Phenotypic Variation in Hereditary Breast Cancer

Cancer Control Implications

Henry T. Lynch, MD; Theresa A. Conway; Jane F. Lynch

Objective: To study hereditary breast cancer pedi- grees as models for the elucidation of the natural his- tory of the disease, including early onset; bilaterality; and tumor associations in the heterogeneous variants to aid in its diagnosis and in targeting surveillance and treat- ment strategies.

Setting: Oncology clinic and hereditary cancer institute.

Patients and Other Participants: Physician- or self- referred probands.

Interventions: None.

Main Outcome Measure: Greater understanding of the diagnosis, surveillance, and treatment of hereditary breast cancer.

Results: Eighteen pedigrees displayed the phenotypic variation of breast cancer’s natural history, obligate gene carriers, and patterns of tumor combinations consonant with genetic heterogeneity.

Conclusion: Hereditary breast cancer is an important public health problem accounting for about 9% of breast cancer cases. The recent identification of a molecular ba- sis for the identification of a subset of hereditary breast cancer, and thereby the likelihood of the sequencing and cloning of the susceptibility gene(s), will enable target- ing of surveillance and treatment measures toward pa- tients at an inordinately high risk of developing cancer. Central to this entire process will be the identification of families with hereditary breast cancer.

(Arch Surg. 1994;129:806-813)

From the Department of Preventive Medicine and Public Health, Creighton University School of Medicine, Omaha, Neb.

A LL WOMEN are not at equal risk of developing breast cancer.1.2 Host factors play an extremely important role in determining who will or will not show manifestations of this disease.2 The highest known risk of de- veloping breast cancer occurs among women from hereditary breast cancer (HBC)-prone families, who are carriers of highly penetrant genes that predispose them to breast cancer. Hereditary breast cancer is not rare. Approximately 5% to 9% of breast cancer occurrences may have a primary genetic etiologic factor.3

The ability to identify members of breast cancer-prone families with the del- eterious gene was improved significantly when Hall et al+ discovered linkage to a gene, BRCAI (designated 1137055), on chromosome 17q12-q21 among early- onset breast cancer-prone families.6 Narod et al7 linked the hereditary breast-

ovarian cancer (HBOC) syndrome also to the 17q chromosome locus. We now know that only about one half of families with site-specific HBC show linkage to the 17q locus, while about 90% of families with HBOC are linked to 17q.8,9 A much rarer breast cancer-associated disorder, the Li- Fraumeni/sarcoma; breast and brain tu- mor; lung, leukemia, lymphoma; and ad- renal cortical carcinoma (SBLA) syndrome, has been linked to a heritable mutation in the p53 gene on chromosome 17p,10 mak- ing presymptomatic identification of car- riers possible in this disorder as well.

Our purpose was to discuss differ- ent HBC pedigrees and show how an HBC

See Materials and Methods on next page

MATERIALS AND METHODS

The study was approved by the Institutional Review Board of Creighton University, Omaha, Neb. The pro- bands were either self referred or physician re- ferred. The goals and objectives of this research were explained to the patients, and they were told that they could withdraw from the study at any time without prejudice.

Whenever possible, a personal interview was conducted with the proband. Cancers of all ana- tomic sites in the patient and his or her first- and sec- ond-degree relatives were recorded. Permission was sought to contact the cancer-affected relatives or, if deceased, their next of kin and other family mem- bers who could furnish information to extend the pedigree through remote relationships. Question- naires on genealogy and cancer history and permis- sion forms for release of primary medical and pa- thology documents were sent to these relatives. This information enabled us to cross-reference historical accounts, thereby improving accuracy. Whenever pos- sible, our collaborating pathologist reviewed slides and tissue blocks from the cancers.

diagnosis can be established when faced with genetic het- erogeneity, variable expressivity of phenotype, particu- larly tumor spectra and age at cancer onset, in addition to reduced gene penetrance.

RESULTS

There is a plethora of phenotypic and genotypic nu- ances in the 18 pedigrees (Figure 1 through Figure 5) that we have described. Since the possible permutations of these findings are legion, we only focused on those fea- tures that were believed to be clinically relevant.

PUTATIVE SITE-SPECIFIC HBC FAMILIES

The pedigrees in Figure 1 show examples of putative site- specific HBC, although the presence of other cancers may suggest other HBC syndromes. Some of these families show marked variation in age at the onset of breast cancer.

Family 3395 was a pedigree that depicted late- onset breast cancer in the proband (III-8), her two sis- ters (III-4,12), and their mother (II-1). The proband’s daughter (IV-6) had premenopausal breast cancer, as also did the proband’s niece (IV-2), whose father (III-1), by virtue of his position in the pedigree, was a likely obli- gate gene carrier (OGC). Thus, this pedigree shows both late and early onset of breast cancer.

