Customize your JAMA Network experience by selecting one or more topics from the list below.
Xu K, Shi ZM, Veeck LL, Hughes MR, Rosenwaks Z. First Unaffected Pregnancy Using Preimplantation Genetic Diagnosis for Sickle Cell Anemia. JAMA. 1999;281(18):1701–1706. doi:10.1001/jama.281.18.1701
Author Affiliations: The Center for Reproductive Medicine and Infertility and the Department of Obstetrics and Gynecology, Weill Medical College of Cornell University, New York, NY (Drs Xu, Shi, Veeck, and Rosenwaks); and the Department of Reproductive Genetics, Wayne State University, Detroit Medical Center, Detroit, Mich (Dr Hughes).
Context Sickle cell anemia is a common autosomal recessive
disorder. However, preimplantation genetic diagnosis (PGD) for this
severe genetic disorder previously has not been successful.
Objective To achieve pregnancy with an unaffected embryo using in
vitro fertilization (IVF) and PGD.
Design Laboratory analysis of DNA from single cells obtained by
biopsy from embryos in 2 IVF attempts, 1 in 1996 and 1 in 1997, to
determine the genetic status of each embryo before intrauterine
Setting University hospital in a large metropolitan area.
Patients A couple, both carriers of the recessive mutation for
sickle cell disease.
Interventions Standard IVF treatment, intracytoplasmic sperm
injection, embryo biopsy, single-cell polymerase chain reaction and DNA
analyses, embryo transfer to uterus, pregnancy confirmation, and
prenatal diagnosis by amniocentesis at 16.5 weeks' gestation.
Main Outcome Measure DNA analysis of single blastomeres indicating
whether embryos carried the sickle cell mutation, allowing only
unaffected or carrier embryos to be transferred.
Results The first IVF attempt failed to produce a pregnancy. Of
the 7 embryos analyzed in the second attempt, PGD indicated that 4 were
normal and 2 were carriers; diagnosis was not possible in 1. Three
embryos were transferred to the uterus on the fourth day after oocyte
retrieval. A twin pregnancy was confirmed by ultrasonography, and
subsequent amniocentesis revealed that both fetuses were unaffected and
were not carriers of the sickle cell mutation. The patient delivered
healthy twins at 39 weeks' gestation.
Conclusion This first unaffected pregnancy resulting from PGD for
sickle cell anemia demonstrates that the technique can be a powerful
diagnostic tool for carrier couples who desire a healthy child but wish
to avoid the difficult decision of whether to abort an affected
anemia is one of the most common human autosomal recessive disorders.
It is caused by a mutation substituting thymine for adenine in the
sixth codon (GAG to GTG) of the gene for the β-globin chain on
chromosome 11p, thereby encoding valine instead of glutamic acid in the
sixth position of the globin chain. The frequency of sickle cell trait
(carrier status) among the African American population at birth is
about 8%, and the incidence of sickle cell anemia at birth is 0.16%,
or 1 per 625 births.1 Furthermore, the widespread presence of the sickle gene in other ethnic groups has also been
confirmed.2 For example, in urban centers in the United States, nearly 10% of patients with various sickling disorders
identify themselves as non–African American.3
Children affected with sickle cell anemia experience recurrent
episodes of pain (during sickle cell crises) and increased
susceptibility to potentially life-threatening conditions, including
bacterial infections, cerebrovascular accidents, and organ failure.
According to US statistics collected between 1981 and 1992, there were
6.8 deaths per 1000 African American children aged 1 to 4 years due to
sickle cell disease.4 Life expectancy for persons with
sickle cell disease varies and there is an age-related pattern in
mortality rates: a peak in patients younger than 5 years, with a
gradual increase starting in late adolescence.5 Progress in
the treatment of sickle cell disease has been slow.6 At present,
there is no satisfactory treatment for the
sickling condition, although blood transfusion may reduce the risk of a
first stroke in children7 and gene therapy holds promise for a curative approach.8
Early prenatal diagnosis of the disease is critical because it allows a
couple to consider pregnancy termination as an option. The first DNA
diagnostic procedure for prenatal purposes was reported 20 years
ago.9 Subsequently, it was recognized that the mutation itself affected the cleavage site of a restriction enzyme,
DdeI, that could recognize the DNA sequence of CTNAG
(N=A, T, C, or G). While DNA from a normal allele
(CTGAG) would be digested by the enzyme, DNA from an affected allele in
which A is substituted by T (CTGTG) would not.10,11 The
resulting differences between DNA fragment sizes can then be recognized
by electrophoresis, thus forming the basis for diagnosis. With the
advent of polymerase chain reaction (PCR), rapid DNA analysis methods
have become available, and these techniques are now widely used for
An alternative and powerful diagnostic tool for identifying
sickle cell status in embryos is preimplantation genetic diagnosis
(PGD), which became possible nearly a decade ago.17
Preimplantation genetic diagnosis takes advantage of assisted
reproductive techniques in conjunction with modern molecular methods.
