1,3,5-Trialkyl-2,4,6-triiodobenzenes: Novel x-ray contrast agents for gastrointestinal imaging

Article abstract of DOI:10.1021/jm990407i

Examination of the gastrointestinal (GI) tract has been performed for decades using barium sulfate. Although this agent has many recognized limitations including extreme radiopacity, poor intrinsic affinity for the GI mucosa, and very high density, no alternative contrast agents have emerged which produce comparable or better contrast visualization. In fact, the various techniques of the GI radiologic examination (i.e., single contrast, double contrast, biphasic) were developed to compensate for its limitations. Each of these techniques requires complex patient manipulation to achieve adequate mucosal coating or compression to overcome the marked radiopacity of barium sulfate in order to obtain a diagnostically useful examination. A series of novel radiopaque oils, the 1,3,5-trialkyl-2,4,6-triiodobenzenes, was designed to improve the efficacy, stability, and safety of barium formulations. These substances were prepared in two steps from 1,3,5- trichlorobenzene. Compound 17 (1,3,5-tri-n-hexyl-2,4,6-triiodobenzene), formulated as an oil-in-water emulsion, was found to be well-tolerated in rodents (mice, hamsters, rats) following acute oral and/or intraperitoneal administrations at 4 times the anticipated human clinical dose. No metabolism of 17 was detected in rat, hamster, dog, monkey, or human hepatic microsomes, suggesting the lack of oral toxicity was a consequence of poor absorption. In imaging experiments in dogs, emulsions of 17 have demonstrated excellent mucosal coating and improved radiodensity relative to barium sulfate suspensions. On the basis of the preliminary imaging and toxicity data, compound 17 was selected as a potential development candidate.

Full text of DOI:10.1021/jm990407i

1940
J . Med. Chem. 2000, 43, 1940-1948
1,3,5-Tr ia lk yl-2,4,6-tr iiod oben zen es: Novel X-r a y Con tr a st Agen ts for
Ga str oin testin a l Im a gin g1
Kimberly G. Estep,*,2 Kurt A. J osef,2 Edward R. Bacon,2 Carl R. Illig,2 J ohn L. Toner,2 Dinesh Mishra,3
William F. Blazak,4 Dennis M. Miller,4 David K. J ohnson,5 J ack M. Allen,6 Andy Spencer,4 and Susan A. Wilson4
Nycomed Amersham Imaging, 466 Devon Park Drive, P.O. Box 6630, Wayne, Pennsylvania 19087-8630
Received August 11, 1999
Examination of the gastrointestinal (GI) tract has been performed for decades using barium
sulfate. Although this agent has many recognized limitations including extreme radiopacity,
poor intrinsic affinity for the GI mucosa, and very high density, no alternative contrast agents
have emerged which produce comparable or better contrast visualization. In fact, the various
techniques of the GI radiologic examination (i.e., single contrast, double contrast, biphasic)
were developed to compensate for its limitations. Each of these techniques requires complex
patient manipulation to achieve adequate mucosal coating or compression to overcome the
marked radiopacity of barium sulfate in order to obtain a diagnostically useful examination. A
series of novel radiopaque oils, the 1,3,5-trialkyl-2,4,6-triiodobenzenes, was designed to improve
the efficacy, stability, and safety of barium formulations. These substances were prepared in
two steps from 1,3,5-trichlorobenzene. Compound 17 (1,3,5-tri-n-hexyl-2,4,6-triiodobenzene),
formulated as an oil-in-water emulsion, was found to be well-tolerated in rodents (mice,
hamsters, rats) following acute oral and/or intraperitoneal administrations at 4 times the
anticipated human clinical dose. No metabolism of 17 was detected in rat, hamster, dog, monkey,
or human hepatic microsomes, suggesting the lack of oral toxicity was a consequence of poor
absorption. In imaging experiments in dogs, emulsions of 17 have demonstrated excellent
mucosal coating and improved radiodensity relative to barium sulfate suspensions. On the
basis of the preliminary imaging and toxicity data, compound 17 was selected as a potential
development candidate.
In tr od u ction
curacy. In fact, a number of techniques (single contrast,
double contrast, etc.) of the radiologic GI exam were
developed around the limitations of barium sulfate.
Each technique required manipulation of the patient to
achieve adequate mucosal coating and compensate for
the high radiopacity of barium sulfate. Moreover, barium
sulfate is highly toxic in the peritoneal cavity and is
therefore contraindicated in patients with suspected
bowel perforations or when postimaging surgery or CT
examination is anticipated.9 In such cases, soluble
iodinated contrast media are the preferred agents.
X-ray visualization of the gastrointestinal (GI) tract
requires the use of a radiopaque contrast agent, as the
differences in radiodensity between the bowel and
surrounding soft tissues are too small to allow sufficient
delineation. Suspensions of barium sulfate have been
used for radiographic imaging of the GI tract since 1910
and have remained the most widely used radiocontrast
media to detect pathologic conditions such as ulcers,
inflammatory bowel disease, colonic polyps, and gastric
cancer.7 Despite the wide-spread use of barium sulfate
for both upper and lower GI examinations, this agent
has many recognized limitations. Its marked radioden-
sity and poor intrinsic affinity for the GI mucosa lead
to inaccurate diagnoses resulting from poor coating of
the contrast agent in the GI tract. Since the degree of
opacification produced by a barium preparation is
directly proportional to the density of the medium and
the thickness of the coating, in certain GI examinations
requiring a high-density barium preparation, the ex-
tremely radiopaque contrast medium may not permit
X-ray penetration of the intestinal lumen, thereby
interfering with the ability to discriminate lesions.8
Suspensions of barium sulfate settle quickly due their
high density, and the limited stability of the suspension
in vivo gives rise to pooling and dispersion of the
formulation, which also interferes with diagnostic ac-
Water-soluble iodinated organic compounds have been
used extensively as contrast media due to the ability of
iodine to attenuate X-rays.10 None of these agents has
been routinely applied to GI imaging. Aqueous solutions
of water-soluble iodinated contrast agents have been
used for opacification of the GI tract when barium is
contraindicated and under certain conditions offer
advantages over barium sulfate.11 However, the hyper-
osmolarity of these parenteral formulations gives rise
to a new set of complications regarding efficacy and
safety. Further issues with respect to poor palatability,
hypertonicity, low radiodensity made poorer by osmotic
dilution, and poor intrinsic coating of the GI mucosa
have prevented their general acceptance as viable
alternatives to barium for routine GI examinations.
Emulsions of water-insoluble iodinated organic oils
have also been studied in GI radiology, although no
commercially viable agent has yet been identified. In
1964 Teplich and co-workers reported that an oil-in-
* Corresponding author, currently at Pfizer Central Research,
Eastern Point Road, P.O. Box 8002, Groton, CT 06340-8002. Tel: 860-
441-1757. Fax: 860-715-7495. E-mail: estep@pfizer.com.
