Nine RCTs were included in the review (n=2,174 patients). There were three trials of transdermal glyceryl trinitrate, three trials of sublingual, two trials of topical and one trial of intravenous. Trial sample sizes ranged from 74 to 806 patients. Trial quality was good (4 out of 5) to excellent (5 out of 5).
Post-endoscopic retrograde cholangiopancreatography pancreatitis (PEP) (seven trials, n=1,841 patients): Compared with placebo, there was a statistically significantly lower incidence of PEP with glyceryl trinitrate (OR 0.56, 95% CI 0.40 to 0.79, I2=1.4%; seven RCTs). Subgroup analysis indicated that results were more significant with sublingual administration (OR 0.34, 95% CI 0.16 to 0.75, I2=0%; two RCTs). Results were not significant with transdermal glyceryl trinitrate (OR 0.64, 95% CI 0.40 to 1.01, I2=47%; three RCTs). Sensitivity analysis revealed that the definition of PEP used in the RCTs and the incidence of PEP in the placebo arm affected the significance of results.
Bile duct cannulation (seven trials, n=1,294 patients): Compared with placebo, there was no statistically significant difference in the primary rate of bile duct cannulation (OR 0.99, 95% CI 0.93 to 1.06, I2=0%; five RCTs). The type of administration did not alter the results.
Adverse events: Hypotension and headache were more common in the glyceryl trinitrate group.