Family 3207 was a small family with variable ages at breast cancer diagnosis. We consider this to be a site-

specific HBC family on clinical grounds. Given the oc- currence of breast cancer in the proband at age 38 years, we consider her three daughters (V-1) and sister (IV-3) to be at increased risk of developing breast cancer.

Family 3442 had a strikingly early onset of breast cancer in the proband’s two daughters (IV-4,5), both of whom underwent prophylactic contralateral mastec- tomy.

POSSIBLE HBOC SYNDROME FAMILIES

Breast cancer was the predominant lesion in the pedi- grees in Figure 2. However, each family had one geneti- cally informed woman in the pedigree who had ovarian carcinoma, suggesting the possibility of the HBOC syn- drome. For example, in family 3300, we considered pa- tient IV-1, who had ovarian carcinoma, to be an OGC, based on breast cancer occurrences in her mother (III- 1), sister (IV-2), and daughter (V-2).

In family 3312, patient II-2, a man who had laryn- geal carcinoma, was considered to be an OGC because of his position in the pedigree; specifically, his mother (I-2) and four of his daughters (III-3,6,7,11) had breast cancer and a fifth daughter (III-5) had ovarian cancer at an early age.

Family 3302 included five sisters, of whom three had late-onset breast cancer (II-2,3,5), one had urinary blad- der cancer (II-4), and the fourth (II-1) had ovarian can- cer at the late age of 71 years. She and one of her sisters (II-2) appear to be OGCs, because they had daughters (III-2,4) with breast cancer.

Family 3268 had originally reported “uterine can- cer” in patient II-2. Pathology documents verified the ovary to be the primary site. Her sister (III-1), mother (II-1), and two maternal aunts (II-2,3) had all shown manifestations of breast cancer.

Family 2970 had two OGC men (III-2,6). Both of their mothers (II-3,4), who were sisters, had breast can- cer, and each of these men had a daughter (IV-2,5) with breast cancer, of whom one (IV-2) also had a synchro- nous ovarian carcinoma.

Family 3511 shows father-to-son transmission of breast cancer. The proband (III-4) had early onset (age 46 years) breast cancer and his father (II-4) had breast can- cer, carcinoma of the prostate, and carcinoma of the lar- ynx. Carcinoma of the breast and ovary occurred in one of the father’s sisters (II-3). One of the proband’s sisters (III-3) had breast cancer, while another sister (III-2) had a metastatic lesion in the lung (adenocarcinoma), pri- mary site unknown, which contained psammoma bod- ies. The pathologist could not exclude an ovarian origin.

HBOC SYNDROME FAMILIES

The pedigrees in Figure 3 provide more secure evidence for an HBOC syndrome diagnosis. Families 3294 and 2973

Family 3395

Family 3207

Family 3442

1

2,

1

I

1

I

2,

I

2

St83

d.84

d.

d.

St.

d.83

d.

d.

1

2

3

4

5

Ii

6

7,

HI

1

2

2)

II

1

2

Br70

d.77

d.55

83

Pro

Cx

d.84

d.79

Pro70

Br58

St

St79

d.78

85

95

d.53

d.83

d.78

d.60

d.47

11

III

2

3

4

5

6

7

Y

8

9

10

1

12

3

14

15

16

II

1

2

3

1

Y

2

4,

3

8

5

III

3

5

O

82

Co49

LU74

Br70

d.Inf

En67

d.58

Br67

67

d.20

63

Br59 50-64 60

70

45

Abd/

Br63

r48

Br51

Br47

d.69

Pro74

77

d.Inf

72

68

48

Sar11

71

Br50

48-61

76

Mou64

67

d.60

₫.20

1.

78

EL

d.20

d.11

sp61

Br65

1/

73

1

2

3

4

5

6

7

8

9

10

11

12

d.61

Y

1

2

3

4

1

2

3

4

6

IV

4

16

2

5

+

5

2

2

3

IV

5

7

2

3

IV

2

49

Br41

24-48

42

37-43 Br40

30-45

37

32

32

34-37

Br38

48

38

40

Br52

50

50

Br31

Br33

50

51

49

46

39

54

52

49

-40’s

45

49

54

49

35

53

10

54

1.

1

2

V

2

1

V

3

9

4

3

6

~20-30

13-23

23

19-25

. Inf

28

30

MaleFemaleMaleFemale
12Individual No.
Unaffected☐ *☐ *Multiple Primary Cancers by Pathology
2833Current age, y
Cancer by Pathology Report+Cancer by Death Certificate or
Lu53Br45Age at Diagnosis, yMedical Records
5547Current age, y
Cancer by Family History.No. of Unaffected Progeny (Both Sexes)
d.54d.86Age at Death, y
☐ AIn Situ Carcinoma by Pathology Report.Proband
Multiple Primary Cancers UnverifiedTwins Identical Twins
☒ 1☒ AMultiple Primary Cancers by Medical Records or Death Certificates/No Children = Never Married
Cancer Sites
AbdAbdominalLyxLarynx
AdAdrenal corticalMmelMalignant melanoma
BIBladderMouMouth
BrBreastOvOvarian
BtBrain tumorPanPancreas
CoColonProProstate
CsuCancer site unknownPsuPrimary site unknown
CxCerixRhRhabdomyosarcoma
EnEndometriumSarSarcoma
EsoEsophagusSigSigmoid colon
HoHodgkinsSkSkin
LeiLeiomyosarcomaStStomach
LeuLeukemiaThyThyroid
LiLiverUtUterus
LuLungWtWilm's tumor
LymLymphoma

