Genetics Workshop, Section of General Medicine

April 11, 2001 - by Jerry Sobieraj, MD

Information in part derived from Genetics in Primary Care (GPC) Materials, which is an NIH funded demonstration project

Introduction to Genetics in Primaary Care, January 17, 2001

• Is the hypothesis valid, i.e. is an understanding of genetics important in primary care? If not now, then when?
• Case identification. Using family history and symptom based clues.
• Non-Directive Counseling. Should primary care doctors modify how they present information and recommend a plan of action with their patients when discussing genetic matters?

Non-Directive Counseling Revisited

This term appears to be heading toward replacement by "shared decision making". Recent work by Gail Geller at Johns Hopkins, suggest patients feel empowered if the are in control of the "final" decision. Some of these issues are summarized in a 1997 report by The Task Force on Informed Consent is part of the Cancer Genetics Studies Consortium (CGSC).

Some comments by GPC members regarding "non-directive counseling". I believe they summarize some of the issues related to this issue. Both authors suggest its importance in the prenatal arena. In terms of adults, the main issue appears to be our ability to provide adequate information to our patients about the genetic test at issue, so that an informed, hopefully shared decision may be made.

• Susie Ball, MS, Genetic Counselor, Univeristy of Washington team
  While there are certainly situations where the physician's knowledge, wisdom and experience are extremely valuable to the patient, geneticists are concerned about those situations where the results will not impact the patient's healthcare and thus the physicians' medical knowledge and training may become less important, while the patient's values become the deciding factor. In this situation, the physician should not be telling the patient what to do, but rather assisting the patient to explore their own values and beliefs and make a decision concordant with those.
  This process may also be called informed decision making, which appears to be the new paradigm in medicine. It includes meeting the ethical obligation of promoting patient understanding and shared decision making. It is a dynamic process to foster the patient's involvement in a clinical decision (rather than a one way directive from the physician) and includes assessment of patient needs and preferences which arise out of the patient's unique situation, educational background, life experiences and values.
  A few examples: Non-directive counseling is important in pre-natal cases. The physician's bias as to "rightness" or "wrongness" of abortion should not play into the discussion. Whether or not to have an abortion needs to be a personal decision for the woman or couple. If they feel pressured by their physician into making a decision that would not have been their own, their resentment toward their physician may last for a long time. The same is true for a pregnant woman deciding whether or not to have some type of prenatal testing. A woman who feels pressured into having, for example, a triple marker screening to estimate her chance the fetus has Down syndrome, if that information is not really that important to her, may resent having to then worry about the rest of her pregnancy or face the decisions of whether or not to have an amniocentesis and/or abortion.
  This approach is also extremely important in pre-symptomatic (testing for a gene mutation before you have symptoms of the condition, e.g. Huntington's disease) and pre-disposition (testing for a gene mutation to see if you are at an increased risk to develop a condition, e.g. breast cancer) testing. In these situations, the physician's experience and knowledge in medicine is not the important factor, but rather how might this information be of value to the patient and what does the patient intend to do with it. Thus the decision maker must be the patient.


