THE CLINICAL SPECTRUM AND TREATMENT OF PRIMARY NON-HODGKINS-LYMPHOMA OF THE STOMACH

Background: Primary non-Hodgkin's lymphoma (NHL) of the stomach is a r are disease; diagnosis may be difficult and the method of treatment is at debate, especially the role of surgery. We therefore evaluated cli nical findings, in particular endoscopic diagnosis and long-term resul ts of various treatment strategies, especially the effect of the stoma ch-saving procedure with radiotherapy alone, in patients with stage I and II NHL of the stomach over a 20 year period (1970-91). Patients an d methods: A series of 119 consecutive patients (71m + 48f, median age 62 years) were studied: 80 patients in stage I and 39 in stage II. Tr eatment schedules were: group A: standard radiotherapy alone, after cl inical staging (n = 46); group B: gastric resection followed by standa rd radiotherapy (n = 28); group C: various combinations of resection, radiotherapy plus or minus chemotherapy (n = 45). Since 1978 surgery w as no longer advocated: our resection rate dropped from 83% to 38%. Re sults: Symptoms and signs were non-specific: epigastric pain (81%), na usea (33%), vomiting (24%) and weight loss (66%). Haemorrhage or perfo ration at presentation were unusual, in 14 (12%) and 5 (4%) cases, res pectively. At endoscopy 3 main patterns were recognized: ulceration (3 4%); diffuse infiltration (18%) and a polypoid mass (37%). Endoscopic biopsies were diagnostic in 92%, including 11% after repeated endoscop y Treatment results were excellent for stage I: local control in 93%, 5-year survival 70%; in stage II the results were less favorable: 57% and 37%, respectively. Recurrences were only seen in 17 cases, among w hich 10 cases were extra-abdominal. According to treatment schedule, r esults were similar in group A (n = 46), with radiotherapy alone, and group B (n = 28), with resection plus radiotherapy, leading to a 5-yea r overall survival of 71% and 82%, respectively (no significant differ ence). Severe complications of treatment were infrequent: one perforat ion, another with a fatal haemorrhage in group A. In group C (n = 45) various individualized schemes were applied, usually palliative either due to large tumour extent or patients of old age and poor condition, leading to a 5 year survival of only 31%. Conclusions: Clear patterns were recognized at endoscopy, and diagnosis on endoscopic biopsies wa s possible in the vast majority (92%); in clinical stage I radiotherap y alone is effective, safe and well tolerated.