BackgroundTreatment with tenofovir is sometimes associated with renal dysfunction. Limited information is available on this side effect in patients with small body weight, although the use of tenofovir will spread rapidly in Asia and Africa, where patients are likely to be of smaller body weight.MethodsIn a single-center cohort, Japanese patients with HIV infection who started tenofovir-containing antiretroviral therapy were retrospectively analyzed. The incidence of tenofovir-associated renal dysfunction, defined as more than 25% decrement of estimated glomerular filtration rate (eGFR) from the baseline, was determined. The effects of small body weight and body mass index (BMI) on tenofovir-associated renal dysfunction, respectively, were estimated in univariate and multivariate Cox hazards models as the primary exposure. Other possible risk factors were evaluated by univariate analysis and those found significant were entered into the multivariate analysis.ResultsThe median weight of 495 patients was 63 kg. Tenofovir-related renal dysfunction occurred in 97 (19.6%) patients (incidence: 10.5 per 100 person-years). Univariate analysis showed that the incidence of tenofovir-related renal dysfunction was significantly associated with smaller body weight and BMI, respectively (per 5 kg decrement, HR = 1.23; 95% CI, 1.10–1.37; p<0.001)(per 1 kg/m2 decrement, HR = 1.14; 95% CI, 1.05–1.23; p = 0.001). Old age, high baseline eGFR, low serum creatinine, low CD4 count, high HIV viral load, concurrent nephrotoxic drugs, hepatitis C infection, and current smoking were also associated with tenofovir-related renal dysfunction. Multivariate analysis identified small body weight as a significant risk (adjusted HR = 1.13; 95% CI, 1.01–1.27; p = 0.039), while small BMI had marginal significance (adjusted HR = 1.07; 95% CI 1.00–1.16; p = 0.058).ConclusionThe incidence of tenofovir-associated renal dysfunction in Japanese patients was high. Small body weight was identified as an independent risk factor for tenofovir-associated renal dysfunction. Close monitoring of renal function is advocated for patients with small body weight treated with tenofovir.

In this cohort of patients with low body weight, TDF exposure increased the risk of renal dysfunction. Furthermore, the loss in eGFR relative to the control increased continuously up to 5 years.

It became evident that tenofovir DF (TDF) causes a modest and gradual decline in GFR, however, the impact of long-term use of TDF on tubular function has not been fully evaluated. In 40 patients treated with TDF and 23 patients treated with other NRTIs, urine beta(2)-microglobulin (U-BMG), percentage tubular reabsorption of phosphate (%TRP), alkaline phosphatase (ALP), serum creatinine, and calculated GFR were prospectively measured for 96 weeks. In patients receiving TDF, median U-BMG rose from 188 microg/liter at baseline to 555 microg/liter at week 96 (p = 0.02), median %TRP declined from 94% at baseline to 90% at week 96 (p = 0.002), median ALP ratio compared with baseline persistently increased from 1 to 1.278 at week 96 (p = 0.001), and serum creatinine showed significant but minimal change from 0.64 mg/dl to 0.74 mg/dl at week 96 (p = 0.02). The GFR level declined minimally but significantly in TDF-receiving patients (-17 ml/min/1.73 m(2)), whereas it did not change in other NRTI-receiving patients [+ 3 ml/min/1.73 m(2); mixed models analysis of variance (MMANOVA) p = 0.03 for overall change from baseline to week 96]. U-BMG, %TRP, ALP, or serum creatinine did not change significantly in other NRTI-receiving patients during the observation period. In five patients with marked changes in U-BMG (>10,000 microg/liter) and %TRP (<80%), both U-BMG and %TRP immediately recovered in all patients after discontinuing TDF, whereas GFR levels did not fully recover for 6 months in three patients. Prolonged treatment with TDF caused progressive renal tubular dysfunction as well as a modest decline in GFR. If U-BMG levels >10,000 microg/liter and %TRP values <80% are observed, discontinuing TDF may be beneficial.

ObjectiveTo compare the rate of decline of renal function in tenofovir- and abacavir-based antiretroviral therapy (ART) in low-body weight treatment-naïve patients with HIV infection.DesignWe conducted a single-center retrospective cohort study of 503 Japanese patients who commenced on either tenofovir- or abacavir-based initial ART.MethodsThe incidence of renal dysfunction, defined as more than 25% fall in estimated glomerular filtration rate (eGFR) from the baseline, was determined in each group. The effect of tenofovir on renal dysfunction was estimated by univariate and multivariate Cox hazards models as the primary exposure. Changes in eGFR until 96 weeks were estimated in both groups with a repeated measures mixed model.ResultsThe median body weight of the cohort was 64 kg. The estimated incidence of renal dysfunction in the tenofovir and the abacavir arm was 9.84 per 100 and 4.55 per 100 person-years, respectively. Tenofovir was significantly associated with renal dysfunction by univariate and multivariate analysis (HR = 1.747; 95% CI, 1.152–2.648; p = 0.009) (adjusted HR = 2.080; 95% CI, 1.339–3.232; p<0.001). In subgroup analysis of the patients stratified by intertertile baseline body weight, the effect of tenofovir on renal dysfunction was more evident in patients with lower baseline body weight by multivariate analysis (≤60 kg: adjusted HR = 2.771; 95%CI, 1.494–5.139; p = 0.001) (61–68 kg: adjusted HR = 1.908; 95%CI, 0.764–4.768; p = 0.167) (>68 kg: adjusted HR = 0.997; 95%CI, 0.318–3.121; p = 0.995). The fall in eGFR was significantly greater in the tenofovir arm than the abacavir arm after starting ART (p = 0.003).ConclusionThe incidence of renal dysfunction in low body weight patients treated with tenofovir was twice as high as those treated with abacavir. Close monitoring of renal function is recommended for patients with small body weight especially those with baseline body weight <60 kg treated with tenofovir.

