Components of the Risk Assessment Process
Additionally, family histories are dynamic. The occurrence of additional cancers may alter the likelihood of a hereditary predisposition to cancer, and consideration of differential diagnoses or empiric cancer risk estimates may change if additional cancers arise in the family. It is important to advise the consultand to take note of, confirm, and report cancer diagnoses or other pertinent family health history that occurs after completion of the initial risk assessment process. This is especially important if genetic testing was not performed or was uninformative.
Finally, the process of taking the family history has a psychosocial dimension. Discussing and documenting discrete aspects of family relationships and health brings the family into the session symbolically, even when a single person is being counseled. Problems that may be encountered in eliciting a family history and constructing a pedigree include difficulty contacting relatives with whom one has little or no relationship, differing views between family members about the value of genetic information, resistance to discussion of cancer and cancer-related illness, unanticipated discovery of previously unknown medical or family information, and coercion of one relative by another regarding testing decisions. In addition, unexpected emotional distress may be experienced by the consultand in the process of gathering family history information.
Determining Cancer Risk
Analysis of the family history
Because a family history of cancer is one of the important predictors of cancer risk, analysis of the pedigree constitutes an important aspect of risk assessment. This analysis might be thought of as a series of the following questions:
- What is the evidence that a cancer susceptibility syndrome is present in this family?
- If a syndrome is present, what is the most probable diagnosis?
- What could make this family history difficult to interpret?
- What is the most likely mode of inheritance, regardless of whether a specific syndrome diagnosis can be established?
- What is the chance of a member of this family developing cancer, if an inherited susceptibility exists?
- If no recognizable syndrome is present, is there a risk for cancer based on other epidemiological risk factors?
The following sections relate to the way that each of these questions might be addressed:
What is the evidence that a cancer susceptibility syndrome is present in this family?
The clues to a hereditary syndrome are based on pedigree analysis and physical findings. The index of suspicion is raised by the following:
- Multiple cancers in close relatives, particularly in multiple generations.
- Early age of cancer onset (younger than age 40 to 50 years for adult-onset cancers).
- Multiple cancers in a single individual.
- Bilateral cancer in paired organs (e.g., breast, kidney).
- Recognition of the known association between etiologically related cancers in the family.
- Presence of congenital anomalies or precursor lesions that are known to be associated with increased cancer risk (e.g., presence of atypical nevi and risk of malignant melanoma).
- Recognizable Mendelian inheritance pattern.
Clinical characteristics associated with distinctive risk ranges for different cancer genetic syndromes have been clarified by the Society of Gynecologic Oncologists Education Committee.
If a syndrome is present, what is the most probable diagnosis?
Hundreds of inherited conditions are associated with an increased risk of cancer. These have been summarized in texts [44,45,46] and a concise review. Diagnostic criteria for different hereditary syndromes incorporate different features from the list above, depending on the original purpose of defining the syndrome, e.g., for gene mapping, genotype-phenotype studies, epidemiological investigations, population screening, or clinical service. Thus, a syndrome such as Lynch syndrome (also called hereditary nonpolyposis colorectal cancer [HNPCC]) can be defined for research purposes by the Amsterdam criteria as having three related individuals with colorectal cancer, with one person being a first-degree relative of the other two; spanning two generations; and including one person who was younger than age 50 years at cancer diagnosis, better known as the 3-2-1 rule. These criteria have limitations in the clinical setting, however, in that they ignore endometrial and other extracolonic tumors known to be important features of Lynch syndrome. Revised published criteria that consider extracolonic cancers of Lynch syndrome have been subsequently developed and include the Amsterdam criteria II and the revised Bethesda guidelines.
What could make the family history difficult to interpret?
Other factors may complicate recognition of basic inheritance patterns or represent different types of disease etiology. These factors include the following:
- Small family size.
- Gender imbalance (e.g., few women in a family suspected of hereditary breast cancer).
- Deaths at particularly early ages.
- Removal of the at-risk organ, either for risk reduction or as a result of a medical condition (e.g., total abdominal hysterectomy due to history of uterine fibroids or endometriosis).
