`
`A Concise Two-Step Synthesis of Thalidomide
`George W. Muller,* William E. Konnecke, Alison M. Smith, and Vikram D. Khetani
`Celgene Corporation, 7 Powder Horn DriVe, Warren, New Jersey 07059
`
`Abstract:
`A two-step synthesis of thalidomide is presented. The sequence
`requires no purifications. Treatment of L-glutamine with
`N-carbethoxyphthalimide produces N-phthaloyl-L-glutamine.
`Cyclization of N-phthaloyl-L-glutamine to afford thalidomide
`is accomplished by treatment with CDI in the presence of a
`catalytic amount of DMAP.
`
`Thalidomide (2-(2,6-dioxo-3-piperidyl)isoindoline-1,3-di-
`one, 1) was developed as a safe alternative to barbiturates
`in the late 1950s in Germany.1 It was widely used as a
`sedative and to prevent morning sickness in pregnant women.
`By 1962, thalidomide’s teratogenic effects had become a
`tragedy. However, even with its teratogenic effects, thali-
`domide continued to be used clinically. The serendipitous
`discovery of thalidomide’s clinical activity in the treatment
`of erythema nodosum leprosum (ENL) in leprosy led to the
`discovery of its immunomodulating and antiinflammatory
`activities.2 Since then, thalidomide has been used experi-
`mentally with benefit in a number of autoimmune and
`inflammatory diseases. Although thalidomide was not ap-
`proved by the FDA in the 1950s, it has been used clinically
`in the United States on investigational new drug applications
`for over 20 years. However, thalidomide’s immunomodu-
`lating properties led to its approval by the FDA for the
`treatment of ENL in July of 1998 under a strict distribution
`and monitoring program. In a 1991 publication, the Kaplan
`group3 reported that thalidomide was a selective inhibitor
`of tumor necrosis factor-R (TNF-R). Overproduction of
`TNF-R is believed to be involved in a variety of inflamma-
`tory and autoimmune diseases.4 Recent clinical trials using
`TNF-R antibodies have shown efficacy in the treatment of
`rheumatoid arthritis and Crohn’s disease, confirming the role
`of TNF-R in these diseases.5
`Celgene Corporation is interested in the clinical develop-
`ment of thalidomide for its anti-inflammatory and immuno-
`
`(1) Stirling, D.; Sherman, M.; Strauss, S. J. Am. Pharm. Assoc. 1997, 3, NS37.
`(b) Tseng, S.; Pak, G.; Washenik, K.; Pomeranz, M. K.; Shupack, J. L. J.
`Am. Acad. Dermatol. 1996, 35 (6), 969-979.
`(2) Sheskin, J. Clin. Pharmacol. Ther. 1965, 6, 303-311.
`(3) Sampaio, E. P.; Sarno, E. N.; Gallily, R.; Cohn, Z. A.; Kaplan, G. J. Exp.
`Med. 1991, 173, 699-703.
`(4) (a) Tracey, K. J.; Cerami, A. Annu. ReV. Med. 1994, 45, 491-503. (b)
`Sekut, L.; Conolly, K. M. Drug News Perspect. 1996, 261-269. (c) Rink,
`L.; Kirchner, H. Int. Arch. Allergy Immunol. 1996, 111, 199-209.
`(5) (a)Van Hogezand R. A.; Verspaget, H. W. Scand. J. Gastroenterol. 1997,
`32 (Suppl. 223), 105-107. (b) Elliott, M. J.; Feldmann, M.; Maini, R. N.
`Int. J. Immunopharmacol. 1995, 17 (2), 141-145.
`(6) (a) Eriksson, T.; Bjorkman, S.; Roth, B.; Fyge, A.; Hoglund, P. Chirality
`1995, 7, 44-52. (b) Knoche, B.; Blaschke, G. J. Chromatogr. A 1994,
`666, 235-240.
`(7) Reepmeyer, J. C.; Cox, D. C. FDA monograph: Guidelines to Thalidomide
`Synthesis; U.S. Food & Drug Administration: Washington, DC, June 1987.
`(8) Shealy, Y. F.; Opliger, C. E.; Montgomery, J. A. Chem. Ind. 1965, 1030-
`1031.
`
`Scheme1a
`
`a Conditions: (a) L-glutamic acid/pyridine reflux; (b) Ac2O; (c) urea, melt.
