Kim, Y.I., Kim, M.J., Park, S.R., Kim, H.K., Cho, S.J., Lee, J.Y., . . . Choi, I.J. (2017). Effect of a proton pump inhibitor on tumor bleeding prevention in unresectable gastric cancer patients: A double-blind, randomized, placebo-controlled trial. Journal of Gastric Cancer, 17, 120–131.

The purpose of this study was to evaluate the effect of a protein pump inhibitor on tumor bleeding prevention in patients with advanced gastric cancer. The primary endpoint was tumor bleeding; secondary endpoints were the number of transfusions received and overall survival.

Patients with inoperable gastric cancer were randomized into one of two groups, with stratification based on hemoglobin level. The study group received lansoprazole 30 mg PO every day. The control group received placebo. Bleeding rates were assessed at 4 months and at a median follow up of 6.4 months. Bleeding was assessed by blood counts every three weeks.

• N = 127   
• AGE: Median is 56 years.
• MALES: 78.7%  
• FEMALES: 21.3%
• CURRENT TREATMENT: Chemotherapy
• KEY DISEASE CHARACTERISTICS: Unresectable gastric cancer
• OTHER KEY SAMPLE CHARACTERISTICS: Age ≥ 18 years, ECOG 0-2

• SITE: Multisite   
• SETTING TYPE: Outpatient    
• LOCATION: Korea

• PHASE OF CARE: Active antitumor treatment
• APPLICATIONS: Palliative care

Prospective, randomized, double-blind, placebo controlled trial

Bleeding events were defined as the presence of witnessed melena or hematemesis, decrease in hemoglobin by greater than 2 g/dL in one week, or > 3 g/dL in three weeks, with endoscopic evidence of bleeding. Endoscopic bleeds were graded per Forrest classification. Gray’s test and the Fine-Gray regression model were used to evaluate time to tumor bleeding events and the effect of lansoprazole upon bleeding.

There was no significant difference in tumor bleeding between the two groups.

Lansoprazole did not have a preventive effect on tumor bleeding in patients with inoperable gastric cancer, nor was there a difference in the number of transfusions required. The lansoprazole group did have a significantly reduced number of tumor bleeding events compared with placebo, until the four-month mark. At that point forward, the difference became statistically insignificant. This may be related to disease progression and increased tumor burden during palliative chemotherapy treatment.

• Unintended interventions or applicable interventions not described that would influence results
• Other limitations or explanation: The study was underpowered for the primary outcome. Patients were receiving first- or second-line chemotherapy during the study. The study planned to enroll 394 patients but ended because of a low recruitment rate. The study did not look at various dosing options of lansoprazole. Many patients were lost to follow up. Median follow-up duration was brief. The placebo group included more patients who received second-line chemotherapy.

The risk of tumor bleeding in patients with inoperable gastric cancer is not decreased with 30 mg of lansoprazole daily, nor did it decrease the requirement for transfusions.