575

Primary Colorectal Lymphoma in Taiwan Wei-Shou Hwang, M.D.,* John C. T. Yao, M.D.,* Syh-Shen Cheng, M.D.,t and Hui-Hwa Tseng, M.D.t

Background. Sixteen patients with primary lymphoma of the colon and rectum were studied. Methods. The median age of these patients was 34 years, and 13 were men. These patients often experienced abdominal pain, diarrhea, a palpable abdominal mass, weight loss, bloody stools, and tumor of the cecum. Intermediate or high-grade lymphomas occurred in 14 patients, and 5 patients had T-cell lesions. The diagnoses were established by using laparotomy in 14 patients and colonoscopic biopsy in 2 patients. Fourteen patients had surgical resections followed by chemotherapy: cyclophosphamide, doxorubicin, vincristine, and prednisolone in 10; cyclophosphamide, vincristine, and prednisolone (COP) in 2; and cyclophosphamide, vincristine, methotrexate, and prednisolone in 1 patient. Two patients underwent biopsy alone followed by chemotherapy with COP in one and chemotherapy with prednisolone in the other. Results. The median follow-up time was 38 months (range, 2-82 months). Eight patients are alive with no evidence of disease (range, 10-82+ months). Six patients died of disease from 2 to 44 months after diagnosis. One patient who had no evidence of lymphoma died of esophageal carcinoma at 61 months. The median survival time was 59 months. Conclusions. The authors' experience with colorectal lymphoma in Taiwan is different from that reported from Japan and other countries. The patients of this study were significantly younger and many had T-cell lesions. Despite the frequently poor histologic types, surgical resection and adjuvant chemotherapy can result in long-term, disease-free survival in many patients with primary colorectal lymphoma. Cancer 1992; 70:575-580. From the Division of *Hematology-Oncology, Departments of Internal Medicine and tPathology, Tri-Service General Hospital, National Defense Medical Center, Taipei, Taiwan, Republic of China. Supported in part by National Science Council R.O.C. Grant NSC 80-0412-8-016-23. The authors thank Dr. Chin-Yang Li for the hematopathologic review; Dr. Peter S. P. Ho for the radiographs; Dr. Tsung-Yao Huang for the radiologic interpretations; and Drs. Lung T. Yam, Hau C. Kwaan, Wing-Kai Chan, and Munsey S. Wheby for the helpful discussions and criticisms of the manuscript. Address for reprints: Wei-Shou Hwang, M.D., Division of Hematology-Oncology, Tri-Service General Hospital, 622 Ting-Chow Road, Taipei, Taiwan 10713, Republic of China. Accepted for publication November 1, 1991.

Key words: colorectal lymphoma, Taiwan, immunohistochemistry, chemotherapy.

When malignant lymphoma involves the gastrointestinal tract,'-4 the stomach and small intestine are the most commonly affected sites; involvement of the colon and rectum is rareG5-13 Most reports of gastrointestinal lymphomas indicate a predominance of B-cell disease,14 and few studies have dealt with gastrointestinal lymphomas of the colon and rectum.15 Because the frequency of B-cell lymphoma in the Western countries is higher than that in the Far East,16,17the clinical features and outcomes of the gastrointestinal tract lymphomas likewise may be different between these areas. We evaluated 16 cases of primary colorectal lymphoma and compared the clinical and histologic features of these cases with those reported in the literature from Taiwan, Japan, and other parts of the ~ o r l d . ~ - ' ~ , ' ~ Patients and Methods

We reviewed the data from the cancer registry in the Tri-Service General Hospital, Taipei, Taiwan, for the period January 1976 to October 1990. Of the 475 cases of non-Hodgkin lymphoma (NHL) listed, there were 5 1 cases of primary gastrointestinal lymphoma: 22 gastric, 13 small intestinal, and 16 colorectal lymphomas. The clinical and laboratory data of these 16 patients were reviewed. Records and films of radiologic studies, including chest roentgenograph, barium enema, abdominal sonogram, and/or computerized tomography scan of the abdomen, were collected and reviewed by one consultant radiologist. Paraffin-embedded tissues stained with hematoxylin and eosin were reviewed by two pathologists and one hematopathology consultant. All lymphomas were classified according to the International Working Formulation. Immunocytochemical studies were performed with a peroxidase-antiperoxidase technique using monoclonal antibodies specific for kappa light chain, lambda light chain, L-26, UCHL-1, and lysozyme." Fourteen patients underwent surgical resection and 2 underwent biopsy. Ten patients received postoperative chemotherapy consisting of cyclophosphamide

