The In Vivo Response of Novel Buprenorphine Metabolites, M1 and M3, to Antiretroviral Inducers and Inhibitors of Buprenorphine Metabolism

Abstract

Buprenorphine metabolism was recently expanded by in vitro identification of a number of hydroxylated metabolites. The identification of two, M1 and M3, in urine suggests that they may be quantitatively significant metabolites. To further understand the in vivo regulation of this mode of metabolism, we evaluated 24-hr urine from subjects (10 per treatment group) on buprenorphine alone or with the antiretroviral agents: efavirenz, delavirdine, nelfinavir, ritonavir, and lopinavir/ritonavir. Quantitative analysis for buprenorphine and traditional metabolites and semi-quantitative analysis of M1 and M3 in urine were performed by liquid chromatography-electrospray ionization-tandem mass spectrometry. The renal clearance of buprenorphine and traditional metabolites were similar for all treatments except for lopinavir/ritonavir, suggesting that urine amounts of M1 and M3 would adequately reflect systemic changes (except lopinavir/ritonavir). Efavirenz decreased M1 and increased M3 consistent with its ability to induce cytochrome P450 (CYP) 3A. Delavirdine increased M1 and decreased M3 consistent with its ability to inhibit CYP3A. Both nelfinavir and ritonavir decreased both M1 and M3, consistent with their ability to inhibit CYP3A and 2C8. These results provide further information on the in vivo response of novel secondary metabolites of buprenorphine to metabolic inhibitors and inducers.

Buprenorphine is a partial μ-opioid agonist and K-opioid antagonist used at lower doses to treat moderate to severe pain and at higher doses as replacement therapy for opioid dependence. Primary metabolism of buprenorphine in humans occurs through N-dealkylation to norbuprenorphine and glucuronidation of both buprenorphine and norbuprenorphine [1]. Cytochrome P450 (CYP) of the 3A family was first shown to catalyse the N-dealkylation [2,3]; subsequent studies also demonstrated involvement of CYP2C8 [4,5]. Recently, up to five additional hydroxyl metabolites of buprenorphine (and norbuprenorphine) have been found after incubation of buprenorphine with human liver microsomes; two of these have been identified in the urine of subjects receiving therapeutic doses of buprenorphine for treatment of opioid dependence [5,6].

The two hydroxyl metabolites identified in human urine correspond to the hydroxylation of the hydroxyl-trimethyl propyl side chain of buprenorphine and norbuprenorphine and have been given the respective designations of M1 and M3. M1 was produced from buprenorphine by both CYP3A4/5 and 2C8; M3 was produced from both buprenorphine and norbuprenorphine, exclusively by CYP3A4/5. In urine, both are glucuronidated, but M1 much more so than M3. No evidence of sulfate conjugates of M1 or M3 was found [6]. The presence of identifiable amounts of these two metabolites in urine suggests they may be of quantitative importance to the clearance of buprenorphine, yet further characterization of these novel pathways of buprenorphine metabolism has been limited.

Recently, we have conducted experiments on the in vivo interaction between buprenorphine and a number of anti-retroviral agents [7,8]. The four drugs, and one drug combination, produced diverse, but usually significant changes in the pharmacokinetics of buprenorphine. The non-nucleoside reverse transcriptase inhibitors, efavirenz and delaviridine produced respective induction and inhibition of buprenorphine metabolism [7]. With the protease inhibitors, ritonavir increased exposure to buprenorphine, while neither nelfinavir nor the combination of lopinavir/ritonavir had any significant effect on the pharmacokinetics of buprenorphine.

During the course of these studies, 24-hr urines were collected during the pharmacokinetic sessions measuring buprenorphine alone, or buprenorphine and co-treatment with an antiretroviral agent. These urine samples were therefore available to examine the influence of the antiretrovirals on the urinary contents and metabolic ratios of the novel buprenorphine metabolites. Buprenorphine, norbuprenorphine, buprenorphine-3-glucuronide and norbuprenorphine-3-glucuronide were also measured to assess the effect of the antiretroviral agents on renal clearance of buprenorphine and its metabolites. These studies offer some new insights into the regulation of the hydroxylation pathways of buprenorphine and norbuprenorphine.

Materials and Methods

Materials

Buprenorphine, norbuprenorphine, d₄-buprenorphine and d₃-norbuprenorphine were purchased from Cerilliant (Round Rock, TX, USA). β-glucuronidase (from Helix pomatia, which also has sulfatase activity) was purchased from Sigma-Aldrich (St. Louis, MO, USA).

