Challenges with annual TAF implant

Polly Clayden, HIV i-Base

In the first-in-human tenofovir alafenamide (TAF) implant trial, there was lower than planned drug release in most participants and tolerability was suboptimal – according to data presented at CROI 2024. [1]

There is an urgent need for a long-acting – ideally annual or longer – HIV prevention technology to overcome oral PrEP adherence challenges.

The CAPRISA 018 phase 1 trial looked at an annual TAF implant in low-risk adult South African cisgender women without HIV.

The silicone implant, developed by the Oakcrest Institute, has the same dimensions as a contraceptive implant (40 mm length and 2.5 mm OD), except with two delivery channels. It is loaded with 110 mg TAF free-base micro-tablets. Target release 0.25 mg/day.

To assess initial safety, six participants aged 18 to 40 years received a single TAF implant (bicep insertion), 28 days in situ) (Group 1).

After this, 30 were randomised 1:1 to either one or two TAF implants and assigned blinded in a 4:1 active (n=24) to placebo (n=6) ratio, for 48 weeks (Group 2).

Among the 36 participants, median age was 26 years (IQR: 22 to 30) and weight 72 kg (61–80). Five participants had used a contraceptive implant in the past or were using one during the study period. Overall retention was 97%.

Systemic AEs and ISRs were mostly mild: Grade 1, 55.3% and 90.9% respectively. Two of 36 participants (Group 2) had Grade 3 ISRs – both implant site abscesses. These were resolved with early removal.

Overall, 31% (11/36) were removed early (before 48 weeks) at a median of 19 weeks (range: 2 to 27); of these 55% were participant-initiated and the remainder clinician-initiated. These were mostly associated with ISRs.

One third (10/30) of TAF implants and 17% (1/6) of placebo implants were removed early. As were 22% (4/18) of single and 50% (6/12) of double implants.

ISR incidence per person month: 2.1 (95% CI: 1.8 to 2.6) vs 0.8 (95% CI: 0.6 to 0.9) in early vs scheduled removals, respectively.

Mean removal time was 40 minutes early vs 19 minutes scheduled for one implant and 81 minutes early vs 33 minutes scheduled for two implants.

Plasma TAF was detectable in 77% of samples from Group 2 participants in the active arm, with 100% and 50% detection 0.5 and 6 hours post insertion.

But median TFV-DP concentrations were 3.9 (IQR 1.7 to13.3) and 14.8 (IQR 6.0 to 29.1) fmol/million cells in participants with 1 and 2 active implants respectively, with only 15% of samples achieving or exceeding the target concentration of 36 fmol/million cells.

comment

The investigators noted that there was no dose dumping, drug released over one year but residual TAF was high at removal.

There was lower than planned drug release – the implant needs to release more TAF: two implants had a median TFV-DP of 14.8 fmol/million cells which is well below the target of 36 fmol/million cells.

Tolerability was poor and early removals were more difficult and took considerable time.

The challenge will be to increase drug release and lower ISRs. If tolerability can be improved, increasing the release from two implants could achieve target TFV-DP levels for a year.

As Charlie Flexner remarked, in his excellent plenary, this is a prototype system and an important study to inform future implants.

If the challenges described in this initial investigation can be overcome, such an implant could meet an urgent HIV prevention need.

References

Gengiah TN et al. Phase I safety, tolerability and pharmacokinetics of tenofovir alafenamide implants in African women. CROI 2024. Denver, Colorado. 3–6 March 2024. Oral abstract 123.
https://www.croiconference.org/abstract/phase-i-safety-tolerability-and-pharmacokinetics-of-tenofovir-alafenamide-implants-in-african-women/ (abstract)
https://www.croiwebcasts.org/p/2024croi/croi/123 (webcast)
Flexner CW. The end of oral? How long-acting formulations are changing the management of infectious diseases. CROI 2024. Denver, Colorado. 3–6 March 2024. Oral abstract 39.
https://www.croiwebcasts.org/console/player/52288