10.1021/jm990407i CCC: $19.00 © 2000 American Chemical Society
Published on Web 05/02/2000

1,3,5-Trialkyl-2,4,6-triiodobenzenes
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 10 1941
water emulsion of Ethiodol, an iodinated poppy seed oil,
produced mucosal coating of the small bowel superior
to either barium sulfate or water-soluble contrast
media.12 However, between 50% and 70% of the iodi-
nated component was absorbed from the GI tract of
dogs, raising concerns regarding systemic exposure to
large amounts of this substance. Oil-in-water emulsions
of iophendylate (1), a radiopaque oil which was devel-
oped in the 1940s for myelography,13 have also been
studied by Pirkey and co-workers as contrast media for
GI imaging.14 However, they observed no advantage
over conventional contrast agents. More recently, Rubin
reported that emulsions of 1 gave small bowel images
superior to that of conventional barium.15 Although no
adverse effects were observed following repeated oral
administration of iophendylate formulations to dogs,16
this emulsion was highly toxic to mice after acute oral
dosing (vide infra).
As part of our research in novel X-ray contrast agents,
we found the consistency and overall image quality
obtained with emulsions of these iodinated oils were
impressive and reproducible, and we sought to develop
an iodinated organic molecule demonstrating acceptable
acute safety, which when formulated as an oil-in-water
emulsion would reproduce the continuous, uniform
coating and fine detail observed with iophendylate
formulations but without the associated toxicity. Herein
we report the synthesis, formulation, safety evaluation,
and imaging of a series of novel radiopaque compounds,
the 1,3,5-trialkyl-2,4,6-triiodobenzenes, which were de-
signed to overcome both the limitations of barium
sulfate and the liabilities associated with existing
iodinated oils as GI contrast agents.
safety of barium sulfate in an uncompromised bowel is
attributed to extremely low aqueous solubility and
minimal absorption from the GI tract. Although chemi-
cally inert and relatively nonabsorbed orally, extrava-
sation of barium sulfate into the peritoneal cavity
through either perforated colon, duodenum, or stomach
may lead to serious reactions ranging from peritonitis,
formation of dense adhesions, and even death. For these
reasons, barium contrast is generally contraindicated
in patients with known or suspected perforations and
alternative contrast agents are required. Intravascular
exposure of barium sulfate is rare but associated with
serious or fatal complications including liver abscesses
and pulmonary and cardiac embolizations.7 Therefore
to provide an advantage over barium sulfate, an iodi-
nated contrast agent should not only demonstrate poor
oral bioavailability but also be nontoxic upon systemic
exposure, as in the case of a bowel perforation or
accidental aspiration following oral administration.
Issues pertaining to synthetic feasibility also were
considered in the design of a novel radiopaque oil.
Lengthy synthetic processes would render the agent too
costly for routine diagnostic use; thus sequences were
limited to a maximum of three steps from relatively
inexpensive starting materials. From a practical stand-
point, the quantity of drug substance required for our
preclinical screening protocol mandated that synthetic
candidates be prepared readily on multigram scale.
The physical properties of the products also presented
unanticipated purification issues. Target compounds
were noncrystalline substances with molecular weights
> 700 amu. Distillation was difficult, as boiling points
often exceeded 200 °C at <0.001 mmHg. Even if the
compounds proved thermally stable at these tempera-
tures, the vapors were often too dense to permit the use
of conventional distillation procedures. Column chro-
matography was unacceptable for large-scale purifica-
tion. Thus for a radiopaque oil to be considered a viable
candidate, the synthetic process must afford the final
product with >98% purity without the need for column
chromatography or distillation.
The 1,3,5-trialkyl-2,4,6-triiodobenzenes meet all of the
above criteria. They are structurally novel, free-flowing
oils which appear to be chemically and pharmacologi-
cally inert. The flexibility of the long alkyl chains
efficiently offsets the inherent crystallinity of a 1,3,5-
triiodobenzene framework. They are prepared cleanly
in two steps. Most importantly, when formulated as an
oil-in-water emulsion, 1,3,5-tri-n-hexyl-2,4,6-triiodoben-
zene (17) provides excellent images of the GI tract with
no apparent toxicity.
Ch em istr y
Design Con sid er a tion s. In designing an iodinated
organic molecule to use as contrast media for GI
radiographic examinations, a number of issues relating
to chemical composition and synthetic feasibility of the
agent required consideration. X-ray attenuation is
directly proportional to the amount of iodine contained
within a molecule, and doses of contrast media are
calculated based on the weight percent of iodine present
in the radiopaque component. Therefore, to minimize
the quantity of a contrast agent administered during a
GI imaging study, one primary objective was to maxi-
mize the number of iodine atoms contained within a
given molecule. At the same time, to facilitate the
formulation of these agents as stable emulsions, the
melting points of viable compounds could not signifi-
cantly exceed room temperature. Iodine which is bonded
directly to a benzene ring has the highest degree of
chemical stability and biological safety,17 although this
structural framework imparts a high degree of crystal-
linity to substances which should ideally be nonviscous
oils. These two opposing properties, low melting point
and high iodine content, added to the challenge of
successful design.
Ch em ica l Syn th esis. The 1,3,5-trialkyl-2,4,6-tri-
iodobenzenes were synthesized according to the general,
two-step procedure outlined in Scheme 1. In the initial
step, the method of Tamborski18a was utilized to couple
1,3,5-trichlorobenzene (2) with alkyl Grignards under
nickel catalysis. The resulting 1,3,5-trialkylbenzene
intermediates 3 were then iodinated with N-iodosuc-
cinimide (NIS) in the presence of catalytic amounts of
trifluoromethanesulfonic acid (TfOH) to afford the
desired products 4 (Table 1).
Other critical factors such as toxicity also required
consideration. Given the large quantities of contrast
agent administered for routine GI examinations, the
radiographic agent should ideally be nontoxic and
pharmacologically inert. Chemical functionality associ-
ated with biological recognition was to be avoided. The
Several issues related to the success of this synthetic
sequence warrant comment. The use of NIS/TfOH was
critical to obtain the products with >98% purity, and

1942 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 10
Estep et al.
Sch em e 1a
Sch em e 2
Sch em e 3a
a
(a) RMgBr, NiCl2(dppp) (cat.), Et2O; (b) method A: I2,
PhI(O2CCF3)2, CCl4; (c) method B: NIS, TfOH, ClCH2CH2Cl,
reflux; dppp ) 1,3-bis(diphenylphosphino)propane.
over 100 different iodination methods were explored
before identifying these conditions.18b The only previ-
ously reported compound in the series is the trimethyl
derivative 12, which had been synthesized to exemplify
methodologies for the iodination of aromatic rings.19
These conditions included the treatment of mesitylene
with iodine and mineral acids,20 NIS and p-toluene-
sulfonic acid,21 thallium(III) trifluoroacetate and tri-
fluoroacetic acid,22 benzyltrimethylammonium dichloro-
iodate and zinc chloride,23 bis(pyridine)iodonium(I)
tetrafluoroborate and TfOH,24 and anodic oxidation of
iodine.25 However, none of these procedures was judged
adequate for our purposes, due to incomplete iodination
of the trialkylbenzene intermediates. The resulting
diiodo and monoiodo impurities 5 and 6, respectively,
were extremely difficult to remove from the structurally
similar triiodo products, especially on large scale. Thus
any iodination conditions which gave incomplete con-
version to 4 were not acceptable.
a
(a) Method B.
tive sources of 7 were sought. The direct treatment of 3
with NIS and trifluoroacetic acid was disappointing, as
it gave the usual mixture of 4-6. However, contrary to
previous iodination attempts, this mixture was devoid
of decomposition products and the reaction proceeded
in very high yield. We were encouraged and reasoned
that by increasing the reactivity of the hypoiodide
intermediate, the iodination would proceed to comple-
tion. To that end, 3 was treated with NIS and catalytic
TfOH (method B), conditions which consistently afforded
the desired 4 in excellent yields.27 The absence of
iodinated contaminants permitted an abbreviated work-
up, consisting of an aqueous wash and filtration through
a short plug of silica gel.
