Method and compositions for orally administered contrast agents for MR imaging

Abstract

Disclosed is CT or MR contrast agent which comprises a base or carrier scaffold formed of a polyhydroxol compound having a linker to which a Gd-DOTA is covalently bonded. Also disclosed is a method of screening a patient for colon cancer using a CT or MR contrast, which method comprises administering to a patient undergoing screening a compound as above described.

Claims

1. A method of screening a patient for colon cancer using a computed tomographic (CT) or magnetic resonance (MR) colonography, said method comprising: dissolving a water soluble CT or MR contrast agent in an aqueous solution to produce an aqueous solution of colonography contrast agent; administering to a patient said aqueous solution of colonography contrast agent; and screening the patient for colon cancer using a computed tomographic colonography (CT-C) or a magnetic resonance colonography (MR-C), wherein the water soluble CT or MR contrast agent comprises a base or carrier scaffold formed of a polyhydroxol compound having a plurality of polyhydroxol functional groups, wherein each of said polyhydroxol functional group is linked to a moiety of the formula: ##STR00001## and wherein R comprises straight chain polyethylene glycol linker having on its terminal end a Gd-chelated DOTA moiety of the formula: ##STR00002##

2. The method of claim 1, wherein the polyhydroxol compound comprises a disaccharide.

3. The method of claim 2, wherein the polyhydroxol compound comprises sucrose.

4. The method of claim 2, wherein the average number of Gd-DOTA chelates bonded to the sucrose is 8.

5. The method of claim 1, wherein the linker incorporates one or more straight chain hydrocarbon segments.

6. The method of claim 1, wherein the linker incorporates one or more branch chain hydrocarbon segments.

7. The method of claim 1, wherein colorectal cancer (CRC) cells are screened for.

8. The method of claim 1, wherein the molar relaxivity (R.sub.1) of said water soluble CT or MR contrast agent is 213±1.7 mM s.sup.−1.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) FIG. 1 illustrates a reaction scheme for synthesizing a contrast agent in accordance with one embodiment of the disclosure;

(2) FIG. 2 is a mass spectrometric analysis of the contrast of the agent synthesized in accordance with FIG. 1;

(3) FIG. 3A are shaded T.sub.1 maps demonstrating variance in T.sub.1 times between phantoms of various concentrations, and FIG. 3B is a graph showing weighted linear regression of experimental phantom data (R.sub.1 as a function of concentration);

(4) FIGS. 4A-4E are maximum intensity projections and relative contrast images demonstrating passage of the contrast agent of the present disclosure; and FIG. 4F plots change in tumor signal intensity compared to baseline values before contrast;

(5) FIG. 5 is a table of elemental analysis of Gadolinium levels in the colon and the urine;

(6) FIG. 6A shows T.sub.1 weighted Spin Echo Multi Slice images at various times post IV injection of contrast agent of the present disclosure; and FIGS. 6B and 6C are graphs plotting quantitative data showing signal intensity with time in the tumor (FIG. 6B) and the kidneys (FIG. 6C).

DETAILED DESCRIPTION OF THE INVENTION

(7) Contrast Agent

(8) The synthesis of the sucrose-derived contrast agent 3 is depicted in FIG. 1. To an argon-degassed solution of azide 1 (16) (1.85 g, 2.43 mmol) and octaalkyne 2 (18) (271 mg, 0.28 mmol) in 9/1 THF/water (50 mL) were added CuSO.sub.4 (80 mg, 0.5 mmol) and sodium ascorbate (160 mg, 0.8 mmol). The mixture was stirred under argon at room temperature for 24 h. Volatiles were evaporated and the residue was diluted with water (50 mL), the solution was washed with a mixture of dithiazone in chloroform (20 mg/L, 5×100 mL), and lyophilized to give a residue, which was loaded onto a reversed phase (C-18) silica gel column. The column was eluted with acetonitrile/water (80/20) to yield the product 3 as a non-hygroscopic cream-colored solid (1.81 g, 0.25 mmol, 89%). MALDI-TOF MS was performed as shown in FIG. 2, and calculated for C.sub.252H.sub.414N.sub.64O.sub.91Gd.sub.8 was 7056.366 [M+H].sup.+, while we observed at 7056.328, and at 3528[M+2H].sup.2+ confirming the presence of 8 Gd-DOTA chelates per sucrose scaffold.

