The PrEP experience: IAS 2015 and beyond

1 October 2015. Related: Conference reports, HIV prevention and transmission, IAS 8th Vancouver 2015.

Simon Collins, HIV i-Base

Oral pre-exposure prophylaxis (PrEP) was a major conference theme at IAS 2015 and featured in numerous plenary talks, oral abstracts and dozens of posters.

Apart from a few notable exceptions discussed below, the integration of PrEP into many overview talks was perhaps more significant than the new data from individual studies. PrEP is now expected to play a key role in reducing HIV incidence globally, being included as a recommendation in WHO treatment guidelines and as part of the UNAIDS 90:90:90 campaign. [1, 2]

In one of the first plenary talks at the conference, Francois Venter from the University of Witwatersrand, gave a compelling perspective from a doctor on the comparative roles for different biomedical prevention options. He described the new data on PrEP as an “electrifying moment” remarkable for working far better in real life settings that in the randomised regulatory studies: the higher someone’s risk of HIV, the more likely they are to take PrEP and benefit. The earlier preoccupation about risk compensation has now been largely debunked and drug resistance occurred at very low rates in studies. With the more realistic acceptance that condoms alone are clearly not enough for many people at high risk, the cost of PrEP is now a major barrier to wider access, especially in higher income countries. [3] In the UK, this has partly been overcome by increased off-label use of generic PrEP over the last year – imported for personal use from internet pharmacies – that is both cheap and legal.

Given that the potential for using tenofovir disoproxil fumarate (TDF) before exposure to prevent HIV infection has been known for more than two decades, [4] the timeline for PrEP has been very slow.

Irrespective of scientific questions of efficacy, several studies at IAS 2015 also touched on why historical attitudes to HIV prevention, including HIV-related stigma, continue to block access to this option for better health.

PrEP research also involves more difficult challenges than treatment research. On the one hand, with good adherence, daily or near-daily dosing achieves greater than 99% protection against HIV infection. This is a remarkable fact. On the other hand, the unpredictability of real life, and the short but important delay before oral PrEP offers protection, highlights the urgency of finding alternatives to daily dosing that will have comparable close-to-100% efficacy.

The two drugs used in current oral PrEP – TDF and emtricitabine (FTC) – have different absorption times, for different body sites, that includes important dosing differences between men and women. The difficulty with interpreting pharmacokinetic (PK) data was nicely understated by Dave Glidden from University of California in a talk at IAS 2015 that looked at how quickly PrEP becomes protective: “We don’t totally understand how much drug there needs to be, where it needs to be, and when it needs to be there”. [5]

The main dosing difference is driven by the 10- to 100-fold higher intracellular tenofovir diphosphate levels in rectal compared to vaginal tissue. This translates to guidance for optimal adherence to daily dosing as requiring 4-7 doses a week for men and 6-7 doses a week for women. However, this is based on a modelled target IC90 for tenofovir of ≥15.6 fmol/million viable PBMCs, which still needs to be confirmed in clinical studies. It is also unclear how to interpret the wide confidence interval for this target (95%CI: 3.0 to 28.2). [6]

Importance of different dosing strategies based on different HIV risk

On a population level (rather than individual protection), whether or not PrEP has a long-term impact on reducing rates of transmission might be dependent on having options of different dosing strategies that can be flexible and adapted to individual risk and changes in risk.

Someone who is at risk several times a week has different needs from PrEP compared with someone whose HIV risk might be less than once a month. Individuals commonly change their degree of risk, so dosing needs to be adaptable to the dynamic of changing personal circumstances.

For all the best supporting data, daily PrEP is unlikely to be sustainable for many years for someone at very low risk when the risk:benefit ratio might not support broad use based on safety, adherence and cost. But demand for PrEP might actually be highest from people with relatively few risks. This might relate to quality of life issue for people who want to be able to have sex without the fear of HIV, irrespective of whether they use a condom.

Patterns of PrEP taking varies greatly in efficacy studies. Both IPrEX and IPERGAY demonstrated this when adherence was shown as median numbers of pills taken by each individual for each 1-3 month period during the studies. Only a minority of people showed either consistently high or constantly low adherence throughout. Variation in adherence was far more common: consistent high adherence for several months, followed by a break with low or minimal use of PrEP. Both studies suggest that it is common for periods of risk to vary considerably for many people.

