Bioorganic & Medicinal Chemistry Letters 16 (2006) 691–694
`
`Derivatives of tramadol for increased duration of effect
`
`Liming Shao,* Craig Abolin, Michael C. Hewitt, Patrick Koch and Mark Varney
`
`Sepracor Inc., 84 Waterford Drive, Marlborough, MA 01752, USA
`
`Received 23 August 2005; revised 6 October 2005; accepted 7 October 2005
`Available online 27 October 2005
`
`Abstract—Tramadol is a centrally acting opioid analgesic structurally related to codeine and morphine. Analogs of tramadol with
`deuterium-for-hydrogen replacement at metabolically active sites were prepared and evaluated in vitro and in vivo.
`Ó 2005 Elsevier Ltd. All rights reserved.
`
`Tramadol (±)-cis is used for the treatment of moderate
`to moderately severe pain.1 Tramadol has a relatively
`short duration of analgesic effect due to extensive first
`pass metabolism and as such is dosed as frequently as
`100 mg every 4–6 h.2 The analgesic activity of tramadol
`is thought to be the result of a dual mechanism—the
`parent (1) acts as an inhibitor of norepinephrine and
`serotonin reuptake, and the major O-desmethyl metabo-
`lite (M1) is a potent agonist at l-opioid receptors
`(Fig. 1).3
`
`The opioid antagonist naloxone only partially inhibits
`the analgesic activity of tramadol in animal tests. Not
`unlike other opioids, tramadol causes a number of side
`effects including constipation, nausea, dizziness, and
`somnolence. A tramadol analog that had a longer
`half-life (and therefore required less-frequent dosing)
`and was devoid of opioid side effects would have thera-
`peutic benefit. Herein, we wish to disclose our efforts to
`
`develop an improved tramadol analog with a longer
`half-life.
`
`Our approach to tramadol analogs took advantage of
`the primary kinetic isotope effect4 to slow CYP450-med-
`iated metabolism (Fig. 2). Replacing hydrogen with deu-
`terium at metabolically active sites can result in a slower
`metabolism due to the reduced rate of cleavage of a C–D
`bond relative to a C–H bond. This approach has been
`shown to be effective for a number of pharmacological
`agents including amphetamine,5 butethal,6 and mor-
`phine.7 For tramadol, formation of metabolite M1 is
`both the primary route of metabolism and is responsible
`for its opioid-like effects. By slowing down metabolism
`of tramadol, it was anticipated that the formation of
`M1 would be reduced, providing a longer-acting drug
`with potentially fewer opioid-related side effects. To-
`ward that end, the hydrogen atoms on the O-methyl
`and N-methyl groups were replaced by deuterium as
`shown in structures D3, D6, D9, and D6 M1.
`
`OCH3
`
`OH
`
`OH
`
`CH3
`N
`CH3
`
`OH
`
`CH3
`N
`CH3
`
`1
`
`M1
`
`Figure 1. (±)-Tramadol 1 and metabolite (±)-M1.
`
`Keywords: Tramadol; Metabolic stability; Deuterium isotope effect;
`Opioids.
`* Corresponding author. Tel.: +1 5083577467; fax: +1 5083577467;
`e-mail: liming.shao@sepracor.com
`
`0960-894X/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved.
`doi:10.1016/j.bmcl.2005.10.024
`
`OR1
`
`OH
`
`R2
`
`N
`R2
`
`D3 R1 = CD3, R2 = CH3
`D6 R1 = CH3, R2 = CD3
`D6 M1 R1 = H, R2 = CD3
`D9 R1 = CD3, R2 = CD3
`Dramadol analogs
`
`Figure 2. D-tramadol (dramadol) analogs.
`
`Auspex Exhibit 2011
`Apotex v. Auspex
`IPR2021-01507
`Page 1
`
`

`

`692
`
`L. Shao et al. / Bioorg. Med. Chem. Lett. 16 (2006) 691–694
`
`R1
`
`1. Mg, THF, 4 or 5
`2. HCl, Et2O
`
`OR2
`
`O
`
`N
`R1
`2 R1 = CH3
`3 R1 = CD3
`
`OR2
`
`Br
`4 R2 = CH3
`5 R2 = CD3
`
`1. n-BuLi, PHPh2, THF
`2. HCl, Et2O
`
`HCl
`
`OH
`
`R1
`
`N
`R1
`
`6 (D3) R1 = CH3, R2 = CD3
`7 (D6) R1 = CD3, R2 = CH3
`
`OH
`
`OH
`
`R1
`
`N
`R1
`
`HCl
`
`8 (D6 M1) R1 = CD3
`
`Scheme 1. Synthesis of D3, D6, and D6 M1 metabolites.
