Survey of genetic screening for oocytedonors

Vivian Lewis, M.D.,* Devereux N. Saller, Jr., M.D.,*† Ruth W. Kouides, M.D., M.P.H.,‡

and Judy Garza, M.S.,†

University of Rochester Medical Center, Rochester, New York

Objective: To describe current screening practices of oocyte donation programs in the Society for AssistedReproductive Technologies (SART).

Design: Descriptive data from a mailed questionnaire.

Setting: Academic medical center.

Participant(s): In vitro fertilization programs in SART.

Intervention(s): Survey mailed to IVF programs in the SART registry.

Main Outcome Measure(s): Current practices, opinions, and genetic criteria for oocyte donor selection.

Result(s): Results from 159 of 229 (69%) eligible oocyte donation programs are described. Most centers(70%) completed fewer than 25 cycles and used both anonymous and directed donors. There was considerablevariability in limits on numbers of cycles or births allowed per donor, with many centers having no limits. Aswell, the use of well-established screening tests for genetic disorders, such as sickle cell anemia and cysticfibrosis, varied considerably. Consultation with a geneticist was possible at most (89%) centers and specif-ically mentioned by some centers as a means to help make decisions.

Conclusion(s): Most programs follow recommendations made by the American Society of ReproductiveMedicine (ASRM) for screening of gamete donors, but a significant percentage does not use well-establishedtesting. The widespread availability of genetic consultation should promote responsible screening practices.(Fertil Sterilt 1999;71:278–81. ©1999 by American Society for Reproductive Medicine.)

Key Words: Oocyte donation, guidelines, genetic screening

The American Society for ReproductiveMedicine (ASRM) and the American Collegeof Obstetricians and Gynecologists (ACOG)have issued guidelines for the use of oocytedonation and the screening methodology (1, 2).These guidelines include recommendations toexclude donors with a history of a major Men-delian disorder, a significant congenital malfor-mation of complex cause, or a history of afamilial disease with a significant genetic com-ponent. Both sets of guidelines advise exclud-ing donors who are heterozygous for an auto-somal recessive gene, which causes a diseasethat is common among members of the donor’sethnic group, such as sickle cell anemia inAfrican Americans and Tay-Sachs disease inJews of eastern European descent. They furtherrecommend exclusion of donors with a familyhistory of significant malformations (heart dis-ease, neural tube defects), major Mendelian dis-orders, or chromosomal abnormalities (1, 2). Be-

cause these criteria were quite general, weasked how most oocyte donation programswere implementing screening.

MATERIALS AND METHODS

We mailed a three-page questionnaire to all311 IVF programs registered with the Societyfor Assisted Reproductive Technologies(SART) in November 1996. A follow-up mail-ing went to nonresponders in January 1997. Wephoned each remaining center in September1997 to inquire whether they had an oocytedonation program.

The questionnaire asked about programsize, types of oocyte donation, allowable cyclenumbers, and limitations based on donor’s ageand family history. Screening policies, includ-ing use of specific tests and exclusion criteria,also were surveyed. Last, we inquired aboutphysician training and practice affiliation. All

Received May 26, 1998;revised and acceptedSeptember 30, 1998.Presented at the 53rdAnnual Meeting of theAmerican Society forReproductive Medicine,Cincinnati, Ohio, October18–22, 1997.Reprint requests: VivianLewis, M.D., University ofRochester Medical Center,601 Elmwood Avenue, Box668, Rochester, New York14642 (FAX: 716-273-4167;E-mail: vivian_lewis@urmc.rochester.edu.)* Department of Obstetricsand Gynecology.† Department of Genetics.‡ Department ofCommunity and PreventiveMedicine.

FERTILITY AND STERILITY tVOL. 71, NO. 2, FEBRUARY 1999

Copyright © 1999 American Society for Reproductive MedicinePublished by Elsevier Science Inc.

Printed on acid-free paper in U.S.A.

0015-0282/99/$20.00PII S0015-0282(98)00463-4

278

forms were coded, and results were tabulated by someonewho was blinded to program identity. We used the MicrosoftAccess database software package for data analysis andentry.

The Institutional Review Board of the University ofRochester Medical Center exempted this study from formalreview because of the nature of the data collected and thedegree of anonymity of the subjects.

RESULTS

We received responses from 159 programs indicating thatthey performed oocyte donation during the previous calendaryear. An additional 82 centers replied that they did not offerthis service or that they had not yet performed any oocytedonation. Assuming that all 70 of the centers that did notrespond actually did perform oocyte donation, then 69%(159 of 229) of all possible active oocyte donation programsprovided data on their screening practices.