Treatment

Detailed descriptions of the treatments with buprenorphine and the antiretroviral agents are presented in our previous publications [7,8], and will only briefly be described at this point. Ten subjects participated in each interaction study. All had a history of opioid dependence and were treated with buprenorphine/naloxone. With one exception, all were maintained on daily doses of 16 mg buprenorphine, 4 mg naloxone; a single subject in the efavirenz study was maintained on 20 mg buprenorphine/5 mg naloxone. All subjects were on a stable maintenance dose for at least 2 weeks before the first (buprenorphine only) 24-hr pharmacokinetic session. This was followed by initiation of standard clinical doses of anti-retroviral medications used in the treatment of HIV disease including: efavirenz 600 mg daily for 15 days, delavirdine 600 mg twice daily for 5 days, nelfinavir 1250 mg twice daily for 5 days, ritonavir 100 mg twice daily for 7 days, and 400/100 mg lopinavir/ritonavir twice daily for 7 days. On several occasions, the first antiretroviral treatment was followed by a 2-week or more wash-out period and then treatment with a second antiretroviral agent. As such, there were only 31 initial buprenorphine only pharmacokinetic sessions. On the last day of antiretroviral treatment, subjects underwent a second 24-hr pharmacokinetic session. Bloods were drawn to prepare heparinized plasma over a 24-hr period as previously described [7,8]. All urine was collected over the same period, the volume measured and aliquots taken for storage at –20°C until analysis. All subjects provided voluntary signed informed consent. The studies were approved by the local institutional review board.

Analyte determination

All analyte concentrations (or relative peak areas for M1 and M3) were determined using liquid chromatography coupled by electrospray ionization to tandem mass spectrometry (LC-ESI-MSMS). A TSQ Quantum triple quadrupole MS, equipped with a Surveyor binary LC pump and a Surveyor autosampler (Thermo Electron, San Jose, CA, USA), was used for all analyses. Quantitative analyses of urine samples for buprenorphine and the traditional metabolites, norbuprenorphine, buprenorphine-3-glucuronide and norbuprenorphine-3-glucuronide used our previously described method [9].

The semi-quantitation of M1 and M3 was performed as previously described [6]. Urine samples first underwent hydrolysis with a β-glucuronidase and then underwent liquid/liquid extraction and LC-ESI-MSMS using conditions previously described for buprenorphine and norbuprenorphine [4]. Buprenorphine-d₄ was used as the internal standard.

Pharmacokinetic determinations and statistics

The amount of urine metabolites was first determined from concentration times the volume; renal clearance was calculated as amount in urine_0–24 divided by the AUC_0–24 determined in the previous studies [7,8]. Student's paired t-test was used for comparison of buprenorphine to buprenorphine plus antiretroviral groups. With statistical significance set at P < 0.05 (two-tailed).The metabolic ratiosforbuprenorphine and the traditional metabolites were determined to measure oxidative metabolism (buprenorphine+buprenorphine-3-glucuronide)/(norbuprenorphine+norbuprenorphine-3-glucuronide)orglucuronidation(buprenorphine+norbuprenorphine)/(buprenorphine-3-glucuronide + norbuprenorphine-3-glucuronide); that for M1 was determined versus its parent compound, (M1/buprenorphine); that for M3 versus its two precursors (M3/nor buprenorphine + norbuprenorphine).

Results and Discussion

Our previous presentation of urine contents of buprenorphine and metabolites was limited to results from five subjects for buprenorphine, norbuprenorphine, buprenorphine-3-glucuronide and norbuprenorphine-3-glucuronide [9], and seven subjects for M1 and M3 [6]. Due to increased use of urinalysis to monitor buprenorphine use, these results are useful; at this time, we have expanded their presentation to the 31 control (buprenorphine only) results from this study. Figure 1A shows the distribution of the amounts of these buprenorphine, norbuprenorphine, buprenorphine-3-glucuronide and norbuprenorphine-3-glucuronide in the thirty-one control (buprenorphine only) 24-hr urine collections. A wide variation in amounts is seen. The amounts of buprenorphine and norbuprenorphine-3-glucuronide represent the minimal and maximal amounts of analytes found in urine, as previously described [9]. The amounts of norbuprenorphine and buprenorphine-3-glucuronide do not differ by much. In this larger population, the mean amount of buprenorphine-3-glucuronide is now slightly greater than that of norbuprenorphine; this is due in part to the extraordinarily high amounts of buprenorphine-3-glucuronide in 4 samples (fig. 1A). Without standard reference material for M1 and M3, we can only provide semi-quantitative amounts (peak area ratios times urine volume). Variation in M1 and M3 contents is also seen, with two to four samples providing a majority of the variance (fig. 1B).

