Endoscopic management of gastrointestinal bleeding in cancer patients with severe thrombocytopenia


Session type:


Paul Hampel1,Guilherme Piovezani Ramos2,Moritz Binder2,Manuel Braga Neto2,Dharma Sunjaya2,Badr Al Bawardy2,Elizabeth Rajan2
1Mayo School of Graduate Medical Education,2Department of Internal Medicine, Mayo Clinic, Rochester, MN, USA



Coagulation deficiencies, common in cancer patients, can lead to gastrointestinal bleeding (GIB) via vulnerable existing lesions. We aimed to describe the clinical presentation and corresponding locations of bleeding, as well as assess the utility of endoscopic management and the associated post-procedural course in cancer patients with severe thrombocytopenia (ST) and GIB.


Retrospective study; adult patients with active malignancy or post-bone marrow transplant and with a platelet count (PC) <50x103/mL at the time of GIB who underwent an inpatient endoscopic procedure within 24-hours of GIB were included. The primary outcomes of interest were median number of packed red blood cell (PRBC) and platelet transfusions pre- and post-procedure, rate of re-bleeding, and length of hospital stay. The secondary endpoints included 30-day mortality following the procedure.


59 patients were included in the study. The mean age was 59-years (19-92); 61% men. The most common etiology of ST was hematological malignancy (61%). The most common clinical presentation was melena (53%). Median PC and hemoglobin at time of GIB and prior to endoscopy was 39 x103/mL(6-49x103/mL) and 7.9g/dL(3.9-11.7g/dL), respectively. Bleeding location was identified in 58%, most commonly gastric ulcer (14%). Endoscopic intervention was performed in 52% of cases, most commonly clipping. There was a decrease in the median platelet and PRBC transfusion requirements before and after endoscopy (2vs1,p<0.001) and (3vs2,p=0.001), respectively. Re-bleeding rate within 30 days was 22%. Median hospital stay was 9 days. All-cause and GIB mortality at 30 days were 22% and 3.4%, respectively.


In our cohort, endoscopic evaluation was able to be safely performed in cancer patients with GIB in the context of ST and was associated with a reduction in transfusion requirements (PRBC and platelets). These findings support the consideration of including endoscopic therapy as part of management in this clinical scenario.