Therapeutic drug monitoring (TDM) has been advocated as a means to optimize the safety and efficacy of antiretroviral therapy. Nevertheless, apart from indinavir and nelfinavir in treatment-naive HIV-infected patients,1,2 there is no evidence from randomized controlled trials that TDM improves therapeutic outcome.3
Standard dosing of efavirenz (EFV), a currently preferred first-line antiretroviral agent, leads to therapeutic plasma concentrations in at least 80% of the HIV-infected individuals,4,5 compared with just 40% for older agents such as nelfinavir.1 As a consequence, large and expensive trials, with more than 500 patients per treatment arm,6 are required to obtain adequate statistical power to judge the potential benefits of the application of TDM as a routine measurement in all patients taking EFV. Therefore, we agree with Khoo et al6 that in the present situation, the value of TDM is best assessed by performing “utilitarian” studies. The goal of these studies is not to provide evidence for routine use of TDM but to explore the use of TDM in specific clinical situations. These studies should for instance focus on the use of TDM during pregnancy or the use of TDM in patients with severe liver impairment. In this article, we describe the use of TDM to manage EFV-related toxicity.
Central nervous system (CNS) side effects are a well known and frequently occurring complication of EFV therapy.7 Several reports have demonstrated the relationship of these side effects to high EFV plasma concentrations.4,8-11 In addition, there is international consensus on a therapeutic window for EFV plasma concentrations: 1.0-4.0 mg/L.12
At our TDM practice, we regularly receive requests for TDM in patients using EFV who suffer from CNS side effects. In case these patients are found to have high EFV plasma concentrations (≥4.0 mg/L), our advice to the clinicians is to reduce the dose of EFV to 400 mg once daily under the guidance of TDM. However, we have no formal evidence that this intervention improves the clinical outcome of these patients.
The objective of the present study was to establish whether EFV dose reduction prevents toxicity-induced EFV discontinuations in patients with high plasma concentrations. In addition, we aimed to evaluate whether dose reduction affects virological efficacy.
All 25 Dutch hospitals that provided antiretroviral treatment participated in the AIDS Therapy Evaluation in The Netherlands (ATHENA) observational cohort study. Currently, data from more than 11,000 patients have been anonymously recorded in a central database that is maintained by the HIV Monitoring Foundation.13
We selected all patients in ATHENA who had a high EFV plasma concentration (ie, ≥4.0 mg/L) recorded within 48 weeks after commencing EFV-based antiretroviral combination therapy. This cohort was subsequently classified into 2 groups. The reduced-dose (RD) group consisted of those who underwent dose reduction after the high plasma concentration determination. The date of dose reduction was considered baseline. The standard-dose (SD) group consisted of patients who continued the standard EFV dosage (ie, 600 mg once daily). For them, baseline was the first documented clinic visit after the high EFV plasma concentration measurement.
Patient characteristics at the time of starting EFV were tabulated for patients in the RD group and the SD group. Differences between groups were compared using χ2 or Fisher exact tests for categorical data and Mann-Whitney tests for continuous data. All tests were 2 sided, and a P value of less than 0.05 was considered statistically significant.
EFV plasma concentrations before and after baseline were compared by using the Wilcoxon matched pairs signed rank sum test. For patients who underwent dose reduction, the EFV plasma concentration after baseline had to be taken at least 10 days after the date of dose reduction to have achieved new steady state conditions.
Reasons for discontinuation of antiretroviral agents are collected in the ATHENA database. Kaplan-Meier and Cox proportional hazards analysis were used to assess the impact of dose reduction on toxicity-induced EFV discontinuations. Patients who discontinued EFV for reasons other than toxicity were censored from the moment of discontinuation. Possible effect-measure modification and confounding were assessed for the following parameters: gender, region of origin (as a surrogate for ethnicity), age, body mass index (calculated as the weight in kilograms divided by the square of the height in meters), hepatitis B and C status, HIV transmission risk group, CD4 count at the start of EFV, specific nucleoside reverse transcriptase inhibitor backbone, the EFV concentration before baseline, and pretreated status at the start of EFV. Pretreated status was categorized as follows: (1) treatment-naive patients who started EFV; (2) treatment-experienced patients with an undetectectable viral load (<50 copies/mL) at the start of EFV; (3) treatment-experienced patients with a detectectable viral load at the start of EFV.
