ras and c-myc protein expression in colorectal carcinoma

r a s and c - m y c Protein Expression in

Colorectal Carcinoma Study of Cancer-Prone Patients Felice Miller, M.D.,* Tomas M. Heimann, M.D.,) Astrid Quish, M.D.,* Daniel J. Pyo, M.D.,} Arnold Szporn, M.D.,* Giorgio Martinelli, Ph.D.,-~ Thomas M. Fasy, M.D., Ph.D.* From the D e p a r t m e n t s o f } Surgery a n d *Pathology, M o u n t Sinai School o f Medicine, City University of N e w York, N e w York, N e w York

This study was performed to determine the correlation of tumor ras and c-myc oncogene expression with clinical and prognostic variables in patients prone to develop colorectal cancer. One hundred eighteen patients with colorectal cancer were studied; mean age was 40 years. Fifty-three were young patients (age 40 or less), 49 had ulcerative colitis, and 16 had multiple polyposis coli. Immunoperoxidase stains of paraffin-embedded cancer sections were performed for the c-myc and ras proteins. ras staining was found to correlate with Dukes stage and prognosis. Patients with tumors negative for ras protein stain had an actuarial five-year survival of 61 percent versus 44 percent for those tumors with a positive stain (P < 0.05). This correlation was not seen with the c-myc stain. Positive ras oncogene stain appears to be a useful indicator of advanced stage and poor prognosis in colorectal cancer occurring in cancer-prone patients. [Key words: Colorectal cancer; Multiple polyposis coli; Ulcerative colitis; Young patients; Oncogenes; ras oncogene; c-myc oncogene; Prognosis] Miller F, Heimann TM, Quish A, Pyo DJ, Szporn A, Martinelli G, Fasy TM. ras and c-myc protein expression in colorectal carcinoma: study of cancer-prone patients. Dis Colon Rectum 1992;35:430-435. arcinoma of the c o l o n and r e c t u m is m o s t c o m m o n l y s e e n in o l d e r patients, with the p e a k incidence occurring in the seventh d e c a d e of life. W h e n these cancers d e v e l o p in y o u n g patients, they are often associated with p r e d i s p o s i n g conditions such as multiple polyposis coli and ulcerative colitis.L, a In patients with multiple polyposis coli, the cumulative risk of d e v e l o p i n g colorectal cancer increases with age and eventually reaches nearly 100 percent. Although patients with ulcerative colitis have a lower incidence of colorectal cancer than do those with multiple polyposis coli, their cancer risk also increases in direct p r o p o r t i o n to the duration of disease. 3'4 Colorectal c a r c i n o m a occasionally occurs in y o u n g patients without a

C

p r e d i s p o s i n g condition, s'6 Although s o m e y o u n g patients have a positive family history of colorectal or other cancers, the reason for early cancer develo p m e n t in the majority is often unclear. Several studies have d e m o n s t r a t e d that ras and c - m y c o n c o g e n e e x p r e s s i o n is frequently present in colorectal a d e n o c a r c i n o m a . I m m u n o h i s t o c h e m ical stains of tissue s a m p l e s from colorectal cancer occurring in the general p o p u l a t i o n have s h o w n that ras and c - m y c protein p r o d u c t s are often f o u n d in greater concentrations than in normal colonic mucosa. S o m e studies have also r e p o r t e d an association b e t w e e n o n c o g e n e e x p r e s s i o n and prognosis. 7-~~ The e x p r e s s i o n of these o n c o g e n e s in tumors occurring in p o p u l a t i o n s p r o n e to d e v e l o p colorectal cancer, however, has not b e e n well studied. The p r e s e n t study was u n d e r t a k e n to determ i n e the prognostic significance of ras and c - m y c o n c o g e n e e x p r e s s i o n in colorectal cancer occurring in c a n c e r - p r o n e patients. MATERIALS AND

Address reprint requests to Dr. Heiman.n: Department of Surgery, Box 1259, Mount Sinai Medical Center, One Gustave L. LeW Place, New York, New York 10029. 430

METHODS

O n e h u n d r e d e i g h t e e n patients with colorectal a d e n o c a r c i n o m a u n d e r g o i n g surgical t r e a t m e n t at Mount Sinai Hospital b e t w e e n 1968 and 1985 w e r e studied retrospectively. Forty-nine patients had ulcerative colitis, 16 had multiple polyposis coli, and 53 w e r e y o u n g patients (age 40 or less) without a p r e d i s p o s i n g cause. Histologic slides w e r e rev i e w e d by a single observer. T u m o r differentiation was d e t e r m i n e d according to the classification described by Jass et al., ~ and p r e s e n c e of colloid lakes and signet ring cells was recorded. T u m o r stage was c a t e g o r i z e d according to the AstlerColler modification of the Dukes classification, x2 In patients with multiple tumors, histologic findings and the Dukes stage of the m o s t a d v a n c e d

