Most ‘genetic amniocentesis’ procedures are done at 15-18 weeks gestation. The relatively low risks at this point in the pregnancy have been well-documented as discussed in my previous posts. Different folks have different criteria for what they consider to be an ‘early amniocentesis’, I have always defined it to be any amniocentesis done prior to 14 weeks (while the pregnancy is still in the first trimester). We first started doing these about 20-odd years ago at a time when ultrasound technology had improved to the point that we could perform the amniocentesis under ‘direct, real-time, ultrasound guidance.’ Our naïve assumption at the time was that since the procedure seemed to be relatively safe in midtrimester, now that we could see better, it should be just as safe earlier in pregnancy.
Early amniocentesis turned out not only to be technically more difficult than expected, but also less reliable in terms of getting a final fetal chromosome result, and riskier to the pregnancy. I know one of the problems I frequently encountered was difficulty in penetrating the membranes surrounding the baby. The ‘bag of waters’ is actually composed of two layers of membranes, the amnion (closest to the baby) and the chorion. During the first part of the pregnancy, the amnion and the chorion are separated and may not fuse into one sac until the end of the first trimester, and even then, the ‘fusion’ can be weak until later in midtrimester (and sometimes remains that way if a baby has a chromosomal abnormality such as trisomy 21). It was not at all unusual during an attempt at early amniocentesis to see ‘tenting’ of the membranes (separation from each other or from the wall of the uterus) as we tried to push the needle into the space around the baby. When chorionic villus sampling (CVS) came along, providers gravitated to offering that in first trimester and amniocentesis at 15+ weeks rather than performing early amniocenteses routinely.
The only study I am aware of that has actually systematically looked at early amniocentesis was done in Canada (Canadian Early and Mid-Trimester Amniocentesis Trial (CEMAT) Group) published in 1998 (Lancet 1998;351:242-47). In this trial, 4,374 women were randomized to either early amniocentesis (between 11 and 12 6/7 weeks) or midtrimester amniocentesis (between 15 and 16 6/7 weeks). This and subsequent reports from the trial demonstrated that compared to midtrimester amniocentesis, early amniocentesis was associated with a 4-fold risk of a technically difficult (twice the risk of requiring multiple needle insertions) or unsuccessful procedure (1.6% vs. 0.4%), a 10-fold risk of chromosome culture failure (2.4% vs. 0.25%), a higher rate of fluid leakage following the procedure (3.5% vs. 1.7%), a greater risk for pregnancy losses (7.6% vs. 5.9%), and a significantly higher risk (1.3% vs. 0.1%) of having a baby with talipes equinovarus (club foot). Unfortunately, this study did not provide data for amniocenteses done between 13-14 6/7 weeks (when I did most of my ‘early amniocenteses’), but it is doubtful at this point, in view of the results of the CEMAT trial, that this study ever will be done. Today, when a patient requests an earlier diagnosis, but is reluctant to undergo CVS, I will still offer an ‘early amniocentesis’ but counsel them regarding the increased risks and request that they delay this until 14 weeks.
Anyway, after a nice discussion, my patient today decided she would go with ‘combined first trimester screening’ for aneuploidy and probably wait to have an amniocentesis done until 15-16 weeks, “regardless of whether or not the ‘screening’ test is reassuring.” For now, that is the right decision for her…As I have learned in the past, however, with further information, she might still opt for another approach after she gets back today’s final results! At least she now knows what options are available to her.