Characterization of two new degradation products of atorvastatin calcium formed upon treatment with strong acids

Article abstract of DOI:10.3762/bjoc.15.206

Atorvastatin calcium (Lipitor, Sortis) is a well-established cholesterol synthesis enzyme (CSE) inhibitor commonly used in the therapy of hypercholesterolemia. This drug is known to be sensitive to acid treatment, but only little data has been published on the structures of the degradation products. Here we report the identification of two novel degradation products of atorvastatin, which are formed only under drastic acidic conditions. While treatment with conc. sulfuric acid led to a loss of the carboxanilide residue (accompanied by an expectable lactonization/dehydration process in the side chain), treatment with conc. aqueous hydrochloric acid gave a complex, bridged molecule under C–C-bond formation of the lactone moiety with the pyrrole, migration of the isopropyl group and loss of the carboxanilide residue. The novel degradation products were characterized by NMR spectroscopy, HRMS data and X-ray crystal structure analysis.

Full text of DOI:10.3762/bjoc.15.206

Characterization of two new degradation products of
atorvastatin calcium formed upon treatment with strong acids
Jürgen Krauß1, Monika Klimt1, Markus Luber1, Peter Mayer2 and Franz Bracher*1
Full Research Paper
Open Access
Address:
Beilstein J. Org. Chem. 2019, 15, 2085–2091.
1
Department of Pharmacy – Center for Drug Research,
doi:10.3762/bjoc.15.206
Ludwig-Maximilians University Munich, Butenandtstr. 5-13, 81377
Munich, Germany and 2Department of Chemistry, Ludwig-Maximilians
University Munich, Butenandtstr. 5-13, 81377 Munich, Germany
Received: 27 May 2019
Accepted: 22 August 2019
Published: 02 September 2019
Email:
Franz Bracher* - Franz.Bracher@cup.uni-muenchen.de
This article is part of the thematic issue "Reactive intermediates –
carbocations" and is dedicated to the memory of Prof. Dr. Gerhard
Rücker, Bonn.
*
Corresponding author
Keywords:
Guest Editor: S. R. Hare
atorvastatin; crystal structure; cyclization; degradation products;
fragmentation; stress test
© 2019 Krauß et al.; licensee Beilstein-Institut.
License and terms: see end of document.
Abstract
Atorvastatin calcium (Lipitor®, Sortis®) is a well-established cholesterol synthesis enzyme (CSE) inhibitor commonly used in the
therapy of hypercholesterolemia. This drug is known to be sensitive to acid treatment, but only little data has been published on the
structures of the degradation products. Here we report the identification of two novel degradation products of atorvastatin, which
are formed only under drastic acidic conditions. While treatment with conc. sulfuric acid led to a loss of the carboxanilide residue
(
accompanied by an expectable lactonization/dehydration process in the side chain), treatment with conc. aqueous hydrochloric
acid gave a complex, bridged molecule under C–C-bond formation of the lactone moiety with the pyrrole, migration of the iso-
propyl group and loss of the carboxanilide residue. The novel degradation products were characterized by NMR spectroscopy,
HRMS data and X-ray crystal structure analysis.
Introduction
Over the past decades, the general trend toward globalization of all imaginable impurities, on the other hand it is desirable to
the supply chains for active pharmaceutical ingredients has identify as much as possible degradation products of drugs re-
created new challenges for the authorities in ensuring the safety sulting from inappropriate exposition to potentially harmful
and quality of the drug supply [1]. Unprecedented impurities conditions during production, manufacturing and storage. For
can appear, most likely if limited information is available about being able to provide relevant data in a manageable time
details (or alterations) of production processes of drugs. On the frame, two kinds of stress tests have found broad application:
one hand, it is impossible to check drug substances routinely for accelerated storage conditions (higher temperatures, higher
2
085

Beilstein J. Org. Chem. 2019, 15, 2085–2091.
