No fewer than 11 posters and two oral presentations at the Eighth Glasgow Congress on Drug Therapy in HIV Infection dealt with the issue of renal (kidney) toxicity in patients taking tenofovir (TDF, Viread), a measure of the concern that has been raised about this issue.
The bottom-line message from the studies is that kidney malfunction does seem to occur at a slightly higher rate on tenofovir than on other drugs. However it only seems to affect a minority (no more than 4%) of patients, many of whom have histories of mild renal impairment, and is usually reversible.
However the occurrence of two documented deaths due to kidney failure in the studies underlined the necessity of continuing to monitor for this problem.
One of the other things that stood out was that many patients with HIV are on other drugs that may cause kidney toxicity and that no study has been done to see if patients taking tenofovir are for one reason or another more likely to take them.
Interestingly, in a couple of the studies, including ones by tenofovir's manufacturer Gilead, there were hints that we have possibly been worrying about the wrong toxicity with tenofovir and that a gradual loss of bone mineral density in patients on the drug may come to have more significance in the long term.
There is only room here for a brief summary of all the presentations.
In poster 147 (Tordato) 316 Italian patients with HIV (77% on HAART, of which 31% were taking TDF) were monitored for their Glomerular Filtration Rate (GFR), a measurement of how efficiently kidneys filter the blood. A GFR of below 90 ml/minute indicates mild, below 60 indicates moderate, and below 30 severe, renal impairment. Below 15 indicates kidney failure.
The study found no association between HAART and a decline in GFR. It did find associations with sex (women being nearly three times more likely to have a reduced GFR), age, with a doubling of the rate for every 10 years’ increase in age, and having a higher CD4 rise above baseline, though this was of borderline significance. It did not separate out TDF recipients from other patients taking HAART, however.
Poster 148 from Portugal (Rocha) looked at 219 patients taking TDF that had normal GFR at baseline. It found a significant decline in GFR of 12.7% during the first year on TDF and 37% of patients developed a GFR of below 90. Two (0.9%) developed GFRs between 30 and 60 and two below 30. Age was again a significant risk factor.
Poster 149 (Davies) looked specifically at patients referred to a specialist renal clinic for TDF-related toxicity.
Altogether 20 patients were referred, 19 with protein in the urine which is regarded as a sign of possible renal failure. Eleven had renal impairment according to the standard measurements of GFR and raised levels of creatinine, a waste substance normally excreted by the kidneys. (Normal creatinine levels vary between 70 and 120 mmols/litre, with lower levels for women.)
Seven out of 14 that had complete blood chemistry measurements satisfied the diagnostic criteria for Fanconi’s syndrome, a sign of kidney failure.
The patients had spent an average time of 2.3 years on TDF, though one had only been on TDF for two months. The patients had their creatinine and GFR levels measured before TDF therapy, during and after drug withdrawal and the good news is that the renal toxicity appeared reversible. Average creatinine went from 86 mmols/litre before TDF to 164 during TDF and then went back to 109 mmols/l after withdrawal of TDF. GFR did not completely normalise: it was 92 ml/minute before TDF, 54 during and 70 after withdrawal.
One other finding was that levels of alkaline phosphatase (ALP) also went up on TDF. Raised levels of this enzyme may indicate liver disease or bone mineral loss as well as kidney malfunction. A normal value is between 36 and 126 International Units per litre. In these patients it went from 81 IU/l before TDF, to 178 during and 114 afterwards.
Poster 150 (Smith) compared the risks of certain toxicities between patients taking TDF and ones taking the alternative NRTI drug abacavir (ABC). A total of 868 patients attending the HIV clinics at two hospitals in London and one in Brighton, 610 of them on TDF and 358 on ABC, were compared.
Patients had a somewhat higher rate of virological failure on abacavir and conversely a somewhat higher rate of CD4 increase, but neither difference was significant.
Patients on abacavir did have a 130% higher rate of raised cholesterol. However patients on tenofovir had a 110% higher rate of raised ALP, a 49% higher rate of sub-normal GFR, a 61% higher rate of raised creatinine and a 66% higher rate of raised liver enzymes. These differences were all statistically significant. Age and a lower absolute CD4 count were also risk factors for a below-normal GFR.
Poster 151 from Spain (Gutierrez) looked at 249 patients taking TDF and ddI and compared ones that switched from this regimen to ones that continued taking it. There was no significant change in the average GFR of these patients over 72 weeks (from 97.8 ml/minute at baseline to 92.9 at 72 weeks and no difference, in the majority of patients, between ones that stayed on TDF/ddI and ones that changed.
