Monitoring and control of genotoxic impurity acetamide in the synthesis of zaurategrast sulfate

Article abstract of DOI:10.1021/op900330e

In this article we describe the strategy adopted to minimize the risk of acetamide presence in zaurategrast sulfate drug candidate. A risk of acetamide formation (a potential genotoxic impurity) was identified in the API formation step of the process during the early development phase. In order to keep the project development timelines unchanged and without having the appropriate analytical method ready developed, we chose to minimize the risk of acetamide impurity presence by applying an adequate chemical process design. The implementation of a workup sequence involving initially three aqueous washes was later proven to be successful when an appropriate analytical method to detect acetamide below ppm levels was available. Additionally the analytical tool gave us the opportunity to assess and fine-tune the designed process for acetamide elimination by spiking experiments. Data acquired during this evaluation showed that a single aqueous wash associated with two efficient crystallization steps were finally enough to deliver API with a content of acetamide below the level defined as the acceptance criterion.

Full text of DOI:10.1021/op900330e

Organic Process Research & Development 2010, 14, 1008–1014
Monitoring and Control of Genotoxic Impurity Acetamide in the Synthesis of
Zaurategrast Sulfate
Arnaud Schu¨le´,*,† Ce´lal Ates,*,† Magali Palacio,† Jimmy Stofferis,† Jean-Pierre Delatinne,† Bruno Martin,‡ and Steven Lloyd§,⊥
UCB Pharma, Chemical Process DeVelopment & Industrialisation and Analytical DeVelopment & Industrialisation,
Chemin du Foriest, 1420 Braine l’Alleud, Belgium, and UCB Pharma, Pharmaceutical Sciences, Granta Park, Great Abington,
Cambridge CB1 6GS, United Kingdom
Abstract:
Working Party agreed with the use of a staged TTC concept
during clinical development.3
In this article we describe the strategy adopted to minimize the
risk of acetamide presence in zaurategrast sulfate drug candidate.
A risk of acetamide formation (a potential genotoxic impurity)
was identified in the API formation step of the process during the
early development phase. In order to keep the project development
timelines unchanged and without having the appropriate analytical
method ready developed, we chose to minimize the risk of
acetamide impurity presence by applying an adequate chemical
process design. The implementation of a workup sequence involv-
ing initially three aqueous washes was later proven to be successful
when an appropriate analytical method to detect acetamide below
ppm levels was available. Additionally the analytical tool gave us
the opportunity to assess and fine-tune the designed process for
acetamide elimination by spiking experiments. Data acquired
during this evaluation showed that a single aqueous wash associ-
ated with two efficient crystallization steps were finally enough to
deliver API with a content of acetamide below the level defined
as the acceptance criterion.
The avoidance of potential GTI impurities4 during the
development of chemical processes may not only limit the
efficiency (yield, selectivity, scalability) of some chemical
transformations, but it may also entail increased development
costs to perform a full chemical and analytical process assess-
ment. A complementary approach to avoiding GTI impurity is
therefore to assess and manage the risk through appropriate
application of chemical process design and analytical testing.
We share in this report an example of this strategy.
Introduction to Zaurategrast Sulfate. Zaurategrast sulfate
1 is a drug that had been under phase II clinical development
and was indicated in the field of inflammation and more
particularly for multiple sclerosis treatment.5 The synthetic route
to zaurategrast sulfate API 1 is a five cGMP step process in
which the last two steps include a key bromination of the final
intermediate UCB1193394 2 leading to crude API 3 followed
by a recrystallization from ethanol-water, giving rise to pure
active pharmaceutical ingredient (API) 1 as shown in Scheme
1. The presence of the bromine atom was found essential for
drug activity. Additionally, the ethyl ester is a prodrug of the
corresponding carboxylic acid.
Introduction to Genotoxic Impurities
The control of genotoxic impurities (GTI) during the
development of drug substances is a growing concern in the
pharmaceutical industry.1 The European Medicines Agency
(EMEA) issued in June 2006 a guideline for GTI limits that
came into force in January 2007. This guideline included the
concept of threshold of toxicological concern (TTC) to define
an acceptable risk for new active substances. A TTC of 1.5
µg/day is given at a level at which exposure will not pose a
significant carcinogenic risk.2 Additionally, the CHMP Safety
The bromination step as developed during early development
phase raised three questions in terms of potential GTI impurity.
First, the carry-over of methanesulfonic acid (MsOH) used
during the bromination reaction may lead to the formation of
the known GTI ethyl methanesulfonate in the presence of
ethanol during the subsequent salt formation.6 Second, the
isolation of the required sulfate salt in ethanol may lead to the
formation of monoethyl sulfate and/or diethyl sulfate impurity.7
And finally, the use of aqueous acetonitrile (MeCN) associated
to a strong acid may generate acetamide by acetonitrile
hydrolysis.8
* Authors for correspondence. E-mail: Arnaud.Schule@ucb.com; Celal.Ates@
ucb.com.
† Chemical Process Development
& Industrialisation, Braine l’Alleud,
Belgium.
‡ Analytical Development & Industrialisation, Braine l’Alleud, Belgium.
§ Pharmaceutical Sciences, Cambridge, United Kingdom.
(3) Questions & Answers on the CHMP Guideline on the Limits of
Genotoxic Impurities. EMA/CHMP/SWP/431994/2007, revision 2;
European Medicines Agency (EMEA), CHMP SAFETY WORKING
PARTY (SWP): London, December 2009.
