Writing Genetic Testing Guidelines Complex Endeavor

June 1, 1997

FORT LAUDERDALE, Fla--Writing guidelines for cancer screening that include genetic testing--the next major guidelines initiative for the National Comprehensive Cancer Network (NCCN)--will be complex because "every cancer, every gene linked to a cancer, and every patient situation is going to present different kinds of problems," Henry Greely, JD, said at the NCCN's second annual conference.

FORT LAUDERDALE, Fla--Writing guidelines for cancer screening that includegenetic testing--the next major guidelines initiative for the NationalComprehensive Cancer Network (NCCN)--will be complex because "everycancer, every gene linked to a cancer, and every patient situation is goingto present different kinds of problems," Henry Greely, JD, said atthe NCCN's second annual conference.

Mr. Greely is a professor of law at Stanford University and is uniquelyqualified to speak to this issue in his position as chair of the steeringcommittee for Stanford's Center for Biomedical Ethics and codirector ofStanford's Program on Genomics, Ethics and Society.

Mr. Greely described seven areas in which genetic diseases and genetictesting can differ, and which must be taken into account when writing guidelines:

1. The nature of the genetic inheritance. Genetic diseases canbe dominant (requiring only one bad copy of the gene to put a person athigher risk); recessive (requiring two bad copies); or several other types."The nature of the genetic inheritance will affect the meaning ofthe patient's result for their children, parents, and siblings," Mr.Greely said.

2. The relationship between the genetic defect and the disease.The key issue here is penetrance, ie, what percentage of people with agiven genetic variation are actually going to get the disease, he said,noting that penetrance of different diseases can vary widely.

For example, the lifetime likelihood of developing breast cancer ina woman with a BRCA1 or BRCA2 mutation is somewhere between 60% and 85%."That represents an absolute rate, which must be compared to the lifetimerisk in the population as a whole, the relative rate, which is about 9%to 12%," he said. In some diseases, the relative risk is fairly smallbut the absolute risk is great and vice versa.

In addition, some genetic diseases require defects in two or more differentgenes for the disease to develop, while in others a defect in any one ofa number of genes can lead to an increase in risk.

3. The type of genetic testing available. There are at leastthree different ways of doing genetic tests. Linkage analysis requiressamples from some affected family members, "and it's never entirelyaccurate because there is always the possibility of recombination withinthe relevant region of the chromosome."

For diseases in which the culprit gene has been identified and located,genetic tests may be used to look for a particular mutation that accountsfor all or almost all of the disease, as with sickle cell anemia, or tolook for many possible mutations, as with BRCA1, which may require totalgene sequencing.

4. The severity of the disease and the interventions available.Some diseases are inevitably fatal, some are often fatal, and some liein between. "A test for Huntington's disease means something differentthan a test for susceptibility to squamous cell skin cancer, because theconse-quences are so different. A test for breast cancer is somewhere inbetween," Mr. Greely said.

The consequences of testing also differ depending on the kinds of interventionsavailable. Prophylactic and screening measures are available, for example,for patients at high risk of colon cancer, "but it's much harder toscreen for, say, ovarian or pancreatic cancer," he noted.

5. The time in a person's life the testing is performed. Genetictesting can be done before conception by couples who want to know if theyare carriers of specific genes, postconception but preimplantation by patientsusing in vitro fertilization, or prenatally via amniocentesis. Testingcan be done in children or in adults presymptomatically or as diagnostictesting. "Although diagnostic genetic testing is not perhaps significantin cancer, it is turning out to be important in Alzheimer's disease,"Mr. Greely said.

6. The costs of testing-financial and psychological. The financialcosts will range from $100 to $200 for a simple single mutation probe,to several thousand dollars or more for more complicated linkage analysisor total gene sequencing. But patients also face social and psychologicalcosts in dealing with their family, their potential insurability, and possibleemployment discrimination, and these will differ depending on the diseaseand the test result.

7. Patient differences. Patients from different cultures havedifferent feelings about genetic testing and the cancers that underliethem. "A patient from a culture where cancer is never mentioned islikely to feel differently about genetic testing for breast cancer thana patient from a culture that deals with cancer in a more straightforwardway," he said.

The patient's relationship with his or her family is likely to be affectedby testing, whether the result is positive or negative. In either case,it may bring a family closer together or tear it apart. "The patient'srelationship with his family will be an important consideration in testing,"Mr. Greely said.

Finally, the patient's personality will make a difference. "Somepeople want to know and others don't," he said. "These are personalitydifferences that you can't necessarily predict until you get down to thelevel of talking to individual patients."

Looking at the Process

Mr. Greely estimates that these seven areas of differences in genetictesting with their many permutations produce hundreds of thousands of differentcombinations of outcomes. "Since it is inconceivable that one couldwrite 200,000 different practice guidelines or even 1,000 different practiceguidelines," he said, the answer is to look at process rather thansubstance.

He outlined six aspects of the process of genetic testing for whichguidelines should be written.

1. Counseling patients about medical issues. Patients need tounderstand something about the cancer for which their genetic risk is beingtested and the genetics of the disease. "This is particularly important,"he said, "since for most genetic tests for cancer risk, the medicalanswer won't be straightforward, but will fall into a gray zone where thepatient's preferences and situation will play a larger-than-usual rolein the medical decision making."

Strong Psychological Reactions

2. Counseling patients about nonmedical issues. Patients needto know before they get genetic testing that the results may, given ourcurrent health care system and laws, lead to discrimination in health insuranceor employment.

They need to know that some people have very strong psychological reactionsto test results, including severe depression and in some cases suicide;that family members often have mixed reactions to results; and that suchreactions are not always predictable.

3. Updating of scientific and medical content. Because out knowledgeof genetics is changing so rapidly, there needs to be an ongoing processto evaluate new data. Mr. Greely recommends ongoing panels to review theindications for genetic testing and intervention in specific diseases.

4. Competency standards for counselors. Currently, there areonly about 1,500 genetic counselors in the United States, and most of thesefocus on prenatal diseases, not cancer. On the other hand, an oncologistwho specializes in breast cancer isn't necessarily equipped to counselpatients about the genetics of the disease or the family dynamics of genetictesting.

"We need programs that train people to a level where they can givecompetent advice about the disease, the genetics, and the possible familial,psychological, and social reactions," he said.

5. Patient follow-up. Follow-up is crucial after genetic testing.In some cases, it involves medical interventions such as an increased screeningschedule or a prophylactic regimen.

In all cases, there should be at least some follow-up with respect tothe psychological and familial consequences of genetic testing.

6. Linking counseling to ongoing research. Guidelines shouldrecommend that testing programs be linked to research, so that patients,both those who test positive and those who test negative, are followedto see how the information affects their lives and their outcomes in termsof developing cancer.

"Only with this kind of database," Mr. Greely said, "drawnfrom a broader spectrum than the very high risk families that are the subjectsof gene discovery studies, will we learn what genetic testing means toindividual patients."