Introduction
One of the most important considerations in the drug discovery process is safety, not only of the drug itself, but also impurities and degradation products. Impurities present in the active pharmaceutical ingredient (API) have to be identified to make sure no mutagenic or toxic substances will be administered to patients. Drug product degradation profiles need to be established to guide stable formulation and provide suitable drug shelf life assessment. Drug regulatory agencies also have requirements for characterization of the impurity profile of a pharmaceutical. Structural characterization of impurities and degradation products in bulk drug substances is an integral part of pharmaceutical product development. The analysis of these low level unknown impurities and degradants can be very challenging [1].
Mass spectrometry (MS) has become an important tool for elucidating the structures of low level unknown impurities and pharmaceutical degradants because of its unique analytical features [2]. The first step in a general analytical strategy for analysis of impurities and degradants is to measure the molecular weight (MW) of the unknown by suitable ionization methods, commonly carried out either by electrospray ionization (ESI) [3, 4] or by atmospheric pressure chemical ionization (APCI) [5]. A second step is to determine the elemental composition of molecular ions for the unknown by high resolution (HR) MS experiments using HR/MS instruments, including time-of-flight (TOF) [6, 7], Fourier transform ion cyclotron resonance (FT-ICR) [8, 9] or Orbitrap [10-13]. Tandem MS experiments (MS/MS) can be performed to generate fragmentation information of the unknown and API [14, 15]. In the case of complex mixtures, hyphenated analytical techniques such as high performance liquid chromatography (HPLC) / MS (LC/MS) and gas chromatography / MS (GC/MS) can be utilized to obtain MW information of individual components including their elemental compositions in HR/MS mode, as well as fragmentation information from LC/MS/MS experiments. Once a tentative structure is proposed, one can perform chemical derivatization including hydrogen / deuterium (H/D) exchange experiments [16-25] to facilitate structural elucidation studies. The final structural assignment is confirmed by nuclear magnetic resonance (NMR) experiments on an isolated sample fraction or a synthetic standard. It is important to note that the use of LC-HR/MS and LC-HR/ MS/MS techniques is critical in providing accurate mass data for both molecular ions and fragment ions in structural assignments. Another approach using on-line H/D exchange LC/MS is also very effective in structural elucidation studies. This method measures the difference in molecular weight of a compound before and after the deuterium exchange to determine the number of exchangeable hydrogen atoms in a molecule to assist structural elucidation. In general, the exchangeable hydrogen atoms are bound to N-, O-, or S-atoms in functional groups such as OH-, NH-, NH2-, COOH-. Deuterated mobile phases such as D2O or CH3OD can be readily deployed for on-line LC/MS analysis of mixtures. The combination of HR/MS/MS and online H/D exchange LC/MS methods provides a practical way for rapid structural characterization of unknowns.
Analysis of Impurities
The traditional approach in impurity identification involves isolation and purification by off-line HPLC, followed by characterization using spectroscopy or MS methods. A relatively large amount of sample is needed for analysis and the process can be very labor-intensive. In contrast, LC/MS and LC/MS/MS techniques require a small amount of material for analysis (typically less than 1 μg). HR/MS/MS and H/D exchange LC/MS approaches can be readily applied with speedy method development. As an illustration, impurity identification in mometasone furoate (di-chlorinated corticosteroid, MW 520 Dalton (Da)) drug substance will be discussed below [1, 26, 27].
In the course of large scale production of mometasone furoate, several low level impurities were detected by LC/MS technique (data not shown). The LC/MS experiments were performed using a C18 column under isocratic conditions with methanol and 2 mM ammonium acetate in water (75/25) at a flow rate of 1 mL/min. The impurity of main interest eluted at a retention time of 9 minute has two co-eluting components with [M+H]+ ions at m/z 535 and 581. The MS data suggest molecular weights of these two components to be 534 Da and 580 Da, respectively. Their isotopic patterns indicate two chlorine atoms for m/z 535 and one chlorine atom for m/z 581.