Multivariable logistic regression modeling was used to investigate the effect of dose reduction on virological response, which was defined as having a viral load below 50 copies per milliliter at week 24 after baseline. We used an observed failure approach in which patients who discontinued EFV due to virological failure were considered failures at subsequent time points, whereas patients who discontinued EFV due to other reasons (eg, pregnancy wish) were censored from that moment onwards. The same factors that we used as independent covariables in the Cox regression analysis for toxicity-related discontinuations were investigated (see above). In addition, plasma viral load at baseline was used as an independent covariable. We used a stepwise selection procedure to identify parameters that were significantly (P < 0.10) associated with virological response. The EFV dose (RD or SD) was a fixed parameter in all models.
Apart from the primary analysis of the virological response 24 weeks after baseline, several sensitivity analyses were carried out to see whether dose reduction affected virological suppression at week 48 after baseline and at week 24 and week 48 after starting EFV. In addition, the analyses for the above mentioned 4 virological efficacy endpoints were repeated stratified for pretreated status at the start of EFV-based antiretroviral therapy. All data were analyzed with SPSS for MS Windows, version 16.0.1.
We identified 180 subjects who had high plasma EFV levels after the start of an EFV-containing antiretroviral regimen. The date of starting EFV ranged from July 1998 to February 2007. Forty-nine patients of the 180 patients (27.2%) underwent a dose reduction (RD group) and 131 patients (72.8%) continued the standard dosage regimen (SD group).
The patient characteristics between these groups differed significantly with regard to gender, HIV transmission risk group, region of origin, and EFV plasma concentration. The median [interquartile range (IQR)] EFV plasma concentration was significantly higher in patients who underwent dose reduction [6.8 (5.7-9.6) mg/L], compared with patients who continued the SD [5.1 (4.3-6.4) mg/L, P < 0.001]. Furthermore, there were more females, heterosexually infected patients, and patients originating from sub Saharan Africa in the RD group (Table 1).
Magnitude of Dose Reduction and Pharmacokinetic Outcome
The EFV dose was reduced from 600 to 400 mg in 47 of the 49 patients in the RD group. As a result, the median (IQR) EFV plasma concentration decreased from 6.8 (5.6-9.5) to 4.0 (3.0-5.6) mg/L (P < 0.001) in these 47 patients.
Two patients underwent a dose reduction directly from 600 mg once daily to 200 mg once daily. In one of these, the plasma concentration decreased from 6.4 mg/L to 2.7 mg/L. In the other patient, the EFV plasma concentration decreased from 27.7 mg/L to 11.4 mg/L. The dosage was further reduced in this latter patient to 100 mg once daily, which resulted in a therapeutic concentration of 2.7 mg/L. Both patients had viral loads below 50 copies per milliliter at all subsequent time points and were included in all further analyses.
The EFV plasma concentration remained above the threshold for efficacy (ie, 1.0 mg/L) in all 42 patients who had a second plasma concentration available after dose reduction. Despite the significant reduction in EFV plasma concentrations in the RD group, 22 patients still had EFV plasma concentrations above 4.0 mg/L. Seven of them had their dosage subsequently further reduced to 200 mg once daily, which resulted in a therapeutic EFV concentration in 3 patients. Two patients still had an elevated EFV plasma concentration of 5.1 mg/L after the dose reduction to 200 mg once daily, and for 2 patients, no measured EFV concentrations were available.
Half of the patients (n = 68, 52%) who remained on the SD had no EFV plasma concentration measured after baseline. In the 63 patients who had a second EFV plasma concentration available, the median (IQR) EFV plasma concentrations were 5.3 (4.3-6.4) and 4.6 (3.5-7.3) mg/L before and after baseline, respectively (P = 0.12).
Toxicity-Induced Discontinuations of EFV
At week 48, 14 patients from the SD group had discontinued EFV due to toxicity, compared with 1 patient in the RD group; Figure 1 shows the Kaplan-Meier curves for toxicity-induced discontinuations in both groups. At week 48, the estimated cumulative incidence of toxicity-induced discontinuations was 11.5% for patients in the SD group compared with 2.3% for patients in the RD group; P = 0.066 (log-rank test). In a Cox proportional hazards model, patients from the RD group had a lower risk (hazard ratio: 0.18, 95% confidence interval 0.02 to 1.40) of toxicity-related EFV discontinuations compared with patients from the SD group. Further explorations using multivariable Cox proportional hazards models showed that no other parameter was significantly associated with the risk of discontinuation of EFV due to toxicity, or significantly (>10%) modified the observed effect of the dose reduction.
