• © 2009 by American Society of Clinical Oncology

Prenatal Diagnosis, Preimplantation Genetic Diagnosis, and Cancer: Was Hamlet Wrong?

Shakespeare's Hamlet proclaimed to his friend Horatio that “nature cannot choose his origin….” Who we are is predetermined and beyond our control. It has frequently been said that one cannot choose one's parents, yet with advances in genetic diagnostic technologies, through prenatal diagnosis (PND) and preimplantation genetic testing (PGD), it is possible to select one's children. Prenatal diagnosis and preimplantation genetic testing have been established for a wide spectrum of single-gene conditions including congenital disorders, inborn errors of metabolism, and chronic neurodegenerative diseases of childhood, such as Duchenne muscular dystrophy, or progressive conditions such as cystic fibrosis, or aneusomy syndromes such as Down's syndrome. New tests for single-gene disorders continue to be added to the armamentarium of the prenatal diagnostician, regardless of the choices offered by prospective parents' physicians. Whether to terminate a pregnancy or carry it to term, the fact remains that this ever-increasing array of tests challenges our medical and moral values. There are now almost 1,500 gene tests available in Clinical Laboratory Improvement Amendments (CLIA)–certified laboratories worldwide. Both PND and PGD offer prospective parents opportunities to select out embryos that would inevitably carry a universally lethal or medically devastating gene. For those families whose pedigrees have been riddled with cancer, these therapeutic options might be seen as a possibility to provide a prospect for a better future for their children. With these opportunities come great challenges. For which conditions should prenatal testing be offered? Should conditions of later age of onset be included? And what is considered later age of onset? Should testing only be available for genetic diseases with 100% penetrance, and when a lethal outcome is inevitable? These questions and many more have been raised in the context of adult-onset cancers including BRCA1/2 testing, where it remains generally accepted that, in light of the exclusively adult-onset nature of the disease, PND is not recommended. However, the work outlined in the study of Julian-Reynier¹ in this issue of Journal of Clinical Oncology brings such recommendations into debate.

Genetic testing for cancer susceptibility risk has until recently been primarily undertaken in the purview of geneticists with an interest in oncology, or in familial cancer genetics programs. Much of the activity in this field has been focused on hereditary breast and ovarian cancer and hereditary nonpolyposis colon cancer syndromes, both of which are almost exclusively diseases of adult onset. However, genetic testing for a wide spectrum of cancer predisposition disorders with a high frequency of high-penetrant childhood onset malignancies is attracting more attention from the genetics, pediatric oncology, and bioethics communities. With the advent of high-throughput rapid whole gene (and whole genome) sequencing, clinical molecular diagnostic laboratories will rapidly expand their ability to offer testing of virtually any disease susceptibility gene. With the potential to implement various clinical surveillance strategies for early detection of pediatric and adult cancers, the role for presymptomatic genetic testing becomes a topic that requires thorough evaluation. While presymptomatic testing and accompanying measures of clinical surveillance will not generally prevent cancers from developing outright, at least in the case of multiple endocrine neoplasia type 2A, prophylactic thyroidectomy at an early age (< 6 years) has been demonstrated to eliminate medullary thyroid carcinoma risk. Even in this circumstance, C-cell hyperplasia has been reported. Carrier testing for retinoblastoma susceptibility has led to early postnatal detection and even early delivery to minimize the devastating clinical effect of aggressive tumor formation.

The Julian-Reynier study¹ sheds new light on the complex algorithm that is required to effectively clarify the decision-making process for parents at extraordinarily high risk of transmitting a cancer-causing gene to their offspring. It is undoubtedly appropriate that this work be published in the mainstream oncology literature, for treatment of individuals/patients at risk of cancer should very much be the concern of both the community and the academic oncologist. Beyond the diagnosis, treatment, and follow-up of patients with cancer, the oncologist and patient develop a unique bond. Based on a strong family history or the documentation of a mutation in a cancer predisposition gene in a practitioner's patient with cancer, it is most likely that the patient's oncologist will be the first to be asked about risks to offspring, and what options may be available to minimize or eliminate these risks. While it may be the case that the oncologist is not expert in providing such genetic counseling, it is incumbent on them to appropriately guide the patient to the best resource to address these concerns. Ideally, such discussions should incorporate expertise of comprehensive multidisciplinary teams of oncologists, pediatric or medical geneticists, psychologists, genetic counselors, and bioethicists. In particular, while genetic testing and counseling remain in the framework of the genetics community, oncologists must share the burden of management decisions.

