Report Card on Molecular Genetic Testing: Title and subTitle BreakRoom for Improvement?

Few areas of medicine are evolving as rapidly as molecular testing, and virtually no area holds so much promise for altering the practice of medicine. Of course, testing based on molecules has been conducted from at least the time when physicians tasted urine to detect glycosuria. Today multiparameter automated chemical analyzers have greatly extended the ability to test for hundreds of compounds in any sample of the human body. But the term molecular genetic testing now refers specifically to nucleic acid analysis, usually involving DNA. The techniques have become so sophisticated and numerous that increasingly the average physician can have no more understanding of molecular genetic testing than of the inner workings of the instrument that performs a chem-18. However, several points warrant emphasis: the scope and impact of molecular testing continue to expand, organized medicine must ensure the highest quality of the molecular testing process, and what comes before and after most molecular tests (the preanalytic and postanalytic aspects) are often just as important as the test itself and do involve a health professional, often a physician, to a high degree.

Testing based on exogenous nucleic acids, as for the RNA of human immunodeficiency virus or the DNA of human papillomavirus, can address presence, viral load, viral subtype, or, through sequence analysis, specific mutations for understanding chain of infection, pathogenicity, and drug resistance. Endogenous human DNA from tumor cells is routinely examined for somatic alterations that enhance both diagnosis and prognosis. Nonpathologic differences in DNA sequence among individuals in a population (polymorphisms) are useful for forensic analysis, paternity determination, HLA haplotyping, and identification of enzyme variants that may predict patient-specific chemotherapeutic responses, dosing requirements, and adverse effects.

The type of molecular genetic testing that is generating the most intense interest involves alterations of germline DNA associated with disease. A number of clinical applications exist, each with its own particular scientific, medical, and ethical concerns. Molecular tests can confirm the diagnosis in a symptomatic patient (eg, fragile X syndrome in a mentally retarded male) or prenatally in the fetus of a pregnant woman who has had a previous child or relative with a disorder.¹ Tests of DNA can be used to identify heterozygote carriers of mutations for autosomal recessive diseases (eg, cystic fibrosis [CF], thalassemias) that would alert an individual to the risk of producing affected offspring if he or she were to mate with another carrier, thus enabling informed reproductive choices. Such carrier testing can be offered to relatives of affected individuals, people of an ethnic group at particularly high risk for 1 or more conditions because of high frequency of mutant allele(s) in that population,² or everyone of childbearing age in the general population. Indeed, a recent National Institutes of Health Consensus Conference³ concluded that all people interested in having children be offered testing for CF heterozygosity, a recommendation that still requires development before full implementation.⁴ Perhaps most difficult from medical and psychosocial perspectives, are predictive DNA tests for presymptomatic identification of diseases of adult onset (eg, Huntington disease) or predisposition to such diseases (eg, familial breast or ovarian cancer associated with mutations in BRCA1) in offspring who are at 50% risk because 1 of their parents had the disease.

Professional societies have moved to establish quality assurance criteria for the actual process of molecular testing, including laboratory accreditation, director certification, and proficiency testing. The article by McGovern and colleagues⁵ in this issue of THE JOURNAL serves as a report card on the current status of laboratories offering molecular testing in the United States and suggests that some have not made the grade. Factors associated with practices judged optimal, based on standards established by the Laboratory Practice Committee of the American College of Medical Genetics (ACMG),⁶ included a larger menu of tests offered, number of tests performed, a dedicated clinical laboratory certified by Clinical Laboratory Improvement Amendments of 1988, and board certification of the laboratory director. However, this study only monitored adherence to professional laboratory practice guidelines established by the ACMG, but laboratory performance in actual testing situations was not measured. The latter capability is better monitored through on-site inspections and proficiency testing programs such as those presently administered jointly nationwide by the College of American Pathologists (CAP) and the ACMG. Such surveys have highlighted recently a need for standardization of such tests as fragile X mutation analysis and CF carrier screening.⁷

While it is important to remain vigilant regarding the concerns raised by McGovern and colleagues,⁵ there is no direct evidence, either from their survey or anecdotally, that serious diagnostic errors are being made on a large scale in either molecular genetics or cytogenetics laboratories. For example, even if a laboratory does not follow every recommended procedure for polymerase chain reaction (PCR) containment, it does not necessarily follow that PCR contamination problems will occur, or if they do, that such problems would go undetected and lead to false results.

The survey aimed for a broad, comprehensive snapshot of the entire world of genetic and molecular testing, which could make it difficult to discern specific trends. Personnel qualifications and procedures in cytogenetics laboratories are markedly different from those in molecular genetics laboratories. Accordingly, some of the perceived poor technical scores of laboratories performing fluorescent in situ hybridization, which skew the overall results, should not be used to infer pervasive problems in the molecular genetics arena. Similarly, a few of the laboratories in the survey were ascertained through performance of molecular oncology testing, which mainly involves molecular clonality studies in lymphomas and leukemias, and should not be graded on having an association with a genetic counselor, since these diseases are by and large not heritable. Another source of low scores, the requirement for informed consent before testing, remains a controversial issue in medical genetics, with many feeling that it should only be necessary for certain presymptomatic or susceptibility tests but not for the vast majority of diagnostic tests.^{8 - 9} It is disturbing that McGovern et al⁵ found low compliance rates (44% and 34%) even for the former.

