Hypophosphataemia with non-tenofovir-containing antiretroviral therapy

INTRODUCTION

Tenofovir (TDF) is recommended as a preferred agent in first-line antiretroviral therapy (ART) regimens by the British HIV Association and is being increasingly used across the UK. ¹ However, a life-threatening renal tubular disorder, Fanconi's syndrome, is a rare but well recognized complication of TDF use. ² Fanconi's syndrome is characterized by impaired re-absorption of phosphate, glucose, urate, bicarbonate and amino acids in the kidney, resulting in hypophosphataemia and metabolic acidosis.

In addition, a number of studies have shown an increased incidence of hypophosphataemia and reduced glomerular filtration rate (GFR) with long-term TDF exposure in the absence of Fanconi's syndrome. Manosuthi et al. ³ monitored phosphate levels in patients switching from stavudine to TDF-based ART. In those switching to TDF plus nevirapine there was a significant fall in phosphate level and in GFR. Maggi et al. ⁴ evaluated patients who were starting a TDF-based regimen and those starting an abacavir-based regimen. There was a significant increase in urinary excretion of phosphate and uric acid in those on TDF and the levels of serum potassium and markers of mitochondrial toxicity were higher. A recent study has shown that renal impairment may be only partially reversible on stopping TDF. ⁵

The Summary of Product Characteristics (SPC, the official medicinal product description that is required for a marketing authorization within the European Union) for TDF recommends measurement of renal function and phosphate levels every four weeks for the first year and then at three-monthly intervals. ⁶

Hypophosphataemia in association with TDF therapy may be secondary to proximal tubule dysfunction; the underlying mechanism has not been fully elucidated but may involve mitochondrial toxicity. ⁷ However, several studies have shown little or no significant change in phosphate levels or GFR in people switching from a non-TDF-containing regimen to a TDF-containing regimen, suggesting that phosphate wasting may be multifactorial. ^{8 –11}

Furthermore, there is a higher risk of renal toxicity and hypophosphataemia with increasing age, co-morbidities such as hypertension and diabetes and in patients taking other nephrotoxic medications or protease inhibitors or nevirapine. ^3,12

Studies comparing phosphate levels in ART-treated and ART-naïve persons are scarce. To address this we determined the prevalence of hypophosphataemia in ART-treated and ART-naïve HIV-positive patients attending a teaching hospital outpatient clinic.

METHODS

Plasma phosphate was measured prospectively in patients registered for HIV management at our clinic as part of their routine care. As phosphate levels can be affected by food intake, sampling after fasting is recommended. ⁷ Data were therefore collected from 123 consecutive patients who attended after an overnight fast. None of the patients in this study were taking phosphate supplementation. The phosphate measurements were performed at a single accredited UK National Health Service laboratory.

Data were analysed using GraphPadInstat software (GraphPad Software, San Diego, CA, USA) and STATA version 11 (Stata Corp LP, College Station, TX, USA). Group comparisons were analysed using a two-sample t-test. The contribution of patient- and treatment-specific characteristics to plasma phosphate levels was explored using multiple linear regression. A P value of <0.05 was considered significant.

RESULTS

The study participants comprised 72 (59%) men and 51 (41%) women, of whom 53 (43%) were of white British ethnicity and 57 (46%) were of black African ethnicity (Table 1). Forty-two (34%) were ART-naïve and 81 (66%) were receiving ART (Figure 1). The mean age of the study participants was 40 years (SD 9.379).

OPEN IN VIEWER

Table 1

Patient characteristics and biochemical characteristics

	ART treated								ART-naïve
	All ART-treated	TDF-containing	Non-TDF-containing	NRTI	EFZ	NVP	PI	II
Patient characteristics
Male	43	21	22	42	14	21	8	1	29
Female	38	12	26	38	13	16	9	0	13
Mean age (years)	41	42	41	41	40	44	38	47	37
White British	31	17	14	30	9	15	7	1	22
Black African	41	12	29	41	14	19	8	0	16
Other ethnicity	9	4	5	9	4	3	2	0	4
Co-morbidities
Cardiovascular disease/VTE	3	2	1	3	0	2	1	0	0
Hepatitis C	1	0	1	1	0	1	0	0	0
Hepatitis B	1	1	0	1	0	0	1	0	0
Diabetes mellitus	1	1	0	1	0	1	0	0	0
Hypertension	5	3	2	5	0	4	1	0	3
Biochemical characteristics
Grade 1 HP (>0.65–0.8 mmol/L)	16	5	11	16	6	9	1	0	3
Grade 2 or lower HP (0.65 mmol/L or lower)	12	2	10	12	2	8	2	0	1
eGFR <60 mL/min/1.73m2	2	1	1	2	0	2	0	0	0

OPEN IN VIEWER

Figure 1

Antiretroviral categories. ART, antiretroviral therapy

The mean plasma phosphate concentration was 0.91 mmol/L (SD 0.1923) (normal range 0.8–1.4 mmol/L). Thirty-two patients (26%) had a phosphate level below the lower limit of normal, of whom 13 had hypophosphataemia of grade 2–4. Among ART-treated patients, the mean duration of ART was five years (SD 3.919). Seven of the 123 patients (6%) had a raised alkaline phosphatase (ALP), of whom only three were also hypophosphataemic.