Results and Discussion
humidity, …) typically provide reliable data on the stability
of a drug, but are still time-consuming; on the other hand Stress tests
in “forced degradation experiments” the drug is submitted to Since the lability of atorvastatin towards moderately acidic
more drastic conditions (e.g., strong acid or base, strong conditions is well-documented, we aimed at investigating the
oxidant, very high temperature), and potential degradation outcome of incubation with acids under more drastic conditions.
products can be identified in very short time [2,3]. However, Treatment of atorvastatin calcium trihydrate (1) with 2 M
forced degradation experiments are highly artificial in nature, aqueous hydrochloric acid at room temperature (Table 1, entry
and thus one has to keep in mind that these extremely drastic 1) gave, in accordance with previous reports, only hydroxy-
conditions are prone to lead to results that might be out of lactone 2 (55% yield). This outcome was confirmed by compar-
proportion for daily quality control [3]. Nevertheless, know- ison with published NMR data [18,19]. At elevated tempera-
ledge about the outcome of stress tests under extreme condi- ture (reflux, 4 h; Table 1, entry 2) a mixture of lactone 2 and
tions helps to get insight into the overall reactivity of drug known unsaturated lactone 3 [15] (arising from acid-catalyzed
substances.
dehydration of 2) was obtained. Under even more drastic acidic
conditions (refluxing with 6 M hydrochloric acid for 3 h, with
Atorvastatin calcium (1, marketed as the trihydrate in Lipitor®, 20% aqueous H2SO4 for 2 h, or with p-toluenesulfonic acid in
Sortis®), is a well-established drug for treatment of hypercho- toluene for 5 h; Table 1, entries 3–5) unsaturated lactone 3 was
lesterolemia [4]. This drug is monographed in the leading phar- formed exclusively and in high to almost quantitative yields
macopoeias (Ph. Eur., USP), and a couple of impurities are (Table 1).
listed there. Most of these impurities result from the synthesis
process (stereoisomers, products resulting from impure starting When atorvastatin calcium trihydrate (1) was submitted to
materials or side reactions), and only one of these impurities, extremely strong acidic conditions by refluxing with concen-
lactone 2, is most likely a degradation product, resulting from trated (37%) aqueous hydrochloric acid (entry 6), a new prod-
acid-mediated lactonization of the 3,5-dihydroxyheptanoate uct 6 was formed in almost quantitative yield. The 1H NMR
side chain.
analysis clearly indicated that the entire carboxanilide partial
structure got lost under these conditions. However, no signal
A couple of previous publications deal with stress tests on ator- was observed which could be attributed to a C–H group at the
vastatin (and its salts), and an overview has been published by pyrrole ring. The 13C NMR data showed one carbonyl reso-
Sirén [5]. Hereby, atorvastatin was found to be sensitive to nance at 170.3 ppm, assignable to a lactone moiety. The HMBC
acidic, oxidative, photochemical and thermal stress. Acidic deg- experiment showed a cross peak between the proposed lactone
radation of atorvastatin was reported to follow first order carbonyl carbon and a neighboring CH-O group, confirming the
kinetics, but decomposition products were not characterized in lactone moiety, and the DEPT spectrum showed a new aliphat-
this [6] and several other reports, which only determined the ic methine resonance at 25.2 ppm. By HRESIMS mass data
downsizing of the atorvastatin peak in HPLC after treatment (found: 404.2020 for [M + H]+) a molecular formula of
with acid [7-10]. The most prominent decomposition product C26H26FNO2 was confirmed, excluding incorporation of HCl
upon acidic treatment, compound 2, results from lactonization into this artefact. Finally, X-ray crystallography structure analy-
of the 3,5-dihydroxyheptanoate side chain under moderately sis (see Figure 2 and Supporting Information File 1) disclosed
acidic conditions (0.1 M HCl) [11-14]. Shah et al. [15] identi- the structure of 6, bearing a novel, bridged tricyclic 1,5-
fied six additional decomposition products upon treatment with methanopyrrolo[1,2-e][1,5]oxazonin-3-one ring system
0
.1 M HCl at 80 °C for 24 h, among which the dehydrated (Scheme 1).
