Article Text
Abstract
Introduction Nicorandil is an anti-anginal drug with a double mechanism of action, featuring a nicotinamide portion which acts on KATPchannels and a nitric oxide group, which relaxes vascular smooth cells. Apart from common gastrointestinal adverse drug reaction (GI-ADR), its safety and tolerability are satisfactory. No ulceration involving the gastrointestinal tract was officially associated with it before 1997 but several reports have suggested a link with gastrointestinal ulceration and fistulas. The aim of this systematic review was to quantify and characterise nicorandil-associated GI-ADRs by identifying all relevant literature reporting side effects of Nicorandil therapy on the GI system per anatomical location.
Method This study complied with the PRISMA statement and comprised a search of Medline, CINAHL, EMBASE, Biomedical Reference Collection, WHO ADR Database, MHRA reports and Cochrane Library using the keywords: Nicorandil, gastrointestinal, side-effect, adverse reaction, ulcer, ulceration, mucosa, cutaneous, fistula. Parameters recorded include median age, median dose, history of symptoms, length of therapy, healing time after withdrawal of the drug. Data from literature is synthesised calculating a median and interquartile range (IQR) for all medians of observations. Difference in distribution of quantitative data is analysed with a Mann-Whitney test. Correlation between quantitative variables is assessed with a Spearman’s correlation coefficient. A p value <0.05 was considered significant.
Results Oral and distal GI involvements are the most common ADR (28–29% and 27–31% of all gastrointestinal ADR, respectively). The hepatobiliary system, the pancreatic and salivary glands were not affected by Nicorandil exposure. The median time to develop oral ulcerations was 74.0 weeks among people on <30 mg/die (IQR 19.4–197.7) versus only 7.55 weeks (IQR 5.25–48.7) in individuals on higher regimens of Nicorandil (p = 0.47). There was a significant correlation between Nicorandil dose and ulcer healing time across all the data (Spearman’s rho 0.525, p < 0.001).
Conclusion The ulcerogenic metabolite hypothesis is the most convincing theory of why nicorandil causes ulceration and this is supported by the data analysis. Nicorandil seems to act as a necessary, but not always sufficient, cause of the ulcerations. A triggering element should coexist, whether that be a minor aphthosis, diverticulitis, ischaemic colitis, surgical or mechanical trauma, an anal abscess or even a simple pruritus ani. Whether the action of the metabolites relies on a specific mechanism or a simple chemical ulceration still has to be established.
Disclosure of interest None Declared.