Figure 1. Pedigrees depicting kindred with diagnoses of putative site-specific hereditary breast cancer syndrome. Symbols and abbreviations are explained at bottom.

were consistent with HBOC, but the genetic status of the probands (V-1) in both families remained unclear since at the time of this writing their mothers (IV-1) were free of cancer, and they may not have inherited the breast can- cer susceptibility gene; however, the probands’ mater- nal grandmothers (III-1) had ovarian cancer, and we noted the occurrence of both ovarian and breast cancer through- out the families. The probands’ hereditary cancer risk was therefore 25%, based on a postulated Mendelian auto- somal dominant mode of inheritance in these families. If their mothers develop breast or ovarian cancer, then the probands’ risk would become 50%.

Given the occurrence of ovarian cancer in family 3350, it is possible that one or all of the proband’s rela- tives with “abdominal cancer” or cancer site unknown (1-2, II-1,2) had ovarian cancer.

Family 3286 had ovarian cancer in two genetically informed patients (II-1, I1-5) who had daughters (III-1, III-6) with early onset breast cancer.

LI-FRAUMENI/SBLA SYNDROME FAMILIES

The findings in families 2152 and 2943 were consistent with the Li-Fraumeni/SBLA syndrome. Family 2152 was an abridged pedigree of one of the original SBLA fami- lies that was first described by Bottomley et al11 and sub- sequently updated by Lynch et al.12,13 The progenitors (I- 1,2) had 10 children; half of them, by family report, had cancer. One of these 10 children, noted on the pedigree

as II-1, had descendants with a remarkable excess of sar- comas, as evidenced in the proband (V-9), in three of his siblings (V-7,8 and V-10), and in his maternal grandfa- ther (III-5). A rhabdomyosarcoma was seen in the pro- band’s daughter (VI-1). Bilateral breast cancer occurred in the proband’s maternal aunt (IV-4) at age 37 years and leukemia at age 47 years. This lady had three affected chil- dren, two with leukemia (V-1,2) and one with adrenal cortical carcinoma (V-3).

Family 2943 consisted of relatively few members, but showed the occurrence of cancer sites consonant with the SBLA syndrome. The proband (IV-3) had early on- set pancreatic cancer. However, two of his brothers (IV- 1,2) had leukemia and his sister (IV-6) had a rhabdo- myosarcoma at age 2 years and intraductal carcinoma of the breast at age 24 years. Another significant finding was that of a Wilms’ tumor in the proband’s daughter (V-2), who died at age 2 years, and adrenal cortical carcinoma in his son (V-3), who died at age 3 years. The proband’s mother (III-1) had early onset breast cancer.

Although family 3472 was relatively “small” at its initial ascertainment (top), it nevertheless showed pre- sumptive evidence of HBOC. Information on the family was extended (bottom) when the proband’s maternal aunt (III-3), who had been affected with ovarian carcinoma at age 56 years, provided information that her paternal aunt (II-6) and two of her daughters (III-11,13) had ovar- ian cancer and a third daughter (III-12) had both breast and ovarian carcinoma. Another aunt (II-5) of the infor-

Figure 2. Pedigrees showing predominance of breast cancer, as well as single occurrence of ovarian cancer in a genetically informed person. See Figure 1, bottom, for an explanation of symbols and abbreviations.

Family 3300

Family 3312

1

2,

1

2

1

I

Li86

d.

d.

Br

d.86

d.32

1

2

3

4

1

3

H1

5

II

Pro60

d.

d.

Csu

Br

Lyx48

d.66

d.55

d.78

d.50

1

2

3

4,

5

6

17

1

2

Y

3

4

5

6

7

8

9

10

11

III

2

2

111

4

#

2

Br42

d.58

Co79

d.65

d.

d.

d.59

d.54

~41-51

Br39

Co49

Ov39

Br45

Ut24

42

41

41

Br29

d.43

85

d.69

Br41

52

d.40

47

Br33

45

d.36

53

Br39

t

2

3

4

5

6

7

8

9

10

11

3

4

0.40

6

7

IV

3

=

IV

1

2

5

3

2

3

3

5

6

4

Ov47

Br4

52

57

Br36

43

d.

Pan64

BI54

d.19

Csu

15-36

20

28

24

20

15-24

14-20

6.55

60

55

39

75

66

-70-80

24

27

22

57

27

30

29

24

V

1

Y

2

3

2

2

33

Br35

30

38

36

31

Family 3302

Family 3268

Family 2970

Family 3511

1

2,

1

2

1

2,

1

2,

I

I

İ

I

d.73

d.97

d.

d.

d.79

d.80

d.

d.

