OPINIONs 010: June 6, 2017
A new form of preimplantation genetic screening (PGS), named “preimplantation genetic diagnosis for aneuploidy screening (PGD-AS)” is now being promoted heavily while skepticism over the efficacy of the not-so-different form of PGS spreads throughout the IVF field, particularly in Europe. A look back at the problem-ridden history of PGS calls for caution when evaluating whether an infertility patient really benefits from PGD-AS.
Norbert Gleicher, MD, CHR’s Medical Director and Chief Scientist was invited to give in September a talk under above noted title at the annual Ovarian Club Meeting, this year to be held in Barcelona, Spain. We here present a summary of his upcoming presentation as a part of the OPINIONs series.
It has been estimated that in 2016 approximately 20% of IVF cycles in the U.S. were accompanied by preimplantation genetic screening (PGS). Assuming approximately 250,000 IVF cycles annually in the U.S., this would represent an added annual cost of ca. U.S. $200 million to IVF for a procedure that even PGS proponents no longer claim to improve pregnancy and live birth chances; but now is alleged to warrant effort and expense because it reduces miscarriage risks and time to conception. That, like prior claims by PGS proponents of improving pregnancy and live birth rates, these two new claims also are unsupported, should not surprise since the PGS industry, from its founding in the 1990s, based on the seemingly unassailable hypothesis that exclusion of aneuploid embryos prior to embryo transfer must improve IVF outcomes, never hesitated to make unsupported claims. As PGS has amply demonstrated, even seemingly unassailable hypotheses do, however, not always work out in clinical practice.
When an earlier form of PGS (PGS 1.0), performed on cleavage-stage embryos, was, finally, declared ineffective by professional societies like the American Society for Reproductive Medicine (ASRM), the solution for the PGS industry was simple: It instantly created PGS 2.0, with even bigger promises of IVF outcome improvements and, therefore, not surprisingly, with even higher cost and wider utilization. But, like with PGS 1.0, there, again, were no validation studies to confirm the industry’s hypothetical claims before PGS 2.0 was brought to market. Remarkably, despite the obvious earlier failure of PGS 1.0 to meet expectations, IVF community and patients accepted renewed promises by the PGS industry with even greater enthusiasm.
Like it took years to overcome opposition from the PGS industry before PGS 1.0 was formally declared clinically useless, again strongly opposed by the PGS industry, opposition to PGS 2.0 has been growing for years. The burden of proof that PGS 2.0 did not work was, however, once again placed fully on opponents rather than, as one would expect, proponents of PGS, who would have to prove that PGS 2.0, indeed, did improve IVF outcomes.
When more recently evidence became irrefutable that trophectoderm at blastocyst stage was much more mosaic than represented by the PGS industry, and that, therefore, false-positive rates of trophectoderm biopsies had to be substantially higher than the industry has claimed, without any “mea-culpa,” the industry once again, literally overnight, radically changed course, simply declaring all prior pronouncements and the prior reporting system nil and void, and establishing brand-new diagnostic criteria for PGS: Suddenly, embryos were no longer either normal (euploid) or abnormal (aneuploid), but embryos (likely in a majority of cases) could fall into a third diagnostic category of being mosaic and, even more remarkably, such embryos could also, selectively, be transferred (we, going forward describe this change as the introduction of PGS 3.0).
What makes the overnight appearance of PGS 3.0 so remarkable is not only that publication of newly recommended guidelines for diagnostic laboratory criteria and clinical practice by the Preimplantation Genetic Diagnosis International Society (PGDIS) in late 2016 did not comment on why the PGDIS, so suddenly, radically changed how PGS was to be reported out (i.e., recognition that the high prevalence of trophectoderm mosaicism did not allow for only diagnostic criteria of euploid and aneuploid and, at minimum, mandated a third category of “mosaic-normal”) but that the PGDIS (in speaking for the PGS industry) did not have the courage to acknowledge that under the prior PGS 2.0 criteria, likely tens of thousands of potentially transferrable embryos had been erroneously discarded and, with it, pregnancy chances for thousands of women.
In a typical marketing strategy for recovery of a commercial product that has fallen into disrepute, and to separate it from its two prior incarnations, this now allegedly again “renewed” PGS product we call PGS 3.0 has in recent publications (May issue of Fertility and Sterility) been given the completely new acronym, preimplantation genetic diagnosis for aneuploidy screening (PGD-AS). Despite these marketing efforts, it is, however, once again only yet another unvalidated diagnostic test with completely unsupported claims by evidence: According to new PGDIS guidelines, embryos with up to 19% aneuploid cells (per biopsy, meaning ca. 1.2 cells in an average 6-cell biopsy) are normal-euploid and should be the preferred embryos for transfer. Embryos with 20-80% aneuploid cells (1.2-4.8 cell in an average 6-cell biopsy) are euploid-mosaic, and are potential candidates for transfer, following a hierarchical selection process. Embryos with over 80% aneuploidy (4.8-6.0 embryos) are considered fully aneuploid, and should not be transferred (i.e., should be discarded).
The new PGDIS guidelines, however, do not address where these cut offs come from. Indeed, there are no data in the literature to support them. The 20% cut off between euploid and euploid-mosaic is obviously based on high-resolution next generation sequencing (hr-NGS), the only diagnostic platform capable of detecting mosaicism with reasonable accuracy, though only above 20% (The new PGDIS guidelines claim that only NGS should be used in testing embryos under their guidelines). The PGS industry, thus, does not separate normal-euploid from normal-mosaic based on clinical validations of differences in IVF outcomes but solely based on technical limitations of hr-NGS, which cannot reliably detect mosaicism levels below 20%. An embryo with 19% aneuploid under PGDIS criteria is, thus, automatically transferrable, but an embryo with 20% or 21% aneuploidy (i.e., being designated as “euploid-mosaic”) should only be transferred with considerable caution, and selectively. The same complete lack of clinical evidence also underlies the 80% cut off between euploid-mosaic and full aneuploidy, as defined by the PGDIS. 80% aneuploid embryos, therefore, can be selectively transferred but 81% aneuploid embryos should be discarded.
The newly promoted PGS 3.0 (PGD-AS) is, therefore, as spurious a diagnostic product as PGS 1.0 and PGS 2.0 before. CHR does not see biological or clinical rationale in these new, yet again unvalidated PGDIS guidelines, as they will still lead to discarding of potentially healthy embryos in large numbers, and will not improve IVF outcomes. Increasing evidence supports that trophectoderm mosaicism at blastocyst stage is a completely normal physiological phenomenon. The extent of trophectoderm mosaicism, lack of congruency in ploidy between trophectoderm and inner cell mass and, finally, increasing evidence for significant self-correction of embryos downstream from blastocyst stage, make it biologically and technically highly unlikely that the ploidy of blastocyst-stage embryos can reliably be determined by a single trophectoderm biopsy. PGS in all forms, therefore, should be clinically abandoned.