Rapid Clinical Induction of Hepatic Cytochrome P4502B6 Activity by Ritonavir

MATERIALS AND METHODS

RESULTS

One subject withdrew (for scheduling reasons) during the steady-state ritonavir phase of the protocol; thus, data were not obtained for this subject session. Results are presented for 13 subjects (control and 3 days of ritonavir) and for 12 subjects (2.5 weeks of ritonavir).

Ritonavir caused significant changes in bupropion disposition. Plasma concentrations of both (R)- and (S)-bupropion were diminished after both 3 days of ritonavir treatment and steady-state ritonavir treatment (Fig. 1). The ratios of plasma AUC from 0 h to infinity (AUC_0-∞) (ritonavir/control) for racemic, (R)-, and (S)-bupropion were significantly reduced to 0.84, 0.86, and 0.80, respectively, after 3 days of ritonavir treatment and to 0.67, 0.69, and 0.60 after more than 2 weeks of ritonavir treatment (Table 1). (R)-, (S)-, and racemic bupropion CL/Fs all were significantly increased by 1.2-fold after 3 days of ritonavir treatment and by 1.4-, 1.7, and 1.5-fold after steady-state ritonavir treatment. Short-term ritonavir decreased the plasma AUC_0-∞ of total and (R,R)-hydroxybupropion but not (S,S)-hydroxybupropion, while steady-state ritonavir treatment decreased the plasma AUC_0-∞ of total, (R,R)-, and (S,S)-hydroxybupropion (Fig. 2). The median apparent elimination half-life of total, (R,R)-, and (S,S)-hydroxybupropion was reduced from 32, 33, and 18 h, respectively, to 17, 18, and 13 h by 3 days of ritonavir treatment and to 11 h for all three by steady-state ritonavir treatment. The plasma hydroxybupropion/bupropion AUC_0-48 ratios for racemic and (R)-bupropion were minimally changed or were unchanged by both short-term and steady-state ritonavir treatments, while the metabolite/parent ratio for (S)-bupropion was increased by both short-term and steady-state ritonavir treatments. Ritonavir caused a 1.8-fold increase in the formation clearance of both (R,R)- and (S,S)-hydroxybupropion after 3 days and a 2.1-fold increase after 2 weeks (Table 2). Similar results were observed for the 0- to 24-h urine hydroxybupropion/bupropion molar ratio. Ritonavir had no effect on bupropion enantiomer renal clearance.

• Open in new tab
• Download powerpoint

FIG. 1.

Plasma concentrations of racemic bupropion (A) and bupropion enantiomers (B and C). Results are the means ± SD (n = 13 for controls and acute ritonavir treatment; n = 12 for steady-state ritonavir treatment). Some SD are omitted for clarity.

• Open in new tab
• Download powerpoint

FIG. 2.

Plasma concentrations of racemic hydroxybupropion (A) and hydroxybupropion diastereomers (B and C). Results are the means ± SD (n = 13 for controls and acute ritonavir treatment; n = 12 for steady-state ritonavir treatment). Some SDs are omitted for clarity.

DISCUSSION

This investigation demonstrates that short-term (3 days) and steady-state ritonavir treatments significantly increased (S)-bupropion hydroxylation. Specifically, ritonavir increased (S,S)-hydroxybupropion apparent formation clearance, 0- to 24-h urine (S,S)-hydroxybupropion/(S)-bupropion molar ratios, plasma (S,S)-hydroxybupropion/(S)-bupropion AUC ratios, and (S)-bupropion CL/F. Changes in (S)-bupropion clearance were not due to increased bupropion renal clearance. (S)-Bupropion hydroxylation is catalyzed selectively by CYP2B6 and an in vivo probe for CYP2B6 (32). Therefore, these results strongly support the conclusion that human hepatic CYP2B6 activity in vivo is induced by both short-term and steady-state ritonavir treatments, with the latter causing an approximately twofold increase. Although ritonavir also increased apparent formation clearances, urine hydroxybupropion/bupropion molar ratios, and systemic clearance of (R,R)- and racemic hydroxybupropion, plasma hydroxybupropion/bupropion AUC ratios were decreased rather than increased. There is insufficient understanding of hydroxybupropion elimination to explain this last observation (32).

