Download Inherited Cancer and the Primary Care Physician

2583
American Cancer Society Second National Conference on
Cancer Genetics
Supplement to Cancer
Inherited Cancer and the Primary Care Physician
Barriers and Strategies
Hilary G. Worthen,
M.D.
Cambridge Health Alliance, Cambridge, Massachusetts.
Difficulties faced by primary care physicians as they increase their responsibility
for the diagnosis of inherited cancer risk include issues of cognitive strategy, the
context of care, and cultural and institutional factors. Charateristics common to
many genetic disorders—such as rarity, variability, implications for relatives, and
temporal pattern—render our usual cognitive strategies less effective. Constraints
of managed care, care teams, and high turnover of panels create a particularly
difficult context for the care of people at risk for inherited cancer. Echoes of the
eugenics movement, the implications of expanding genetic knowledge, and concerns about discrimination all complicate collaborative clinical decision making.
Eight strategies are suggested to cope with these barriers to diagnosis. Primary care
physicians also face challenges managing patients identified as at increased risk for
inherited cancer. These include confidentiality, coordination and communication.
Concerns for protecting the patient’s confidentiality can inadvertently leave primary care physicians with partial information. Coordination is complicated when
multiple organ systems and individuals are at risk, and knowledgable specialty
centers may be distant. Communication requires sensitivity and skill in translating
complex concepts from molecular biology and statistics into lay terms. Seven
strategies are suggested to help with management. Cancer 1999;86:2583– 8.
© 1999 American Cancer Society.
KEYWORDS: inherited cancer, primary care.
I
nherited forms of cancer are thought to account for about 5% of all
cancer cases.1 Although the presence of a genetic predisposition to
cancer substantially raises an individual’s risk for the disease, it also
offers opportunities, through pedigree analysis and genetic testing,
for surveillance, prophylaxis, and treatment, which can significantly
improve the outcome for many individuals. Much of the responsibility for identifying and pursuing these opportunities falls to primary
care physicians (PCPs). This article discusses the role of PCPs in the
care of people at risk for inherited cancer, some barriers to effective
implementation of that role, and strategies for overcoming these
barriers.
Presented at the American Cancer Society Second
National Conference on Cancer Genetics, San
Francisco, CA, June 26 –28, 1998.
Address for reprints: Hilary G. Worthen, Cambridge
Family Health, 237 Hampshire Street, Cambridge,
MA 02139.
Received May 25, 1999; accepted June 3, 1999
© 1999 American Cancer Society
CASE
A 43-year-old woman was referred to a neurologist by her psychiatrist
because of headache. Computed tomography (CT) of the head
showed the top of a fusiform swelling of the cervical cord, which on
further imaging was consistent with an ependymoma or a hemangioblastoma. Neurosurgical consultation was obtained and observation
recommended. Family history showed that the patient’s sister had
died at age 19 of a brain hemorrhage, and her father had died at age
2584
CANCER Supplement December 1, 1999 / Volume 86 / Number 11
48 of pancreatic cancer. She was referred to a PCP for
primary care and routine heath maintenance. A second neurologist advised a renal ultrasound to rule out
von Hippel-Lindau (VHL) disease. The findings on
ultrasound were normal. Follow up CT scans over the
next 2 years showed no change in the cord lesion. The
patient’s depression improved, and she made contact
with a daughter whom she had given up for adoption
at birth. The PCP was coincidentally involved with
organizing a conference on genetics that included a
panel on VHL disease. At the next annual check-up,
the PCP became concerned that VHL disease had not
been adequately ruled out and that the patient’s pedigree was strongly suggestive of it despite the absence
of renal cell carcinoma and cerebellar or retinal hemangioblastomas. Ultrasound now demonstrated
complex renal cysts and a pancreatic mass, which on
exploration was an inoperable islet cell tumor. Subsequent magnetic resonance imaging of the head
showed development of at least five cerebellar hemangioblastomas. The patient informed her daughter,
who broke off contact. It is not known whether the
daughter and her family have been evaluated for VHL
disease.