To obtain structural information on the unknown impurities, HR tandem MS experiments were performed on mometasone furoate molecular ion at m/z 521 (C27H31O6Cl2, [M+H]+) using a HR-MS mass spectrometer at a resolution of 30,000 in the FTMS mode. Common fragmentations observed for mometasone furoate include the loss of water, furoate ring, HCl and the cleavages of the steroid rings. The base peak in the product ion mass spectrum is the fragment ion at m/z 503 due to the loss of water. Further loss of furoate ring from m/z 503 yields the fragment ion at m/z 391. The loss of HCl group from m/z 391 generates m/z 355. Two product ions at m/z 279 and 319 are likely produced due to the loss of ClCHCO moiety and HCl from m/z 355, respectively. Another fragmentation pathway involves m/z 521 -> 485 (loss of HCl from m/z 521) -> 373 (loss of furoate ring from m/z 485) -> 355 (loss of water from m/z 373). A low abundant fragment ion at m/z 147 is likely formed as a result of the cleavages of steroid ring. Figure 1A summarizes fragmentation patterns for mometasone furoate with mass accuracy of 3 ppm or less for all product ions. For the impurity ion at m/z 535, its HR/MS data gives the elemental composition of C27H29O7Cl2 ([M+H]+, -0.41 ppm), suggesting the addition of one oxygen atom with the reduction of two hydrogen atoms in mometasone furoate molecule. HR/MS/MS data shows a highly abundant fragment ion at m/z 135 (absent in the product ion mass spectrum of m/z 521 under similar activation conditions), indicating a stabilized product ion and suggesting a 6-keto structure for this impurity, as displayed in Figure 1B. The 6-keto structure was further confirmed by a synthetic standard [26].
Accurate mass measurements on impurity ion at m/z 581 yield the best possible elemental composition as C28H34O9ClS with a mass accuracy of 0.18 ppm. Compared with the elemental composition of mometasone furoate (C27H31O6Cl2, [M+H]+), the net addition for m/z 581 is the moiety of CH3O3S with reduction of one chlorine atom. Two possible structures are proposed with one open structure involving replacement of 21-Cl and one close structure involving 20- keto and replacement of 21-Cl (Figure 2). Clearly, the number of exchangeable hydrogen atoms is different for these two structures. The open structure would have one exchangeable hydrogen atom and the close structure would have two exchangeable hydrogen atoms. On-line H/D exchange LC/MS experiments were performed and molecular ions were found to be shifted to m/z 583.1731 (C28H32D2O9ClS, [M+D]+, -0.27 ppm), suggesting one exchangeable hydrogen atom in the molecule. The result is consistent with this impurity being the open structure. Further LC-HR/MS/MS experiments on m/z 581 support its structural assignment as the replacement of 21-Cl with the sulfur-moiety.
Characterization of Degradation Products
As a part of drug development process, various stress-testing methods have been developed to simulate stresses the compound might experience during production processes and storage. These methods often expose drug candidates to forced degradation conditions such as acid, base, heat, oxidation and exposure to light. A rapid and successful identification of degradation products can help us to understand the degradation mechanism of drug candidate. A case study on antifungal agent posaconazole is shown below.
Posaconazole is a novel triazole antifungal agent [28]. Compared with the existing antifungal drugs such as amphotericin B, itraconazole and fluconazole, posaconazole has been found to exhibit higher potency against a broad range of fungal pathogens including Asperigillus, Candida and Cryptococcus [29-33]. The drug substance is stable at ambient conditions, but the compound starts to degrade under stress conditions such as prolonged exposure to heat and light. As a part of stability studies, degradation products of posaconazole were characterized in order to understand its degradation pathway [34].
