of a HCV polymerase inhibitor

Transcription

1 Multikilogram enantioselective synthesis of a HCV polymerase inhibitor Mr Andrew Gibb Global Process Chemistry Merck Sharp and Dohme. Hertfordshire, UK SCI Young Chemist in Industry 8 May

2 Hepatitis C Virus is a chronic viral infection that can cause liver disease RA virus Primarily infects liver cells Multiple l genotypes Primarily blood borne transmission RA-dependent RA polymerase Frequent mutation and no proofreading Selection of HCV variants 2

3 Hepatitis C: A global health problem U.S. 3-4 M Americas M E. Europe WE W. Europe 10M 5M Africa 30-40M Far East Asia 60M SE Asia 30-35M 170 million people are chronically infected with HCV 3

4 Current HCV Treatment Pegylated interferon alpha (PEG-IF) plus ribavirin (RBV) for weeks Weekly subcutaneous injections of PEG-IF Twice daily oral ribavirin Goal is eradication of virus Sustained Viral Response (SVR): undetectable HCV RA in plasma 6 months after completion of therapy Significant adverse experiences associated with both ribaviran and PEG-IF 4

5 2011: A new dawn rises for HCV patients Victrelis (Merck) and Incivek (Vertex) received FDA approvals in st generation S3/4A inhibitors used with existing standard of care to substantially improve treatment rates for hepatitis C. The frequent mutation and genetic heterogeneity of HCV requires that new therapies continue to be developed. Long term goal is shift to all oral direct acting antiviral therapy. 5

6 HCV S5B polymerase candidate 2 Single enantiomer 2 Unusual 8-membered dihydroindolobenzoxazocine ring 2,3-Disubstituted pyrrole Zwitterionic arjes, F. and co-workers J. Med. Chem. 2011, 54, 289 6

7 Medicinal chemistry approaches Me 2 Me H 2 H H 2 C 6steps Me 2 C 3steps Me 2 C ,3-Dielectrophile construction of 8-membered ring. Boc 2 2 or S S 5 6 1st generation Me 2 C + H H 2nd generation Thermal instability of aziridine route via 5 unworkable and multiple steps with no crystalline intermediates to prepare. 4 Epoxide route via 6 attractive for further exploration. 7

8 Racemic synthesis of desired target Used to access multi-gram amounts. Preparative a e separation a not viable for further kilogram scale-up. 8

9 Strategy for first kilogram scale delivery Me 2 Me H 2 H H 2 C Me 2 C Me 2 C 1 6steps 2 a 3 3, DMF o C Me 2 C H H + 4 S 6 (S)-glycidol tosylate Employ commercially available chiral pool starting material (S)-6. Replace high temperature azide displacement with alternative protocol. Develop expedited delaboration lb of triethylethylenediamine hl hl i sidechain hi from primary amine 2. 9

10 Substituted indole core synthesis Readily scaled and straightforward chemistry. 2-Hydroxyphenylboronic acid expensive and not widely available. 10

11 8-Membered benzoxazocine ring construction Heavily optimised to minimise dimeric and dialkylated impurities. First step: Slow, reverse addition of preformed phenoxide into epoxide at 65 ºC. Addition of EtAc prior to water addition gave smooth direct crystallisation of 10. MTBE swish of product to remove unreacted glycidol tosylate. Typical 10A% of 3 formed under reaction conditions. Second step: Slow reverse addition of 10 into cesium carbonate at 65 ºC to promote intravs intermolecular cyclisation. Direct crystallisation by water addition following addition of IPAc as co-solvent. ne-pot 2-step telescoped through process suffered from low yields. (S)-Epichlorohydrin cheaper but led to racemic 3. 11

12 Transformation of alcohol to primary amine : medchem High temperature displacement : safety concerns. -elimination also a competitive side reaction. What about Mitsunobu based protocol using diphenylphosphoryl azide? 12

13 Mitsunobu inversion with diphenylphosphoryl azide (DPPA) DIAD : Diisopropylazodicarboxylate Dramatic reactivity enhancement : inversion takes place at 15 ºC vs 110 ºC! Telescoped Staudinger azide reduction through addition of further Ph 3 P then water to avoid any handling of intermediate azide. Simple direct crystallisation of HCl salt from IPA/MeH sheds all of the 2 mol Ph 3 P byproduct. o distillation required at any point. ther -nucleophiles p evaluated did not afford desired inversion products. 13

14 Hydrazoic acid headspace measurements 14

15 Hydrazoic acid headspace measurements Process run in the presence of 1.2 equiv. of i-pr 2 Et to avoid H 3 in the headspace as well as with nitrogen sweep. 15