Ideally, all individuals at any increased risk for
inherited cancer should have the opportunity to find
out, with the maximum certainty, their actual risk and
should have access to the most effective management
for their situation. This must be done with a high
degree of respect for the autonomy of individuals and
the profound implications of genetic information not
only for the individual, but also for relatives and members of the same ethnic group. The PCP’s role in this
process can be considered in two parts: screening and
diagnosis, which includes both raising the possibility
of inherited cancer risk and taking steps to confirm it,
and management, which includes surveillance, preventive measures, and treatment.
SCREENING AND DIAGNOSIS
Aristotle identified and ranked six methods by which
discovery could take place in a tragedy.2 Of these,
bodily signs and marks (congenital or acquired) he
considered the least artistic and discoveries that come
about with surprise through the unfolding of events
the most powerful. Unlike the dramatist, the physician’s art is to avoid high tragedy, and to this end the
most mundane types of discoveries are far preferable.
Inherited cancer risk, however, as in the case above,
can all too often be tragic for the patient and devastating for the physician.
The process seems simple: the patient comes to
the PCP with or without specific concerns, the PCP
elicits a history, including a pedigree, and performs a
FIGURE 1. Schematic depiction of the process of identifying people in the
general population who carry a genetic predisposition to cancer.
physical examination and possibly some tests. Someone, either the PCP or the patient, raises the question
of inherited cancer risk, which leads to the gathering
of more detailed data about the family and more targeted examination and testing, until a very rough risk
assessment can be made. The PCP and the patient
discuss the risks and options, and a decision is made
about referral to a high-risk clinic for a full evaluation.
This process, applied over and over, should identify
those people in the general population who actually
carry a genetic predisposition to cancer, and allow
them to move through a series of subpopulations toward the opportunity for genetic testing, while filtering out those who are not at risk. (Fig. 1).
Experience suggests, however, that this process is
fraught with difficulties, which do not derive simply
from lack of information or poor clinical skills on the
part of the PCP. Setting aside the limitations of genetic
testing itself, the ideal scenario actually relies on a
collaboration between patient, PCP, family, and specialist, which is in reality poorly supported by the
training, culture, and institutions most of us work with
today. Several barriers to effective screening and diagnosis can be identified, and grouped into issues of
clinical reasoning, the context of care, and cultural and
institutional issues.
Clinical Reasoning
Genetic problems, including hereditary cancer, expose
some of the vulnerabilities of the cognitive processes
we rely on in diagnosis. Expert clinicians have been
shown to use an iterative process of hypothesis generation followed by refinement, until candidate hypotheses have been refined into working hypotheses
that can support further testing or treatment.3 Much
clinical knowledge is stored, it turns out, not in tables
Inherited Cancer and the Primary Care Physician/Worthen
or graphs, as in a textbook, but as clinical vignettes to
which facts are attached, and hypotheses are actually
generated by making quick comparisons with data
stored in memory as clinical stories. Several “heuristics,” or cognitive shortcut strategies, are commonly
used to allow efficient access to these data. These
include the “availability” heuristic (“I just saw a case
like this last week . . .”), the “representativeness” heuristic (“That is exactly like a case I recall where . . .”),
and the “criticalness” heuristic (“It is only a remote
possibility, but the consequences would be so bad that
I’d better consider it . . .”). These heuristics are like
cognitive assistants, which take the data about a patient, go back into the stacks of memory, and select
which hypotheses to float into consciousness.
