Improved process for preparation of gemfibrozil, an antihypolipidemic

Article abstract of DOI:10.1021/op400034f

An improved process for the preparation of gemfibrozil, an antihypolipodimic drug substance, with an overall yield of 80% and ～99.9% purity (including three chemical reactions) is reported. Formation and control of possible impurities are also described.

Full text of DOI:10.1021/op400034f

Article
pubs.acs.org/OPRD
Improved Process for Preparation of Gemfibrozil, an
Antihypolipidemic
Suri Babu Madasu,†,‡ N. A. Vekariya,*,† Hero Velladurai,† Aminul Islam,† Paul Douglas Sanasi,‡
and Raghu Babu Korupolu‡
†Chemical Research and Development, Aurobindo Pharma Ltd., Survey No. 71 and 72, Indrakaran (V), Sangareddy (M), Medak
District-502329, Andhra Pradesh, India
‡Engineering Chemistry Department, AU College of Engineering, Andhra University, Visakhapatnam-530003, Andhra Pradesh, India
ABSTRACT: An improved process for the preparation of gemfibrozil, an antihypolipodimic drug substance, with an overall
yield of 80% and ∼99.9% purity (including three chemical reactions) is reported. Formation and control of possible impurities
are also described. Finally, gemfibrozil is isolated from water without any additional solvent purification.
need to develop an alternative process for gemfibrozil, meeting
with all regulatory aspects.
INTRODUCTION
■
Gemfibrozil is classified as a fibric acid derivative and is used in
the treatment of hyperlipidaemias. It has effects on plasma-lipid
concentrations similar to those described under bezafibrate.
The major effects of gemfibrozil have been a reduction in
plasma-triglyceride concentrations and an increase in high-
density lipoprotein (HDL) cholesterol concentrations. A
reduction in very-low-density lipoprotein (VLDL)-triglyceride
appears to be largely responsible for the fall in plasma
triglyceride although reductions in HDL and low-density
lipoprotein (LDL)-triglycerides have also been reported.1,2
The effects of gemfibrozil on total cholesterol have been
more variable: in general, LDL-cholesterol may be decreased in
patients with pre-existing high concentrations and raised in
those with low concentrations.3 The increase in HDL-
cholesterol concentrations has resulted in complementary
changes to the ratios of HDL-cholesterol to LDL-cholesterol
and to total cholesterol.4,5 Gemfibrozil has successfully raised
HDL-cholesterol concentrations in patients with isolated low
levels of HDL-cholesterol but otherwise normal cholesterol
concentrations.6 The Helsinki heart study7 assessed gemfibrozil
for the primary prevention of ischaemic heart disease in middle-
aged men with hyperlipidaemia. The usual dose, by mouth, is
1.2 g daily in two divided doses given 30 min before the
morning and evening meals. Gemfibrozil is available as tablets
for oral administration (Lopid: USP).
Reports are available in the literature related to the various
synthetic routes for the preparation of gemfibrozil. These
processes8 have restricted application in the industry because of
less overall yield, number of purifications, and the stringent
regulatory requirement to meet the quality of a finished drug
substance. Moreover, the maximum daily dose of gemfibrozil is
1.2 g in divided doses. As per ICH guidelines, the acceptable
level for known and unknown related impurities in the drug
substance should be not more than 0.15% and 0.08%,
respectively,9 based on maximum daily dose. To minimize the
cost constraints and number of purifications, we avoided
tedious work-up processes. In view of the high-volume
requirement, huge revenues associated with this molecule,
and disadvantages from the reported processes, there arises a
RESULTS AND DISCUSSION
■
Our improved process also starts with the O-alkylation using
commercially available materials 2,5-dimethylphenol (2) and
isobutyl 5-chloro-2,2-dimethylpentanoate (4b). In the literature
O-alkylation of 2,5-dimethylphenol (2) with the methyl or
isobutyl ester of 5-chloro-2,2-dimethylpentanoic acid (chloro
alkyl ester side chain, 4) was reported10 by using o-xylene or
toluene with polar solvent DMSO in the presence of the
catalyst sodium iodide (Scheme 1). The addition rate of the
a
Scheme 1. Reported Synthetic Route for Gemfibrozil
a
4a: R = −CH3; 4b: R = −CH2CH(CH3)2; 5a: R = −CH3; 5b: R =
−CH2CH(CH3)2.
