From the vantage point of 2010, it is hard to imagine a time when cancer was not widely accepted as a genetic disease, in the most basic sense of being caused by alterations in the structure and function of genes. It is also sometimes difficult to remember that the hereditable nature of some common cancers only became broadly understood in the second half of the 20th century. Broca described a family with a strong predisposition to cancer as long ago as 1866, and Warthin noted another such family in 1913. But it was not until the middle of the last century that clinical investigators began to ascertain significant numbers of families with what we would recognize today as familial adenomatous polyposis, Lynch syndrome, Li-Fraumeni syndrome, and, of course, hereditary breast and ovarian cancer.

Clinically, these kindreds seemed to be transmitting single, highly penetrant alleles in an autosomal dominant fashion. The technology of the day could not identify the causative genes, but Knudson was able to provide empiric statistical support that at least one disorder, hereditary retinoblastoma, could be explained by a model in which disease resulted from an inherited mutation and a metachronous somatic mutation of the wild-type allele. In the 1980’s, complex segregation analyses of colorectal and breast cancer kindreds provided further support for the idea that highly penetrant alleles cause a small but significant fraction of these common cancers.

These genealogic and epidemiologic observations set the stage for a burst of scientific activity in the 1990s, when developments in positional cloning facilitated the discovery of the genes responsible for the most common cancer predisposition syndromes, and for several less common ones. These discoveries can rightly be considered triumphs of cancer genetics, providing important insights into the biology of not only inherited cancer but also sporadic disease.

However, at the time these discoveries were made, only a limited clinical evidence base was available to guide the management of individuals with an inherited predisposition. Guidelines were developed by various expert panels, but much of the clinical research to support and refine those guidelines had not been completed. The evidence base arose largely because of a firm commitment by those providing clinical genetic testing to do so mainly in the context of institutional review board–approved research studies, a commitment mirrored by the position statements of the American Society of Clinical Oncology and other groups. Reviews of the results of these clinical investigations form a significant part of this issue. These results illustrate why genetic testing has become an accepted (even critical) element in the management of subsets of patients with certain cancers, some of which are very common (eg, breast, colorectal), whereas others are not (eg, retinoblastoma, paraganglioma, medullary thyroid cancer).

Although genetic testing for cancer predisposition is broadly accepted for certain diseases, its clinical usefulness in others is less clear. For example, in this issue Gerstenblith and colleagues eloquently describe the limitations of testing for mutations in CDKN2/p16 as part of an evaluation of melanoma predisposition. One group defines the clinical utility of a genetic test as “…its usefulness and added value to patient decision-making compared with current management without genetic testing.” In other words, a test has clinical utility if the result is likely to influence diagnosis or management. This clinical utility can manifest in several ways.

First, a genetic test may have clinical utility if it helps diagnose a condition, and this has certainly been a major role of testing in nononcologic clinical genetics. However, diagnostic testing is less directly relevant in cancer clinical genetics because neoplastic diagnoses are tissue-based and usually do not require germline information for interpretation. However, cancers that arise as the result of a strong germline predisposition may be managed differently from those that arise from sporadic disease. Patients may undergo more extensive local therapy if they are at increased risk for metachronous malignancy. Obvious examples include the consideration of bilateral mastectomy instead of breast conservation in patients with breast cancer carrying a BRCA1 or BRCA2 mutation, and subtotal colectomy instead of limited resection in patients with colorectal cancer who have Lynch syndrome. Although genetic testing is not absolutely necessary for consideration of more extensive surgery, it certainly informs patient decision-making about whether to undergo these preventive procedures.

Germline tests could also have clinical utility if the results were either prognostic or predictive of response to treatment. Although some predispositions may be associated with differential outcomes (eg, microsatellite instability and improved prognosis in colorectal cancer), germline testing has not yet found a place in prognostication for either early or advanced disease. Similarly, germline changes associated with susceptibility have not, until recently, been predictive of differential response to treatment approaches. Recent success of inhibitors of poly(ADP-ribose) polymerase inhibitors in BRCA -mutated breast and ovarian cancers indicates a possible role for this testing in the future, as does the suggestion of sensitivity to cis-platinum in the neoadjuvant treatment of BRCA1 -mutated breast cancer. At the time of this writing, however, routine adjustment of systemic treatment is not based on germline status.

Although the role of germline testing in the treatment of established cancer is important to consider, the greatest clinical utility obviously flows from the ability of genetic testing to assess individual risk for developing cancer in the future. The usefulness of this genetically determined risk estimate depends on several factors. The estimate must be reasonably robust (clinically valid), in the sense that a mutation should be associated with a sufficiently consistent level of risk among individuals based on published absolute risk estimates which can be used to guide care for the patient being tested. If significant heterogeneity of risk exists among mutation carriers, then average estimates are difficult for patients to apply to their own situation, because their own risk may be either greater or less than the average. Even if the risk estimate is clinically valid, it must exceed an action threshold.

