Download Simultaneous fluorescence tomographic imaging of

Simultaneous fluorescence tomographic imaging of efficacy and
toxicity following acute 5-FU treatment of HT-29 tumor xenografts
Kristine O. Vasquez and Jeffrey D. Peterson. Life Sciences and Technology, PerkinElmer Inc., Hopkinton MA
Non-Invasive IVIS Imaging & Quantification
Liver Region
Tumor Radiance
Fold AMT750 Increase
B. 5-FU Treatment: FL Tomography
Tumor Volume
AMT750
AS680
Control
AMT 750 Cocktail
50 mg/kg
100 mg/kg
Increase:
Heart
Liver
Tumor
#p
< 0.05
*p < 0.01
**p < 0.001
Quantification of A. bioluminescence and B. fluorescence imaging datasets from Figure 3 reveal that even a single
administration of 5-FU can cause statistically significant biomarker changes in tumor, heart, liver, lungs, stomach, kidneys,
and intestines when imaging agents are dosed 2h post-5-FU. These biomarker changes occurred in the absence of overt
changes in tumor size. C. As expected, these acute tissue effects seen by imaging were transient and could not be seen if
imaging agents were administered 24h after 5-FU (t-test, n = 5, bar: s.e.m.).
5 Biological Changes in the Absence of Histologic Changes
Tumor
Heart
Kidney
Lungs
Liver
Tumor Bioluminescence
3.0
Lungs
2.0
Heart
Liver
Kidneys
Radiance
Brain
Increase:
Stomach
Tumor
Intestines
Increase:
Kidney
Stomach
C. 2-Color Tomographic ROI Slices
Whole Animal Images
Control
100 mg/kg
Control
5-FU
1.0
Bladder
C. Reversibility of 2h 5-FU Effects on Tissues 24h After Treatment
5-FU
x109
Intestines
**p < 0.001
% of Untreated Control
A. 5-FU Treatment: Tumor BLI
4.0
~ 3 weeks
*p < 0.01
3 Noninvasive Imaging of Anti-Tumor Efficacy and
Biological Effects on Normal Tissue Simultaneously
Biodistribution Analysis Model
3D FL
Localization &
Quantification
< 0.05
Normal BALB/c female mice were injected IP with a single 300 mg/kg dose of rifampicin (RMP) or thioacetic acid (TAA),
known to induce toxicity in various tissues. The AMT 750 cocktail was administered 24h after drug dosing and mice were
imaged on the IVIS Spectrum CT by epifluorescence. In vivo images revealed increased signal in the liver region, and ex
vivo tissue imaging identified the specific tissue localization of signal indicative of biological change/injury (t-test, bar:
s.e.m.). Note the different tissue profiles induced by 2 different toxic drugs.
5.0
Tox/Efficacy
Screening
TAA
AngioSense 680
B. Whole Body Fluorescence Tomography
RMP
B. 2h Fluorescence Tomography Quantification
Tumor
Bioluminescence
Select Tissues
#p
1 Biodistribution and Tox/Efficacy Imaging Protocols
A. Mouse Tumor Model
Ex Vivo Tissue Quantification
A. IVIS Spectrum BLI
% of Untreated Control
A. Epifluorescence Screening for Tissue Injury using the imaging cocktail AMT750
4 Quantification of Single Dose 5-FU Tumor Efficacy and
Profiling for Potential Toxicity in Normal Tissues
AngioSense 680
AMT750 Cocktail
Cancer chemotherapy can produce severe side effects such as suppression of immune function and
damage to heart muscle, gastrointestinal tract, and liver. If serious enough, tissue injury can be a major
reason for late stage termination of drug discovery research projects, so it is becoming more important to
integrate safety/toxicology assessments earlier in the drug development process. There are a variety of
traditional serum markers, tailored mechanistically to specific tissues, however there are no current noninvasive assessment tools that are capable of looking broadly at in situ biological changes in target and
non-target tissue induced by chemical insult.
We used non-invasive near infrared (NIR) fluorescence tomography for whole-body imaging of
luciferase-expressing HT-29 human colon adenocarcinoma tumors implanted in nude mice. Both tumor
and host tissue responses were imaged using a cocktail of near infrared fluorescent imaging agents,
specific for cell death (Annexin-Vivo 750TM [AV750]), inflammatory matrix metalloproteases (MMPSenseTM
750 FASTTM [MMP750]), and metabolic changes in transferrin receptor expression (Transferrin-VivoTM 750
[TfV750]). Vascular changes were imaged using AngioSense® 680 (AS680). As a means of validating this
efficacy/tox screening approach, HT-29-bearing nude mice were dosed with 5-Fluorouracil (5-FU). 5-FU
has been a mainstay in the treatment of many cancers, including colorectal, but is associated with several
peripheral toxicities, including gastrointestinal, hepatic, renal, vascular, and (less frequently) cardiac.
