Discontinuation of TDF due to renal dysfunction in Japanese HIV-1-infected patients

Table 1 shows the general background of the 149 included subjects, who were divided into those who discontinued TDF within 5 years (discontinuation group) and those who continued administration for at least 5 years (continuation group). The discontinuation group included 34 patients (23%), with a median duration until discontinuation of 967 days (interquartile range: 552–1491). While the median age of the patients in the discontinuation group was 45 years, that of the continuation group was 37 years, which was significantly younger (p < 0.001). In addition, there were fewer hepatitis B cases in the discontinuation group than that in the continuation group. There was no significant difference in the number of cases of trimethoprim/sulfamethoxazole given to prevent of pneumocystis pneumonia, which causes an apparent rise in SCr, and there were no cases in which a similarly acting combination drug was added during the observation period.

Table 1 Demographics of participantsRelationship between TDF discontinuation due to renal function-related adverse events and plasma TFV trough concentrations

The median plasma TFV trough concentrations of all 149 patients was 75 ng/mL (interquartile range: 57–97). Figure 1 shows the TFV trough concentrations of the discontinuation and continuation groups. The median TFV trough concentration of the discontinuation group was 88 ng/mL, which was significantly higher (p = 0.0041) than that of the continuation group (72 ng/mL).

Fig. 1

Comparison of Tenofovir plasma-trough concentrations and discontinuation due to impaired renal function. Horizontal straight line indicates median value. The Mann–Whitney U test showed significant differences (** p = 0.0041)

Figure 2 shows the rates of TDF discontinuation due to renal function-related adverse events for each TFV trough concentration quartile. The rate of TDF discontinuation tended to be higher when the TFV trough concentration was greater than or equal to the third quartile value (p = 0.001).

Fig. 2

Association between trough Tenofovir concentration and discontinuation of TDF. P value by Cochran-Armitage test is shown

The cut-off TFV trough concentration for discontinuation of TDF calculated using an ROC curve was 98 ng/mL (AUC, 0.660; sensitivity, 0.471; specificity, 0.817).

Factors associated with TDF discontinuation due to renal function-related adverse events

To investigate factors associated with the discontinuation of TDF due to renal function-related adverse events, univariate and logistic multivariate analyses were performed using explanatory variables, such as age ≥ 50 years, which is a previously reported risk factor for renal function decline due to TDF administration [10], eGFR levels < 80 (mL/min /1.73m2) at the start of TDF administration, median weight < 60 kg, and TFV trough concentration of ≥ 98 ng/mL, which is the cut-off value for TDF discontinuation calculated from the ROC curve. Hepatitis B was not listed as an explanatory variable because of the younger age (Median; 36 y.o.) and higher eGFR levels (Median; 105 mL/min /1.73m2) of the continuing patients. The Variance inflation factor (VIF) calculated by converting to dummy variables ranged from 1.1 to 1.2.

We found that being ≥ 50 years old (odds ratio (OR) 2.96; 95% confidence interval (CI), 1.01–8.64), having eGFR < 80 mL/min/1.73m2 (OR 5.51; 95% CI, 1.83–17.5), and having a TFV trough concentration of ≥ 98 ng/mL (OR 2.96; 95% CI, 1.16–7.60) were significant independent factors for TFV discontinuation due to renal function-related adverse events (Table 2).

Table 2 Associations between parameters and discontinuation of TDFRelationship between plasma TFV trough concentration and decline in renal function over time due to TDF administration

The 115 patients in the continuation group were divided by the TFV trough concentration quartiles for all 149 cases to examine annual changes in eGFR from the start of TDF administration till after five years. Figure 3 shows the median decline in eGFR per year, from the start of TDF administration. In the TFV trough concentration quartile, after five years, eGFR declined by 25, 17, 20, and 31 mL/min, respectively. eGFR was significantly lower after three years for the first and second quartiles, and after one year for the third and fourth quartiles than that at the start of administration. Further, for the fourth quartile, eGFR level after five years was significantly lower than that after one and two years. Also, the fourth quartile coincidentally matched the cutoff value by ROC.

Fig. 3

Association between trough Tenofovir concentration and median change in eGFR from baseline to 5 years. The x-axis is labeled with months to make the figure visually understandable; 30 days are used to represent 1 month. eGFR, estimated glomerular filtration rate. Filled circles, < 57 (ng/mL, n = 33); filled triangles, 57–75 (ng/mL, n = 28); filled diamonds, 75–98 (ng/mL, n = 33); filled squares, ≥ 98 (ng/mL, n = 21). The Steel–Dwass test showed significant differences (*p < 0.05, **p < 0.01)