lactone 3 was dominating, accompanied by minor amounts of
products arising from dehydration of the δ-hydroxy group and In contrast, submission of atorvastatin calcium trihydrate (1) to
some epimers resulting from acid-catalyzed isomerization reac- concentrated sulfuric acid for two hours at 60 °C (Table 1,
tions. In contrast, Vukkum et al. [13] describe, besides lactones entry 7) led to the degradation product 7 in low yield (18%)
2
and 3, an α,β-unsaturated carboxylic acid 4. Treatment under (Scheme 1). No further decomposition products could be
more drastic conditions (6 M HCl, reflux, 3 h) was reported to isolated. Here, lactonization and dehydration steps in the
result mainly in hydrolysis of the anilide moiety to give side chain took place as observed before in the other acid
carboxylic acid 5 [16,17] (Figure 1).
treatments, however, under these extremely strong, virtually
anhydrous acid conditions, the entire carboxanilide residue
Here we report on the results of our investigations on the de- was removed to give the (S)-configured 4-unsubstituted
composition of atorvastatin calcium (1) under strongly acid pyrrole 7, as exemplified by a typical CH resonance at
conditions.
6.20 ppm in the 1H NMR spectrum. This structure was further
2
086

Beilstein J. Org. Chem. 2019, 15, 2085–2091.
Figure 1: Atorvastatin calcium trihydrate (1) and previously published decomposition products arising from treatment with acids: lactone 2, dehy-
drated lactone 3, α,β-unsaturated carboxylic acid 4, and carboxylic acid 5 (resulting from postulated anilide hydrolysis).
Table 1: Acidic stress conditions and decomposition products formed.
Entry no.
Acidic conditions
Decomposition products (yield)
2
3
6
7
1
2
3
4
5
6
7
2 M HCl, 20 °C, 2 h
2 M HCl, reflux, 4 h
6 M HCl, reflux, 3 h
20% H2SO4, reflux, 2 h
p-toluenesulfonic acid, toluene, reflux, 5 h
37% HCl, reflux, 5 h
55%
65%
–
14%
70%
>98%
95%
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
96%
–
–
18%
conc. H2SO4, 60 °C, 2 h
–
confirmed by X-ray data (see Figure 2 and Supporting Informa- we worked out an isocratic HPLC protocol, which prettily
tion File 1).
separates the four artefacts 2, 3, 6 and 7 from atorvastatin (1).
This method uses an RP18 stationary phase (Eurospher
HPLC method for the detection of the novel
impurities
1
00–C18), isocratic elution with 0.01 M ammonium
acetate buffer (pH 4)-acetonitrile 54:46 (v/v) at a flow
In order to provide a convenient method for including our rate of 1 mL/min at 40 °C, with UV detection at 246 nm
new findings into quality control of atorvastatin batches, (Figure 3).
2
087

Beilstein J. Org. Chem. 2019, 15, 2085–2091.
Scheme 1: Formation of novel artefacts 6 and 7 under extremely
strong acidic conditions.
Discussion
In this investigation we first confirmed some pathways of de-
composition of atorvastatin under acidic conditions. With dilute
mineral acids at room temperature, atorvastatin is conveniently
converted into the lactone 2 under retention at the C5–O bond
of the aliphatic chain [13,20], whereas treatment under more
drastic conditions (e.g., 6 M HCl or heating) causes expectable
subsequent dehydration to give the unsaturated lactone 3
[
13,15]. In contrast to previous reports [16,17] we could not
Figure 2: Top: Molecular structure of artefact 6. Shown here is the mo-
lecular structure of one of three independent molecules in 6 drawn at
the 50% ellipsoid probability level. Bottom: Molecular structure of arte-
fact 7 (drawn at the 50% ellipsoid probability level).
find any indication for a cleavage of the carboxanilide partial
structure to give free pyrrolecarboxylic acid 5 under treatment
with 6 M HCl under reflux.