⊥ Current address. E-mail: steve.lloyd@domainex.co.uk.
(1) Pierson, D. A.; Olsen, B. A.; Robbins, D. K.; DeVries, K. M.; Varie,
D. L. Org. Process Res. DeV. 2009, 13, 285–291. Teasdale, A.; Eyley,
S. C.; Delaney, E.; Jacq, C.; Taylor-Worth, K.; Lipczynski, A.; Reif,
V.; Elder, D. P.; Facchine, K. L.; Golec, S.; Schulte Oestrich, R.;
Sandra, P.; David, F. Org. Process Res. DeV. 2009, 13, 429–433. Yang,
Q.; Haney, B. P.; Vaux, A.; Riley, D. A.; Heidrich, L.; He, P.; Mason,
P.; Tehim, A.; Fisher, L. E.; Maag, H.; Anderson, N. G. Org. Process
Res. DeV. 2009, 13, 786–791.
(4) GTI impurity may be a reactant, an intermediate, or a by-product.
(5) NTC00484536. Double-blind, Placebo-controlled, Randomized, Paral-
lel-group Phase II Study in Subjects With Relapsing Forms of Multiple
Sclerosis (MS) to Evaluate the Safety, Tolerability, and Effects
of CDP323. http://clinicaltrials.gov/ct2/show/NCT00484536?term)
CDP323&rank)1; accessed 12/11/2009
(6) Glowienke, S.; Frieauff, W.; Allmendinger, Th.; Martus, H.-J.; Sutter,
W.; Mueller, L. Mutat. Res. 2005, 581, 23–34.
(2) Guidelines on the Limits of Genotoxic Impurities. CPMP/SWP/5199/
02 EMEA/CHMP/QWP/251344/2006; European Medicines Agency
(EMEA), Committee for Medicinal Products for Human Use (CHMP):
London, June 28, 2006.
(7) Summary and EValuation Diethyl Sulfate; International Agency for
Research on Cancer (IARC): 1992; Vol. 54, p 213, http://www.
inchem.org/documents/iarc/vol54/04-diethyl-sulfate.html.
1008
•
Vol. 14, No. 4, 2010 / Organic Process Research & Development
10.1021/op900330e  2010 American Chemical Society
Published on Web 02/11/2010

Scheme 1. Last two steps of zaurategrast sulfate synthesis
Scheme 2. Major impurity from the bromination process
The potential risk of generating ethyl methanesulfonate was
readily eliminated by changing the acid. A rapid screen showed
that hydrochloric acid gave similar results and was therefore
selected instead of methanesulfonic acid.
The choice of the sulfate as the counterion salt was driven
by solid-state consideration: polymorphism and particle size
control. The hydrate hemisulfate salt is formed by using half
an equivalent of sulfuric acid in 11 volumes of EtOH/H2O (70/
30) mixture. The experimental protocol that we developed for
zaurategrast sulfate formation is very different from the
experimental conditions that lead to formation of monoethyl-
sulfate and diethylsulfate.9 Additionally the monitoring of sulfate
esters is not in the scope of this article.
The potential presence of acetamide in zaurategrast sulfate
1 has received a detailed examination. Although the FDA lists
acetamide as a food additive,10 the International Agency for
Research on Cancer (IARC) has classified acetamide as possibly
carcinogenic to humans (Group 2B)11 based on rodent toxicity
data and thus is controlled to a threshold level of <5 µg/day.
The compound is negative in the Ames test but shows
genotoxicity in ViVo.12 For all these reasons it was decided to
control the acetamide level in the drug candidate following
guidelines for GTI control (<1.5 µg/day exposure for clinical
trials with duration longer than 12 months).
Synthesis of Zaurategrast Sulfate. The selective bromi-
nation of UCB1193394 2 into crude API 3 appeared rapidly as
the key step in the development of zaurategrast sulfate 1. In
general, brominations of advanced pharmaceutical intermediates
on industrial scale remain a challenge, more particularly in cases
where overbromination can occur13 and come to a head when
the overbrominated product cannot be easily eliminated by
conventional means (i.e., crystallization). This is the situation
we faced in this bromination reaction which produced the
overbrominated process impurity UCB1191133 5 at a typical
level of 0.10-0.30%-w/w (NMT 0.40%-w/w specified) as
shown in Scheme 2. Developing a satisfactory industrial
bromination process required in this case, simultaneous control
of reaction conversion and overbromination. Amongst the
brominating agents that were screened, NBS proved to be the
(10) (a) According to FDA’s EAFUS (Everything Added to Food in the
US) database, “The EAFUS list of substances contains ingredients
added directly to food that FDA has either approved as foods additives
or listed or affirmed as GRAS (Generally Recognized as Safe)”. In
this table acetamide is listed in the EAF category, meaning “There is
reported use of the substance, but it has not yet been assigned for
toxicology literature search”. (b) The Flavor and Extracts Manufactur-
ers Association Expert Panel, an independent organization, has
included acetamide in the GRAS (Generally Recognized as Safe) list
with the average maximum use level of 5 ppm for baked goods.
(11) IARC (1999) Monogr. Eval. Carcinog. Risks Hum.: Acetamide, Vol
71, 1211-1220.