HPTN 067 – three dosing strategies to achieve PrEP protection in relation to risk

Several different analyses from the HPTN 067 ADAPT study were presented at IAS 2015 in symposia, oral abstracts and posters. This study is important because it randomised three socially distinct groups of people at high risk of HIV infection, each to three different PrEP dosing strategies.

The primary outcome of the study was adherence based on drug coverage in relation to risk – ie not just as a measure of pills taken. This combined adherence endpoint was measured using electronic pill boxes and weekly phone calls from a researcher. Coverage was defined as someone having taken one or more pills in the four days before sex and one or more pills in the 24 hours after sex. Partial coverage was when only one of the components were taken.

The three groups were: gay men and transgender women in Harlem (US), gay men and transgender women in Bangkok (Thailand), and heterosexual women in an informal settlement near Cape Town (South Africa). Median age was approximately 30 years in all studies with up to one third of participants younger than 25. All three groups were at high risk, with 12%, 19% and 7% of potential participants testing HIV positive during screening in the Harlem, Bangkok and Cape Town groups respectively. A small number of transgender women were enrolled in the Harlem and Bangkok sites.

The dosing strategies were: daily PrEP (D); time-based PrEP (T) – defined as two pills a week, plus one pill after sex; and event-based PrEP (E) – defined as one pill before sex and one after. The study design included: a four week lead-in period where once weekly (ie suboptimal) dosing was directly observed (DOT); a two week PK and washout period; 24 weeks of self-administered dosing; and a final 4 week washout period that included quantitative research. Participants were told that efficacy for PrEP had only been proven for daily dosing.

Bob Grant from UCSF presented the overall study design, [7] including the results from Cape Town, previously presented earlier this year at CROI 2015 [8].

Results from the Harlem and Bangkok groups were presented for the first time at IAS 2015, both in the symposium on PrEP (that was webcast) and later the same afternoon as oral late breaker abstracts (that were not). PK sub-studies were only included for the Cape Town and Bangkok sites.

Although HPTN 067 was not powered for efficacy, the few infections that occurred were all in people with either very low or no detectable drug levels at seroconversion. There were 1, 2 and 2 new infections at the Harlem, Bangkok and Cape Town sites respectively during the suboptimal lead-in DOT phase and 1, 0 and 5 new infections during the 24-week main study.

For the women in Cape Town, daily dosing regimen resulted in greater drug coverage – in relation to times someone had sex – than either time-based or event-based dosing (75% vs 56% vs 52% of events covered, p=0.0006 D/T, p<0.0001 D/E, p=NS T/E). In the PK sub-study, daily dosing resulted in a higher percentage (of participants who had sex in the previous seven days) with detectable intracellular tenofovir (in 81% vs 52% vs 54% by group at week 10 and 56% vs 46% vs 32% at week 30). This was defined as levels >9.1 fmol/million PBMCs, which approximating to two doses/week. [7]

In the Harlem group, daily PrEP again resulted in significantly greater coverage for the times people had sex (66% vs 47% vs 52%, p=0.001 D/T and D/E, p=NS T/E). [9, 10] As in the Cape Town site, when people reported partial coverage, it was the post-dose that was generally missed.

In contrast, the Bangkok group reported much higher drug coverage for all strategies (85% vs 84% vs 74%). For people reporting partial coverage, the pre-dose was taken more often than the post-dose. In the PK sub-study, intracellular tenofovir was detectable in >85% of participants who had sex in the previous seven days, (in 100%, 87% and 93% at week 10 and 91%, 95% and 86% at week 30, in the D, T and E groups respectively). In the context of high adherence, the long intracellular half-lives of both drugs (40 to 60 hours), probably contributed to showing similar protection to daily dosing. In this context, event-based dosing achieved similar coverage with significantly fewer doses. [11, 12]

Although not explicitly reported, the difficulty with the post-sex dose reported in all three sites, might have been related to a misunderstanding that this dose does not have to be immediately afterwards, but anytime over the subsequent 24 hours.

Qualitative results from all three sites provided important insight into some of the social determinants relating to PrEP. HIV-related stigma against people who were HIV negative was commonly reported based on the false assumption that having HIV meds meant they were positive. [13, 14, 15]

Stigma and misconceptions were reported in all countries, again when people were taking PrEP at the time of sex – which misses a key benefit of PrEP actually being better if taken the day before, at any time on the day of sex and anytime the day after.