576

CANCER August 1,1992, Volume 70, No. 3 Table 1. Distribution of Patients With Colorectal Lymphoma According to Age, Sex, and Anatomic Location Local

Current series No. of patients Age (Y4 Range Mean Sex M:F Ratio Location Cecum Rectum

16

Regional--Japan'** 22

Western

135

109

3.5-76 34.1

10-47 37.6

3-83 51.9

3-82 46.0

13:3 4

13:9 1.4

99:36 2.7

68:41 1.7

9 (60%) 3 (20%)

12 (55%) 1(4%)

97 (72%) 22 (16%)

76 (70%) 5 (5%)

(750 mg/m2), doxorubicin (50 mg/m2), and vincristine (1.4 mg/m2) given intravenously on day 1, and prednisolone (60 mg/m2) given orally on days 1-5; this regimen was repeated every 21 days. Two patients received chemotherapy consisting of cyclophosphamide (650 mg/m2) and vincristine (1.4 mg/m2) administered intravenously on days 1 and 8, and prednisolone (40 mg/ m2)administered orally on days 1-14; this regimen was repeated every 28 days. One patient received cyclophosphamide, vincristine, methotrexate, and prednisolone. Of the two patients whose cases were diagnosed by biopsy alone, one received cyclophosphamide, vincristine, and prednisolone (COP) and the other received prednisolone (60 mg/d). The patient who received prednisolone was lost to follow-up after 3 weeks and was excluded from the survival analysis. The follow-up period ranged from 2 to 82 months (median, 38 months). The significance of differences in survival was determined using the Mantel-Cox test for unpaired data.

Radiologic and Colonoscopic Studies The following radiologic and colonoscopic studies were performed: barium enema, 11patients; abdominal computerized tomography scan, 7 patients; abdominal sonogram, 6 patients; lymphangiogram, 1 patient; and colonoscopic examination, 3 patients. The characteristic findings of barium enema included narrowing of the lumen, a filling defect, a mass lesion, ulceration, or irregularity of mucosa (Fig. 1). Two pediatric patients had intussusception. Abdominal sonograms of six patients and computerized tomography scans of seven patients showed soft tissue masses with or without lymph node enlargement. The lymphangiogram of Patient 15 showed enlargement of the paraaortic lymph nodes. In three cases, the results of a colonoscopic examination showed a tumor mass with or without ulceration. Neither the imaging nor the endoscopic visualization allowed definitive diagnosis of colorectal lymphoma.

Histologic and Immunologic Studies Results

Clinical Features The distribution according to age, sex, anatomic location, symptoms, and signs of our patients and others reported in the l i t e r a t ~ r e ~ , ~ -is~ ,shown ' ~ , ' ~ in Tables 1 and 2. The laboratory findings of our patient are shown in Table 3. The results of hematologic studies were frequently normal, except for elevated alkaline phosphatase levels in eight and elevated lactic dehydrogenase levels in four patients. All four patients with elevated lactic dehydrogenase levels died of disease within 2 to 44 months. These clinical findings are not specific for colorectal lymphoma. Definitive diagnoses were made using laparotomy in 1 4 and colonoscopic biopsy in 2 cases.

The tumor sites, histologic types, and immunophenotypic characteristics are shown in Table 4. Eleven patients had intermediate grade NHL and 3 patients had high-grade NHL, including 1 Burkitt, 1 non-Burkitt, and 1 immunoblastic NHL; only 2 patients had lowgrade NHL. Immunohistochemistry studies were performed in 14 patients, the results of which demonstrated a B-cell origin of the tumor in 8 and a T-cell origin in 5 patients. The T-cell lesions had a variety of histologic types, which sometimes made it difficult to classify them according to the International Working Formulation. In one case (Case 13), the histologic picture showed a diffuse large cell pattern (Fig. 2) and immunologic staining showed a positive reaction for Tcell markers and a negative reaction for B-cell markers. In another case (Case 9), the neoplastic cells stained

Colorectal Lymphoma/Hwang et al.