Figure 3: Separation of atorvastatin (1; retention time: 5.8 min) from the four decomposition products 2 (retention time: 9.2 min), 3 (retention time:
5.6 min), 6 (retention time: 21.4 min) and 7 (retention time: 24.9 min). Chromatogram obtained with a solution containing 10 mg each in 5.0 mL DMF
retention time 2.3 min), diluted 1:5 with the eluent buffer before injection.
1
(
2
088

Beilstein J. Org. Chem. 2019, 15, 2085–2091.
However, upon treatment with concentrated sulfuric acid, quent reactions, the carboxanilide moiety at C-3 is eliminated
lactonization/dehydration is accompanied by complete loss of and the isopropyl residue, originally located at C-2 of the
the carboxanilide residue to give pyrrole 7. Complete one-step pyrrole, is shifted to C-3, rendering the annulated, tetrasubstitut-
removal of carboxamide residues from aromatic rings has been ed pyrrole 6. The structure of 6 was confirmed by X-ray crystal
observed before in investigations of fragmentations of proto- structure analysis. However, most likely cleavage of the carbox-
nated species in mass spectrometry [21,22]. For benzanilides Tu anilide moiety (compare formation of 7 from atorvastatin with
[
21] proposed a mechanism involving protonation of the amide concentrated sulfuric acid) is not the initial step in the cascade
oxygen, followed by 1,3 proton shift to the ring carbon next to of reactions leading from intermediate 3 to product 6. In a
the amide carbonyl group, followed by elimination of proto- control experiment, we treated compound 7 with 37% hydro-
nated phenyl isocyanate under re-aromatization. For the pyrrole chloric acid under the above mentioned conditions, but only a
derived substrate investigated here, even direct protonation at complex mixture of decomposition products was obtained, with
C-3 of the pyrrole by strong acid is most likely, due to the sig- no indication for formation of 6. As the intermediate occur-
nificant basicity of pyrroles. Delocalization of the positive rence of a positive charge at C-2 of the pyrrole ring and a sp3-
charge (A ↔ A’) as shown in Scheme 2 will support the initial hybridized C-3 are most likely triggering the elimination of
ring protonation step. The X-ray analysis of compound 7 phenyl isocyanate from the pyrrole (see postulated mechanism
revealed that the asymmetric center in the lactone ring is (S)- shown in Scheme 2), we propose the formation of an intermedi-
configured, indicating that once again the lactonization step ate C, which is formed via initial acid-mediated electrophilic
took place with retention at the remaining stereocenter (the shift attack of the unsaturated lactone at C-2 of the pyrrole ring. The
from (R) to (S) configuration is only a nomenclatory result of resulting carbenium ion should be stabilized as shown in
altered priorities of residues around the stereocenter upon dehy- Scheme 3 (B ↔ B’). Subsequent shift of the isopropyl group
dration). A comparable fragmentation has been observed in the from C-2 to C-3 then would give carbenium(-iminium) ion C,
CID (collision-induced dissociation) mass spectrum of atorvas- which can eliminate protonated phenyl isocyanate under forma-
tatin, where the base peak observed at m/z 440 clearly corre- tion of bridged pyrrole 6. A comparable shift of the isopropyl
sponds to a loss of phenyl isocyanate [22].
group in atorvastatin has previously been observed only under
oxidative conditions, giving rise to pyrrolidone-type degrada-
In contrast, refluxing 2 with concentrated aqueous hydrochloric tion products [23].