Unanswered questions relating to IPERGAY event-based dosing

The French/Canadian IPERGAY study set out to look at event-based dosing using a randomised placebo-controlled design that was powered for efficacy. The IPERGAY dosing is increasingly important because it is currently being used in the UK as a basis for taking PrEP, even though the study itself produced only limited data from people who only had sex infrequently.

IPERGAY has some similarities to the UK PROUD study. Both studies were in a similar high-risk population of gay men, and both reported 86% reductions in HIV infection from PrEP in presentations at CROI 2015 following a DSMB recommendation to offer immediate PrEP to all participants. [16, 17]

But while PROUD used daily dosing, IPERGAY used event-based dosing that involved a minimum of four pills from having sex once. This schedule was:

• A double pre dose (2 x coformulated TDF/FTC taken 24 to 2 hours before sex).
• A single TDF/FTC taken within 24 hours of having sex – and also on each subsequent day if they had sex for a few days – and
• A final single TDF/FTC taken the day after the last sex.

Lead investigator, Jean-Michel Molina from Saint-Louis Hospital Paris covered some of the outstanding issues about the IPERGAY dosing in an oral symposia session. This talk provided background information about PK evidence supporting event-based dosing together with new PK results. [18]

The results from IPERGAY need to be interpreted knowing that study participants were generally a sexually active cohort. The median number of pills taken per month in IPERGAY was 16 (IQR: 12 to 24 pills). Someone having sex once every week therefore more closely approximated someone taking four doses a week, rather than just relying on starting from zero drug levels using the dosing strategy in isolation. The iPrEX study has already shown that four doses a week achieves >95% protection in a similar high-risk group of gay men [6]. This means that the results from IPERGAY cannot (yet) be extrapolated to less frequent PrEP use.

Evidence used when planning IPERGAY, included macaque data using TDF/FTC with a 2-hour pre-dose and 24-hour post dose. Over 14 weeks, 5/6 animals were protected from weekly rectal exposure, with one infection occurring at week 5. By comparison, all 6/6 animals in the control group became infected. The same study showed that drug levels of tenofovir were detectable in plasma after two hours but not in rectal tissue until 24 hours. Levels of FTC were detectable at the 2-, 5- and 24-hour time points in both plasma and rectal secretions. With only these three time points, we do not know when tenofovir became detectable between the 5- and 24-hour time points. [19]

New data from IPERGAY was presented from a PK sub-study involving 12 participants who received a single double- dose (2 x TDF/FTC) with plasma, PBMC, dried blood spot, saliva and rectal tissue samples taken at 0, 0.5, 1, 2, 4, 8 and 24 hours (2 participants per time point). Similar to data from HIV positive people on ART, drug levels of FTC triphosphate were rapidly detected in rectal tissue after 30 minutes, but tenofovir diphosphate was still not detected after 8 hours and was only detectable at 24 hours.

As with the earlier macaque study, there is no information from between the 8- and 24-hour time points – something that clearly would have practical importance when recommending dosing strategies. It might also be important that levels following the single double-dose were significantly lower compared to control participants at steady state from daily dosing (approximately 5 vs 10 ng/mg).

Another point is that the double-dose approximately doubles the Cmax, AUC and Cmin drug levels of both tenofovir and FTC compared to historical single dose controls. But the double-dose has little if any impact on the Tmax – ie how quickly levels are reached. Although the double-dose intuitively implies a faster potency, it is unlikely to overcome the 24-hour time for tenofovir to achieve protective levels in rectal tissue. This is a critical parameter when recommending minimum timing for the pre-exposure dose.

When the pre-dose is taken less than 24 hours before sex, ie 2-8 hours before, it is difficult to know how much FTC alone contributes to PrEP. Macaque data supports partial protection from FTC alone [20] but the results from the Partners PrEP study reported no significant difference between TDF monotherapy and TDF/FTC dual therapy. [21]

In the PK subsidy of IPERGAY, a trend to partial protection at early time points from using the double-dose was reported from ex vivo rectal tissue using a quantitative infectivity score, but this had no control group using a single TDF/FTC dose for this indirect marker for protection.