In addition, the low compliance to standards observed in research laboratories is something well known yet treated rather sympathetically in the report of the National Institutes of Health–Department of Energy Task Force on Genetic Testing,⁸ which recognized that these laboratories perform a unique and valuable service for diagnosis and management of orphan diseases. Research laboratories usually do not have the resources to meet all the regulatory requirements designed for a large clinical laboratory. The tests performed in research laboratories are of low volume, thus not affecting overall quality performance in the field; in most cases it is likely that the laboratory directors, even though they may be investigators rather than clinical laboratorians, are expert on the particular disorder being addressed and certainly are best qualified to oversee the testing.

What can be done to improve the likelihood that a specimen obtained anywhere in the United States will be analyzed competently? McGovern and colleagues⁵ suggest that regulation offers 1 solution; others argue that competency is best fostered by professional societies rather than by the government, with support systems in place to enable compliance. The joint proficiency testing program and laboratory inspection activities of CAP and ACMG are useful components of such a system.¹⁰ Both the American Board of Pathology and the American Board of Medical Genetics provide certification for clinical laboratory directors, and both boards are working to establish a joint certification in molecular testing; we encourage the early conclusion of these negotiations. A relatively new organization of DNA diagnostic laboratory directors and technologists, the Association for Molecular Pathology, is working to promulgate technical improvements through investigative studies and exchange of validated protocols.

Molecular genetic testing is so new and advancing so rapidly that neither the laboratory reagent manufacturers nor the US Food and Drug Administration have been able to keep abreast of the latest developments. As a consequence, molecular genetics laboratories, more so than any other sections of the clinical laboratory, rely on assays developed in-house rather than on commercially sold, Food and Drug Administration–approved kits. This approach further promotes fragmentation of the field and impedes efforts toward national uniformity in procedures. While there is no reason to believe that a properly validated in-house assay should be any less accurate than a manufactured one, the Food and Drug Administration has recently taken steps to address a situation it perceived as spiraling out of control by stipulating certain accreditation and disclaimer requirements for laboratories that construct their own tests using analyte-specific reagents, such as DNA probes.¹¹ One impediment to the introduction of new tests, particularly for laboratories in academic centers, is the high costs of licensure and patent royalties for both the procedures used (such as PCR) and the DNA sequences. This is 1 issue that may well require governmental intervention, and all the aforementioned professional organizations are supporting this approach.

Molecular testing, particularly for hereditary disorders and susceptibilities, requires much more pretest and posttest involvement of health care professionals than do most other tests performed using blood or tissue specimens. In addition to issues of test sensitivity and specificity (which, despite the high precision or reproducibility of the test, are often far from ideal), the patient needs to be informed about a seeming plethora of additional concerns. For example, what is learned about a parent often applies, with some definable probability, to offspring, grandparents and other relatives. This raises the question of how confidentiality of the relatives is guarded. Knowledge of susceptibility based on genotype can potentially lead to discrimination in insurance coverage, both fairly and unfairly,¹² although actual instances of abuse seem rare.¹³ Furthermore, the ability to detect susceptibility currently outstrips the ability to modulate the risk; the association of certain genotypes at the apolipoprotein E locus with the risk of Alzheimer disease is an example.^{14 - 16} For these reasons, informed consent before initiating certain types of moleculargenetic testing has much to recommend it, as stated among others by the National Society of Genetic Counselors,¹⁷ the Institute of Medicine,¹⁸ and the National Institutes of Health–Department of Energy Task Force on Genetic Testing⁸ and as may be required in states with genetic privacy laws.

How a health professional balances these responsibilities of confidentiality with the desire, and some would argue obligation, to warn relatives of a potentially lethal hereditary risk is often not clear. Many busy physicians who, if they studied genetics at all in their training, often did so before the molecular age, when faced with all of these nuances of molecular testing might despair, or worse, ignore the pretest and posttest aspects entirely. Laboratories can help with this process by accepting certain samples only when pretest counseling and, where appropriate, informed consent are documented. Similarly, laboratories can provide reports that detail in understandable language many of the issues that need to be addressed when the results are interpreted for the patient. Considering that not enough individuals skilled in the pretest and posttest aspects of molecular testing, such as medical geneticists and genetic counselors, are available to handle the expected caseload of large-scale programs such as CF carrier screening,^{3 ,8 ,19} let alone yet-to-be-discovered mutations and polymorphisms associated with common diseases such as diabetes and hypertension, it is imperative that physicians of all specialties and ages become conversant with these gene-based medicine concepts.

While still in its infancy, molecular testing holds every promise of emerging as the predominant paradigm by which physicians will diagnose and manage patients well into the 21st century. But if it is to fulfill this promise rather than stumble before it ever leaves the gate, molecular genetic testing will need to earn the confidence of both the profession and the public. The findings of McGovern et al⁵ suggest that this is best achieved by promoting uniform, consistent, and widely accepted standards of laboratory practice across the country. Cooperative efforts between professional organizations, such as those between the ACMG and CAP, and governmental agencies, already have come a long way toward these goals, and we look to the future with optimism and excitement for the benefits these services will afford patients.