Resolution of thrombocytopenia after treatment for Helicobacter pylori: a case report.

Sir, We report the case of a 51-year-old man with idiopathic thrombocytopenic purpura who experienced a complete and lasting normalization of platelet count after antibiotic treatment for a Helicobacter pylori infection. A 51-year-old man with previously untreated chronic idiopathic thrombocytopenic purpura (ITP) was admitted because of melena. An upper gastrointestinal endoscopy showed a small bleeding ulcer in the paracardial region of the stomach. The patient was treated with i.v. ranitidine until endoscopically documented resolution of bleeding, and with a short course of i.v. immunoglobulins (40 g/die for 4 days), which allowed the platelet count to increase to a maximum of 1623109/L. Soon after a laparosplenectomy was performed without complications. The patient was discharged three days after surgery with a platelet count of 403109/L. He continued to take ranitidine orally (150 mg daily) for 2 months. Six months after his splenectomy he was admitted again because of a second episode of endoscopically documented gastric bleeding manifested by melena. Over this period his platelet count never exceeded 1003109/L (Figure 1). Considering that about 80% of gastric ulcers are associated with Helicobacter pylori infection1,2 and that our patient had a positive serologic test for Helicobacter (35 IU/mL, normal values less than 15 IU/mL) (Enzygnost anti-Helicobacter pylori, Boehringer, Germany) he was treated with amoxicillin (1 g three times a day) and omeprazole (40 mg daily) for 14 days. Afterwards he continued assuming oral omeprazole (20 mg daily). His platelet count, which was monitored every two weeks, started to increase at the end of antibiotic therapy and reached normal and stable values after 5 weeks (Figure 1). The first endoscopic control was performed 6 weeks after the end of treatment. No ulcers were seen, nor was Helicobacter pylori found, but a polyclonal B lymphocytic infiltration of the mucosa was detected. Three subsequent endoscopic controls with multiple biopsies were performed every three months. Histologic specimens remained negative for Helicobacter pylori and the lymphocytic infiltrate was not documented again. A serologic test for Helicobacter pylori was 20 IU/mL 6 months after the end of therapy. Helicobacter pylori infection can be associated with development of lymphoid follicles in the stomach, which may disappear after eradication of the micro-organism.3,4 This kind of infection has been hypothesized to be related with the development of gastric MALT lymphoma;5,6 its possible involvement in some autoimmune disorders is still an open question.7,8 Recently it was reported that Helicobacter pylori eradication can be followed by a significant increase in platelet count in patients with ITP and that the prevalence of such infection in patients with ITP can be high.9 Diagnosis of Helicobacter pylori infection in our case was based only on epidemiological probability and serologic data, since it was not possible to obtain histologic documentation of infection and other non-invasive tests were not available. The diagnostic accuracy of serology is, however, high10 and the infection was very probable in our patient who had a gastric ulcer and a positive serologic test. The normalization of his platelet count following antibiotic treatment was impressive: on the basis of this observation and of the recent report by Franceschi et al.9 a relationship between Helicobacter pylori infection and ITP can be hypothesized. This deserves further investigation in a larger number of patients.