Participants taking tenofovir/emtricitabine (Truvada) for pre-exposure prophylaxis (PrEP) in two major studies experienced modest declines in kidney function that were associated with higher tenofovir drug levels and older age, according to studies presented at the Conference on Retroviruses and Opportunistic Infections (CROI 2016) in Boston. Together, these findings indicate that while Truvada PrEP is safe for most people, ongoing kidney function monitoring is important to promptly catch any problems that may occur.
Tenofovir disoproxil (Viread, also in the Atripla, Eviplera or Complera, and Stribild single-tablet regimens) has been used in HIV treatment for more than a decade and is generally considered safe and well-tolerated, but it can cause kidney damage and bone loss in some people.
For this reason, regular kidney function tests are recommended for people taking Truvada for PrEP and those with pre-existing kidney problems are advised not to use it. The US product label for Truvada states that it should not be used for PrEP in people with creatinine clearance below 60 ml/min, and if a decrease is observed providers should “evaluate potential causes and re-assess potential risks and benefits of continued use.”
Reports to date from PrEP clinical trials – which tried to exclude people with known kidney impairment – have not revealed notable kidney problems. But sometimes uncommon toxicities only show up after many more people are using a drug.
Monica Gandhi from the University of California, San Francisco presented follow-up findings on kidney function in the pivotal iPrEx trial, the first to show that once-daily Truvada PrEP was highly effective for gay and bisexual men and transgender women who took it consistently. In an open-label extension (iPrEx OLE), in which everyone knew they were taking active Truvada, no one who took the pills at least four times a week became infected with HIV.
The iPrEx team looked at the relationship between tenofovir and emtricitabine concentrations in hair samples – which reflect cumulative drug exposure – and kidney function over time in iPrEx OLE. Hair samples were collected, serum creatinine (an indicator of kidney function) was measured and estimated glomerular filtration rate (eGFR, an indicator of creatinine clearance) was calculated every 12 weeks.
Of the more than 1000 iPrEx OLE participants, hair drug levels and creatinine measurements were available for 220 people. A majority of participants in this international study were Latino or mixed race/ethnicity and the median age was 29 years (range 19-70). At baseline, the mean creatinine level was 0.89 mg/dl (0.6 to 1.3 is considered normal) and the median estimated eGFR was 112 ml/min (>90 is normal, 60 is a moderate decrease and <30 is a severe decrease).
While creatinine clearance for participants on Truvada decreased slightly overall – by 2.5% over 18 months – there was a consistent and statistically significant relationship between percentage eGFR decline and increasing levels of tenofovir or emtricitabine in hair samples. Among people in the highest quartile of hair drug levels, indicating they took all seven doses a week, the eGFR decrease was 5.6%.
Older participants were more likely to see their eGFR fall below 70 ml/min, considered a clinically significant decrease. Among people older than 50, the proportion with eGFR below this threshold rose from 6% in the lowest hair level quartile (indicating about two doses a week) to 24% in the highest quartile.
Among participants under age 40, however, the proportion never rose above 5% even with daily dosing. People with slightly low eGFR (<90ml/min) at baseline were also more likely to fall below 70 ml/min, with 27% doing so.
Establishing thresholds of tenofovir/emtricitabine exposure that minimise adverse events while maximizing efficacy “are essential to the real-world roll-out of PrEP,” the researchers wrote.
Albert Liu of the San Francisco Department of Public Health then presented findings on changes in kidney function among participants in the US Demo Project.The Demo Project enrolled more than 500 gay and bisexual men and transgender women at risk for HIV at sexual health clinics and community health centers in San Francisco, Miami and Washington, DC. All received once-daily Truvada PrEP on an open-label basis for a year.
Nearly half of participants were white, about 7% were black, the median age was 35 years (range 18-65) and 12% had hypertension or diabetes. To be eligible, participants had to have baseline creatinine clearance of at least 60 ml/min, most had at least 70 ml/min and the median was 97 ml/min.
In this study, tenofovir diphosphate (the metabolised form of the drug in the body) was measured in dried blood spots from a subset of participants, and kidney function was again monitored every 12 weeks; 82% of participants had tenofovir diphosphate levels consistent with taking at least four Truvada doses a week.
Overall, eGFR declined by an average of 2.8% from baseline to week 12 – similar to the proportion in iPrEx OLE – and then remained stable through the end of the study at week 48. Only a small number (11 people or 2.4%) had their creatinine clearance fall by 10% or more and their serum creatinine rise by 0.2 mg/dl or more on more than two occasions. Three people did have their Truvada suspended due to an elevated creatinine reading, but these were not confirmed with repeat testing and all restarted PrEP with no further interruptions.