Of the active programs, 95% (151 of 159) of respondentshave completed an obstetrics and gynecology residency, and85% (136 of 159) were fellowship trained in reproductiveendocrinology. These residencies were completed since1980 in 67% of the programs. Eighty-nine percent (142 of159) of centers have genetic consultation available to them.Most physicians (105 of 159) have university appointments.However, 46% of these appointments are part-time. Manyphysicians are in solo or group private practice (58%); 14%of these physicians also are employed by the university orhospital.

Seventy-one percent (113 of 159) of respondents reportedfewer than 25 oocyte donation cycles in the previous year.The other 46 programs performed 25–150 cycles in theprevious calendar year. No center reported performing morethan 150 cycles. Both anonymous and directed donor cycleswere performed by 90% (144 of 159) of centers. There were147 programs that performed anonymous donation, withonly 3 of these offering anonymous donation, exclusively.These centers performed fewer than 75 cycles per year (2 ofthe 3 performed,25 cycles).

Directed oocyte donation was performed by 156 pro-grams. For 12 of these programs, directed donation was theonly type performed. All of these were small programs (,25cycles per year). Most (140 of 159) centers reported thatthere were no babies born with congenital anomalies duringthe previous calendar year; the remaining centers saw a rateof 1%–3% of babies with birth defects, born to mothers whohad used donated oocytes.

Age limits were almost uniformly applied (156 of 159centers). The three programs that did not apply age limitswere small (,25 cycles/y), and two of the three performedexclusively directed oocyte donation. The age limits tendedto be more strict for anonymous donors than for directed

donors. An upper age limit of 30–34 years was used by mostprograms (85%) for anonymous donors, whereas 63% ofprograms used 35 years of age for directed donors.

Lower age limits were likely to be the same for directedand anonymous donors. Approximately half of all programsused 19–22 years of age as the lower limit (56% for anon-ymous and 48% for directed). A few programs had no lowerage limit for donors (15 directed and 4 anonymous).

The widespread application of age limits was in starkcontrast to the use of limits on cycle number and birthnumber. Only 49% (78 of 159) of programs had a policy oflimiting the number of births allowed per donor. The numberof births allowed per donor ranged from 1 to 10, with 56%of these programs using a limit of 3 or fewer. A large numberof programs (51%) did not limit the number of births perdonor.

Limitations on the number of births per donor also couldbe achieved through limiting the number of allowable dona-tions per donor. Ninety percent (132 of 147) of programsperforming anonymous donation had a policy limiting thenumber of donations per donor. Sixty-six percent (83 of 132)of these programs used a limit of one to three cycles peranonymous donor. There were 49 programs that set the cyclelimit at four or more.

Only 10% (15 of 147) of programs that offer anonymousdonation had no policy restricting the number of donationsallowed per donor. Most (13 of 15) of these programs set nolimit on birth number either. The programs with no policy oneither birth number or cycle number were mostly small, with11 of these programs completing,25 cycles each per yearand only 2 such programs completing 75–150 cycles each,per year. It was much more common to have no formalpolicy restricting birth number or cycle number for directeddonors. Forty percent (62 of 156) of centers had no limit oncycle numbers for directed donors, and 51% (80 of 156) hadno limit on number of births for directed donors.

We inquired about screening for specific genetic disor-ders. Most (72%) programs used standard tests for sickle cellanemia in potential donors who were African American. Asimilar percentage (77%) reported the use of Tay-Sachsscreening in women of eastern European Jewish descent.Only 22% of programs tested for cystic fibrosis, in theabsence of family history. Karyotypes were used to screendonors by a similar number of programs (19%).

We asked about the likelihood of using tests that are stillbeing developed for clinical use. Almost one third of thecenters indicated that they are very likely or somewhat likelyto use a test for genetic susceptibility to breast cancer. Asimilar proportion (28%) would use a test for a colon cancergene. The likelihood of using a test for fragile X syndromewas more popular, with 46% of centers indicating that theywere very likely or somewhat likely to use such a test.