Unfortunately, these heuristics are much less effective for most genetically determined disorders. The
large number of possible disorders makes the availability heuristic problematic even for geneticists, and
because of the enormous variability of genetic disorders (pleiotropy, variable expressivity and penetrance), the resemblance heuristic may be limited,
even if the very same mutation is present. A case of
inherited cancer is likely to bear most resemblance to
the much more common sporadic version. Findings,
which together would constitute a representative
“classic case,” may be spread out over a pedigree
rather than clustered in any one individual. Furthermore, the pedigree is a less available part of the patient database, infrequently revisited after it is initially
recorded and buried in the back of the chart. Finally,
the seriousness of many genetic disorders, although
profound, is not often acute and may be most important for individuals who are not even in the PCP’s
panel. This diminishes the utility of the criticalness
heuristic. Add to this the general infrequency with
which we seek genetic explanations for adult onset
diseases, and one can start to become skeptical of the
infallibility of our heuristic assistants in bringing forth
the hypothesis of an inherited cancer syndrome when
it is needed.
In addition to heuristics, experts also employ
“causal” reasoning, using knowledge of anatomy,
physiology, and pathology to reason out hypotheses
by a kind of brute force method. This route is also
especially vulnerable in genetic disorders: the defect at
the DNA level is many levels deeper than the actual
symptom or finding; there are innumerable opportunities for other genes and environment to exert effects;
and the molecular biology, biochemistry, and physiology that could explain the effects of a mutation are
only beginning to be understood.
This mismatch between cognitive strategies and
genetic problems may be one reason that patients
2585
with genetic disorders frequently remain undiagnosed
for years despite multiple encounters with clinicians.
For patients at risk for inherited cancer, such delays
can be costly.
Context of Care
The context within which primary care is increasingly
delivered creates additional barriers to effective
screening for inherited cancer risk. Many of those at
risk are not even in the population of people getting
medical care. Of those who are getting care, many are
receiving only episodic care for acute illness. Others
have frequent breaks in continuity of care caused by
mobility, job and insurance changes, and the increasing use of coverage “teams” in managed care settings,
which can diffuse both long-term responsibility for
and in-depth knowledge of a patient’s issues. More
and more of the PCP’s time with the patient is used in
negotiating insurance-related issues, such as switching to formulary drugs, seeing in-network specialists,
and managing referrals.
Most full time PCPs carry a panel of between 1000
and 2000 patients and see between 3000 and 5000
visits a year. There is up to a 25% yearly turnover of
patients in some health maintenance organization
(HMO) plans. Much of the utilization of visits is by the
elderly, with patients in the age groups when issues of
inherited cancer might be most important to address
getting care less frequently or primarily for family
planning and obstetrics. PCPs tend to do their work
with patients in short pieces over multiple visits:4 most
visits are booked for 10 to 20 minutes, and the physician who blocks out 30 minutes to see a patient these
days is not likely to remain in practice long.
The denominator of responsibility for PCPs these
days has increasingly become the “panel,” a group of
patients, some of whom the PCP may never have met,
who have chosen that PCP from a list. This panel may
have only marginal overlap with other denominators
that are important for genetic considerations: the family and the ethnic group. When the PCP’s responsibilities to that panel are discussed and evaluated, the
usual parameters considered are the HEDIS (Health
Plan Employer Data and Information Set) parameters,
by which insurance companies and HMOs display
their quality to employers who may be potential customers.5 These parameters, though laudable, do not
yet include specific measures to promote evaluation
for inherited cancer risk.
Confusion persists as well between screening of
the general population or a panel of patients who have
enrolled for medical care and case finding within a
group already identified to be at increased risk. Several
professional and consumer organizations have devel-
2586
CANCER Supplement December 1, 1999 / Volume 86 / Number 11
oped policy statements about screening that counter
the push by commercial interests to market genetic
tests.6 These policies range from advising access to
genetic testing for those at increased risk by family or
personal history for a select group of disorders, to
advising no use of genetic testing for cancer outside
the research environment. Although these policies distinguish DNA testing from the more general and more
broadly accepted cancer risk counseling process to
which it is an occasional adjunct, there is a tendency
to extrapolate their skepticism about DNA testing to
pedigree analysis and risk assessment for individuals.
This context, although yielding cost-effective care
for many problems, presents barriers to evaluation for
hereditary cancer risk. First, patients often have incomplete data about their families, and if this issue is
not pursued over time, the original incomplete pedigree may be left to stand in the chart as a misleading
signal that no more attention need be focused on it.