1

2

3

4

5

1

2

3

4

1

2

3

4

1

U

2

3

4

II

=

H1

II

II

·

Ov71

Br67

Br60

BI62

Br63

Br41

Br

Br

d.

Br

d.Inf

Br

Br35

Lu

81

Br

Br65

d.72

76

d.66

d.64

66

d.78

d.

d.

~d.52

d.49

d.37

d.80

Ov

Pro66

Lyx67

11

79

0.68

3

5

1

3

1

2

3

4

5

6

1

2

3

4

HII

1

Y

2

4

3

III

2

4

2

III

H

I!I

2

Y

Thy46 48

Br33

47

Br29

42-46

Br47

-60

~82

80

Pan

Br57

Psu68

57

Br52

Br4

45

36

d.52

0v47

d.Inf

d.47

d.79

d.

Br58

58

Psu52

d.68

d.54

52

51

1/L

64

IV

Y

1

2

13

IV

1

2,

3

4

Y

5

6

IV

1

2

3

2

3

4

3

34

30

~50

51

Br41

51

32

Br37

30

30-36

14-28

23

Ov41

34

34

d.43

41

36

mant who was cancer-free but was assumed to be an OGC, had four sons (III-6,7,9,10) with prostate cancer and a daughter (III-8) with breast cancer. This new informa- tion shown in Figure 5, bottom, supported our suspi- cions that patient II-1 was, indeed, an OGC for HBOC. It is an excellent example of how information can change over time (in this case, only 1 month) and help in mak- ing a genetic diagnosis.

COMMENT

Hereditary breast cancer is extremely heteroge- neous.12,14 It is essential that the physician comprehend the clinical implications of this heterogeneity. A well- orchestrated family history of cancer that includes ma- ternal and paternal lineages, with particular attention to cancer of all anatomic sites and any variation in its phe- notypic expression, particularly age at onset and pat- tern of multiple primary cancers, is mandatory for the elucidation of the significance of the pedigrees. At first glance, accomplishing this in the clinical setting may ap- pear to be a formidable task; however, it can be per- formed successfully by trained paramedical personnel. This will then free up the physician to devote his or her time to pedigree interpretation and, when indicated, to the implementation of a highly targeted surveillance and treatment program.

Occasionally the clinical findings in a single af- fected patient may provide important clues to the exist- ence of a potential primary genetic etiologic factor. For example, consider a woman with bilateral breast cancer in her mid-20s or early 30s, but whose family history is “negative” for breast cancer. Because of the early onset and bilaterality, her sisters and daughters should be con- sidered to be at potentially increased risk of developing breast cancer. The genetic significance, if it exists, may have been obfuscated by a number of possibilities in- cluding small family size, paternal transmission, nonpa- ternity, reduced gene penetrance, variable expressivity in age at onset of breast cancer (her mother or other key relatives may become affected at a later age and only then become genetically informative), incomplete documen- tation of cancer in the family, or a new germ line muta- tion. Confidence in a diagnosis of HBC syndrome will be strengthened, however, as additional cases of cancer are identified, as occurred in one of our families (family 3472). Once a diagnosis of HBC is established, appro- priate surveillance and treatment protocols can then be initiated (Table).

The HBC pedigrees in Figures 1 through 5 display a variety of clinical nuances that require careful atten- tion when assessing the diagnostic possibilities in breast cancer-prone families. These include such natural his- tory features as early onset (but in some families we find

Figure 3. Pedigrees of kindred with more secure evidence for hereditary breast-ovarian cancer syndrome. See Figure 1, bottom, for an explanation of symbols and abbreviations.

Family 3294

Family 3350

1

2

1

2

1

I

d.

d.

d.

Abd/

Csu

d.69

Il

1

2

3

4

II

1

2

3

A

Ov47

Ov54

Br61

BI

Abd/

Csu

d.80

Csu 46

Br56

d.49

d.55

d.62

d.39

Br56

d.49

IF0

d.75

III

1

2

3

4

5

6

7

8

9

10

11

III

2

3

.

4

5,

11

2

=

*

P

0v69

BI

Ov

Ov45

d.63 d.64

70

Br55

Ov

Br48

d.61

Lu47

Br41

Csu52

61

Br36

Ov45

71

d.64

d.51

d.46

68

d.69

d.48

d.48

d.44

d.52

Sig56

Ov56

d.54

57

4

6

IV

1

2

3

5

7

8

9

10

11

12

13

14

15

1

2

3

4

5

6

va

.