One method which routinely afforded complete tri-
iodination involved the treatment of 3 with bis(trifluo-
roacetoxy)iodobenzene and iodine in carbon tetrachlo-
ride (method A).26 While these conditions afforded 4 in
high yields at room temperature, the products were
contaminated with 1,4-diiodobenzene (8) which arose
from the competing iodination of the iodobenzene gener-
ated during the course of the reaction (Scheme 2). While
8 could be separated via recrystallization from the
products which were solids (i.e., 12-16, 21, 22), 8 was
readily soluble in analogues which were oils. Methods
to physically remove this crystalline impurity from the
desired oils via sublimation, steam distillation, or low-
temperature crystallization were not successful. Chemi-
cal methods, such as selective reduction, were also
unsuccessful. It was possible to obtain analytically pure
oily analogues (17, 18) via method A after repeated
chromatography of the product mixture; however, this
was not suitable for large-scale synthesis.
The iodinated impurities 5 and 6 were not the only
contaminants to impact the synthetic process used for
the 1,3,5-trialkyl-2,4,6-triiodobenzenes. When method
B was first applied to a 100-g batch of 1,3,5-tri-n-
hexylbenzene (9; Scheme 3), HPLC analysis of the
resulting 17 revealed unacceptable levels of another
impurity (10%) which could not be separated from the
desired product. This contaminant was subsequently
identified as the chlorinated compound 10, which ulti-
mately originated from the incomplete coupling of 2 with
n-hexylmagnesium bromide (Scheme 1). The amount of
11 formed in the coupling could be minimized by using
excess Grignard; however, even the use of >10 equiv
did not effect complete conversion of 2 to 9. It was
absolutely critical that the purity of 9 was at least 98%
prior to the iodination step, as impurities could not be
The complete conversion of 3 to 4 via method A was
noteworthy and encouraging and was attributed to the
reactive intermediate trifluoroacetyl hypoiodide (7). To
avoid the formation of 8 or similar byproducts, alterna-

1,3,5-Trialkyl-2,4,6-triiodobenzenes
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 10 1943
Ta ble 1. Syntheses and Physical Properties of
1,3,5-Trialkyl-2,4,6-triiodobenzenes
removed from the product 17 (vide supra). Although the
boiling points of 9 and 11 were similar, it was possible
to separate them via fractional distillation.28 Unfortu-
nately, the increase in purity of 9 (88.6% of the crude
mixture, increased to 98.7% after distillation) was
achieved at the expense of chemical yield (>99% crude
yield, decreased to 48% after distillation). However, once
sufficient purity was attained, 9 (>98% pure) was
readily converted to 17 (>98% pure) using method B.
This iodination proceeded in 96% yield on a 2.5-kg scale.
Several 1,3,5-trialkyl-2,4,6-triiodobenzenes were pre-
pared in which the alkyl chains contained a terminal
methyl branch (21-23). Attempts to triiodinate 1,3,5-
triisopropylbenzene or 1,3,5-triisobutylbenzene via
method B afforded mixtures with varying degrees of
iodination. These reactions could not be driven to
completion with excess reagents or increased temper-
ature, presumably due to steric reasons. When the
methyl branch was a minimum of three bonds removed
from the benzene ring (21-23), complete triiodination
could be achieved. Interestingly, it was necessary to
increase the amount of NIS to over 4 equiv for this
reaction to proceed to completion. Conditions which
consistently afforded 17 gave only diiodinated product
when applied to the isooctyl precursor of 23. The source
of this reactivity difference is unclear, but it appears to
be common to branched analogues.
compd
R
method % yielda formula % I
mp, °C
76.5 209-11f
12
13
14
15
16
17
18
19
20
21
22
23
CH3
C2H5
B
B
A
A
A
B
A
B
B
A
A
B
80b
85b,c
5d
C9H9I3
C12H15I3 70.6 195-96
C15H21I3 65.5 95-96
C18H27I3 61.1 92-93
C21H33I3 57.2 71-72
C24H39I3 53.8 oil
C27H45I3 50.8 oil
C30H51I3 48.1 oil
C33H57I3 45.7 oil
n-C3H7
n-C4H9
n-C5H11
n-C6H13
n-C7H15
n-C8H17
n-C9H19
i-C5H11
i-C6H13
i-C8H17
8d
40d
96e
35
70
88
5d
C21H33I3 57.2 86-87
C24H39I3 53.8 87-89
C30H51I3 48.1 oil
31d
81
a
Yields for solids based on first crop after recrystallization.
Recrystallized from xylenes. c Compound 13 obtained in 56%
b
yield (method A) following recrystallization from cyclohexane.
d
Recrystallized from ethanol. e Compound 17 obtained in 55%
yield using method A. f Lit.20b mp 208 °C.
Ta ble 2. Murine Toxicity Following a Single Oral Dose
dosea,g
mortality,
compdb
% Ic vol, mL/kg g compd/kg g I/kg
day 7d
1
30.5
10
5.34
2.80
6.26
12.5
8.49
7.00
1.63
0.853
1.91
3.81
4.56
4.56
3/3e
1/3
1/3
1 (neat) 30.5
1 (neat) 30.5
1 (neat) 30.5
17
24
Resu lts a n d Discu ssion
3/3
P h ysica l P r op er ties of 1,3,5-Tr ia lk yl-2,4,6-tr i-
iod oben zen es. As noted in Table 1, the melting points
of the 1,3,5-trialkyl-2,4,6-triiodobenzenes decrease as
the length of the alkyl chain increases. The n-hexyl
derivative 17 and analogues with longer n-alkyl chains
(18-20) are free-flowing oils at room temperature,
although the 5-methylhexyl analogue 22 is a solid. None
of the 1,3,5-trialkyl-2,4,6-triiodobenzenes has measur-
able solubility in water, and the oily compounds are
insoluble in alcohols or dipolar aprotic solvents. They
have slight solubility in acetone and are readily soluble
in nonpolar solvents such as hexanes and dichloro-
ethane. The compound with the highest iodine content
(weight percent) to remain an oil at room temperature
was 17. Thus, this was chosen as the prototypical
analogue to examine in safety and imaging studies.
Rod en t Sa fety Eva lu a tion a n d Th yr oid Toxicity.
