Two cases of PrEP failure on solo tenofovir pose significant research questions

Gus Cairns
Published: 14 January 2016

A report originally presented to the 2015 BHIVA conference last year details two cases where therapeutic levels of solo tenofovir unequivocally failed to prevent HIV infection in gay men. In one case, despite the tenofovir apparently suppressing the man’s HIV viral load in his blood plasma, it failed to prevent HIV infecting the cells of his immune system. Solo tenofovir has been tested as PrEP in two major studies , Partners PrEP and the Bangkok Tenofovir Study, with moderate to good results, and it has been assumed on the basis of solo animal studies that tenofovir might be enough the prevent HIV, though results have been mixed.

No unequivocal case of the failure of Truvada (tenofovir + emtricitabine) has yet been reported, but the cases raise a number of interesting questions: whether the levels of tenofovir required to prevent infection need to be higher than those used for treatment; whether hepatitis B co-infection may have made HIV infection more likely; whether the men would have been infected if they had been taking tenofovir + emtricitabine (Truvada); and if not, what are the exact contributions to prevention of the two drugs.

The men were not taking tenofovir specifically as pre-exposure prophylaxis (PrEP) but instead were taking it as treatment for chronic hepatitis B infection. One had had persistent hepatitis B infection for six years and had been on tenofovir for four years; the other had had hepatitis B for seven years and had taken tenofovir for three years.

In the case of the first patient (patient A), the date of HIV infection can be pinpointed almost exactly as he tested HIV-antibody positive only twelve days after a confirmatory HIV western blot antibody test found him HIV negative. Although this is an unusually short time in which to develop HIV antibodies, it is not unknown. It was estimated that the most likely date for his actual exposure to HIV, based on his report of condomless receptive anal sex with a casual male partner, was one day after his HIV-negative test result. Six days later, he reported mild flu-like symptoms suggestive of HIV seroconversion illness and was tested again.

The second patient (patient B) had not had a recent negative HIV test but was hospitalised with a severe flu-like illness with fatigue and muscle pain, suggestive of HIV seroconversion. Based on his first positive HIV test, where he was HIV antibody-negative but p24-positive, and also on his account of condomless receptive anal sex, it was estimated that his likely time of infection was about two weeks previous to the test.

Both men appeared to have excellent adherence to tenofovir based on pill counts. More to the point, tenofovir drug levels were taken on the day they tested HIV positive. Patient A had last taken tenofovir 24 hours before his drug level test and was therefore at the ‘trough’ or lowest blood level one would expect. His level of tenofovir was about 75 nanograms per millilitre of blood (ng/ml), meaning that he had lower trough levels than about 80% of patients, but still well within the therapeutic level to treat both hepatitis B and HIV. (The IC95 of tenofovir, which is the amount of drug in the blood sufficient to reduce HIV replication by 95%, is 6.8 to 35 ng/ml.)

Patient B had taken tenofovir seven hours before his HIV test result and had a blood level of 230 ng/ml, or higher than 75% of the average patients, seven hours after a dose. Both results indicate that their absorption of tenofovir was normal.

Both patients had CD4 cells counts of 550-600 cells/mm3 – notably lower than typical CD4 counts seen in HIV-negative people – and while patient A had a relatively normal CD4:CD8 ratio of 1.16, patient B had the ‘inverted’ ratio typically seen in people with HIV – 0.49. [HIV-negative people typically have more CD4 than CD8 cells: in people with HIV this ratio is usually inverted immediately after acute infection and stays that way without treatment.] Either way, these figures show that despite the tenofovir, both patients had suffered significant immune damage during acute HIV infection – as usually happens.

The biggest difference between the two patients was that in patient A the tenofovir, despite not preventing HIV infection, did seem to be suppressing his HIV viral load in the blood. Although testing HIV-positive and having HIV integrated into his cell (see below) at no point did he have a viral load over 50 copies/ml. This ‘blunting’ of the HIV viral load has been seen before in cases of PrEP failure, notably in the animal studies that established its efficacy. This means his HIV could not be tested to see if it had acquired, or already had, drug resistance to tenofovir. This patient was switched to Eviplera (rilpivirine/tenofovir/emtricitabine) as soon as he tested HIV positive.

