Abstract
Since the introduction of techniques to diagnose genetic diseases in the unborn child, the main application has been to test for severe, untreatable conditions as early in the pregnancy as possible, to allow parents to consider termination of pregnancy.
Although still rare, an increasing number of genetic diseases can now be treated prenatally, or their prognosis can be improved by prenatal or early postnatal interventions. These promising developments radically change the prenatal screening and diagnosis landscape. Acceptance of prenatal screening and diagnosis may positively change, when the focus moves away from just termination as an option in case of an abnormal result. However, pretest and posttest choices for prospective parents become more complicated. In this chapter, we discuss the currently available and newly emerging methods for intervention in fetal genetic diseases, with emphasis on the published human experiments. We also highlight the complex counseling, decision-making, and ethical implications of these innovations.
Keywords
Fetal therapy, Genetic disease, Stem cell therapy, Gene therapy, Pregnancy, Prenatal diagnosis, Prenatal screening, Prenatal treatment
Introduction: Benefits of Prenatal Genetic Diagnosis Beyond the Option of Termination
For several decades, prenatal genetic testing was available to pregnant women in three different situations. Traditional karyotyping was offered to a large group of women at increased risk for trisomy 21, based on population screening using maternal age, serum marker testing, ultrasound or a combination of these. Chorionic villus sampling (CVS) with DNA analysis was offered to couples at high risk for a fetus with dominant or recessive genetic condition. The third situation was karyotyping after the diagnosis of fetal structural anomalies on ultrasound examination. In the majority of cases in all three groups, an abnormal result led to the request by the parents to terminate the pregnancy. In case of potentially treatable fetal structural anomalies, an abnormal karyotype was considered a contraindication to fetal intervention.
Many people believed that with the completion of the Human Genome Project, in the year 2003, we would enter an era of diagnosing, and even curing, many or most genetic diseases. Although since then, enormous progress has been made in technology, leading to both much faster and much cheaper sequencing, clinical application is still mostly limited to testing predisposition for genetic diseases, including certain types of cancer. Major advances have been made in genetic testing methods, and most new modalities are now available for prenatal diagnosis and screening as well. Although a small role for traditional karyotyping remains, testing for fetal diseases is now mostly done using QF-PCR for rapid aneuploidy evaluation, and chromosomal microarray for detailed analysis. Whole exome sequencing is currently introduced in perinatal clinical practice as well, further refining our diagnostic capabilities. Hudecova published a method for noninvasive prenatal diagnosis of inherited single-gene disorders using droplet digital PCR from circulating cell-free DNA in maternal plasma . Noninvasive genetic diagnostic testing, safe and therefore more acceptable to offer to a broad group of pregnant women, is entering the clinic.
The development of treatment options often lag behind the advances in diagnostic testing. This is especially true for fetal diseases, where both high definition ultrasound and high-resolution chromosome analysis enable early and accurate detection of many pathologic conditions. If at all possible, fetal diseases are preferably treated after birth, when direct access to the patient is possible, and risks for the pregnant woman are absent. In a number of fetal conditions however, progressive disease may lead to demise before birth, or to irreversible damage and handicaps at the time of birth. In such diseases, prenatal intervention can be considered to save the child’s life or to improve the prognosis for its future health. Any prenatal intervention has the potential to harm the mother, or to inadvertently harm the fetus in case of complications or failure. In particular, the possible harm to the mother makes fetal therapy a challenging subspecialty. In addition, unlike in most other areas of medicine, termination of the pregnancy may be an alternative to prenatal or postnatal treatment. Thus when considering fetal intervention, potential benefits to the fetus also need to be balanced against potential harm to the fetus and to risks to the pregnant woman. If an attempt to save the life of a fetus, or to prevent irreversible damage, leads to the birth of a severely handicapped child, either because of complications of failure of the prenatal treatment, the child and its family may suffer more than with expectant management. The risk of “trading mortality for morbidity” is a valid argument against some forms of fetal therapy. We will explore this aspect in more depth in the part on ethics.
The “classic” forms of fetal therapy are directed at acquired, and not genetic, fetal diseases, such as Rh-alloimmunization or twin-twin transfusion syndrome. In these conditions, the fetus is essentially healthy, only threatened to die from a “hostile environment.” Successful prenatal intervention in these conditions often leads to the birth of healthy children with a normal long-term outcome. In genetic diseases, the underlying cause, a DNA mutation, is in most cases present in every cell of the body. Until recently, treatment options were limited to reducing the consequences of the abnormal organ functions or alleviating symptoms. Early versions of gene- and stem-cell therapy were aimed mostly at addition of genes or cells with normal function to fetal organs, and/or inducing tolerance for safer postnatal transplantation. Repairing the DNA itself, gene correction , is now on the horizon, using DNA cutting techniques such as Cas9, zinc-finger, or TALE nucleases. In this chapter, we will focus on currently or likely soon available prenatal treatments for fetal genetic diseases.