One previous investigation evaluated the effects of ritonavir on bupropion disposition (24). Subjects received bupropion (presumably the immediate-release variety, although it was not specified) on the second day of treatment with 200 mg ritonavir twice daily. Racemic bupropion and metabolite plasma concentrations were measured for 24 h. Neither bupropion nor hydroxybupropion concentrations were significantly altered by ritonavir, and the hydroxybupropion/bupropion ratio was insignificantly decreased by 15%, which was interpreted as suggesting no effect on CYP2B6 activity. There are, however, differences between previous studies and the present study. Hesse et al. evaluated racemic bupropion plasma concentrations and the hydroxybupropion/bupropion AUC ratio (24). Nevertheless, plasma bupropion concentrations may be insensitive to altered CYP2B6 activity, because CYP2B6-catalyzed hydroxylation is only a minor route of overall bupropion elimination (36, 50). Additionally, the plasma hydroxybupropion/bupropion AUC ratio may be confounded, because racemic hydroxybupropion is elimination rate rather than formation rate limited, and hydroxybupropion elimination also may be altered (32, 36). Hesse et al. also evaluated bupropion metabolism after fewer and lower doses of ritonavir (three to four doses of 200 mg) than those used herein (3 days of two doses of 200 mg and three to four doses of 300 mg and 2.5 weeks of 400 mg twice daily) and speculated that longer ritonavir exposure might induce CYP2B6. The present investigation demonstrates that both short-term and steady-state ritonavir treatments do induce hepatic CYP2B6 activity.

The ritonavir induction of hepatic CYP2B6 is consistent with previous in vitro findings. Ritonavir increased CYP2B6 mRNA expression in LS180 adenocarcinoma cells (23) and induced CYP2B6 mRNA expression, immunoreactive protein expression, and catalytic activity in primary human hepatocytes (8, 15). Ritonavir was classified as a strong in vitro CYP2B6 inducer, along with rifampin, phenobarbital, phenytoin, and clotrimazole (15).

Other CYP2B6 inducers also have been shown to induce the metabolism of racemic bupropion and other CYP2B6 substrates. The prototypic CYP2B6 inducer rifampin induced bupropion clearance by threefold and increased peak plasma hydroxybupropion concentrations (36). Lopinavir/ritonavir (400 mg lopinavir/100 mg ritonavir twice daily for 2 weeks) decreased the plasma AUC for bupropion and hydroxybupropion by 57 and 50%, respectively (27). Ritonavir (500 mg twice daily for 10 days) decreased the plasma meperidine AUC and increased the normeperidine AUC, suggesting the induction of metabolism (43). Although this was initially attributed to CYP3A induction (43), subsequent reports showing that meperidine is primarily a CYP2B6 substrate (44) and that ritonavir does not induce CYP3A4 (see below) suggest that the ritonavir enhancement of meperidine metabolism and clearance likely reflects CYP2B6 induction.

The ritonavir induction of hepatic CYP2B6 activity has clinical implications, specifically with regard to the disposition of numerous CYP2B6 substrates, including antineoplastic agents, anesthetics and analgesics, drugs of abuse, antiretrovirals, pesticides, and other environmental contaminants, together encompassing as much as 8% of all therapeutic drugs (39, 49, 54). One example is the ritonavir induction of nelfinavir metabolism, evidenced by increased plasma ratios of metabolite M8/nelfinavir (1). Another notable example is methadone, a mainstay in the treatment of opiate addiction and acute and chronic pain. Ritonavir increases methadone clearance and can cause acute withdrawal (19, 30, 34, 38). Both short-term and steady-state ritonavir treatments recently were shown to increase intravenous and oral methadone metabolism and clearance by 1.5- to 2-fold (30, 34). This occurred despite the profound ritonavir inhibition of CYP3A. The mechanism by which ritonavir induces methadone clearance is unknown. Methadone metabolism and clearance in vivo traditionally have been attributed to CYP3A (5, 10, 18) and ritonavir effects on methadone disposition to CYP3A induction (19, 38). Notwithstanding years of the clinical use of these drugs, the paradox of the ritonavir induction of methadone clearance despite CYP3A inhibition remains unexplained. Recent evidence suggests a significant, if not predominant, role for CYP2B6 in methadone metabolism and clearance (20, 31, 47, 48). CYP3A inhibition failed to diminish methadone metabolism and clearance (30, 31, 34). Conversely, rifampin, which induces CYP2B6, increased methadone metabolism and clearance at a level similar to that by ritonavir. This suggests that the ritonavir induction of methadone metabolism and clearance may be mediated by hepatic CYP2B6 induction.

The clinical effects of ritonavir on human hepatic CYP2B6 are diametrically opposite to those on hepatic and intestinal CYP3A. Ritonavir is one of the most potent and efficacious mechanism-based inhibitors of CYP3A isoforms in vitro (13). Clinically, acute and steady-state ritonavir treatments profoundly inhibit first-pass and hepatic CYP3A. Several recent investigations, using well-characterized CYP3A probes such as midazolam, triazolam, and alfentanil, demonstrated >95% inhibition of first-pass CYP3A activity by various ritonavir regimens, such as 100 to 200 mg/day, 200 mg twice daily, and 400 mg twice daily (6, 17, 22, 30, 34). Although early clinical investigations suggested that steady-state ritonavir treatment induced CYP3A based on the autoinduction of ritonavir clearance (29) and decreased ethinyl estradiol AUC and half-life (40), nevertheless it was recognized that such effects also could emanate from ritonavir induction of efflux transporters (6).