The sample was weighed into a flask covered with aluminum foil and stored in an oven at 150°C for 12 days. The heat-stressed drug substance generated four major degradation products at retention times of 13 minutes (D), 16 minutes (B), 19.5 minutes (C), and 25 minutes (A). Initial structural elucidation work was carried out by LC/NMR and low resolution LC/MS/MS experiments. To gain further confidence in the structural assignments, LC-HR/MS and LCHR/ MS/MS experiments were performed on these four degradants using a HR/MS mass spectrometer at a resolution of 15,000. Accuratemass measurements were obtained on all degradants with excellent mass accuracy (< 2 ppm) (Figure 3). Clearly, the results eliminate the ambiguities for the elemental compositions of degradants obtained with low resolution mass data. Another advantage is confident structural assignment for fragment ions using accurate mass data in LC-HR/MS/MS mode [35]. For example, the major product ions from HR/MS/MS experiments for degradant B are m/z 713, 703, 685, 441, 372, 317 and 299 (Figure 4). Two abundant fragment ions at m/z 441 and 317 are results of the breakdown of the center piperazine ring. The loss of the triazole group (-C2H2N3) from the ion at m/z 441 results in the fragment ion at m/z 372. Further loss of water from the ion at m/z 317 leads to the formation of the fragment ion at m/z 299. The HR/MS/MS data suggest that posaconazole is cleaved into NN’-formyl diamine at the piperazine ring in the center of the molecule. The accurate mass data greatly enhanced structural characterization capability for the unknowns. In addition, on-line H/D exchange LC-HR/MS experiments were employed to facilitate structural identifications of four degradants, as shown in Figure 3. The measured number of exchangeable hydrogen atoms in the molecules was consistent with assigned structures.
Based on the degradant structures determined from the studies, an oxidative degradation pathway of posaconazole has been proposed (Figure 3). The air oxidation of posaconazole would initially yield a mixture of oxidation products (degradant A). Further oxidation leads to the formation of N N’-diformyl structure in the degradant B. Deformylation of the degradant B leads to degradant C with one N-formyl group remained and one secondary amine. Subsequent oxidative cleavage in the degradant C causes the breakdown of the structure into the degradant D. The proposed oxidative degradation pathway is believed to be a major one under the stress conditions in the studies.
Conclusions
Tremendous advances in LC/MS technologies have occurred in the last two decades. The use of LC-HR/MS and LC-HR/MS/MS methodologies has been widely accepted in the pharmaceutical industry for structural analysis of unknown impurities and degradation products in pharmaceuticals. Accurate mass measurements on molecular ions provide critical information on elemental compositions of unknowns. Subsequent accurate mass data on fragment ions is increasingly important to confirm structural assignments of fragment ions and unknowns. The ability to identify exchangeable hydrogen atoms through the use of the H/D exchange LC/MS approach adds an additional dimension in facilitating structural identification of unknowns. The use of HR/MS/ MS and on-line H/D exchange LC/MS methods will continue to play important roles in analysis of impurities and degradants in small molecule pharmaceuticals.
References
1. Chen G, Pramanik BN, Liu YH, Mirza UA (2007): Applications of LC/MS in structural identifications of small molecules and proteins in drug discovery. Journal of Mass Spectrometry 42: 279-287.
2. Chen G, Zhang LK, Pramanik BN (2007). LC/MS: Theory, Instrumentation and Applications to Small Molecules, In: Kazakevich Y, LoBrutto R (eds.) HPLC for Pharmaceutical Scientists, Wiley, Inc., New York, pp 281-346.
3. Yamashita M, Fenn JB (1984): Electrospray ion source. Another variation on the free-jet theme. Journal of Physical Chemistry 88: 4451-4459.
4. Fenn JB, Mann M, Meng CK, Wong SF, Whitehouse CM (1989): Electrospray ionization for mass spectrometry of large biomolecules. Science 246: 64-71.
5. Horning EC, Horning MG, Carroll DI, Dzidic I, Stillwell RN (1973): New picogram detection system based on a mass spectrometer with an external ionization source at atmospheric pressure. Analytical Chemistry 45: 936-943.
6. Mamyrin BA, Karataev VI, Shmikk DV, Zagulin VA (1973): Mass reflectron. New nonmagnetic time-of-flight high-resolution mass spectrometer. ZhurnalEksperimental’noi i Teoreticheskoi Fiziki 64: 82-89.
7. Guilhaus M (1995): Principles and instrumentation in time-of-flight mass spectrometry. Physical and instrumental concepts. Journal of Mass Spectrometry 30: 1519-1532.
8. Lawrence EO, Edlefsen NE (1930): Science 72:376.
9. Comisarow MB, Marshall AG (1974): Fourier transform ion cyclotron resonance spectroscopy. Chemical Physics Letters 25: 282-283.