(8) Basu, M. K.; Luo, F.-T. Efficient Transformation of Nitrile into Amide
under Mild Conditions. Tetrahedron Lett. 1998, 39, 3005–3006.
Wilgus, C. P.; Downing, S.; Molitor, E.; Bains, S.; Pagni, R. M.;
Kabalka, G. W. The Acid-Catalyzed and Uncatalyzed Hydrolysis of
Nitriles on Unactivated Alumina. Tetrahedron Lett. 1995, 36, 3469–
3472. Krieble, V. K.; Smellie, R. H. Process for hydrolyzing organic
nitriles and dehydrating ethers. U.S. Patent 2,441,114, 1948. Liler,
M.; Kosanovic, D. J. J. Chem. Soc. 1958, 1084–1090. Travagli, G.
Gazz. Chim. Ital. 1957, 87, 682–687 and 830–836. Barbosa, L. A.
M. M.; Van Santen, R. A. Study of Hydrolysis of Acetonitrile Using
Different Bronsted Acid Models: Zeolite-Type and HCl(H2O)x Clusters.
J. Catal. 2000, 191, 200–217. Lee, G. R.; Crayston, J. A. Hydrolysis
of acetonitrile in the presence of NbCl5. Polyhedron 1996, 15 (11),
1817–1821.
(12) Dybing, E.; Soderlung, E. G.; Gordon, W. P.; Holme, J. A.;
Christensen, T.; Becher, G.; Rivedal, E.; Thorgeirsson, S. S. Studies
on the mechanism of acetamide hepatocarcinogenicity. Pharmacol.
Toxicol. 1987, 60, 9–16.
(13) Frutos, R. P.; Rodriguez, S.; Patel, N.; Johnson, J.; Saha, A.;
Krisshnamurthy, D.; Senanayake, C. H. Org. Process Res. DeV. 2007,
11, 1076–1078. Li, B.; Buzon, R. A.; Zhang, Z. Org. Process Res.
DeV. 2007, 11, 951–955. Li, X.; Jain, N.; Russell, R. K.; Ma, R.;
Branum, S.; Xu, J.; Sui, Z. Org. Process Res. DeV. 2006, 10, 354–
360. Zanka, A.; Kubota, A.; Hirabayashi, S.; Nakamura, H. Org.
Process Res. DeV. 1988, 2, 71–77.
(9) Kazansky, V. B. React. Kinet. Catal. Lett. 1999, 68 (1), 35–43. In
this paper, the authors show that a 1:1 volume mixture of 70% H2SO4
with ethanol leads to the detection of the monoethylsulfate by 13C
NMR. In addition, diethylsulfate was only detected in harsher
conditions when using a 10:1 volume mixture of 95% H2SO4 with
ethanol.
Vol. 14, No. 4, 2010 / Organic Process Research & Development
•
1009

Figure 1. Aqueous washes to eliminate acetamide.
Table 1. Acetamide content in crude and pure APIs in the
best for this reaction.14 NBS is a common and cheap, stable,
and relatively safe industrial brominating agent for scaling-up
procedures.15 It has, however, two major disadvantages, a low
solubility in most solvents and a relatively high reactivity to
most of the functional groups, that makes it incompatible with
many solvents.16
250 kg campaign
zaurategrast acetamide
zaurategrast
sulfate
pure API
acetamide
content
(ppm)
sulfate
content
(ppm)
crude API
batch 1
batch 2
batch 3
BLQ
BLQ
BLQ
batch 4 (from batch 1)
batch 5 (from batch 2)
batch 6 (from batch 3)
0.18
BLQ
BLQ
Local high concentrations of NBS in the reaction mixture
during the addition step have been hypothesized to be the cause
of the overbromination reaction.17 In order to avoid or minimize
this overbromination side reaction, NBS was slowly added as
a solution in acetonitrile instead of a solid addition. Moreover,
water was added to acetonitrile to increase NBS solubility.18
The NBS acetonitrile/water solution thus obtained proved to
be stable for at least 12 h at 25 °C, which was sufficient for
performing our slow addition on industrial scale.19 The most
robust reaction conditions identified to deliver API within the
established UCB1191133 5 specification was a mixture of
acetonitrile/water/hydrochloric acid maintaining the reaction
temperature below 0 °C. Since the clinical study plan was
designed to cover up to 2 g of API as a daily dose, the level of
acetamide tolerated in the API was adapted according to
regulatory guidelines to <0.75 ppm.2
Once the risk of the presence of acetamide had been
identified and assessed, we set up a risk management plan to
reduce the probability of occurrence of this risk. The precise
evaluation of the acetamide carry-over to the API by the
chemical process in place was not possible at the time of the
risk identification. The tight development timelines did not allow
time to develop a sufficiently sensitive analytical method (with
the ability to detect acetamide below the ppm level) before
starting the planned production campaigns, nor to change the
process solvent. These two options would have jeopardized the
clinical program. The strategy adopted to reduce the risk of
acetamide presence in the API was therefore a rapid design of
this key step. The implementation of an aqueous workup
between the bromination step and the sulfate salt isolation was
selected as shown in Figure 1. This decision was motivated
after having identified in the literature and verified in the lab
the very high solubility of acetamide in water (>2 g/mL at r.t.).20
Results and Discussion
The workup was readily implemented (Figure 1). After
reaction completion, acetonitrile was efficiently removed by
azeotropic distillation. The zaurategrast hydrochloride salt in
water was partitioned between ethyl acetate and potassium
bicarbonate solution. The free base UCB1184197 4 was then
washed three times with 5%-w/w sodium chloride. A solvent
swap from ethyl acetate to ethanol followed by addition of
sulfuric acid finally led to the sulfate salt formation and isolation.
