Predicting Spontaneous Ovulation in Hyperstimulation

Click here for Audio/Video Version  *requires RealPlayer - free download

Hans van der Slikke, MD: "It's October of 2001 and we're in Orlando, Florida at the ASRM meeting. Next to me is Kaylen Silverberg who is the Medical Director of the Texas Medical Center. Welcome, Kaylen."

Kaylen Silverberg, MD: "Thank you. How are you?"

Hans van der Slikke, MD: "Fine. You did a presentation about hyperstimulation and spontaneous ovulation. I think it was a very interesting story but it's a problem we all deal with and we want to predict spontaneous ovulation. Could you tell us how you set up your study?"

Kaylen Silverberg, MD: "Sure, we designed an analysis to only look at poor responders and we know that poor responders ovulate approximately 10% of the time spontaneously. That's a real problem because we know that pregnancy rates are maximized by completing the stimulation, getting mature follicles, administering HCG, and then doing two consecutive intrauterine inseminations on the next two days. The problem is what do you do when patients ovulate spontaneously? It really markedly lowers the chance for pregnancy. Now that there are new drugs on the market called GnRH antagonists we would like to be able to identify patients who are at risk for ovulating spontaneously because if we could have a good way to predict who is going to ovulate spontaneously, we would know who we should give these medicines to. So what we did was an analysis to try and determine what the risk factors are and, if it's possible to predict, who really is at risk of ovulating spontaneously."

Hans van der Slikke, MD: "How many cycles did you cover with your research?"

Kaylen Silverberg, MD: "We looked at 762 cycles and of those 762 patients they were all undergoing their initial cycle of ovulation induction for intrauterine insemination. They were freshly presented to our office; we had no background information on them other than their prior history but they hadn't been stimulated before so we had no real triggers or real tip off that they might be a risk. It was really a very fair analysis looking at everybody. Of those 762, we ended up with 54 patients who ovulated spontaneously - about 7%, and the remainder did not. They actually made it to HCG administration and did what they should have done. So our numbers don't differ too much with what's published in the literature, implying about 10%, we were about 7.1%."

Hans van der Slikke, MD: "Then you compared these two groups, this group with poor responders and spontaneous ovulation with the others, and you looked at the number of follicles and at the estradiol level."

Kaylen Silverberg, MD: "Exactly right. We looked at a lot of different factors. One of the first things we looked at were demographic factors because we wanted to see if these patients were older because there's a lot of evidence that the older patient is more likely to be a poor responder and could potentially ovulate spontaneously. They were not older, in fact, the patients who ovulated spontaneously were 36 ½, and the patients who didn't ovulate spontaneously were 35 - no statistically significant difference. No difference in weight because some people believe that there are potential weight issues; they were all about 148 pounds so no real differences in weight at all. When we looked at stimulation parameters, there were no differences there either. The days of stimulation were the same, the total gonadotrophin dosage was the same, and the number of ampules with medication was the same so you couldn't predict even as you were going through the stimulation. It's very disheartening."

Hans van der Slikke, MD: "And you used two types of FSH?"

Kaylen Silverberg, MD: "We did, the recombinant FSH and we used highly purified FSH. The reason we switched was because our practice switched so initially we were using highly purified FSH and then we switched to recombinant. We did a separate analysis to analyze patients who got recombinant and patients who got highly purified and we analyzed them separately and no difference."

Hans van der Slikke, MD: "There was no difference."

Kaylen Silverberg, MD: "We were able to then combine those two groups when the separate analysis proved to be no different."

Hans van der Slikke, MD: "What were your conclusions?"

Kaylen Silverberg, MD: "It's interesting because what we concluded was the patients who ovulated spontaneously had fewer large follicles and fewer follicles greater than 18 mm so when they ovulated spontaneously they did so much earlier than you would have predicted otherwise. In our practice we try and administrate CG when we have follicles in excess of 19 mm so, in other words, we try to get to 20 which is pretty much the standard, and these patients didn't even have 1 total follicle greater than 18 so they would ovulate very early. In addition, they had fewer follicles greater than 14 mm; they had an average of 2 follicles compared to 4 follicles in the group that didn't ovulate spontaneously and also their serum Estradial levels were lower statistically, significantly lower. So we were able to then do a receiver operating characteristic curve analysis, a ROC analysis, to try and determine a threshold. What we found was that patients who have Estradial levels less than 800 with few mature follicles are at a much greater risk of ovulating spontaneously."

Hans van der Slikke, MD: "Did you now set up a new study to see if you could prevent it in this group."

Kaylen Silverberg, MD: "That's an excellent question, and that's what we're doing now. We're taking these patients who are at high risk and starting them on GnRH antagonists to be able to see how they do now and see if we can block spontaneous ovulation."

Hans van der Slikke, MD: "I think I understand. Thank you very much for this interview."

Kaylen Silverberg, MD: "Thank you."