Dosing was performed as a single IP bolus administration, using doses (50 and 100 mg/kg) known to have
minimal overt effects on body weight or tumor mass using this acute dosing regimen. At 2h and 24h post-5FU, independent cohorts of mice were injected IV with the AV750/MMP750/TfV 750 cocktail (AMT 750)
combined with AS680. Imaging was performed 24h later. No apparent effects were seen on tumor mass
with this very short treatment regimen, although there was a trend for decreased bioluminescence 2h
following 5-FU, however both tumor vascular leak and AMT750 signal increased at the high 5-FU dose.
Tumors, heart, liver and lungs showed predominant changes in AMT750 signal, with the heart showing the
most dramatic increase (>20-fold). The acute nature of the response, and the absence of histologic
inflammation, suggests that AV750 was detecting tissue apoptotic changes, perhaps in the vascular
endothelium. The stomach, kidneys, and Intestines showed predominant increases in AS680, indicating
changes in vascular permeability in these tissues (known to occur with 5-FU). The reversible nature of the
biological changes at 24h suggests that this could be a sensitive imaging approach for detecting early
tissue toxicity. These results agree well with observations in the literature that have seen 5-FU effects in the
same tissues in both preclinical studies and in humans undergoing treatment.
2 Validating a Cocktail of Imaging Agents for Detection of
Patterns of Drug-Induced Tissue Injury
Tumor Volume (mm3)
Abstract
Tissue Region Slice Images
Tumor
Control
50 mg/kg 100 mg/kg
Kidney
Control
50 mg/kg 100 mg/kg
After the animals were imaged in vivo, they were sacrificed by carbon dioxide asphyxiation. The organs (brain, heart,
lungs, liver, pancreas, spleen, stomach, intestines, kidneys, fat, skin, and tumor) were removed post-mortem then fixed
with 10% formalin for histological assessment by H&E.
No overt evidence of toxicity was seen despite measurable BLI
decreases in tumors from treated animals and biomarker changes in tumors, heart, liver, lungs, stomach, kidneys, and
intestines.
Luciferin
injection
FMT 4000 Fluorescent Tomography
Intestines
12-15 min
Heart
Stomach
Liver
Lungs
front
side
Bioluminescence
The Bioware® Brite Cell Line HT-29 Red-FLuc (PerkinElmer, Waltham, MA) was implanted in the lower right mammary fat pad of female
nu/nu mice (8 weeks old, Charles River Laboratory, Wilmington, MA). After the tumors have grown to approximately 130 mm3 (~3
weeks), the animals were randomized into groups. Designated groups of animals were injected IP with either 50, 100, or 100 BID
mg/kg of 5-Fluorouracil (5-FU) or vehicle (PBS with 2.5% DMSO). Specific groups of animals were injected with the imaging agent
cocktail AMT-750 (0.5x of Annexin-Vivo 750, 2x MMPSense 750 FAST, and 0.25x Transferrin-Vivo) combined with AngioSense 680 EX
and imaged at 2 or 24 h post-drug injection. Imaging of both A. bioluminescence (IVIS ® Spectrum CT system, PerkinElmer) and B.
whole body biodistribution (FMT® 4000 system, PerkinElmer) were performed 24 h post- imaging cocktail injection (12-15 minutes after
luciferin injection).
AngioSense 680
AMT750 Cocktail
680/750 overlap
HT-29 tumor bearing mice were treated with PBS, 50 mg/kg, or 100 mg/kd of 5-fluorouracil, and 2h later they were
injected with fluorescence imaging agents. A. Mice were injected with luciferin 12-15 min prior to imaging (at 24h) for
tumor bioluminescence (IVIS Spectrum CT). B. AMT750/AS680 fluorescent signal distribution was assessed using
fluorescence tomography (FMT 4000). C. 2-color representations of imaging datasets were rendered using 1 voxel slices
through the fluorescence tomographic datasets of indicated tissue regions. Red outlined images indicate tissues
showing obvious changes in signal intensity relative to controls.
Summary
The present studies provide evidence for the utility of a cocktail of imaging agents, detecting cell
death, MMP activity, and transferrin receptor upregulation, in the detection of both acute druginduced tissue changes as well as anti-tumor efficacy. A variety of toxicities are associated with 5-FU
treatment regimens, including oral and gastrointestinal mucositis, stomach pain, nephrotoxicity,
hepatotoxicity, and cardiotoxicity. It is remarkable that in these studies as little as a single
administration of 5-FU induced detectable biomarker changes in tissues previously characterized to
show injury following much longer 5-FU dosing regimens. Many of these toxicities are poorly
understood in patients and very difficult to recapitulate in mouse tumor models, however optical
imaging could readily detected biological changes in the absence of overt histologic changes.
In conclusion, imaging fluorescent biomarkers of tissue biological changes should provide a robust
approach to the simultaneous assessment of drug-induced efficacy and tissue injury in mice, allowing
early screens of therapeutics in drug discovery.