The choice of three aqueous washes was balanced between
expected wash efficiency to remove acetamide and manufactur-
ing cost consideration (cycle time). The pilot-plant campaign
proceeded successfully and afforded three batches of ∼80 kg
ofzaurategrastsulfate1withinthespecificationsforUCB1191133
5 dibrominated impurity.
Meanwhile, the analytical department developed a suitable
analytical method for detecting acetamide with a reported limit
of 0.10 ppm on the isolated sulfate salt. The results obtained
for both crude and pure APIs are reported in Table 1 (BLQ )
below the limit of quantification).
All API batches obtained were well below the level of 0.75
ppm acetamide defined as the acceptance criterion. Moreover
the results were mostly below the reported limit of 0.10 ppm
associated to the analytical method. The discrepancy observed
between batch 4 (0.18 ppm) and batch 1 (<BLQ) could be
explained by analytical uncertainty of the method at the time
of development.21 Indeed the appearance of acetamide during
the last recrystallization step by either a concentration effect or
a simple formation is very unlikely due to the experimental
conditions avoiding the presence of acetonitrile and acetamide.
These results demonstrated that the appropriate design of both
chemical process and analytical testing is very efficient to
control the level of GTI in the API.
(14) The following bromination reagents were tested in various industrial
pharmaceutical solvents: NBS, 1,3-dibromo-5,5-dimethylhydantoin,
HBr/Br2, and V2O5/H2O2/NH4Br.
(15) NBS could be purchased at an approximate price of 15 €/kg at ton
scale.
(16) The Merck Index, 14th ed.; https://themerckindex.cambridgesoft.com/
TheMerckIndex (accessed 12/11/2009). Synthetic Reagents; John
Wiley: New York, 1974; 2, pp 1-63.
(17) A FTIR monitoring study comparison between NBS solution
addition and portion-wise solid NBS addition showed that solid
NBS addition favored the formation of the over-brominated
impurity, UCB1191133.
(18) The NBS solution was prepared by dissolving 1 mol equiv of NBS in
a mixture of acetonitrile/water (2.6 vol/0.4 vol) with respect to starting
material UCB1193394.
(19) A batch size of 245 kg of UCB1193394 involved a NBS solution
addition time of 3 h 38 min at a rate of 3.0 kg/min, maintaining the
internal temperature at -2.3 °C to-2.6 °C.
(20) International Chemical Safety Cards, www.cdc.gov/niosh/ipcsnfrn/
nfrn0233.html (accessed 12/01/2009).
(21) At the time these data were generated, the analytical method was
neither fully developed nor qualified yet.
1010
•
Vol. 14, No. 4, 2010 / Organic Process Research & Development

Table 2. Acetamide content in crude and pure APIs in the
750 kg campaign
zaurategrast acetamide
zaurategrast
sulfate
pure API
acetamide
content
(ppm)
sulfate
content
(ppm)
crude API
batch 7
batch 8
batch 9
0.36
0.22
0.22
batch 10 (from batch 7)
batch 11 (from batch 8)
batch 12 (from batch 9)
BLQ
BLQ
BLQ
The robustness of the designed process (scalability and
efficiency to supply API within the specification and acetamide
acceptance criterion) was confirmed during a subsequent
manufacturing campaign of 750 kg of API. The bromination
step was run in three batches of ∼230 kg. The analytical results
for acetamide content in both crude and pure APIs are reported
in Table 2.
Process Performance Understanding. The availability of
an appropriate analytical method to track acetamide allowed
performing an in-depth study on the capacity of the designed
chemical process to eliminate acetamide potentially formed
during the bromination step.
In the designed process used during the manufacturing
campaigns, four operations were theoretically identified as
efficient to eliminate the potential acetamide formed during the
bromination. These four operations were (i) the potassium
bicarbonate neutralization which can be considered as a very
first aqueous wash, (ii) the three aqueous washes themselves,
(iii) the sulfate salt formation, and (iv) the final recrystallization
step. These substeps are depicted in Figure 2.
From an analytical point of view, the in situ monitoring of
potential acetamide formed during the reaction itself was
unfortunately not possible due to the low levels present and
because of several analytical interferences (zaurategrast itself
and the inorganic salt content). An alternative way to proceed
was to quantify the presence of acetamide in the aqueous and
organic phases from the implemented aqueous washes. Once
again, due to analytical interferences with the API free base in
the organic layers and salt contents in the aqueous ones, it was
impossible to generate reliable data on the acetamide content.
The evaluation of the capacity of the process to eliminate
acetamide potentially formed during the bromination step was
finally performed by spiking acetamide at key process points
highlighted in Figure 2 (see spiking points 1, 2, and 3) and by
analyzing the obtained crude and pure APIs.
Figure 2. Process flow diagram of zaurategrast sulfate: steps
4 and 5.