The contradictory effects of ritonavir on human hepatic CYP2B6 and CYP3A activities are paradoxical, particularly since these isoforms share common regulatory pathways. The orphan nuclear receptors pregnane X receptor (PXR) and constitutive androstane receptor (CAR) both regulate the expression of CYP2B and CYP3A (51). Early investigations suggested that PXR and CAR were the predominant activators of CYP3A and CYP2B, respectively, although later findings of cross-talk between PXR and CAR demonstrated the reciprocal activation of their response elements (52). Cross-talk is asymmetrical, so that the activation of human PXR, which has a large and accommodating active site that binds numerous and diverse ligands (2, 11), results in the nonselective activation of both CYP2B6 and CYP3A4 (14). Conversely, human CAR, which has a smaller binding site (53), preferentially activates CYP2B6 over CYP3A4 (14). A recent screen of CYP2B6 and/or CYP3A4 inducers identified three classes: selective human PXR activators (i.e., rifampin), mixed human PXR and CAR activators (i.e., troglitazone), and preferential human CAR activators (i.e., efavirenz and nevirapine) (16). One possible explanation for reciprocal CYP2B6 induction/CYP3A inhibition by ritonavir is PXR activation combined with mechanism-based CYP3A inhibition. In human hepatocytes, ritonavir strongly activated PXR, upregulated CYP2B6 and CYP3A mRNA levels, and induced both CYP3A and CYP2B6 protein expression (8, 9, 15, 23, 37). Ritonavir was the most potent and efficacious PXR activator (15, 23), although it was not as efficacious in inducing CYP3A4 as predicted by PXR activation (23, 37). More importantly, while ritonavir induced both CYP2B6 and CYP3A4 protein expression, CYP3A activity was completely inhibited (8, 15, 37). Thus, the PXR-mediated coordinate upregulation of CYP2B6 and CYP3A4 expression by ritonavir was offset by the selective inhibition of CYP3A4 and not CYP2B6. Ritonavir is a potent (K_i < 0.1 μM) mechanism-based CYP3A4 inhibitor (13, 21). It appears to be a less potent inhibitor of CYP2B6 (50% inhibitory concentration, 2 μM) (25). A similar in vivo scenario could result in the upregulation of both CYP3A4 and CYP2B6 expression, with the selective inhibition of CYP3A4 activity. Another possible explanation for the reciprocal clinical ritonavir induction of CYP2B6 and inhibition of CYP3A is that ritonavir is a preferential human CAR activator, resulting in selective CYP2B6 upregulation and likely also CYP3A4 inhibition. The antiretrovirals efavirenz and nevirapine have been characterized as preferential human CAR activators (16); however, the effects of ritonavir on CAR have not been reported.

One potential limitation of this investigation is that the steady-state ritonavir dose differs from that used in typical contemporary antiretroviral regimens (42). In early clinical use, ritonavir was dosed at 600 mg twice daily, although this dose was associated with considerable drug interactions and generally is no longer used for treating HIV-infected patients. Contemporary use is predominately in lower doses, typically 100 to 200 mg once or twice daily. The present protocol was designed as a mechanistic investigation specifically to provide insights into the mechanism of the ritonavir induction of methadone clearance in our previous investigation using the same dose of 400 mg twice daily (30, 34). Nonetheless, CYP2B6 induction occurred at higher (400 mg twice daily) as well as lower doses (after only 2 days of 200 to 300 mg twice daily) and also after longer periods at 100 mg twice daily (27). Thus, the present investigation has both mechanistic and clinical relevance to ritonavir use at lower doses.

Ritonavir prescribing guidelines suggest that concomitant use with bupropion increases, or is predicted to increase, bupropion concentrations. Caution is advised if bupropion is administered with ritonavir, with a dosage reduction potentially being necessary (Norvir [ritonavir] soft gelatin capsules and oral solution prescribing information, Abbott Laboratories, North Chicago, IL). This may be based in part on extrapolation from the in vitro ritonavir inhibition of human liver microsomal bupropion hydroxylation (25). The present study demonstrates the need for caution in extrapolating in vitro observations and the necessity of performing clinical investigations to verify such observations, and it suggests that a revision of the ritonavir dosing guidelines is appropriate.

In summary, the present investigation shows that short-term (2 to 3 days) and steady-state ritonavir treatments increase the hydroxylation and clearance of the CYP2B6 probe (S)-bupropion, suggesting that ritonavir induces human CYP2B6 activity. CYP2B6 induction occurs despite the profound inhibition of CYP3A4 activity.