`
`Scheme2a
`
`a Conditions: (a) (1) L-glutamine, Na2CO3, (2) 4 N HCl(aq) (67%); (b) CDI,
`DMAP (91%).
`
`modulating properties. For this development and for further
`research into thalidomide’s biological properties and mech-
`anism of action, we wanted to develop a concise synthesis
`of thalidomide which could be used for preparing multigram
`to multikilogram quantities in standard glassware and pilot
`plant equipment.
`Thalidomide has always been used clinically as a race-
`mate. It is the classically quoted example of a drug developed
`as a racemate in which only one isomer, the S-isomer, carries
`the negative side effect, teratogenicity. However, recent
`publications have shown that thalidomide is chirally unstable
`in vitro and in vivo, thus making the chiral issue moot.6
`To take advantage of previous preclinical and clinical
`experience with thalidomide, we were interested in develop-
`ing a racemic synthesis. Thalidomide has traditionally been
`
`10.1021/op980201b CCC: $18.00 © 1999 American Chemical Society and Royal Society of Chemistry
`Published on Web 03/19/1999
`
`Vol. 3, No. 2, 1999 / Organic Process Research & Development
`
`¥
`
`139
`
`AstraZeneca Exhibit 2090
`Mylan v. AstraZeneca
`IPR2015-01340
`
`Page 1 of 2
`
`
`
`prepared as outlined in Scheme 1.7 This is a fairly simple
`three-step sequence. The last step in this synthesis involves
`a high-temperature melt reaction that affords a crude thali-
`domide requiring multiple recrystallizations. In our synthesis,
`we wished to avoid the melt reaction, which is not amenable
`to standard equipment.
`The synthesis in Scheme 1 begins with L-glutamic acid,
`but a synthesis starting with L-glutamine would allow a more
`direct two-step synthesis to be developed. In a 1965 paper,
`the use of 1,1¢ -carbonyldiimidazole (CDI) for 4 days at room
`temperature in DMF for the cyclization of N-phthaloyl-L-
`isoglutamine to form (S)-thalidomide in low yield (41%) was
`reported.8 In the same paper, under similar conditions,
`N-phthaloyl-L-glutamine was cyclized to afford a 31% yield
`of racemic thalidomide. In an earlier paper, acetic anhydride
`was used for this cyclization; however, again the cyclization
`afforded a low yield (33%) of thalidomide.9
`
`We felt the approach starting from L-glutamine instead
`of L-glutamic acid afforded a more direct route (Scheme 2).
`Glutamine was chosen over isoglutamine because of cost
`and the desire for racemic material. N-Phthaloyl-L-glutamine
`(4) was prepared by a standard technique using N-carbe-
`thoxyphthalimide. Treatment of L-glutamine with Na2CO3
`in water followed by the addition of N-carbethoxyphthalimide
`afforded, after workup, a 50-70% yield of 4 as a white
`powder. During the acidification, the reaction mixture is
`seeded with solid 4 to ensure solidification of the product.
`This material requires no purification. Use of N-carbethox-
`yphthalimide produces chirally pure 4. This is as expected
`from literature precedent10 and was confirmed in our
`laboratories by conversion of the material to (S)-thalidomide
`using the previously published cyclization method of Casini
`and Ferappi.11 The cyclization of 4 is accomplished using
`CDI in THF. Racemization of the product occurs in this step.
`THF was used since thalidomide has a low solubility in THF
`and the imidazole byproduct from CDI is soluble in THF.
`A stirred mixture of 4 and CDI (1.05 equiv) in the presence
`of a catalytic amount of DMAP in THF is heated to reflux
`for 15-18 h. Further work demonstrated that DMAP is not
`required for this cyclization to occur. Thalidomide crystallizes
`out of the reaction mixture during reflux. The cooled reaction
`mixture is filtered to produce an 85-93% yield of thalido-
`
`(9) Kig, F. E.; Clark-Lewis, J. W.; Wade, R.; Swindon, W. A. J. Chem. Soc.
`1957, 873-880.
`(10) Bodanszky, M.; Bodanszky, A. The Practice of Peptide Synthesis; Springer-
`Verlag: New York-Heidelburg-Berlin-Tokyo, 1984; pp 10-11.
`(11) Casini, G.; Ferappi, M. Farmaco, Ed. Sci. 1964, 563-56.
`(12) Additional material can be obtained from the mother liquor.
`(13) The major impurity is 4.