(9) Phantom Evaluations

(10) Initial Mill characterization of the Gd-DOTA-Sucrose compound was accomplished in phantoms of varying concentrations according to previously described protocol (16). Similarly, these phantoms were studied using progressive saturation experiments (PS), with 11 TR values exponentially spaced from 30 s to 60 ms. Nonlinear least squares regression was employed to determine the relaxation time- and relaxation rate constants, T.sub.1 and R.sub.1=1/T.sub.1, respectively. Weighted linear regression was used to determine the R.sub.1 value as a function of concentration, and the relationship between R.sub.1 and [Gd-DOTA-sucrose CA] was estimated for the average multimer. The weights used in the fit were inversely proportional to the variance in R.sub.1 for each concentration, (1/s.sup.2.sub.R1,i).

EXAMPLES

(11) In vivo Experiments

(12) A human colorectal cell line engineered to express luciferase (HCT-116/Luc) was used to generate orthotopic CRC xenograft tumors. Eight female SCID/beige mice each received injections into the lining of the rectum, 4-5 mm beyond the anal verge, of 1×10.sup.6 HCT 116/luc cells in a 10 μl volume (mixed in a solution of X uL PBS). Mice were allowed to recuperate for two weeks, while being weighed daily, and then monitored for tumor growth by bioluminescence imaging in the IV IS 200 (Perkin Elmer) 2 times per week. After approximately 2 weeks, tumor growth also was monitored via T.sub.2-weighted MRI (spin echo multi slice sequences with TE/TR=72/1000 ms, slice thickness of 0.785 mm and resolution of 350×350×100 um over 6 minutes). Upon tumor detection (˜3 weeks after injection), contrast enhanced imaging experiments were initiated.

(13) Compound 3 was dissolved in 100 mM sodium phosphate buffer, pH 7.4, at a concentration of 2.5 mM for gavage administration. Gavage was given at 10 μl per gram of body weight. Imaging began within 30 minutes of contrast agent administration. For i.v. contrast administration, compound 3 was dissolved in PBS at a concentration of 25 μmol/kg of body weight, and administered in 150 μl volume via tail vein catheter during the imaging session, with image acquisition occurring before, during and after injection.

(14) Imaging Protocol & Analysis

(15) MRI was performed at baseline and at 30 minutes and at 5, 24 and 48 hours following gavage administration. All imaging data were acquired using a 7-T horizontal magnet (ASR 310, Agilent Technologies) equipped with nested 305/120/HDS gradient set. Prior to acquisitions, animals were placed in an induction chamber and anesthetized using 2% isoflurane. The mice were then restrained in a specific holder and inserted into the magnet and 35-mm quadrature coil (Doty Scientific) with a constant supply of isoflurane and heated air gas in order to maintain a temperature of 37±1° C. Body temperature and respiratory functions were monitored using the SAII system (Small Animals Instrument Inc, Stony Brook, N.Y.) and temperature control of the imaging gradients was achieved by means of a water chiller (Neslab Waters) and maintained at 12° C. for all acquisitions. Applying a coronal slice orientation, 3-dimensional T.sub.1-weighted spin-echo (SE3D) sequences were acquired with TE/TR=10.6/31.6 ms, field of view=90×45×16 mm, matrix=256×128×16 and four averages in 4.3 min. Images were acquired prior to contrast and at 30 min, 5, 25 and 48 hours post gavage.

(16) For the dynamic scans, contrast was administered via tail vein catheter during imaging with two dimensional spin-echo multislice (SEMS) with T.sub.1-weighting using TE/TR=10/233 ms, field of view=90×45 mm, matrix=256×128 and 9 slices with 1 mm thickness. The total scan time equaled 90 min (5 min prior to contrast and 85 min post). For both scans, spatial resolution was hence 351×351×1000 μm.

(17) Signal intensity (SI) in the tumors and kidneys was calculated using manually drawn regions of interest (ROIs) in VnmrJ (Agilent Technologies, Inc.) and normalized to surrounding tissue intensity. Each mouse was used as its own control and the percent increase in SI was calculated individually and then averaged.

(18) Statistical analysis was performed using Graphpad Prism software (Graphpad, San Diego, Calif., USA) and one-way analysis of variance (ANOVA) followed by the Tukey test for comparison of mean values. A confidence interval of 95% was chosen and thus statistical significance was pre-determined at p<0.05.

