The widespread use of tenofovir disoproxil fumarate (TDF) may somewhat be limited by its potential renal toxicity as suggested by a growing number of case reports [1,2]. In addition to being freely filtered by the glomerulus, TDF and other nucleotide reverse transcriptase inhibitors (NtRTIs) are actively and largely secreted at the level of the proximal tubule where they can induce a toxic mitochondriopathy . As a result of the reduced energy outflow, the massive fluid and electrolytes proximal reabsorption becomes insufficient, leading to a renal leak of phosphates, small proteins, glucose, uric acid, amino acids and bicarbonates. This acute or sub-acute dysfunction, which is known as ‘Fanconi syndrome’, is potentially very deleterious, especially for bone mineralization [4,5]. For this reason, it is recommended to closely monitor renal function during NtRTIs treatment.
Cystatin C (CysC) is a low-molecular weight (LMW) protein (13 kDa), which has been widely investigated as a glomerular filtration rate marker . Conversely, CysC has been little studied as a tubular marker. In 2002, Uchida and Gotoh  were the first to perform a large-scale study suggesting that the urinary CysC to urinary creatinine ratio (UCysC/UCrea) may serve as an index of tubular impairment from a very early stage to severe damage of the tubular cells. More recent reports confirmed that UCysC was an accurate marker, either to identify patients requiring kidney replacement therapy after acute tubular necrosis  or to detect tubular dysfunction among pure and mixed nephropathies .
In this study, we assessed the diagnostic performance of the UCysC/UCrea ratio to detect or rule out a Fanconi syndrome induced by TDF.
Thirty-seven patients infected with HIV were referred for a suspicion of TDF (n = 36) or adefovir-induced nephropathy (n = 1) on the basis of hypophosphoremia (n = 15) and/or severe bone demineralization (n = 8) and/or decreased evaluated glomerular filtration rate (GFR; n = 20). They underwent an extensive renal investigation in order to confirm the diagnosis of proximal tubule nephropathy. Eleven patients met the diagnostic criteria of a Fanconi syndrome: hypophosphatemia (median 0.51 mmol/l; range 0.24–0.76) and renal leak of phosphates (tubular reabsorption: median 47.8%; range 25.0–75.5); LMW proteinuria; normoglycemic glucosuria; aminoaciduria; and, for most of them, hypouricemia and renal leak of uric acid and proximal tubular acidosis.
UCysC was measured retrospectively to renal investigation on remaining samples of freshly voided urine using a particle-enhanced nephelometric immunoassay (PENIA; Dade-Behring-Siemens, Marburg, Germany). In the Fanconi negative group (n = 26), eight samples had an UCysC concentration below the detection limit (<48 μg/l) and 18 had very low but detectable levels of UCysC. These 18 concentrations were within the physiologic ranges proposed by Conti et al. (<270 μg/l)  or by Herget-Rosenthal et al. (<180 μg/l) . However, we cannot rule out a minimal tubular dysfunction. Indeed, in a population of 38 GFR, sex and age-matched TDF-naive patients without symptoms of proximal tubule injury, the UCysC was under the detection limit for all urine specimens (data not shown). In the Fanconi positive group (n = 11), all samples had detectable UCysC. Measurable UCysC/UCrea levels in the Fanconi negative group ranged from 3–39 μg/mmol with a median of 7 μg/mmol, whereas in the Fanconi positive group, UCysC/UCrea levels ranged from 12–4970 μg/mmol with a median of 79 μg/mmol (P = 0.0001). A receiver operating curve (ROC) analysis was performed to define the best threshold and assess the diagnostic performance of the UCysC/UCrea ratio . The UCysC/UCrea ratio had a high ROC area under curve (0.929), which means that, in 92.9% of the cases, it was able to distinguish a Fanconi positive patient from a Fanconi negative one. The highest accuracy was obtained with a threshold of 14 μg/mmol. This threshold was associated with a positive predictive value of 76.9% and a negative predictive value of 95.8% (Fig. 1).
Hypophosphatemia is believed to be a sensitive sign of a Fanconi syndrome . Indeed, in our group of Fanconi positive patients, all of them had hypophosphatemia. However, hypophosphatemia was present in eight of the 26 patients for whom a Fanconi syndrome had been ruled out by the renal function tests. Hypophosphatemia is one of the most common electrolyte disorders in HIV patients and may be related to abnormalities in vitamin D metabolism and to digestive losses [11,12]. This low specificity also concerns proteinuria and glucosuria. Yet, almost 25% of HIV patients have glomerulopathy , and proteinuria has been shown to occur in 32% of HIV-infected female individuals . Concerning nondiabetic glucosuria, its specificity as a tubular marker may be limited by the intermittent glucosuria related to the insulin resistance frequently observed in HIV patients . Therefore, as a specific marker was unavailable until now, the occurrence of hypophosphatemia and/or proteinuria and/or normoglycemic glucosuria imposed a further extensive renal exploration to confirm or rule out a Fanconi syndrome.