However five patients did stop taking TDF/ddI due to ‘renal events’. All had had a subnormal GFR at baseline (from 51-89 ml/min) and this decreased considerably during TDF/ddI therapy, with rates of 27-53 ml/min at the time of TDF/ddI withdrawal. In four patients GFR went back to relatively normal – but one patient died of kidney failure.
Poster 152 (Cassetti) was Gilead’s own look at renal toxicity and bone mineral loss in patients who were on their 903 Study. This study randomised 172 drug-naïve patients either to TDF/3TC/efavirenz plus a d4T placebo or to d4T/3TC/efavirenz plus a TDF placebo. It kept the d4T patients on the drug for two years and then went open-label and gave all patients TDF for another three.
Poster 152 only looked at the patients who’d been on TDF throughout: however it could be read in conjunction with poster 120 (Madruga) which looked at bone mineral changes in the patients that switched from d4T.
The average GFR rate in the 903 Study patients remained unchanged over five years. What was more interesting, and significant, however, were the changes in bone mineral density. In patients that were on TDF throughout, the Z score (a measure of bone mineral density or BMD) declined by 1.5 in the spine and 2.7 in the hip. BMD immediately started to decline and reached a plateau at the end of the first year after which it did not decline much further. The decline in both spine and hip was statistically significant.
In patients that switched from d4T, the BMD Z score went down by 0.6 in the spine and 1.0 in the hip on d4T, which was not a significant decrease. After the switch to TDF it declined further in the spine to a total decrease of 1.1 which was still not significant. However BMD in the hip decreased to a total decrease of 3.3, more than the patients who had been on TDF throughout and statistically significant, and with no evidence so far of a plateau effect.
Poster 153 (Wolf) had an odd design comparing baseline and follow-up measurements of creatinine and creatinine clearance (a measure similar to GFR of how fast the kidneys filter waste) in 500 patients taking TDF, and comparing them to a single cross-sectional snapshot of a control group of 100 patients on non-TDF-containing HAART.
It found significantly lower creatinine clearance in patients who had been taking TDF but no significant difference in the number of patients with mildly raised creatinine.
What it did find, in common with other studies, was a minority of patients with significant problems. Eighteen patients (4%) of patients on TDF had moderately raised creatinine (grade 2) compared to zero taking non-TDF regimens and 1% of the TDF patients before they started TDF. And it found three patients that had severely (grade 3) raised creatinine on TDF.
In terms of the rises as opposed to the absolute levels of creatinine, it found 10 patients (2%) with mild grade 1 creatinine rises, four (0.8%) with moderate grade 2 rises – and three (0.6%) with very severe (grade 4) rises, indicative of renal failure. Ten out of these 17 patients had mild renal impairment at baseline.
However the authors also found that patients on TDF were very much more likely to be taking other potentially nephrotoxic (kidney-damaging) drugs – with no fewer than 70% of patients on TDF also taking something else that could damage kidneys compared with 24% at baseline and 12% on non-TDF HAART. “The relationship with TDF could not be excluded in these patients,” they say, “but other cofactors were likely to contribute to renal failure.”
Poster 154 (Madeddu) looked at the SCOLTA cohort in Italy. SCOLTA is a nationwide pharmacovigilance programme which monitors adverse events in patients on seven ARV drugs including tenofovir. The TDF cohort numbered 754 patients and creatinine figures were available for 354. It found cases of significantly raised creatinine in only nine of these patients (2.5%) and calculated an annual incidence of raised creatinine of 1.6%, with a 0.5% incidence rate for grade 3 and above. The nine patients who developed raised creatinine were all male and all but one of them were former injecting drug users co-infected with hepatitis C. Seven had cirrhosis.
Poster 155 (Fernando) was a retrospective look at 93 Edinburgh patients who had been on TDF for over a year. It looked at their creatinine clearance rates from one month to five years after the start of TDF therapy and found a statistically significant reduction in mean creatinine clearance by 2.83ml/minute after a year on TDF. Interestingly although creatinine clearance got worse in white patients, it actually significantly improved in African patients, though there were not enough of them to do a thorough analysis of why this might be so.
Poster 156 (Rosenvinge) looked retrospectively at 90 patients at St George’s Hospital in London on TDF and compared them with 97 not on TDF. This also found that being African appeared to be protective against kidney damage. It found that creatinine clearance improved by 10.2 mls/min in non-TDF-taking patients after a year on therapy and declined in TDF-taking patients by 1.8 mls/minute. When the patient group was split into white and African patients (regardless of TDF) it was found that there was a non-significant rise in creatinine clearance of 0.5 ml/min in white patients but a significant rise of 7.5ml/min in black African ones.