`
`CD3
`N
`CD3
`
`1. t-BuLi, Et2O, 5
`2. HCl, Et2O
`OCD3
`
`3
`
`OCD3
`
`OH
`
`HCl
`
`CD
`N
`CD3
`
`Br
`
`9 (D9)
`
`Scheme 2. Synthesis of 9 (D9).
`
`the deuterated tramadol analogs
`The syntheses of
`were relatively straightforward and began with Man-
`nich base 2 or 3 (Scheme 1). Specifically, addition of
`the Grignard reagent 4 or 5 to ketones 2 or 3 provid-
`ed crude 6 (D3) and 7 (D6) as mixtures of diastereo-
`mers (76% cis for D6). The abundance of the cis
`isomer could be improved to 93% after formation of
`the HCl salt and multiple recrystallizations from iso-
`propanol.8 Demethylation of 7 (D6) occurred without
`incident to provide phenol 8 (D6 M1).9 Purification
`was effected via formation of the HCl salt and recrys-
`tallization from isopropanol/ethyl acetate to give the
`phenol salt as an analytically pure white powder.
`The reaction to form D9 was attempted several times
`with starting materials 3 and 5, with little product for-
`
`mation seen by LCMS and HPLC analysis. This led
`to an alternate strategy to synthesize D9 that relied
`on a lithium–halogen exchange reaction that had been
`outlined previously in the literature,10 and was suc-
`cessfully employed using Mannich base 3 and aromat-
`ic bromide 5 (Scheme 2).
`
`Crude 9 (D9) was also converted to its HCl salt and
`recrystallized several times in isopropanol to provide a
`93% pure cis isomer.11
`
`The receptor binding affinity and functional monoamine
`uptake inhibition by tramadol and the deuterated tram-
`adol analogs are shown in Table 1.
`
`As expected, the deuterated analog of M1 retained the
`l-opioid receptor binding affinity present in metabolite
`M1. In general, the deuterated analogs retained in vitro
`activity comparable to their non-deuterated parent mol-
`ecules, with a slightly diminished reuptake inhibition of
`norepinephrine compared to tramadol.
`
`The in vitro metabolic stability results for the deuterated
`tramadol compounds are detailed in Table 2.12,13 Based
`on disappearance of parent compound in incubation
`media, the most stable compound of the six tested in hu-
`man liver microsomes and hepatocytes was clearly the
`D9-tramadol derivative 9. Tramadol and D6-tramadol
`
`Table 1. l-Opioid binding, and 5-HT and NE reuptake inhibition of (±)-cis tramadol and deuterated derivatives
`OR2
`
`Compound
`
`1 Tramadol
`Metabolite M1
`6 (D3)
`7 (D6)
`8 (D6 M1)
`9 (D9)
`
`R1
`
`CH3
`CH3
`CH3
`CD3
`CD3
`CD3
`
`R2
`
`CH3
`H
`CD3
`CH3
`H
`CD3
`
`OR3
`
`R1
`
`N
`R1
`
`R3
`
`H
`H
`H
`H
`H
`H
`
`l
`
`7600
`47
`>10,000
`>10,000
`43
`5300
`
`IC50 (nM)
`
`5-HT
`
`4300
`4600
`1900
`3100
`9900
`1100
`
`NE
`
`790
`>10,000
`3600
`3200
`6700
`2000
`
`Note: IC50 is the concentration required to inhibit 50% binding to l-opioid receptors or reuptake of 5-hydroxytryptamine (5-HT) and norepinephrine
`(NE).
`
`Auspex Exhibit 2011
`Apotex v. Auspex
`IPR2021-01507
`Page 2
`
`

`

`L. Shao et al. / Bioorg. Med. Chem. Lett. 16 (2006) 691–694
`
`693
`
`hepatocytes and microsomes in Figures 3 and 4, respec-
`tively. Consistent with a deuterium isotope effect on the
`metabolically labile O-methyl group, 6 (D3) and 9 (D9)
`slowed the formation of the primary metabolite M1 by
`approximately 5-fold. The formation rate of metabolite
`M1 from 7 (D6) was nearly indistinguishable from that
`of tramadol in human hepatocytes.