Patient B had a viral load of just over 100,000 copies/ml – not especially high for acute or recent infection. Despite having HIV actively reproducing in the presence of high tenofovir levels, he had no drug resistance mutations, illustrating, as other studies have done, that resistance to tenofovir only rarely develops in cases where people with HIV take PrEP. However to forestall resistance, patient B was put on to an intensified three-class antiretroviral regimen of Truvada (tenofovir/emtricitabine), boosted darunavir and raltegravir, which resulted in successful suppression of his viral load.

Both patients had significant intracellular HIV infection, and HIV integrated into the genetic material of their immune cells – 587 copies per million CD4 cells in the age of patient A and 1432 copies in patient B. This indicates quite enough integrated HIV for an ongoing, productive infection in both patients and indeed transcribed RNA, evidence that cells were actively producing new HIV viral particles, was found in both patients, though at one-tenth the level in patients A as patient B.

These cases may be the first ones where it can be shown without doubt that a daily dose of tenofovir has failed to prevent HIV infection. In one of the largest randomised studies of PrEP, Partners PrEP, there were six cases of HIV infection (Donnell) in people who, at the visit they were diagnosed with HIV, had tenofovir levels consistent with daily dosing (two of these were also taking emtricitabine). However in all cases but one, the participants could have caught HIV at any time in the previous three months since their last visit, so may not have been taking PrEP at the actual time they were infected. In one case, the participant had a level of tenofovir consistent with daily dosing one month before being diagnosed with HIV and at the visit she was diagnosed: this is the only previous case where the failure of PrEP to protect against infection despite what should have been protective levels looks like the most likely interpretation of the data.

The researchers comment that their cases show that “tenofovir monotherapy PrEP in men who have sex with men has limited efficacy data and that HIV acquisition can occur in the presence of tenofovir drug levels within the therapeutic range required to treat HIV.” They point out that it has never really been established whether the level of tenofovir (or any other single drug) sufficient to treat HIV is sufficient to prevent it. They also hypothesise that hepatitis B infection could increase susceptibility to HIV, even where it appears largely suppressed virologically. It may also cast further doubt over the use of solo tenofovir as PrEP in any patient, and should spur further research into even more protective PrEP regimens, and may add to nervousness about recommending some intermittent PrEP regimens.

Nonetheless, as the researchers note, the overwhelming majority of cases of so-called ‘PrEP failure’ are due to people not actually taking PrEP: the fact that these two cases have been reported underlines that HIV infection in situations where it looks like people have been taking PrEP consistently is extremely rare.


Fox J et al. Tenofovir disoproxil fumarate fails to prevent HIV acquisition or the establishment of a viral reservoir: two case reports. Journal of Infectious Disease and Therapy, early online publication. DOI 10.1007/s40121-015-0102-x. See abstract here. 2016.

Fox J et al. Pre-exposure prophylaxis fails to prevent HIV-1 infection or the establishment of a significant viral reservoir. 21st annual BHIVA conference, Brighton. Abstract P10. 2015. See for abstracts.

Donnell D et al. HIV Protective Efficacy and Correlates of Tenofovir Blood Concentrations in a Clinical Trial of PrEP for HIV Prevention. JAIDS 66(3): 340–348. 2014.

Community Consensus Statement on Access to HIV Treatment and its Use for Prevention

Together, we can make it happen

We can end HIV soon if people have equal access to HIV drugs as treatment and as PrEP, and have free choice over whether to take them.

Launched today, the Community Consensus Statement is a basic set of principles aimed at making sure that happens.

The Community Consensus Statement is a joint initiative of AVAC, EATG, MSMGF, GNP+, HIV i-Base, the International HIV/AIDS Alliance, ITPC and NAM/aidsmap