16 RC-1 Calorimetry measurements BATCH PERATI HEAT F REACTI (kj mol -1 of alcohol 3) ADIABATIC T (ºC) CMMETS Addition of DIAD Addition rate controlled. Accumulation <5% (addition over 20 minutes at 10 C) Addition of DPPA % accumulation (addition over 6 minutes at 15 C) Addition of THF ~50% accumulation (addition over solution of Ph 3 P 12 minutes at 25 C) Addition of water

17 Trimethylethylene diamine sidechain installation: medchem H 2 H H H MeC F 3 CCH 2 H MeC BH 3. SMe 2, THF MeC THF (95%) 1 2 (75-81%) 3 Boc H H MeC Boc H TFA / DCM MeC H 2 aac / HAc H 2,Pd/C(10%) MeH (95%) 4 (87%) 5 HCH aq, aac H 2, Pd/C (10%) MeC ah HC MeH MeH/THF/H 2 (85%) 6 (67%) 7 6 steps, linear with Boc glycine aldehyde not readily available 17

18 Trimethylethylene diamine sidechain installation Crystallisation of 18 from acetonitrile leads to ee upgrade to 98% through rejection of racemic mother liquors. Hydrazinolysis developed to address issues of genotoxicity and headspace liberation during batch concentration. 18

19 ne-pot Mitsunobu/Staudinger/Aza-Wittig Me 2 C H PPh 3 3, DIAD DPPA, THF Me 2 C Ph 3 P 17 H H Me Phth H 3 then PPh 3 21 STAB 18 ne-pot telescoped through process demonstrated to be viable but not developed due to time constraints. 19

20 Reductive trimethylation and final isolation High pressure required for triple methylation (90 psi H 2 ). Direct isolation by ph adjustment to crystallise 20. >5 kg drug substance prepared at >99 A%, >99% ee as tosylate salt. Jeremy P. Scott* and co-workers rg. Process Res. Dev. 2011, 15, 1116 (Special Issue: Asymmetric Synthesis on Large Scale 2011) 20

21 Long term route development : indole core Me 1 H Et 3 SiH Me H BS DCM Me H Br TFA, MeC 2 3 (H) 2 B H Me 2 C H H 4 a 2 C 3, dioxane, PdCl 2 2( (PPh 3 3) 2 Cost basis too high to support long term manufacture. Supply chain for phenylhydroxyboronic acid unreliable and slow. 21

22 Smiles rearrangement to prepare indole core Br Mg THF BrMg Me 3 S 3 H HBr (>95%) 2 (>95%) 3 H 1 F K 2 C 3 DMF Me 2 C 2 4 Pd/C C 2 Me 2 C 2 Me Me 2 C 7 H H AcH, THF (85%) H Gi Grignard prepared using elemental lmg. Ketone formation and demethylation high yielding allowing for 4-step telescoped through process to the desired indole product. Raw material cost basis significantly lower vs previous route. > 50 kg of indole, 59% overall from cyclohexylacetic acid. A Gibb* and co-workers, rg. Process Res. Dev. 2012, 16,

23 Asymmetric approaches based on enamide reduction Ketone 1 fully converted to imine in the presence of TiCl 4 (0.5 equiv.) and,-dimethyl ethylenediamine. Acylation gave the phenyl, methyl and trimethylacyl enamides. Enamine and enamide regiochemistry confirmed by MR. 23

24 Proof of concept for enamide reduction Me 2 C 34 R=Me 35 R= t Bu 36 R=Ph R H 2 Rh(nbd) 2 BF 4 TFE 38 or 39 Me 2 C 37 R=Me R * t-bu t-bu t-bu P t-bu P t-bu t-bu t-bu t-bu SL-A t-bu Me 2 t-bu t-bu Me 2 R=Me; Rh(nbd) 2 BF 4 with ligands 38 P t-bu P t-bu t-bu Me 2 SL-A Me 2 t-bu t-bu R Me; Rh(nbd) 2 BF 4 with ligands 38 or 39 each gave full conversion and 95% ee. Enamine reductions gave only low ee s. 24

25 Summary Efficient construction of the 8-membered dihydroindolobenzoxazocine ring. Practical room temperature azidation under Mitsunobu conditions. Expedited construction of the trimethylethylenediamine sidechain. Multikilogram demonstration to prepare >5 kg of drug substance. Alternative indole core synthesis via Smiles rearrangement demonstrated. Alternative enantioselective route evaluted by asymmetric enamide hydrogenation. 25

26 Acknowledgements Mahbub Alam Jos Brands Robert Wilson adine Bremeyer Antony Alorati Andrew Gibb Adrian Goodyear eil Strotman Tony Davies Frank arjes Bob Reamer Peter Mullens Gavin Stewart George Zhou Thientu Lam 26