chloro alkyl ester side chain was very critical. It should be
completed in a specified time. If the addition rate is faster, more
unreacted 2 will be left due to the conversion of the chloro alkyl
ester side chain into the acid derivative. If the addition rate is
slow, formation of unknown impurities will be greater due to
the higher temperature of the reaction. Moreover, iodo
gemfibrozil impurity 7 (Figure 1) is forming in this route of
synthesis in 0.25−0.50%, due to use of sodium iodide, and it is
Received: February 11, 2013
Published: June 18, 2013
© 2013 American Chemical Society
963
dx.doi.org/10.1021/op400034f | Org. Process Res. Dev. 2013, 17, 963−966

Organic Process Research & Development
Article
aqueous toluene at reflux temperature, but the reaction was not
completed. It was identified that the effect of water dilution
plays an important role in the ester hydrolysis reaction with
base. After doing a number of experiments, the ester hydrolysis
reaction was optimized in toluene and highly concentrated
aqueous sodium hydroxide solution at reflux temperature
(Table 4). Moreover, the quantity of base can have a significant
impact on the quality of the hydrolysis product. In this regard,
we explored the mole ratio of sodium hydroxide required to
accomplish the complete deprotection. The results obtained
from this study showed the required mole ratio of sodium
hydroxide. It was found that 4.7 mol of sodium hydroxide was
optimum to bring about an efficient deprotection of the ester
group (entry 5, Table 3).
After completion of ester hydrolysis, workup was very
simplified, as separation of the aqueous sodium hydroxide layer
from the bottom, followed by addition of water and layers
separation. The aqueous layer contained gemfibrozil sodium
salt 6, which was further washed with toluene to remove
nonpolar impurities which were carried from the condensation
reaction, viz. gemfibrozil dimer 9 and gemfibrozil side chain
dimer 10 (Figure 2; isolated by column chromatography and
characterised). Thereafter, toluene was removed completely.
Many processes were reported11,12 for isolation of gemfibrozil
from gemfibrozil sodium in solvents such as methanol, hexanes,
etc. We have significantly simplified isolation of gemfibrozil
from water for obtaining a more cost-effective process without
having any significant changes in the solid state parameters of
the gemfibrozil drug substance. Gemfibrozil was isolated from
the aqueous layer by addition of aqueous hydrochloric acid at
pH 2.8−3.2 at room temperature and had a purity of ∼100%
and high yields.
Figure 1. Structure of iodo gemfibrozil 7.
difficult to remove this impurity even after crystallisation in
methanol−water followed by hexanes.
To overcome this, experiments were carried out in a biphasic
system by using tetrabutylammonium bromide (TBAB) as the
phase transfer catalyst. On the basis of experimental results
(Table 1) it was found that higher temperature in closed
autoclave conditions are required for O-alkylation. Experiments
for O-alkylation of 2 were performed with a chloro methyl ester
side chain (4a) as well as a chloro isobutyl ester side chain (4b)
in aqueous sodium hydroxide using TBAB as phase transfer
catalyst. Reaction with 4b is completed in 5 h, whereas, in the
reaction with 4a, the amount of starting material 2 left
unreacted was 45% (Table 2). It was observed that 4a itself
became hydrolysed in aqueous basic medium, gave compound
8 (Scheme 3), and conformed by mass analysis (MS m/z:
165.31, 187.25 corresponding to M + 1 and M + 22 with chloro
pattern). If excess sodium hydroxide is added to complete the
O-alkylation reaction, gemfibrozil methyl ester (5b) is hydro-
lyzed into gemfibrozil and it is very difficult to isolate the
gemfibrozil along with the gemfibrozil methyl ester during
work.
It was also found that hydrolysis of the gemfibrozil methyl
ester required only 1.2 mol (less base) of sodium hydroxide but
hydrolysis of the gemfibrozil isobutyl ester required 4.7 mol
more (more base) of sodium hydroxide. It was concluded from
above that half a portion of the chloro methyl ester side chain is
becoming hydrolysed itself in aqueous basic medium at higher
temperatures (under pressurised conditions) during the O-
alkylation reaction in the presence of base. This is the reason
for incomplete O-alkylation of 4a with 2, whereas O-alkylation
of 4b with 2 gave fruitful results (Scheme 2).
After optimizing O-alkylation of 2 with the chloro isobutyl
ester side chain (4b), gemfibrozil isobutyl ester extracted with
toluene and unreacted 2,5-dimethyl phenol (2) was recovered
by washing with aqueous sodium hydroxide solution. There-
after, the toluene layer was washed with 2% aqueous sodium
metabisulphite solution followed by water. The toluene layer
was as such taken for deprotection of the isobutyl group.
