This Week in the Democratization of IVF
first draft dispatches from the battle for a billion babies
It was hectic week in the battle for a billion babies: board meetings, term sheets, cap tables, some pretty eye-opening under-NDA data. So its a good weekend to look far back, to gain some perspective on how far we’ve come, and how reachable our goals are.
Let’s rewind to 1989. It’s 3AM and I’m the intern covering M8, the labor and delivery floor at New York Hospital- Cornell Medical Center, what the senior docs call The Old Lying-In.
As I move from room to room, looking at fetal monitor strips and talking to the patients and nurses, I hear a gurney being pushed quickly down the hall. Anattending and a senior resident rush one of the IVF patients into the procedure room to do an egg retrieval. The patient had been collecting and checking urine LH levels every few hours, lookign for the luteinizing hormone (LH) rise that signalled that the most mature eggs were about to ovulate on their own, and we needed to retrieve them and get them into the lab first.
In the 1980’s we were just transitioning from laparoscopic egg retrieval under general anesthesia to trans-vaginal, ultrasound-guided egg retrievals. The old method had a lot of disadvantages. Laparoscopy was more invasive, had a more painful post-operative period, and was a less efficient way to identify where the eggs were. Since eggs develop in fluid-filled follicles, and ultrasound detects and lights up the interfaces between fluid and tissue, ultrasound is an ideal way to figure out where to point the aspiration needle. If you played video games as a kid, you intuitively know how to do a transvaginal, ultrasound guided egg retrieval.
Looking through a laparoscope directly at the ovary, on the other hand, you blindly puncture the inside of the ovary again and again and hope for the best.
Like today, our patient had gone through a regimen of fertility drug injections; in this case the drug was probably Pergonal, a 50:50 mix of follicle stimulating hormone (FSH) and luteinizing hormone (LH) extracted from the urine of menopausal nuns in a convent in Switzerland (look it up.) Pergonal was an effective drug; too effective in many cases. The excess LH in the drug differentially stimulates the stroma of the ovaries to make hormones like testosterone at levels far above what the maturing eggs and thickening uterine lining need. This excess stimulation put the patient at higher risk for ovarian hyperstimulation syndrome (OHSS) and may have decreased the pregnancy potential for the embryos that resulted from the stimulation.
Since Pergonal was a urine-derived drug, it was more irritating to the body when injected; instead of the 1/4” subcutaneous injections that patients endure today, Pergonal needed to be injected 2” deep, the needle aimed where the upper, outer rivet on the back pocket of a pair of jeans, twice a day.
The bruising was epic, as was the inconvenience of needing to recruit an injector to give the shot whether at home, at work, or wherever.
There were probably one or two patients in the hospital at the time under observation for OHSS; with ovaries three or four times the normal size and at risk for twisting on their vascular pedicles and choking off the blood supply (a surgical emergency) and an abdomen rapidly filling with fluid, the patients could become extremely unstable. Intensive care admission were not uncommon and rare deaths occurred. The key time period for women with OHSS was ten days to two weeks after the egg retrieval, when the early pregnancy starts to produce HCG, the hormone that we mean when we refer to a “pregnancy test.” The HCG acts as fuel on the fire for the enlarged, supercharged ovaries. When a patient with OHSS told us that her symptoms suddenly got worse, we had conflicting emotions: joy that the patient was pregnant, but worry that she could get very, very sick.
Fast forward to today. What’s different?
To start, the stimulation is easier. We’ve replaced Pergonal with drugs manufactured using laboratory techniques not requiring urine; the increased purity lets us inject them just below the skin, a procedure that the patient can usually do herself and, while not painless, one that is less irritating, even when done twice a day for two weeks. We also found that by adding a second hormone to the regimen, one that minimizes the risk of the eggs ovulating on their own before they can be retrieved for the lab, we gain control over the retrieval schedule and no longer need the mad rush to the operating room.
The development of vaginal ultrasound-guided egg retrieval made everyone’s lives better. For patients it meant less pain and less anesthesia; for the clinical staff it meant one less nurse (laparoscopy needs a scrub and circulating curse) and dispensing with the cumbersome incision-fill the abdomen with carbon dioxide-setup up and insert the scope, awkwardly immobilize the ovaries and blindly stab away where at the bumps in the surface where you think the eggs might be. For the embryologists, it meant that retrieval could be done adjacent to the laboratory itself rather than at the site of whatever operating suite was used in the particular facility.
We can’t finish the discussing the egg retrieval procedure without mentioning as IVF variant that were still widely practiced in the late 1980’s and early 90’s: gamete intrafallopian transfer, or GIFT. In these early decades, with embryo implantation rates in the low single digits, many thought that minimizing the time the eggs and embryos spent in the laboratory was a good idea, so in many cases, during the laparoscopy we combined the just retrieved eggs with a fresh sperm prep, threaded a catheter through the end of one or both of the fallopian tubes and deposited the eggs and sperm about midway through the tube, close to where fertilization occurs in unassisted reproduction. Variants of this procedure included zygote intrafallopian transfer (ZIFT) and tubal embryo transfer (TET), depending on where in the embryo development cycle the transfer occurred.
There was an ongoing and occasionally fierce debate as to whether IVF or GIFT produced better outcomes. Unfortunately, the pregnancy rates for both were too low for anyone to feel confident about the outcome of a cycle. To compensate for the low probability of success for each egg or embryo returned to the body, we returned a lot of embryos all at once, so when it did work it worked too well: rates of twins, triplets and higher order multiple pregnancies jumped, each one representing much higher risk for both mother and babies — risks of preterm delivery, risks of pregnancy-associated diabetes and blood pressure complications, difficult deliveries and risks of bleeding after delivery. Newborn intensive care units filled with infants prematurely delivered 8, 10, 12 or more weeks early, the incubator-connected babies often in a row with their siblings. (This demographic bump later challenged local school boards.)
Clearly, IVF patients and their medical teams in 1989 faced a difficult balance: to have even anywhere near a decent chance of having a child from the cycle, what might turn out to be a too-high number of embryos needed to be transferred. Compounding that dilemma was the fact that we were not all that great at assessing which embryos to put back. The typical IVF embryo transfer in 1989 might happen two days after the egg retrieval, when an “ideal” embryo contained only four cells. As we now know, embryos often “declare themselves” in the subsequent three days as they progress to what is known as the blastocyst stage; many decent looking embryos at day four stop along the way, and it difficult to predict which embryos would. We lacked the capability to continue to nourish embryos, so on day two we put back a bunch of them and hoped for the best.
Yet another big factor in the transfer decision was that we were just figuring out how to freeze and store embryos to use later. So faced with a one-shot only chance of a good outcome from the stimulation (20+ deep shots with a long needle, high risk of hyperstimulation syndrome, a surgical laparoscopy under general anesthesia), it was easy to justify a more is better than less approach to embryo transfer, up to and including “what the hell, just put ’em all back.”
We have a long way to go to reach the 30 million five thousand dollar IVF cycles to produce 25 million babies a year model that would address all of the good that we can do with this remarkable procedure. That said, looking at how far we have come makes it easier to envision how continued innovation will get us there. Those of us who remember just how quickly the world changed when we figured out that by injecting sperm directly into the eggs themselves we could treat almost any case of male factor infertility — how on a Tuesday in 1993 we could only offer the use of donor sperm because IVF had truly dismal results with very low sperm counts or motility — and then on Wednesday we called the same patients back, telling them “never mind what we told you yesterday. We can help you now.”
Millions of people worldwide are waiting for that phone call.