`
`The in vivo activities of tramadol (1), 7 (D6), and 9 (D9)
`were tested in the rat tail-flick model. Tramadol (1), 7
`(D6), and 9 (D9) were administered at a dose of
`50 mg/kg intraperitoneally. Intraperitoneal morphine
`(5 mg/kg) and saline were used as controls. Analgesic
`activity (tail-flick latency) was evaluated at 1, 2, 4, and
`8 h after the dose. The results of efficacy (tail-flick laten-
`cy) are shown in Figure 5.
`
`Tramadol and two of its deuterated derivatives (7 (D6)
`and 9 (D9)) significantly lengthened tail-flick latency
`when tested at 1 and 2 h after intraperitoneal adminis-
`tration of 50 mg/kg. Tramadol and 9 (D9) were also
`effective when tested 4 h after dosing. By 8 h the effects
`of the test compounds were no longer statistically signif-
`icant. As expected, morphine increased tail-flick latency
`at the 1 h time point, but at later times the effect was not
`significantly greater than that of vehicle. The results
`indicate that the test compounds effectively reduced
`acute nociception for 2–4 h following a single acute
`intraperitoneal dose.
`
`In conclusion, the substitution of deuterium for hydro-
`gen in the methyl groups of tramadol did not adversely
`affect the in vitro binding affinity. In vitro testing in hu-
`man microsomes confirmed that replacing hydrogen
`with deuterium in the metabolically labile O-methyl po-
`sition of 1 slowed the formation of the primary metabo-
`lite M1 by approximately 5-fold. The deuterated
`derivatives 7 (D6) and 9 (D9) were active analgesics in
`the rat tail-flick model, but were not superior to trama-
`dol in terms of potency or duration of effect. Deuterium
`
`**
`
`**
`
`**
`
`**
`
`****
`
`**
`
`**
`
`*
`
`60
`
`120
`
`240
`
`Minutes
`D6
`Tramadol
`Morphine Control
`Saline Control
`Dunnett's test: *, **/ p<0.05, p<0.01 (n = 9 - 10 per group)
`
`480
`
`D9
`
`25
`
`20
`
`15
`
`10
`
`05
`
`Latency (s)
`
`Table 2. Summary of in vitro metabolic stability of tramadol and
`deuterated derivatives
`
`Compound
`
`Half-lifea (min)/rankb
`HLMc
`
`Hepatocytes
`
`9 (D9), (+/)-cis
`7210/1
`3610/1
`8 (M1 D6), (+/)-cis
`747/5
`1650/2
`Metabolite M1, (+/)-cis
`2820/2
`1560/3
`6 (D3), (+/)-cis
`744/4
`1330/4
`1 Tramadol, (+/)-cis
`741/3
`968/5
`7 (D6), (+/)-cis
`547/6
`852/6
`a Calculated from samples taken at 0, 180, and 360 min for hepatocytes
`and 0, 90, and 180 min for microsomes.
`b Ranked from 1, the most stable to 6, the least stable.
`c HLM = human liver microsomes.
`
`7 were the least stable compounds in human liver micro-
`somes and hepatocytes.
`
`Metabolic stability was also evaluated based on the for-
`mation rate of metabolite M1 from tramadol (1) and
`three tramadol deuterated derivatives capable of pro-
`ducing metabolite M1. The results are summarized for
`
`D6 Tramadol
`
`D3 Tramadol
`
`D9 Tramadol
`
`180
`
`Incubation Time (min)
`
`360
`
`100
`
`80
`
`60
`
`40
`
`20
`
`0
`
`M1 Formation Rate (% Control)
`
`Figure 3. Stability of tramadol analogs in human hepatocytes.
`
`D6 Tramadol
`
`D3 Tramadol
`
`D9 Tramadol
`
`90
`
`Incubation Time (min)
`
`180
`
`100
`
`80
`
`60
`
`40
`
`20
`
`M1 Formation Rate (% Control)
`
`Figure 4. Stability of tramadol analogs in human liver microsomes.
