To the Editor:
We are very pleased to note Dr. Meldrum’s “Conceptions” piece (1), following our recent report (2), reaffirming our main conclusion that the current IVF outcome reporting system requires a major overhaul. Dr. Meldrum, furthermore, emphasizes other points raised in our article, including the need for ART reports to incorporate maternal and perinatal outcomes rather than just pregnancy rates in defining assisted reproductive technology (ART) success, to better align reports with patient interests (3). We also agree with basing outcome reports on the total reproductive potential (TRP) of each initiated in vitro fertilization (IVF) cycle, which includes the initial fresh cycle plus subsequent frozen transfers, a process feasible from the current Society for Assisted Reproductive Technology (SART) database (4).
Dr. Meldrum’s main proposal of reporting outcome data on a per embryo transfer basis will, however, not solve the problem of outcome data manipulation in IVF, which our report brought into the open (2). In fact, reporting outcome data on a per embryo transfer basis may actually further exacerbate exclusion of poor prognosis patients who start cycles but, for various reasons (embryo banking with unsuccessful subsequent thaw, all embryos degenerate in extended day-5/6 culture, all embryos aneuploid after preimplantation genetic screening (PGS), etc.), never reach embryo transfer. Exactly these kinds of exclusions currently allow a small minority of reporting centers to disproportionately select out patients with low pregnancy potential and, thereby, to manipulate pregnancy rates in the centers’ favor (2).
The only statistically correct way to prevent such data manipulations is outcome reporting based on prospective data collection at cycle start, and calculation of success rates based on intention to treat rather than per embryo transfer. Therefore, reporting ART outcomes on a per embryo transfer basis as proposed by Dr. Meldrum would perpetuate the flaws of the current system, further misleading the public as well as professional colleagues, who accept these reports at face value. Similarly, clinical trials involving emerging ART techniques like embryo banking and PGS should report outcomes on an intention to treat basis.
Furthermore, outcome data should reflect patient selection criteria, which identify how adversely selected a patient population is at any given center. Patient characteristics for individual IVF centers can be easily defined by age, FSH and AMH levels at cycle start, and by other prognosticators, such as number of prior failed cycles. Outcome reports by individual centers then can be centrally statistically adjusted by reporting agencies for severity of treated patient populations.
Data representations by individual IVF centers, of course, also need to be properly verified through a credible inspection process, with appropriate penalties for intentional misrepresentations.
Vitaly A. Kushnir, M.D.
David H. Barad, M.D., M.S.
Norbert Gleicher, M.D.
Center for Human Reproduction (CHR) (V.A.K, D.H.B., N.G.), New York, New York
Foundation for Reproductive Medicine (D.H.B., N.G.), New York, New York
1. Meldrum DR. Pregnancies and deliveries per fresh cycle are no longer adequate indicators of in vitro fertilization program quality: how should registries adapt? Fertil Steril. Published online Aug 2, 2013. doi:pii: S0015-0282(13)02748-9.
2. Kushnir VA, Vidali A, Barad DH, Gleicher N. The status of public reporting of clinical outcomes in assisted reproductive technology. Fertil Steril. Published online June 8, 2013. doi:pii: S0015-0282(13)00610-9.
3. Min JK, Breheny SA, MacLachlan V, Healy DL. What is the most relevant standard of success in assisted reproduction? The singleton, term gestation, live birth rate per cycle initiated: the BESST endpoint for assisted reproduction. Hum Reprod 2004;19:3-7.
4. Stern JE, Hickman TN, Kinzer D, Penzias AS, Ball GD, Gibbons WE. Can the Society for Assisted Reproductive Technology Clinic Outcome Reporting System (SART CORS) be used to accurately report clinic total reproductive potential (TRP)? Fertil Steril 2012; 97:886-9.
Published online in Fertility and Sterility doi:10.1016/j.fertnstert.2013.08.036
The author responds:
Concerns for a sorely needed overhaul of Society for Assisted Reproductive Technology (SART) registry reporting appear to be mainly coming from two directions. I have proposed changes aimed primarily at recognizing recent advances and maximizing the potential for a positive outcome for both good and poor prognosis couples (1) , whereas others may have the primary goal of trying to eliminate what may be intentional manipulation of the reporting system by competing programs. Both are valid concerns, but let’s be careful not to literally “throw the baby out with the bathwater.” To some extent both concerns can be answered by a carefully thought out restructuring of reported clinic-specific data, but our primary concerns must be to maximize every couple’s chances of taking home a baby and to not discourage innovative approaches to in vitro fertilization (IVF) treatments.
Ever since the first clinic-specific report, for which I accept responsibility, its structure has actively discouraged “natural” or low-stimulation cycles. When added to centers’ total fresh pregnancy rates, those cycles would drag down their rates, making them less competitive. This has promulgated a high-dose, intensive cycle approach of generating multiple embryos to maximize embryo selection for transfer. Patients with a normal prognosis do well with that approach. However, the number one reason for failure in states and countries where IVF is a covered procedure is couples dropping out of treatment because they are unable to cope with intensive failed cycles (2) (that is, it is the failed transfers and not the stimulations that primarily impact their ultimate chance of success). Not only may high-dose gonadotropin further compromise the already reduced oocyte quality of poor prognosis patients, but with their lower success rate, early dropout from treatment magnifies their poor outcome. This was nicely shown by Cobo et al. (2), who first reported the concept of accumulation of oocytes in poor responders using a more restrained stimulation together with vitrification, resulting in a significantly improved outcome by avoiding the high drop-out rate in the controls. That group of investigators also had shown that letrozole together with gonadotropin not only improved ovarian response but was associated with increased follicular fluid testosterone (T) and a dramatic improvement of the embryo implantation rate (3). Because ovarian T is reduced in poor responders and T improves granulosa cell proliferation and function (4), such an approach together with low-dose gonadotropin may be one of the strategies that can be applied for oocyte/embryo (O/E) accumulation cycles.