Higher tenofovir diphosphate levels in blood spots were associated with significantly greater declines in eGFR. Again, kidney function decline was associated with older age and a lower baseline level. Among people with baseline eGFR
Finally, Kenneth Mugwanya of the University of Washington in Seattle presented findings on the occurrence of more serious kidney problems among participants in the Partners PrEP trial, which tested once-daily Truvada or tenofovir alone as PrEP for HIV-negative partners in serodiscordant heterosexual couples in Africa.
This study looked at proximal tubulopathy, which occurs when proximal tubule cells in the kidneys are unable to reabsorb substances such as glucose, salts, minerals and bicarbonate. Sometimes this is sub-clinical and detected by protein or glucose in the urine, but more serious conditions such as Fanconi syndrome can lead to disruption of the body’s pH balance and bone loss. Drugs like tenofovir that are excreted by the kidneys can build up and cause tubule damage, usually in people who have a genetic predisposition or other risk factors. Mugwanya noted that proximal tubulopathy can sometimes occur without an accompanying severe decline in GFR.
This analysis included 776 people randomised to Truvada and 773 randomly assigned to a placebo. Two-thirds were men and the median age was 37 years. Proximal tubulopathy was defined as having any two of the following: protein in the urine, glucose in the urine, increased urinary phosphate or uric acid excretion. The frequency of proximal tubulopathy was 1.7% (13 people) in the Truvada arm compared with 1.3% (10 people) in the placebo arm, not a statistically significant difference. Protein in the urine alone was seen in 7.3 vs 4.0%, respectively, and this was statistically significant.
A related case-control analysis revealed that two of the 52 people (3.8%) who experienced at least a 25% decline in eGFR developed tubulopathy, compared with three of 208 people (1.4%) who did not, again not a statistically significant difference. There was “no significant association between tubulopathy and clinically relevant decline in eGFR,” the researchers concluded, suggesting that “using routine markers of proximal dysfunction will not be an efficient approach to predict the rare events of kidney toxicity in persons using PrEP.”
Taken together, these studies indicate that Truvada is safe and well-tolerated for most people using it for PrEP, but kidney problems – usually mild or moderate – may occur in a small proportion of people, especially if they have other risk factors.
“It is often said that these declines are so low it doesn’t matter, but PrEP studies have not monitored people as long as treatment studies,” Gandhi said at a CROI press conference. She noted that even minor toxicities can be a concern in a healthy population.
These findings support the US Centres for Disease Control and Prevention recommendation that creatinine clearance should be monitored at least every six months, with more frequent monitoring and additional tests for people with other kidney safety risk factors such as high blood pressure or diabetes.
Gandhi and Liu agreed that monitoring more often may be appropriate for older people and those who start PrEP with an eGFR
PrEP is proven to reduce the risk of HIV acquisition. Drug levels (as markers of adherence) have been critical to interpreting disparate outcomes in PrEP trials, but can also be assessed (as markers of exposure) in relationship to adverse effects. Concentrations of tenofovir (TFV) and emtricitabine (FTC) in hair represent cumulative exposure and may be associated with toxicities in HIV-uninfected persons. We report for the first time, in a large PrEP demonstration study, the relationship between TFV/FTC levels in hair and renal function over time.
The iPrEx Open Label Extension (OLE) study enrolled HIV-negative MSM and transwomen and all were on PrEP. Hair samples were collected every 12 weeks and levels of TFV/ FTC measured via liquid chromatography/tandem mass spectrometry. Serum creatinine (Cr) was measured every 12 weeks and glomerular renal function (eGFR) estimated by Cockcroft-Gault (CG) or the MDRD equation. The association between change in eGFR over time (adjusted for baseline eGFR) and TFV/FTC levels (categorized into quartiles) was analyzed by generalized estimating equations.
Hair data and creatinine measures were available for 1144 person-visits in 202 participants followed for a median of 16.8 months. Median age 29 years (19-70); 91% MSM; 22% White, 11% Black, 6% Asian, 60% Latino/mixed. Baseline mean Cr level was 0.89mg/dL with a median baseline eGFR of 112mL/min (99-128). The eGFR for all participants on TFV/FTC decreased over 18 months, but there was a monotonic relationship between % decrease in eGFR with increasing quartile of hair level for TFV (p 0.008) and FTC (p 0.006) (Figure). For instance, mean % change in eGFR from baseline was -2.6ml/min (SE 0.8) in person-visits with TFV levels in the 1st quartile, but -5.6 (SE 0.7) when hair levels were in the 4th quartile. The odds of eGFR falling below 70 ml/min (6.1% of sample) increased with increasing quartile of TFV/FTC concentration (OR 4.4 (1.1-17.4) for 4th TFV hair quartile, p trend 0.045; OR 4.0 (0.9-17.2) for 4th FTC quartile, p trend 0.027).