(19) Elemental Analysis

(20) For elemental analysis studies, four female SCID mice weighing between 24 and 26 grams each were administered 200 μl of compound 3 in 0.1 M sodium phosphate buffer via oral gavage. This delivers 3.52 mg of 3 (GdSucrose) into each mouse. Mice were then immediately transferred individually to metabolic caging, designed to separately capture urine and feces. After 72 hours, mice were euthanized and colons were removed, lightly rinsed with PBS and collected in individual tubes. Urine and feces were collected separately. All samples were lyophilized, and subsequently sent to Elemental Analysis, Inc., (Lexington, Ky.) for detection of gadolinium via instrumental neutron activation analysis (INAA).

(21) Histology

(22) Following MRI, animals were euthanized and tumor sections fixed, mounted in paraffin, cut into sections, mounted on slides and stained with Hematoxylin & Eosin (H & E) for histology examination.

(23) Results

(24) The calculated mass for compound 3 (C.sub.252H.sub.414N.sub.64O.sub.91Gd.sub.8) is 7056.366 [M+H].sub.+. By MALDI-TOF MS (FIG. 2), the observed mass was 7056.328, and the average number of Gd-DOTA chelates per sucrose was determined to be 8. During the phantom portion of this study, it was noted that the solubility of this current compound 3 differed substantially from the CA previously presented by Martinez et al (19). Specifically, we observed that while 3 dissolved rather easily in water or aqueous buffer, the previous compound was not soluble in water and required 15% ethanol to achieve suspension.

(25) MR relaxivity measurements clearly demonstrated differences in T.sub.1-shortening and thus contrast enhancement in the various phantoms (i.e. concentrations) could readily be observed in shaded T.sub.1 maps (FIG. 3A). The shaded T.sub.1 maps demonstrate the variance in T.sub.1-times between phantoms of various concentrations. Phantom concentrations range as following [μM]; a=2500; b=1250; c=625; d=313; e-156; f=78; g=39; h=20 and w=0 (i.e. pure water). Note that a, b and c had to be excluded due to oversaturation effects. Due to shortening of spin-spin (T.sub.2) relaxation, three of the highest concentration phantoms were excluded for quantitative analysis of relaxivity. To determine the molar relaxivity (r.sub.1) of this improved agent, 3, a weighted linear regression model was applied to the phantom data (FIG. 3B). Impressively, the fitted value for molar relaxivity (R.sub.1) was 213±1.7 mM s.sup.−1 with R.sup.2=0.983, which is an 8-fold increase compared to our previous compound (16). Relaxation experiments were accomplished with multiple TR spin echo experiments on two separate Eppendorf tube phantoms prepared with identical concentrations of Gd-DOTA-sucrose CA. Error bars denote standard deviations.

(26) In vivo, the passage of the CA and potential tumor uptake in CRC xenografts were monitored by a series of two- and three-dimensional T.sub.1-weighted sequences (see protocol). Herein clearly visualized by maximum intensity projections (MIP), the agent showed rather fast movement throughout the mouse GI-tract (FIG. 4A). Maximum Intensity Projections (MIP) generated from T.sub.1-weighted Spin Echo sequences (3D) demonstrate the passage of CA throughout the gastrointestinal tract of a representative animal at baseline (BL) and 30 minutes at 5, 25 and 48 hours post gavage. In fact, as early as 30 min following gavage administration, the agent had moved from the stomach to the upper intestinal areas. At the 5-hour time point it appeared that the agent had partially cleared the animal with notable enhancement remaining in the intestines. By 24 hours, only small amounts of compound could be detected in the intestines and this contrast appeared to be associated with feces. By 48 hours post CA, we could no longer detect any enhancement.

(27) More importantly, and although non-targeted at this point, the contrast agent of the present disclosure induced notable signal enhancement in the tumors at 5 hours post-gavage. Zoomed in on the tumor solely, 3D Spin Echo images of pre-contrast FIG. 4B, and 5 hours post-gavage FIG. 4C, show the relative contrast enhancement. This enhancement was particularly noticeable in the tumor rim, and revealed a potentially necrotic core. To investigate this darker inner core further, H & E stained tumor sections were evaluated, and did indeed demonstrate substantial cellular necrosis in the tumor center versus the outer rim (FIGS. 4D & 4E). Additionally, enhancement was observed in the bladder at the 5 hour time-point, which was co-incident with tumor enhancement. Since we had previously not observed uptake of the original compound in the bladder (19), and the original study did not include tumors, while not wishing to be bound by theory, we believe that this could be due to systemic uptake through the orthotopic xenograft tumor.

