Evaluating the functional net value of pharmacologic agents in treating irritable bowel syndrome

The recent FDA provisional endpoint incorporates a one-tailed measure of improvement for IBS based on the underlying motility complaint. However, motility exists along a spectrum. Patients may experience diarrhoea resulting from therapy for their constipation-predominant IBS (IBS-C) or constipation during treatment for diarrhoea-predominant IBS (IBS-D), but still meet a unidirectional motility-based FDA endpoint.

To weigh the reported efficacy of existing therapies based on patient-reported outcomes with negative intestinal side effects in controlled clinical trial data.

We analysed the difference between 'attributable risk' of efficacy based on number needed to treat (NNT) in the literature and percentage of adverse events (AE) of opposite intestinal complaints in placebo-controlled trials identified through a literature search of IBS trials. This calculation was coined 'functional net value' (FNV) or net benefit of the given drug.

For treating IBS-C, lubiprostone caused diarrhoea in excess of placebo in 3.9% of patients, leading to a FNV of 3.9 percentage units. Linaclotide caused diarrhoea in 15.3% resulting in negative FNV (-1.0 percentage unit). For IBS-D, alosetron and tricyclic anti-depressants caused constipation among a respective 16.9% and 13.0% resulting in a FNV of -3.6 and -0.5 percentage units. Among all therapies, only rifaximin did not cause the adverse event opposite the underlying motility complaint and the drug only had benefit, not detriment.

Functional net value (FNV) offers a method of evaluating the net benefit of a drug in IBS. Most IBS treatments have a negative effect on IBS that exceeds the benefits.