`
`Figure 5. Effect of test compounds on tail-flick latency.
`
`Auspex Exhibit 2011
`Apotex v. Auspex
`IPR2021-01507
`Page 3
`
`

`

`694
`
`L. Shao et al. / Bioorg. Med. Chem. Lett. 16 (2006) 691–694
`
`for hydrogen replacement at metabolically active sites
`had no obvious deleterious effects in vivo but did not
`result
`in a longer duration of effect. In this case,
`deuteration at metabolically active sites produced a
`pharmacological
`agent
`equipotent
`in vivo with
`tramadol.
`
`References and notes
`
`1. (a) Raffa, R. B.; Friderichs, E.; Reimann, W.; Shank, R.
`P.; Good, E. E.; Vaught, J. L. J. Pharmacol. Exp. Ther.
`1992, 260, 275; (b) Gibson, T. P. Am. J. Med. 1996,
`101(Suppl 1A), 47S; (c) Radbruch, L.; Grond, S.; Leh-
`mann, K. Drug Safety 1996, 15, 8; (d) Raffa, R. B. Am. J.
`Med. 1996, 101(Suppl 1A), 41S; (e) Garrido, M.; Valle,
`M.; Campanero, M. A.; Calvo, R.; Troconiz, I. F. J.
`Pharmacol. Exp. Ther. 2000, 295, 352; (f) Subrahmanyam,
`V.; Renwick, A. B.; Walters, D. G.; Young, P. J.; Price, R.
`J.; Tonelli, A. P.; Lake, B. G. Drug Metab. Dispos. 2001,
`29, 1146; (g) Wu, W. N.; McKown, L. A.; Liao, S.
`Xenobiotica 2002, 32, 411; (h) Leppert, W.; Luczak, J.
`Support Care Cancer 2005, 13, 5; (i) Duhmke, R. M.;
`Cornblath, D. D.; Hollingshead, J. R. Cochrane Database
`Syst. Rev. 2004, 2, CD003726; (j) Grond, S.; Sablotzki, A.
`Clin. Pharmacol. Ther. 2004, 43, 879;
`(k) Klotz, U.
`Arzneim.-Forsch. 2003, 53, 681.
`2. PDR Electronic Library, Thomson Scientific, http://
`www.thomsonhc.com/pdrel/librarian,
`reference
`for
`UltramÒ tablets (Ortho-McNeil).
`3. Raffa, R. B.; Nayak, R. K.; Liao, S.; Minn, F. L. Rev.
`Contemp. Pharmacother. 1995, 6, 485.
`4. (a) Wilberg, K. B. Chem. Rev. 1955, 55, 713;
`Westheimer, F. H. Chem. Rev. 1961, 61, 265.
`5. Parli, C. J.; McMahon, R. E. Drug Metab. Dispos. 1973, 1,
`337.
`6. Tanabe, M.; Yasuda, D.; LeValley, S.; Mitoma, C. Life
`Sci. 1969, 8, 1123.
`7. Elison, C.; Rapoport, H.; Laursen, R.; Elliott, H. W.
`Science 1961, 134, 1078.
`8. NMR data for 16 (D3): 1H NMR (400 MHz, CD3OD) 7.36–
`7.31 (m, 1H), 7.16–7.11 (m, 2H), 6.85–6.83 (m, 1H), 3.04–
`2.97 (m, 1H), 2.74–2.64 (m, 7H), 2.34–2.26 (m, 1H), 1.97–
`1.60 (m, 8H); 13C NMR (100 MHz, CD3OD) 161.3, 150.5,
`130.6, 118.2, 113.0, 112.2, 75.8, 61.8, 46.2, 42.8, 42.6, 41.6,
`27.3, 26.0, 22.4; mp: 180–184 °C; TLC (10% MeOH in
`CH2Cl2) Rf = 0.48. NMR data for 17 (D6): 1H NMR
`(400 MHz, CD3OD) 7.30–7.27 (m, 1H), 7.10–7.05 (m, 2H),
`6.83–6.80 (m, 1H), 3.79 (s, 3H), 2.98–2.93 (m, 1H), 2.66–
`2.62 (m, 1H), 2.25–2.20 (m, 1H), 1.94–1.51 (m, 8H); 13C
`NMR (100 MHz, CD3OD) 162.3, 151.4, 131.6, 119.1, 113.8,
`113.2, 76.7, 62.6, 56.5, 43.8, 42.4, 27.9, 27.0, 23.3; mp: 175–
`180 °C; TLC (10% MeOH in CH2Cl2) Rf = 0.43.