acid (37%) led to the formation of the complex, bridged prod-
uct 6. Most likely, this decomposition starts again with Conclusion
lactonization of the 3,5-dihydroxyheptanoate side chain, fol- Atorvastatin calcium trihydrate (1) is known to be an acid-labile
lowed by dehydration to give the unsaturated lactone 3. This drug, and incubation with dilute mineral acids gives the known
hypothesis was confirmed by refluxing pure lactone 3 with 37% lactone 2, which was further dehydrated to unsaturated lactone
hydrochloric acid, resulting in clean conversion into 6 as well. 3 under more drastic conditions. Treatment with very strong
In the following, this unsaturated lactone (Michael system) acids gave two hitherto unknown degradation products, whose
most likely performs an acid-mediated intramolecular attack at structures were elucidated by NMR and crystal structure analy-
C-2 of the electron-rich pyrrole ring. In a cascade of subse- sis. The bridged tricyclic product 6 was formed with concen-
Scheme 2: Proposed mechanism for the formation of desamidated product 7.
2
089

Beilstein J. Org. Chem. 2019, 15, 2085–2091.
Scheme 3: Proposed mechanism for the formation of bridged product 6 under cyclization, isopropyl migration and carboxanilide fragmentation.
trated hydrochloric acid, whereas lactone 7 resulted from treat- ORCID® iDs
ment with concentrated sulfuric acid. We propose mechanisms
Franz Bracher - https://orcid.org/0000-0003-0009-8629
for the formation of the novel artefacts 6 and 7 here. But it has
to be mentioned that these new artefacts, which are formed only Preprint
under extremely drastic conditions, are not likely to be relevant
in terms of drug safety and control of impurities in launched
atorvastatin batches.
A non-peer-reviewed version of this article has been previously published
as a preprint doi:10.3762/bxiv.2019.31.v1
References
1
.
Woo, J.; Wolfgang, S.; Batista, H.
Clin. Pharmacol. Ther. (Hoboken, NJ, U. S.) 2008, 83, 494–497.
doi:10.1038/sj.clpt.6100493
Klick, S.; Muijselaar, P. G.; Waterval, J.; Eichinger, T.; Korn, C.;
Gerding, T. K.; Debets, A. J.; Vandingenen, J.; Sänger, C.;
van den Beld, C.; Somsen, G.; de Jong, G. Pharm. Technol. 2004, 29,
Supporting Information
Supporting Information File 1
Materials and methods; stress tests and analytical data of
the products obtained thereby; details of characterization of
2.
4
8–66.
Blessy, M.; Patel, R. D.; Prajapati, P. N.; Agrawal, Y. K. J. Pharm. Anal.
014, 4, 159–165. doi:10.1016/j.jpha.2013.09.003
6
and 7 by X-ray data; crystallographic data for 6 and 7,
NMR spectra of artefacts 2, 3, 6, and 7.
https://www.beilstein-journals.org/bjoc/content/
3.
2
[
4
5
.
.
Haque, T.; Khan, B. V. Clin. Lipidol. 2010, 5, 615–625.
Sirén, H. M. M. J. Biomed. Res. Pract. 2017, 1, 100003.
http://hdl.handle.net/10138/236735
supplementary/1860-5397-15-206-S1.pdf]
Supporting Information File 2
checkCIF/PLATON report (Structure factors for artefacts 6
6.
Oliveira, M. A.; Yoshida, M. I.; Belinelo, V. J.; Valotto, R. S. Molecules
2013, 18, 1447–1456. doi:10.3390/molecules18021447
(
[
wq033) and 7 (wv633)).
https://www.beilstein-journals.org/bjoc/content/
7. Chowdhury, T.; Jakaria, M.; Bhuiyan, D.; Arifujjaman; Mamur, A. J.
Sci. Innovation Res. 2014, 3, 598–601.
8
9
1
.
.
Zaheer, Z.; Farooqui, M.; Mangle, A. A.; Nikalje, A.
Afr. J. Pharm. Pharmacol. 2008, 2, 204–210.