2

4

2

5

2

2

2

2

3

IV

3

2

3

3

46

47

47

48

Br24

-40

~40

Br41

36

31-43

38

39

-50’s

50’s

20

Br38

33

23

26

22

Br24

8r41

48

46

40

43

23

39

36

31

27

26

42

1|1

26

d.26

1|”

23

30

1/

Y

1

2

3

4

5

7

8

V

6

2

3

3

5

3

26

26

23

13-2816-23

29

20

21

~20

23

22

26

26

Family 2973

Family 3286

1

2

1

I

1

2

d.

d.

d.37

Abd/

Psu

d.46

HI

1

2

3,

4

II

1

2

3

4

3)

5

5

2

E

Br56

d.

Ov

d.

Ov55

70

Br42

Leu

d.78

d .~ 59

d.60

72

Ov43

d.46

d.62

d.45

1

ıp

I

III

1

2

3

4

5

6

7

8,

1

5

6

7

2

III

Y

2

3

4

4

4

4

Ov40

d.43

Br45

67

Br47

Ov51

Pro76

d.44

d.38 d.44

Br35

~30’s

46-51

42-50

Br31

Br33

42

‘‘

d.51

d.50

d.57

76

39

40

44

IV

1

2

3

4

5

6

7

8

9

10

11

12

13

2

T

3

2

2

61

d.Inf

Ov52

Br60

58

38

46

Br29

Br36

44

43

d.54

-40-50

61

Ov39

d.41

41

d.39

48

48

44

59

Y

1

2

3

4

5

6

7

8

9

10

V

2

2

4

3

3

3

5

35

d.Inf

29

27

33

7

17

24

13-20

36

29

0 1.23

35

17

21

27

39

22

22

38

40

Figure 4. Pedigrees of two families with diagnoses of Li-Fraumeni/sarcoma; breast and brain tumor; lung, leukemia, lymphoma; and adrenal cortical carcinoma syndrome. See Figure 1, bottom, for an explanation of symbols and abbreviations.

Family 2152

Family 2943

I

1

2,

I

1

2,

d.

d.

d.

d.

II

1

II

y

2

3

4

1

3

Pro

d.62

d.76

d.

Lu

~70’s

-70’s

III

1

2

3

4

5

1

3

3)

III

d.10

90

d.Inf

Br40

Sar46

Br42

d.26

d.69

IF

d.9

d.51

d.61

d.42

d.28

IV

U

2

3

4

5

6

1

2

.

3

4

5

6

IV

2

=

d.Inf

59

66

Br37

B+37

Lei59

60

Leu

Leu17

Pan37

40

35

Rh2

72

Leu47

d.59

d.33

d.17

d.37

44

Br24

d.47

11

28

V

1

2

3

4

5

6

7

8,

9

10

11

12

1

2

3

4

5

6

2

2

Y

*

5

V

4

2

3

Leu14

.eu10

Ad

38

d.Inf

35

Sar10

Sar14

Sar33

Sar22

29-40

Ho28

10

Wt

Ad2

8

6

-4

d.15

d.10

d.inf

40

d.11

d.16

39

Sar27

38

12

d.2

d.3

11

22

31

19

24

1

2

3

20

VI

P 2

Rh2

12

12

8

17

both early and late onset),15 an excess of contralateral breast cancer,16,17 and specific tumor associations con- sonant with the HBOC17-19 and the Li-Fraumeni20/SBLA syndromes.21 It is important to realize that phenotypic variation is a natural accompaniment of all forms of he- reditary disorders (of which there are more than 3000),5

particularly those with autosomal dominant inherit- ance patterns.

In certain pedigrees, as already acknowledged, there may be insufficient information to establish a heredi- tary diagnosis, but on clinical grounds, surveillance rec- ommendations can often be made for first-degree rela-

Figure 5. Depiction of pedigree at time of initial ascertainment (top), and extension of pedigree (bottom) after information received from key family member III-3, which led to the identification of hereditary breast-ovarian cancer syndrome. See Figure 1, bottom, for an explanation of symbols and abbreviations.

Family 3472

1

2

I

Pro

d.78

d.84

2

3

4

II

1

3

Ov40

Br38

Ov56

55-61

d.41

Br42

64

d.71

1

2

3

4

5

HII

6

7

8

2

10

50

41

1

41

Br26

Br46

50

46

28-40

33

51

48

1

IV

1

2

3

4

5

6.

7

3

29

20

17

9

27

d.18

32

22

13

29

32

Family 3472

1

I

2,

d.83

Ov

d.53

1

2

3

4

5

6

7

8

10

II

19

11

1

Pro

d.84

72

Cx

d.80

Ov

d.80

Lu63

Pro74

Ov56

Csu

d.78

d.66

d.68

d.65

d.82

d.56

d.66

III

1

2

3

4

11

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

3

2

I

2

6

Ov40

Br38

Ov56

55-61

73

Pro74

Pro63

Br62

Pro65

Pro63

Ov42

Br

Ov63

49-70

d.55

Co54

49

48

39-60

-50’s

d.41

Br42

64

79

77

d.66

71

66

68

8.43

Ov

d.65

d.73

60

50

d.71

77

IV

F

2

3

4

5

6

7

8

2

50

41

41

Br26

Br46

50

46

28-40

33

51

48

1

2

3

4

5

6,

7

V

2

2

3

29

20

17

9

27

d.18

32

22

13

29

32

tives of patients affected with breast cancer (Table) as seen in three daughters of the proband (IV-1) in family 3207 (Figure 1).

