ATLANTA — A new embryo morphology grading system could revolutionize the way in which patients and physicians compare infertility clinics, investigators reported.
Until now, data from national reporting of in vitro fertilization (IVF) success rates have been used to compare clinics. However, that data can be imprecise, because the severity of patient infertility varies between IVF clinics. Centers that are required to accept all comers inherently have lower success rates than those that can cherry-pick their clientele.
In 2004, several clinics approached the Society for Assisted Reproductive Technology (SART) to consider having clinics report morphologic grades for their embryos, because embryo quality is thought to reflect fertility potential. Some clinics have been collecting these data, but the quality measurements differed between laboratories.
SART developed its own system that grades embryos according to visual characteristics into one of three categories: good, fair, or poor. The society also introduced the morphologic characteristics of cellular symmetry and percentage fragmentation as potential hallmarks of embryonic well-being.
From June 2006 to January 2008, SART asked clinics to voluntarily report their morphology data. Those data have been obtained for 70,293 embryos from 28,186 transfers, representing 19% of all the fresh autologous embryo transfers reported to SART's Clinical Outcomes Reporting System in 2006 and in 32% of all transfers in 2007.
Analyses showed significant differences between good, fair, and poor embryos and live birth rates. Furthermore, the relationship between live births and embryo grade was negatively correlated with increasing maternal age, according to Michael Vernon, Ph.D., who helped develop the system with seven other embryologists led by Catherine Racowsky, Ph.D.
Embryo transfers were performed from 1 to more than 7 days post insemination, with 62% of transfers performed on day 3 and 29% on day 5. The majority of embryos transferred were classified as good on day 3 (70%) and day 5 (78%). Few fair and even fewer poor embryos were transferred on day 3 (24% and 5.5%, respectively) and day 5 (18.6% and 3.6%).
Among women who received two embryos of the same grade, the live birth rate for good embryos was 45% on day 3 and 56% on day 5, compared with 35% and 42% for fair embryos, and just 21% and 30% for poor embryos, Dr. Vernon, chair and professor of obstetrics and gynecology, West Virginia University in Morgantown, and his associates reported in a poster at the annual meeting of the American Society for Reproductive Medicine (ASRM).
The data are so encouraging that SART is considering mandating that clinics report their embryo morphology information. If that mandate does come to be, clinics will have critical information to aid their quality control and quality assurance activities.
In the current data set, more than 670 embryo transfers contained more than 6 embryos, which is not within SART guidelines of acceptability. In some cases, more than 10 embryos were transferred. In all, 48% of labs transferred only one to two embryos on day 3, and 79% did so on day 5.
That falls within the recently revised SART/ASRM embryo transfer guidelines recommending that only one more embryo be transferred than called for in four age-based prognostic categories.
Related data reported at the same meeting by Dr. Racowsky associate professor of obstetrics and gynecology at Brigham and Women's Hospital, Boston, showed a strong positive correlation for cellular symmetry and percentage fragmentation.
The analysis she presented showed that live birth rate increased from from 2.9% for embryos with less than six cells on day 3, to 24.3% for those with eight cells, but decreased to 16.2% for those with with more than eight cells. The live birth rate decreased from 23% for embryos with perfect symmetry to 11.3% with moderate asymmetry and 4.5% for severe asymmetry. The live birth rate was 21% for embryos with no fragmentation, 11% for those with 10%-25% fragmentation, and just 2.5% for those with greater than 25% fragmentation.
The regression equation derived from this analysis revealed that with a cut-off of 0.3, 76.4% of embryos were classified correctly as either not resulting in a live birth, or giving rise to a live birth.
The authors noted that future analysis of a larger SART data set could increase the accuracy of the morphologic classification system and lead to a Web-based regression equation enabling ranking of embryo viability.
Such an equation would enhance the selection of fewer embryos at embryo transfer and reduce the potential for multiple births.
Moreover, standardization of a national embryo morphology system should assist clinics with quality control and quality assurance activities, thereby improving overall care of infertility patients.