There were three cases of acute renal failure and one of Fanconi’s syndrome out of 90 TDF patients - and one of the three patients with renal failure died. Two had had a previous history of kidney stones.
Poster 157 (Staszewski) compared creatinine changes in 87 patients taking TDF and 92 not taking it who either had a history of mild renal impairment (57 patients) or who had high blood pressure and/or diabetes (122 patients), as these are also risk factors for kidney failure. It found cases of raised creatinine in patients with mild renal impairment and none in patients with the other risk factors, but no difference between patients on TDF and ones who were not.
Finally, oral presentation 13.2 (Harris) was a retrospective cohort study of raised creatinine in 1182 patients taking TDF in British Columbia. It found that 7.5% of patients had at least one measure of a 30% or greater rise in creatinine above baseline, which Harris said was equivalent to a 20ml/min or greater decrease in GFR, and 4.2% of patients with two or more such measurements. Amongst the cohort 20% discontinued TDF-containing regimens but only 3.2% discontinued TDF alone while keeping their other drugs, an indicator for discontinuation due to toxicity.
References (in order of citation)
Tordato F et al. Evaluation of renal function in 316 HIV-infected patients. Eighth Congress on Drug Therapy in HIV Infection, Glasgow. Abstract P 147. 2006.
Rocha S et al. Tenofovir-associated nephrotoxicity in the first year of therapy. Eighth Congress on Drug Therapy in HIV Infection, Glasgow. Abstract P 148. 2006.
Davies C et al. Tenofovir-associated renal toxicity. Eighth Congress on Drug Therapy in HIV Infection, Glasgow. Abstract P 149. 2006.
Smith C et al. Comparison of efficacy and toxicity profile of abacavir (ABA) and tenofovir (TDF) containing regimens. Eighth Congress on Drug Therapy in HIV Infection, Glasgow. Abstract P 150. 2006.
Gutierrez F et al. Renal safety profile of tenofovir plus didanosine containing regimens. Results from the DIDITEN cohort. Eighth Congress on Drug Therapy in HIV Infection, Glasgow. Abstract P 151. 2006.
Cassetti I et al. The safety and efficacy of tenofovir DF (TDF) in combination with lamivudine (3TC) and efavirenz (EFV) through five years in antiretroviral-naïve patients. Eighth Congress on Drug Therapy in HIV Infection, Glasgow. Abstract P 152. 2006.
Madruga R et al. Switch from stavudine (d4T) to tenofovir DF (TDF) in combination with lamivudine (3TC) and efavirenz (EFV) resulted in continuing virological suppression and improvement in lipoatrophy though 2 years in HIV-infected patients. Eighth Congress on Drug Therapy in HIV Infection, Glasgow. Abstract P 120. 2006.
Wolf F et al. Tenofovir (TDF)- versus non-TDF-containing ART: renal toxicity in daily practice. Eighth Congress on Drug Therapy in HIV Infection, Glasgow. Abstract P 153. 2006.
Madeddu G et al. Nephrotoxicity in HIV patients receiving combination antiretroviral therapy including tenofovir: results from the SCOLTA Project. Eighth Congress on Drug Therapy in HIV Infection, Glasgow. Abstract P 154. 2006.
Fernando I, Clutterbuck D. Changes in creatinine clearance over time in a patient cohort on tenofovir as part of combined antiretroviral therapy. Eighth Congress on Drug Therapy in HIV Infection, Glasgow. Abstract P 155. 2006.
Rosenvinge M M et al. Renal function in patients taking tenofovir and non-tenofovir based antiretroviral regimens. Eighth Congress on Drug Therapy in HIV Infection, Glasgow. Abstract P 156. 2006.
Staszewski S et al. Renal safety profile of tenofovir DF (TDF)-containing compared to non-TDF-containing regimens in antiretroviral-naïve patients with mild renal impairment or hypertension and/or diabetes mellitus. Eighth Congress on Drug Therapy in HIV Infection, Glasgow. Abstract P 157. 2006.
Harris M et al. Predictors of creatinine (Cr) increase and drug discontinuation in patients receiving tenofovir DF (TDF). Eighth Congress on Drug Therapy in HIV Infection, Glasgow. Abstract PL 13.2. 2006.