10. Hardman M, Makarov AA (2003): Interfacing the orbitrap mass analyzer to an electrospray ion source. Analytical Chemistry 75: 1699-1705.
11. Hu Q, Noll RJ, Li H, Makarov A, Hardman M, Cooks RG: The orbitrap: a new mass spectrometer. Journal of Mass Spectrometry 40: 430-443.
12. Makarov A, Denisov E, Kholomeev A, Balschun W, Lange O, Strupat K, Horning S (2006): Performance evaluation of a hybrid linear ion trap/orbitrap mass spectrometer. Analytical Chemistry 78: 2113-2120.
13. Makarov A, Denisov E, Lange O, Horning S (2006): Dynamic range of mass accuracy in LTQ orbitrap hybrid mass spectrometer. Journal of the American Society for Mass Spectrometry 17: 977-982.
14. Cooks RG, Beynon JH, Caprioli RM, Lester GR. (1973). Metastable Ions, Elsevier, Amsterdam, The Netherlands.
15. McLafferty FW (1981): Tandem mass spectrometry. Science 214: 280-287.
16. Ohashi N, Furuuchi S, Yoshikawa M (1998): Usefulness of the hydrogen-deuterium exchange method in the study of drug metabolism using liquid chromatography-tandem mass spectrometry. Journal of Pharmaceutical and Biomedical Analysis 18: 325-334.
17. Olsen MA, Cummings PG, Kennedy-Gabb S, Wagner BM, Nicol GR, Munson B (2000): The use of deuterium oxide as a mobile phase for structural elucidation by HPLC/UV/ESI/MS. Analytical Chemistry 72: 5070-5078.
18. Liu DQ, Hop CE, Beconi MG, Mao A, Chiu SH (2001): Use of on-line hydrogen/ deuterium exchange to facilitate metabolite identification. Rapid Communication in Mass Spectrometry 15: 1832-1839.
19. Lam W, Ramanathan R (2002): In electrospray ionization source hydrogen/ deuterium exchange LC-MS and LC-MS/MS for characterization of metabolites. Journal of the American Society for Mass Spectrometry 13: 345-353.
20. Tolonen A, Turpeinen M, Uusitalo J, Pelkonen O (2005): A simple method for differentiation of monoisotopic drug metabolites with hydrogen-deuterium exchange liquid chromatography/electrospray mass spectrometry. European Journal of Pharmaceutical Science 25: 155-162.
21. Wolff JC, Laures AM (2006): ‘On-the-fly’ hydrogen/deuterium exchange liquid chromatography/mass spectrometry using a dual-sprayer atmospheric pressure ionisation source. Rapid Communication in Mass Spectrometry 20: 3769-3779.
22. Novak TJ, Helmy R, Santos I (2005): Liquid chromatography-mass spectrometry using the hydrogen/deuterium exchange reactions as a tool for impurity identification in pharmaceutical process development. Journal of Chromatography B 825: 161-168.
23. Liu DQ, Hop CECA (2005): Strategies for characterization of drug metabolites using liquid chromatography-tandem mass spectrometry in conjunction with chemical derivatization and on-Line H/D exchange approaches. Journal of Pharmaceutical and Biomedical Analysis 37: 1-18.
24. Chen G, Khusid A, Daaro I, Irish P, Pramanik BN (2007): Structural identification of trace level enol tautomer impurity by on-line hydrogen/deuterium exchange HRLC/ MS in a LTQ-Orbitrap hybrid mass spectrometer. Journal of Mass Spectrometry 42: 967-970.
25. Chen G, Daaro I, Pramanik BN, Piwinski JJ (2009): Structural characterization of in vitro rat liver microsomal metabolites of antihistamine desloratadine using LTQ-Orbitrap hybrid mass spectrometer in combination with online hydrogen/ deuterium exchange HR-LC/MS. Journal of Mass Spectrometry 44: 203-213.
26. Pramanik BN, Bartner PL, Chen G (1999): The role of mass spectrometry in the drug discovery process. Current Opinion in Drug Discovery & Development 2: 401-417.