577

Table 2. Distribution of Patients With Colorectal Lymphoma According to Symptoms and Signs Other Current series in series No. of patients Abdominal pain (%) Diarrhea (%) Body weight loss (%) Bleeding (%) Vomiting (%) Weakness (%) Anorexia (%) Change of bowel habits (%) Abdominal mass f%I

16 75 50 37.5 25 19 12.5 6.3

22 86.4 22.7 59.1 18.2 9.1 9.1 -

6.3 43.8

18.2 63.6

Literatu~e"'~ 114 65 31 51 38 46 29 35 52 20

negatively for B-cells (L26, kappa, and lambda), but there was mild infiltration of small T-lymphocytes (MT1+ and UCHL-1+). This case was cIassified as null cell type. For two patients (Patients 15 and 16),the original paraffin blocks were not available for reexamination and the final diagnosis reported here was established by reclassifying the original histologic diagnosis according to the International Working Formulation.

Clinical Stage, Treatment, and Survival The results of treatment as related to age, tumor size, stage, and cell type are shown in Table 5. Four patients had Stage IE disease, 11 had Stage IIE, and 1 had Stage IIIB. The median survival time was 59 months and the 5 year-survival rate was 53%. Eight patients are alive with no evidence of disease at 10 to 82+ months. Six patients died of disease at 2 to 44 months. One patient

Figure 1. A barium enema showed a "thumbprinting" submucosa mass (arrows) in one patient with malignant lymphoma of the cecum.

died of esophageal carcinoma at 61 months. There were no significant differences in survival related to age, sex, histologic grade, tumor size, or stage. There appeared to be a survival advantage in patients without "B" symptoms, although this did not reach statistical significance ( P > 0.05) (Fig. 3) There was a significant difference in the survival rate between B-cell and T-cell lymphomas ( P < 0.05) (Fig. 4). Patients with elevated lactic dehydrogenase levels had a poorer prognosis. Discussion

Primary lymphoma of the colon and rectum is rare, accounting for 0.65-2.1% of the malignant diseases of the

Table 3. Results of Hematologic and Biochemical Studies

Hemogram Hemoglobin (13-16 g/dl)* Leukocytes (5000-10,000/pl) Platelets (200,000-400,000/pl) Bone marrow aspiration/biopsy Albumin (3.5-5.0 g/dl) Serum glutamic oxaloacetic transaminase (7-40 U/1) Serum glutamic pyruvic transaminase (7-40 U/1) Lactic dehydrogenase (49-140 U/1) Alkaline phosphatase (20-90 U/1) Carcinoembryonic antigen (< 3.5 ng/ml)

* Normal range.

No. of patients tested

Increased

Normal

Decreased

16 16 15 15 15

0 1 0 0 0

7 12 15 15 11

9 3 0 0 4

14

1

13

0

14 8 15 5

1 4 8 0

13 4 7 5

0 0 0 0

CANCER August 2, 2992, Volume 70, No. 3

578

Table 4. Tumor Sites, Histologic Types, and Immunophenotypic Characteristics Patient no. Tumor sites

Immunophenotype Histologic type

L-26

Cecum, ileum Decending colon, cecum, ileum, mesentery LN Transverse colon Cecum, mesentery LN Cecum, ileum, regional LN Ascending colon, ileum, regional LN Cecum, ileum, appendix, paraaortic LN Ascending, transverse colon; paraaortic LN Cecum, mesentery LN Sigmoid colon Transverse defending sigmoid colon; rectum; right supraclavicle LN

Diffuse small cleaved cell Small lymphocytic plasmacytoid

12 13 14

Cecum, mesentery LN Ascending colon, cecum Rectum

15

Ascending colon, cecum, appendix, mesentery LN Rectum

Diffuse large cell Diffuse large cell Diffuse, mixed small and large cell Diffuse small lymphocytic Diffuse large cell

1 2 3 4 5 6 7

8 9 10 11

16

Diffuse large cleaved cell Small noncleaved, non-Burkitt Diffuse large cleaved cell Diffuse large cleaved cell Small noncleaved Burkitt Diffuse large noncleaved cell

K

X

UCHL-1 MT1 Lysozyme Others

+ t

+ + t t

+ + -

Diffuse small cleaved cell Diffuse mixed small and large cell Diffuse large immunoblastic cell ND ND ND

ND

-

+ + +

+ +

-

ND ND ND

ND

ND

ND

ND ND ND

ND

ND

ND

-

-

-

-

-

-

t

Leu-l(+) Leu-Za(+) Leu-3Al(+) Leu-4(+) Tac(+)

fW-1

LN: lymph node; ND: not done; K: kappa light chain; A: lambda light chain.

colon and approximately 15% of the lymphomas of the gastrointestinal tracL2j7All age groups can be affected, but the average age at onset is 50 years. Men are affected more often than women. Common presenting

clinical features include abdominal pain, weight loss, and a palpable abdominal mass. Primary lesions are most often located in the cecum (Table 1).In most patients, the diagnoses are established using laparotomy.