Also, although it did not directly inform the IPERGAY strategy, an HIV negative study was also referenced to show that daily dosing achieves EC90 drug levels in rectal PBMCs associated with >90% protection (95%CI: 60% to 96%) after three days and >99% (95%CI 69% to 100%) after 5 days. After achieving steady state with daily dosing, levels remained high for seven days after stopping PrEP, but this would not be the case with less than daily dosing. [22]

This is important practical information for people using daily dosing, as clinical studies including PROUD have previously recommended daily dosing for 1 to 2 weeks before assuming protective drugs levels are achieved.

Jean-Michel Molina concluded his talk with an important clarification: “Clearly the effectiveness of the IPERGAY dosing strategy in people having frequent sex taking multiple pills cannot yet be extrapolated to people who have less frequent sex”. He also noted that he hoped that additional follow-up from the ongoing open label phase might provide additional data to address this issue.

Conclusions

PrEP is a fast changing field and many more studies at IAS 2015 were presented than could have been included in this report.

Numerous posters were notable for reporting little or no evidence that access to PrEP increased high-risk behaviour. These studies often noted that risk factors among the people using PrEP are already high and that resources might be more appropriately focused on increasing awareness of PrEP and supporting adherence (including perhaps with text messaging). [23, 24, 25, 26, 27, 28, 29, 30]

Several of these studies also looked at barriers to PrEP in different populations, including transgender people, who although included in some PrEP studies are still under represented. [31]

Comment

These studies show not only continued efficacy and safety but that flexibility for dosing is likely to be essential if PrEP is to become widely used.

While the NHS has been deciding on criteria for free access, off-label generic use has already become common in the UK. [32]

Doctors are therefore increasingly likely to be asked for advice about PrEP and about access to monitoring. This advice is likely to be different depending on whether cost is a factor that limits daily use.

The NHS clinic at 56 Dean Street should be commended for responding to demand by offering a PrEP service. Initial costs have already been reduced and services include an option of drug level testing to confirm that internet-sourced drugs are genuine. [33]

Off-label use is also currently destabilising HIV post-exposure prophylaxis (PEP) services in London. [34] Given the inconvenience of “clinic hopping” this could be reduced by better awareness that buying generic PrEP online for personal use is both legal in the UK and is dramatically cheaper than Gilead’s Truvada at £400 a month. Cipla’s Tenvir-EM costs approximately £40 to £70 for 30 tablets, depending on online supplier. [32]

Buying generic PrEP online makes it more affordable – approximating to condoms and lube or a few pints – even using daily dosing. As access to PrEP on the NHS, when it becomes available, is likely to be very limited while TDF/FTC remains patent protected, this model approximates that for access to sildenafil. This drug was not available on the NHS when it was first approved but the huge demand was managed by private prescriptions (for the licensed drug).

When daily generic dosing is not affordable, alternate day dosing to achieve four doses a week, is likely to produce similar (>95%) reductions in risk and enable a month supply to cover two months. This dosing strategy halves the generic costs.

Infrequent event-based dosing is supported by less data, but aiming for a 24-hour rather than 2-hour pre-dose, will enable tenofovir to be active in rectal tissue. Post-exposure doses are essential: daily dosing through the period of sexual activity, plus additional doses in the day or days afterwards. This should be included in information about PrEP, until new data shows otherwise.

It is unclear how much the double pre-dose of TDF/FTC in IPERGAY contributes to protection. Although FTC rapidly reaches rectal tissue it is difficult to estimate the level of protection this provides in absence of tenofovir.

Although the post-dose is essential, HPTN 067 and other studies have reported that this is more likely to be missed than the pre-dose. This should be a focus for community education and support, given that intuitively it should be easier to do something after an event, because the event itself should be the prompt, and it is more difficult to successfully plan for something that has yet to happen. Communication about the timing of the post dose seems to require further research.

HPTN 067 showed that in the context of good adherence (>85%), both time- and event-based dosing strategies provided similar coverage/protection to daily dosing, but with significantly fewer pills, therefore reducing both drug exposure and cost. But for people with lower (<65%) adherence, daily dosing is a significantly better strategy, probably due to greater flexibility to miss doses and still retain coverage. Although in HPTN 067 none of the HIV infections occurred in people with drug levels that supported recent adherence, it was not powered to show efficacy. In the context of lower adherence, any cost saving from using fewer drugs using time- or event-based dosing is likely to be cancelled out by the poorer efficacy.