• Robert Hopkin, MD, Division of Human Genetics, U. of Cincinnati
  I think at least part of the difference between primary care givers and providers of genetic services is how we use language. In primary care the concept of patient centered care is well established and may be very useful in initiating discussion of the issues related to genetic testing including the issues such as insurability and implications for other family members.
  In addition I don't think genetic counseling or genetic testing are literally "non-directive" our very existence as service providers implies at least to some clients that action should be taken if testing is pursued. Furthermore some genetic testing is virtually forced on the population, eg. new born screening.
  The concept of nondirective counselling is very useful in situations where there is controversy about the best management of the findings. The best example being prenatal testing where with the same scenerio one couple may opt for testing and termination, another may opt not to test, and a third may opt to test and continue the pregnancy knowing that they will have a child with multiple medical and developmental problems. No one of these is universally the right or wrong choice. An example of a bad out come with directive counseling in this type of case is a family that has a triple screen " because my doctor told me I should." If an abnormality is detected they may feel obligated to follow through with amnio or even termination. I have seen families who felt very guilty and resentful in this situation and know of some who have avooided early prenatal care to avoid being asked to consider penatal testing after this type of experience. I have also worked with families who did not get testing because it wasn't offered, but would have wanted it. Some have experienced years of anger following the birth of a child with severe disabilities.
  Another example of possible poor outcomes due to insufficient counselling is the family in which a woman with breast cancer seeks information on breast cancer risk. She is found to have inceased risk and testing is done with little discussion of what the result will mean other than that it is a test for "the breast cancer gene." She wanted the information "for her daughters and neices." However, when the results are back and positive she is very upset and resentful about the risk for ovarian cancer and the need to have screening or consider oopherectomy. The daughters are angry at their mother because they didn't want to know their risk and now they refuse to discuss the matter let alone seek screening, and the neices simply write the whole thing off as a scam without really considering it. Again we have dealt with situations like this that could have been avoided if the involved familiy mambers had been given the information and participated in the decision making process.
  An example of directive genetics is screening for PKU. Not only do we test whether or not the family wishes testing, but we will often seek a court order or foster placement if the parents don't make at least a token effort to follow recommendations.
  As a genetics community we are aware that in some situations the information we give has definite implications, and that in other situations there are complex and difficult choices that should be explored but no path preferentially recommended because we deal with these questions daily. We sometimes forget that all information comes with implied value, and potentially with perceptions of pressure to act. I think the goal in working with our collegues in primary care should be to assist them in understanding the important issues involved in a decision to pursue or not pursue testing. They will then be able to refer patients for counselling or offer counseling in a way that addresses the issues openly and completely. The concepts of patient centered care already imply a duty to assist the patient in evaluating a variety of options even if the care provider thinks one option is more desirable than the others. Thus it incorporates the essential elements of nondirective counselling. I believe and hope that the bigger issue is that many primary care providers are not aware of some issues and implications of testing, and that they lack the time to review all the issues in detail with their patients.

When making a genetic diagnosis, factors that need to be considered include:

• Family issues, that is the dynamics of testing others in the family
• Family planning, how does the diagnosis impact upon child bearing
• Monitoring the patient at increased risk (i.e. a person who tests positive on the genetic test). This will include clinical testing and surveillance, helping the patient to identify support groups, and the need to remember that your patient will still have a myriad of other non-genetic issues to deal with.

Breast Cancer Genetics

The major issue is to define wether there is likely to be a hereditary basis. Non-hereditary factors include:

• Age: median age at dx is 55 yo. 20% occurs in women less than 40yo. Only 6-8% of woman with breast cancer onset at less than 35yo have identifiable BRCA1 mutation.
• Reproductive Hx: inverse relationship between breast cancer risk and parity. If first birth at less than 20yo, about half the risk versus a nulliparous woman.
• Menstrual Hx: early menarche and late menopause increase risk (relative risk of about 2).
• Histologic features of prior breast bx: atypical hyperplasia.
• Other factors: moderate alcohol intake and weight gain after menopause increase risk.

Yet the major factor in defining risk of hereditary breast cancer is family history. About 5-10% of breast cancers have strong hereditary component. The 6 features of family hx which affect risk of breast cancer are:

• Number of affected relatives
• Closeness of biologic relationship to affected relatives
• Presence of bilateral breast cancer in an affected relative
• Breast cancer in a male relative (especially true for BRCA2)
• Age at diagnosis of breast cancer
• Presence of ovarian cancer

These may be summarized the Amersterdam Criteria, often referred to as the 3, 2, 1 rule. Three affected relatives, 2 generations, 1 affected relative at a young age (e.g. <50 years old).