Hypophosphataemia was significantly more frequent in patients receiving ART than ART-naïve patients (35% versus 10%, P = 0.0001 for any grade and 15% versus 2%, P = 0.002 for grade 2 or higher). There was also a trend towards more frequent hypophosphataemia in patients receiving a non-TDF regimen than those receiving TDF-containing ART (44% versus 21%; P = 0.056).

There was no statistically significant difference between the prevalence of hypophosphataemia in patients taking other groups of antiretrovirals, such as protease inhibitors (PIs) versus non-nucleoside reverse transcriptase inhibitors (NNRTIs) and efavirenz versus nevirapine.

We determined the eGFR (using the 4-parameter modification of diet in renal disease equation) in all subjects in order to ascertain whether renal impairment of any aetiology was a confounder for abnormal plasma phosphate levels. Only two patients had an eGFR less than 60 mL/min/1.73 m²; both were receiving ART (one with TDF and one with a non-TDF regimen) and had plasma phosphate levels in the normal range. None of the ART-naïve subjects had an eGFR less than 60 mL/min/1.73 m².

Using multiple linear regression, there was no significant association between phosphate level and gender, duration of ART, duration of HIV infection and eGFR when accounting for the other variables. However, increasing age was significantly associated with a rise in phosphate level (coefficient: 0.00458; P value: 0.04). Treatment with ART (both TDF-containing and non-TDF containing) remained significantly associated with a fall in phosphate level when accounting for the other variables (TDF-containing: coefficient 0.125, P value 0.01; non-TDF: coefficient 0.116, P value 0.008).

DISCUSSION

Our study has shown a similar overall prevalence of hypophosphataemia and a similar, very low, prevalence of renal impairment to other studies. ^7,8 We did not find a significantly higher prevalence of hypophosphataemia in patients taking TDF-containing regimens, consistent with previous reports. ^{7 –11}

We explored possible confounding factors that have been reported previously in association with hypophosphataemia, including male gender, NNRTI use, length of time on ART, PI use, nevirapine use and old age. ^3,7,11 These associations were not found in our study. In fact, increasing age was associated with increasing phosphate levels. However, the coefficient, and therefore the magnitude of the rise, was very small, the association has not been found in other studies and the clinical significance of this is uncertain.

Our study clearly shows an increased prevalence of hypophosphataemia in ART-treated individuals compared with ART-naive subjects. Although hypophosphataemia of grade 2 or higher was observed in a minority of patients receiving any ART, it was surprisingly frequent among patients receiving a non-TDF-containing ART regimen and was not related to impaired renal function. This finding was not explained by prior exposure to TDF, since the majority of patients were on their first ART regimen for the period during which the study was conducted. As current practice for many physicians is to monitor the phosphate level only of those patients taking a TDF-containing regimen, our findings suggest that clinically significant hypophosphataemia may be underestimated in patients on ART.

Chronic hypophosphataemia is of potential clinical significance as it may contribute to decreased bone density and osteomalacia. Low bone mineral density is increasingly recognized as an important co-morbidity in HIV infection. ¹³ Hypophosphataemia due to chronic renal wasting of phosphate could be an indication of previously unrecognized renal toxicity with non-TDF-containing ART. However, further studies are needed to determine whether renal wasting is the primary mechanism underlying hypophosphataemia in HIV-infected individuals as this will have important implications for its management.

Our study was limited by the cross-sectional design, involving a single phosphate measurement, therefore, these findings require confirmation in a larger prospective cohort. In addition, we were not able to examine the contribution of vitamin D status to hypophosphataemia in this study as vitamin D measurements were not available for all patients. Vitamin D deficiency is a possible cause of hypophosphataemia. There may have been a channelling bias in our study as patients with known renal impairment may have been preferentially treated with a non-TDF-containing regimen. However, the low number of patients with an eGFR less than 60 mL/min/1.73 m² suggests that this is unlikely.

In conclusion, our findings raise two issues: first, that current guidelines for monitoring of hypophosphataemia in patients receiving TDF should be reviewed with regard to its utility in predicting TDF-related toxicity; secondly, monitoring of plasma phosphate should be considered in patients on any ART regimen irrespective of whether there is concomitant renal impairment.