References:

Unless stated otherwise, references are to the Programme and Abstracts of the 8th IAS Conference on HIV Pathogenesis, Treatment and Prevention (IAS 2015), 19 – 22 July 2015, Vancouver.

http://pag.ias2015.org

1. Hirnschall G. Panel discussion following START presentation. IAS 2015, Session MOSY03.
   https://www.youtube.com/watch?v=miM3umStJFE (Webcast)
2. UNAIDS. Oral pre-exposure prophylaxis: putting a new perspective. July 2015.
   http://www.unaids.org/sites/default/files/media_asset/UNAIDS_JC2764_en.pdf (PDF)
3. Venter F. Treatment as Prevention versus other biomedical prevention: contrasting implementation challenges. Plenary session MOPL0101. Monday 8.15-10.30 Ballroom C-D.
   http://pag.ias2015.org/PAGMaterial/Webcast/68_12486/webcast.mp4
4. Tsai C-C et al, Prevention of SIV Infection in Macaques by (R)-9-(2-Phosphonylmethoxypropyl)adenine. Science 1995.
5. Glidden D. How to start and stop PrEP. a pharmacology perspective (HPTN 067 and more). MOSY0109.
   http://pag.ias2015.org/PAGMaterial/Webcast/2018_12920/webcast.mp4
6. Grant R et al. Uptake of pre-exposure prophylaxis, sexual practices, and HIV incidence in men and transgender women who have sex with men: a cohort study. Lancet Infect Dis 2014; 14: 820–29.
   http://www.thelancet.com/journals/laninf/article/PIIS1473-3099(14)70847-3/abstract  (Abstract)
   http://www.natap.org/2014/HIV/309914708473.pdf (PDF)
7. Grant R et al. HPTN 067 ADAPT methods and results from women in Cape Town. Symposia session MOSY01. Monday 2.30-4.00 pm, Ballroom C-D. IAS 2015 19 – 22 July 2015, Vancouver.
   http://pag.ias2015.org/PAGMaterial/Webcast/2011_12555/webcast.mp4
8. Bekker L-G et al. HPTN 067/ADAPT Cape Town: A comparison of daily and nondaily PrEP dosing in African women. CROI 2015, 23-26 February, Seattle. Late breaker poster abstract 978LB.
   http://www.croiconference.org/sessions/hptn-067adapt-cape-town-comparison-daily-and-nondaily-prep-dosing-african-women  (Abstract)
   http://www.croiconference.org/sites/default/files/posters-2015/978LB.pdf (PDF)
9. Mannheimer S et al. HPTN 067 ADAPT results from MSM in Harlem. Symposia session MOSY01. Monday 2.30-4.00 pm. Ballroom C-D.
   http://pag.ias2015.org/PAGMaterial/Webcast/2013_11079/webcast.mp4 (Webcast)
10. Mannheimer S et al. HPTN067/ADAPT study: a comparison of daily and intermittent pre-exposure prophylaxis dosing for HIV prevention in men who have sex with men and transgender women in New York city. IAS 2015, 19 – 22 July 2015, Vancouver. Late breaker oral abstract MOAC0305LB.
11. Holtz T et al. HPTN 067 ADAPT results from MSM in Bangkok. Symposia session MOSY01. Monday 2.30-4.00 pm, Ballroom C-D.
    http://pag.ias2015.org/PAGMaterial/Webcast/2012_13093/webcast.mp4 (Webcast)
12. Holtz T et al. HPTN 067/ADAPT study: a comparison of daily and non-daily pre-exposure prophylaxis dosing in Thai men who have sex with men, Bangkok, Thailand. IAS 2015 19 – 22 July 2015, Vancouver. Late breaker oral abstract MOAC0306LB.
13. Amico R et al. PrEP experiences among South African women in the HPTN067 study: Healthyparanoia,Ubuntu champions and challenges to resolving PrEP dissonance. Symposia session MOSY0106.
    http://pag.ias2015.org/PAGMaterial/Webcast/2014_12870/webcast.mp4
14. Chemnasiri T et al. Patterns of Sex and PrEP in Bangkok MSM. Symposia session MOSY0107.
    http://pag.ias2015.org/PAGMaterial/Webcast/2015_11080/webcast.mp4
15. Franks J et al. Patterns of sex and PrEP in Harlem MSM: a qualitative study. Symposia session MOSY0108.
    http://pag.ias2015.org/PAGMaterial/Webcast/2017_11500/webcast.mp4
16. Molina JM et al. On demand PrEP with oral TDF-FTC in MSM: results of the ANRS Ipergay trial. 2015 Conference on Retroviruses and Opportunistic Infections, 23-26 February 2015, Seattle, Washington. Oral late breaker abstract 23LB.
