Introduction Cancer of the upper gastrointestinal tract accounts for approximately 4% of acute upper gastrointestinal bleeding (UGIB) related admissions. The treatment of malignancy related bleeding is challenging due to the diffuse and widespread nature of bleeding. Hemospray (Cook Medical, Winston-Salem, North Carolina, USA), an inorganic haemostatic powder that allows a wide area of therapeutic coverage. To date little is known of its use in tumour related UGIB. The aim of this study was to assess feasibility and effectiveness of Hemospray in the treatment and palliation of malignancy related upper GI bleeding.

Method Retrospective analysis of all malignancy related UGIB treated with Hemospray at Russells Hall Hospital. Patients were identified from the endoscopy database. Patient demographics, Hemospray use and the use of other therapeutic modalities were collected. Outcomes including primary haemostasis, rebleeding and mortality were obtained.

Results Hemospray use was identified in 15 patients, on 19 occasions (range 1–3 per patient), with malignancy related UGIB. Mean patient age was 71 years, range 43–96 years; male gender 66.6% (n = 10). Mean haemoglobin on presentation was 72 ± 17 g/L and mean urea 10.9 ± 6.6 mmol/L.

The bleeding was diffuse and widespread ooze in all cases. The tumour site was oesophageal 40% (n = 6), gastric 26.7% (n = 4) and duodenal 33.3% (n = 5). Hemospray was used as the primary haemostatic treatment on 89.5% of occasions (n = 17). On 2 occasions adrenaline was used first line. Primary haemostasis was achieved in 95% of cases (n = 18). In 1 patient the Hemospray delivery system failed; in this case Endoclot was used successfully. The median procedure duration was 18 min (range 9–50 min).

The 7 day and 30 day rebleeding rate was 10.5% (n = 2) and 15.8% (n = 3) respectively. None of the patients required surgery or radiological intervention to arrest the bleeding. There were no documented complications of Hemospray therapy. Mortality was 5% (n = 1) and 16% (n = 3) at 7 and 30 days respectively. There were no bleeding related deaths; death was due to the underlying condition in all patients. Average length of stay was 19 days (range 3–101 days); excluding the outlier of 101 days the average length of stay was 9 days (range 3–27).

One patient with gastric cancer had Hemospray applied on 3 occasions which palliated bleeding for approximately 2 months duration each time.

Conclusion Overall Hemospray appears to be an effective haemostatic therapy for malignancy related UGIB and should be considered as a primary therapeutic modality in this setting. It may be used as a bridge to more definitive therapies such as radiotherapy or drug therapy (e.g. thalidomide, chemotherapy).

Disclosure of interest None Declared.