On the basis of considerations described above, the
safety profile needed for a novel emulsion-based iodi-
nated oral contrast agent to gain regulatory approval
was expected to be high. While the ideal agent would
be poorly absorbed, nontoxic, and pharmacologically
inert, a rigorous screening program was established
early in the discovery program to help guide synthetic
efforts. Toxicity studies in the contrast media field are
performed to determine drug-related effects that cannot
be evaluated in standard pharmacology profiles or that
may occur after repeat administration. This takes on
particular significance given the large (multigram)
quantities of contrast agent that are often administered
in a radiographic procedure. A two-tiered approach was
utilized to assess the acute toxicity of drug candidates.
An initial characterization of gross oral toxicity was
obtained through a 7-day, single administration, as-
cending dose test in mice. The novel radiopaque oils,
53.7
65.2
40
40
0/3f
0/3f
a
Dosage volume and amount of iodinated compound contained
b
therein. Formulated as mineral oil-in-water emulsions, unless
noted. c Weight percent iodine contained in compound. Number
d
of decedents over the total number of animals in each dosage group
7 days after oral administration. e 100% mortality was observed
at 7 days in both the 20 mL/kg (3.27 g I/kg) and 40 mL/kg (6.53 g
I/kg) dose groups. f No signs of toxicity were observed in either
the low-dose (10 mL/kg; 1.14 g I/kg) or middose (20 mL/kg; 2.28 g
I/kg) groups. g Emulsion vehicle, dosed at 40 mL/kg, served as the
control group. No signs of toxicity were observed at this dose level.
which possessed acceptable physical properties (i.e.,
chemical stability, mp < 30 °C, emulsifiable) were
formulated as light mineral oil-in-water emulsions at a
concentration of 100-160 mg I/mL and were adminis-
tered by oral gavage to mice at three volumes: 10, 20,
and 40 mL/kg body weight, corresponding to ca. 1-, 2-,
and 4-fold the anticipated clinical dosage (114 mg I/mL)
for humans. The emulsion vehicle (without contrast
agent) served as a control group and was dosed at 40
mL/kg, the maximum volume that could be reasonably
administered orally. Animals were observed for 7 days,
at which point the surviving mice were killed and a
necropsy was performed. Compounds which produced
mortality, adverse clinical symptoms (i.e., CNS effects),
weight loss, or gross lesions at necropsy were removed
from further consideration. The results of the acute
murine toxicity screen for compounds 1, 17, and 24 are
summarized in Table 2. The emulsion of iophendylate
(1) caused death in all animals at all doses within 3
days. These results were surprising given the apparent
lack of toxicity in dogs,15 and the cause of this discrep-
ancy is unclear. To probe the relationship between the
formulation of this agent and toxicity, neat 1 was

1944 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 10
Estep et al.
Ta ble 3. Hamster Thyroid Toxicity 48 h Following a Single
Oral Dose
bations than rats, an acute hamster assay was devised
to provide a quick indication of thyroid toxicity. Fur-
thermore, the hamster was a preferred species for
toxicity testing because it lacks a forestomach, unlike
rats. In this assay, hamsters received a single dose of
test article orally, and after 48 h, thyroid glands were
removed and examined histomorphologically. In the
screening protocol, compounds which passed the acute
murine toxicity screen now proceeded directly into this
assay for thyroid toxicity.
sex
(n)
dose,
mitotic
figuresc
compda
g I/kgb
necrosisd
dilatione
vehicle
vehicle
17
17
23
M (6)
F (6)
M (3)
F (3)
M (5)
M (3)
F (3)
0
0
0/6
0/6
1/3f
0/3
0/5
3/3
3/3
0/6
0/6
0/3
0/6
0/5
2/3
0/3
0/6
0/6
1/3f
0/3
0/5
3/3
3/3
4.56
4.56
4.56
4.56
4.56
24
24
a
Formulated as mineral oil-in-water emulsions at 114 mg I/mL
The acute hamster thyroid assay proved a significant
hurdle for organic radiopaque oils to pass. Representa-
tive compounds from 10 different structural classes of
iodinated benzenes were evaluated for thyroid toxicity
in the hamster, including iodinated phenyl esters,
phenyl ethers, benzoates, and isophthalates. While the
radiopaque oils share some structural similarities to
thyroxine (T4) and related thyroid hormones, the mech-
anism by which 24 induced thyroid toxicity is unclear,
and more importantly, this is not a property shared by
barium sulfate. The only molecules which did not induce
thyroid toxicity were 17 and 23, both of which belong to
the class of 1,3,5-trialkyl-2,4,6-triiodobenzenes (Table
3). Compound 17 could now be differentiated from 24
in a rodent model of toxicity. Analysis of plasma samples
drawn from hamsters treated orally with 4.56 g I/kg 17
or 23 failed to reveal significant levels of parent drug
(minimum quantifiable level (MQL) ) 0.19 µg/mL). The
lack of acute toxicity of 17 and 23 therefore appeared
due to low bioavailability, although it was not clear
whether this was because of poor absorption, extensive
first-pass metabolism, or a combination of the two.
b
(see the Experimental Section). Dosage volume ) 40 mL/kg.
c Number of hamsters with follicular cell mitotic figures above
control values over total number of animals in each dosage group.
d
Number of hamsters with follicular cell necrosis over total
number of animals in each dosage. e Number of hamsters with
follicular cell dilation over total number of animals in each dosage
group. f Changes were minimal in severity.
examined in this assay. While toxicity was diminished
compared to that of the emulsion, neat 1 induced
mortality in one of the three mice tested at the lowest
dose. These results were sufficient to suspend further
evaluation of this agent. The remaining candidates (17,
24) could not be readily differentiated in the mouse and
were then evaluated in other acute rodent models for
which the hamster became the second tier study.
To further address this issue, the rates of in vitro
metabolism of 17 and 23 were determined in rat,
hamster, and human liver microsomes. Since 24 had
previously been shown to be extensively metabolized in
all three species, [14C]24 was used as a positive control
in this study. Compounds (substrate concentrations of
4 and 40 µM) were incubated for 60 min with mi-
crosomes, and aliquots were removed and then analyzed
by isocratic reverse-phase HPLC. At the lower substrate
concentration (4 µM), the rate of metabolism (half-life)
of 24 in the hamster was 15 min compared to the rat
(36 min) and human (132 min). At the higher substrate
concentration, metabolism of 24 was generally less
extensive with 73-100% of intact drug remaining after
60 min in the three test systems. In contrast, both 17
and 23 exhibited little or no metabolism in any of the
microsomal preparations after a 60-min incubation at
either substrate concentration. Intact drug accounted
for between 99% and 111% of the control values (0 min),
indicating low rates of metabolism across a range of
species. These data indicate that 17 and 23 were poorly
metabolized in the three microsome preparations and
that the lack of thyroid toxicity observed in the hamster
may be a result of poor absorption rather than rapid
and extensive systemic clearance following metabolism.