- Misidentified parentage.
- Late or variable onset.
- Variable expression.
- Genetic heterogeneity.
- Genomic imprinting.
- De novo mutation.
- Mosaicism (somatic or germline).
- Mitochondrial inheritance.
- Assisted reproductive technology (e.g., donor egg or sperm, in vitro fertilization).
What is the most likely mode of inheritance, regardless of whether a syndrome diagnosis can be established?
The mode of inheritance refers to the way that genetic traits are transmitted in the family. Mendel's laws of inheritance posit that genetic factors are transmitted from parents to offspring as discrete units known as genes that are inherited independently from each other and are passed on from an older generation to the following generation. The most common forms of Mendelian inheritance are autosomal dominant, autosomal recessive, and X-linked. Non-Mendelian forms of inheritance include chromosomal, complex, and mitochondrial. Researchers have learned from cancer and other inherited diseases that even Mendelian inheritance is modified by environmental and other genetic factors and that there are variations in the ways that the laws of inheritance work.[48,49,50]
Most commonly, Mendelian inheritance is established by a combination of clinical diagnosis with a compatible, but not in itself conclusive, pedigree pattern. Below is a list of inheritance patterns with clues to their recognition in the pedigree, followed by a list of situations that may complicate pedigree interpretation.
- Autosomal dominant inheritance refers to disorders that are expressed in the heterozygote, i.e., the affected person has one copy of a mutated allele and one allele that is functioning normally. Autosomal dominant inheritance is characterized by the following:
- Vertical occurrence (i.e., seen in successive generations).
- Usually seen only on one side of the family (i.e., unipaternal or unimaternal).
- Males and females may inherit and transmit the disorder to offspring.
- Male-to-male transmission may be seen.
- Offspring have a 50% chance of inheriting a mutation and a 50% chance of inheriting the normal allele.
- The condition may appear to skip a generation due to incomplete penetrance, early death due to other causes, delayed age of onset, or paucity of females or males when the at-risk organ is gender-specific (e.g., prostate, ovary).
- Most currently known cancer susceptibility syndromes follow an autosomal dominant inheritance pattern. Examples include hereditary breast and ovarian cancer syndrome, Lynch syndrome, familial adenomatous polyposis, and von Hippel Lindau disease.
- It is possible for an individual to have a mutation in a gene that has not previously been expressed as an autosomal dominant family history of cancer due to a variety of factors discussed above (see question #3).
- It is possible for an individual to have a de novo (new) mutation. This person would be the first affected member of his or her family, but could transmit this trait in the usual autosomal dominant manner to their offspring.
- Autosomal recessive inheritance refers to an inheritance pattern in which an affected person must be homozygous, i.e., carry two copies of a mutant gene, one from each parent. Autosomal recessive inheritance is characterized by the following:
- Horizontal occurrence, i.e., seen in one generation only (affected siblings in the absence of affected parents); generally not seen in successive generations.
- Mutated genes must come from both sides of the family, i.e., biparental inheritance.
- Parents are heterozygous carriers; each carries one mutated copy of the gene and one functional copy.
- Parents usually do not express the trait or the full syndrome; in some cases, parents may show a mild version of some features.
- Two heterozygous parents together have a 25% recurrence risk for future offspring being affected.
- Some well-defined cancer susceptibility syndromes with an autosomal recessive inheritance pattern include Bloom syndrome, Ataxia Telangiectasia, MYH-associated polyposis, and Fanconi anemia.
- X-linked inheritance refers to inheritance of genes located on the X chromosome. Because males carry one Y and one X chromosome, genes on their X chromosome are hemizygous and may be expressed, regardless of whether the trait is dominant or recessive in females. X-linked recessive inheritance is more common than X-linked dominant and is characterized by the following:
- Male and female offspring have a 50% chance of inheriting the mutated allele from a female carrier.
- Males in the maternal lineage (brothers and maternal uncles) are affected.
- Females are rarely affected, but when they are, the effects are usually milder than the effects in males.
