A total of 618 samples (12%) from 243 patients (18%) had uPCR ≥ 30 mg/mmol on at least one measurement. At the time

the first uPCR sample was measured, the median duration of infection was 6.4 years (IQR 2.5–11.8 years) and 88% were cART-experienced. Sixty-seven RNA Synthesis inhibitor patients with at least one measurement of uPCR ≥ 30 mg/mmol had concurrent urine albumin and total protein measurements, and thus uAPR could be calculated. Paired measurements were also more likely to be taken among patients who were cART-experienced (P = 0.02), or who were on a boosted PI either before (P < 0.001) or at the time (P < 0.001) the paired samples were measured, but were less likely to be taken on patients who were on TDF at the time of sampling (P < 0.001). Forty-six (69%) of these 67 patients had been taking TDF at the time of sampling. There were no significant differences in age, duration of HIV infection, nadir CD4 count, plasma creatinine concentration, eGFR, plasma phosphate concentration, fractional excretion of phosphate or uPCR (all P > 0.05) between patients who were taking TDF and those NVP-BKM120 nmr who were not taking TDF at the time of sampling. Patients on TDF also had a lower uACR (median 10 vs. 33 mg/mmol,

respectively; P < 0.01) and a lower uAPR (median 0.18 vs. 0.69, respectively; P < 0.01). Of these 67 proteinuric patients, 46 (32%) had TP, while 21 (15%) had GP. There was no difference between the TP and GP groups with regard to age, sex, ethnicity, sexuality, duration of HIV infection, nadir CD4 count, plasma creatinine or eGFR (Table 1). Plasma phosphate was lower, whereas fractional excretion of phosphate was higher in the TP group (Table 1). uPCR was significantly lower in the TP group compared with the GP group (median 49 vs. 102 mg/mmol, respectively; P < 0.01). uACR was significantly lower in the TP group compared with the GP group (median 9 vs. 72 mg/mmol, respectively; P < 0.01). Patients in the TP group were more likely Etoposide to have been on TDF or a boosted

PI prior to sampling, and to have been taking TDF and/or a boosted PI at the time of sampling (Table 1). There were 18 patients (14%) with heavy proteinuria (uPCR > 100 mg/mmol), two of whom had diagnoses of TDF-related renal injury, both of which improved after switching from TDF (Table 2). An additional patient was on TDF because he was hepatitis B virus coinfected. Eight patients with heavy proteinuria had a renal biopsy; all the biopsy results correlated with the definitions of proteinuria using uPCR and uACR. There were three patients who were thought to have tubular dysfunction. None of these patients has undergone a renal biopsy, and in some the proteinuria resolved on switching antiretroviral agents. When uAPR was calculated in these 18 patients, there was a significant difference between TP and GP pathologies (P = 0.001) (Fig. 1).