Compound 2429 was initially identified as an alterna-
tive to 1 for GI imaging.30 Although no signs of toxicity
(Table 2) were evident in mice after a single oral
administration of 24 at 4-fold the anticipated clinical
dosage (4.56 g I/kg; see Preclinical Imaging Studies),
24 induced histomorphologic changes in the thyroids of
hamsters 48 h after administration of this dose (Table
3). These changes included follicular cell mitosis, ne-
crosis, and dilation. Thyroid gland changes also were
evident in rats after a single administration of 24 at
1.14 g I/kg, although these effects were not as pro-
nounced as in the hamster. To assess the possible
specificity of thyroid toxicity for rodents, followup stud-
ies were performed with 24 in the dog. While no
abnormalities were found after a single administration
at 4-fold the clinical dosage, dogs receiving this dosage
six times over a 2-week period showed increased thyroid
weight accompanied by histomorphologic abnormalities.
Therefore, 24 induced thyroid toxicity in all species
tested, although the sensitivity greatly varied. Further
development of 24 for GI imaging was therefore termi-
nated. Despite the very similar profile of compounds 17
and 24 in the acute murine screen, thyroid toxicity
became the principal hurdle to surmount for 17.
While poor absorption is a desirable property for a
GI contrast agent, the issue of toxicity upon systemic
exposure remained for these compounds. An agent
which is nontoxic upon leakage into the peritoneum
would offer a distinct advantage over barium sulfate.
To investigate this possibility with 17, the hamster
thyroid protocol was modified to mimic a bowel perfora-
Given the significance of thyroid toxicity with respect
to advancement of an iodinated contrast agent, it
became desirable to screen compounds for these effects
at a very early stage in the discovery effort. Since
hamsters had proven more sensitive to thyroid pertur-

1,3,5-Trialkyl-2,4,6-triiodobenzenes
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 10 1945
F igu r e 1. Plasma concentrations (µg/mL) after ip administra-
tion of 17 to hamsters. Bar indicates 1 SEM; the MQL for this
study was 0.19 µg/mL.
tion. In this study, hamsters were dosed intraperito-
neally (ip) with an emulsion of 17 at 57 and 570 mg
I/kg and orally at 570 and 4560 mg I/kg for comparison.
Plasma concentrations were monitored for 72 h, after
which time the animals were examined for thyroid
perturbations. As shown in Figure 1, increased systemic
exposure of 17 was evident after ip dosing at 570 mg
I/kg while plasma levels of 17 following oral dosing were
< MQL (0.19 µg/mL) at the same dosage level. Absorp-
tion and/or elimination of drug was slow after ip dosing
with detectable concentrations in plasma still present
at 72 h in the high-dose group. Significantly, no changes
in thyroid pathology were detected in animals treated
with 17 compared to controls, while hamsters which
received 24 ip exhibited thyroid toxicity similar to that
observed after oral administration of that agent. There
were no other indications of major organ toxicity in
hamsters treated with 17.
P r eclin ica l Im a gin g Stu d ies. Imaging studies were
performed in rats and dogs to assess the radiographic
properties of test formulations, designed to optimize
both formulation parameters (i.e., emulsion stability)
and imaging efficacy. Imaging efficacy was evaluated
in the upper GI tract using the degree and distribution
of radiopacity as a visual indicator of mucosal surface
coating. Preliminary imaging studies were conducted in
pentobarbital (125 mg/kg im) anesthetized rats. The
contrast agent was formulated at a concentration of 114
mg I/mL (see below) and introduced via gastric intuba-
tion at a dose of 10 mL/kg. Manipulations of the animals
were purposely avoided, thereby emphasizing the innate
ability of the test formulation to adhere to the mucosal
surface of the GI tract. The rodent radiographs provided
a means to screen formulations based on uniformity,
50% or greater mucosal coating of the small intestine,
F igu r e 2. Representative upper GI canine radiograph fol-
lowing oral administration of a conventional barium sulfate
suspension (Polibar, formulated at 70% w/v) taken 15 min
postoral administration at a dose of 10 mL/kg in a 12-14-kg
beagle. Air was administered just prior to the film.
and transradiation as evidenced by the ability to visual-
ize loops of the bowel that overlap each other. Promising
formulations were then prepared in larger volumes and
evaluated in a canine model. Criteria of success in the
dogs were based on (1) appearance of uniform mucosal
coating in at least 50% of the small intestine, (2)
persistence of mucosal coating for a minimum of 30 min
postadministration, (3) appearance of mucosal detail,
(4) multiple levels of transradiation, and (5) a high
degree of reproducible images among dogs studied. The
anticipated human clinical imaging dosage (iodine
concentration) was derived from dose ranging studies
(25-150 mg I/mL) in dogs comparing 24 (and 1 as a
prototype) with barium sulfate and Gastrograffin, a
water-soluble, iodinated contrast agent. The optimal
iodine concentration necessary to provide sufficiently
opaque images was found to be 114 mg I/mL (i.e., 228
mg of compound containing 50 wt % iodine). The current
formulation of 17 was found to satisfy all of the
aforementioned criteria.
Representative canine radiographs which compare a
conventional barium sulfate formulation with that of 17
are shown in Figures 2 and 3, respectively. The barium
sulfate used was a commercial Polibar preparation,

1946 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 10
Estep et al.
Con clu sion
With the advent of managed health care, the role of
routine radiological procedures is likely to depend on
the efficiency and cost-effectiveness of these examina-
tions. The labor-intensive techniques of GI examinations
were developed almost entirely around the limitations
of barium sulfate and involve manipulation of the
patient to achieve adequate mucosal coating and/or
abdominal compression to overcome the radiopacity of
barium. A contrast agent which offers greater sensitivity
for detecting subtle tissue abnormalities while increas-
ing procedural ease and patient throughput would
present advantages over current procedures.31 We have
identified a novel series of radiopaque oils which appear
to fulfill these criteria. Oil-in-water emulsions of 1,3,5-
trialkyl-2,4,6-triiodobenzenes have demonstrated excel-
lent mucosal coating and optimal radiodensity without
subject manipulation and offer improved physical sta-
bility in vivo when compared to barium sulfate suspen-
sions. The n-hexyl analogue 17 was minimally absorbed
from the GI tract, was resistant to metabolism, and
appeared safe following oral and intraperitoneal ad-
ministration. On the basis of these properties, 17 was
selected for additional studies.
Exp er im en ta l Section
Melting points were determined on a Thomas-Hoover melt-
ing point apparatus in open capillaries and are uncorrected.
Proton NMR (GE QE300), IR (Nicolet 10DX or 550 FT-IR
spectrophotometer), and mass spectra (Kratos Concept or a
Nermag R10-10C spectrometer) were consistent with the
assigned structures. 1H NMR multiplicity data are denoted
by s (singlet), d (doublet), dd (doublet of doublets), t (triplet),
m (multiplet), and br (broad). Coupling constants are in hertz
(Hz). Infrared spectra (IR) were measured as 1% KBr pellets
or neat oils. Carbon, hydrogen, and iodine elemental analyses
were performed by Quantitative Technologies, Inc., White-
house, NJ , and were within (0.4% of theoretical values.
