Secondary Logo

Journal Logo

Serum Hypophosphatemia in Tenofovir Disoproxil Fumarate Recipients Is Multifactorial in Origin, Questioning the Utility of Its Monitoring in Clinical Practice

Day, Sara Louise MBChB*; Leake Date, Heather A*; Bannister, Alan*; Hankins, Matthew; Fisher, Martin MBBS*

JAIDS Journal of Acquired Immune Deficiency Syndromes: March 1st, 2005 - Volume 38 - Issue 3 - p 301-304
doi: 10.1097/01.qai.0000149790.29581.5a
Brief Report: Clinical Science

Tenofovir disoproxil fumarate (TDF) has been anecdotally associated with isolated hypophosphatemia (HP) as well as proximal tubular toxicity and renal dysfunction in which HP has consistently been a feature. Consequently, routine phosphate measurements in TDF recipients have been recommended. We identified and compared the frequency of HP in TDF recipients with that in non-TDF recipients; assessed the reproducibility of HP; identified the incidence of renal dysfunction in hypophosphatemic patients; and evaluated associations between HP and host, HIV infection, or treatment factors. This prospective observational study measured serum phosphate, urea, and creatinine in HIV-positive individuals among the following treatment groups: TDF-containing highly active antiretroviral therapy (HAART, group A), TDF-sparing HAART (group B), HAART naive (group C), and off HAART but treatment experienced (group D). Phosphate measurements were obtained in 252 patients. Seventy-two percent of patients prescribed TDF received a phosphate measurement. The frequency of HP in groups A, B, C, and D was 31%, 22%, 10%, and 14%, respectively. Seventy-eight percent of phosphate measurements were reproducible. Kaletra (P = 0.016) administration and duration of antiretroviral therapy (P = 0.023) were independently associated with HP, but elevated creatinine and urea or use of TDF was not. The etiology of HP seems to be multifactorial and unrelated to TDF or renal dysfunction. This questions the utility of routine phosphate testing, in isolation, in TDF recipients.

From the *Brighton and Sussex University Hospitals, National Health Service Trust, Brighton, Sussex, United Kingdom; and †Centre for Health Care Research, Brighton University, Sussex, United Kingdom.

Received for publication June 1, 2004; accepted October 21, 2004.

Statistical support funded by Gilead Sciences, Foster City, CA.

Reprints: Martin Fisher, Lawson Unit, Royal Sussex County Hospital, Eastern Road, Brighton BN2 5BE, United Kingdom (e-mail:

Tenofovir disoproxil fumarate (TDF) is a nucleotide analogue and a normally well-tolerated agent used for HIV treatment. Recently, however, some patients receiving TDF have experienced hypophosphatemia (HP) isolated1 or in association with renal dysfunction, particularly Fanconi syndrome, a condition of proximal tubular dysfunction (PRTD).2-4 These anecdotal reports provide cause for concern, given that adefovir, a structurally similar compound to TDF, was discontinued as HIV therapy because of a high incidence of PRTD mediated via mitochondrial toxicity5 or through interaction with human organic anion transporter protein 1 (hOAT 1).6 It had been anticipated that these issues would not be seen with TDF because of minimal mitochondrial toxicity observed in vitro7 and minimal interaction with hOAT 1.6 Preclinical studies of TDF observed renal toxicity in animals taking high doses,8 yet this was not demonstrated during pivotal clinical studies.9,10 Furthermore, TDF use has been associated with reduced bone mineral density and osteomalacia potentially mediated via HP.8,11 Because of concern that HP may reflect TDF-mediated renal toxicity or lead to significant bone pathologic changes in clinical practice, this department has routinely measured serum phosphate in TDF recipients, as has been recommended in the summary of product characteristics (SPC) for TDF as well as by other authors.2,8

The study aims were to identify and compare the range, severity, and frequency of HP in TDF recipients versus non-TDF recipients; to assess the reproducibility of HP in TDF recipients in the absence of intervention; to identify any association between renal dysfunction and HP; and to evaluate any other associations with HP, including host factors and HIV disease or treatment factors.

Back to Top | Article Outline


This prospective observational study performed serum phosphate, urea, and creatinine measurements in the following 4 treatment groups of HIV-positive individuals attending the Brighton HIV treatment center during 2003: TDF-containing highly active antiretroviral therapy (HAART, group A), TDF-sparing HAART (having never received TDF or not received it for 3 months, group B), naive to antiretroviral therapy (group C), and HAART experienced (but not having taken it within the past 3 months, group D). Patients who were not receiving TDF (groups B, C, and D) and who attended the clinic during a 2-month period for routine venipuncture received an isolated serum phosphate measurement, whereas those taking TDF (group A) were monitored regularly as part of their routine clinical care over a 6-month period. Patient consent and ethical approval were considered unnecessary by the local research ethics committee.

