A case with gastric purpura induced acute gastrointestinal hemorrhage

Abstract

A 72 years old male, with medical histories of hypertension and gout with herb use, was admitted because of tarry stool passage for one day. On admission, the vital signs were relatively stable. Physical exam didn't show significant findings. Blood tests revealed leukocytosis (WBC: 20810/ul), thrombocytopenia (platelet: 2000/ul), coagulopathy (activated partial thromboplastin time: 55.1 second), abnormal liver and renal function (AST: 390U/L, ALT: 155U/L, Cre: 1.8 mg/dl). Abdominal ultrasound revealed parenchymal liver disease without splenomegaly. Esophagogastroduodenoscopy (EGD) disclosed multiple rosette-like hyperemic patches, some with exposed vessels, at the antrum (Fig. 1). Hemoclipping was applied for hemostasis. Biopsy showed no evidence of Helicobacter pylori infection. Esophagogastroduodenoscopy (EGD) in our elderly patient. (A) Multiple gastric purpura presented as rosette-like hyperemic patches at the antrum. (B) Exposed vessel on the lesions. Blood transfusion was performed for severe thrombocytopenia correction. Medication (herb) was suspected to be the possible cause of thrombocytopenia after work-up and discussing with an experienced hematologist. After improvement of thrombocytopenia, liver and renal function, he was discharged few days later. There was no more episode of gastrointestinal hemorrhage. Follow-up EGD one month later showed dramatic improvement of previous gastric purpura (Fig. 2). Follow-up EGD one month later. (A) Shallow ulcer with previous hemoclip at the antrum. (B) Gastric purpura resolved completely. Gastric purpura was initially described in 1997 with the feature of subepithelial hemorrhage [1]. It was a rarely cause of gastric bleeding with only few cases [1–3]. The cause of gastric purpura was attributed to severe thrombocytopenia, probably due to immune-mediated [2]. In our patient, EGD showed multiple rosette-like hyperemic patches in the stomach. This unique picture was the typical feature of subepithelial hemorrhage. Spontaneous resolution after correction of thrombocytopenia seemed to be an evidence of thrombocytopenia induced gastric purpura. Interestingly, there were no purpura-like lesions seen in the esophagus or the duodenum from literatures. The actual reason needed to be further clarified. In summary, severe thrombocytopenia may lead to spontaneous subepithelial hemorrhage with manifestation as purpura in stomach and resulted in gastrointestinal bleeding. Fortunately it was expected to resolve after thrombocytopenia correction. All authors declare no conflicts of interest.