Nonaqueous reactions were generally carried out under a N2
or Ar atmosphere.
F igu r e 3. Representative upper GI canine radiograph fol-
lowing oral administration of an oil-in-water emulsion of
compound 17 (formulated at 22.0% w/v oil, 118 mg I/mL) taken
30 min postoral administration at a dose of 10 mL/kg in a 12-
14-kg beagle. Air was administered 15 min postadministration
of the emulsion.
The 1,3,5-trialkylbenzenes were purchased from Aldrich
(1,3,5-trimethyl- and 1,3,5-triethylbenzene) or synthesized as
outlined in Scheme 1.32
Gen er a l P r ep a r a tion of 1,3,5-Tr ia lk ylben zen es 3. The
following procedure illustrates the general method utilized for
the preparation of all 1,3,5-trialkylbenzenes used to prepare
14-23. The crude products were obtained as colorless to straw-
colored, mobile oils and were used directly in the next step,
following workup, filtration through a short pad of silica gel,
and concentration in vacuo.
formulated as a 70% w/v suspension and administered
at a dose of 10 mL/kg to a 12-14-kg beagle dog.
Compound 17 was formulated as an oil-in-water emul-
sion (22.0% w/v oil, 118 mg I/mL) and administered at
10 mL/kg to a 12-14-kg beagle dog. In both studies, air
was administered prior to contrast agent administra-
tion. These images are typical and highly reproducible
for both agents. The emulsion formulation of 17 pro-
vided extensive, uniform, and persistent coating of the
small intestines of both rats and dogs after oral admin-
istration. Transradiation was clearly evident in most of
the animals studied, and mucosal detail was also
evident in most animals. However, in contrast to the
sharp granular appearance of the mucosal surface
typically seen with barium formulations, the mucosal
surfaces in the case of the emulsion were delineated by
a smooth continuous zone of enhancement at the
tangential margins of the lumen. Although the diag-
nostic implications of such a “halo” effect are unknown
at the present, it is hypothesized that these new views
may provide for significant advantages in the early
diagnosis of brush-border abnormalities.
1,3,5-Tr i-n -h exylben zen e. 1-Bromohexane (1 mL) was
added to a stirred mixture of magnesium turnings (5.88 g, 242
mmol) in 25 mL of anhydrous diethyl ether at room temper-
ature under an atmosphere of nitrogen. A crystal of iodine was
added and the mixture was heated to reflux until the color
faded. A solution of the remaining 1-bromohexane (33 mL,
234.8 mmol) in ether (100 mL) was added at a rate sufficient
to maintain gentle reflux. After refluxing for 2 h, the solution
was transferred by cannula to a solution of 1,3,5-trichloroben-
zene (14.2 g, 77.4 mmol) in ether (200 mL) containing 0.32 g
(0.25 mol %) of [1,3-bis(diphenylphosphino)propane]nickel(II)
chloride (NiCl2(dppp)). The solution was gently refluxed for
several hours whereupon a solid began to precipitate from
solution. The reaction mixture was carefully poured into 2000
mL of cold 1 N aqueous HCl and the layers were separated.
The aqueous layer was extracted with additional ether (250
mL) and the ether extracts were combined. The organic layer
was dried (MgSO4), filtered, and evaporated to give 23 g of a
brown oil. The oil was dissolved in hexane and filtered through
a short pad of silica gel, eluting with hexane, and then

1,3,5-Trialkyl-2,4,6-triiodobenzenes
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 10 1947
concentrated to dryness under vacuum to give 19.3 g (58.5
mmol, 75%) of 1,3,5-tri-n-hexylbenzene as a colorless, mobile
oil (C24H42; M+ 330): 1H NMR (CDCl3) δ 0.78 (m, 9H), 1.23
(m, 18H), 1.53 (m, 6H), 2.47 (m, 6H), 6.52 (s, 3H). The product
was used in the subsequent step without additional purifica-
tion.
Ha m ster Th yr oid Toxicity Assa y. Male golden Syrian
hamsters (SASCO Inc., Madison, WI) weighing 90-120 g (6-8
weeks) were used for the oral administration study. The test
articles were formulated as described above, and hamsters
received either the control emulsion or test emulsion by oral
gavage. The treatments were administered at dosage volumes
of 40 mL/kg, which was the highest volume that could be
reasonably administered, representing 4-fold the anticipated
clinical dosage of 10 mL/kg. Animals received either a single
administration of the test emulsion or four administrations
over 8 days. The latter dosing schedule was selected as a
surrogate for increasing drug exposure to the animals in a
realistic manner, since volume restrictions limited the maxi-
mum dose that could be given by oral gavage. Body weights
were recorded prior to dosing and necropsy. After an overnight
fast, hamsters were killed by exsanguination following the
administration of an intraperitoneal overdose of pentobarbi-
tone. The study was thus terminated on 48 h after dosing for
animals receiving a single dose and 1 week after dosing for
animals receiving four doses. Plasma samples were collected,
and all hamsters were subject to a gross necropsy. Thyroid
glands were weighed following fixation. Stomach, thyroid
glands and tissues containing macroscopic abnormalities were
preserved in 10% neutral buffered formalin and examined
histomorphologically using hematoxylin and eosin stained
tissue sections.
1,3,5-Tr i-n -h exyl-2,4,6-tr iiod oben zen e (17). Meth od A.
This procedure illustrates the general method for the prepara-
tion of 13-18, 21 and 22.32 A mixture of 1,3,5-tri-n-hexylben-
zene (81.8 g, 248 mmol), iodine (102.1 g, 402 mmol), and
[bis(trifluoroacetoxy)iodo]benzene (213.3 g, 496 mmol) in 1.5
L of carbon tetrachloride was stirred at room temperature
overnight. The solvent was evaporated, CH2Cl2 (400 mL) and
H2O (400 mL) were added and the layers were separated. The
organic layer was washed with 400 mL of 10% aqueous sodium
thiosulfate and then dried over magnesium sulfate. After
concentration, the mixture of oil and solids was then subjected
to flash chromatography over a silica gel plug (hexanes). The
crude product was dissolved in hexanes (150 mL) and chilled;
the precipitated solid (1,4-diiodobenzene) was removed. This
crystallization procedure was repeated three times followed
by repeated silica gel chromatography (heptane elution, 5 kg
SiO2 total) of the oily residue. After concentrating under
vacuum, the mobile oil was dried with warming under high
vacuum to afford 96.5 g (55%) of 17.
For solids 13-16, 21, and 22: The mixture of product and
1,4-diiodobenzene was dissolved in EtOH,33 and the 1,3,5-
trialkyl-2,4,6-triiodobenzene was precipitated. Recrystalliza-
tion from EtOH afforded the crystalline products.