After preparation of highly pure gemfibrozil isobutyl ester,
our next target was deprotection of the isobutyl ester group.
We used sodium hydroxide as a base for the deprotection of the
ester group due to lower cost. Initially, the hydrolysis reaction
was carried out in water at reflux with 2.5 mol of sodium
hydroxide, but the reaction was not completed even after 10 h.
Another trial for the hydrolysis reaction was done by using
Finally, the redesigned process furnished gemfibrozil (1)
with an overall yield of 80% from the chloro isobutyl ester side
chain after three synthetic reactions with about 99.9% purity
and meeting all other quality parameters (Table 4).
CONCLUSION
■
An improved process has been developed for synthesis of
gemfibrozil with an overall yield of 80% and ∼99.9% purity.
The process described in this article has certain advantages over
the reported processes, with the primary one being the work-up
procedure during isolation of gemfibrozil. It is isolated from
water without any additional solvent purifications.
EXPERIMENTAL SECTION
■
1H NMR, 13C NMR, and DEPT spectral data were performed
in dimethyl sulfoxide (DMSO-d6) with 500 and 300 MHz
spectrometers. The chemical shift values were reported on the
δ scale in parts per million (ppm), downfield from
tetramethylsilane (TMS, δ = 0.0) as an internal standard.
Spin multiplicities are given as s (singlet), d (doublet), dd
Table 1. O-Alkylation of 4b with 2 in Presence of TBAB
a
b
entry
solvents
NaOH (m/r)
TBAB (catalyst, % w/w)
temp (°C)
time (h)
purity by HPLC (%)
yield (%)
1
2
3
4
5
o-xylene
o-xylene
toluene
water
1.03
1.03
1.03
1.05
1.05
7
7
7
7
7
140−145
130−135
110−115
95−100
6
6
63
61
72
65
88
57
56
65
60
80
10
10
5
c
water
125−130
a
b
c
It is conversion of product during reaction monitoring. Isolated yields of gemfibrozil. Under pressure.
964
dx.doi.org/10.1021/op400034f | Org. Process Res. Dev. 2013, 17, 963−966

Organic Process Research & Development
Article
Table 2. O-Alkylation of 4a/4b with 2 in Presence of TBAB in Water
c
a
b
entry
4
NaOH (m/r)
TBAB (catalyst, % w/w)
temp (°C)
time (h)
purity by HPLC (%)
yield (%)
1
2
3
4
methyl (4a)
methyl (4a)
isobutyl (4b)
isobutyl (4b)
1.05
1.05
1.05
1.05
7
10
7
125−130
125−130
125−130
125−130
6
6
5
5
47
45
88
88
45
43
80
80
10
a
b
c
It is conversion of product during reaction monitoring. Isolated yields of gemfibrozil. Under pressure.
Scheme 2. Synthetic Route for Gemfibrozil with New Process Conditions
Scheme 3. Self Hydrolysis of Chloro Methyl Ester Side
Chain (4a)
temperature. Sodium hydroxide (26.6 g, 0.665 mol) was added
to the reaction mass slowly in 15−25 min at 25−40 °C. The
resultant mixture was heated to 128−132 °C to complete the
reaction in a closed autoclave; ∼2.0 kg/cm2 pressure developed.
Thereafter, the reaction mass was allowed to return to room
temperature, and the gemfibrozil isobutyl ester (5b) was
extracted with toluene (1 × 280 mL; 1 × 140 mL). Unreacted
2,5-dimethylphenol (2) was removed by washing with 10% w/v
aqueous sodium hydroxide solution (3 × 140 mL). Thereafter,
the organic layer was washed with 2% w/v aqueous sodium
metabisulphite solution (140 mL) and finally washed with
water (140 mL). This toluene layer was taken as such for
deprotection of the isobutyl group. A small sample was taken,
and the toluene was evaporated completely under reduced
pressure at 60−70 °C to get oily gemfibrozil isobutyl ester
(5b), which was analyzed. Purity by HPLC: 99.0%; IR (film,
cm−1): 3047.20, 3024.13, 2961.22, 2873.87, 1728.75, 1615.45,
1585.90, 1509.29, 1472.37, 1414.50, 1388.22, 1310.58, 1264.93,
(doublet of doublet), t (triplet), and m (multiplet) as well as
brs (broad). Coupling constants (J) are given in hertz. IR
spectra were recorded in the solid state as KBr dispersions
using a Perkin-Elmer spectrum one Fourier transform (FT)-IR
spectrophotometer. The mass spectrum was recorded using a
Perkin-Elmer PE SCIEX-API 2000, equipped with an ESI
source used online with a HPLC system after the ultraviolet
(UV) detector. HPLC chromatographic purity was determined
by using the area normalization method. The thermal analysis
was carried out on a DSC Q 1000 TA. The thermogram was
recorded from 40 to 320 °C. The solvents and reagents were
used without purification.