Symptomatic Treatment
Inherited or de novo mutations affecting erythropoiesis, red cell morphology, or the structure/function of hemoglobin may all lead to fetal anemia, which as a final common pathway can cause fetal heart failure, hydrops, and death. Irrespective of the underlying cause, fetal anemia can effectively and quite safely be treated by fetal blood transfusion, a method developed already in the early 1960s. This purely symptomatic treatment, often given multiple times during pregnancy, can be lifesaving. The diagnosis of genetic hematologic diseases in a fetus is made either by testing because of a family history or by workup after the incidental finding of fetal hydrops. To illustrate the complex choices and consequences doctors and parents can be confronted with, the example of fetal treatment of homozygous alpha-thalassemia will be explored further.
Alpha-thalassemia major or Hb Bart’s disease is the most common inherited disorder of hemoglobin synthesis in Asia. It is caused by a deletion of all four “alpha genes” on chromosome 16. The gene mutation (a deletion) frequency in Southeast Asia is around 4.5%, resulting in a high prevalence of the homozygous mutation . Traditionally, this is considered a lethal disease due to deletion of all alpha-globin genes. In early fetal life, embryonic hemoglobin types can deliver sufficient oxygen to the tissues, however later on, hemoglobin consisting of four γ chains (Bart’s) becomes the dominant fetal hemoglobin. Hemoglobin Bart’s has is a very high oxygen affinity, compromising oxygen delivery to the tissues. Screening programs identify carriers, and, in at risk couples, noninvasive fetal testing can be performed by ultrasonography aiming at finding early markers such as cardiomegaly, or by invasive testing and genetic analysis of chorionic villi. Cell-free DNA testing from maternal plasma for thalassemia is being developed . In the vast majority of cases, detection of fetal homozygous alpha-thalassemia within such screening programs will lead to termination of pregnancy.
Some early examples of fetal (intrauterine) intravascular blood transfusion for homozygous alpha-thalassemia were cases where the cause of fetal anemic hydrops was unknown, and diagnostic fetal blood sampling was combined with blood transfusion, to gain time for the diagnostic workup . Then when the diagnosis was made, parents requested continuation of the transfusion therapy, both prenatally and postnatally, to keep their children alive. Technically, this has been a relatively safe and, in fetal therapy centers, common procedure for fetal anemia treatment already for decades. Most often however, it is performed for red cell alloimmunization and parvovirus infection, two acquired fetal diseases carrying an excellent prognosis when treated timely. The main dilemma in alpha-thalassemia is the quality of life of the surviving children. As will be discussed in more detail later in this chapter, doctors feel a responsibility for the burden of the future child. The general view of fetal medicine specialists around the world is that homozygous alpha-thalassemia should still be regarded as a lethal disease, for which termination of pregnancy is justified.
A literature review identified 15 children with homozygous alpha-thalassemia surviving after fetal transfusions . In 5/15, hypospadia was present. Long-term neurodevelopmental outcome, with follow-up between 3 months and 7 years, revealed four children had mild and one had marked psychomotor impairment. Since insufficient tissue oxygenation is already present early in fetal life, fetal transfusion therapy, which at the earliest can be performed from 16 to 18 weeks’ gestation, may not be able to prevent neurologic damage . After birth all children required blood transfusions at 3–4-week intervals, and 9/15 were treated with bone marrow/stem cell transplantation. The morbidity related to iron overload and side effects of medication accompanying transplantation is considerable and significantly worse than in β-thalassemia.
This example of a technically feasible symptomatic treatment of fetuses with a serious gene mutation illustrates the main dilemma; the fetal symptomatic treatment may save the child’s life, but it is no cure, and the survivors remain patients needing lifelong treatment, with far from normal lives.