Chaudhari, B. G.; Patel, N. M.; Shah, P. B. Chem. Pharm. Bull. 2007,
supplementary/1860-5397-15-206-S2.pdf]
55, 241–246. doi:10.1248/cpb.55.241
0.Chaudhari, B. G.; Patel, N. M.; Shah, P. B.; Patel, L. J.; Patel, V. P.
Acknowledgements
J. AOAC Int. 2007, 90, 1539–1546.
The authors thank Prof. Dr. Herbert Mayr for helpful discus-
sions.
2
090

Beilstein J. Org. Chem. 2019, 15, 2085–2091.
1
1
1.Mohammadi, A.; Rezanour, N.; Ansari Dogaheh, M.;
Ghorbani Bidkorbeh, F.; Hashem, M.; Walker, R. B. J. Chromatogr., B
2007, 846, 215–221. doi:10.1016/j.jchromb.2006.09.007
2.Seshadri, R. K.; Desai, M. M.; Raghavaraju, T. V.; Krishnan, D.;
Rao, D. V.; Chakravarthy, I. E. Sci. Pharm. 2010, 78, 821–834.
doi:10.3797/scipharm.1004-14
1
1
1
1
3.Vukkum, P.; Babu, J. M.; Muralikrishna, R. Sci. Pharm. 2013, 81,
9
3–114. doi:10.3797/scipharm.1208-06
4.Rakibe, U.; Tiwari, R.; Rane, V.; Wakte, P. Acta Chromatogr. 2019, 31,
3–44. doi:10.1556/1326.2017.00333
5.Shah, R. P.; Kumar, V.; Singh, S. Rapid Commun. Mass Spectrom.
008, 22, 613–622. doi:10.1002/rcm.3403
3
2
6.Darwish, H. W.; Hassan, S. A.; Salem, M. Y.; El-Zeany, B. A.
Spectrochim. Acta, Part A 2016, 154, 58–66.
doi:10.1016/j.saa.2015.10.007
1
1
1
7.Hassan, S. A.; Elzanfaly, E. S.; El-Zeany, S. B. A.; Salem, M. Y.
Acta Pharm. 2016, 66, 479–490. doi:10.1515/acph-2016-0040
8.Sawant, P.; Maier, M. E. Tetrahedron 2010, 66, 9738–9744.
doi:10.1016/j.tet.2010.10.028
9.Stach, J.; Havlíček, J.; Plaček, L.; Rádl, S.
Collect. Czech. Chem. Commun. 2008, 73, 229–246.
doi:10.1135/cccc20080229
2
2
2
2
0.Kearney, A. S.; Crawford, L. F.; Mehta, S. C.; Radebaugh, G. W.
Pharm. Res. 1993, 10, 1461–1465. doi:10.1023/a:1018923325359
1.Tu, Y.-P. Rapid Commun. Mass Spectrom. 2004, 18, 1345–1351.
doi:10.1002/rcm.1475
2.Guo, C.; Jiang, K.; Yue, L.; Xia, Z.; Wang, X.; Pan, Y.
Org. Biomol. Chem. 2012, 10, 7070–7077. doi:10.1039/c2ob26011e
3.Cermola, F.; DellaGreca, M.; Iesce, M. R.; Montanaro, S.; Previtera, L.;
Temussi, F. Tetrahedron 2006, 62, 7390–7395.
doi:10.1016/j.tet.2006.05.027
License and Terms
This is an Open Access article under the terms of the
Creative Commons Attribution License
(
http://creativecommons.org/licenses/by/4.0). Please note
that the reuse, redistribution and reproduction in particular
requires that the authors and source are credited.
The license is subject to the Beilstein Journal of Organic
Chemistry terms and conditions:
(
https://www.beilstein-journals.org/bjoc)
The definitive version of this article is the electronic one
which can be found at:
doi:10.3762/bjoc.15.206
2
091