Table 3. Purification effect of the isolation of crude and
pure APIs: spiking experiment with 200 ppm acetamide
acetamide content mass balance percentage
step 4
(mg/mL)
(mg)
(%)
acetamide spiked
mother liquors
1st cake wash
(EtOH 70%)
2nd cake wash
(H2O 100%)
zaurategrast
200 ppm
0.0159
0.0079
6.675 mg
5.424 mg
0.750 mg
100
81.3
11.2
0.0002
0.005 mg
0.090 mg
0.1
1.4
2.7 ppm
sulfate crude 3
Recovery 94%
step 5
acetamide content
The overall magnitude of acetamide purge was initially
unknown. For that reason we started with the assessment of
the two crystallizations (closest operations to the API). Aceta-
mide was added in the aqueous ethanolic solution of the free
base before the addition of sulfuric acid to form the required
crude sulfate (spiking point 3, Figure 2). All amounts of
acetamide spiked in our studies were related to the theoretical
quantity of crude API obtained with an 85% yield for the
bromination step obtained routinely in the laboratory and
confirmed at larger scales.
An initial spike of 200 ppm of acetamide was performed.
The results are described in Table 3. The experiment afforded
the crude API containing 2.7 ppm acetamide only. Most of the
spiked acetamide was eliminated in the mother liquors (81%)
then during the first cake wash (11%). The mass balance for
mother liquors
1st cake wash
(EtOH 70%)
2nd cake wash
(EtOH 15%)
zaurategrast
sulfate 1
not detected22
not detected
not detected
BLQ
acetamide indicated a recovery of 94%. The recrystallization
of the corresponding crude API containing 2.7 ppm acetamide
gave a pure API with acetamide below the reported limit (<0.10
ppm) and by consequence well below the level of 0.75 ppm
defined as the acceptance criterion.
The good results obtained encouraged us to challenge the
capacity of both crystallizations to eliminate acetamide. Larger
amounts of acetamides1000 ppm and 5000 ppmswere spiked,
Vol. 14, No. 4, 2010 / Organic Process Research & Development
•
1011

Table 4. Purification effect of the isolation of crude and
pure APIs: spiking experiments with 1000 ppm and 5000
ppm acetamide
The set of experiments entries 2-4 dealt with the potassium
bicarbonate neutralization influence. Entry 2 gave a significant
result concerning the usefulness of the aqueous-wash sequence
when the bromination reaction was run as developed. When
no acetamide spike was done, skipping all three aqueous washes
led to crude API containing 0.11 ppm acetamide which was
already in the expected range. It could retrospectively be
concluded that in the designed chemical process, the introduc-
tion of three aqueous washes did not bring any measurable
improvement and was therefore unnecessary. If some acetamide
is generated during the reaction, it will be removed by the single
or combined actions of the neutralization and the crystallization
steps.
Entry 3 dealt with the 10000 ppm spiking before neutraliza-
tion in the reference process. The combined effects of the
neutralization, washes, and crystallizations removed all the
initially spiked amount of acetamide to afford a crude API
containing 0.31 ppm acetamide. Entry 4 showed that in these
engineered conditions, skipping all three aqueous washes gave
a crude API containing 15 ppm acetamide. In such a case, when
a larger amount of acetamide is present in the reaction mixture,
the combined actions of the neutralization and the crystallization
steps are not enough to ensure a complete elimination of
acetamide.
The second set of experiments (entries 5-7) dealt with the
evaluation of aqueous washes for which the spike was done
before the wash sequence. Entry 5 involved only one aqueous
wash and afforded a crude API containing 20 ppm acetamide
only. Considering the crystallization evaluation where a 1000
ppm spike gave a crude API containing 54 ppm acetamide, it
can be concluded that a single aqueous wash is able to eliminate
more than 9000 ppm of acetamide in the conditions used.
Two aqueous washes (entry 6) gave a crude API containing
3 ppm acetamide, and three aqueous washes (entry 7) afforded
a crude API whose acetamide content was 0.65 ppm. These
results showed that for the removal of a large quantity of
acetamide the first aqueous wash is the most effective.
The purification process implemented (potassium bicarbonate
neutralization and three aqueous washes) proved to be highly
efficient for acetamide elimination. Moreover, these purification
steps are complemented by two very effective crystallization
and recrystallization steps as demonstrated in Tables 3 and 4.
In that respect, a recommendation to optimize the cycle times
of the modified chemical process on the plant was to skip the
unnecessary second and third aqueous washes. It was, however,
recommended to keep only one aqueous wash to ensure the
elimination of all salts brought during the workup (sodium
sulfite and potassium bicarbonate).
Acetamide Input from Industrial Acetonitrile. The aceto-
nitrile used during the pilot-plant campaign was analyzed to
evaluate its acetamide content and the potential influence on
API acetamide level. The analysis showed a concentration of
0.018 mg/L of acetamide in the acetonitrile used in the
bromination step. The significance of this result can be easily
translated for a bromination batch run during the campaign.
Even if no acetamide was removed during the process, the
acetamide level within industrial acetonitrile is such that only
64 ppm would remain in the API.
1000 ppm
spike
5000 ppm
spike
Step 4
cake after filtration
cake after EtOH/water
(70/30) wash
155.64
113.21
696.50a
382.36a
cake after water wash
(crude API)
54.00
158.83
Step 5
cake after filtration
cake after EtOH/water
(70/30) wash
4.32
1.27
31.17
12.57
cake after EtOH/water
(15/85) wash (pure API)
0.77
5.88
a Underestimated values due to detector saturation.