Secondly, families often do not collaborate in exactly
the way that would best facilitate the gathering of such
data, and sometimes the help of the PCP is needed.
This may require the PCP to become involved with
people outside his or her panel. Thirdly, the issues
involved are extremely complex and serious, and the
task of empowering patients without under- or overpowering them is nowhere more delicate and time
consuming. Finally, for the PCP trying obediently to
stay “in-network,”there may be a hesitancy to explore
referrals to a high-risk cancer clinic that may be unaffiliated and appear likely to perform expensive tests
for dubious gain. Yet the usual in-network specialists
may not be knowledgeable about unusual genetic disorders or equipped to perform the appropriate evaluation, a process that usually involves a team including
genetic counselors.
Cultural and Institutional Barriers
The legacy of the eugenics movement, with its history
of a small number of people imposing unwanted reproductive choices on a large number and justifying it
with crude and usually inaccurate science, is still a
force in the thinking of many health care providers,
consumers, and ethnic and advocacy groups.7 Many
thoughtful people also share a queasy feeling that
there is something antidemocratic, anti-American,
and antireligious about a science that seems to demonstrate that we are not all born equal, that what
happens to you has a lot to do with the family you
were born into, and that diseases such as cancer may
not always be a punishment for bad behavior or moral
failings. One of the core paradoxes of American culture, that we are all unique, independent, and autonomous, yet are all identical in our rights and before the
law, seems in danger of being rewritten: we are apparently all closely related and remarkably the same,
across families, ethnic groups, and even species, yet
have simultaneously so much variation that no one is
invulnerable to disease or discrimination. Institutions,
from the law to insurance companies and employers,
have been slow to catch up to the burgeoning knowledge of genetics and its implications and to develop
safeguards against discriminatory use of information.
The very language of genetics is a kind of creole, a
mixture of words and concepts from molecular biology, biochemistry, population genetics, all the way to
ethics and philosophy. Many terms, such as “clone” or
“hybrid” are used differently in each of these contexts.
This leaves PCPs trying to communicate with patients
using metaphors and analogies, while they themselves
are often wondering what geneticists are talking
about. This may also contribute to the observation
that people with high educational levels are overrepresented among those undergoing genetic testing for
hereditary cancer
Finally, the concepts of “resonance” and “coherence” have been used to explain why it is more difficult and stressful for patients and PCPs when a patient’s story and situation do not resonate with
commonly known stories and may even appear incoherent because of an incomplete scientific explanation.8 In the cultural analogue to the “availability”
heuristic, unusual forms of disease, such as the inherited cancer syndromes, are less likely to resonate or
seem coherent compared with more common forms.
This catalog of difficulties can make it challenging
for the PCP to accomplish the goal of being sure that
patients with hereditary cancer syndromes have the
opportunity to understand their risk and make informed choices about diagnosis and management.
The information needed to trigger the right hypothesis
may not be forthcoming; the PCP may not generate
the right hypothesis; and the time, resources, and
continuity to pursue that hypothesis may not be available. The potential for misuse of genetic information
may deter both patient and PCP. Skepticism about
population screening and genetic testing in general
may affect readiness to examine pedigrees and pursue
risk assessment. In this process, opportunities to prevent illness and death may be missed.
Strategies for Diagnosis
Once aware of these barriers, the PCP can use several
strategies to overcome them (Table 1). To encourage
the surfacing of a genetic hypothesis, simply asking
“could this be genetic?” when making a diagnosis can
yield occasional dramatic benefits. The extra effort to
try to develop a single unifying hypothesis to explain
Inherited Cancer and the Primary Care Physician/Worthen
TABLE 1
Strategies for Raising a Genetic Hypothesis
Set a low threshold for genetic hypotheses
Try to develop unifying hypothesis for disparate findings
Maintain and update the pedigree
Look for clues in presentation and setting
Distinguish sporadic, familial, and inherited cases
Consider variation
Become familiar with resources
Allow time
disparate and apparently unrelated findings is another
exercise that geneticists have found useful. With minimal instructions, patients can often develop a very
full pedigree, which can be reviewed for new developments at the time of annual preventive health checkups. The PCP can be alert to unusual presentations of
common problems, such as bilateral cancers in paired
organs or onset of a tumor in an unusually young
person. The setting, such as ethnic group or habits,
can also be revealing.