The present study shows that the UCysC/UCrea ratio is an efficient marker; the high negative predictive value for a UCysC/UCrea ratio less than 14 μg/mmol enables a NtRTI-induced Fanconi syndrome to be ruled out in 95.8% of the cases, whatever the results for proteinuria, glucosuria and phosphatemia. Although more than 75%, the positive predictive value of the UCysC/UCrea ratio is a little lower. Thus, an UCysC/UCrea ratio of at least 14 μg/mmol should not be interpreted alone. If proteinuria, glucosuria and phosphatemia are normal, an intensified renal follow-up is probably needed. If one or more of the three usual markers are also pathological, then a renal function exploration is indicated to confirm or rule out a Fanconi syndrome.
In conclusion, the measurement of the UCysC/UCrea ratio on a single urine voiding is a simple and efficient marker to exclude a Fanconi syndrome induced by TDF and it can reasonably be assumed by other NtRTIs. Further studies are needed to define the decision threshold and the positive predictive value of this marker. Nevertheless, as its measurement can be performed easily and quickly (with a cost similar that of a CD4 cell count), we have already proposed to measure the UCysC/UCrea ratio every time an NtRTI-induced tubulopathy is suspected.
Professor I. Tack and Dr S. Séronie-Vivien supervised this work. Professor P. Massip referred the patients to Professor I. Tack and Dr A. Jaafar, who performed the renal function explorations. Dr A. Jaafar and Dr L. Malard performed the biochemical analysis of the urine specimens, and Dr S. Séronie-Vivien carried out the statistical analysis and preparation of the article that was critically read by Professor E. Chatelut. We thank Mrs C. Franques for her helpful technical assistance and Professor John Woodley for his help with the English. This study was carried out with the support of Siemens, who provided the analyzer and the reagents for urinary cystatin C measurements.
1. Izzedine H, Isnard-Bagnis C, Hulot JS, Vittecoq D, Cheng A, Jais CK, et al
. Renal safety of tenofovir in HIV treatment-experienced patients. AIDS 2004; 7:1074–1076.
2. Zimmermann AE, Pizzoferrato T, Bedford J, Morris A, Hoffman R, Braden G. Tenofovir-associated acute and chronic kidney disease: a case of multiple drug interactions. Clin Infect Dis 2006; 2:283–290.
3. Cote HC, Magil AB, Harris M, Scarth BJ, Gadawski I, Wang N, et al
. Exploring mitochondrial nephrotoxicity as a potential mechanism of kidney dysfunction among HIV-infected patients on highly active antiretroviral therapy. Antivir Ther 2006; 1:79–86.
4. Brim NM, Cu-Uvin S, Hu SL, O'Bell JW. Bone disease and pathologic fractures in a patient with tenofovir-induced Fanconi syndrome. AIDS Read 2007; 6:322–328. C3.
5. Earle KE, Seneviratne T, Shaker J, Shoback D. Fanconi's syndrome in HIV+ adults: report of three cases and literature review. J Bone Miner Res 2004; 5:714–721.
6. Roos JF, Doust J, Tett SE, Kirkpatrick CM. Diagnostic accuracy of cystatin C compared to serum creatinine for the estimation of renal dysfunction in adults and children – a meta-analysis. Clin Biochem 2007; 40:383–391.
7. Uchida K, Gotoh A. Measurement of cystatin-C and creatinine in urine. Clin Chim Acta 2002; 323:121–128.
8. Herget-Rosenthal S, Poppen D, Husing J, Marggraf G, Pietruck F, Jakob HG, et al
. Prognostic value of tubular proteinuria and enzymuria in nonoliguric acute tubular necrosis. Clin Chem 2004; 3:552–558.
9. Conti M, Moutereau S, Zater M, Lallali K, Durrbach A, Manivet P, et al
. Urinary cystatin C as a specific marker of tubular dysfunction. Clin Chem Lab Med 2006; 3:288–291.
10. Zweig MH, Campbell G. Receiver-operating characteristic (ROC) plots: a fundamental evaluation tool in clinical medicine. Clin Chem 1993; 4:561–577.
11. Day SL, Leake Date HA, Bannister A, Hankins M, Fisher M. Serum hypophosphatemia in tenofovir disoproxil fumarate recipients is multifactorial in origin, questioning the utility of its monitoring in clinical practice. J Acquir Immune Defic Syndr 2005; 3:301–304.
12. Isnard BC, Du Montcel ST, Fonfrede M, Jaudon MC, Thibault V, Carcelain G, et al
. Changing electrolyte and acido-basic profile in HIV-infected patients in the HAART era. Nephron Physiol 2006; 3:131–138.
13. Fernando SK, Finkelstein FO, Moore BA, Weissman S. Prevalence of chronic kidney disease in an urban HIV infected population. Am J Med Sci 2008; 2:89–94.
14. Szczech LA, Gange SJ, van der HC, Bartlett JA, Young M, Cohen MH, et al
. Predictors of proteinuria and renal failure among women with HIV infection. Kidney Int 2002; 1:195–202.
15. Florescu D, Kotler DP. Insulin resistance, glucose intolerance and diabetes mellitus in HIV-infected patients. Antivir Ther 2007; 2:149–162.