Table 5. Purification effect of the bicarbonate neutralization
and aqueous washes
10000 ppm
acetamide
spike
number of acetamide content (ppm)
aqueous
washes
in crude zaurategrast
sulfate
entry
1
2
3
4
5
6
7
no spiking
3
0
3
0
1
2
3
0.21
0.11
no spiking
spiking point 2
spiking point 2
spiking point 3
spiking point 3
spiking point 3
0.31
14.98
20.31
3.02
0.65
and the results for sulfate salt samples are reported in Table 4.
These experiments afforded crude APIs containing 54 ppm and
159 ppm acetamide, respectively. The corresponding pure APIs
contained 0.77 ppm and 5.88 ppm acetamide, respectively.
These results demonstrated the very high potential of acetamide
elimination for both crystallization and recrystallization steps.
The second phase was the evaluation of the perfor-
mance of the aqueous washes implemented. Depending
on the operation studied (potassium bicarbonate neutral-
ization or aqueous washes) the spiking point was set either
before the bicarbonate neutralization (spiking point 1,
Figure 2) or before the wash sequence (spiking point 2,
Figure 2). The amount of acetamide selected for spiking
experiments was 10000 ppm in order to take account of
downstream crystallization efficiencies demonstrated
previously.
The results obtained are described in Table 5. Entry 1 was
the reference-designed process without any acetamide spike and
with the implemented neutralization and three aqueous washes.
The amount of acetamide contained in the crude API (0.21 ppm)
was in the expected range and well below the limit of 0.75
ppm acetamide defined as the acceptance criterion.
(22) The 2.7 ppm acetamide contained in the crude API is likely to be
removed in the mother liquors of the step 5 final recrystallization based
on data obtained for the crude API isolation at step 4. In that case
80% of acetamide spiked was removed in the mother liquors (11 vol).
For step 5, 2.7 ppm (1.4% acetamide spiked) diluted in approximately
the same volume of solvent (9 vol) makes its quantification not possible
by the analytical method. This explains why the amount of 2.7 ppm
acetamide is neither detected nor reported.
1012
•
Vol. 14, No. 4, 2010 / Organic Process Research & Development

The hypothesis that this amount of acetamide present in the
solvent would pass through the entire process is unlikely as
demonstrated in this paper with the huge capability of the
neutralization, the aqueous washes, and both crystallizations to
purge out acetamide. Therefore, the acetamide content in
industrial acetonitrile is marginal and can be considered totally
removed by the designed bromination process put in place.
with 5975B mass spectrometer operating under Chemstation
software control. The column is a Phenomenex ZB-Wax Plus
GC column, 30 m × 0.25 mm with 0.25 µm film thickness,
fitted with 2-3 m retention gap of deactivated silica tubing.
The column selection has been highly influenced by the possible
presence of significant amount of residual water in the final
samples injected. Carrier gas: helium; initial head pressure: 200
kPa, constant mode pressure; injector: temperature 300 °C,
pulsed splitless mode, pulse pressure 250 kPa, pulse time 1.0
min, purge time 1.0 min; injection linear: Agilent 5185-8818
or equivalent (straight, 2 mm, ID, 250 µL); injection volume:
1 µL, (10 µL syringe); injection program: slow plunger speed,
post injection dwell ) 1.00 min; injector wash: ethanol/water
95/5 v/v; purge flow: 100 mL/min; oven temperature: initial
temperature ) 100 °C, hold 1 min, then 10 °C/min until 180
°C is reached. Solid samples are typically dissolved in dichlo-
romethane, and the acetamide is separated from the matrix by
liquid/liquid extraction principle toward an alkaline aqueous
layer. The use of ultrapure NH3 is required since NH3 of poor
quality has been identified as primary source of acetamide in
blanks. Additional solid-phase extraction and concentration steps
have been added to increase sensitivity of the method. Samples
are quantified against an acetamide external standard, but the
complexity of sample preparation has required the use of an
additional internal standard. D3-Acetamide has been naturally
selected as internal standard considering the detection technique
used. In terms of qualification, the method has been validated
for specificity, precision, sample stability, linearity, and accuracy
from 0.10 ppm to 1.50 ppm.
The detailed sample preparation procedure is described
hereafter: 3.00 g of sample is accurately weighed into a 40 mL
vial. Three milliliters of internal standard solution (3 µg (1 ppm)
D3-acetamide) and 6 mL of dichloromethane are added fol-
lowed by the addition of 3 mL of 7% aqueous NH3 solution.
The vial is thoroughly shaken to extract analytes. A sample of
the aqueous layer, 1.5-2 mL, is placed into a 20 mL vial; 1.5
mL of clean aqueous solution is eluted onto a conditioned HLB
cartridge. One milliliter of eluent is diluted with 19 mL of
ethanol and left to stand for a few minutes. The solution is
filtered through a 25 mm, 0.45 µm preconditioned GDX-XP
(PVDF) filter into a 50 mL round-bottom flask. The solution is
evaporated under moderate vacuum (30 mbar/40 °C) until a
volume of approximately 1 mL remains. If volume is less than
1 mL, make up to 1 mL with ethanol/water 95/5 v/v. The
solution is transferred into a 2 mL autosampler vial and left for
1 h. The 1 mL solution is filtered through a preconditioned 0.2
mm PTFE filter into a fresh vial for injection.