We show for the first time that greater long-term exposure to TFV or FTC in patients on PrEP is associated with declining renal function over time. Hair levels of TFV/FTC were associated with decreases in eGFR and a higher likelihood of eGFR falling to <70mL/min in a monotonic fashion in iPrEX OLE. Establishing thresholds of TFV/FTC exposure that protect from HIV, but minimize the risk of toxicity, is essential to the real-world roll-out of PrEP.
Several trials have demonstrated the safety and efficacy of TDF/FTC pre-exposure prophylaxis (PrEP). Renal toxicity was uncommon in randomized trials of healthy individuals, but has not been assessed in clinical settings. We evaluated changes in renal function among participants enrolled in the open-label US PrEP Demonstration Project. The Demo Project enrolled HIV-negative MSM and transwomen (TGW) in STI clinics and a community health center. Eligible participants [creatinine clearance (CrCl) ≥60 ml/min] were offered 48 weeks of TDF/FTC PrEP. Creatinine (Cr) was measured every 12 weeks and CrCl estimated by the Cockcroft Gault equation. Tenofovir diphosphate (TFV-DP) levels in dried blood spots (DBS) were measured in a subset of subjects. The associations of time-dependent factors with visit-to-visit changes in CrCl were assessed using linear mixed models. From October 2012 to January 2014, 557 MSM and TGW enrolled. Median age was 35 (range 18-65); 48% were White, 35% Latino, 7% Black, 5% Asian, and 6% other; 98% were MSM. Baseline median Cr was 0.92 (0.59-1.55), with a median CrCl of 124 ml/min (71-309). DBS were tested at 1,067 person-visits among 294 participants, with TFV-DP levels consistent with ≥4 doses/week in 82% of person-visits. Median CrCl declined 6 ml/min (5%) from baseline to week 12 and remained stable through week 48 (p=0.96), with no differences by race/ethnicity, weight, or NSAID use. However, 30% had >10% decline in CrCl at week 12. TFV-DP levels ≥2 vs. 2 doses/week. Younger PrEP users and those taking medications for hypertension or diabetes had greater decreases in CrCl and may warrant additional monitoring during PrEP use.
Tenofovir disoproxil fumarate (TDF) is infrequently associated with proximal tubular dysfunction in HIV-infected persons when used as part of combination antiretroviral therapy, but limited data are available for HIV-uninfected persons on TDF for pre-exposure prophylaxis (PrEP).
Data are from the Partners PrEP study, a randomized trial of daily oral TDF and emtricitabine (FTC)-TDF PrEP among African HIV-uninfected men and women (ClinicalTrials.gov:NCT00557245). We conducted: 1) a cohort analysis to determine whether FTC-TDF PrEP causes proximal tubular dysfunction among HIV-uninfected persons randomized to FTC-TDF versus placebo, and 2) a nested case-control analysis of persons on TDF or FTC-TDF to determine whether tubular dysfunction predicts subsequent clinically relevant decline in estimated glomerular filtration rate (eGFR). The primary outcome was subclinical proximal tubulopathy (PT), pre-defined as any two of the following markers of tubular dysfunction: tubular proteinuria, euglycemic glycosuria, increased urinary phosphate excretion, or increased urinary uric acid excretion. PT was assessed in concurrently obtained urine and serum samples at the 24-month visit or last on-treatment visit. For the nested case-control analysis, cases were persons on TDF or FTC-TDF with confirmed ≥25% eGFR decline from baseline and controls were persons with similar drug exposure without the ≥25% eGFR decline.
Of 1549 persons included in the cohort (776 on FTC-TDF, and 773 on placebo), 64% were male. Median age was 37 years (range 18-64) and median duration of study drug exposure was 24 months (range 3-27). The frequency of PT was 1.7% in FTC-TDF versus 1.3% in the placebo arm [odds ratio 95% confidence interval: 1.30 (0.52, 3.33); p=0.68]. PT occurred in 2 of 52 (3.8%) persons who experienced ≥25% eGFR decline versus 3 of 208 (1.4%) controls (adjusted odds ratio, 95% confidence interval: 1.40 (0.10, 14.1); p >0.99]. One person on FTC-TDF and potentially nephrotoxic co-medications developed features indicative of Fanconi syndrome.
In this large placebo-controlled study, proximal tubular dysfunction was rare and was not significantly associated with daily oral FTC-TDF PrEP over up to 24 months of observation, nor did it predict a subsequent clinically relevant decline in eGFR. These findings support the safety of TDF-based PrEP as a component of HIV prevention in healthy HIV-uninfected individuals.