With PGD, the genetic status of an embryo can be determined before
transfer into the uterus after in vitro fertilization (IVF), thus
eliminating the risks of bearing a child with the disease.
While PGD for sickle cell anemia has been performed in the mouse
model,18 routine clinical application in the human
previously has not been successful. In this article, we describe our
experience using PGD to determine the precise genetic status of embryos
generated by assisted reproduction for a couple who are heterozygous
carriers of the sickle cell mutation.
A 34-year-old female patient had undergone 2 previous induced abortions
because she was carrying fetuses affected with sickle cell anemia.
Genetic diagnosis indicated that both female and male partners were
carriers of the sickle cell mutation. After extensive counseling, the
couple gave informed consent and elected to undergo preimplantation
genetic diagnosis. The study was approved by the Weill Medical College
of Cornell University (New York, NY) Institutional Review Board.
To confirm the genetic status of the couple and establish a protocol
for single-copy gene amplification from single cells, blood was
collected from both partners. Lymphocytes were isolated using
Ficoll-Paque density gradient separation (Pharmacia Biotech Inc,
Piscataway, NJ) with the protocol provided by the manufacturer. Single
lymphocytes were loaded into 0.5-mL tubes containing 5 µL of lysis
buffer19 and stored at −20°C before trial testing.
On the day of preliminary trial testing, sample tubes were
removed from the freezer and heated to 65°C for 10 minutes before
they were placed back on ice. Five microliters of neutralization buffer
was added to each tube. A nested PCR approach was used for the
amplification of the region surrounding the sickle cell mutation. The
primers used have been described previously.17,19
Polymerase chain reaction was performed after adding a standard mixture
of all components, including 2.5 mmol of Mg2+, 0.2 mmol of
dNTPs (containing dATP, dCTP, dGTP, and dTTP, Perkin Elmer, Foster
City, Calif), 100 ng of primers, and 2 U of Taq polymerase (AmpliTaq,
Perkin-Elmer). A hot start at 95°C was applied for 3 minutes to
ensure complete denaturation of the template. For the PCR profile, the
following parameters were used: 93°C denaturation for 30 seconds;
50°C for 40 seconds for annealing; and 72°C for 45 seconds for
extension. A total of 20 amplification cycles were applied for outer
primers. For the inner primer set, identical parameters were used,
except that the annealing temperature was raised to 55°C and a total
of 40 cycles were used.
Following the nested PCR amplification, 18 µL of amplified product
was digested with the restriction enzyme DdeI (GIBCO/BRL,
Rockville, Md) for 3 hours. Subsequently, 10 µL of digested product
was run on a 10% acrylamide gel. On completion of electrophoresis, the
gel was stained with ethidium bromide and photographed by UV
transillumination. As predicted, unaffected DNA showed 3 bands (201,
90, and 74 base pairs [bp]), carrier DNA showed 4 bands (291, 201,
90, and 74 bp), and an affected homozygous sample showed only 2 bands
(291 and 74 bp). Testing of single lymphocytes from the male and female
subjects showed the same predicted patterns (4 bands of predicted
sizes), together with an unaffected DNA control (3 bands) (Figure 1).
The IVF procedure has been described previously.20 Briefly,
to ensure that several embryos would become available for DNA analysis,
multiple ovarian follicular development was initiated with gonadotropin
therapy. After pituitary desensitization with gonadotropin
hormone–releasing hormone agonist (leuprolide acetate, TAP
Pharmaceutical, Chicago, Ill), ovarian stimulation was begun on day 3
of the ensuing menstrual cycle using intramuscular administration of a
combination of urofollitropin (75 U of pure follicle-stimulating
hormone) and menotropins (150 U of follicle-stimulating hormone and
luteinizing hormone) (Serono Laboratories, Norwell, Mass). Follicular
growth was monitored by daily serum estradiol levels and pelvic
ultrasonograms. To induce final oocyte maturation, 3300 IU of human
chorionic gonadotropin was administered when 2 follicles of 18 mm in
average diameter were observed on ultrasonogram. Transvaginal oocyte
retrieval was performed 35 hours later. To avoid sperm contamination
and possible amplification of sperm DNA, intracytoplasmic sperm
injection was used. After 16 hours of incubation, fertilization was
confirmed by the identification of 2 pronuclei. Normally fertilized
concepti were then transferred to droplets of human tubal fluid (made
on site) supplemented with 15%
maternal serum under mineral oil (ER Squibb & Sons
Inc, Princeton, NJ). Biopsy was performed on the morning of the third
day after harvest. All embryos were maintained at 37°C in an
atmosphere of 5% carbon dioxide. Cleavage rate and morphologic
appearance were assessed daily.