27. Chen G, Daaro I, Pramanik BN (2007): Structural identifications of mometasone furoate steroid related impurities in a LTQ-Orbitrap hybrid mass spectrometer. Proceedings of the 55th ASMS Conference on Mass Spectrometry and Allied Topics, Abstract # A074176, Indianapolis, Indiana, June 3 – 7.
28. Nomeir AA, Pramanik BN, Heimark L, Bennett F, Veals J, Bartner P, Hilbert M, Saksena A, McNamara P, Girijavallabhan V, Ganguly AK, Lovey R, Pike R, Wang H, Liu YT, Kumari P, Korfmacher W, Lin CC, Cacciapuoti A, Loebenberg D, Hare R, Miller G, Pickett C (2008): Posaconazole (Noxafil, SCH 56592), a new azole antifungal drug, was a discovery based on the isolation and mass spectral characterization of a circulating metabolite of an earlier lead (SCH 51048). Journal of Mass Spectrometry 43: 509-517.
29. Perfect JR, Cox GM, Dodge RK, Schell WA (1996): In vitro and in vivo efficacies of the azole SCH56592 against Cryptococcus neoformans. Antimicrobial Agents and Chemotherapy 40: 1910-1913.
30. Galgiani JN, Lewis ML.In vitro studies of activities of the antifungal triazoles SCH56592 and itraconazole against Candida albicans, Cryptococcus neoformans, and other pathogenic yeasts (1997): Antimicrobial Agents and Chemotherapy 41: 180-183.
31. Graybill JR, Bocanegra R, Najvar LK, Luther MF, Loebenberg D (1998): SCH56592 treatment of murine invasive aspergillosis. Journal of Antimicrobial Chemotherapy 42: 539-542
32. Uchida K, Yokota N, Yamaguchi H (2001): In vitro antifungal activity of posaconazole against various pathogenic fungi. International Journal of Antimicrobial Agents 18: 167-172.
33. Frampton JE, Scott LJ (2008): Posaconazole: a review of its use in the prophylaxis of invasive fungal infections. Drugs 68: 993-1016.
34. Feng W, Liu H, Chen G, Malchow R, Bennett F, Lin E, Pramanik B, Chan TM (2001): Structural characterization of the oxidative degradation products of an antifungal agent SCH 56592 by LC-NMR and LC-MS. Journal of Pharmaceutical and Biomedical Analysis 25: 545-557.
35. Daaro I, Chen G, Schwartzburg I, Pramanik BN (2009): Characterization of Degradation Products of A Triazole Antifungal Agent by HR-LC/MS/MS and On-Line H/D Exchange LC/MS in LTQ-Orbitrap Mass Spectrometer. Proceedings of the 57th ASMS Conference on Mass Spectrometry and Allied Topics, Abstract # A090128, Philadelphia, PA, May 31 – June 4, 2009
Dr. Guodong Chen has extensive pharmaceutical research experience in major pharmaceutical companies, including Eli Lilly, Schering-Plough (now Merck) and Bristol-Myers Squibb. He is currently heading a mass spectrometry group at Bristol- Myers Squibb’s Princeton site, providing mass spectrometric/ analytical support to drug discovery and development programs in small molecule pharmaceuticals and biologics. He has over 50 research publications and over 55 presentations. He received his Ph.D. in Chemistry from Purdue University under the direction of Professor R. Graham Cooks.
Bethanne M. Warrack is a Senior Research Scientist II in Pharmaceutical Candidate Optimization at Bristol-Myers Squibb. She began her career at the Squibb Institute for Medical Research and has over 25 years experience in the use of advanced mass spectrometric techniques for the analysis of pharmaceutical molecules, endogenous and xenobiotic metabolites, proteins, and peptides.
Dr. Angela K. Goodenough is currently a Research Investigator II in the department of Bioanalytical and Discovery Sciences at Bristol-Myers Squibb, Princeton, NJ. She specializes in the development of MS-based approaches for the characterization and quantification of small molecule biomarkers and large biomolecules (biologics and protein biomarkers) in support of Discovery programs. She earned her Ph.D. from Vanderbilt University and completed her post-doctoral studies at the Wadsworth Center (NYSDOH); both in the area of DNA adduct research.