Removable PrEP implant produces high drug levels in rabbits; further studies planned

A small removable plastic implant resembling a tiny piece of dry spaghetti is capable of delivering sustained levels of the drug tenofovir, researchers told last month’s HIV Research for Prevention conference (HIVR4P 2018) in Madrid.

The drug delivery capability of the implants was tested on rabbits and will now advance into tests on dogs and non-human primates before phase 1 trials in humans, researcher Ariane van der Straten told the conference. Van der Straten belongs to an initiative called the Women’s Global Health Imperative, which promotes research into sexual health and safe motherhood in women and girls globally.

There has been considerable interest in developing removable implants for HIV pre-exposure prophylaxis (PrEP). Long-term injectable versions of anti-HIV drugs have already shown efficacy as treatments and have reached phase II trials as PrEP. PrEP injections are popular among users but suffer from the fact that cabotegravir, the drug used in the trials, persists in the body for an extraordinarily long time – years in some cases, as another study presented at Madrid showed.

This sets up the conditions for the ex-user of injectable PrEP to develop drug-resistant HIV if they are exposed to an infection, and one case of this happening was reported in 2014. Even though an injectable version of the drug concerned, rilpivirine, is no longer being taken forward, the possibility still exists with other drugs. The only way to ensure this does not happen it to cover the “long tail” of drug level decay with oral PrEP – which obviously makes no sense if people choose injectables because they prefer them to pills.

For this reason, there is intense interest in developing PrEP as a removable implant. The technology exists for these, and removable contraceptive implants containing levonorgestrel have been in use for 30 years; one implant can be effective for three to five years.

However, all drugs have different distribution and excretion characteristics, so trials need to be done of the most effective formulation of a PrEP implant.

Even before these are done, as previous PrEP studies have shown, it is important to do acceptability research to establish what people want from a PrEP implant. Van der Straten told the conference that 14 focus groups were conducted with 105 potential implant users and 30 healthcare providers were interviewed, all in South Africa.

A preference was expressed for a smaller and thinner implant than the one already devised, discreetly matched to skin tone, and participants said they would be willing to have two implants at once if it meant changing them less often. As a result, the initial implant, which was more like a drinking straw containing drug, was reformulated as a thinner but stiffer extruded-polymer version measuring 40mm by 2.5mm.

The implant is inserted with a trocar, a medical device commonly used to insert cannulas, and can be removed easily with one too, though it is planned to make the implants with a biodegradable plastic that will be absorbed by the body. Users also said they would like smaller ‘test’ implants to be made that can be used on a trial basis, and also asked if they could be made with some kind of monitoring system that would show when they had run out of drug.

The Women’s Global Health Imperative have been testing rabbits with drug implants that deliver the ‘newer’ version of tenofovir, tenofovir alafenamide or TAF. This drug was chosen because it gets delivered so efficiently into tissue cells. They used three different versions designed to deliver a low, medium or high dose of TAF, which means that 0.2, 0.4 or 0.8 milligrams of drug per day are delivered – the high dose requiring two implants.

In total 24 rabbits received one of the three doses. Because the high dose necessitated two implants, the rabbits on the lower doses received one ‘dummy’ implant as well as a real one.

The implants were inserted into the rabbits' front legs for nine weeks and then removed. Each week, drug levels were measured weekly in blood and in PBMCs – T-lymphocytes in the bloodstream. Two of the eight rabbits in each group were humanely killed and had drug levels in vaginal, cervical and rectal tissues biopsied at week six, as were three rabbits in each group at week nine.

The rate that drug diffused out of the implant into saline in a lab dish was also measured to find if the way it diffused in animal tissue was different: it was not significantly so.

Drug levels in blood reached steady state – above 8.2 nanograms per millilitre – in blood after seven days and were maintained throughout, or had even slightly increased by week nine. Drug levels in PBMCs did not build up to a steady state (maximum) till week six or even week seven but were already well above the expected effective concentration for PrEP at the first measurement, seven days after the implants were inserted.

The expected effective concentration for PrEP in PBMCs is 48 femtomols per million (fmols/10⁶) cells. Levels were well above this at all time points; the average level for high, medium and low doses was 1203, 990 and 440 fmols/10⁶ cells averaged from week one to week nine, and the mean lowest level seen was 790, 655 and 175 fmols/10⁶ cells respectively.

Levels in tissues were expressed as fmols per milligram. In the rabbits receiving the lowest dose they were about 10 fmols/mg at week five, 15-60 fmols/mg at the medium dose and 15-70 fmols/mg at the high dose. It was notable that tissue concentrations in the rectum tended to be lower than concentrations in the vagina or cervix, contrary to some other PrEP studies. At week nine the levels in the low and medium doses were only slightly higher, but levels in the animals receiving the high dose were markedly higher at about 75-320 fmols/mg in the vagina, 65-120 fmols/mg in the cervix and 25-75 fmols/mg in the rectum (though it is worth noting these were not the same animals).

All the implants were retrieved intact. Some reactivity to a foreign body under the skin is to be expected and has been seen in people using other implants, ranging from slight redness to mild fibrosis (scarring). There was no reaction in the rabbits that gave cause for concern.

The next step in the implant studies is to do longer studies – six months or more – in rabbits, dogs and monkeys before progressing to phase I human trials.