Conclusion
The synthesis of zaurategrast sulfate 1 is an example of a
process for which fulfilling the specifications involved condi-
tions that potentially lead to GTI presence in the API. At an
early stage of development, the management of such a potential
risk is often in conflict with the tight timelines allocated to the
drug development program. In such cases, the development of
an appropriate analytical method in order to track ppm levels
of impurities, where there is also a high degree of interferences,
takes time and resources.
A valuable solution (when applicable) is therefore to manage
the risk through the appropriate application of chemical process
design as presented in this paper. Based on a strong rationale
the implementation of aqueous washes to eliminate acetamide
proved retrospectively to be successful when an appropriate
analytical method was available. Two manufacturing campaigns
of 250 kg and 750 kg of zaurategrast sulfate 1 were successfully
accomplished, giving API within the required specification of
the overbrominated impurity and below the acceptance criterion
for acetamide content.
The availability of the analytical method gave us then the
opportunity to assess and challenge the designed process. This
evaluation unambiguously demonstrated its high performance
to eliminate acetamide potentially formed during the bromina-
tion reaction. Additionally, this evaluation allowed the adjust-
ment of the designed process to a high level of confidence by
implementing a single aqueous wash instead of the three ones
initially introduced. This process adjustment was unfortunately
not performed at industrial scale due to the discontinuation of
the project at the end of phase II clinical trials.
Finally, from a commercial API production perspective, the
validated, sensitive GC/MS method used to detect acetamide
at ppm levels was not suitable for QC routine analysis.
The foreseen strategy therefore was to combine a less sensitive
detection at the crude stage with our knowledge of the acetamide
rejection capacity of the recrystallization. An initial spike of
1000 ppm acetamide (see Table 4) resulted in a level of 54
ppm in crude API which in turn gave 0.77 ppm in the
recrystallized product. Given that the acceptance criterion for
acetamide in API was 0.75 ppm, a level of 25-30 ppm
acetamide could have been proposed for the crude API. Details
of the overall chemical process for zaurategrast sulfate 1
synthesis will be presented in a separate communication.
NMR spectra were recorded on an Oxford AS400 instru-
ment. Elemental analyses were performed on an Elementar
Vario EL3 apparatus. All reagent quantities quoted are relative
to starting UCB1193394, assuming 100% purity, unless oth-
erwise stated.
Zaurategrast Sulfate 1 Formation. A double-jacketed glass
reactor equipped with anchor is charged with UCB1193394 (1
equiv), acetonitrile (1 vol), and water (1 vol). The suspension
is stirred at 210 rpm at 20 °C. To that suspension is added
hydrochloric acid (1.1 equiv) maintaining T e 25 °C, and the
resulting mixture is stirred until complete dissolution. The
Experimental Section
Analytical Method for the Determination of Acetamide
Content. The analytical method for the determination of
acetamide content in crude and pure APIs consists of a GC/
MS determination using single ion monitoring (SIM) for mass
detection. The GC/MS system is an Agilent 6890 GC system
Vol. 14, No. 4, 2010 / Organic Process Research & Development
•
1013

solution is cooled to 0 °C, and a cold solution of NBS (1 equiv)
in acetonitrile/water (2.6/0.4 vol) at 8-10 °C (prepared in a
separate vessel equipped with stirring) is added, maintaining
the reaction temperature at 0 °C. IPC is performed 30 min after
the end of the addition. UCB1193394 IPC specification is
<0.5% PAR left. The reaction is quenched with Na2SO3 (0.04
equiv) in water (1 vol). The temperature is increased to 40 °C.
Acetonitrile is then removed by azeotropic distillation (40 °C/
vacuum) to end up with protonated free base in water (1.6 vol
theor.). EtOAc (4 vol) is added at 40 °C and maintained for
the following workup: a solution of KHCO3 (1.1 equiv) in water
(1 vol) is added. Phases are separated and the aqueous one
discarded. The organic layer is washed with 5% w/w aqueous
NaCl solution (3 vol) three times. Ethyl acetate is switched to
EtOH (8 vol)a by azeotropic distillation. The temperature is set
to 45 °C, and water (2.2 vol)a is added followed by a solution
of H2SO4 (0.48 equiv) in water (1 vol)a maintaining T at 45
°C. The solution is seeded with 1 (0.32% w/w). The crystal-
lization appears within 5 min. The suspension is stirred
vigorously an extra 10 min and cooled to 0 °C in 4 h. The
product is isolated by filtration, washed with cold EtOH/water
(70/30, 2.4 vol)a and cold water (2 vol)a to obtain zaurategrast
sulfate 1 wet considered at 35% LoD (water).
5.8, 1H), 7.22 (d, J ) 8.6, 2H), 7.13 (d, J ) 5.8, 1H), 4.36 (s,
1H), 4.21-4.12 (m, 3H), 3.13 (dd, J ) 13.9, J ) 5.3, 1H),
2.98 (dd, J ) 13.6, J ) 9.3, 1H), 1.75-1.44 (m, 9H), 1.19 (t,
J ) 7.1, 3H), 1.17 (m, 1H). 13C NMR (100 MHz, DMSO-d6):
δ ) 190.8, 177.1, 170.6, 153.2, 148.1, 147.0, 145.4, 140.3,
139.0, 130.5, 129.2, 120.8, 119.1, 113.7, 110.3, 94.7, 61.7, 61.0,
59.1, 36.1, 30.9, 25.3, 23.6, 14.0. Anal. Calcd for C28H30N4O3:
C, 71.47; H, 6.43; N, 11.91. Found: C, 71.62; H, 6.55; N, 11.85.