Fig. 1.

Scatter plot of: A) individual amounts (μg) of buprenorphine (Bup), norbuprenorphine (Nor), buprenorphine-3-glucuronide (B3G) and norbuprenorphine-3-glucuronide (N3G) and B) relative amount (peak area ratio times volume) of novel buprenorphine metabolites M1 and M3 in 24-hr urines collected from subjects taking only buprenorphine (N = 31). These values represent the total results for all buprenorphine only sessions. Mean results for appropriate treatment comparisons are presented in table 1.

Since M1 and M3 cannot be measured in plasma, use of urine findings is important, but must still be approached with appropriate cautions in regard to other factors that may alter urine excretion. Calculation of renal clearance provides a comparison of the plasma AUC and urine content of the drug or metabolite. No significant changes in renal clearance of buprenorphine, norbuprenorphine, buprenorphine-3-glucuronide, or norbuprenorphine-3-glucuronide were caused by co-treatment with efavirenz, delavirdine, nelfinavir or ritonavir (data not shown). The lopinavir/ritonavir mixture, however, significantly increased the renal clearance of buprenorphine, norbuprenorphine and norbuprenorphine-3-glucuronide (data not shown). The mechanism for increased renal clearance by the combined antiretroviral, lopinavir/ritonavir is not known. The same combination altered renal clearance of the nucleotide reverse transcriptase inhibitor, tenofovir [10] in a manner that could not be attributed to inhibition of the characterized human organic anion transporter 1 secretion of tenofovir [11]. The change in drug clearance in this case was a decrease. So while the Kisar et al. study and our current study both show that lopinavir/ritonavir can alter renal clearance of drugs or drug metabolites, neither demonstrates a mechanism. There was a lack of similar effect by ritonavir alone in this study. This does point out that the induction of renal clearance seen here requires either lopinavir or the combination of lopinavir/ritonavir.

No co-treatment had any effect on the volume of urine excreted over the 24-hr collection period (mean volumes in buprenorphine only groups ranged from 1.86 to 2.36 l; those for the buprenorphine plus antiretroviral groups ranged from 1.95 to 2.50 l). These findings suggest that the urinary amounts of M1 and M3 will accurately reflect their systemic changes, with the exception of lopinavir/ritonavir treatment. The effect of the lopinavir/ritonavir on M1 and M3 were therefore not presented, since other factors could be involved in changes of their urine amounts.

Efavirenz significantly decreased the amount of buprenorphine, norbuprenorphine, buprenorphine-3-glucuronide and norbuprenorphine-3-glucuronide in urine. These changes are consistent with the previously reported changes in plasma AUC [7]. Changes in the metabolic ratios suggest that efavirenz had a more significant effect on relative glucuronidation (table 1). After efavirenz, M1 trended downward while M3 amount trended upward, these changes were not significant (table 1). The decrease in M1 mirrored a decrease in buprenorphine, such that the metabolic ratio (M1/buprenorphine) was not significantly changed; the increase trend in M3 was coupled with decreases in buprenorphine and norbuprenorphine, such that the metabolic ratio [(M3)/(buprenorphine + norbuprenorphine)] was significantly increased (table 1). While efavirenz has shown some ability to inhibit the activity of CYP3A4/5 and 2B6 in human liver microsomes [12,13], it is an inducer of CYP 3A4/5 and 2B6 in cultured human hepatocytes [14,15]. The responses of M1 and M3 to efavirenz are consistent with an induction of all pathways leading to M3 (Scheme 1A).

Table 1.

Response of buprenorphine and metabolites in urine to antiretroviral treatments.

