Ranitidine and cancer: most of the adverse-event signal looks like reporting bias

Design

Retrospective disproportionality analysis of FAERS and JADER with a systematic review of comparative studies

Population

Spontaneous reports in FAERS (140,985 ranitidine cancer reports in 2020 to 2021) and JADER

Primary outcome

Ranitidine-associated cancer signals, with active-comparator and time-restricted analyses

Effect

JADER PRR 3.54 among suspected-drug reports, 0.92 with all involvement types

Ranitidine was withdrawn worldwide in 2020 after contamination with N-nitrosodimethylamine, a probable human carcinogen. This Drug Safety analysis examined whether cancer signals in two spontaneous-reporting databases reflect real risk. It used the FDA Adverse Event Reporting System (FAERS) and the Japanese JADER database, four standard disproportionality algorithms, famotidine as an active comparator, time-restricted analyses before and after the recall, and a systematic review of comparative studies.

FAERS recorded 140,985 cancer reports for ranitidine in 2020 to 2021, more than two thirds of all its cancer reports. Most were submitted by consumers and clustered in regions affected by regulatory actions and litigation. Unrestricted analysis gave high-magnitude signals for several cancers, but these weakened or disappeared in active-comparator and time-restricted analyses, and ranitidine's and famotidine's reporting profiles converged. In JADER, a signal among suspected-drug reports (PRR 3.54) disappeared when all involvement types were included (0.92). Most large cohort studies in the review found no increased risk of overall or site-specific cancer.

The authors attribute the signals mainly to the contamination alerts rather than intrinsic carcinogenicity, and they note protopathic bias. This is a methods lesson: a disproportionality signal is a hypothesis, and publicity and litigation can generate a signal without any change in risk. It does not show that ranitidine could never raise cancer risk, and it does not reverse the withdrawal.

• Treat a disproportionality signal as a hypothesis, not a risk estimate.
• Ask whether an alert, a recall or litigation could have driven reporting before reading a signal as causal.
• Look for an active comparator and time-restricted analysis when judging a pharmacovigilance claim.
• Counsel patients who took ranitidine that large cohort studies generally did not show increased cancer risk.

Why it matters

It shows how a regulatory alert can generate its own safety signal, which matters whenever a new warning draws reports.

Don't overread it

The analysis explains the reporting signal; it does not prove that ranitidine never causes cancer or that the withdrawal was unwarranted.

The statistics, in plain English

A proportional reporting ratio of 3.54 means a drug is reported with an event 3.5 times as often as the background rate, which says nothing about how many people take the drug. Once all reports are counted, as in JADER, it fell to 0.92, showing how sensitive it is to which reports are included.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.