Blastomere biopsy was carried out in the early morning, approximately
65 hours after oocyte collection. Briefly, a holding pipette was used
to stabilize the embryo (at the 9 o'clock position). A hole was made
at the 3 o'clock position by expelling a small amount of acidified
Tyrode solution (pH, 2.35) onto the zona pellucida through a small-bore
pipette. Reverse suction was applied as soon as a hole of appropriate
size was created to reduce possible damage caused by exposing the
blastomeres to the acidic solution. A large inner-diameter biopsy
pipette replaced the pipette containing the acidified Tyrode solution.
Subsequently, 1 or 2 cells were aspirated, depending on the total cell
number of the embryo. Blastomeres were rinsed in biopsy medium 3 times
before loading into a PCR tube containing 5 µL of lysis buffer. Tubes
were processed and PCR amplification and restriction enzyme analysis
were performed as described herein.
Micromanipulated embryos were further cultured in medium droplets
overnight. Embryo transfer was performed in the afternoon of day 4.
Pregnancy was determined by serum β–human chorionic gonadotropin
measurement on cycle days 28 and 35, followed by ultrasonographic
assessment at 7 weeks' gestation. The genetic status of the fetuses
was confirmed after amniocentesis by an independent laboratory.
Cultured amniocytes were also sent to our laboratory for follow-up PCR
Polymerase chain reaction and restriction enzyme analysis of single
lymphocytes from both the female and male partners clearly indicated
that each carried the sickle cell mutation. Forty-six of 48 single
lymphocytes, 24 from the female and 24 from the male, were successfully
amplified. As predicted, 4 bands of correct size were obtained from
both partners. An example of the gel is shown in Figure 1.
During the first IVF attempt in November 1996, 18 oocytes were
retrieved. Four of 6 mature oocytes were normally fertilized after
single-sperm injection by intracytoplasmic sperm injection, yielding 4
Embryo biopsy was performed on all 4 embryos on day 3 by removing
a single blastomere from each conceptus. Polymerase chain reaction and
restriction enzyme analysis revealed that 1 was homozygous unaffected,
2 were carriers, and 1 was homozygous affected. Transfer of 1
unaffected embryo on day 4 failed to result in a pregnancy.
The second IVF attempt was initiated in August 1997, during which 16
oocytes were retrieved. Of those, 8 were mature and underwent
intracytoplasmic sperm injection. Seven concepti cleaved at least once
by the following day. On the morning of the third day, 2 cells were
removed from 2 embryos and 1 cell from 5 embryos. Polymerase chain
reaction amplification was successful in 7 of 9 blastomeres (5 of 6
embryos). Amplification failed in 1 cell from embryo 5 and in 1 cell
from embryo 6. Restriction enzyme digestion demonstrated that 4 were
homozygous unaffected, 2 were carriers (Figure 2), and 1 was of unknown status due to PCR
amplification failure. On the afternoon of day 4, all concepti that
underwent biopsy demonstrated further cleavage. Selection of embryos
for transfer was based on PGD diagnosis, growth rate, and morphology.
Two unaffected embryos were of poor quality, displaying slow cleavage
and multinucleation (embryo 2) or high fragmentation (embryo 4) and
therefore were not suitable for transfer (Table 1). Because there were only 3
high-quality transferable embryos—2 unaffected (embryo 1 with 15
cells; embryo 6 with 8 cells) and 1 carrier (embryo 8 with 10 cells) (Figure 3 and Table 1)—and because
the patient was willing to accept a fetus of carrier status, all 3
embryos were transferred.
A twin pregnancy was confirmed by ultrasonography at 7 weeks.