In fact, the scenario for genetic testing (pre- or postnatally) of offspring, or potential offspring, of adult patients with cancer present a somewhat unique dilemma to the practitioner. Generally, adult oncologists treat adults with cancer and pediatric oncologists treat children with cancer. However, in the setting of determining cancer risk and reducing cancer burden for the offspring, the pediatric and adult oncologists must find common ground. Ultimately, it is the recognition of family cancer history that leads to a decision to offer genetic testing, and the recognition that, in fact, numerous adult-onset cancers are often associated with a wide spectrum of pediatric cancers in close relatives. It is very much part of a geneticist's upbringing to take detailed family histories and to recognize nonrandom patterns of inheritance or associations of phenotypes; oncologists are focused on caring for patients with cancer. While a family history may be taken initially by an oncologist, it is rarely routinely updated unless unsolicited information is provided by the patient, and recognizing characteristics of familial cancer syndromes is not yet widely appreciated. This is where the strength of the multidisciplinary teams in France is notable.

But what of the option for PGD? How reasonable is this option or PND for a particular family? When is it reasonable and who should ultimately determine the value of such a test? At what point does a family devastated with early-onset cancers have the right to seek PGD or PND? And at what point do these tests become part of the battery of diagnostic tests that all patients with cancer experience during their battle with the disease? In France, PGD and PND for cancer susceptibility are covered by the national insurer with adjudication on a case-by-case basis and with the expert advice of multidisciplinary teams. Do such teams exist in other jurisdictions? What influence do they have on convincing medical insurers, be they national universal insurers or individual health care management organizations, of the medical necessity of these procedures? These are tough questions, and it is likely that this article will open dialogue among all the relevant stakeholders to try to reach some consensus. The four theoretical clinical scenarios presented to the 123 cancer geneticists (CG) and 47 multidisciplinary prenatal diagnosis teams (MPDT) represent realistic situations that practicing oncologists face. The clear differences in uptake of PGD/PND between the CGs and MPDTs highlights differences in training, experience with cancer (in contrast to more conventional genetic diseases), and availability of the particular service in the unit's jurisdiction. As pointed out by the authors, the opinions likely also reflect the fact that the scenarios presented are theoretical, and in a real situation, many other factors might play into the recommendations and decisions. The observations deal with the genetic practitioners' perspectives, but any dialogue of the value and validity of PND or PGD in the setting of cancer susceptibility must also include oncologists and parents—particularly those who carry the gene, whose children carry the gene, and who are facing the daily concerns and future uncertainty of childhood cancer. Furthermore, while in France it would appear that PGD for these conditions would be covered by the universal insurer on recommendation of MPDTs, this is not the case in all countries, where costs can be prohibitive for most families. Furthermore, while the risks of delivering a gene mutation carrier child may be eliminated, inherent risks of the procedures themselves—both psychological as well as medical—cannot be underestimated.

Cancer is devastating to the individual patient, to the health care team, to family and friends, as well as the community, society, and the global health care system. It does not seem unreasonable that, when provided an almost guaranteed certainty that offspring will develop cancer, a parent would wish to prevent it at all costs. After all, a parent sacrifices virtually anything to protect her child and to eliminate all reasonable risk to that child's health and safety. PGD is one new approach to offer parent's the hope of eliminating such risk.

Imagine prenatal genetic testing or preimplantation genetic diagnosis for cancer susceptibility as the hole down which Lewis Carroll's character Alice chased the White Rabbit. It leads to a complex world that offers many twists and turns and dark tunnels, but at the same time raises at least as many new questions. As oncologists, we must not simply defer the problem to our genetics colleagues to come up with solutions to guide patients/parents through these tunnels; rather we must work together with them, and engage patients as well as insurance providers, to assure improvement of patient outcomes. This is what the article by Julian-Reynier is telling us, and what makes it important reading for us.

The study does not in itself dictate policy or even offer a particular set of guidelines for practitioners to follow. But the study is important in acknowledging the need to develop practical guidelines and determining what is reasonable. The issue of cancer susceptibility testing is not going to disappear; with the almost weekly identification of new cancer susceptibility genes—many of which are directly relevant to childhood cancers—access to options for parents to decide the fate of their offspring must be made available. But unlike the melancholy Hamlet, who shunned those who would help him face what fates his stars held for him, parents at risk of transmitting cancer-causing genes—and, in so doing, sealing the fates of their children—should know they can engage a multidisciplinary team of geneticists, psychologists, ethicists, and (most importantly) their oncologist. Having done so, they can then hope to live in peace with their decision.