    http://www.croiconference.org/sessions/demand-prep-oral-tdf-ftc-msm-results-anrs-ipergay-trial (Abstract)
    http://www.croiwebcasts.org/console/player/25540 (Webcast)
17. McCormack S et al on behalf of the PROUD Study. Pragmatic open-label randomised trial of preexposure prophylaxis: the PROUD study. Conference on Retroviruses and Opportunistic Infections, 23-26 February 2015, Seattle, Washington. Oral late breaker abstract 22LB.
    http://www.proud.mrc.ac.uk/pdf/PROUD_CROI%202015_Final.pdf
    http://www.croiconference.org/sessions/pragmatic-open-label-randomised-trial- preexposure-prophylaxis-proud-study (Abstract)
    http://www.croiwebcasts.org/console/player/25539(Webcast)
18. Molena JM et al. Updates on PrEP efficacy in Ipergay. Symposia session MOSY0102.
    http://pag.ias2015.org/PAGMaterial/Webcast/2010_12769/webcast.mp4
19. Garcia-Lerma JG et al. Intermittent prophylaxis with oral Truvada protects macaques from rectal SHIV infection. ScienceTransMed2010,Jan13;2:14ra4. doi: 10.1126/scitranslmed.3000391.
    http://stm.sciencemag.org/content/2/14/14ra4
    http://www.ncbi.nlm.nih.gov/pubmed/20371467
20. García-Lerma JG et al. Prevention of rectal SHIV transmission in macaques by daily or intermittent prophylaxis with emtricitabine and tenofovir. PLoS Med. 2008 Feb;5:e28. doi: 10.1371/journal.pmed.0050028.
    http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2225435
21. Baeten JM et al. Antiretroviral prophylaxis for HIV prevention in heterosexual men and women. NEnglJMed2012;367:399-410.
    http://www.nejm.org/doi/full/10.1056/NEJMoa1108524
22. Seifert S et al. ClinicalInfDis, March 2015:60;804-10.
    http://cid.oxfordjournals.org/content/60/5/804.abstract
23. Thomas R et al. PrEP in real-life settings: good adherence and no increase in high-risk behavior. Poster abstract TUPEC503.
24. Krakow DS et al. Patient-provider communication about sexual behaviors and pre-exposure prophylaxis: results from a national online survey of men who have sex with men in the United States. Poster abstract TUPEC506
25. Elsesser S et al. Absence of sexual behavioral disinhibition in a PrEP adherence trial: considerations for medical providers who prescribe PrEP for men who have sex with men. Poster abstract TUPEC507.
26. Mayer K et al. Significant increases in HIV pre-exposure prophylaxis (PrEP) uptake in Boston, a Boston Community Health Center in 2014: who are the recent users? Poster abstract TUPEC508.
27. Wanna A et al. Pre-exposure prophylaxis (PrEP) knowledge and use in a population-based sample of younger Black men who have sex with men (YBMSM) in Chicago. Poster abstract TUPEC509.
28. Corales R et al. TUPEC510 PrEP in the real world: implementation of PrEP. Poster abstract TUPEC510.
29. Blumenthal J et al.TUPEC513Text messaging responses correlate with tenofovir-diphosphate dried blood spot concentrations among men who have sex with men on pre-exposure prophylaxis. Poster abstract TUPEC513.
30. Milam J et al. Risk compensation among men who have sex with men in Southern California following the initiation of pre-exposure prophylaxis. Poster abstract TUPEC518.
31. Keatley J. PrEP for Transwoman: what are the opportunities and challenges. Symposia session MOSY0505.
32. i-Base Q&As on PrEP: How do I safely use PrEP if I buy it online? Can I check PrEP from the Internet is genuine? Can I get PrEP privately in the UK? etc
    https://i-base.info/qa/category/prep
33. Private PrEP clinicat 56 Dean Street.
    http://getprep.uk
34. Owens G. Get free P(r)EP on the NHS – NOW. Web blog (12 October 2015).
    https://gregowenblog.wordpress.com/2015/10/12/get-free-prep-on-the-nhs-now