The following data were accessed from the pharmacy and HIV clinic databases and recorded for all groups: CD4 cell count, HIV RNA viral load, and serum urea and creatinine levels (all taken at the same time as the serum phosphate level); nadir CD4 cell count as an estimate of HIV stage; date of HIV diagnosis; duration receiving TDF and HAART; concomitant use of Lopinavir-ritonavir (Kaletra); and concurrent or prior use of nephrotoxic agents.

The Pearson χ2 and Fisher exact tests were used for univariate analyses, and logistic regression was used for multivariate analyses with SPSS software.

Back to Top | Article Outline


Of 140 patients prescribed TDF over the study period, 72% had received a serum phosphate measurement. Overall serum phosphate measurements were obtained in 252 patients. Groups A, B, C, and D comprised 101, 86, 51, and 14 patients, respectively. Group-specific patient demographics, surrogate markers, and duration taking TDF and/or HAART are shown in Table 1.



The frequency of HP (serum phosphate <0.8 mmol/L) in groups A, B, C, and D was 31%, 22%, 10%, and 14%, respectively. Group-specific phosphate ranges are shown in Table 1. There was no significant difference in HP range or severity (severe HP defined as ≤0.4 mmol/L) between the 4 groups or between TDF and non-TDF recipients.

Forty-four percent (44 of 101) of patients taking TDF received more than 1 serum phosphate measurement. Of these patients, 78% of initial measurements were reproducible: 11% (n = 5) were initially hypophosphatemic but subsequently normalized, 37% (n = 16) were consistently hypophosphatemic, 11% (n = 5) were initially normophosphatemic but became hypophosphatemic, and 41% (n = 18) were consistently normophosphatemic.

Univariate analyses pertaining to frequency of HP are shown in Table 2. Using multivariate analysis, only cumulative time on HAART and current use of Kaletra were independently associated with HP (see Table 2). There was no independent association found between the frequency of HP and TDF use, and there was no synergistic or additive relation demonstrated between Kaletra and TDF.



Back to Top | Article Outline


This study observed a higher frequency of HP in TDF recipients compared with clinical studies,9 but no independent association was found between TDF use and frequency or severity of HP.

The independent association between HP and length of time on HAART may explain the differing rates of HP reported among HAART-naive versus HAART-experienced patients9,10 and suggests that the high frequency of HP among TDF recipients is a reflection of the frequent use of TDF as part of “salvage” therapy.

Other studies found HP to be a transient biochemical abnormality in TDF recipients, resolving without treatment interruption.9 Our reproducibility figures suggest that HP is likely to persist, although it may not be related to TDF per se.

No association was observed between HP and abnormal renal function. This may suggest that other mechanisms, such as intestinal phosphate loss, are involved in the development of HP.8 Serum phosphate homeostasis, however, is primarily controlled at the renal proximal tubule,12 and although HP represents an effective tool signifying Fanconi syndrome or PRTD,13 it may be that the number of patients in this cohort who had PRTD was small or that serum urea and creatinine are insensitive markers of associated PRTD and a combination of other markers, such as serum and urinary phosphate/creatinine ratios, glycosuria, proteinuria, and urinary pH, may be more useful measurements. Serum creatinine may also be an insensitive marker of general renal dysfunction in HIV-positive patients because of reduced muscle mass, and creatinine clearance may thus be more appropriate to measure. Furthermore, the absence of data collection regarding the extensive range of alternative causes of renal dysfunction could limit detection of an association between HP and renal dysfunction.

Previous authors have found that a prior history of renal toxicity with adefovir or cidofovir may predispose TDF recipients to recurrent HP.1 Our data have shown no association between frequency of HP and these agents or other nephrotoxic agents.

Protease inhibitor therapy, especially Kaletra, has consistently been a concurrent medication among TDF recipients who developed renal dysfunction and HP.1,2 These agents can theoretically emphasize TDF-induced tubulopathy and, consequently, HP.4 This study found HP to be independently associated with Kaletra administration, something that requires further study, but no synergism with TDF was demonstrated.

In a nonfasted state, insulin induces intracellular phosphate uptake, which can lower serum values by 0.16 mmol/L.12 A diurnal variation in serum phosphate levels of 0.2 mmol/L may mean that most phosphate measurements are lower, given that clinic venipuncture takes place 3 hours on either side of the 11 o'clock nadir.12 These potential confounding factors were not recorded but probably affect all groups equally. Finally, the median time to anecdotal occurrence of PRTD among TDF recipients is 6.89 ± 5.51 months,13 whereas the median time taking TDF among our study group A was 5.26 months. The estimated incidence of renal toxicity in TDF recipients has recently been reported to be 0.85%.14 This study may thus not have been powered sufficiently or followed up for long enough to detect TDF-mediated renal toxicity or significant HP. Therefore, longer term monitoring of proximal tubular function among larger numbers of TDF recipients may be required.

