Ectopic pregnancy rate increases with the number of retrieved oocytes in autologous in vitro fertilization with non tubal infertility but not donor recipient cycles An analysis of 109140 clinical pregnancies from the Society for Assisted Reproductive Technology registry

In autologous in vitro fertilization but not in donor/recipient in vitro fertilization cycles, ectopic pregnancy rates are higher in cycles with a higher number of oocytes retrieved.

Kelly S. Acharya, M.D., Chaitanya R. Acharya, M.S., Meredith P. Provost, M.D., Ph.D., Jason S. Yeh, M.D., Ryan G. Steward, M.D., Jennifer L. Eaton, M.D., M.S.C.I., Suheil J. Muasher, M.D.

Volume 104, Issue 4, Pages 873-878


To study the impact of controlled ovarian stimulation on ectopic pregnancy (EP) rate as a function of the number of oocytes retrieved, using donor IVF cycles as a control.

Retrospective cohort study using a large national database.

Not applicable.

Data from 109,140 cycles from the 2008–2010 SART registry, including 91,504 autologous cycles and 17,636 donor cycles in patients with non-tubal infertility.

Varying amounts of oocytes retrieved in autologous and donor IVF.

Main Outcome Measure(s):
Ectopic pregnancy rates.

In autologous cycles, the EP rate significantly increased as oocyte yield increased. This association was not found in oocyte recipients.

In autologous IVF cycles, increasing oocyte yield is correlated with a significantly increased EP rate. This association is not found in oocyte recipients, indicating that the increased EP rate may be due to the supraphysiologic hormone levels achieved with controlled ovarian hyperstimulation.

  • Suheil Muasher

    Dr Quaas,
    Thank you for your comments. We acknowledge that the overall effect size was relatively small (1.92 vs 1.65% ectopic pregnancies) and that the clinical significance is debatable. In our discussion, we were not suggesting that all cycles should be freeze-all, but mentioning that there may a high-risk population which would benefit from freeze-all for multiple reasons. As you mentioned, noting the slightly lower ectopic pregnancy rates, especially in presumably higher-risk patients such as high responders, may very well add to the “package” of benefits that a freeze-all cycle could offer.

    Dr. Franasiak,
    Thank you as well for your comments; these were also issues which we considered while analyzing our data and writing the paper. As may be expected, the blastocyst percentage was seen to increase with increasing oocyte yield, and we wanted to address this in some way. To investigate whether the blastocyst vs 3-day percentage had any effect on the ectopic pregnancy rate, we calculated the overall percentage of ectopic pregnancies in the blast versus cleavage transfers. To correct for the number of embryos transferred (as this is often higher in cleavage-stage transfers), we then calculated the percentage of ectopic pregnancies in all blast and cleavage-stage transfers in which 2 embryos were transferred. We found that cleavage-stage transfers had a higher rate of ectopic pregnancy, which shows that this was not the factor driving the ectopic pregnancy rates. In other words, we would expect the EP rate to be higher in the cycles with higher number of cleavage transfers, but we actually found the EP rate to be higher in the cycles with higher oocytes retrieved, in which there were a higher number of blastocysts transferred. The P value for the EP rate between blastocysts and cleavage-stage embryos was significant, but we believe that it actually skewed our data towards the null given the above-listed findings.

    Thank you both again for your interest and comments; they are greatly appreciated.

    Kelly Acharya, MD

  • Jason M. Franasiak

    Fascinating interpretation of a very large data set. It appears as though the cleavage vs. blastocyst transfer rates as well as the transfer order differed in the two groups. The transfer order was controlled in the sensitivity analysis, but was the stage of transfer also accounted for? It did not appear as though a p-value was presented when the DET analysis between cleavage and blastocyst transfer analysis was completed at the end of the results. Given that stage of transfer and transfer order are also related to oocyte yield it would be of interest to see both factors controlled for in the analysis.

  • The finding of opposite trends of EP rates in relationship to number of oocytes retrieved between autologous and donor cycles is intriguing. In the discussion of the article the authors therefore suggest that a “freeze-all cycle followed by later FET may be a safer option” in good responders with higher numbers of oocytes retrieved.
    Theoretically this is true. However when considering doing a freeze-all for this indication alone, the following needs to be considered: From your figure, it appears that the difference between the highest and lowest ectopic rates is (1.92-1.65 =) 0.27%.
    If you believe that doing freeze-all’s on patients with high oocytes numbers returns the EP rate to the EP rate in fresh cycles with low oocyte numbers, then the “number needed to treat” (“number needed to freeze”) would be 370- which does not appear worth it. If the EP number is generally reduced (even more) in FET cycles, for example to the 1.01% found in a prior study (reference 12 from your article), the NNTs would be 110 for women with high oocyte counts and 156 for women with low oocyte counts. This appears to be a lot of extra effort and cost for a modest benefit in either group- if done for this reason alone.
    However, given that recent research has shown that there are other benefits to FETs, maybe the findings from this study could be included in the “counseling package” on this issue, and factor into the decision-making in conjunction with other FET benefits.

Translate »