VERIFICATION OF PATHOLOGY REPORTS AND EXTENSION OF FAMILY HISTORY FOR HBC SYNDROME DIAGNOSIS

The effort of verifying a diagnosis and extending the fam- ily history can be of benefit. For example, extending the family, as seen in family 3472 (Figure 5) and persisting in the search for pathology verification of cancer as evi- denced in family 3268 (patient III-2, Figure 2) lent credence to a putative diagnosis of HBOC in these kindred.

VARIATION IN AGE AT ONSET

The average age at the onset of breast cancer in HBC is the mid-40s, about 15 to 20 years earlier than its spo- radic counterpart.15 In our pedigrees, there was signifi- cant variation in age at onset in several of the families, particularly families 3395 (Figure 1), 3302, 2970 (Fig- ure 2), and 2973 (Figure 3). On the other hand, a pat- tern of extremely early onset breast cancer occurrences (in the mid-20s and early 30s) may characterize some HBC kindred,22 while in others there is much later onset.23,24

Families 3305 (Figure 1) and 3302 (Figure 2) include examples of mothers with late onset of cancer who have daughters with remarkably early onset. In these cases, daughters may show manifestations of cancer prior to its onset in their mothers. Therefore, before the mothers de- veloped breast cancer, they may have been considered OGCs, depending on their position in the pedigrees. Be- cause early onset is predominant in these pedigrees, sur- veillance should logically begin early. In turn, it is nec- essary to continue screening through the later years of life, because age, in addition to genetics, is a significant contributing factor to all breast cancer occurrences.1,25 Whether later onset occurrences of HBC are due to causes other than primary genetic factors will remain elusive, because even with a positive linkage to the susceptibil- ity locus at chromosome 17q4,6-9 or a p53 germ line mu- tation indicative of Li-Fraumeni syndrome,10 one can- not exclude reduced gene penetrance, environmental factors, and/or their interaction.

The etiologic significance of early onset breast can- cer in HBC has been supported by Roseman et al.26 They studied the influence of age on breast cancer in 9000 women, of whom 2712 had breast cancer, 2972 had be- nign breast disease, and 3316 were control subjects. They found that a family history of breast cancer was an im- portant determinant of its occurrence only among women younger than 60 years.

Surveillance and Treatment Strategies in Kindreds With Hereditary Breast Cancer, Hereditary Breast and Ovarian Cancer (HBOC), and Li-Fraumeni Syndromes*

IndicationAction/Recommendation
Late teens Age 20 yBegin education on the natural history of cancer and genetic risk, including counseling on rapidly evolving genetic testing technology, which may lead to sequencing and cloning of the BRCA-1 gene (and possibly other genes) Give instruction on breast self-examination; semiannual breast examination by physician
Age 25 yInitiate mammography, repeat every other year through age 35 y, and annually thereafter
Initial breast cancer in a stage I or Il lesionConsider prophylactic contralateral mastectomy
Cancer phobia, difficult-to-examine breasts, markedly increased density on mammograms, multiple biopsy specimens showing benign breast disease, or histopathologic study showing proliferative breast disease with atypiaConsider prophylactic bilateral mastectomy
Age 30 y; HBOC syndrome kindred; and first-degree relative has breast or ovarian cancerInitiate ovarian cancer surveillance to include annual pelvic examination, transvaginal ultrasonography, Doppler color blood flow imagery, and CA-125
Completed family history; HBOC syndrome kindred; and has first-degree relative with breast or ovarian cancerDiscuss prophylactic oophorectomy and potential for developing intraperitoneal cystadenocarcinoma consonant with ovarian origin despite ovaries being considered histologically normal at time of surgery

*For kindred with the Li-Fraumeni sarcoma; breast and brain tumor; lung, leukemia, lymphoma; and adrenal cortical carcinoma syndrome, we follow the above surveillance and treatment recommendations and also explain the limitations of screening for the other integral cancers.

TUMOR SPECTRUM AND HBC SYNDROME DIAGNOSIS

In addition to recognizing the cardinal features of the natu- ral history of HBC, the diagnostician needs to under- stand the significance of patterns of tumor combina- tions in specific HBC syndromes such as HBOC and Li- Fraumeni. For example, in hereditary site-specific breast cancer (Figure 1), patients are at increased risk for only breast cancer, while breast and ovarian cancer are the tu- mors of concern in the HBOC syndrome (Figure 3). A more complex array of tumors is evidenced in the Li- Fraumeni/SBLA syndrome as shown in Figure 4. Be- cause of these differing breast cancer-prone genotypes, breast cancer-prone families cannot be accurately evalu- ated based merely on evidence for breast cancer. It is im-

perative that cancers of all anatomic sites be assessed. This knowledge can then be used to plan and implement sur- veillance and treatment programs to screen for cancers of those integral at-risk organs that characterize the HBC syndrome of concern (Table).