Twenty-four weeks after baseline, 95.2% of the patients in the RD group had a viral load below 50 copies per milliliter, compared with 86.1% of the patients who continued the SD (P = 0.15). Univariable logistic regression models showed that the pretreated status, the plasma viral load at baseline, the HIV transmission risk group, and the EFV concentration before baseline significantly affected whether or not patients had virological response. In the multivariable analysis, pretreated status and plasma viral load at baseline remained significantly associated with virological outcome. Patients who had undergone dose reduction had an adjusted odds ratio of 3.76 (95% confidence interval: 0.69 to 20.49, P = 0.13) for virological response when compared with patients who continued the SD.
Table 2 shows the adjusted odds ratios for achieving virological response at week 48 after baseline and at week 24 and week 48 after starting EFV. At most time points, patients in the RD group trended toward better virological response, although statistically significant differences were not observed. After stratification for pretreated status, dose reduction still had no negative impact on virological response (data not shown).
This study demonstrates that TDM-guided EFV dose reduction may prevent toxicity-induced discontinuations in patients with high plasma concentrations. This result is conforming our expectations. We hypothesized that by reducing EFV plasma concentrations, CNS toxicity would diminish, which would thereupon prevent toxicity-induced EFV discontinuations.
In addition, we did not observe any detrimental effect of dose reduction on virological response (Table 2). Pretreated status seemed to be the most important factor predicting virological response. Treatment-experienced patients with a detectable viral load (>50 copies/mL) performed significantly worse in all virological analyses, compared with either treatment-naive patients or treatment-experienced patients who had no detectable viral load when switching to EFV. Therefore, we repeated all virological analyses stratifying for pretreated status at the start of EFV-based antiretroviral therapy. Again, dose reduction was not associated with virological response in any of these strata. These results demonstrate that dose reduction is safe in patients with high EFV plasma concentrations, regardless of the pretreated status at the start of EFV.
Of importance, no patient decreased to subtherapeutic EFV plasma concentrations (ie, <1.0 mg/L) after dose reduction, which also confirms the safety of the dose reduction strategy. It is often stated that EFV trough levels should be at least 1.0 mg/L, but this is in fact not in line with the work of Marzolini et al4, who established the therapeutic window of EFV based on mid-dose interval plasma levels which were taken during the day, 8-20 hours after EFV administration at bedtime. In this study, we also used mid-dose interval plasma levels, which were taken on average 13 hours postdose, both before and after baseline.
Patients who underwent dose reduction had higher baseline EFV plasma concentrations (median 6.8 mg/L) than patients who continued the SD (median 5.1 mg/L). This difference may be caused by a tendency of physicians to decrease the dose with increasing plasma concentrations, but it may also be explained by increased EFV toxicity at higher plasma concentrations, resulting in a higher clinical necessity to adjust the dose. Despite this unbalance at baseline favoring the patients who continued the SD, the proportion of patients who stopped EFV because of toxicity was still higher in the latter group, demonstrating the effectiveness of the dose reduction strategy in those who are most in need of it.
Women and patients originating from sub Saharan Africa were overrepresented in the RD group. Previous studies indeed demonstrated higher EFV plasma concentrations in women5 and black African patients5,14,15. EFV is metabolized by the polymorphic cytochrome P450 2B6 enzyme,16,17 and black patients are known to have higher frequencies of certain cytochrome P450 2B6 polymorphisms (eg, 516 G > T and 983T > C), which are clearly associated with elevated EFV plasma concentrations.11,17,18 Possible causes for higher plasma concentrations in women are differences in body weight (in this cohort, female patients had an average weight of 63 kg compared with 73 kg for men), hormonal influences, and body composition.
The ATHENA cohort study does not collect data on the seriousness of drug toxicity. Thus, the only reliable endpoint available to evaluate the pharmacodynamic consequence of dose reduction was discontinuation of EFV due to toxicity. Because there were only 15 toxicity-induced discontinuations, we had limited statistical power, which is a limitation of our analysis. A study in which CNS toxicity had been scored systematically before and after dose reduction could have been more powerful. Nonetheless, discontinuation of a drug due to toxicity is the ultimate consequence of drug toxicity, and we consider this a clinically relevant endpoint.