`
`(b)
`
`9. NMR data for 18 (D6 M1): 1H NMR (400 MHz, CD3OD)
`7.14–7.11 (m, 1H), 6.90–6.88 (m, 2H), 6.62–6.60 (m, 1H),
`2.91–2.85 (m, 1H), 2.61 (dd, J = 2.57, 13.2 Hz, 1H), 2.13–
`2.07 (m, 1H), 1.94–1.43 (m, 8H); 13C NMR (100 MHz,
`CD3OD) 158.8, 150.5, 130.6, 117.1, 114.7, 113.2, 75.8,
`61.8, 43.1, 41.5, 27.1, 26.2, 22.4; mp: 221–226 °C; TLC
`(20% MeOH in CH2Cl2) Rf = 0.23.
`10. Draper, R. W.; Hou, D.; Iyer, R.; Lee, G. M.; Liang, J. T.;
`Mas, J. L.; Vater, E. J. Org. Process Res. Dev. 1998, 2, 186.
`11. NMR data for 19 (D9): 1H NMR (400 MHz, CD3OD)
`7.27–7.24 (m, 1H), 7.09–7.04 (m, 2H), 6.80 (d, J = 8.07 Hz,
`1H), 2.94 (dd, J = 9.16, 13.2 Hz, 1H), 2.21–2.16 (m, 1H),
`1.95–1.50 (m, 8H); 13C NMR (100 MHz, CD3OD) 161.0,
`150.1, 130.5, 117.9, 112.7, 112.0, 75.6, 61.5, 42.8, 41.3,
`27.0, 25.9, 22.1; mp: 176–182 °C; TLC (10% MeOH in
`CH2Cl2) Rf = 0.45.
`12. General procedure for microsomal stability assay in
`Human liver microsomes: The test compounds at 23 lM
`were incubated separately with human liver microsomes
`(4 mg protein/mL) at 37 ± 1 °C in 3-mL incubation
`mixtures
`containing
`potassium phosphate
`buffer
`(50 mM, pH 7.4), MgCl2 (3 mM), and EDTA (1 mM).
`Reactions were started by the addition of the NADPH-
`generating system. At designated times (0, 90, and
`180 min), a 500-lL aliquot was
`removed from the
`incubation and added to 500 lL acetonitrile to terminate
`the reaction. The amount of unchanged test compound
`was quantified by LC/MS/MS. Where appropriate, the
`amount of O-desmethyl metabolite formed was also
`assessed by LC/MS/MS.
`13. General procedure for in vitro stability assay in human
`hepatocytes. Test compounds (23 lM) were incubated
`separately with a pool (n = 2) of cryopreserved human
`hepatocytes (2 million cells/mL) in WaymouthÕs+ (Way-
`mouthÕs medium [without phenol
`red]
`supplemented
`(5.6 lg/mL),
`with FBS
`(4.5%),
`insulin
`glutamine
`(3.6 mM), sodium pyruvate (4.5 mM), and dexametha-
`sone (0.9 lM)) at
`the final concentrations indicated.
`Each test article was added to incubations in 2.5 lL
`(1%) of methanol. Reactions were started when placed
`in the incubator. At designated times (0, 180, and
`360 min), a 250-lL aliquot was
`removed from the
`incubation and added to 250 lL acetonitrile to terminate
`the reaction. Precipitated protein was
`removed by
`centrifugation (920g for 10 min at 10 °C). Amount of
`unchanged parent compound was quantified as follows;
`(20 lL) of
`an aliquot
`the supernatant
`fraction was
`transferred to 1 mL of
`internal
`standard (50 ng/mL
`dextrorphan in 1:1 methanol/water, 51-fold dilution) and
`analyzed by LC/MS/MS. Additional
`incubations were
`performed with hepatocytes in which the test article was
`replaced with 7-ethoxycoumarin (500 lM, marker sub-
`strate) to determine if the hepatocytes were metaboli-
`cally competent.
`
`Auspex Exhibit 2011
`Apotex v. Auspex
`IPR2021-01507
`Page 4
`
`