in these studies was between 18 and 49. The age range in the Lupron studies was 18-35.
What happens to ovarian function after the age of 40? Well, it goes down. So by including patients up to the age of 49, I worry that some of the improvement if any that was seen with elagolix, might have just been related to the fact that they were not having very good estrogen production as they approach 50 years of age. The surgical diagnosis of endometriosis was required to have been done in the last 10 years. Previous surgical treatment was not disallowed, other than hysterectomy.
They didn't allow complex cysts over 3 cm so this was not a study of advanced endometriosis. The target was six months of treatment and see what happens.
Efficacy. Recall that the Biberoglu and pain scale had the main end points that you see back in the 1980s with Lupron, but now the main end points have been reduced to two, dysmenorrhea and nonmenstrual pelvic pain. This was measured by a composite pelvic score.
The other three, dyspareunia, pelvic tenderness and induration. Dyspareunia, they were all collected by protocol, the dyspareunia improved a little bit at the upper dose, but the fact that induration and tenderness, nodularity and tenderness on exam, the fact that the data was collected, but not reported makes me wonder did those signs stay the same or perhaps even worsen? You can be sure that if there was any slight improvement, it would have shown up in the article.
Dysmenorrhea. The response options that a patient had as far as I could look at the other literature with the study, no discomfort; mild discomfort, but I was easily able to do the things I usually do; moderate discomfort or pain, I had some difficulty and so on. And again, as I asked yesterday, is it important to try to tease out a true uterine component? Do you have uterine cramps? Does it radiate to the typical areas? So it's one of those things that you get two things mixed together and what's really going on?
Dyspareunia is not a primary end point of the elagolix study. Where did it go? Was it demoted as a primary end point? It was a primary end point in the Lupron studies. Was it demoted after examination of preliminary studies showed that there was only a limited response of dyspareunia at the higher dose? Does that mean that secondary end points are less important? Dyspareunia is a very important symptom to patients and physicians.
Tenderness and nodularity, where did they go? The data was collected, but not reported and like I said, one might conclude that these did not improve or perhaps worsened.
Dropouts before the six months. There were up to 26, over 20% dropouts across the way. Adverse events, 4-10% of [inaudible 00:14:18] experienced adverse events. Another thing that was interesting in here is withdrew consent was a reason 6-9% of the patients dropped out of the study. What I saw in the Lupron study was frequently patients would be given two or three or more reasons to check off why do you want to stop.
In the Lupron studies, if a patient had checked something like my preference and Lupron is not working, my preference would become the first reason given for why the patient stopped the Lupron rather than the Lupron wasn't working. So withdrew consent in this makes me wonder what's hidden in there? Are there some more treatment failures than what was experienced?
You get some degree of amenorrhea, but not extremely good, I guess. Ovulation isn't inhibited, so you need backup birth control.
The protocol for ovarian function said essentially report at the estradiol level sinks to less than 10 pg/mL. It's like the estradiol requirements are going down over time. I've only seen one patient in my entire career with an estradiol level of less than 10. So if you make it that low, you'll never get any reports, perhaps, about low estrogen concerns.
Here's the graph that was published in the New England Journal of Medicine on efficacy for the two studies. And one of the things that gets confusing, is two identical studies, two almost identical-appearing graphs. It's almost like if you say the same thing twice, it becomes more powerful. But then when you correct for placebo, it becomes a little bit less impressive. And then when you correct for dropouts, it becomes less impressive.
And then what about the corrections for previous surgery? Maybe some of those patients didn't have endometriosis anymore, just had pain for some other reason. What about the women who approach the age of 50? Is some of their improvement related just to the fact that they don't produce much estrogen anymore?
This is from the supplemental appendix. This is table 5. This is the intent to treat analysis, combined both studies the [inaudible 00:16:48] 1 and 2 into one graph. And this shows the basic response of these symptoms to elagolix. You can see that dysmenorrhea responds in less than half and nonmenstrual pelvic pain responds in less than that.
One of the interesting things about response was if the patient continued to take the same amount of pain medicine, that was considered a plus and as long as they had some improvement on some other pain scale or some other instrument.
The study seemed overpowered. There were hundreds and hundreds of patients compared to the few dozen in the Lupron studies. One of the tactics used by drug companies is to overpower studies with large numbers of patients so that small, arithmetic differences can be shown to be statistically significant. And were these studies overpowered for this reason? I don't know and again, I'm just going through what I as a general gynecologist had to think about this as I read it. Overpowering lowers the statistical bar.
So my conclusion was that elagolix is better than nothing for three supported indications, temporary reduction during treatment of dysmenorrhea and nonmenstrual pelvic pain and at a higher dose, dyspareunia can improve. But it seems clear to me from the data that I reviewed that most women don't seem like they're going to respond and it seems no better than a birth control pill.
By the power invested in me by evidence-based medicine, therefore, I'm able to say that the article had no support for treating patients with mild pain because the patient's were moderate-to-severe pain. There is no support for severe or deep endometriosis, ovarian endometriomas greater than 3 cm, intestinal endometriosis, urinary tract endometriosis, diaphragm, tenderness on pelvic exam. If you have a patient with tenderness on exam, don't tell them that they will improve with elagolix. There is no evidence for that. Pelvic nodularity, deep endometriosis doesn't improve.
Interestingly, tenderness improved with Lupron so the history is there that it can improve and there is no support for brain tumors. But since off-label use is allowed, you could potentially use it for any of these.
Birth control pills seem like a plausible option and the typical pattern would be somebody has already been on birth control pills and they start elagolix as a second or third-line treatment, but I'm not impressed by the apparent efficacy of the drug. I don't know if there is going to be a future comparison of Orilissa to birth control pills, but that would be the obvious thing to know.
A former Editor in Chief of the New England Journal of Medicine said it's simply no longer possible to believe much of the clinical research that is published. The data is massaged and pummeled until it says something that is good.
I had an editorial in the British Medical Journal about evidence of endometriosis from randomized control trials. We said that elitism is prevalent and inappropriate because it makes statistical significance into clinical significance sometimes. If you torture the data, the data will confess.
All these medical therapies are all hormonal at the end of the day and they were all introduced with fanfare. And it's almost like we're just repeating the same steps with newer and more expensive medicines that are no more effective than what we already have. Medical treatments are cliches in the sense that they treat symptoms, but not the disease.
A bigger problem, though, I feel, is that science is being controlled by drug and equipment companies increasingly. This is a real concern to many, many people and patients alike. And what could go wrong if companies are in control of science? Well elagolix is targeted at about 844 a month, 20,000 for two years of treatment. And so for long-term care of patients with endometriosis, it's something that is beginning to approach the expense of surgery.
So in a perverse way, I'm happy to see elagolix have them charge as much as they can possibly get because the more it costs, the more economically competitive surgery becomes with much better results. [inaudible 00:22:17] used to make contributions to the Endometriosis Foundation of America, but [inaudible 00:22:23] has said that he is going to be severing that relationship, which I applaud.
One of the things that [inaudible 00:22:32] is trying to do is do data mining, I think, from patients by for instance saying that we'll give a donation to the EFA if you send us some information. So that's gone.
This is what endometriosis treatment looks like. This is a patient who had excision of endometriosis from several areas. And you can do more. Even if your surgery lasts 10 hours, you can do more in 10 hours of excision than in six months or two years of any medical therapy.
I'm happy for those that improve with any kind of medical therapy, but at the same time, I'm sad because how many people are kept from having good surgery because they go to medical therapy for whatever reason, whether the physician can do the surgery or not or who knows what.
So that's all I've got to say. Thank you.