Frequency of hereditary colorectal carcinoma

GASTROENTEROLOGY

Frequency Carcinoma JUKKA-PEKKA Second

Department

of Hereditary

1987;9’~:‘1021-5

Colorectal

MECKLIN of Surgery,

Helsinki

University

The frequency of hereditary colorectal carcinoma was evaluated in a study group consisting of all colorectal carcinoma patients (n = 468) diagnosed in one Finnish county (0.25 million inhabitants) during the period 1970-1979. The cancer family syndrome type of hereditary nonpoIyposis colorectal carcinoma emerged as the most common verifiable risk factor for colon cancer involving 3.8% 5.5% of all colorectal carcinoma patients in this study. The frequencies uffamilial adenomatosis and ulcerative colitis were 0.2% and 0.6’S, respectively. As the diagnostic method of this study was based on the family history of the patients, only those families with inherited cancer cases in two or more generations could be identified. The cancer family syndrome cannot be diagnosed on the basis of a single patient, nnd so the observed frequency of 4%6% for the syndrome may still represent an underestimate. The patients with cancer family syndrome were young, accounting for 290/0-390/0 of the patients under 50 yr of age, and their tumors were located predominuntly (65%) in the right hemicolon. The high frequency of hereditary cases among colorectal carcinoma patients indicates the importance of .&dying the family history of every new patient. It has been estimated that 0.5~&50% or even more of the patients with colorectal carcinoma inherit their disease as an autosomal dominant trait (1,Z). These inherited syndromes have been divided into two categories: (a) hereditary colorectal carcinoma with polyposis and (b) hereditary nonpolyposis colotectal carcinoma (3). The most common type of hereditary Received August 13, 1986. Accepted May 19, 1987. Address requests for reprints to: J-P. Mecklin, M.D., II Department of Surgery, University Central Hospital, SF-00290 Helsinki, Finland. This work was supported by the Finnish Cancer Foundation and the Finnish Academy. The author thanks Dr. H.J. Jarvinen for his cooperation and critical review of the manuscript and both Mrs. K. PitkBnen and M. Kirsi for excellent technical assistance. (~11987 by the American Gastroenterological Association 001%5085i8763.50

Central

Hospital,

Helsinki,

Finland

colorectal carcinoma with polyposis is familial adenomatous polyposis coli, representing
and Methods

The study population consisted of all coloreLta1 carcjnoma patients (n = 468) who were living ill the county of Middle-Finland (0.25 million inhabitants) and whose cancer was diagnosed between 1970 and 1979. Only cases of colorectal adenocarcinoma were inc,luded. The patients’ biographical data and the addresses of the hospitals where they were treated were obtained from the Finnish Cancer Registry. The hospital records were reviewed in an attempt to discover aspects with possible etiologic significance with regard to carcinoma, for example, ulcerative colitis, familial adenomatosis, or previous malignancies. Additionally, for all patients under 70 yr of age at the time of diagnosis of the cancer (n = 241: 52’s,), biographical data and the causes of death of their parents were obtained from parish registers and the Register of Cause of Death of the Finnish Central Statistical Office. If one of the parents had died of cancer, the family was studied further until a generation with no incidence of cancer was reached. This information was not obtained for the patients over 70 yr of age (11 = 227; 4R0/r,) for two reasons: (a) hereditary colorectal carcinoma syndromes Ahbreviotion

used in this pnpcr:

CFS, cancer

family

syndrome.

1022

Table

MECKLIN

1.

GASTROENTEROLOGY

Frequency of Inherited Colorectal Carcinoma and of Colitic Carcinoma in Different Age Groups

of Patients

With Colorectal

Age (Yd

n”

20-29 30-39 40-49 50-59 60-69

7 13 31 77 117

70-79 '80

152 75

Total Percent

472 100.0

CFS 2 5 8 3

18 3.8

FPC

Carcinoma

Probable

Total

CFS

(percent)

UC

1

1 1 1

1 1 3 3 0

1 0.2

3 0.6

8 1.7

3 6 12 8 1

(42.9) (46.1) (38.7) (10.4) (0.91

30 6.4

CFS, cancer family syndrome; FPC, familial adenomatous polyposis coli; UC, ulcerative colitis. ‘n = 472 (468 patients with colorectal carcinoma plus 4 patients with metachronous colon cancer diagnosed between 1970 and 1979 and registered earlier in the Finnish Cancer Registry).

arise very seldom in patients over 70 yr of age (3,~) and (b) the death certificates of the parents of these patients, which dated back to the beginning of this century, were considered to be uninformative in many instances. A patient was considered to have inherited nonpolyposis colorectal carcinoma if he belonged to a family with at least three first-degree relatives with colorectal carcinoma. The original hospital records of the family members affected with cancer were reviewed whenever possible (72%) of the cases), but in the remaining 28% only the cause of

death (from parish registers or the Finnish Central Statistical Office) was available. In the present study no attempt

Figure

I. Pedigrees

of CFS kindreds 15-19. (1. female: intraabdominal cancer: 0, other malignancy:?, healthy children;m, cause of death unknown.