The above hamster protocol was modified for single dose,
intraperitoneal administration to mimic drug exposure as a
result of a perforated bowel. Compound 24 was administered
over the same dose range as the positive control, particularly
for the thyroid. Male golden Syrian hamsters received a single
intraperitoneal administration of the test article (day 1)
formulated at 114 mg I/mL concentration. Control groups
received 0.9% w/v sodium chloride at a dosage volume of 5.0
mL/kg or the emulsion vehicle (without iodinated agent) at
5.0 mL/kg. Hamsters were subjected to a full necropsy at the
end of the study. Thyroid glands were examined histologically
from all animals. In the absence of an overt effect on the
thyroid, the following tissues were examined histologically
from control hamsters and the high-dosage groups: adrenal
glands, cecum, colon, duodenum, ileum, jejunum, heart, kid-
ney, liver, lung, mesenteric lymph nodes, omentum, pancreas,
rectum, and stomach.
In Vitr o Meta bolism . The in vitro metabolism of 17, 23,
and [14C]24 (positive control) was investigated by HPLC
following incubation with rat (male, Sprague-Dawley; Charles
River (U.K.) Ltd.), hamster (male golden Syrian; Charles River
(U.K.) Ltd., Manston Rd., Margate, Kent), and human liver
microsomes at substrate concentrations of 4 and 40 µM.
Aliquots (0.5 mL) from duplicate incubates were taken at 0,
15, 30 and 60 min, mixed with saturated zinc sulfate solution
(0.5 mL), and extracted with propan-2-ol (0.7 mL). The
resultant supernatants, including appropriate blank and
control samples, were then analyzed by isocratic reversed-
phase HPLC (IBSIL C8, 65% i-PrOH/water with 0.1% TFA).
Recovery of drug in the microsomal extracts varied between
71-82% for the control, 71-89% for 17 and 65-105% for 23.
Results were normalized to 100% of the control values (0 min)
in order to determine the half-life of each drug in the incubate
(n ) 2).
Ca n in e Im a gin g Stu d ies. A group of 12 beagle dogs
(Marshall Farms, North Rose, NY) weighing 12-15 kg were
assigned to this study. A Microfluidizer 110T (5000 psi) was
used to prepare the contrast agent as an emulsion containing
22% 17 with 5% surfactants (0.8% Span 80, 0.2% Tween 80,
4% Pluronic F127), 1% methocel K4M as a viscosity enhancer,
and 0.1% simethicone as an antifoaming agent. The dogs were
radiographed using a Siemens Polyphos 50 at a maximum
interval of once per week. This time period provided adequate
time for the formulation to clear from the GI tract ()24 h)
plus a minimum of 6 days of rest before dosing with the next
formulation. Contrast agents were administered orally using
a gavage tube placed into the distal portion of the esophagus.
For the first few months, the dogs were lightly sedated with
1,3,5-Tr i-n -h exyl-2,4,6-tr iiod oben zen e (17). Meth od B.
This procedure illustrates the general method for the prepara-
tion of 12, 13, 19, 20, and 23. A mixture of the 1,3,5-tri-n-
hexylbenzene (1.4 kg, 3.64 mol) and N-iodosuccinimide (2.86
kg, 12.72 mol, freshly recrystallized from acetone/diethyl ether)
was slurried in 20 L of 1,2-dichloroethane. Trifluoromethane-
sulfonic acid (0.55 kg, 3.64 mol) was added at a moderate rate
and the reaction was heated under reflux for 2 h. Additional
N-iodosuccinimide (0.41 kg, 1.82 mol) was added and reflux
was maintained for 60 min. The solvent was evaporated and
the residue was dissolved in 4 L hexanes and washed with 4
L water. The organic layer was separated, and then flash
chromatographed over a silica gel plug (2 L of hexanes). The
organics were washed with 6 L of 5% aqueous sodium
thiosulfate. After concentrating under vacuum, the mobile oil
was dried with warming (100 °C) under high vacuum to afford
2.45 kg (96%) of 17: 1H NMR (CDCl3) δ 0.89 (m, 9H), 1.35 (m,
12H), 1.46 (m, 12H), 3.28 (m, 6H).
Biologica l Stu d ies. The animal care, use of tissue, and in
vivo experimentation conformed to the NIH Guide for the Care
and Use of Laboratory Animals (NIH Publication No. 86-23,
1985) and the Animal Welfare Act (P.L. 89-544, as amended).
All research involving animals described in this publication
was performed in accord with the Sterling Winthrop Pharma-
ceuticals Research Division’s (SWPRD) Policy on Animal Use
and all national and federal legislation. All SWPRD animal
facilities and programs are accredited by the American As-
sociation for Accreditation of Laboratory Animal Care Inter-
national (AAALAC International).
Acu te Mu r in e Toxicity Assa y. Male ICR mice (Harlan
Sprague-Dawley) weighing 20-25 g received a single admin-
istration of the test articles via oral gavage. The contrast
agents were formulated as oil-in-water emulsions at a con-
centration of 114 mg I/mL. Emulsions were prepared im-
mediately prior to dosing by adding the appropriate % w/v of
test article to a mixture of 12.5% w/v light mineral oil, 3.365%
w/v Tween 80, 1.635% w/v Span 80, 0.5% w/v Avicel RC591
and water to 100 vol %. A control emulsion was prepared in
the same manner by omitting the test article and adjusting
the volume of water required. The mice were observed for
clinical signs of toxicity and mortality several times on the
day of dosing and once a day thereafter for a total of 7 days.
Changes in body weight were assessed 24 h and 1 week after
treatment. Seven days after dosing all mice were subjected to
a gross necropsy of the abdominal and thoracic cavities.

1948 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 10
Estep et al.
(16) D. Rubin, private communication.
acepromazine and placed in a ventral dorsal (VD) position on
the radiographic table for a series of radiographs taken at 15,
30, 45, 60, 120, and 240 min. Air (100 mL) was administered
immediately after the 15-min radiograph so as to distend the
lumen and provide a double contrast examination for the
subsequent radiographs. No movement or manipulation was
performed on the dogs during the experiment. The dogs quickly
became acclimated to this procedure such that the acepro-
mazine sedation was unnecessary after 4-6 weeks.
(17) Wallingford, V. H. The Development of Organic Iodine Com-
pounds as X-ray Contrast Media. J . Am. Pharmacol. Assoc. 1953,
42, 721-728.
(18) (a) Eapen, K. C.; Dua, S. S.; Tamborski, C. Regiospecific
Synthesis of Aromatic Compounds via Organometallic Interme-
diates. J . Org. Chem. 1984, 49, 478-482. (b) K. J osef, unpub-
lished results.
(19) Merkushev, E. B. Review of aromatic iodinations: Advances in
the Synthesis of Iodoaromatic Compounds. Synthesis 1988, 923-
937.
(20) (a) Merkushev, E. B.; Sedov, A. M.; Simakhina, N. D. A Simple
Method for the Synthesis of Iodomethylbenzenes. J . Org. Chem.
USSR 1978, 14, 1040. (b) Varma, P. S.; Subrahmanian, T. S.
Halogenation. Part XVI. Bromination and Iodination of Mesi-
tylene J . Indian Chem. Soc. 1936, 13, 192-193; Chem. Abstr.
1936, 10, 59496.
(21) Bovonsombat, P.; McNelis, E.; Ring Halogenations of Polyalkyl-
benzenes with N-Halosuccinimide and Acidic Catalysts. Syn-
thesis 1993, 237-241.