		Antiretroviral in session 2
Analyte(s)	Session	Efavirenz	Delavirdine	Nelfinavir	Ritonavir
		Amounts (μg)
Bup	1	4.73 ± 2.62	3.06 ± 2.72	3.76 ± 3.07	2.58 ± 1.88
	2	2.63 ± 3.44*	10.5 ± 4.6‡	5.09 ± 2.75	5.08 ± 4.39*
Nor	1	231 ± 116	215 ± 148	224 ± 136	203 ± 152
	2	63.2 ± 50.3‡	65.5 ± 53.4†	192 ± 103	209 ± 136
B3G	1	263 ± 114	281 ± 216	311 ± 188	257 ± 219
	2	110 ± 62‡	714 ± 408*	274 ± 149	327 ± 171
N3G	1	1495 ± 635	1079 ± 638	1423 ± 594	921 ± 694
	2	968 ± 861*	217 ± 143†	1392 ± 736	1153 ± 590
		Amounts (peak area ratio times volume)
M1	1	190 ± 108	191 ± 183	199 ± 198	195 ± 184
	2	104 ± 124	345 ± 236	54.7 ± 57.0*	53.9 ± 35.5*
M3	1	206 ± 241	183 ± 175	177 ± 128	197 ± 169
	2	264 ± 278	16.6 ± 29.8*	48.5 ± 75.4*	26.9 ± 28.4*
		Metabolic Ratios
(B + B3G)/(N + N3G)	1	0.16 ± 0.07	0.21 ± 0.06	0.20 ± 0.08	0.23 ± 0.11
	2	0.14 ± 0.06	3.51 ± 3.27*	0.18 ± 0.05	0.26 ± 0.13
(B + N)/(B3G + N3G)	1	0.14 ± 0.06	0.19 ± 0.12	0.13 ± 0.06	0.21 ± 0.11
	2	0.07 ± 0.02*	0.09 ± 0.08*	0.12 ± 0.07	0.16 ± 0.08
M1/B	1	88 ± 120	175 ± 177	96 ± 105	175 ± 175
	2	295 ± 399	42.6 ± 49.6*	11.3 ± 13.1*	14.3 ± 8.9*
M3/(B + N)	1	0.88 ± 0.9	21.2 ± 1.6	1.21 ± 1.64	1.30 ± 1.55
	2	4.82 ± 3.66*	0.05 ± 0.08*	0.18 ± 0.23*	0.09 ± 0.10*

Note: Values are mean ± SD. Subjects participated in session 1 (buprenorphine alone) and session 2 (buprenorphine plus antiretroviral) that were separated by 5 to 7 days of treatment with the antiretroviral; treatment times that were sufficient to reach steady state. Subjects served as their own controls; statistical differences were determined using the paired t-test

^*P < 0.05

^†P < 0.005

^‡P < 0.0005.

Scheme 1.

Speculative changes in metabolism of M1 and M3 that may arise from treatment with A) efavirenz, B) delavirdine or C) nelfinavir and ritonavir.

Following delavirdine, urine buprenorphine and buprenorphine-3-glucuronide were significantly increased while urine norbuprenorphine and norbuprenorphine-3-glucuronide were significantly decreased (table 1). These changes are consistent with the previously reported changes in plasma AUC [7], and the metabolic ratio changes are consistent with a greater impact on oxidative metabolism (table 1). M1 showed a tendency to increase, while M3 was significantly decreased, as were the metabolic ratios (M1/buprenorphine) and [(M3)/(buprenorphine + norbuprenorphine)] (table 1). Delavirdine inhibited buprenorphine N-demethylation to norbuprenorphine, as it did the conversion of M1, or norbuprenorphine, to M3. Inhibition of M3 formation is consistent with the ability of delavirdine to inhibit CYP3A4/5 activities in human liver microsomes [12,16,17]. Both norbuprenorphine and M1 are produced by CYP3A4/5 and 2C8; 2C8 appeared more important for M1 [6]. Since formation of norbuprenorphine is significantly reduced while that of M1 is not, this is consistent with a specificity of delavirdine to inhibit CYP3A4 and the greater formation of M1 by CYP2C8 (see Scheme 1B).

Both nelfinavir and ritonavir show a trend to increase urine buprenorphine (significant for ritonavir) without a major effect on the other traditional metabolites, and with no significant effects on the metabolic ratios (table 1), consistent with previous findings in plasma [8]. With both nelfinavir and ritonavir, M1 and M3 content decrease, as do their metabolic ratios (table 1). The decreases of M1 and M3 in response to ritonavir and nelfinavir are consistent with the ability of these protease inhibitors to inhibit both CYP3A4 and 2C8 [18–20]. In this case, the decreased formation of M1 but not of buprenorphine would have to be explained by a greater potency of inhibition of CYP2C8 by these protease inhibitors (Scheme 1C). This is consistent with the findings of Dixit et al. [20] where the protease inhibitors show mixed inhibition and induction of CYP3A4 and 2C8, with a greater tendency for inhibition of 2C8.

This study furthers our knowledge concerning the metabolism of buprenorphine. The lack of standard reference material for M1 and M3 precludes quantitative determination of their content, which in turn diminishes the clinical significance of these findings. For the first time, however, we have examined the response of the novel metabolites, M1 and M3, to in vivo modulators of CYP metabolism. The response of M1 and M3 in the urine of subjects taking both buprenorphine and the inducer and inhibitor antiretroviral agents supports our earlier studies on the involvement of CYP3A4/5 and 2C8 in their formation.