Although risk is a continuous variable, decisions about whether to pursue incremental screening or preventive surgeries are dichotomous. Modest elevations in risk may be perfectly reproducible and homogeneous, but not of sufficient magnitude to justify differential action. Similarly, effective interventions must be available to offer individuals at increased risk, which would not be offered without a documented genetic predisposition. The widespread acceptance of BRCA1 and BRCA2 testing is largely based on its ability to separate individuals into those who are or are not at increased risk for ovarian cancer, and the availability of effective and acceptable preventive surgery to address that risk. In other situations, such as Li-Fraumeni syndrome, the clinical utility of testing is limited because of the limited screening available for many of the component tumors of that syndrome.

Although highly penetrant genes are the ones most likely to be clinically relevant, they only explain a small portion of the hereditability of cancer. Several intermediate penetrance genes are associated with risks that do not clearly exceed an action threshold. And now, genome-wide association studies (GWAS) are identifying large numbers of genetic variants that are, individually, associated with very limited increases in risk. Notwithstanding the premature dissemination of “genomic profiles” of these variations by commercial entrepreneurs, the clinical utility of identifying these variants is completely unclear, and also poses real risks of false alarm and false reassurance.

This limitation is not to deny the role genomic profiles will ultimately have in preventive and therapeutic oncology. In fact, we anticipate that a future Hematology/Oncology Clinics volume on personalized cancer genomics will contain many articles on pathway analysis of the interplay of thousands of sequence and structural variants that will define the bulk of inherited cancer susceptibility and pharmacogenomic response. However, significant clinical and regulatory challenges remain as barriers to the effective integration of these new genomic discoveries into clinical care.

Accompanying the accelerated pace of genetic discovery and clinical translation has been an equally rapid evolution in thinking about the legal, social, and ethical aspects of the genetics revolution. Two decades ago, at the dawn of clinical cancer genetics, there was broad apprehension about a range of adverse social implications. These untoward consequences were averted by legal judgments and timely legislation, and progressive policies. Foremost among these concerns was the risk of genetic discrimination by health insurers and employers, precluded by federal legislation in 2008. Rather than discriminate against BRCA mutation carriers, for example, insurers now use these tests to justify payment for risk-reducing surgeries. When our group described the most common BRCA2 mutation in Ashkenazi Jews in 1996, concern was expressed about the risk of group stigmatization. Instead, as other groups were noted to have “founder” mutations in disease predisposition genes, these discoveries were used to empower these groups to take preventive actions. Equally unimaginable two decades ago, the uses of cancer genetic tests are now routinely discussed in the context of reproductive planning through techniques such as preimplantation genetics.

The further progress of clinical cancer genetics may also be influenced by recent and upcoming legal decisions that may result in loss or limitation of intellectual property protections afforded to patents on gene sequences. The outcome of those decisions, some argue, may dampen the pace of venture capital investment in needed biomedical research in the decades ahead, whereas others see the possibility that genomic information will become less expensive and more widely available. In any event, an intimate interplay exists among science, law, and medicine that will determine the ultimate future application of genomic information.

Following this introduction to the historical and societal context of two decades of discovery in cancer genetics, the articles in this issue provide the reader with expert and comprehensive reviews of the state of knowledge regarding genetic predisposition to cancer affecting several different sites. These reviews include detailed discussion of the genetics and management of individuals at risk for hereditary cancer of the breast and ovaries, upper and lower gastrointestinal tracts, genitourinary tract, genodermatoses, endocrine system, and pediatric tumors, breast and ovaries, upper and lower gastrointestinal tracts, genitourinary tract, genodermatoses, endocrine system, and pediatric tumors. A concluding article will review the emerging genomics of low-penetrance cancer susceptibility. Two articles will appear in the next issue of Hematology/Oncology Clinics of North America (24:6) that were originally planned to publish in this issue; one article will review mouse models of human cancer susceptibility and the other will focus in-depth on colorectal cancer predisposition syndromes. A common theme of these articles is that the clinical utility of testing for mutations in the genes described ranges widely, although this is an evolving field and new findings may well alter the perspective on the clinical utility of testing for different genes.

A detailed review of the current status of whole genome approaches to cancer predisposition concludes that it is not yet clear how newer findings from genomic profiling technologies should be used. Appropriate clinical research studies must be performed to answer that question. In this regard, the situation is not that different from what it was in the 1990s after discovery of BRCA1 , BRCA2 , and the mismatch repair genes responsible for Lynch syndrome. Currently, the major limitation to the clinical application of the new genomic results is that the magnitude of associated risk is insufficient to clearly justify differential action, although this may change with the advent of new screening technologies or biologically targeted pharmacologic prevention approaches. Furthermore, whole genome sequencing studies may identify less-common variants associated with sufficient risks to warrant intervention with available technologies. Much remains to be done to build on the 20th century successes of clinical cancer genetics so as to gain the maximum benefit from the genomic advances of the 21st century.