Vol. 93, No. 5

was made to subdivide inherited nonpolyposis carcinoma [or cancer family syndrome (CFS)] firm basis for subdivision is still lacking (8).

colorectal because a

Results The cause of death of the parents of 241 patients under 70 yr of age could be verified in 84% (405 of 482), revealing 54 died of cancer (14%). Further analysis of these families indicative of inherited cancer disclosed that 14 patients (3.00/u) belonged to 13 unrelated CFS families with 23 cases of colorectal carcinoma in first-degree relatives. One patient (0.2%) had familial adenomatous polyposis, and in 3 cases carcinoma (0.6%) complicated ulcerative colitis [Table 1). Eight of the 13 families had already been identified earlier and their pedigrees have been published previously (CFS kindreds No. 1, 2, 5, 7: 9, 10, 12, 14) (7), whereas five CFS families were newly diagnosed in this study (Figure 1). The distribution of all malignancies in these 13 kindreds is presented in Table 2. Pedigree analysis of the CFS families revealed 4 additional patients with colorectal carcinoma (0.8%) whose second carcinoma was diagnosed in the study period. These 4 cases had not been registered in the Finnish Cancer Registry as new cases because the primary colorectal carcinoma in each use had been diagnosed and registered before the study period. Inclusion of these 4 verified CFS cases increased the frequency of CFS to 3.8% of all colorectal carcinomas. In addition to the 18 verified CFS cases, there were

0, male; 0, colorectal Lancer; 0, endometrial cancer; colorectal cancer with other primary malignant neoplasms;

@, undefined 0. number of

November

1987

HEREDITARY

8 patients with colorectal carcinoma (1.7%)who had two or more first-degree relatives with other intraabdominal malignancy (e.g., uterus, stomach, pancreas) (Table 3). Identification of CFS in these families was not possible with certainty, but inclusion of them would raise the frequency of CFS to 5.5% of all colorectal carcinomas. Some essential clinical data on these patients (n = 26) with verified or probable CFS are presented in Table 4. In comparison with the remaining colorectal carcinoma patients (n = 442), excluding patients with familial adenomatous polyposis coli and ulcerative colitis, patients with CFS comprised 29’%-39% of all the patients under 50 yr of age, depending on whether only verified CFS cases or both verified and probable CFS cases were included.

Table

3.

COLORECTAI,

CARCINOMA

Distribution of Malignancies in Probable Cancer Family Syndrome Kindreds Number of malignancies --Other intraOther abdominal Colorectal

Kindred

carcinoma

(No.1

1023

cancer

carcinoma

I

1

II III IV V VI VII VIII

1 2 1 1 1 1 1

2 2 2 2 2 2 2 2

Total

9

16”

Number of faniI4 members”

-~___

1 1

8 6 17 7 8 6 16 12

6’

80

1 I 1

I

” All family members in affected generations. ” Stomach 11, pan1, bladder 1, creas or biliary tract 5. ’ Lung 2, breast 1, esophagus leukemia

1.

Discussion The present study confirms that CFS [or hereditary nonpolyposis colorectal carcinoma) is the most common identifiable risk factor for large bowel cancer. The estimation that about 4%-6% of all colorectal carcinoma patients seem to inherit their disease corresponds well with the previous work of Lynch and his colleagues (1,2), and exceeds nearly lo-fold the preliminary calculations based on limited data for the whole of Finland (7). Analogy with previous estimations, together with the increasing number of reports about new CFS materials all over the world (g--14), supports the view that the present observation is not limited only to the Finnish county concerned.

Table

i’.

Distribution Syndrome

of Malignancies Kindreds

in Cancer

Family

It could be that even the present estimation is still low. First, the most obvious deficiency of the method of this study is its inability to identify CFS or1 the basis of a single patient. In the present study it was possible to identify only those CFS families with inherited cancer cases in two or more generations. Solitary cases, however, must occur as they do in familial adenomatous polyposis coli in about 30%45% of the families (15,161. Further, these solitary patients could have a propensity to transfer the trait to their children and have a predisposition to metachronous colorectal tumors. If we believe that such undiagnosed CFS cases occur among the young patients with no hereditary background, it would be advisable to arrange appropriate life-long follow-up examinations for all young patients who have had a

Number of malignancies intraabdominal Other carcinoma cancer

Other

Kindred (No.)