Isobutyl 5-(2,5-Dimethylphenoxy)-2,2-dimethylpen-
tanoate (Gemfibrozil Isobutylester, 5b). Chloro isobutyl
ester side chain (4b) (140 g, 0.635 mol) and 2,5-
dimethylphenol (2) (79.8 g, 0.654 mol) were suspended in
water (175 mL) containing TBAB (9.8 g, 7% w/w) at room
1
1193.17, 1145.75, 1048.85, 995.50, 946.27, 803.05; H NMR
(DMSO, 500 MHz, δ ppm): 0.89 (d, 6H), 1.16 (s, 3H), 1.63
and 1.67 (m, 4H), 1.85 (q, 1H), 2.08 (s, 3H), 2.24 (s, 3H), 3.78
(d, 2H), 3.90 (t, 3H), 6.62 (d, 1H), 6.69 (s, 1H), 6.97 (d, 1H);
13C NMR and DEPT (DMSO, 500 MHz, δ ppm): 15.34
Table 3. Optimization of Sodium Hydroxide Mole Ratio (m/r) and Water Dilution for 5b Hydrolysis
a
solvents ratio
purity by HPLC (%)
b
entry
NaOH (m/r)
water (vol)
toluene (vol)
temp (°C)
time (h)
5b
1
yield (%)
1
2
3
4
5
6
2.5
2.5
3.0
4.0
4.7
6.0
1.5
100
7
9
7
7
5
5
95.40
93.76
2.96
0.37
0.55
1.0
2.0
3.25
3.0
3.0
3.0
100
0.25
0.30
0.30
0.30
100−110
100−110
100−110
100−110
95.76
97.60
99.84
99.53
75
77
80
80
0.46
0.07
c
ND
a
b
c
It is conversion of product during reaction monitoring. Isolated yields of gemfibrozil. Not detected.
965
dx.doi.org/10.1021/op400034f | Org. Process Res. Dev. 2013, 17, 963−966

Organic Process Research & Development
Article
Figure 2. Structures of nonpolar impurities 9 and 10.
Table 4. Purity Data of 1
a
purity by HPLC (%)
RS by GC (ppm)
b
entry
1
5b
2
any other
toluene isobutanol
yield (%)
DSC (°C)
c
c
c
c
c
c
c
c
c
c
c
c
1
2
3
100
100
ND
ND
ND
ND
ND
ND
ND
ND
105
ND
ND
ND
80.30
80.30
80.30
61.54
61.49
61.81
c
ND
c
100
ND
ND
a
b
c
Residual solvents. Overall yield. Not detected.
(CH3), 18.72 (CH3, CH3), 20.90 (CH3), 24.56 (CH2), 24.75
(CH3, CH3), 27.23 (CH), 36.47 (CH2), 41.47 (C), 67.39
(CH2), 69.65 (CH2), 112.00 (CH), 120.43 (CH), 122.40 (C),
129.94 (CH), 135.90 (C), 156.35 (C), 176.56 (C); MS m/z
(ESI): 307.23 [(MH)+].
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
The authors are grateful to colleagues in the Analytical
Research Department of APL Research Centre, a division of
Aurobindo Pharma Limited, Hyderabad for their valuable
contribution to this work. The authors also thank the
management for giving permission to publish this work.