Supplementation Therapy
Genetic diseases generally described as inborn errors of metabolism often originate from (single) gene mutations leading to absent or abnormal enzyme, hormone, or protein function. The mode of inheritance is often autosomal recessive. In particular in metabolic diseases where one specific essential substance is lacking, replacement therapy may solve the clinical problem. In others, preventing the individual to take food containing substances the body cannot process normally, leading to dangerous levels of a toxic product, “nutrition management” may be the solution. The best-known example is phenylalanine hydroxylase deficiency, leading to phenylketonuria (PKU). During pregnancy, the unaffected (heterozygous) carrier mother may provide essential substances via the placenta, and in this way reduce the risk of harm for the fetus with this metabolic disorder. After birth however, completely asymptomatic but affected (homozygous) neonates may soon suffer from their metabolic disorder. Likewise pregnant women with PKU need strict dietary control at least 1 month prior to conception and during the whole of pregnancy, as well as monitoring of phenylalanine levels, to avoid the severe and irreversible teratogenic effects of hyperphenylalaninemia . Most of the rare diseases screened for with the neonatal dried blood spot test by the “heel prick” fall in this category. In some genetic mutations leading to errors of metabolism, the abnormal function of an enzyme or hormone may lead to irreversible damage already before birth. In such cases, prenatal intervention may improve the outcome and prognosis. Some have argued that several of the diseases screened for by the heel prick should actually be tested for already during pregnancy, to enable earlier intervention . Examples of prenatal fetal treatment of metabolic disorders such as methylmalonic acidemia and multiple carboxylase deficiency have already been given in the 1970s and 1980s . With our increasing ability to screen for fetal genetic alterations during pregnancy, it is expected that a much larger number of “inborn errors of metabolism” will be detected, followed by more research into suitable treatments. We speculate that very few parents (and Institutional Review Boards) are willing to go the uncertain route of experimental fetal treatment. The options and dilemmas will be illustrated by some examples.
Methylmalonic acidemia (MMA) is due to deficient synthesis of 5′-deoxyadenosylcobalamin, which can lead to fetal growth restriction and dilated cardiomyopathy. Children suffer from vomiting, failure to thrive, acidosis, and mental retardation. In 1975, Ampola et al. reported prenatal treatment by vitamin B12 in an attempt to diminish the accumulation of methylmalonic acid to a pregnant woman who lost a previous child to MMA . Dose and response were monitored using maternal urinary excretion of methylmalonic acid. Postnatally, the child was treated with dietary measures. At 19 months, the child was developing normally. In the following decades, a few more apparently successful cases have been published. In 2016, the long-term follow-up was published of a girl with Cobalamin C deficiency (leading to MMA as well as hyperhomocysteinemia) treated in utero by giving the pregnant mother a weekly dose of hydroxycobalamin in 2005, and her untreated older sister . At the age of 11 years the girl had a normal IQ of 103, attended secondary school, and only mild ophthalmic findings. The affected older sister, with the same genotype and similar residual enzymatic activity, received postnatal treatment only and was severely intellectually disabled, cannot walk or talk, shows aggressive behavior, and has severe ophthalmic symptoms. Due to the rarity of the disease and the many genetic variants, it remains difficult to establish optimal dosage and to predict clinical outcome, especially on the long term.
Multiple carboxylase synthetase deficiency presents in the newborn with severe metabolic acidosis. It is treated with lifelong biotin therapy. However, the fetus may already suffer from growth restriction as well as ventriculomegaly . The first case report of prenatal treatment, by maternal oral biotin supplementation, was published in 1982 . Dosage is still a problem, as illustrated by the latest published case where the investigators used 10 mg biotin daily based on a few previous reports. Fetal growth accelerated after biotin treatment, but the child was born with significant lactic acidemia and metabolic acidosis .
3-Phosphoglycerate-dehydrogenase deficiency is an amino acid (serine) synthesis disorder, associated with microcephaly, severe psychomotor retardation, and intractable seizures. In 2004, de Koning et al. from Utrecht, The Netherlands published a case of prenatal maternal administration of supplements of l -serine, three doses of 5 g l -serine (190 mg/kg) per day, starting at 26 weeks when the growth of the head circumference was decelerating . The child had low levels of serine at birth, but was born with a normal head size, and at the age of 2 developed normally, in sharp contrast with its two affected older siblings.