We studied the use of family history as a screening

FERTILITY & STERILITY t 279

method, through a checklist of various disorders with agenetic component (Table 1). Some of these disorders werespecifically drawn from the ASRM guidelines. Many (54 of159) respondents wrote qualifying comments where yes orno answers were requested. Most often (35 of 54) the re-spondents wrote “it depends,” “maybe,” or “test further.”Not all programs have formulated a policy for several of thedisorders listed. Routine consultation with a clinical geneti-cist or a genetics counselor was employed by four programs.Some respondents (n5 6) indicated a desire to know howmany first-degree relatives or generations were affected.Disclosure of family history of genetic problems to therecipient was indicated by several programs, particularly indealing with the more controversial disorders with a strong

environmental component, such as alcoholism or problems thatdo not have severe health consequences, such as baldness.

Family history of a major autosomal dominant disease,such as Huntington’s disease or neurofibromatosis, wouldexclude a potential donor in most programs (76% and 67%,respectively). Hyperlipidemia in the family would exclude adonor from only 28% of programs. However, five additionalrespondents thought that it depends on the type of hyperlip-idemia or whether the donor herself was affected. Anotherfour programs allowed the recipient to choose after disclo-sure. Family history of a serious autosomal recessive disor-der, such as polycystic kidney disease or cystic fibrosis,would exclude a donor from approximately two thirds of theprograms (62% and 67%) polled. Many of the respondentssuggested testing the donor for carrier status.

The greatest variability among programs was seen inpolicies relating to family history of complex genetic dis-eases, in which there are multiple inheritance patterns. Fewerthan one third of the programs would exclude a donor for afamily history of seizures (17%), colon cancer (22%), orrecurrent spontaneous abortions (13%). Almost half of thosepolled would exclude donors for family history of chromo-somal abnormalities (48%) or congenital heart disease(42%). Family history of deafness or blindness was checkedas an exclusion by 32% and 46%, respectively. Severalprograms commented that they would like more informationon the etiology of the hearing or visual deficit. Familyhistory of schizophrenia or other major psychiatric diseasewas the only complex genetic disorder commonly agreed on(62%) as an exclusion.

DISCUSSION

Oocyte donation is a commonly performed procedure forwhich there are general guidelines covering the need toscreen potential donors for infectious and genetic disorders(1, 2). Because these guidelines are quite broad in someareas, we asked each SART center for their policies. Wefound that most centers follow recommendations fromASRM and ACOG for excluding donors on the basis of ageand family history of several common serious, geneticallytransmitted disorders.

Most centers reported that the incidence of congenitalanomalies did not exceed the rate seen in the general popu-lation. The reliability of these data is limited by the fact thatmost infertility practices do not deliver the women whoconceive through therapy nor do they routinely refer thewomen to high-risk obstetric centers. Nonetheless, theSART registry consistently has requested information onpregnancy outcome. The SART data do not suggest anincrease in the rate of congenital anomalies among the off-spring of oocyte donors either (3). Presumably, each report-ing program has established means of ascertaining the inci-

T A B L E 1

Genetically transmitted disorders, which constitute groundsfor exclusion from oocyte donation.

Disorder

Percentage ofprograms thatexclude for

family history*

Autosomal dominantBreast cancer† 34Colon cancer† 23Huntington disease 76Hyperlipidemia 28Neurofibromatosis 67Polycystic kidney disease (adult) 62

Autosomal recessiveCystic fibrosis 67Hemochromatosis 46Hemoglobinopathies 53Polycystic kidney disease (infantile) 62

X-linkedColor blindness 24Hemophilia or bleeding disorder 69Muscular dystrophy 70

MultifactorialCleft lip and/or palate 42Neural tube defects 55

ComplexAlcoholism 32Alzheimers 41Baldness 3Blindness 32Breast cancer 34Chromosomal abnormality 48Colon cancer 23Congenital deafness 47Congenital heart defects 42Diabetes Mellitus, before age 40 36Mental retardation (other than Down’s) 49Psoriasis 6Repeated spontaneous abortions 13Seizures 17Schizophrenia, or other major psychiatric disease 62

* First-degree relatives only.† Disorders that sometimes exhibit different modes of inheritance.

280 Lewis et al. Genetic screening of oocyte donors Vol. 71, No. 2, February 1999

dence of congenital anomalies in infants born as a result ofassisted reproduction techniques.

A significant minority of centers does not use well-estab-lished screening tests, such as testing for Tay-Sachs or sicklecell disease in a population at risk. In a survey of 16 spermbanks, Conrad et al. found a similar percentage (81%) ofprograms tested at risk populations for these disorders (4).Whether this is because of a desire to limit costs or becauseof a lack of information or understanding about the guide-lines or frequency of these problems cannot be determinedfrom available data.

