High Prevalence of Helicobacter pylori (Hp) Infection in Ocular Adnexal Lymphoma (OAL).

Association between infectious agents and non-Hodgkin’s lymphomas, such as Hp infection and MALT-type gastric lymphomas or Chlamydia psittaci and OAL, has been previously reported. Therefore, in a large monocentric cohort of OAL patients, we investigated the prevalence of Hp infection in gastric biopsy specimens obtained during the initial staging period. The detection of Hp infection was determined by Giemsa staining of gastric biopsies and by Hp-specific PCR. Among the patients treated at the Institut Curie between 1970 and 2006 for an OAL and for whom gastric tissues were available for PCR analyses, 83 cases were selected for the study. Pathological review according to the WHO classification showed 5 patients with lymphocytic lymphoma, 14 cases with lymphoplasmocytic lymphoma, 6 patients with follicular NHL, 47 cases with MALT type lymphoma (57%), 3 patients with unclassified low-grade NHL, 2 patients with mantle-cell lymphoma, and 6 cases with diffuse large B-cell lymphomas. Initial characteristics were: median age 66 years (range: 26–90); M/F sex ratio 0.9; B symptoms 1%; PS ≥ 2 in 2%; Stages I and IV 59% and 41%, respectively; nodal involvement 24%; 2 or more extra-nodal localizations 34% (10% of bone marrow involvement, 8% of gastric involvement, and 5% of lung localizations); elevated LDH in 9%; serum albumin lower than 40g/l in 32% of patients; and IPI score of 0–1/2/3/4–5 in 60%/21%/18% and 1% of cases, respectively. Ocular localization were intra-orbital in 32 patients (39%), conjunctival in 36 (43%), palpebral in 4 cases, lacrymal in 10 patients, and intra-ocular in 1 case; bilateral ocular sites were found in 10% of cases. The research of Hp infection was performed in 51 patients by Giemsa staining of gastric biopsies with 21 positive cases (41%), and in the 83 selected patients by PCR analysis with 29 positive cases (35%), with an overall number of 37 Hp positive patients (45%). Correlation between histopathological and PCR analyses evaluated on the 51 patients for whom both analyses have been performed is presented in Table 1 (p = 0.011). Finally, we observed that the presence of gastric Hp infection did not correlate with patients’ characteristics of the overall population and of the 47 MALT-type cases. However, on the 83 studied patients, we found a trend of significant difference in Hp positive cases including low albumin levels (p = 0;09), elevated stage (p = 0.08), 2 or more extra-nodal localizations (p = 0.08), and high IPI score (p = 0.06). In conclusion, we observed a high prevalence of gastric Hp infection at the initial diagnosis of OAL, namely 45% of selected cases, with a possible but not significant association of Hp infection with pejorative clinical characteristics. In order to well define this observation and clearly evaluate the impact of Hp infection on the clinical presentation, we are now investigating the prevalence of gastric Hp infection in patients having different histopathological subtypes of nodal malignant lymphomas.