Annals of Surgical Oncology, 13(11):1372–1378

DOI: 10.1245/s10434-006-9069-x

Bleeding, Obstruction, and Perforation in a Series of Patients With Aggressive Gastric Lymphoma Treated With Primary Chemotherapy Galia Spectre, MD,1 Diana Libster, MD,1 Sigal Grisariu, MD,1 Nael DaÕas, MD,4 Dina Ben Yehuda, MD,1 Zvi Gimmon, MD,2 and Ora Paltiel, MDCM1,3

1

Department of Hematology, Hadassah Hebrew University Medical Center, P.O. Box 12000, Jerusalem, Israel 91120 2 Department of Surgery, Hadassah Hebrew University Medical Center, P.O. Box 12000, Jerusalem, Israel 91120 3 School of Public Health, Hadassah Hebrew University Medical Center, P.O. Box 12000, Jerusalem, Israel 91120 4 Department of Internal Medicine, Hematology Unit, Bikur Cholim Hospital, Jerusalem, Israel

Background: The management of patients with gastric lymphoma has evolved, with a shift toward nonsurgical treatment. The rates of surgical complications in patients receiving chemotherapy have been insufficiently studied. The objective of this study was to assess the frequency of bleeding, perforation, and gastric outlet obstruction in patients who received chemotherapy as primary treatment for gastric diffuse large B cell lymphoma (DLBCL). Methods: We reviewed files of all patients with gastric DLBCL who were diagnosed and treated primarily with chemotherapy in our hospital between 1990 and 2005. Results: Eighteen (25%) of 73 patients experienced surgical complications, of whom 6 (8%) underwent surgery. Eight patients (11%), six with active lymphoma, experienced gastric bleeding; one required gastrectomy. Eight patients (11%) developed gastric outlet obstruction, of whom three were treated conservatively, three required surgery, one stopped treatment, and one received further chemotherapy. Six of the eight patients had no evidence of active lymphoma at the time of obstruction. Two additional patients underwent gastrectomy due to resistant or relapsed disease. Gastric perforation was not observed. Median survival was 90 months for the entire series, 94 months for patients with gastric outlet obstruction, and 11.5 months for patients with gastric bleeding. Conclusions: Given the rate of surgical complications, especially gastric bleeding and gastric outlet obstruction, there is still an important role for the surgical consultant in the treatment of patients with gastric DLBCL receiving chemotherapy. Gastric perforation, although frequently cited as a complication, is in fact rarely observed. Key Words: Gastric—Diffuse large B cell lymphoma—Chemotherapy—Surgical complications.

The stomach is the most common extranodal site for lymphoma.1 Diffuse large B cell lymphoma (DLBCL) may arise from mucosa associated lymphoid tissues (MALT) lymphoma or as a primary tumor and is the most common histological subtype.2

The presence of reactive lymphoid follicles and lymphoepithelial lesions is suggestive of transformation of MALT lymphoma to DLBCL. However, in the absence of these lesions, the tumor may be histologically and cytologically indistinguishable from DLBCL arising at nodal sites.3 Controversy in the literature remains regarding the optimal treatment for early stages of gastric lymphoma. Historically, surgery was the initial and sole therapy for these tumors. Despite increasing evidence supporting treatment that is based on systemic mul-

Received June 27, 2006; accepted June 28, 2006; published online September 30, 2006. Address correspondence and reprint requests to: Galia Spectre, MD; E-mail: [email protected] Published by Springer Science+Business Media, Inc.
2006 The Society of Surgical Oncology, Inc.

1372

1373

SURGICAL COMPLICATIONS OF GASTRIC LYMPHOMA

tiagent chemotherapy,4–7 some centers continue to operate in early-stage gastric lymphoma.8–11 Complications of gastrectomy are both short- and longterm and include early satiety, abdominal discomfort, afferent loop syndrome, malabsorption, and dumping syndrome.12–14 However, surgical complications also occur in patients receiving chemotherapy. The rate of these complications is underreported in the literature. In the era of evidence-based medicine, treatment decisions should be made on the basis of the best available evidence. Recently a large prospective randomized controlled clinical trial established that chemotherapy is essential in the treatment of the early stages of the disease, and not only in advanced stages, because patients who did not receive chemotherapy had a far lower event-free survival and overall survival compared with those who did receive chemotherapy.14 In that study, only fatal complications were reported. The aim of the current study was to assess the frequency of bleeding, perforation, and gastric outlet obstruction in patients with gastric DLBCL who received chemotherapy as the primary treatment for their disease.

MATERIALS AND METHODS We retrospectively analyzed files of all patients with gastric lymphoma who were treated in our hospital from 1990 to 2005. Only patients with DLBCL (World Health Organization classification) or transformed lymphoma were included. We included all patients with stage I to IV disease.15 Computed tomographic scan, gallium or positron emission tomography (PET) scans, bone marrow examinations, and, for some cases, endoscopic ultrasound were used for staging. Patients with primary gastric lymphoma, recurrent disease, or transformed lymphoma were included in the study. Surgical complications were defined as bleeding (melena or hematemesis that occurred during chemotherapy treatment and required hospital monitoring and blood transfusions); gastric perforation; and gastric outlet obstruction resulting in symptoms of early satiety, eating difficulty, or vomiting, and proven by endoscopy and upper gastrointestinal imaging. Response to therapy was assessed with diagnostic procedures such as computed tomography, endoscopy with biopsies, and gallium and PET scans. If a residual mass was noted but biopsy findings were negative, and gallium or PET scans were also negative, the patient was considered to be in remission.

We compared the frequencies of categorical variables for patients with and without surgical complications by v2 test, or by FisherÕs exact test when expected cell size was