Advances in Understanding Genetic Changes in Cancer: Impact on Diagnosis and Treatment Decisions in the 1990s (1992)

The diseases for which analyses of genetic changes are most important are summarized in Table 2-1. Specific genetic changes, tests that can be used now or that will be available in the near future for their detection, and the way in which a positive test can be clinically useful are listed. While the table is up-to-date, it should be understood that this is a field of rapid advancement. Because all cancers contain acquired mutations, the number of tumors for which genetic analysis is part of the standard workup will undoubtedly expand in the next several years. Moreover, the technology employed to detect these changes is also evolving quickly.

Given that assessment of genetic changes is going to have an increasing impact on clinical medicine in the future, physicians will need to have a working knowledge of the molecular pathogenesis of various neoplasms. Continuing medical education of physicians and incorporation of recent developments in understanding the genetic mechanisms of neoplasia into medical school curricula are essential if physicians are to acquire the necessary information. In addition to the education of clinicians, several other potential stumbling blocks must be overcome before genetic analysis can be considered a routine part of caring for the patient with cancer.

For the aforementioned tests to be useful, internists, oncologists, surgeons, and pathologists will need to know how to collect, handle, and store specimens. Close communication among clinicians, pathologists, and laboratory personnel must occur if a specimen is to be handled in optimal fashion. In situations where a biopsy of surgical procedure is required for procurement of diagnostic material, it is important that tissue can be divided for various tests while fresh. If special tests are to be cost effective, they should be employed only in specific settings. It should be emphasized that morphology will continue to be the primary method for diagnosis and will dictate which additional special tests are required. Since permanent sections are often unavailable at the time decisions are being made, examination of frozen sections may be desirable to guide the ordering of special tests.

In cases where lymphoma is suspected, the clinician should provide the pathologist with sufficient fresh, unfixed tissue to allow for routine histology, immunohistochemistry, and gene rearrangement studies. Many of the most useful immunohistochemical stains, including those for immunoglobulin light chain, work well only on snap-frozen tissue, and it is therefore critical that residents and staff pathologists know how to freeze tissue in a way that minimizes ice crystal artifact and preserves morphology. If immunohistochemistry does not yield definitive diagnosis, a portion of the remaining frozen tissue can be used for DNA preparation and gene rearrangement studies. Occasionally, atypical lymphoid infiltrates will be discovered unexpectedly in a specimen after routine processing in formalin; this is best handled with a repeat biopsy. Ideally, all pathology departments should have a-70°C storage facility for lymphoid specimens

Soft tissue tumors with spindle cell or round blue cell morphology constitute an indication for cytogenetic analysis, both for precise diagnosis and, in the special case of neuroblastoma, to help predict natural history. This will require transport of tissue to a central cytogenetics laboratory and may lead to some difficulties in the isolated community hospital, since even with prompt refrigeration in physiologic solutions (e.g., Hank's buffered saline), the yield of informative metaphase cells decreases after three to six hours. The future ability to detect most important cytogenetic abnormalities in interphase nuclei by fluorescence in situ hybridization (FISH) may partially eliminate this problem, since cells need only be preserved, not viable.

Special studies should also be considered in some other types of solid tumors, most notably ploidy and ERBB2 analysis in breast carcinoma and MYCN analysis in neuroblastoma. Genetic analysis of lung and colorectal carcinoma and other types of tumors is currently mainly of research inter-

est. Undoubtedly, however, important molecular markers will be discovered in these more common tumors and ultimately utilized clinically. Banking of tissue in academic centers is strongly recommended, because it will provide an invaluable resource for initial retrospective studies designed to test the usefulness of various molecular lesions as they are discovered. While snap-freezing and storage at-70°C or lower are preferable for many purposes, fixation and storage in 95% ethanol at room temperature provides a cheap, simple alternative, since DNA is easily isolated from tissue preserved in this way.

Cytogenetic analysis is typically performed in an established laboratory under various guidelines, such as those of New York State, and commonly run by a director certified by the American Board of Medical Genetics. In contrast, specific molecular diagnostic tests are being performed largely in individual research laboratories at large academic centers. In some cases, dedicated molecular diagnostic laboratories (MDLs) have been organized as distinct entities; however, there is as yet no group responsible for overseeing the performance of such laboratories. Training of technical support staff is often on-the-job. Individuals with diverse backgrounds are often responsible for supervision of these laboratories, and uniform standards for quality control have not been set for these facilities. A number of professional organizations including the College of American Pathology, as well as individual members of the American Society of Human Genetics, are considering guidelines for various aspects of MDLs.

Training of MDL directors and technical staff is also an unsettled issue. Two of the logical places for M.D.s and Ph.D.s to acquire MDL technical and management skills are in clinical pathology programs or in hospital-based postdoctoral fellowships in molecular diagnosis that are certified by the American Board of Medical Genetics. Basic laboratory practices, such as quality control and laboratory management, are essential for MDLs and would be an important part of this training. Unfortunately, few training programs currently offer such instruction. Thus, the training of professional and technical support staff and the establishment of national standards for laboratories remain two outstanding issues to be addressed.

Another, more general issue that needs to be addressed concerns the commercialization of genetic tests in oncology and the availability of tests for rarer forms of cancer. Genetic testing for cancer has already become the subject of commercial activities, both in the form of reference laboratories and in the marketing of testing kits. An incentive apparently exists for

private industry to develop and provide diagnostic tests for more common forms of cancer, such as lung, breast, colonic, and various hematopoietic malignancies. However, there is a chance that tests for less common cancers will go undeveloped if research in these areas is not subsidized. Congress has recently recognized the need to support research and development of drugs aimed at treating uncommon diseases. By analogy, similar support might be required to enable and encourage the development of diagnostic tests for less common types of cancer.

For molecular oncology to come to full fruition, further advances in understanding of the molecular mechanisms of carcinogenesis are needed, particularly those involved in the more common tumors of man. Improved cytogenetic techniques for solid tumors and almost weekly announcements of new mutations in genes such as TP53, RB1, RAS, and MCC in a variety of solid tumors are evidence that many mutations of potential clinical relevance are going to be discovered soon.

To prove that these mutations are clinically useful, however, will not be easy. National or cooperative tumor banks will be essential to this end, because they will allow retrospective correlation of the presence of newly identified molecular lesions with disease outcome. The procurement service established by the National Institutes of Health, which is designed to provide specimens to appropriate researchers, should aid this purpose, and this activity should be expanded, optimally with some standardized procedures for tumor characterization and clinical follow-up. Single reference laboratories employing well-controlled and reproducible methods in multi-institutional studies should be utilized to increase the probability of finding significant associations.

Use of genetic analysis in clinical management of patients places a premium on testing accuracy and speed. Tests based on polymerase chain reaction (PCR) need to be standardized, and means of ensuring an acceptably low false-positive and false-negative test rate need to be developed. Southern blotting is often too slow to provide clinical guidance; increased automation is needed. Better means of preserving tissue for FISH (and possibly cytogenetics) during shipping or initial morphologic analysis would be helpful. It is also evident that many of the important lesions are likely to be point mutations occurring at scattered locations within proto-oncogenes and tumor suppressor genes. A variety of techniques have been developed to analyze these types of changes, but many are technically difficult and too time consuming for routine clinical use. Development of new technology to speed and simplify detection of point mutations and other small genetic changes within fairly large regions of DNA will greatly increase the number of tumors from which clinically relevant molecular data can be gathered.