There is no detailed information on the association between age, time of disease, and HIV-associated neurocognitive disorders (HAND). In this prospective study involving 17 medical facilities across Japan, we recruited HIV-infected patients to complete a 14-test neuropsychological battery that assess eight neurocognitive domains. HAND were diagnosed by the Frascati criteria. Of 1399 recruited patients, 728 were enrolled. The prevalence of HAND was 25.3% [13.5% asymptomatic neurocognitive impairment, 10.6% mild neurocognitive disorder (MND), and 1.2% HIV-associated dementia (HAD)]. Tests that assess executive and visuospatial functions showed better diagnostic accuracy than other tests for HAND. Multivariate analysis identified age (≥ 50 years) and incomplete virological suppression as risk factors for MND and HAD and current ART as a protective factor. The prevalence of MND and HAD was low in the early stage of infection (6.3% in ≥ 2 to < 6 years), then increased in the later stage [17.3% in ≥ 11 years, p = 0.001 (vs. ≥ 2 to < 6 years)], independent of age or treatment. Older patients were more likely to show MND or HAD in the early stage of HIV infection (26.7 vs. 8.7% for < 2 years and 17.4 vs. 3.1% for ≥ 2 to < 6 years, p = 0.040 and 0.004, respectively) compared to younger ones. In conclusion, MND and HAD were more commonly found in later years since diagnosis of HIV infection and older patients are at risk of neurocognitive impairment at the early stage of HIV infection. Tests for executive and visuospatial functions seem more sensitive than other tests for diagnosing HAND.

Despite its wide use, the renal tubular toxicity of tenofovir has not been fully evaluated. Twelve weeks after initiating a tenofovir-containing HAART regimen, a high urine-beta 2 microglobulin level was observed in 12 out of 17 patients, the percentage of tubular reabsorption of phosphate decreased from 96.0 to 91.1% and alkaline phosphatase increased from 294 to 365 U/l, whereas serum creatinine and phosphorus remained largely unchanged. Patients with the above findings should be monitored carefully for renal tubular toxicity.

HIV-infected patients are at high risk for bone mineral density (BMD) loss. The present study was designed to provide information on characteristics of BMD abnormalities in Japanese HIV-1-infected patients and risk factors involved in worsening of BMD. A total of 184 Japanese HIV-1-infected men were studied with a dual-energy X-ray absorptiometry scan (DXA) at the lumbar spine and femoral neck. Multivariate logistic regression models were used for comparison of the impact of risk factors on BMD loss. Osteopenia and osteoporosis were diagnosed in 46% and 10% of the patients at lumbar spine, and 54% and 12% at femoral neck, respectively. In logistic analysis, factors associated with low BMD at both lumbar spine and femoral neck were long-term treatment with a protease inhibitor (PI) [odds ratio (OR) 1.100 and 1.187 per 1 year increase of PI use; 95% confidence interval (CI) 1.003-1.207 and 1.043-1.351; p=0.042 and 0.009, respectively] and a low body mass index [OR: 0.938 and 0.852, CI 0.892-0.992 and 0.783-0.927; p=0.024 and <0.001, respectively]. Patients who discontinued PI had a significantly higher BMD than those who currently use PI at lumbar spine (t score -0.8 vs. -1.3, p=0.04) but not at femoral neck (-1.3 vs. -1.5, p=0.38). In HIV-infected Japanese patients, the duration of treatment with PI correlated significantly with BMD loss. Discontinuation of PI is a promising option in the treatment of BMD loss since it allows recovery of BMD, especially in the lumbar spine.

BackgroundThe epidemiology of incident syphilis infection among HIV-1-infected men who have sex with men (MSM) largely remains unknown.MethodsThe incidence and risk factors for incident syphilis (positive TPHA and RPR> = 1:8) among HIV-1-infected MSM who visited a large HIV clinic in Tokyo for the first time between 2008 and 2013 were determined, using clinical data and stored blood samples taken every three months for screening and determination of the date of incident syphilis. Poisson regression compared the incidence of syphilis at different observation periods.ResultsOf 885 HIV-1-infected MSM with baseline data, 34% either presented with active syphilis at baseline (21%) or became infected with syphilis during follow-up (13%). After excluding 214 patients (MSM with syphilis at baseline (n = 190) and no follow-up syphilis test (n = 24)), of 671 men, 112 (17%) developed incident syphilis with an incidence of 43.7/1,000 person-years [95% CI, 36.5–52.3]. The incidence decreased slightly during observation period although the trend was not significant (2008–2009: 48.2/1,000 person-years, 2010–2011: 51.1/1,000 person-years, 2012–2013: 42.6/1,000 person-years, 2014 to 2015: 37.9/1,000 person-years, p = 0.315). Multivariable analysis identified young age (<33 years versus >40, HR 4.0, 95%CI 2.22–7.18, p<0.001), history of syphilis at baseline (HR 3.0, 95%CI 2.03–4.47, p<0.001), positive anti-amoeba antibody (HR 1.8, 95%CI 1.17–2.68, p = 0.006), and high baseline CD4 count (CD4 ≥350 /μL versus CD4 <200, HR 1.6, 95%CI 1.00–2.53, p = 0.050) as risk factors for incident syphilis. Incidence of syphilis was particularly high among young patients (age <33 years: 60.1/1,000 person-years). Interestingly, 37% of patients with incident syphilis were asymptomatic.ConclusionsAlthough incidence of syphilis did not increase during the observation period, it was high among HIV-1-infected MSM, especially among young HIV-1-infected MSM and those with history of syphilis, in Tokyo. Regular screening for syphilis needs to be strictly applied to this population.