UCB1184197 4: 1H NMR (400 MHz, DMSO-d6): δ ) 9.84
(s, 1H), 9.57 (s, 1H), 8.98 (d, J ) 9.1, 1H), 8.66 (d, J ) 5.5,
1H), 8.16 (d, J ) 5.5, 1H), 7.81 (d, J ) 8.6, 2H), 7.69 (d, J )
5.8, 1H), 7.23 (d, J ) 8.6, 2H), 7.14 (d, J ) 5.5, 1H), 4.80 (m,
1H), 4.20 (q, J ) 7.1, 2H), 3.20 (dd, J ) 13.8, J ) 4.5, 1H),
3.00 (dd, J ) 13.8, J ) 9.9, 1H), 1.70-1.65 (m, 2H), 1.65-1.48
(m, 6H), 1.40 (m, 1H), 1.23 (t, J ) 7.1, 3H), 1.13 (m, 1H). 13C
NMR (100 MHz, DMSO-d6): δ ) 186.5, 173.2, 170.1, 153.1,
148.1, 147.0, 145.3, 140.2, 139.0, 130.0, 129.2, 120.8, 119.1,
113.7, 110.3, 67.3, 61.6, 61.2, 57.4, 37.0, 30.3, 24.9, 23.4, 13.9.
Anal. Calcd for C28H29BrN4O3: C, 61.21; H, 5.32; N, 10.02.
Found: C, 61.16; H, 5.42; N, 10.11.
UCB1191133 5: 1H NMR (400 MHz, DMSO-d6): δ ) 9.85
(s, 1H), 9.77 (s, 1H), 8.98 (d, J ) 9.1, 1H), 8.81 (d, J ) 5.5,
1H), 8.34 (s, 1H), 7.73 (dapp, J ) 7.1, 3H), 7.24 (d, J ) 8.6,
2H), 4.80 (m, 1H), 4.20 (q, J ) 7.1, 2H), 3.21 (dd, J ) 13.8,
J ) 4.5, 1H), 3.01 (dd, J ) 13.9, J ) 10.1, 1H), 1.75-1.66
(m, 2H), 1.66-1.48 (m, 6H), 1.40 (m, 1H), 1.23 (t, J ) 7.1,
3H), 1.13 (m, 1H). 13C NMR (100 MHz, DMSO-d6): δ ) 186.4,
173.9, 170.2, 152.8, 148.7, 146.5, 138.6, 138.5, 130.8, 129.4,
117.5, 115.0, 105.1, 67.4, 61.7, 61.3, 57.4, 37.0, 30.5, 25.0,
23.5, 14.0. Anal. Calcd for C28H28Br2N4O3: C, 53.52; H, 4.49;
N, 8.92. Found: C, 53.15; H, 4.53; N, 8.89.
aVolumes relative to 100% theoretical yield of zaurategrast
from this point.
The wet zaurategrast sulfate 1 crude considered at 35% LoD
is suspended in a mixture of EtOH (6.4 volb) and water (2.5
volb). The suspension is heated to 70 °C to obtain a solution,
and polish-filtered to a clean vessel. The temperature is adjusted
to 53 °C in -5 °C/h and seed material added (0.2% w/wb).
The temperature is held at 53 °C for one hour, allowing
crystallization to set, then cooled according to the following
controlled temperature profile: 53 °C to 50 °C in -5 °C/h,
temperature held at 50 °C for one hour, 50 °C to 45 °C in -5
°C/h, temperature held at 45 °C for one hour, 45 °C to 40 °C
in -5 °C/h, temperature held at 40 °C for one hour, 40 °C to
-10 °C in 8 h. The suspension is filtered. The cake is washed
with cold EtOH/water (70/30, 1 volb), then with EtOH/water
(15/85, 1 volb) so as to reduce the EtOH content of the wet
cake. The product is then dried under vacuum at 40 °C until
the LOD is ∼4-5% (in the lab: dried then rehydrated in a 58%
RH chamber overnight using saturated NaBr solution at 25 °C).
bCalculated on theoretical zaurategrast sulfate 1 crude content
with the stated 35% LoD.
Acknowledgment
We thank Petra Gartmann and Marc Speleers for acetamide
analytical support during process performance assessment and
Renaud Brysse for elemental analysis. We also thank Dolors
Terricabras for her project management support during this
work.
Supporting Information Available
1H and 13C NMR spectra are available for UCB1193394 (2),
zaurategrast (4) and UCB1191133 (5). This material is available
free of charge via the Internet at http://pubs.acs.org.
UCB1193394 2: 1H NMR (400 MHz, DMSO-d6): δ ) 9.84
(s, 1H), 9.55 (s, 1H), 8.65 (d, J ) 5.5, 1H), 8.42 (d, J ) 8.6,
1H), 8.16 (d, J ) 5.8, 1H), 7.78 (d, J ) 8.6, 2H), 7.68 (d, J )
Received for review December 17, 2009.
OP900330E
1014
•
Vol. 14, No. 4, 2010 / Organic Process Research & Development