GENE PENETRANCE

It is essential to understand that the penetrance of breast cancer genes is not 100%.27,28 In some cases, we assume women to be OGCs because of their position in the pedi- gree. However, given variable expressivity and the lack of complete gene penetrance, some will develop breast cancer, albeit at an older age, while the disease will never be manifested in others. Also, given the extremely com- mon occurrence of breast cancer in western countries, a patient who is an OGC for HBC may have developed breast cancer due to nongenetic factors. At this point in our un- derstanding of breast cancer genetics, the ability to dis- tinguish between primary genetic etiologic and environ- mental factors, and/or their interaction, remains elusive.

OBLIGATE GENE CARRIERS

Patients in an HBC pedigree who are likely gene carriers display extremely important implications for following surveillance and treatment strategies. Given reduced gene penetrance and variable expressivity, particularly with respect to age at onset, such patients-in spite of their gene carrier status-may never develop the cancer phe- notype. The problem faced by the clinician is that it is not possible to determine with certainty whether cancer will be expressed. Hence it is appropriate to err on the side of caution and assume that the patient is destined to ultimately show manifestations of breast cancer. Ex- amples of OGCs are seen in many of the pedigrees that we have described.

THE CONTRALATERAL BREAST

If a high-risk patient develops breast cancer, then atten- tion must be given to the opposite breast since bilater- ality is also a cardinal feature of HBC, as shown in our families 3207 and 3386 (Figure 1) and 2970 (Figure 2). Therefore, we recommend consideration of contralat- eral prophylactic mastectomy in patients from families with HBC with initial stage I or II breast cancer16 (Table).

TWINS

Identical twins who are considered to be at 50% risk be- cause of their position in an HBC pedigree present a unique situation. If one twin develops breast cancer, her co-twin’s risk is increased to near 100%, consonant with the gene’s penetrance. In contrast, in identical twin sis- ters of whom one develops breast cancer, but who are

not from a family with HBC, the concordance rates for breast cancer diminish significantly.29,30

MALE BREAST CANCER

An issue that may not be appreciated by physicians when assessing genetic risk for breast cancer is the fact that men as well as women in the direct genetic lineage (first- degree relatives of affected patients) have an equal (ap- proximately 50%) chance of inheriting a deleterious breast cancer-prone gene. However, although men will rarely develop breast cancer, there are noteworthy exceptions as seen in our family 3511 (Figure 2), in which a father and son had breast cancer.

Kozak et al31 and Demeter et al32 also described fami- lies in which several male as well as female relatives had breast cancer. In a case-control study, Rosenblatt et al33 found breast cancer to be more frequent among young men with first-degree relatives with breast cancer than among older men with first-degree affected relatives. The findings were similar to those observed for breast can- cer in women. Breast cancer in men with Klinefelter syn- drome34 occurs 2035 to 6636 times more often than in men without this cytogenetic disorder.

In conclusion, a thoroughly compiled family his- tory of cancer, coupled with knowledge of the clinical- genetic nuances of HBC, can aid the physician signifi- cantly in the diagnosis, treatment, and genetic counseling of patients who are at increased genetic risk of develop- ing breast cancer.37

Reprints not available.

REFERENCES 0

1. Harris JR. Medical progress: breast cancer. N Engl J Med. 1992;327:319-328.

2. Lynch HT. Genetics and Breast Cancer. New York, NY: VN Reinhold Co; 1981.

3. Lynch HT, Lynch JF. Breast cancer genetics in an oncology clinic: 328 con- secutive patients. Cancer Genet Cytogenet. 1986;23:369-372.

4. Hall JM, Ming K, Lee MK, et al. Linkage of early-onset breast cancer to chro- mosome 17q21. Science. 1990;250:1684-1689.

5. McKusick VA. Mendelian Inheritance in Man. Baltimore, Md: The Johns Hop- kins University Press; 1990.

6. Hall JM, Friedman L, Guenther C, et al. Closing in on a breast cancer gene on chromosome 17q. Am J Hum Genet. 1992;50:1235-1242.

7. Narod SA, Feunteun J, Lynch HT, et al. Familial breast-ovarian cancer locus on chromosome 17q12-q23. Lancet. 1991;2:82-83.

8. Easton DF, Bishop DT, Ford D, Crockford GP, the Breast Cancer Linkage Con- sortium. Genetic linkage analysis in familial breast and ovarian cancer: results from 214 families. Am J Hum Genet. 1993;52:678-701.

9. Feunteun J, Narod SA, Lynch HT, Watson P, Conway T, Lynch J. A breast- ovarian cancer susceptibility gene maps to chromosome 17q21. Am J Hum Genet. 1993;52:736-742.

