Update on nelfinavir recall: plan for safety registries

Simon Collins, HIV i-Base

We included several pages in the last issue of HTB covering the recall, due to contamination, of the protease inhibitor nelfinavir, manufactured by Roche Laboratories. This recall later resulted in suspension of Roche’s license to distribute the drug.

The following update was provided in two meetings organised by Roche in Sydney that were outside of the main IAS conference programme.

The company has now tracked down which countries received contaminated batches of nelfinavir: Botswana, Burkina-Faso, Cameroon, Egypt, France, Germany, Iran, Italy, Kenya, Mali, Mexico, Mozambique, Nigeria, Portugal, South Africa, Spain, Taiwan, Uganda, Ukraine and the UK.

At the meeting Roche could only give a rough estimate of how many people this might affect because they only know the quantity of drug sold – not whether it is used for continuous treatment, short course PEP or pregnancy etc. The estimate given at the meetings however said that this may affect up to 20,000 people.

Roche commit to providing a breakdown of what percentage of the total batches were contaminated in each country above, though this had not happened by the time we went to press.

The problem in the manufacturing process arose from a main drum that was not cleaned properly and this resulted in 8 batches having much higher levels of methane sulfonic acid ethyl ester (EMS).

Larger quantities of drug was affected because each of these batches contributed to a wider range of final batches. Nevertheless, all contaminated shipments from this period have been tracked though to final drugs produced.

However, given that Roche’s license has been suspended by the EMEA, distribution of nelfinavir has been stopped worldwide and nelfinavir already distributed has been recalled even from countries where there was no contamination. This is clearly causing major problems for people in any country when nelfinavir was widely used. For many countries this is their main second line treatment and was estimated at the meeting to be affecting up to another 20,000 people.

There is very little data on toxicity risk of EMS, and none in humans. In rats and mice at exposures 200 times the maximum expected from this manufacturing problem, the compound is genotoxic (it gets into DNA and mistakes made by DNA can induce cancers). However, EMS is quickly degraded by the body and can’t be detected after a few hours.

The meeting referred to an in vitro study published earlier this year in human cells looking a similar compound. This showed that toxicity may not be immediate from any exposed dose. Instead, low exposures may have no effect until a certain minimum concentration is reached and then toxicity increases. Even ‘worst-case’ exposure is calculated as being below this predicted minimum concentration.

Two registries are being set up to monitor for whether anyone was harmed from the exposure to EMS:

  1. the first registry will include all patients who used NFV from March to June 2007 in the countries affected
  2. the second will look at all NFV-use in pregnancy, children and exposed but not infected children since the drug was approved in 1998. This is because there is a chance that previous batches of NFV included EMS. Apparently, the EMEA warned of this risk back in 2001, and when production was checked, levels were below any minimum concern (less than 1 part per million). It is not clear why this didn’t become part of the routine batch of safety tests from then on.

Paediatric formulations – powder: none of the contaminated batches, by chance, were used in production of paediatric powder – though many children split or crush tables rather than use the powder which is very bulky.

Slides from this meeting and further information

Direct link for PDF

Further links

Links to other websites are current at date of posting but not maintained.