Outcome of Gastric Diffuse Large B-Cell Lymphoma and Usefulness of PET-CT in Detecting Gastric Lesion

Background:

The standard therapy for primary gastric diffuse large B-cell lymphoma (DLBCL) is chemotherapy of R-CHOP with/without involved-field radiation therapy. Although some reports indicated that gastrointestinal complications after chemotherapy for DLBCL occur at a rate of 0-26%, little is known about riskfactors for the complications. In addition, to detect DLBCL in gastric lesion, both of positron emission tomography-computed tomography (PET-CT) and esophagogastroduodenoscopy (EGD) are useful tools. However, there have been few reports comparing them. The aim of this study is to show the outcomes including treatment-related complications in patients with gastric DLBCL and risk factors for the gastric complications. Moreover, we evaluated whether PET-CT is sufficient to detect DLBCL in gastric lesion by comparing with EGD.

Patients and methods:

This retrospective study included consecutive patients with newly diagnosed DLBCL between October 2003 and July 2014 who underwent EGD and were treated with R-CHOP in our hospital. We classified the patients into three groups. Group A-1: patients who had documented DLBCL in gastric lesion by EGD and underwent PET-CT; group A-2: patients who had documented DLBCL in gastric lesion by EGD and did not underwent PET-CT; group B: patients who had no documented DLBCL in gastric lesion by EGD and underwent PET-CT. Suspected lymphomatous lesions by EGD were biopsied and immunopathologically examined. Gastric DLBCL was defined only when pathologically confirmed. In PET-CT, gastric lesions with SUV max ≥ 5 were considered positive. Outcomes and risk factors for complications among group A were analyzed using the logistic regression model. We evaluated significance of PET-CT and EGD in group A-1 and B by the positive predictive value (PPV) and the negative predictive value (NPV).

Results:

Among 448 patients diagnosed with DLBCL, 178 patients were enrolled for our study: 55 in group A-1, 28 in group A-2 and 95 in group B. Among 83 patients with gastric DLBCL (group A), the median age was 69 years (range, 29-85). The numbers of patients with clinical stage (Ann Arbor classification) I, II, III, and IV were 27, 18, 5, and 33, respectively. The rate of complete remission was 87%, and the median 3- and 5-year over survival (OS) were 81% and 75%, respectively. The median 3-year OS of patients with very good, good, and poor grade of Revised International Prognostic Index (R-IPI) was 100%, 77%, and 63%, respectively (Figure, p=0.025). Ten patients had gastric complications: 6 with bleeding that needed blood transfusion and 3 with gastrointestinal stenosis defined as ordinary endoscopy could not pass, no patients had gastrointestinal perforation. Most of bleeding (66.7%) occurred during the first cycle of R-CHOP (median, 15 days; range, 1-206). A multivariate analysis showed that low serum albumin (ALB) at diagnosis was an independent risk factor for gastric complications (odds ratio 10.75, p<0.001). The numbers of patients with positive or negative results examined by PET-CT or EGD in group A-1 and B were shown in Table. PPV and NPV of PET-CT were 0.90 and 0.97, respectively.

Conclusions:

The present study showed that R-IPI was also predictive of survival in gastric DLBCL and low ALB at diagnosis as a significant risk factor for gastric complications following R-CHOP. In addition, our data suggested that PET-CT may be sufficient in the role of detecting gastric lesion of DLBCL because of high PPV and NPV.