It is important to distinguish among sporadic
cases, familial cases (at least one relative with the
disorder but no clear mendelian pattern or known
mutation), and inherited cases (a single gene mutation
playing a causative role). Most situations in which a
concern is raised about inherited cancer actually fall
into the familial category, felt to reflect the effects of
several genes, genes plus environment, or the occasional gene mutation of low penetrance. Awareness of
this fact keeps the PCP from being overly suspicious of
inherited cancer or becoming dissuaded from looking
for it by frequent negative results. Recalling that “inherited” means “genetic” alerts the PCP to common
characteristics of genetic disorders, such as rarity,
pleiotropy, and variable penetrance.
Electronic knowledge bases are effective tools for
broadening the range of hypotheses and preventing
one from overlooking an entire category of etiology.
Online Mendelian Inheritance in Man (OMIM)9 can be
used to get quick answers to questions such as, “could
this finding be due to a genetic disorder?” or “what
other features should I look for in evaluating this problem?” All PCPs should become aware of an appropriate high-risk cancer evaluation center and of an expert
to call for guidance should concern arise about inherited cancer. Even a rough risk assessment can be quite
complicated and may require the advice of a genetic
counselor. It is also easier for the PCP to discuss a
referral with a patient if the PCP has a firm idea of
whom the patient will see and what will take place.
Keeping in mind the distinctions between population screening and assessment of risk for an individual who has sought medical advice will keep skepti-
2587
cism about widespread or premature use of DNA
testing from inhibiting pedigree analysis or consultation with a high-risk cancer clinic. Risk assessment
does not equal DNA testing: only a subset of those
referred for evaluation will be found to be appropriate
for testing and an even smaller subset will choose to
have it.
Finally, there is no substitute for time. The key
piece of information may not emerge from the woods
until the noise of structured data collecting has quieted down and the patient and PCP have a moment to
reflect with each other as two people trying to solve a
problem.
MANAGEMENT
A different set of challenges faces the PCP once a
patient is found to be carrying a mutation for an
inherited cancer syndrome. These fall into the categories of confidentiality, coordination, and communication.
Confidentiality
The sensitive nature of genetic information, with its
implications for the future health of the patient and
others as well, suggests that it should be handled
differently from other medical information. Even
speculation in the chart about a possible inherited
cancer syndrome, let alone a genetic testing result,
could have a negative impact if the patient applies for
insurance and signs an innocuous-looking records release form. However, any suppression or sequestering
of information runs the risk of defeating the original
purpose of the medical record to assist the physician
in the care of the patient. The problem can become
paradoxically worse when the patient is seen at a
high-risk clinic under a research protocol, which can
maintain higher standards of confidentiality. This may
result in the PCP being given partial or indirect information or being left entirely in the dark. The PCP may
also need to develop strategies with the patient for
discussing the results with other family members
whose attitudes toward the information may vary.
Coordination
Although the established referral networks may be
adequate for most problems, inherited cancer syndromes pose special difficulties. Family members
from different geographic areas and with different insurance coverages often want to be observed by the
same specialists. Many syndromes cause problems in
a variety of organ systems that require the coordination of multiple specialists. For the less common syndromes, there may be only a few groups in the country
with extensive experience.
2588
CANCER Supplement December 1, 1999 / Volume 86 / Number 11
TABLE 2
Strategies for Management of Inherited Cancer
Encourage patient to become expert
Use voluntary organizations
Maintain a summary with references
Be sure the patient has full data set
Keep a log of communications with consultants
Locate and engage the most expert resources
Deal prospectively with managed care issues
Communication
Not only must the PCP learn some of the language of
molecular biology to understand the patient’s problem, but he or she must also become adept at communicating about it with a variety of people, ranging
from patients and family who may have little education or major cultural differences from the PCP, to
specialists and reviewers who may be quite unfamiliar
with the problem. The PCP may have to function as a
clearing house for information coming in from a farflung group of consultants, all the while assuring that
the patient has control over the dissemination of any
information about the patient’s condition.