Colorectal carcinoma

1

19

2

17 5 4 5 5 5 4 4 3 3 3 9

5 7 9 10 12 14 15 16 17 18 19

Total

86

1

10 11

3 -

1 2 2 5 50b

14’

Total 30 31 5 5 8 8 7 11 5 5 5 4 26 150

Number of family members” 143 140 19 18 15 18 26 23 13 15 13 34 107 584

a All family members in affected generations. ’ Undefined 24, uterus 15, stomach 6, pancreas or biliary tract 5. ’ Breast 6, lung 4. skin 2, brain 1. lymphoma 1.

Table

4.

Comparison of Clinical Data Between the Patients With and Without a Hereditary Background Patients with verified or probable CFS In = 261

Mean age (yr) Tumor site in right hemicolon Metachronous colorectal cancer Other malignancies; before or after

44.4 17 (65.4) 6 (23.1) 1

(3.81”

patients without evidence of CFS (n = 4421 63.7 98 (22.0) 2 (0.5) :31 (6.9)”

CFS. cancer family syndrome. Numbers in parentheses are per” Other malignancies: skin 6. cents. ‘Carcinoma of the uterus. prostate 4, uterus 3. breast 3, sarcoma :% (chondro-, lipo-, and myosarcoma), lung 2, stomach 2, kidney 2. bladder I, esophagus 1, pancreas 1, penis 1, myeloma 1, pldsmacytoma 1.

1024

GASTROENTEROLOGY

MECKLIN

curative operation for colorectal carcinoma. Second, Finland is considered as a low-risk country with regard to colorectal carcinoma. Its incidence in Finland is about half that in other Scandinavian countries and the United States (17). It could be supposed that the frequency of cases of CFS would also be higher in high-risk countries. There is no evidence, however, that the proportion of hereditary cases is dependent on the overall incidence of cancer. For example, in carcinoma of the stomach there is obviously no association between these two factors. The frequency of the diffuse type of gastric cancer, which has a strong familial tendency, has remained stationary, whereas the overall incidence of carcinoma of the stomach has decreased remarkably (18). The identification of patients with a positive family history of colorectal carcinoma was as complete as possible in the present study. Reliance was not placed on the medical records alone. The official causes of death of parents were sought for all patients under 70 yr of age, and this was successful in 84%. Even this method may leave some CFS cases unrevealed: e.g., if a parent with the CFS trait died of some other cause than malignancy and the possible cancer remained undiagnosed or if the fate of a parent was unknown due to illegitimacy. Thus, the ultimate frequency of CFS in different populations can be reliably determined only after discovering methods to identify single CFS patients by genetic or biochemical markers. Such new methods are intensively sought as they would also clarify the possible heterogeneity of hereditary nonpolyposis colorectal carcinoma syndromes (19). It is remarkable that among the more than 30 Finnish CFS kindreds presently identified, we have not yet been able to find a large family where colorectal carcinoma would be the only type of cancer, corresponding to the subtype, hereditary site-specific colorectal carcinoma of Lynch syndrome I (20). The patients with a hereditary background were young (all under 60 yr of age) in this study. This same observation has been made in earlier studies (3,6,7) and the probability of the occurrence of hereditary cases in the group over 70 yr of age, not screened in this study, would be low. Young age together with a proximally located tumor helps to direct attention to the patients with the highest risk of inherited cancer. Other features typical of CFS are multiple primary malignancies (21) and colorectal carcinomas with a mutinous histology (22). These features are not specific for hereditary colorectal carcinoma, however, and an accurate family history should be taken in every new case of large bowel cancer. The clustering of malignancies in certain families is a well-documented observation. It has been dem-

Vol. 93, No. 5

onstrated that 4% of the patients in an oncology clinic had features compatible with hereditary cancer syndromes (23). Carcinoma of the breast has been considered to have the greatest predisposition to heredity (24). Lynch et al. verified the hereditary cancer syndrome in 5% of all patients with carcinoma of the breast (25). The result of the present study indicates that such a hereditary background is at least as important etiologically in colorectal carcinoma (4%-6%). If only colon cancer is included, the proportion of inherited cases would be doubled on the basis of the present study. Cancer family syndrome seems to be the most common verifiable risk factor for colorectal carcinoma, and its frequency indicates that the family history should be detailed in all carcinoma patients. Cancer family syndrome cases are found with the greatest frequency among young patients with proximally located colon cancer. The identification of CFS could lead to early detection of carcinoma in asymptomatic family members by appropriate screening procedures.