Another limitation is the retrospective design of this analysis. The best evidence for the dose reduction strategy would come from a controlled trial with a prospective design, in which one would randomly assign subjects to a TDM group in which the results of EFV concentration measurements plus advice (eg, dose reduction) were reported to the treating physician or to a control group for whom TDM results were not reported. Because it is quite improbable that such a trial will be ever organized, we must rely on alternative evaluations of the potential benefits of TDM in HIV disease management.
In conclusion, our study demonstrates that TDM-guided dose reduction can be considered in patients who have high EFV plasma concentrations. Dose reduction does not negatively affect virological efficacy and may prevent toxicity-induced discontinuations.
1. Burger D, Hugen P, Reiss P, et al, for the ATHENA Cohort Study Group. Therapeutic drug monitoring of nelfinavir and indinavir in treatment-naive HIV-1-infected individuals. AIDS
2. Fletcher CV, Anderson PL, Kakuda TN, et al. Concentration-controlled compared with conventional antiretroviral therapy for HIV infection. AIDS
3. Best BM, Goicoechea M, Witt MD, et al. A randomized controlled trial of therapeutic drug monitoring in treatment-naive and -experienced HIV-1-Infected Patients. J Acquir Immune Defic Syndr
4. Marzolini C, Telenti A, Decosterd LA, et al. Efavirenz plasma levels can predict treatment failure and central nervous system side effects in HIV-1-infected patients. AIDS
5. Burger D, Van der Heiden I, laPorte C, et al. Interpatient variability in the pharmacokinetics of the HIV non-nucleoside reverse transcriptase inhibitor efavirenz: the effect of gender, race, and CYP2B6 polymorphism. Br J Clin Pharmacol
6. Khoo SH, Lloyd J, Dalton M, et al. Pharmacologic optimization of protease inhibitors and nonnucleoside reverse transcriptase inhibitors (POPIN)-a randomized controlled trial of therapeutic drug monitoring and adherence support. J Acquir Immune Defic Syndr
7. Staszewski S, Morales-Ramirez J, Tashima KT, et al. Efavirenz plus zidovudine and lamivudine, efavirenz plus indinavir, and indinavir plus zidovudine and lamivudine in the treatment of HIV-1 infection in adults. Study 006 Team. N Engl J Med
8. Gutierrez F, Navarro A, Padilla S, et al. Prediction of neuropsychiatric adverse events associated with long-term efavirenz therapy, using plasma drug level monitoring. Clin Infect Dis
9. Nunez M, Gonzalez dR, Gallego L, et al. Higher efavirenz plasma levels correlate with development of insomnia. J Acquir Immune Defic Syndr
10. Gallego L, Barreiro P, del Rio R, et al. Analyzing sleep abnormalities in HIV-infected patients treated with Efavirenz. Clin Infect Dis
11. Haas DW, Ribaudo HJ, Kim RB, et al. Pharmacogenetics of efavirenz and central nervous system side effects: an Adult AIDS Clinical Trials Group study. AIDS
12. La Porte CJ, Back DJ, Blaschke T, et al. Updated guideline to perform therapeutic drug monitoring for antiretroviral agents. Rev Antivir Ther
13. Gras L, van Sighem AI, Smit C, et al. Monitoring of Human Immunodeficiency Virus (HIV) Infection in The Netherlands. Report 2008. Available at: http://www.hiv-monitoring.nl
14. Stohr W, Back D, Dunn D, et al. Factors influencing efavirenz and nevirapine plasma concentration: effect of ethnicity, weight and co-medication. Antivir Ther
15. Pfister M, Labbe L, Hammer SM, et al. Population pharmacokinetics and pharmacodynamics of efavirenz, nelfinavir, and indinavir: Adult AIDS Clinical Trial Group Study 398. Antimicrob Agents Chemother
16. Ward BA, Gorski JC, Jones DR, et al. The cytochrome P450 2B6 (CYP2B6) is the main catalyst of efavirenz primary and secondary metabolism: implication for HIV/AIDS therapy and utility of efavirenz as a substrate marker of CYP2B6 catalytic activity. J Pharmacol Exp Ther
17. Rotger M, Tegude H, Colombo S, et al. Predictive value of known and novel alleles of CYP2B6 for efavirenz plasma concentrations in HIV-infected individuals. Clin Pharmacol Ther
18. Wyen C, Hendra H, Vogel M, et al. Impact of CYP2B6 983T>C polymorphism on non-nucleoside reverse transcriptase inhibitor plasma concentrations in HIV-infected patients. J Antimicrob Chemother