Amniocentesis performed by an independent laboratory
at 16.5 weeks revealed that neither fetus harbored the sickle cell
mutation. DNA analysis of amniocytes shipped from the prenatal
diagnostic laboratory to our own laboratory also showed that both
fetuses were unaffected (Figure 4). The patient delivered healthy, unaffected fraternal twin girls at 39
After natural conception, couples who carry autosomal recessive
mutations risk a 25% chance of delivering an affected child, and half
of the offspring may carry the mutation. Although prenatal testing is
currently available, some couples have strong personal objections to
aborting affected fetuses. For these couples, PGD provides a realistic
alternative to prenatal testing.
Although the first pregnancy achieved by PGD for sex determination to
avoid the transmission of a sex-linked disorder occurred nearly a
decade ago,17 PGD for single-gene defects is
still in the
experimental stage because of its complexity and technical
difficulties. At present, PGD is primarily applied for severe genetic
disorders for which detailed genetic information is available. Normal
pregnancies following a search for specific mutations have been
reported for only a few genetic diseases, including cystic
fibrosis22-24 and Tay-Sachs disease.25
Despite the fact that sickle cell anemia is one of the most common
genetic disorders and detailed genetic information is
available,1,26 unaffected pregnancies following PGD for
sickle cell anemia previously have not been reported. Lack of previous
success in this area presumably is due to the length of time and effort
required to overcome technical difficulties inherent in these
procedures, as well as lack of available research funding. Our results
demonstrate that sickle cell anemia can be detected in single cells by
PCR and restriction enzyme analysis and that unaffected pregnancies can
be established by the transfer of embryos of known genetic makeup that
have undergone biopsy.
The protocol used in this study was initially developed in the
mouse model by Sheardown et al.18 In their investigation, 4
tandem copies of the human β-globin gene were detected in transgenic
mouse embryos. In humans, β-globin is a single-copy gene. Under
clinical PGD circumstances, single-cell PCR requires an extremely
sensitive protocol. In this study, we initially tested the protocol on
single lymphocytes isolated from each member of the couple at risk.
Lysis buffer, reported to be better for single-cell PCR,19
instead of the freeze-thaw method.18
Using the modified protocol, successful amplification was achieved in
46 (96%) of 48 single lymphocytes tested. This provided invaluable
preliminary technical experience prior to executing the PCR-PGD
technique in the clinical setting. However, PCR failure occurred in 1
of the 9 cells undergoing biopsy. Diagnostic failure in PCR-PGD could
be due to a number of factors, including the absence of a nucleus, loss
of the blastomere during handling, and blastomere
mosaicism.27 While the current technique is apparently
adequate, further modification and enhancement, such as the use of
fluorescent PCR, may further improve the accuracy and efficiency of
We performed biopsies on day 3 and intrauterine transfer on day 4
because the biopsy, PCR, enzyme digestion, and electrophoresis could
not be completed within 10 hours. This change provided additional
valuable hours for accurate diagnosis. Apparently, the extended in
vitro culture did not compromise embryo viability. Because both the
morphology and genetic status of each of the transferred embryos were
known, it is worth examining the characteristics of each embryo. On the
morning of day 3, embryo 6 (Table 1) was composed of 7 blastomeres, 2 of
which were removed for testing. This accounts for more than one fourth
of the embryo volume. At the time of intrauterine transfer on day 4,
the conceptus had reached the 8-cell stage. This embryo implanted, as
indicated by its genetic status (βA/βA). It
is known from animal models that biopsy does not decrease implantation
and subsequent live birth rates28 and that embryo biopsy
does not necessarily impair subsequent in vitro development in
humans.29 This particular case
unequivocally shows that the
removal of 2 blastomeres from a 7-cell conceptus did not compromise its
developmental potential. Furthermore, our results confirm that day-3
biopsy and day-4 transfer is indeed a feasible approach for PGD. Recent
progress in developing culture media30,31 and/or human
autologous endometrial coculture32,33 may even further
enhance embryo viability, thus increasing the pregnancy rate after PGD.
In summary, this is the first unaffected pregnancy and delivery
after successful PGD for sickle cell anemia. Our results demonstrate
that PGD for the detection of sickle cell anemia is a powerful
diagnostic tool for carrier couples who desire a healthy child but wish
to avoid the difficult decision of whether to abort an affected fetus.
The procedure, successfully used in this case, may also be applied to
other monogenic disorders and further supports the notion that PGD is
destined to be an integral
aspect of assisted reproductive technology.
Given the current methods and relatively high cost of the procedure, it
is unlikely that PGD will totally replace prenatal testing. However, it
is conceivable that with further refinements, PGD will certainly become
an invaluable and powerful diagnostic modality.
Create a personal account or sign in to: