To the Editor,
Parsonage et al.  have described two patients with osteomalacia, a rare complication of tenofovir disoproxil fumarate (TDF) therapy. Both cases also had signs of diminished renal function with proximal renal tubular dysfunction (PRTD), a previously reported and usually reversible complication of TDF treatment . As TDF is renally eliminated, dose adjustment is required in chronic kidney disease (CKD) to avoid accumulation that may lead to further renal injury. We believe these two cases are instructive regarding safer use of TDF and wish to expand on the discussion offered by the authors.
Both patients were described as having normal renal function prior to beginning TDF, based on serum creatinine values within the normal range. However, serum creatinine is an imprecise measure of renal function, and may remain within normal limits in the presence of mild or moderate CKD . This is especially true in older patients and those with chronic illnesses, including AIDS . Such patients have higher rates of adverse events to drugs that are renally eliminated . Two equations are commonly used to obtain a more reliable estimate of renal function. The Cockcroft–Gault formula has been used to estimate creatinine clearance for purposes of dose modification of medication in CKD. The Modification of Diet in Renal Disease (MDRD) equation  estimates glomerular filtration rate (GFR) corrected for body surface area and is recommended by the (US) National Kidney Foundation as the preferred tool for staging CKD. Either method is more accurate than serum creatinine alone . Although neither formula has been validated in HIV-infected patients, published guidelines on management of chronic kidney disease recommend all HIV-infected patients should have renal function estimated using one of these methods as part of initial medical evaluation and prior to use of any medications that are renally eliminated, including antiretrovirals .
We could not estimate creatinine clearance from the data provided by Parsonage et al. because weights were not reported. Using the simplified MDRD equation , patient 2 had a mildly reduced GFR of 74 mL/min/1.73 m2 prior to initiating TDF, assuming that this individual, who was from Zambia, was black. Approximating serum creatinine from Fig. 2, GFR declined to 52 mL/min/1.73 m2 by 6 months and to 40 mL/min/1.73 m2 by 9 months. Despite this 46% decline in GFR, full dose TDF was continued for 11 months. Assuming that patient 1 was white, the calculated GFR was normal at 125 mL/min/1.73 m2 prior to TDF therapy. This patient had advanced HIV disease (AIDS diagnosis for 13 years with Kaposi's sarcoma) and hypogonadism. Both conditions would reduce muscle mass and cause overestimation of GFR based on serum creatinine. Approximating serum creatinine from Fig. 1, the calculated GFR declined 47% to 66 mL/min/1.73 m2 at 22 months, and signs of PRTD were present by month 17. In this case, full dose TDF was continued for 25.5 months. Both were also on lopinavir/ritonavir (LPV/r), which increases exposure to tenofovir by 32% . The mechanism of this interaction is not known, but it is unlikely to be attributable to ritonavir (RTV), as suggested by the authors, as pharmacokinetic interaction studies performed with TDF and RTV coadministered with certain other HIV protease inhibitors have not demonstrated this interaction [10,11].
We believe that these two patients developed unusual toxicities (PRTD and osteomalacia) because full doses of TDF were continued for many months despite substantial loss of renal function, allowing the drug to accumulate. We cannot state with certainty if either patient met the recommended criteria for dose reduction of TDF , as dosing guidelines are based on estimation of creatinine clearance (which requires weights) rather than the MDRD estimation of GFR. TDF dose should be reduced when the creatinine clearance falls below 50 mL/min in patients with stable CKD, as specified in the US prescribing information for TDF . Clinicians should also consider dose adjustment when renal function is unstable and has declined substantially by any reliable criteria, and should discontinue TDF if other causes of renal injury are excluded or signs of PRTD are also present. In these cases, the gradual rise in serum creatinine was a subtle but important sign that warranted assessment and intervention.
TDF is a widely prescribed antiretroviral that combines potency, convenient dosing and a favourable safety and tolerability profile. Despite case reports of renal injury, prospective controlled clinical trials have not shown excess renal toxicity attributable to TDF . We feel that patients with normal or mild abnormalities of renal function (creatinine clearance ≥50 mL/min or GFR ≥60 mL/min/1.73 m2) can safely receive TDF, with or without LPV/r, but periodic monitoring of renal function must be carried out routinely  and corrective action taken if significant declines are observed. Patients with more advanced CKD and those with concurrent medical conditions in which serum creatinine will tend to overestimate renal function require closer monitoring and dose adjustment or discontinuation of TDF if progressive renal dysfunction occurs.