- No father-to-son transmission of the mutation occurs, i.e., a father cannot transmit an X-linked condition to his son because he gives the son his Y chromosome and not his X.
- It is unusual for a cancer susceptibility syndrome to show X-linked transmission. One rare example is X-linked lymphoproliferative disorder.
- Chromosomal disorders generally are not inherited conditions except in rare cases of chromosomal translocations. Rather, they occur as a de novo error in meiosis at the time of conception of a given individual. Certain chromosomal anomalies confer a risk of malignancy; thus, inquiries about birth defects and intellectual disability are important for thorough pedigree construction and interpretation. Examples of chromosomal disorders with increased risk of malignancy include leukemia associated with Down syndrome (trisomy 21) and breast cancer associated with Klinefelter syndrome (47,XXY karyotype).
- Complex or multifactorial disease inheritance is used to describe conditions caused by genetic and environmental factors. Thus, a condition may be caused by the expression of multiple genes or by the interaction of genes and environmental factors. Therefore, a single genetic locus is not responsible for the condition. Rather, the net effect of genetic, lifestyle, and environmental factors determines a person's liability to be affected with a condition, such as cancer.
Susceptibility or resistance shows a more or less normal distribution in the population. Most people have an intermediate susceptibility, with those at the tails of the distribution curve having unusually low or unusually high susceptibility. Affected individuals are presumably those who are past a point of threshold for being affected due to their particular combination of risk factors. Outside of the few known Mendelian syndromes that predispose to a high incidence of specific cancer, most cancers are complex in etiology.
Clustering of cancer among relatives is common, but teasing out the underlying causes when there is no clear pattern is more difficult. With many common malignancies, such as lung cancer, an excess of cancers in relatives can be seen. These familial aggregations are seen as being due to combinations of exposures to known carcinogens, such as tobacco smoke, as well as to mutations in high penetrance genes or alterations in genes with low penetrance that affect the metabolism of the carcinogens in question.
The general practitioner is likely to encounter some families with a strong genetic predisposition to cancer and the recognition of cancer susceptibility may have dramatic consequences for a given individual's health and management. Although mutations in major cancer susceptibility genes lead to recognizable Mendelian inheritance patterns, they are uncommon. Nonetheless, cancer susceptibility genes are estimated to contribute to the occurrence of organ-specific cancers from less than 1% to up to 15%. Mutations in these genes confer high relative risk as well as high absolute risk. The attributable risk is low, however, because they are so rare.
In contrast, scientists now know of polymorphisms or alterations in deoxyribonucleic acid which are very common in the general population. Each polymorphism may confer low relative and absolute risks, but collectively they may account for high attributable risk because they are so common. Development of clinically significant disease in the presence of certain genetic polymorphisms may be highly dependent on environmental exposure to a potent carcinogen. People carrying polymorphisms associated with weak disease susceptibility may constitute a target group for whom avoidance of carcinogen exposure may be highly useful in preventing full-blown disease from occurring.
For more information about specific low-penetrance genes, please refer to the summaries on genetics of specific types of cancer.
Complex inheritance might be considered in a pedigree showing the following:
- Males and females affected (unless the target organ is gender-specific).
- A few cancers, without clear-cut vertical transmission or sibship clusters.
- No set pattern of inheritance.
- May appear to skip generations.
What is the chance of developing cancer if an inherited susceptibility exists?
These probabilities vary by syndrome, family, gene, and mutation, with different mutations in the same gene sometimes conferring different cancer risks, or the same mutation being associated with different clinical manifestations in different families. These phenomena relate to issues such as penetrance and expressivity discussed elsewhere.
If no recognizable syndrome is present, is there a risk for cancer based on other epidemiological risk factors?
A positive family history may sometimes provide risk information in the absence of a specific genetically determined cancer syndrome. For example, the risk associated with having a single affected relative with breast or colorectal cancer can be estimated from data derived from epidemiologic and family studies. Examples of empiric risk estimates of this kind are provided in the PDQ summaries on Genetics of Breast and Ovarian Cancer and Genetics of Colorectal Cancer.