`(14) The racemic nature of the material was determined on a Daicel Chemical
`Industries Chiralpak OJ column using ethanol as the eluent.
`
`140
`
`¥
`
`Vol. 3, No. 2, 1999 / Organic Process Research & Development
`
`mide as a white solid.12 This material is normally of greater
`than 99% purity.13 The racemic nature of this material was
`confirmed by chiral HPLC.14
`In conclusion, a synthesis was developed which fulfilled
`our initial requirements of readily available starting materials,
`no purifications, a good yield, and ability be done in standard
`glassware and pilot plant equipment. The procedure can
`easily be used to prepare thalidomide at the 100-g scale in
`the laboratory, and it was successfully scaled up to the
`multikilogram scale.
`
`Experimental Section
`All reactions were run under a nitrogen atmosphere unless
`otherwise noted. All of the final compounds synthesized were
`characterized by 1H and 13C NMR, C, H, N elemental
`analysis, and melting point for solids. 1H and 13C NMR were
`determined on a Bruker AC250 FT instrument
`in an
`appropriate deuterated solvent. Elemental analyses and
`melting point determinations were done by Quantitative
`Technologies Inc., Whitehouse, NJ. Reagents and solvents
`were used as received from commercial suppliers.
`N-Phthaloyl-L-glutamine (4). To a stirred solution of
`L-glutamine (43.8 g, 300 mmol) and Na2CO3 (33.4 g, 315
`mmol) in 750 mL of water was rapidly added N-carbethoxy-
`phthalimide [65.8 g (97% pure, 67.8 g), 300 mmol] as a solid.
`After 1 h, the reaction mixture was filtered to remove
`unreacted N-carbethoxyphthalimide. The pH of the stirred
`filtrate was adjusted to 3-4 with 6 N HCl. The mixture was
`then seeded with N-phthaloyl-L-glutamine and the pH
`adjusted to 1-2 with 6 N HCl. The resulting slurry was
`stirred for 1 h. The slurry was filtered and the solid washed
`with copious amounts of water. The solid was air-dried and
`then dried in vacuo (60 (cid:176)C, <1 mmHg) overnight to afford
`51.8 g (67%) of 4 as a white powder: mp 169-171 (cid:176)C;
`1H
`NMR (dmso-d6) (cid:228) 13.22 (br s, 1 H, COOH), 8.05-7.75 (m,
`4 H, Ar), 7.22 (s, 1 H, CONH2), 6.74 (s, 1 H, CONH2), 4.76
`(dd, 1 H, CH), 2.50-1.95 (m, 4 H, CH2CH2), 2.15-2.00
`(m, 1 H, CH2); 13C NMR (dmso-d6) (cid:228) 173.0, 170.4, 167.3,
`134.7, 131.2, 123.3, 51.2, 31.3, 23.9. Anal. Calcd for
`C13H12N2O5: C, 56.52; H, 4.38; N, 10.14. Found: C, 56.64;
`H, 4.33; N, 10.10.
`Thalidomide (1). A stirred mixture of 4 (125 g, 452
`mmol), CDI (76.1 g, 469 mmol), and 4-DMAP (0.20 g, 1.6
`mmol) in anhydrous THF (750 mL) was heated to reflux
`for 16 h. The reaction slurry was filtered and the solid washed
`with CH2Cl2 (200 mL). The solid was air-dried and then dried
`in vacuo (60 (cid:176)C, <1 mmHg) to afford 106 g (91%) of the
`product as a white powder: mp 274-276 (cid:176)C;
`1H NMR
`(dmso-d6) (cid:228) 11.16 (s, 1 H, NH), 8.05-7.80 (br s, 4 H, Ar),
`5.18 (dd, 1 H, J ) 12, 5 Hz, CHCO), 3.05-2.85 (m, 1 H,
`CH2CO), 2.70-2.45 (m, 2 H, CH2CH2), 2.15-2.00 (m, 1
`H, CH2); 13C NMR (dmso-d6) (cid:228) 172.8, 169.8, 167.1, 134.9,
`131.2, 123.4, 49.0, 30.9, 22.0. Anal. Calcd for C13H10N2O4:
`C, 60.47; H, 3.90; N, 10.85; O, 24.78. Found: C, 60.42; H,
`3.82; N, 10.81; O, 24.98.
`
`Received for review April 10, 1998.
`
`OP980201B
`
`Page 2 of 2