5-(2,5-Dimethylphenoxy)-2,2-dimethylpentanoic Acid
(gemfibrozil, 1). To a solution of gemfibrozil isobutylester
(5b) in toluene (∼570 mL; entire quantity obtained from an
input of 140 g of 4b; 0.635 mol) was added a solution of
sodium hydroxide (119.37 g, 2.984 mol) in water (64 mL) at
room temperature. The resultant mixture was heated to 100−
112 °C for 7 h. After completion of reaction and cooling to
70−80 °C, the lower sodium hydroxide layer was separated,
and the gemfibrozil sodium salt 6 was extracted by addition of
water (840 mL) at 70−80 °C. The aqueous solution with
washed with toluene (3 × 300 mL) at 70−80 °C, and toluene
traces were removed by applying mild vacuum (300−500
mmHg) at 70−80 °C. The reaction mixture was allowed to
cool to room temperature and was diluted with water (1000
mL). The solution pH was adjusted to 8.8−9.2 with
hydrochloric acid at 25−30 °C (assay 10% w/w, ∼34 mL
was consumed). The resultant reaction solution was treated
with carbon enoanticroms (4.2 g; 3% w/w) and bed washed
with water (140 mL), followed by passage through a 0.2 μm
filter. The solution pH was further adjusted to 2.8−3.2 at 25−
30 °C (assay 10% w/w, ∼170 mL was consumed). The
resultant reaction mixture was stirred at room temperature for 1
h, and the precipitated solid was filtered, washed with water (2
× 140 mL; 25−30 °C), and dried under vacuum at 50−60 °C
to furnish 127.5 g (80.3%) of the title compound 1. Purity by
HPLC: 100%; IR (KBr, cm−1): 2959.03, 2919.78, 2877.65,
1709.42, 1613.44, 1586.60, 1511.07, 1473.81, 1414.01, 1387.89,
1317.61, 1286.34, 1271.91, 1214.39, 1159.26, 1048.83, 996.57,
REFERENCES
■
(1) Olsson, A. G.; Rossner, S.; Walldius, G.; Carlson, L. A. Proc. R.
Soc. Med. 1976, 69, 28−31.
(2) Nikkila, E. A.; Ylikahri, R.; Huttunen, J. K. Proc. R. Soc. Med.
1976, 69, 58−63.
(3) Kesaniemi, Y. A.; Grundy, S. M. JAMA 1984, 251, 2241−2246.
(4) Manninen, V.; Elo, M. O.; Frick, M. H.; Happa, K.; Heinonen, O.
P.; Heinsalmi, P.; Helo, P.; Huttunen, J. K.; Kaitaniemi, P.; Koskinen,
P. JAMA, J. Am. Med. Assoc. 1988, 260, 641−651.
(5) Lewis, J. Multicenter. Pract. Cardiol. 1983, 9, 99−118.
(6) Miller, M.; Bachorik, P. S.; McCrindle, B. W.; Kwiterovich, P. O.
Am. J. Med. 1993, 94, 7−12.
(7) Frick, M. H.; Elo, M. O.; Happa, K.; Heinonen, O. P.; Heinsalmi,
P.; Helo, P.; Huttunen, J. K.; Kaitaniemi, P.; Koskinen, P.; Manninen,
V. N. Engl. J. Med. 1987, 317, 1237−1245.
(8) Paul, L. C. 2,2-Dimethyl-ω-aryloxy alkanoic acids and salts and
ester thereof. U.S. 3,674,836, 1972.
(9) ICH guidelines, Q 3A(R); Impurities is new drug product. The
quality guide lines for Active Pharmaceutical Ingredients related to
impurities according to the international conference of harmonization
2002; http://www.ich.org.
(10) Francis, R. K. Process for preparing 5-(2,5-dimethylphenoxy)-
2,2-dimethylpentanoic acid. U.S. 4,665,226, 1987.
(11) Om, P. G.; William, M. P. Process for the production of 2,2-
dimethyl-5-(2,5-xylyloxy) valeric acid. U.S. 4,126,637, 1978.
(12) Hui-Po, W.; On, L.; Chin-Tsai, F. Derivatives and preparation of
2,2-dimethyl-5-substituted phenoxypentanoic acids. U.S. 5,412,112,
1995.
1
803.75; H NMR (DMSO, 500 MHz, δ ppm): 1.12 (s, 6H),
1.60 and 1.67 (m, 4H), 2.08 (s, 3H), 2.24 (s, 3H), 3.90 (t, 2H),
6.62 (d, 1H), 6.70 (s, 1H), 6.97 (d, 1H); 13C NMR and DEPT
(DMSO, 500 MHz, δ ppm): 15.39 (CH3), 20.94 (CH3), 24.67
(CH2), 24.87 (CH3, CH3), 36.43 (CH2), 40.91 (C), 67.57
(CH2), 112.07 (CH), 120.45 (CH), 122.44 (C), 129.96 (CH),
135.93 (C), 156.43 (C), 178.56 (C); MS M/Z (ESI): 251.16
[(MH)+].
AUTHOR INFORMATION
■
Corresponding Author
*E-mail: navekariya1@rediffmail.com.
966
dx.doi.org/10.1021/op400034f | Org. Process Res. Dev. 2013, 17, 963−966