(22) Ishikawa, N.; Sekiya, A. The Iodination of Benzene with Iodine
in the Thallium(III) Trifluoroacetate-Trifluoroacetic Acid Sys-
tem. Bull. Chem. Soc. J pn. 1974, 47, 1680-1682.
(23) Kajigaeshi, S.; Kakinami, T.; Moriwaki, M.; Tanaka, T.; Fujisaki,
S.; Okamato, T. Halogenation Using Quaternary Ammonium
Polyhalides. XIV. Aromatic Bromination and Iodination of
Arenes by Use of Benzyltrimethylammonium Polyhalides-Zinc
Chloride System. Bull. Chem. Soc. J pn. 1989, 62, 439-443.
(24) Barluenga, J .; Gonza´lez, J . M.; Garc´ıa-Mart´ın, M. A.; Campos,
P. J . Polyiodination on Benzene at Room-Temperature A Regio-
selective Synthesis of Derivatives. Tetrahedron Lett. 1993, 34,
3893-3896.
(25) Shono, T.; Matsumura, Y.; Katoh, S.; Ikeda, K.; Kamada, T.
Aromatic Iodination by Positive Iodine Active Species Generated
by Anodic Oxidation in Trimethyl Orthoformate. Tetrahedron
Lett. 1989, 30, 1649-1650.
(26) Merkushev, E. B.; Simakhina, N. D.; Koveshnikova, G. M. A
New, Convenient Iodination Method of Aromatic Compounds.
Synthesis 1980, 486-87.
(27) NIS/TfOH has also been used to couple carbohydrates: Kon-
radsson, P.; Mootoo, D. R.; McDevitt, R. E.; Fraser-Reid, B.
Iodonium Ion Generated In Situ from N-Iodosuccinimide and
Trifluoromethanesulphonic Acid Promotes Direct Linkage of
′Disarmed′ Pent-4-enyl Glycosides. J . Chem. Soc., Chem. Com-
mun. 1990, 270-272.
(28) Purification of 9 from this mixture by fractional distillation
proved to be quite challenging on small scale. Attempts to
facilitate the purification by chemical reduction of 11 to more
volatile 1,3-di-n-hexylbenzene were not successful.
(29) Newbery, G.; Libman, D. D. Secondary Octyl Ether of Tri-iodo
Phenol as an X-ray Contrast Media. U.S. Patent 2,622,100, Dec
16, 1952.
Ack n ow led gm en t. The authors thank J ulio Perez,
Manohar Saindane, and Denise Schwartz of our Chemi-
cal Development Department for their help in the large-
scale synthesis of some analogues; Nancy Fetrow, Mary
Ann Harris, Bonnie Simpkins, and Bill Hoffman for
technical assistance in the acute murine toxicity assays;
and Gregg Stetsko and Steve Ruddy for helpful discus-
sions.
Refer en ces
(1) This research was conducted at Sterling Winthrop Pharmaceu-
ticals Research Division, Collegeville, PA, and Alnwick, Northum-
berland, U.K., prior to its divestiture in October 1994.
(2) Department of Medicinal Chemistry.
(3) Department of Pharmaceutical Sciences.
(4) Department of Drug Safety Assessment.
(5) Department of Laboratory Animal Resources.
(6) Department of Drug Metabolism.
(7) Swanson, D. P.; Halpert, R. D. Gastrointestinal Contrast
Media: Barium Sulfate and Water-Soluble Iodinated Agents. In
Pharmaceuticals in Medical Imaging. Radiopaque Contrast
Media, Radiopharmaceuticals, Enhancement Agents for Mag-
netic Resonance Imaging and Ultrasound; Swanson, D. P.,
Chilton, H. M., Thrall, J . H., Eds.; Macmillan Publishing Co.:
New York, NY, 1990; pp 155-183.
(8) (a) Gelfand, D. W.; Ott, D. J .; Tritco, R. Causes of Error in
Gastrointestinal Radiology. I. Upper Gastrointestinal Examina-
tion. Gastrointest. Radiol. 1980, 5, 91-97. (b) Ott, D. J .; Gelfand,
D. W.; Ramquist, N. A. Causes of Error in Gastrointestinal
Radiology. II. Barium Enema Examination. Gastrointest. Radiol.
1980, 5, 98-105.
(9) Skucas, J . Barium Sulfate: Toxicity and Complications. In
Radiographic Contrast Agents, 2nd ed.; Skucas, J ., Ed.; Aspen
Publishers: Rockville, MD, 1989; pp 62-82.
(10) Hoey, G. B.; Weigart, P. E.; Rands, J r., R. D. Organic Iodine
Compounds as X-ray Contrast Media. In International Encyclo-
pedia of Pharmacology and Therapeutics; Knoefel, P. K., Ed.;
Pergamon Press: Oxford, England, 1971; Vol. 76, pp 23-131.
(11) Margulis, A. R.; Thoeni, R. F. Water-Soluble Agents. In Radio-
graphic Contrast Agents, 2nd ed.; Skucas, J ., Ed.; Aspen
Publishers: Rockville, MD, 1989; pp 91-108.
(12) Teplich, J . G.; Haskin, M. E.; Skelley, J .; Wohl, G. T.; Sanen, F.
Experimental Studies with a New Radiopaque Emulsion. Ra-
diology 1964, 82, 478-485.
(30) Illig, C. R.; Cooper, E. R.; Toner, J . L.; Upson, D. A.; Douty, B.
D.; Caulfield, T. J .; Bacon, E. R.; Estep, K. G.; J osef, K. A.;
Robinson, S.; Spara, P. P. Compositions of Iodophenoxy Alkanes
and Iodophenyl Ethers for Visualization of the Gastrointestinal
Tract. U.S. Patent 5,316,755, May 31, 1994.
(31) Gelfand, D. W.; Ott, D. J . Colon Cancer Screening: Where
Radiology Fits In. Diagnostic Imaging 1993, 75-77.
(32) J osef, K. A.; Bacon, E. R.; Estep, K. G.; Illig, C. R.; Douty, B. D.
Compositions of Alkylbenzenes for Visualization of the Gastro-
intestinal Tract Using X-ray Contrast. U.S. Patent 5,308,607,
May 3, 1994.
(13) Steinhausen, T. B.; Dungan, C. E.; Furst, J . B.; Plati, J . T.;
Smith, S. W.; Darling, A. P.; Wolcott, J r., E. C.; Warren, S. L.;
Strain, W. H. Iodinated Organic Compounds as Contrast Media
for Radiographic Diagnoses. Radiology 1944, 43, 230-235.
(14) Pirkey, E. L.; Reed, E. S.; Smith, W. H. Clinical Studies of the
Use of Emulsion of Ethyl Iodophenylundecylate (Pantopaque).
Radiology 1950, 55, 89-93.
(33) Cyclohexane was used as the recrystallization solvent for 13 (see
Table 1).
(15) Rubin, D. L. Iodinated Aliphatic Emulsions for X-ray Contrast.
U.S. Patent 5,258,172, Nov 2, 1993.
J M990407I