References Lynch PM, Lynch HT. Preface. In: Lynch PM, Lynch HT, eds. Colon cancer genetics. New York: Van Nostrand Reinhold, 1985:1x-XI. Mulvihill JJ. The frequency of hereditary large bowel cancer. In: lngall JR, Mastromarino AJ, eds. Prevention of hereditary large bowel cancer. New York: Alan R Liss. 1983:61-75. Lynch HT, Lynch PM. What is hereditary colon cancer. In: Ingall JR, Mastromarino AJ, eds. Prevention of hereditary large bowel cancer. New York: Alan R Liss, 1983:3-38. McConell RB. Genetics of precancerous conditions of the gastrointestinal tract. In: Sherlock P, Morson BC, Barbara L, Veronesi U, eds. Precancerous lesions of the gastrointestinal tract. New York: Raven, 1983:23-42. 5. Macklin MT. Inheritance of cancer of the stomach and large intestine in man. JNCI 1960;24:551-71. 6. Lovett E. Family studies in carcinoma of the colon and rectum. Br J Surg 1976:63:13-R. 7. Mecklin J-P, Jarvinen HJ, Peltokallio P. Cancer family syndrome: genetic analysis of 22 Finnish kindreds. Gastroenter010gy i986;90:328-33. a. Mecklin J-P, Jarvinen HJ, Peltokallio P. Identification of cancer family syndrome (abstr). Castroenterology 1986;90:1099. 9. Morson BC. Markers for increased risk of colorectal cancer. In: Sherlock P, Morson PC, Barbara L, Veronesi U, eds. Precancerous lesions of the gastrointestinal tract. New York: Raven, 1983:255-g. 10. Svendsen LB, Biilow S. Cancerfamiliesyndromet. Ugeskr Laeger 1981;143:2713-6. 11. Maack P, Rtidiger HW. Familial cancer or cancer family syndrome. Report on a cancer family and consideration of genetic mechanisms. Clin Genet 1983;24:26-40. 12. Nagengast FM, Speth PAJ, van Tongeren JHM. Familiair coloncarcinoom. Ned Tijdschr Geneeskd 1984;128:796-800. 13. Sheehan MP, Metzmaker CO. Cancer family syndrome manifested in an extended kindred. Surg Gynecol Obstet 1984; 158:450-6. 14. Hasselgren P-O, Hulten L, Karlstrom L. Beskrivning av en

November

15.

16.

17.

18.

19.

1987

slakt med hereditar anhopning av maligna tumorer i colon och andra organ. Lakartidningen 1984;81:2451-3. Jarvinen HJ, Husa A, Aukee S, Laitinen S, Matikainen M, Havia T. Finnish registry for familial adenomatosis coli. Scand 1 Gastroenterol 1984:19:941-6. Bussey HJR. Familial polyposis coli. Familial studies, histopathology, differential diagnosis and results of treatment. Baltimore: Johns Hopkins University, 1975. International Agency for Research on Cancer. Cancer incidence in five continents. IARC scientific publication no. 42. Lyon: World Health Organization, 1982. Hill MJ. Environmental and genetic factors in gastrointestinal cancer. In: Sherlock P, Morson BC, Barbara L, Veronesi U, eds. Precancerous lesions of gastrointestinal tract. New York: Raven, 1983:1-22. Lynch HT, Kimberling W, Albano WA, et al. Hereditary

HEREDITARY

20. 21.

22.

23. 24. 25.

COLORECTAL

CARCINOMA

1025

nonpolyposis colorectal cancer (Lynch syndrome I and II). Cancer 1985;56:934-8. Lynch HT. Frequency of hereditary colorectal carcinoma (Lynch syndromes I and II). Gastroenterology 1986;90:486-9. Mecklin J-P, Jarvinen HJ. Clinical features of colorectal carcinoma in cancer family syndrome. Dis Colon Rectum 1986; 29:160-l. Mecklin J-P, Sipponen P, Jarvinen HJ. Histopathology of colorectal carcinomas and adenomas in cancer family syndrome. Dis Colon Rectum 1986;29:849-53. Albano WA, Lynch HT. Recabaren JA, et al. Familial cancer in an oncology clinic. Cancer 1981;47:2113-8. Mulvihill JJ. Clinical ecogenetics: cancer in families. N Engl J Med 1985;312:1569-70. Lynch HT, Albano WA, Danes BS, et al. Genetic predisposition to breast cancer. Cancer 1984:56:612-22.