IVF 2: Acupuncture

The nurse who walked us through the whole plan for in vitro fertilization (IVF) mentioned that acupuncture might be helpful in increasing our (well, my) odds of getting preggo. What's that all about?

Methods: I used PubMed (a search engine of published scholarly articles in the biomedical sciences) to find primary research articles and a couple of commentaries. I read 6 of these, in various states of ebriation. In order to maintain objectivity, I significantly suppressed snarky thoughts about acupuncture (p less than 0.0000001).

Results: First off, few published articles address acupuncture and IVF. PubMed gave me 20-odd hits for "ivf, acupuncture". Many of these hits were irrelevant. This alone indicates a line of research of such infancy that I would not use it to make any medical decisions.

Second, and I'm going to say this in practically every blog entry on infertility treatments, infertility is not a unitary syndrome. I can think of no other medical problem in which patients would be lumped together regardless of whether their problem was hormonal (such as polycystic ovarian syndrome), accidental (damage to the tubes?), unexplained (your eggs just don't implant....), natural (your eggs are getting old), or even non-existent--since, after all, male factors alone explain 20-30% of infertility.

This seems crazy. I can imagine many reasons why a woman with hormonal problems might respond differently to a given treatment than a woman whose husband had a bad soccer accident. Yet they're lumped together for the purposes of this and other studies on IVF. It's clear why: it's hard to recruit enough patients for a clinical trial otherwise. Also, so many fertility problems are unexplained or derive from multiple causes that it'd be a headache to separate them out. But still.

We'll leave that aside for now. What was I reading? Most everything published on the topic is in the journal Fertility and Sterility. I don't know its reputation, although I have some opinions now after reading its publications.

Ok, on to the meat. The original observation was published by Paulus et al in 2002. In a study of 160 patients randomly assigned to acupuncture or control, where control means that the patients just lay quietly, the pregnancy rate was 1.5 x higher in the acupunctured patients (43% vs 26%). Acupuncture was carried out twice, both the same day as the embryo transfer.

Well, that's quite a big increase in pregnancy rates! Naturally, people wanted to follow up. In the May 2006 issue of F&S, three groups published their findings: 2 found significant increases in pregnancy rates following acupuncture, and a third found an increase in pregnancy rates that, while large (50%), did not reach significance.

Let's break these studies down. First, all three were careful to assign patients randomly to the acupuncture vs control groups, with the happy result that in all papers, the characteristics of women in the two groups were highly similar (age, # of previous IVF attempts, # of good embryos retrieved/implanted, etc etc). Good.

What happened to patients in the "control" groups? In a study of 225 total patients, Dieterle et al compared acupuncture to "placebo acupuncture," in which half of the patients were given actual acupuncture but at sites other than those indicated for improving pregnancy odds. This seems like the best way to evaluate whether targeted acupuncture itself is helpful (as opposed to general relaxation associated with the procedure). The "real" acupuncture group had pregnancy rates of 28%, while the placebo acu group got pregnant at a 14% rate.

Those results look great, but the authors acknowledge a strange fact: the average rates (in Germany) for pregnancy success after IVF are ~23%. They do not report the average rate of success at the clinic they studied, which is unfortunate because clinics have highly varying success rates. But if you assume that their clinic was representative, then the study raises the worrisome possibility that real acupuncture did very little, but acupuncture in the wrong spot diminished pregnancy rates. That should give anyone pause. Again, though, we can't know if that's likely or not because they don't report expected success rates at their location.

Ok, what about the other two studies? Westergaard et al have two acupuncture groups, one of which gets one acupuncture treatment, one of which gets two; plus one control group, which gets no treatment at all. This study will not necessarily tell us anything about acupuncture per se, as opposed to stress-reducing effects of the treatment, but at least they're not putting pins in the wrong places. This group finds an increase in pregnancy rates that reaches significance in the one-acu group (from 24% to 39%), but does not reach significance in the two-acu group (36% pregnant). For no apparent reason, all of the pregnancy increases came in women under the age of 37 (the median age of their patients).

The third study, by Smith et al, used "sham" acupuncture as a control, by which they mean that needles were put in slightly "off" locations, about an inch or two away from the traditional sites. Also, sham needles were used. There were 3 sessions of acupuncture or sham acupuncture, and this time the increase in pregnancy rates (from 18 to 28%) did not reach statistical significance.

These studies are overall encouraging regarding acupuncture. All show an increase in pregnancy rates following acupuncture in the "right" locations, and two of the studies even give a hint that the precise acupuncture locations, as opposed to the procedure itself, are important.

The only major concern I would raise is the possibility of publication bias. That is, if your study shows a significant effect it will be much more likely to see the light of day than if your conclusions are negative or murky. This concern was also raised in a commentary at the time by Domar. She claims that the original group (Paulus et al) presented follow-up data at a meeting, this time comparing acupuncture to acupuncture with sham needles; there was no significant increase in associated pregnancy rates (43% for acu, 37% for sham). However, these data remain unpublished, making it impossible for us to evaluate the study. How many other such studies are unpublished, Domar asks?

It wasn't a rhetorical question. Two years later, Domar et al published a paper showing absolutely no increase in pregnancy rates between acupuncture and quiet resting groups (31% acu, 33% resting). I.e., an effort to replicate the original Paulus study failed. The authors do not have any persuasive explanations, although they helpfully point out "It is notable that this is the first study from the United States and perhaps there is something to the expectations or care that differs." (Right, because here in the US we're known for our skeptical attitude towards all things mystical.)

Conclusions
1) Is acupuncture on or around embryo transfer safe? These data all suggest that it is, as long as you don't get punctured in contraindicated locations.

2) Does any aspect of acupuncture lead to increased pregnancy rates for women undergoing IVF? The jury's still out. As I said above, I personally would avoid making medical decisions based on such a small number of studies (although, of course, not deciding means not getting acupuncture, which itself is a decision...). To my surprise, I'm reasonably impressed with the evidence in favor of acupuncture, simply because four distinct groups have seen some form of positive result. However, I remain concerned about publication bias.

The other thing is that all these studies exhibit quite low success rates for IVF overall. Partly this may be due to the age of the patients, which was on average high (37). The clinic Dr Hyde and I are using boasts of 67% success rates for one cycle for women under 35, though. I'd like to see what a similar acupuncture study would find in this group.

3) Am I going to get acupuncture? Not for the first round, at least. There's already going to be enough novel experiences involving needles that I feel no urge to add another. But if you want to get acupuncture, go ahead. I will reconsider acupuncture if the first IVF cycle fails, if for no other reason than that I would want to do something.

4) How might acupuncture work in this instance, if it does work? The hypothesis put forward in several papers is that the sites needled aid in blood flow to the uterus, thereby diminishing uterine contractions and aiding in implantation. Bummer, though, that Paulus et al published another article specifically evaluating uterine contractions and found no effects of acupuncture on them. So the mechanism of the effect, if the effect is real, is a mystery.