Since some may categorically reject the value of accumulation strategies, let’s keep our discussion closely bound to the real world with two examples from our own practice. In a 43-year-old female who failed to respond to maximum stimulation protocols and did not accept the option of egg donation, a strategy of multiple modified natural cycles (5) in conjunction with growth hormone was undertaken with vitrification and deferred transfer of day 2 embryos. The first transfer of three embryos resulted in a miscarriage, and the second transfer of four embryos resulted in a healthy female. The second example was a 41-year-old female with day 3 FSH 13.5 mIU per ml, two prior early losses, and cancellation on maximum stimulation protocols. Because of early reports of success with clomiphene citrate (CC) “minimal stimulation” protocols in abstracts and early publications, since CC has been reported to increase aromatase and follicular fluid estradiol levels, and because of favorable embryo quality recalled from the earlier days of IVF using CC, she was offered accumulation of embryos, and four CC cycles resulted in five embryos. Transfer of two good-quality embryos resulted in the birth of a healthy female.
As I understand, SART is considering future reporting of O/E accumulation cycles as a single success but multiple oocyte retrievals. That would be a strong disincentive for programs to further examine the option of O/E accumulation cycles, and with such a registry structure, these two couples likely would not have become parents.
I proposed in my “Conceptions” paper (1) to combine fresh and frozen cycles to completely remove all considerations of whether to do a fresh or freeze-all cycle except for what is best for the couple’s success, and for the pregnancy and for the resulting offspring. The authors and SART have valid concerns with this approach because reporting success rates per transfer is not transparent in indicating the performance of multiple O/E accumulation cycles in some of the patients. As I understand, SART will likely separate out natural cycles, I presume including modifications required with addition of a GnRH antagonist to allow full oocyte maturation (5). Perhaps the solution to the above concerns is to also separate out O/E accumulation cycles, still reporting results per transfer, but indicating the average number of cycles performed and average number of embryos generated and transferred. The authors also expressed valid concerns that reporting of outcomes per transfer in routine stimulation cycles would give an unfair advantage to programs doing only day 5 transfer with or without preimplantation genetic screening, because cycles not reaching transfer would be excluded. As long as accumulation cycles are reported separately, routine cycles with and without deferred transfer could be combined and reported per stimulation, retrieval, and transfer as in the current report. Separating out O/E accumulation cycles would also allow normal prognosis couples to have a more accurate estimation of their success rate, and low prognosis couples offered O/E accumulation cycles to see the actual experience reported by their respective center.
As the authors point out, ideally registry results would be corrected for duration of infertility, number of failed IVF cycles, whether controlled ovarian hyperstimulation/intrauterine insemination cycles had been done, proportions of ethnic groups documented to have lower outcomes, body mass index, and perhaps even mean antimüllerian hormone levels. However, the complexity of such an endeavor makes it unlikely that it will happen. We must continue to caution patients against blindly comparing rates among centers without realizing how many factors, other than the quality of the IVF service, can impact those rates.
My voice and the authors’ voices are just some of a multitude that have reached the ears of the SART board members and indirectly the Centers for Disease Control officials with whom they must collaborate to produce a new registry format mutually acceptable to both parties. We unfortunately may not be able to expect that our current discussions will influence that process for this go-around, but if not, hopefully our efforts will result in further improvements in reporting in the near future.
This exchange of opinions and ideas need not end with the authors’ letter and this response. I would encourage continuation of the discussion on the online Fertility and Sterility Discussion Forum linked to my “Conceptions” paper (http://fertstertforum.com/meldrumdr-ivf-success-rates-frozen-embryo-transfer/).
David R. Meldrum, M.D.
Reproductive Partners Medical Group,
Redondo Beach, California
1. Meldrum DR. Pregnancies and deliveries per fresh cycle are no longer adequate indicators of in vitro fertilization program quality: how should registries adapt? Fertil Steril 2013.
2. Cobo A, Garrido N, Crespo J, Jose R, Pellicer A. Accumulation of oocytes: a new strategy for managing low-responder patients. Reprod Biomed Online 2012;24:424-32.
3. Garcia-Velasco JA, Moreno L, Pacheco A, Guillen A, Duque L, Requena A, et al. The aromatase inhibitor letrozole increases the concentration of intraovarian androgens and improves in vitro fertilization outcome in low responder patients: a pilot study. Fertil Steril 2005;84:82-7.
4. Meldrum DR, Chang RJ, Giudice LC, Balasch J, Barbieri RL. Role of decreased androgens in the ovarian response to stimulation in older women. Fertil Steril 2013;99:5-11.
5. Meldrum DR, Rivier J, Garzo G, Wisot A, Stubbs C, Hamilton F. Successful pregnancies with unstimulated cycle oocyte donation using an antagonist of gonadotropin-releasing hormone. Fertil Steril 1994;61:556-7.
Published online in Fertility and Sterility doi:10.1016/j.fertnstert.2013.08.037