Vol. 35, No. 5

ONCOGENES AND COLORECTAL CANCER

lesion were used. Follow-up information was obtained by direct patient contact, review of medical records, tumor registry information, or contact with primary physicians. To determine r a s protein expression, a mouse monoclonal antibody, 142-24E05, was raised against a synthetic 23-amino acid peptide corresponding to residues 96 to 118 of the human H - r a s and N - r a s sequence. Similarly, a mouse monoclonal antibody, 155-11C7, was raised against an 18-amino acid peptide corresponding to residues 171 to 188 to the human c - m y c protein (National Cancer Institute Repository, Microbiological Associates, Inc., Bethesda, MD). Immunoperoxidase stains were performed on sections from formalinfixed, paraffin-embedded tumor blocks. Five-micron sections were deparaffinized, rehydrated, placed in a solution of 1 percent H202 for 15 minutes, and then digested with saponin. After washing in phosphate-buffered saline (PBS) at pH 7.4, the specimens were incubated with 10 percent normal goat serum for 40 minutes. The sections were then incubated overnight with the primary antibody against c-myc protein (1:2,000 dilution) or r a s protein (1:8,000 dilution) at 4 ~C. After washing with PBS, biotinylated goat antimouse immunoglobulin was applied for 30 minutes. The sections were then rinsed in PBS and incubated with streptavidin peroxidase complex for 30 minutes. After rinsing with PBS, the sections were placed in a 0.5 percent solution of Triton X-100 (Rohm & Haas Co., Philadelphia, PA) for 30 seconds and then incubated with a diaminobenzidine solution containing 0.01 percent H202 and counterstained with hematoxylin (ras protein) or nuclear fast red ( c - m y c protein). Specificity controls were performed by replacing the primary antibody with PBS. The slides were reviewed by a single observer who was unaware of the patient's outcome. Tumors that showed minimal to no staining were classified as negative. Tumors that stained strongly for the protein products were considered positive. The data obtained were analyzed using the SAS statistical analysis system (SAS Institute Inc., Raleigh, NC). Numerical variables were compared using Student's t-test. Chi-squared test was used for variables grouped into categories. Survival analysis was performed using the BMDP Statistical Software (BMDP Inc., Los Angeles, CA). Actuarial life table statistics (Breslow) and the Cox proportional hazards model using stepwise regression were used TM

431

for survival analysis with single and multiple variables. The covariates studied were age, sex, diagnosis, location of the tumor, Dukes classification, tumor differentiation, colloid content, presence of signet ring cells, and staining of the c - m y c and r a s protein products.

RESULTS One hundred eighteen patients were studied; 57 (48 percent) were men and 61 (52 percent) were women. The mean age was 40 years (range, 21-76 years). The mean age of the young patients was 34 years, and the mean ages for the patients with ulcerative colitis and multiple polyposis coli were nearly identical at 45 and 44 years, respectively. Thirty-eight young patients (72 percent) and 12 with multiple polyposis coli (75 percent) had tumors located in the rectum and sigmoid. In the ulcerative colitis group, 27 (82 percent) of the 33 patients with single tumors had lesions in the rectum and sigmoid, but this number dropped to 25 percent in the 16 patients with synchronous cancers. There was no significant difference in Dukes stage among the three groups (Table 1). The incidence of poorly differentiated tumors was significantly higher in the patients with ulcerative colitis, while in the other two groups the majority of lesions were well and moderately differentiated (P < 0.01). Young patients and those with ulcerative colitis also had a significantly higher incidence of tumors with prominent colloid features when compared with patients with multiple polyposis (47 percent and 55 percent vs. 13 percent; P < 0.05). Seventeen patients (14 percent) had cancers with signet ring cells, and fourteen of these were in the ulcerative colitis group (P < 0.01). The tumors stained for c - m y c protein showed nuclear and cytoplasmic localization, but those stained for r a s protein demonstrated only cytoplasmic staining. The stain for the c-myc protein was negative in 34 tumors (29 percent), and the r a s protein stain was negative in 36 tumors (31 percent). There was no significant difference in the staining pattern for the r a s or c - m y c protein among the three patient groups (Table 2). The r a s protein stain correlated with the Dukes classification (P < 0.05). There was a progressive increase in the percentage of tumors staining positively for the r a s protein with advanced Dukes stage (Table 3). There was no correlation between