Models which have been developed to assess breast cancer risk (e.g. the Gail and Claus Models) were not designed to assess BRCA1 and BRCA2 risk. Both ignore family history of ovarian cancer and paternal history (both parents contribute BRCA allele). In addition, the Gail Model ignores age of breast cancer onset in relatives, while the Claus model can accomodate only 2 family members with breast cancer.

BRCA 1 and 2:
These are large genes (more than 6000 bp) with over 140 mutations defined. Two populations which have had their mutations well characterized are Ashkenazi Jews and Icelanders. The 185delAG and 5382insC founder mutations are the consequence of only two historical mutations. These two and a third common mutation are present in 2.5% of Ashkenazi Jews. Based on these mutations, this population by age 70 has a 56% risk of breast cancer, 16% risk of ovarian cancer and 16% risk of prostate cancer. The Icelandic mutations have lower penetrance (a BRCA2 mutation is found in 0.6% of Icelanders), with 37% breast cancer risk by age 70.

The following excerpt if from recent correspondence in the NEJM 344: 936, 2001: In women with these genetic defects, there is a reduced ability to repair DNA double-strand breaks and other types of genetic damage. Both copies of BRCA1 must be inactivated in the same cell to induce the development of cancer. The BRCA1 gene appears to have a low rate of mutation. New BRCA1 mutations are exceedingly rare. In contrast, p53 mutations are common in a range of tumor types, and new p53 germ-line mutations often occur. In the Ashkenazi Jewish population, the high proportion of familial cases attributed to recurrent BRCA1 mutations reflects the high prevalence of these founder alleles (1 percent for the 185delAG mutation alone), rather than an unusually low frequency of new germ-line mutations.

It is important to remember, that when assessing the risk of your patient, the first person to test is the index case. You need to define the genetic abnormality in the "proband" before looking at relatives of the proband.

Hemochromatosis (HHC): the case for genetic screening

HHC is the most common genetic disorder in European Americans. Iron overload may be difficult to diagnose clincally (i.e. it has a varied phenotype). There are two common mutations which are both autosomal recessives. The C282Y mutation is thought to have occurred about 1000 years ago in the Celtic population. The H63D mutation is much more common (3-5 times) than the C282Y, but due to higher penetrance of the latter, most HHC is related to the C282Y mutation. HHC genotype shows 1 in 10 of European Americans are carriers, while 1/250 are homozygotes (about 2 in 3 will evidence of iron overload).

About 7% of HHC patients have a mutation outside of HFE (the hemochromatosis iron gene). In these people, only a phenotypic diagnosis is possible. Phenotypic characteristics to consider are:

• Mildly elevated transaminases with normal serologies, low suspicion of alcohol abuse and low probability of fatty liver of obesity
• Fatigue and arthralgias in a middle aged male (about 0.5% incidence of HHC, compare with 1-4% incidence of hypothyroidism in similar aged female with fatigue)
• Patients with Fe/TIBC of greater than 50% (the higher the transferrin saturation, the more likely HHC, when over 70%, 97% specificity)

Should an at risk population be screened at the genetic level? The data isn't in, but a recent article suggests many at risk individuals go unrecognized (N Engl J Med 2000;343:1529-35). Zaneta et al., identified 214 homozygous relatives of 291 homozygous probands. Of the 113 men in this group (mean age, 41 years), 96 (85 percent) had iron overload, and 43 (38 percent) had at least one disease-related condition. Of the 52 men over 40 years of age, 27 (52 percent) had at least one disease-related condition. Of the 101 female homozygous relatives (mean age, 44 years), 69 (68 percent) had iron overload, and 10 (10 percent) had at least one disease-related condition. Of the 43 women over 50 years of age, 7 (16 percent) had at least one disease-related condition. The most common disease related condition was hepatic fibrosis/cirrhosis.

It may be premature to screen patients of Irish ancestry for HHC, but the condition is ripe for such screening. As the genotype and risk are well characterized, and the phenotype is hard to recognize.