Back to Top | Article Outline


This study identified a common (23%) occurrence of HP among a moderate-sized clinical cohort of HIV-positive patients. It also demonstrated HP has a multifactorial etiology, with no independent association with TDF use or renal dysfunction. This questions the utility of monitoring serum phosphate in TDF recipients to identify rare cases of PRTD or renal dysfunction, as suggested by the SPC for TDF. The study highlights the issue of introducing and potentially overinterpreting new monitoring tools without thorough evaluation, which may result in the unnecessary or premature discontinuation of effective medication. This is important, because patients taking Kaletra or TDF are often highly treatment experienced with few alternative options.

Because HP is a common finding, its management warrants further study, more so if it is likely to persist. Patients with HP should have repeat serum phosphate measurements performed. If HP is persistent, the more sensitive aforementioned PRTD markers should be measured and other common causes of HP (eg, respiratory alkalosis) ruled out. In the absence of symptoms or clinical sequelae of HP, treatment continuation may be appropriate. In a patient with refractory severe or symptomatic HP and limited HAART options, phosphate replacement may be an option,1 although this is not without complications and toxicity.12

Back to Top | Article Outline


The authors thank the Sussex Eye Hospital staff for collecting pharmacy data.

Back to Top | Article Outline


1. Blick G, Greiger-Zanlungo P, Garton T, et al. Tenofovir may cause severe hypophosphataemia in HIV/AIDS patients with prior adefovir-induced renal tubular acidosis [abstract 718]. Presented at: 10th Conference on Retroviruses and Opportunistic Infections; 2003; Boston.
2. Reynes J, Peyriere H, Merle de Boever C, et al. Renal tubular injury and severe hypophosphataemia (Fanconi syndrome) associated with tenofovir therapy [abstract 717]. Presented at: 10th Conference on Retroviruses and Opportunistic Infections; 2003; Boston.
3. Coca S, Perazella MA. Rapid communication: acute renal failure associated with tenofovir: evidence of drug-induced nephrotoxicity. Am J Med Sci. 2002;324:342-344.
4. Rollot F, Nazal E, Chauvelot-Moachon L, et al. Tenofovir-related Fanconi syndrome with nephrogenic diabetes insipidus in a patient with acquired immune deficiency syndrome: the role of lopinavir-ritonavir-didanosine. Clin Infect Dis. 2003;37:(on-line edition).
5. Tanji N, Tanji K, Kambham N, et al. Adefovir nephrotoxicity: possible role of mitochondrial DNA depletion. Hum Pathol. 2001;32:734-740.
6. Cihlar T, Ho E, Lin D, et al. Human renal organic anion transporter 1(hOAT 1) and its role in the nephrotoxicity of antiviral nucleotide analogs. Nucleosides Nucleotides Nucleic Acids. 2001;20:641-648.
7. Birkus G, Hitchcock M, Cihlar T. Assessment of mitochondrial toxicity in human cells treated with tenofovir: comparison with other nucleoside reverse transcriptase inhibitors. Antimicrob Agents Chemother. 2002;46:716-723.
8. Gilead Sciences International Limited. Viread summary of product characteristics. Foster City, CA: Gilead Sciences International Limited; 2003.
9. Cheng A, Coakley D, Chen S, et al. Safety profile of tenofovir DF in antiretroviral-experienced patients from randomised, double-blind, placebo-controlled clinical trials [abstract TuPo4460]. Presented at: XIV International AIDS Conference; 2002; Barcelona.
10. Gallant J, Pozniak A, Staszewski S, et al. Similar 96-week renal safety profile of tenofovir DF versus stavudine (d4T) when used in combination with lamivudine and efavirenz in antiretroviral naïve patients [abstract H-840]. Presented at: 43rd Interscience Conference on Antimicrobial Agents and Chemotherapy; 2003; Chicago.
11. Gilead Sciences, Inc. Background package for NDA 21-356/S-003: VIREAD (tenofovir disoproxil fumarate). Foster City, CA: Gilead Sciences, Inc.
12. Shiber JR, Mattu A. Serum phosphate abnormalities in the emergency department. J Emerg Med. 2002;23:395-400.
13. Izzedine H, Isnard-Bagnis C, Hulot JS, et al. Renal safety of tenofovir in HIV treatment-experienced patients. AIDS. 2004;18:1074-1076.
14. James C, Steinhaus M, Szabo S, et al. Tenofovir-related nephrotoxicity: case report and review of the literature. Pharmacotherapy. 2004;24:415-418.

phosphate; tenofovir; HIV; Fanconi syndrome; proximal tubule dysfunction

© 2005 Lippincott Williams & Wilkins, Inc.