Figure 2. The histologic picture of cecal tumor showed diffuse large cell lymphoma (Case 13, see text) (H & E, X400).

Colorectal Lymphoma/Hwang et al.

579

Table 5. Results of Treatment as Related to Age, Tumor Size, Stage, Cell Type, Treatment, Response, and Survival Patient no.

Age (yr)

1 2 3 4 5 6 7 8 9 10 11 12

21 57 24 3.5 68 76 6 6 16 67 24 21 49 22 20 65

13 14 15 16

Tumor size (cm) 10 X 8X 8X 5X 10 X 10 X 10 X 10 X 9X 2X

8 8 8 4 8 7 7 3 8

Cell type

Stage

B B B B B B B B

IE(A) IIE(A) IE(B) IIE(A) IlE(A) IIE(A) IIE(A) IIE(B) IIE(B) IIE(B) IIIE(B) IIE(A) IE(B) IIE(B) IIE(B) 1E

1

2 5 10 X 3

N T T T T T

* *

Treatment (cycle)

S S

+ CHOP(6) + CHOP(6)

S + COP(6) S + CHOP(6) S CHOP(6) S CHOP(6) S CHOP(6) S COMP(6) S CHOP(6) S CHOP(6) S CHOP(8) S + CHOP(6)

+ + + + + + +

Response

Status

CR CR CR CR CR CR CR CR CR CR PR CR

NED Died NED NED NED NED DOD NED NED DOD DOD NED

NE PR

DOD DOD -

S

+ COP(1) s + COP(20) Bx + prednisolone

Bx

-

Survival (mo) 82+ 61 69+ 63+ 41+ 37+ 6 13f 56+ 44 15 10+ 3 2 38

-

Comment Died of esophageal carcinoma

Relapse at 5 mo

Relapse at 40 mo

Died of surgical complication Lost to follow-up

S: surgical removal; C R complete remission; NED: no evidence of disease; I'R partial response; DOD: died of disease; N: null cell; Bx: biopsy; NE: not assessable; 8: 8-cell lymphoma; T T-cell lymphoma; CHOP: cyclophosphamide, doxorubicin, vincristine, prednisolone; COP: cyclophosphamide, vincristine, prednisolone; COMP: cyclophosphamide, vincristine, methotrexate, prednisolone. *: tissue was not available for study.

Treatment usually includes resection of the primary leUnited Kingdom, most of the patients were reported to sion followed by chemotherapy or radiation therapy. have polymorphic B-cell 1ymph0rna.l~However, there The 5-year survival rate varies from 20% to 55Yx6r7 is little information concerning the phenotypic characteristics of the tumor cells in the primary lesions. All 45 The results of histologic studies showed that most of the colorectal lymphomas were diffuse NHL. Hodgcases of colorectal lymphoma in the study by Shepherd kin disease and follicular NHL are uncommon. Most of et al." were of B-cell phenotype. The clinical findings of the published studies have used the Rappaport system our cases are similar to those reported by others from Although these cases share many of the for classification. In the Japanese and Indian experi- Tai~an.'~*'~,'' ences, colonic lymphomas are more often histio~ytic.~,~clinical features of colorectal lymphomas reported elsehere,^-^,^^,^^,^^ they are different from the others in Of the literature from the United States, one study rethat the patients were younger and there was a higher ported a slight predominance (40%) of mixed cellular histologic type16whereas another reported lymphocytic percentage of T-cell disease. Surgical excision is the main treatment used to lymphomas in 80% of cases.' In one report from the

S

I

!

c 1 c

po.2-

-

Lu

0.01

0

1

I

1

2

3

1

8

4

5

time ( Y e a r s )

Figure 3. Survival of 15 patients with or without B-symptoms (Mantel-Cox test, P > 0.05).

time (Years)

Figure 4.A comparison of the survival of patients with 8-cell and T-cell lymphomas (Mantel-Cox test, P < 0.05).