10. Malkin D, Li FP, Strong LC, et al. Germ line p53 mutations in a familial syn-

drome of breast cancer, sarcomas, and other neoplasms. Science. 1990;250: 1233-1238.

11. Bottomley RJ, Condit PT, Chanes RE. Cytogentic studies in familial malig- nancy. Clin Res. 1967;15:334.

12. Lynch HT, Krush AJ, Lemon HM, Kaplan AR, Condit PT, Bottomley RH. Tumor variation in families with breast cancer. JAMA. 1972;222:1631-1635.

13. Lynch HT, Radford B, Lynch JF. SBLA syndrome revisited. Oncology. 1990; 47:75-79.

14. Lynch HT, Fitzgibbons RJ Jr, Lynch JF. Heterogeneity and natural history of hereditary breast cancer. Surg Clin of North Am. 1990;70:753-774.

15. Lynch HT, Conway T, Fizgibbons R Jr, et al. Age-of-onset heterogeneity in hereditary breast cancer: minimal clues for diagnosis. Breast Cancer Res Treat. 1988;12:275-285.

16. Harris RE, Lynch HT, Guirgis HA. Familial breast cancer: risk to the contralat- eral breast. J Natl Cancer Inst. 1978;60:955-960.

17. Lynch HT, Harris RE, Organ CH, Lynch JF. Management of familial breast can- cer, Il: case reports, pedigrees, genetic counseling, and team concept. Arch Surg. 1978;113:1061-1067.

18. Lynch HT, Harris RE, Guirgis HA, Maloney K, Carmody LL, Lynch JF. Familial association of breast/ovarian carcinoma. Cancer. 1978;41:1543-1549.

19. Lynch HT, Lynch JF. Hereditary ovarian carcinoma. Hemat Oncol Clin North Am. 1992;6:783-811.

20. Li FP, Fraumeni JF Jr. Soft tissue sarcomas, breast cancer, and other neo- plasms: a familial syndrome? Ann Intern Med. 1969;71:747-752.

21. Lynch HT, Mulcahy GM, Harris RE, et al. Genetic and pathologic findings in a kindred with hereditary sarcoma, breast cancer, brain tumors, leukemia, lung, laryngeal, and adrenal cortical carcinoma. Cancer. 1978;41:2055-2067.

22. Lynch HT, Watson P, Conway T, Fitzsimmons ML, Lynch J. Breast cancer fam- ily history as a risk factor for early onset breast cancer. Breast Cancer Res Treat. 1988;11:263-267.

23. Lynch HT, Watson P, Conway TA. Natural history and age of onset of heredi- tary breast cancer. Cancer. 1992;69:1404-1407.

24. Mettlin C, Croghan I, Natarajan N, et al. The association of age and familial risk in a case-control study of breast cancer. Am J Epidemiol. 1990;131:973-983.

25. Avila MH, Walker AM. Age dependence of cohort phenomena in breast cancer mortality in the United States. Am J Epidemiol. 1987;126:377-384.

26. Roseman DL, Straus AK, Shorey W. A positive family history of breast cancer: does its effect diminish with age? Arch Intern Med. 1990;150:191-194.

27. Go RCP, King MC, Bailey-Wilson J, Elston RC, Lynch HT. Genetic epidemiol- ogy of breast cancer and associated cancers in high-risk families, I: segrega- tion analysis. J Natl Cancer Inst. 1983;71:455-461.

28. King MC, Go RCP, Lynch HT, et al. Genetic epidemiology of breast cancer and associated cancers in high-risk families, Il: linkage analysis. J Natl Cancer Inst. 1983;71:463-467.

29. Heizer WD, Lewison EF. Concordant disease in identical twins. JAMA. 1964; 188:217-220.

30. Holm NV, Hauge M, Harvald B. Etiologic factors of breast cancer elucidated by a study of unselected twins. J Natl Cancer Inst. 1980;65:285-298.

31. Kozak FK, Hall JG, Baird PA. Familial breast cancer in males: a case report and review of the literature. Cancer. 1986;58:2736-2739.

32. Demeter JG, Waterman NG, Verdi GD. Familial male breast carcinoma. Can- cer. 1990;65:2342-2343.

33. Rosenblatt KA, Thomas DB, McTiernan A, et al. Breast cancer in men: aspects of familial aggregation. J Natl Cancer Inst. 1991;83:849-854.

34. Lynch HT, Kaplan AR, Lynch JF. Klinefelter syndrome and cancer: a family study. JAMA. 1974;229:809-811.

35. Everson RN, Fraumeni JF, Wilson RE, et al. Familial male breast cancer. Lan- cet. 1976;1:9-12.

36. Crichlow RW. Carcinoma of the male breast. Surg Gynecol Obstet. 1972;134: 1011-1019.

37. Lynch HT, Watson P, Conway TA, et al. DNA screening for breast/ovarian can- cer susceptibility based on linked markers. Arch Intern Med. 1993;153:1979- 1987.