Fetal Genetic Thyroid Disease
Fetal goitrous hypothyroidism is usually secondary to abnormal maternal iodine intake or autoimmune thyroid disorders. However, the fetus itself may have a genetic mutation causing defects in the thyroglobulin synthesis, leading to primary fetal hypothyroidism. Mutations likely causing congenital hypothyroidism have been described in genes coding for enzymes involved in regulating thyroid functions, such as thyroid oxidase 2. Unless there is a family history, this may go unnoticed before birth. When a fetal goiter develops, complications such as polyhydramnios (due to impaired swallowing) and malposition may occur. Postnatally, the goiter may cause airway problems. Ultrasound can easily detect the thyroid enlargement, and amniocentesis or fetal blood sampling can confirm hypothyroidism. Fetal therapy for goiter was already described in 1980 . The largest single center series of prenatal treatment of fetal goitrous hypothyroidism in euthyroid mothers is from France . The authors described 12 cases where thyroxin was given intraamniotic, with good clinical response defined as reduced size of the goiter. However, most fetuses remained hypothyroid, with sequelae such as jaundice, open posterior fontanel, hypotonia, and delayed bone age. A literature review published in 2017 found 32 in utero treated cases, with evidence of benefit by preventing mechanical difficulties due to the goiter in 28 . The authors confirmed inability of reaching a euthyroid state at birth. In 2017, Vasudevan et al. published a case with fetal goiter and polyhydramnios, in which intraamniotic T3 and T4 was given. The fetus died at 31 weeks. The therapy may not be risk free. Although in a few cases with untreated siblings, the IQ of the prenatally treated children was slightly higher, long-term outcome data are lacking to suggest benefit on neurodevelopmental outcome. Therefore intraamniotic thyroxine injections may have a place, if success of prenatal intervention is defined as size reduction of a large goiter, enabling the fetus to swallow and the neonate to breath at birth. Parents and clinicians should, however, be aware that the fetus likely has suffered from prolonged hypothyroidism, and irreversible sequelae can already be present at the start of fetal treatment.
In conclusion, fetal (and maternal) metabolic disorders may have irreversible effects on fetal health. This justifies attempts to treat the condition before birth. Studies to find the optimal dose of medication needed to reach appropriate fetal tissue levels are difficult to perform, given the rarity and variability of each disorder. Most published cases had a previous severely affected sibling, with better outcome in the prenatally treated pregnancy. Even with the same mutation, clinical presentation may vary between sibs. Improvement may be not only due to the prenatal medication, but also to other aspects of improved care for the second affected child in a family, both pre- and postnatally. Randomized or otherwise properly controlled studies in this field are virtually impossible to perform.
Causal Therapy I: In Utero Stem Cell Transplantation
In addition to the rationale of prenatally treating a disease that may cause death or irreversible damage already before birth, there is the potential advantage of the window of fetal immunologic immaturity enabling transplantation or injection of unmatched donor cells. Stem cell transplantation in children and adults is useful in many diseases, but one of the difficulties is that careful matching and immunosuppression are needed to reduce the risk of rejection and graft-versus-host disease (GVHD). Transplantation early in fetal life may be safe and effective without immunosuppression, and even if engraftment is very low, there may be induction of donor-specific tolerance, enabling safer postnatal transplantations from the same donor. The fetus is so to say an immunologically “naïve” target. Proof for this concept has been given since the 1950s . The use of less immunogenic fetal (liver) stem cells (from abortion material) as donor material was already proposed in the 1980s . Also, the fetus is still small, and only requires a small transplant. It has relatively “empty” bone marrow with space for new transplanted cells. Before birth migration of stem cells to bone marrow and other target organs is a natural phenomenon. The transplanted donor stem cells may “piggyback” on this system, but still need to compete with the (defective) host cells in the target organs.
Although for decades, hopes were high that this type of fetal treatment would be successful in the approach of many genetic diseases, in vitro and animal research have not yet resulted in established clinical applications. One major barrier became apparent, when Peranteau et al. demonstrated that even the young fetus has a functioning immune system . Some examples of human trials of fetal stem cell therapy are discussed as follows.
Immunodeficiency Disorders
Inherited severe immunodeficiency disorders are ideal targets for stem cell transplantation: rejection and GVHD are no issue; the absence of immune reaction is the disease itself.
In 1988, Touraine et al. from Lyon, France did a human proof-of-principle experiment: they injected fetal liver and thymus cells into the umbilical cord of a 30-week fetus, diagnosed with bare lymphocyte syndrome, a combined immunodeficiency disorder with absent expression of HLA antigens . At 1 month of age, 10% of the newborn’s lymphocytes expressed HLA class I. The child received multiple fetal stem cell transplantations and was still in isolation at 7 months of age.
In 1995, Flake et al. published the first successful in utero stem cell transplantation, in a case with X-linked recessive severe combined immunodeficiency (SCID) . The parents’ first son had died from the disease. In the next pregnancy, CVS revealed again an affected son. This fetus received three stem cell transplantations between 16 and 19 weeks’ gestation. The cells were harvested from his fathers’ bone marrow. There was a high level of donor cell engraftment, and normal growth and development at 11 months of age.