(28) Quantitatively, tumor signal intensity (SI) was determined for each animal using manually drawn ROIs and compared to pre-gavage values. Confirming the visual observations, a significant increase of 40% in tumor SI was noted by 5 h post gavage and persisted throughout the 24-hour time point. See FIG. 4F which shows percent change in tumor signal intensity compared to baseline values before contrast. By 48 hours post CA administration, the enhancement had dropped to 20% above baseline values and the CA appeared to largely have cleared the animal. This observation was further confirmed by elemental analysis, which showed that the Gadolinium levels in the colon and urine were below the limits of detection, while a notable portion was found in the feces (see FIG. 5). Since the elemental analysis study was conducted using animals that did not bear tumors, and Gd was not detected in the colon tissue or urine, this observation further supports the idea that the bladder uptake in the MM study was due to systemic uptake via the orthotopic tumor and not uptake through the unaffected GI tract.

(29) To evaluate the properties of this agent following intravenous administration and determine kidney clearance rates, mice (N=4) received contrast (compound 3) by an i.v. tail vein catheter while inside the magnet. As demonstrated in FIG. 6A which shows T.sub.1-weighted Spin Echo Multi Slice (SEMS) showing baseline (BL), and at 2 minutes, 10 minutes, 30 minutes, and 1 hour post IV injection of contrast agent, dramatic signal enhancement within the tumor could be detected immediately following the i.v. injection (lower arrow 10) and the kidneys showed maximum uptake at this point (upper arrow). These observations were in agreement with quantitative analysis showing an increase of signal intensity of tumor contrast by 72% within the first 5 minutes following injection (FIG. 6B). By 15 minutes post CA, the increase in signal intensity reached 93%, following which, the signal intensity appeared to decrease exponentially. By 90 min, the signal intensity had dropped back to 16% above baseline values and the imaging session was terminated. Kidney clearance of the CA was also quantified and showed an increase in signal intensity of 185% by 3 min post injection followed by a fast wash-out of the agent (FIG. 6C).

(30) Neutron activation analysis specifically detected gadolinium in colon tissue, urine and feces collected over the first 72 hours post gavage. Detection of gadolinium in urine and colon tissue was under 5 ppm or less in all samples, which is the limit of detection of the analysis. In feces, the samples averaged 126.5 ppm. This indicates that most, if not all, of the Gd-DOTA-Sucrose probe passed through the GI tract and was excreted in the feces. These non-tumor bearing mice did not have measurable excretion through the urine or uptake in colon tissue, in contrast to what was seen in tumor-bearing mice during imaging sessions.

(31) The present disclosure provides the following features and advantages: Our first generation Gd-DOTA-sucrose contrast agents had superior relaxometric properties compared to Gd-DOTA in its lower limit of detectability and remained in the GI tract throughout the passage. In this study, we improved our first generation Gd-DOTA-sucrose contrast agents by increasing the average number of Gd-DOTA chelates per sucrose to 8. To improve relaxivity, the chelates were shortened to induce stiffness thus increasing the rotational correlation time. These modifications yielded an 8-fold increase in Spin-Lattice-relaxivity which also was observed in vivo (212 from 29). The new molecule has shorter arms and limits motion and rapidly clears the kidney. In vivo, tumor signal intensity enhanced 40% (gavage) and 93% (tail-vein injection). Kidney uptake was maximized by 3 min post i.v. injection and cleared by 10 min (significantly faster than reported Magnevist renal clearance). Rapid kidney uptake could be improved by co-injection with diuretics, etc., which could enhance circulation time and increase tumor uptake. Hence, allowing for detection of smaller or flat lesions with greater sensitivity. Tumor bearing mice experienced uptake via gavage, indicated by bladder enhancement Non tumor bearing mice for INAA had no measurable Gd in urine, indicating no uptake via gavage While not wishing to be bound by theory, it is believed that uptake is by tumor-associated vasculature or other tumor related physiology While enhancement was observed in the bladder in the MM studies with the current compound, but was not observed in the previous study (16). And since Gd was not detected in the colon or urine by elemental analysis in the absence of a tumor. Again, while not wishing to be bound by theory, we hypothesize that the tumor uptake allowed systemic clearance and, hence, bladder contrast enhancement. But this would not be observed in normal, unaffected GI tracts that do not bear tumors. However, the tumor contrast enhancement by both the oral and intravenous routes of administration suggests that the current untargeted contrast agent could provide significant benefit when combined with standard MRI based virtual colonoscopy. The rapid systemic clearance of the agent through the renal system implies that intravenous administration has potential. Additionally, since this novel contrast agent has greater relaxivity than standard contrast, and rapid systemic clearance, also may demonstrate greater utility for contrast-enhanced MRI for a wide range of cancers in addition to colorectal cancers. It is also believed that the larger size, mass, of the current Gd-sucrose compound (3) could restrict extravasation in normal tissues, while allowing extravasation in tumors with more leaky vasculature, allowing for improved contrast with surrounding normal tissues.