These problems of communication, coordination,
and confidentiality can increase and broaden the
workload of the PCP and lead to frustration, blaming,
and less effective care.
depth of experience invaluable. Finally, it pays to deal
proactively with the patient’s managed care plan because many plans can assign a case manager who can
be a facilitator and advocate around any unusual referrals.
SUMMARY
The identification of inherited forms of cancer and the
development of diagnostic and therapeutic tools for
them have opened possibilities for helpful interventions. There are significant barriers for the PCP that
make the processes of screening, diagnosis, and management of inherited cancer syndromes challenging.
More than just knowledge of these syndromes is
needed: the PCP must develop general strategies, be
prepared to use unaccustomed resources, and expect
to invest extra time to optimize the care of people at
risk for inherited cancer. Within the health care system, a growing awareness of these disorders will facilitate the necessary communication and referrals. As a
society, we need continued progress toward destigmatization of genetic disorders and institutional protection against inappropriate use of medical information.
Until then, the major role the PCP plays in diagnosing
and managing patients with these disorders will remain complex and difficult.
REFERENCES
1.
Strategies for Management
Several strategies are helpful in overcoming these barriers (Table 2). It is a worthwhile investment of time as
well as good clinical care to help patients become as
expert as possible about their problems. The voluntary
organizations are an excellent source not only of support but also of information ranging from pragmatic
solutions for everyday problems to recent scientific
advances. Their materials are an invaluable educational resource for both patient and PCP. Keeping an
ongoing summary, complete with pedigree, work-up
and references, that can be printed up as needed
simplifies the task of bringing new providers and consultants up to speed. It is also helpful if the patient
maintains a full set of records and data to take to
consultations. Nowhere is the habit of keeping a running log of discussions with consultants more rewarding because the number of these transactions can
grow geometrically. The PCP should be ready to locate
and seek consultation from a multidisciplinary team
that has extensive experience because the variability
inherent in genetically determined disease renders
2.
3.
4.
5.
6.
7.
8.
9.
Olopade OI, Blackwood MA, Cummings S, Davis JA, Garber
JE, Lynch PM. Cancer genetics and cancer predisposition
testing: an ASCO curriculum. Vol. 1. Alexandria, VA: American Society of Clinical Oncology; 1998.
Aristotle. Poetics. In: Introduction to Aristotle. McKeon R,
editor. New York: McGraw Hill; 1947:644 –5.
Kassirir JP, Kopelman, RI. Learning clinical reasoning. Baltimore: Williams & Wilkins; 1991:1– 42.
Touchette N, Holtzman NA, Davis JG, Feetham S. Toward
the 21st century: incorporating genetics into primary health
care. Plainview, NY: Cold Spring Harbor Laboratory Press;
1997:47.
Epstein AM. Rolling down the runway: the challenges ahead
for quality report cards. JAMA 1998;279:1691– 6.
Kodish E, Weisner GL, Mehlman M, Murray T. Genetic testing for cancer risk: how to reconcile the conflicts. JAMA
1998;279:179 – 81.
Kevles DJ. In the name of eugenics: genetics and the uses of
human heredity. Berkeley: University of California Press;
1986.
Weingarten K, Worthen ME. A narrative approach to understanding the illness experiences of a mother and daughter.
Fam Syst Health 1997;15:41–54.
Online Mendelian Inheritance in Man, OMIM (TM). Center
for Medical Genetics, Johns Hopkins University (Baltimore,
MD) and National Center for Biotechnology Information,
National Library of Medicine (Bethesda, MD), 1997. World
Wide Web URL: http://www.ncbi.nlm.nih.gov/omim/