Genetic Hematologic Disorders
Hemoglobinopathies, in particular sickle cell anemia and thalassemia, were among the first diseases thought to be candidates for in utero stem cell transplantation.
A 2017 review article counted a total of 26 reported cases of intrauterine stem cell transplantations for hematopoietic disorders in human fetuses . Only in the three cases of immune deficiency disorders, where the lack of fetal immunity causes selective advantage for donor cells, there was clinically significant donor cell chimerism None of the other attempts, for various conditions including Rh disease, leukodystrophy, Chediak-Higashi, thalassemias, Niemann-Pick type A, and Hurler syndrome, were deemed successful, success being defined as sustained donor cell chimerism sufficiently high to ameliorate the disease phenotype. Most transplantations, often intraperitoneally, were done after 14 weeks’ gestation, when fetal immunity has already developed, and fetal blood sampling for evaluation of donor cell expression was, with the current knowledge, done too early. Of the 26 fetuses, 4 died unintentionally before birth, and 2 were terminated because of lack of successful engraftment. Current knowledge from animal work suggests that more time may be needed before engraftment can be established, and that intravascular injection (instead of, e.g., intraperitoneally, which was done in some early human cases) is the most successful route of administration. In none of the reported cases overt GVHD occurred. The disappointing engraftment was in part found to be due to fetal immunity, which in contrast to earlier views, is present already early in fetal life. To overcome this barrier, autologous or maternal/paternal transplantation could be used. Another potential improvement could be the use of (fetal donor) mesenchymal stroma cells.
Osteogenesis Imperfecta
Osteogenesis imperfecta (OI) is a group of incurable genetic disorders caused by > 1350 different dominant and > 150 recessive mutations in collagen genes. Seven types have been described. Mutated COL1A1 and COL1A2 are the genes found in most individuals with OI. These mutated genes cause an abnormal production of type I collagen. The abnormal collagen mainly leads to abnormal “brittle bones,” resulting in fractures. Other clinical sequelae are deformities, growth problems, protruding eyes, and deafness. Some forms have the typical “blue sclerae.” Type I is mild, type II is lethal, types III and IV are not lethal but often severe, with fractures already occurring in utero. These can be seen on ultrasound. Postnatal treatment is symptomatic (mostly by bisphosphonates) and aims at preventing fractures.
In 2002, Westgren, Le Blanc and colleagues from the Karolinska Institutet in Stockholm, Sweden performed the first prenatal transplantation of human fetal liver mesenchymal stem cells with a successful outcome in an immunocompetent fetus suffering from OI type III . Fetal mesenchymal stem cells from a fetal liver were injected in the umbilical vein at 31 weeks. There was long-term donor cell engraftment. Until the age of 8, the patient only suffered one fracture and one vertebral compression fracture per year. At the age of 8 years, the girl received another stem cell transplant from the same donor. Over the following 2 years, she did not acquire any fractures and her linear growth and mobility improved. A child with an identical mutation that did not receive stem cell transplantation exhibited a very severe phenotype of OI and died at the age of 5 months. In recent years, two other fetuses with OI were treated with mesenchymal stem cells with assistance of the same group of investigators. Outcomes thus far are promising.
Fetal Transplantation and Immunology Society
In 2015, a consensus statement was published on in utero stem cell transplantation . An international organization was founded, FeTIS (Fetal Transplantation and Immunology Society, www.fetaltherapies.org ), which will develop a registry of patients treated and outcomes. The international expert community agreed that in utero transplantation is a viable option for selective congenital disorders. Beside its success in SCID and similar diseases, the advantage of in utero stem cell transplantation is the development of donor-specific tolerance. A chimerism of 1%–2% may be sufficient to reach this. A few other diseases may also benefit from low-level engraftment. According to FeTIS, the main target diseases now suitable for human phase I trials are sickle cell disease, thalassemia, and osteogenesis imperfecta.
Two human in utero stem cell trials are currently (2018) ongoing. The first is the BOOSTB4 trial, an EU Horizon 2020 program supported international multicenter phase I/II study ( www.boostb4.eu ). The investigators, led by the Karolinska group, aim to perform 15 prenatal and 15 postnatal-only fetal mesenchymal stem cell transplantations for OI type III and severe type IV.
The other study is a single center phase I clinical trial combining maternal stem cell transplantation with intrauterine blood transfusion in alpha-thalassemia major, led by MacKenzie, University of California, San Francisco ( www.fetaltherapies.org ).