(32) Various changes may be made in the above disclosure without departing from the spirit and the scope of the disclosure. As another modification, a Gadolinium-DOTA sucrose-derived agent was synthesized using an argon-degassed solution of azide, octaalkyne in 9/1 THF/water, CuSO.sub.4 and sodium ascorbate. The compound was improved by increasing the number of Gadolinium-DOTA chelates per sucrose to eight and shortening the chelates to induce stiffness and increase the rotational correlation time, inducing an eight-fold increase in Spin-Lattice-relaxivity. For in vivo assessment, SCID mice were intra-rectally injected with human CRC cells. Then, the mice received the Gadolinium-DOTA sucrose-derived agent (dissolved in sodium phosphate buffer at a concentration of 2.5 mM) by oral gavage, which increased tumor visualization by 40% above MRI with no contrast agent. Only mice with tumors (n=4) showed traces of the Gd-DOTA sucrose molecule in the bladder, compared to control mice (n=3) with no signal.

(33) It is also to be understood that the following claims are intended to cover all of the generic and specific features of the disclosure herein.

REFERENCES

(34) 1. R. Siegel, C. DeSantis, K. Virgo, K. Stein, A. Mariotto, T. Smith, D. Cooper, T. Gansler, C. Lerro, S. Fedewa, C. Lin, C. Leach, R. S. Cannady, H. Cho, S. Scoppa, M. Hachey, R. Kirch, A. Jemal, E. Ward, Cancer treatment and survivorship statistics, 2012. CA Cancer J Clin 62, 220-241 (2012); published online Epub2012 July-August (10.3322/caac.21149). 2. O. Ben-Ishay, E. Brauner, Z. Peled, A. Othman, B. Person, Y. Kluger, Diagnosis of colon cancer differs in younger versus older patients despite similar complaints. Isr Med Assoc J 15, 284-287 (2013); published online EpubJun ( 3. A. C. Society, “Cancer Facts & Figures, 2012,” (American Cancer Society, 2012). 4. J. M. Walsh, J. P. Terdiman, Colorectal cancer screening: scientific review. JAMA 289, 1288-1296 (2003); published online EpubMar ( 5. D. K. Rex, C. S. Cutler, G. T. Lemmel, E. Y. Rahmani, D. W. Clark, D. J. Helper, G. A. Lehman, D. G. Mark, Colonoscopic miss rates of adenomas determined by back-to-back colonoscopies. Gastroenterology 112, 24-28 (1997); published online EpubJan ( 6. S. B. Ahn, D. S. Han, J. H. Bae, T. J. Byun, J. P. Kim, C. S. Eun, The Miss Rate for Colorectal Adenoma Determined by Quality-Adjusted, Back-to-Back Colonoscopies. Gut Liver 6, 64-70 (2012); published online EpubJan (10.5009/gn1.2012.6.1.64). 7. W. Luboldt, P. Bauerfeind, S. Wildermuth, B. Marincek, M. Fried, J. F. Debatin, Colonic masses: detection with MR colonography. Radiology 216, 383-388 (2000); published online EpubAug ( 8. J. Yee, CT screening for colorectal cancer. Radiographics 22, 1525-1531 (2002); published online Epub2002 November-December ( 9. P. J. Pickhardt, J. R. Choi, I. Hwang, J. A. Butler, M. L. Puckett, H. A. Hildebrandt, R. K. Wong, P. A. Nugent, P. A. Mysliwiec, W. R. Schindler, Computed tomographic virtual colonoscopy to screen for colorectal neoplasia in asymptomatic adults. N Engl J Med 349, 2191-2200 (2003); published online EpubDec (10.1056/NEJMoa031618). 