Testing for cystic fibrosis is more controversial. Only22% of programs tested donors for cystic fibrosis, in theabsence of family history. This is almost identical to thefrequency of testing for cystic fibrosis among sperm banks(25%) in the Conrad survey (4). Both ACOG and the ASRMrecognize this as a controversial area and make no specificrecommendations (1, 2). Increased availability of this test maymake it more attractive to all gamete donation programs.

There were some common areas of departure from policyrecommendations. Approximately half of all donor oocyteprograms have no limitations on the number of births allow-able per anonymous donor. Many of these programs aresmall, so that the potential magnitude of this problem is notgreat. The number of programs without limits on birth num-ber for directed donors was slightly higher. This is not likelyto pose a problem because most directed donors will havecome to some mutual agreement with the recipient regardinglimits on birth or cycle number. The most recent estimatessuggest that the risk of inadvertent consanguineous matingsdoes not increase in a population until there are approxi-mately 25 offspring per donor, assuming a population of atleast 800,000 in an area served by the donor’s facility (5).

The ASRM guidelines recommend a limit of 10 offspringper donor (1). The ACOG Committee Opinion uses the moreliberal limit of 25 (2). Curie-Cohen (6) concluded that lim-itations on use of a sperm donor should be determinedlocally, based on population size. His mathematical modelsuggests that the risk of inadvertent consanguinity would notincrease unless a given donor fathered at least 292 offspring,in an average American population. As a practical matter,these numbers are far greater than one would see from atypical oocyte donation program; yet, the rationale behindminimizing this possibility is important. Many of the pro-grams without limitations on birth number will never exceedthe recommended limits on birth numbers because they limitnumbers of possible donations for a given donor. The pub-lished surveys on sperm banks note that there is controversyabout an acceptable limit on birth numbers, but almost allbanks recognize the need to have some limit (4).

Our data show that age limits were almost uniformlyapplied in selection of oocyte donors, in keeping with theguidelines from both ACOG and ASRM. Surveys of spermbanks indicate that age limits of some sort are almost uni-versal. The age limits for semen donors vary from 36 to 50years of age, reflecting controversy over evidence linkingpaternal age to autosomal dominant mutations (4). Amongoocyte donation programs, many programs use an upperlimit of 30–34 years of age, which may also reflect the dropin fecundity seen in women in their 30s.

Family history is an effective screening tool for somegenetic disorders and frequently was used to exclude certainpotential donors. Several diseases with classically Mendelianpatterns of inheritance, such as neurofibromatosis, Hunting-ton’s disease, and polycystic kidney disease, were groundsfor donor exclusion in 60%–75% of programs. The Conradsurvey suggests greater variability among sperm banks, witha 44%–96% rejection rate depending on the specific disorder(4). Again, the rationale for not eliminating these donorscannot be determined from available data.

In conclusion, most oocyte donation programs follownationally recommended guidelines, but there is room forimprovement: more liberal use of established testing, carefulattention to family history, and establishment of birth num-ber limits. Greater input from the geneticists affiliated withoocyte donation programs in designing criteria for accep-tance of a given donor probably would help. Such input willbecome increasingly important as the diagnostic tests forgenetic diseases expand in number and complexity.

Acknowledgments:We thank Grace Centola, M.D., for providing informa-tion about screening policies of sperm banks and Ms. Kathleen DeRider andMs. Wanda Rivers for data gathering and entry.

References1. American Society for Reproductive Medicine. ASRM guidelines for

gamete donation. Fertil Steril 1993;59:1S–9S.2. American College of Obstetricians and Gynecologists. Genetic screening

of gamete donors. ACOG Committee Opinion 192, 1997.3. Society for Assisted Reproductive Technology and The American Soci-

ety for Reproductive Medicine. Assisted reproductive technology in theUnited States and Canada: 1995 results generated from the AmericanSociety for Reproductive Medicine/Society for Assisted ReproductiveTechnology Registry. Fertil Steril 1998;69:389–98.

4. Conrad EA, Fine B, Hecht BR, Pergament E. Current practices ofcommercial cryobanks in screening prospective donors for genetic dis-ease and reproductive risk. Int J Fert Menopausal Studies 1996;41:298–303.

5. deBoer A, Oosterwijk JC, Rigters-Aris CA. Determination of maximumnumber of artificial inseminations by donor children per sperm donor.Fertil Steril 1995;63:419–21.

6. Curie-Cohen M. The frequency of consanguineous matings due to mul-tiple use of donors in artificial insemination. Am J Hum Genet1980;32:589–600.

FERTILITY & STERILITY t 281