432

MILLER E T AL Table 1. Clinical Material

Number Mean age Range Sex Male Female Dukes A B

C O Histology Well Moderate Poor Colloid None Present Signet None Present

Young Patients

UC*

FPC1-

Total

53 345 21-40

49 45 22-74

16 44 21-76

118 40 21-76

22 (42) 31 (58)

30 (61) 19 (39)

5 (31) 11 (69)

57 (48) 61 (52)

0 (0)

4 (8)

3 (19)

7 (6)

17 (32)

15 (31)

3 (19)

35 (30)

18 (34) 18 (34)

17 (35) 13 (27)

6 (38) 4 (25)

41 (35) 35 (30)

27 (51) 15 (28)

21 (43) 6 (12)

5 (31) 8 (50)

53 (45) 29 (25)

11 (21)

22 (45)$

3 (19)

36 (31)

28 (53) 25 (47)

22 (45) 27 (55)

14 (88) 2 (13)$

64 (54) 54 (46)

50 (94) 3 (6)

35 (71) 14 (29)$

16 (100) 0 (0)

101 (86) 17 (14)

Values in parentheses are percentages. * UC = ulcerative colitis. t FPC = multiple polyposis coli. 1: P < 0.05.

Dis Colon Rectum, May 1992

Dukes B tumors was 88 percent, 45 percent for Dukes C, and 6 percent for Dukes D (Fig. 1). There was no significant difference in survival between well- and moderately differentiated lesions. Patients with poorly differentiated tumors had a significantly worse outcome (67 percent and 56 percent vs. 19 percent; P < 0.01; Fig. 2). Patients with tumors containing signet ring cells also had a significantly lower five-year survival (29 percent vs. 52 percent; P < 0.05; Fig. 3). The five-year survival rates for patients with colloid and noncolloid lesions were identical (49 percent). Patients with tumors staining positively for the r a s protein had a five-year survival rate of 44 percent, as compared with 61 percent for those with negative tumors (P < 0.05; Fig. 4). There was no difference in survival when c- m y c protein stain was correlated with prognosis. Multivariate analysis identified Dukes stage as the strongest prognostic variable, followed by tuTable 3.

Oncogene Stains by Tumor Stage Dukes A Dukes B Dukes C Dukes D ras stain

Negative Positive c-myc stain Negative Positive

Table 2.

Oncogene Stains by Patient Population Young Patients

UC*

FPCI

Total

20 (38) 33 (62)

14 (29) 35 (71)

2 (13) 14 (88)

36 (31) 82 (69)

19 (36) 34 (64) 53

11 (22) 38 (78) 49

4 (25) 12 (75) 16

34 (29) 84 (72) 118

3 (43) 4 (57)

16 (46) 19 (54)

12 (29) 29 (71)

5 (14) 30 (86)*

1 (14) 6 (86)

11 (31) 24 (69)

12 (29) 29 (71)

10 (29) 25 (71)

Values in parentheses are percentages. * P < 0.05.

ras stain

Negative Positive c-myc stain Negative Positive Total

CANCER PRONE PATIENTS Survivol by Dukes stQge I

Values in parentheses are percentages. * UC = ulcerative colitis. 1 FPC = multiple polyposis coli.

~ I.o l i ~

protein staining and Dukes class. Oncogene expression did not correlate with age, sex, tumor location, differentiation, colloid content, and presence of signet ring cells. The overall actuarial five-year survival rate in this series was 49 percent. There was no difference in five-year survival among the three different groups. There were no deaths in patients with Dukes A lesions. The five-year survival for patients with

I

I

f

rl_.

~" 0 . 8 - h' 'q'L L D m t, z_o

I

J

Dukes

B

I

I

'LL_.L. .Dukes . . . }. C 13

0

I

~,

~~

oarr

c-myc

I

L ..... L. 1

, L

a_ 0.4 _>

"',

0.2

%L . _ .

~

Dukes

D

~--3.. i 0

I

0

I

12

i

I

I

24

i

36

I

k ....... I i I

48

60

MONTHS

Figure 1. Comparison of actuarial five-year survival curves

by Dukes classification. The differences in five-year survival are statistically significant (P < 0.01).

ONCOGENES AND COLORECTAL CANCER

Vol. 35, No. 5

CANCER PRONE PATIENTS Survival by tumor differentiation

I,O i~._~.._ll

i

r

g O. 8

J

I

differentiated

t ~ ~ i _ . , M0dG'~Tt-- ..... 1 O. 6 "differentiated k'7

0 0n-Q_ 0.4 W > I-