10. B. Levin, D. A. Lieberman, B. McFarland, K. S. Andrews, D. Brooks, J. Bond, C. Dash, F. M. Giardiello, S. Glick, D. Johnson, C. D. Johnson, T. R. Levin, P. J. Pickhardt, D. K. Rex, R. A. Smith, A. Thorson, S. J. Winawer, A. C. S. C. C. A. Group, U. M.-S. T. Force, A. C. o. R. C. C. Committee, Screening and surveillance for the early detection of colorectal cancer and adenomatous polyps, 2008: a joint guideline from the American Cancer Society, the US Multi-Society Task Force on Colorectal Cancer, and the American College of Radiology. Gastroenterology 134, 1570-1595 (2008); published online EpubMay (10.1053/j.gastro.2008.02.002). 11. B. Saar, A. Meining, A. Beer, M. Settles, H. Helmberger, E. Frimberger, E. J. Rummeny, T. Misch, Prospective study on bright lumen magnetic resonance colonography in comparison with conventional colonoscopy. Br J Radiol 80, 235-241 (2007); published online EpubApr (10.1259/bjr/83959666). 12. G. Pappalardo, E. Polettini, F. M. Frattaroli, E. Casciani, C. D'Orta, M. D'Amato, G. F. Gualdi, Magnetic resonance colonography versus conventional colonoscopy for the detection of colonic endoluminal lesions. Gastroenterology 119, 300-304 (2000); published online EpubAug ( 13. D. Hartmann, B. Bassler, D. Schilling, H. E. Adamek, R. Jakobs, B. Pfeifer, A. Eickhoff, C. Zindel, J. F. Riemann, G. Layer, Colorectal polyps: detection with dark-lumen MR colonography versus conventional colonoscopy. Radiology 238, 143-149 (2006); published online EpubJan (10.1148/radio1.2381041756). 14. W. Ajaj, G. Pelster, U. Treichel, F. M. Vogt, J. F. Debatin, S. G. Ruehm, T. C. Lauenstein, Dark lumen magnetic resonance colonography: comparison with conventional colonoscopy for the detection of colorectal pathology. Gut 52, 1738-1743 (2003); published online EpubDec 15. S. C. Goehde, E. Descher, A. Boekstegers, T. Lauenstein, C. Kühle, S. G. Ruehm, W. Ajaj, Dark lumen MR colonography based on fecal tagging for detection of colorectal masses: accuracy and patient acceptance. Abdom Imaging 30, 576-583 (2005); published online Epub2005 September-October (10.1007/s00261-004-0290-4). 16. G. V. Martinez, S. Navath, K. Sewda, V. Rao, P. Foroutan, R. Alleti, V. E. Moberg, A. M. Ahad, D. Coppola, M. C. Lloyd, R. J. Gillies, D. L. Morse, E. A. Mash, Demonstration of a sucrose-derived contrast agent for magnetic resonance imaging of the GI tract. BioorgMed Chem Lett 23, 2061-2064 (2013); published online EpubApr (10.1016/j.bmc1.2013.02.008). 17. P. Caravan, Protein-targeted gadolinium-based magnetic resonance imaging (MRI) contrast agents: design and mechanism of action. Acc Chem Res 42, 851-862 (2009); published online EpubJul (10.1021/ar800220p [pii] 10.1021/ar800220p). 18. V. Rao, R. Alleti, L. Xu, N. K. Tafreshi, D. L. Morse, R. J. Gillies, E. A. Mash, A sucrose-derived scaffold for multimerization of bioactive peptides. BioorgMed Chem 19, 6474-6482 (2011); published online EpubNov (10.1016/j.bmc.2011.08.053). 19. Martinez G V, Navath S, Sewda K, Rao V, Foroutan P, Alleti R, Moberg V E, Ahad A M, Coppola D, Lloyd M C, Gillies R J, Morse D L, Mash E A, “Demonstration of a sucrose-derived contrast agent for magnetic resonance imaging of the GI tract”, Bioorganic & Medicinal Chemistry Letters 23 (2013) 2061-2064.