580

CANCER August 1, 1992, Volume 70, No. 3

manage colorectallymphoma. Although adjuvant radiation therapy has been frequently used after surgical resection, recent studies have suggested that adjuvant chemotherapy improves the relapse-free survival in gastrointestinal NHL.’9,20Thirteen patients in our series received adjuvant combination chemotherapy and none received radiation therapy. Eight patients are alive with no evidence of disease; six of them (Patients 1 , 3 , 4 , 5 , 6 , and 9) have survived more than 3 years. Of these longterm survivors, all, except one who had high-grade NHL, had intermediate grade NHL. There were no significant differences in survival related to age, tumor size, grade, or stage. Patients with elevated lactic dehydrogenase levels and B-symptoms had a poorer prognosis. Patients with T-cell colorectal lymphoma had a poorer prognosis than those with B-cell type (Fig. 4). More aggressive chemotherapy may be needed for Tcell colorectal lymphoma. It was reported that the prognosis is worse in the younger age group.’ Our patients were relatively young, but had a better survival rate than that reported by Nirmala et al.9 We recommend that patients with colorectal lymphoma be treated with surgical resection and adjuvant chemotherapy. Some patients may have prolonged survival or may be cured with this approach. References Isaascon PG, Wright DH, Judd MA, Mepham BL. Primary gastrointestinal lymphomas: a classification of 66 cases. Cancer 1979; 43:1805-19. Contreary K, Nance FC, Becker WF. Primary lymphoma of the gastrointestinal tract. Ann Surg 1980; 191:593-8. Dragosics B, Bauer P, Radaszkiewicz T. Primary gastrointestinal non-Hodgkin’s lymphoma: a retrospective clinicopathologic study of 150 cases. Cancer 1985; 55:1060-73. Haber DA, Mayer RJ. Primary gastrointestinal lymphoma. Seniin Oncol 1988; 15:154-69.

5. Glick DD, Soule HE. Primary malignant lymphoma of the colon and appendix. Arch Surg 1966; 92:144-51. 6 . Wychulis AR, Beahrs OH, WooIner LB. Malignant lymphoma of the colon. Arch Surg 1966; 93:215-25. 7. Kashimura A, Murakami T. Malignant lymphoma of large intestine: 15-year experience and review of literature. Gastroenterol Jpn 1976; 11:141-7. 8. Jinnai D, Awasa Z, Watanuke T. Malignant lymphoma of the large intestine: operative result in Japan. Jpn J Surg 1983; 13~331-3. 9. Nirmala V, Thomas JA, Anthony AJ. Primary malignant lymphoma of colon. Zndian J Cancer 1981; 18:47-50. 10. Wang CH, Wang CY, Liu CH, Hsu SC, Ho SW. Malignant lymphoma of the lower gastrointestinal tract. Transaction of the Hematology Society of the Republic of China 1982; 39-45. 11. Heule BV, Taylor CR, Terry R, Lukes RJ. Presentation of malignant lymphoma in the rectum. Cancer 1982; 49:2602-7. 12. Henry CA, Berry RE. Primary lymphoma of the large intestine. A m Surg 1988; 54:262-6. 13. Lin CH, Chen PC. Primary colonic lymphoma: colonoscopic and barium enema study. Chin J Gastroenterol 1988; 5:152-6. 14. Seo IS, Binkley WB, Warner TFCS, Warfel KA. A combined morphologic and immunologic approach to the diagnosis of gastrointestinal lymphoma: I. Malignant lymphoma of the stomach (A clinicopathologic study of 22 cases). Cancer 1982; 49:493-502. 15. Shepherd NA, Hall PA, Coates PJ, Levison DA. Primary malignant lymphoma of colon and rectum: a histopathological and immunohistochemical analysis of 45 cases with clinicopathological correlations. Histopathology 1988; 12:235-52. 16. Kadin ME, Berard CW, Manba WH. Lymphoproliferative disease in Japan and Western countries. Hunz Pathol 1983; 14:74572. 17. Su IJ, Shin LY, Kadin ME, DunP, Hsu SM. Pathologic and immunologic characterization of malignant lymphoma in Taiwan. A m J Clin Pathol 1985; 84:715-23. 18. Hsu SM, Raine L, Fanger H. Use of Avidin-Biotin-Peroxidase complex (ABC) in immunoperoxidase techniques: a comparison between ABC and unlabeled antibody (PAP) procedures. J Histochem Cytochem 1981; 29:577-80. 19. Shepherd FA, Evans WK, Kutas G, Yau JC, Dang P, Scott JG, et al. Chemotherapy following surgery for stage IE and IIE nonHodgkin’s lymphoma of gastrointestinal tract. J Cliiz Oncol 1988; 6:253-60. 20. Auger MI, Allan NC. Primary ileocecal lymphoma: a study of 22 patients. Cancer 1990; 65:358-61.