One week of rifampicin pushed long-acting rilpivirine below target; levels recovered within a week of stopping

Design

Pharmacokinetic case report

Population

One adult with suppressed HIV on long-acting cabotegravir/rilpivirine given 7 days of rifampicin

Primary outcome

Plasma cabotegravir and rilpivirine concentrations

Effect

Cabotegravir −57%, rilpivirine −83% (below target) after 7 days; recovering 1 week after stopping

A case report described a man in his forties with suppressed HIV on long-acting intramuscular cabotegravir/rilpivirine every eight weeks, who needed a seven-day course of rifampicin with levofloxacin for MRSA eradication. The combination is contraindicated, so oral tenofovir/emtricitabine and dolutegravir were given as a bridge. Levels were measured weekly.

After seven days of rifampicin, cabotegravir fell 57% (2.08 to 0.90 mg/L) and rilpivirine 83% (0.126 to 0.022 mg/L), dropping below its proposed target trough. One week after stopping, both were recovering, and continued rising from the injection depot without another dose. The viral load stayed undetectable.

The mechanism is enzyme induction of CYP3A4 and UGT1A1. The case shows the effect is both larger than modelling predicted for rilpivirine and faster than many assume, and that it reverses quickly after a short course.

In India, where rifampicin is the backbone of TB treatment, this matters for any patient on long-acting injectable antiretrovirals. A single case cannot define safe management, but it supports oral bridging when a short rifampicin course is unavoidable.

• Do not co-prescribe rifampicin with long-acting cabotegravir/rilpivirine without specialist HIV pharmacology input
• If a short rifampicin course is unavoidable, cover it with a fully active oral regimen
• Expect rilpivirine levels to fall within days of starting rifampicin
• Levels recovered within about a week of stopping in this case, but monitor where testing is available
• Check for rifampicin in any TB or staphylococcal regimen before giving an injectable antiretroviral

Why it matters

It shows the induction effect on an injectable depot is fast and deep, which is the window where resistance can emerge.

Don't overread it

A single case cannot establish a safe protocol for co-administration.

The statistics, in plain English

These are single-patient drug levels, not a trial. The 83% fall is a measured change in one person and may differ in others, but it was larger than modelling had predicted.

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.