investigating the contribution of collagen to the tumor ... · jin li1, konstantinos zormpas ......

Convergence and Technologies

Investigating the Contribution of Collagen to theTumor Biomechanical Phenotype withNoninvasive Magnetic Resonance ElastographyJin Li1, Konstantinos Zormpas-Petridis1, Jessica K.R. Boult1, Emma L. Reeves1,Andreas Heindl2, Maria Vinci2, Filipa Lopes3, Craig Cummings1, Caroline J. Springer3,Louis Chesler4, Chris Jones2, Jeffrey C. Bamber1,Yinyin Yuan2, Ralph Sinkus5,Yann Jamin1,and Simon P. Robinson1

Abstract

Increased stiffness in the extracellular matrix (ECM) con-tributes to tumor progression and metastasis. Therefore, stro-mal modulating therapies and accompanying biomarkers arebeing developed to target ECM stiffness. Magnetic resonance(MR) elastography can noninvasively and quantitatively mapthe viscoelastic properties of tumors in vivo and thus has clearclinical applications. Herein, we used MR elastography, cou-pled with computational histopathology, to interrogate thecontribution of collagen to the tumor biomechanical pheno-type and to evaluate its sensitivity to collagenase-inducedstromal modulation. Elasticity (Gd) and viscosity (Gl) weresignificantly greater for orthotopic BT-474 (Gd¼5.9�0.2 kPa,Gl¼ 4.7� 0.2 kPa, n¼ 7) and luc-MDA-MB-231-LM2-4 (Gd¼7.9 � 0.4 kPa, Gl ¼ 6.0 � 0.2 kPa, n ¼ 6) breast cancerxenografts, and luc-PANC1 (Gd¼ 6.9� 0.3 kPa,Gl¼ 6.2� 0.2kPa, n ¼ 7) pancreatic cancer xenografts, compared withtumors associated with the nervous system, includingGTML/Trp53KI/KI medulloblastoma (Gd ¼ 3.5 � 0.2 kPa, Gl

¼ 2.3� 0.2 kPa, n¼ 7), orthotopic luc-D-212-MG (Gd¼ 3.5�0.2 kPa, Gl ¼ 2.3 � 0.2 kPa, n ¼ 7), luc-RG2 (Gd ¼ 3.5 � 0.2kPa,Gl¼ 2.3� 0.2 kPa, n¼ 5), and luc-U-87-MG (Gd¼ 3.5�

0.2 kPa, Gl ¼ 2.3 � 0.2 kPa, n ¼ 8) glioblastoma xenografts,intracranially propagated luc-MDA-MB-231-LM2-4 (Gd ¼ 3.7� 0.2 kPa, Gl¼ 2.2� 0.1 kPa, n¼ 7) breast cancer xenografts,andTh-MYCNneuroblastomas (Gd¼3.5�0.2 kPa,Gl¼2.3�0.2 kPa, n¼ 5). Positive correlations between both elasticity (r¼ 0.72, P < 0.0001) and viscosity (r ¼ 0.78, P < 0.0001) weredetermined with collagen fraction, but not with cellular orvascular density. Treatment with collagenase significantlyreduced Gd (P ¼ 0.002) and Gl (P ¼ 0.0006) in orthotopicbreast tumors. Texture analysis of extracted images ofpicrosirius red staining revealed significant negative correla-tions of entropy with Gd (r ¼ �0.69, P < 0.0001) and Gl

(r ¼ �0.76, P < 0.0001), and positive correlations of fractaldimension with Gd (r ¼ 0.75, P < 0.0001) and Gl (r ¼ 0.78,P < 0.0001). MR elastography can thus provide sensitiveimaging biomarkers of tumor collagen deposition and itstherapeutic modulation.

Significance:MR elastography enables noninvasive detec-tion of tumor stiffness and will aid in the development ofECM-targeting therapies.

IntroductionAberrant tensional homeostasis and increased stiffness are

hallmarks of cancer. The origin of elevated tumor stiffness is notfully understood, but may often reflect increased mechanicalstress associated with rapid tissue expansion and compressedvasculature and lymphatics, and extracellular matrix (ECM) rigid-ity. There is much evidence showing that increased tissue stiffness

contributes to malignant transformation, tumor progression andmetastasis (1, 2). The elevated solid stress and interstitial fluidpressure (IFP) that may drive increased tumor stiffness are alsotwo major obstacles to efficient tumor drug delivery (3).

Both breast and pancreatic cancers are characterized by exces-sive desmoplastic stromal reaction and dense ECM (4, 5). Col-lagen, the principal component of the fibrillary protein networkwithin the ECM, is the major contributor to ECM stiffening, is

1Division of Radiotherapy and Imaging, The Institute of Cancer Research,London, United Kingdom. 2Division of Molecular Pathology, The Institute ofCancer Research, London, United Kingdom. 3Cancer Therapeutics Unit, TheInstitute of Cancer Research, London, United Kingdom. 4Division of ClinicalStudies, The Institute of Cancer Research, London, United Kingdom. 5Division ofImaging Sciences and Biomedical Engineering, King's College London, London,United Kingdom.

Note: Supplementary data for this article are available at Cancer ResearchOnline (http://cancerres.aacrjournals.org/).

J. Li and K. Zormpas-Petridis contributed equally to this article.

Y. Jamin and S.P. Robinson are the co-senior authors of this article.

Current address forM. Vinci: BambinoGes�u Children'sHospital-IRCCS, Rome, Italy;and current address for F. Lopes and C.J. Springer, Drug Discovery Unit CancerResearch UK Manchester Institute, University of Manchester, United Kingdom.

Corresponding Authors: Yann Jamin, The Institute of Cancer Research, London,15 Cotswold Road, Sutton, Surrey SM2 5NG, United Kingdom. Phone: 4420-8722-4492; Fax: 4420-8661-0846; E-mail: [email protected]; and Simon P.Robinson, Phone: 4420-8722-4528; E-mail: [email protected]

Cancer Res 2019;79:5874–83

doi: 10.1158/0008-5472.CAN-19-1595

�2019 American Association for Cancer Research.

CancerResearch

Cancer Res; 79(22) November 15, 20195874

on June 8, 2020. © 2019 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst October 11, 2019; DOI: 10.1158/0008-5472.CAN-19-1595

http://crossmark.crossref.org/dialog/?doi=10.1158/0008-5472.CAN-19-1595&domain=pdf&date_stamp=2019-10-25

http://crossmark.crossref.org/dialog/?doi=10.1158/0008-5472.CAN-19-1595&domain=pdf&date_stamp=2019-10-25

http://cancerres.aacrjournals.org/

strongly implicated in tumor evolution and progression, and isassociated with poor patient prognosis (6–9). Increased collagendeposition and enhancedmatrix cross-linking occurs with progres-sive structural remodeling of the ECM scaffold, which facilitatestumor growth (9). Suppression of collagen synthesis inhibitstumor growth andmetastasis, and enhances drug penetration (10).

Significant efforts are currently focused on targeting tumorECMstiffness for therapeutic gain (2, 4, 11–13). The developmentof stromal modulating therapies would benefit from noninvasiveimaging biomarkers to inform on changes associated with ther-apeutic efficacy. A number of innovative magnetic resonance(MR) and ultrasound (US) imaging techniques are beingexploited to image the viscoelastic and other mechanical prop-erties of tissue in vivo (14–17). One approach,MR elastography, isbeing used to visualize andmeasure tissue elasticity and viscosityin vivo. MR elastography yields quantitative images, and thereforeimaging biomarkers, that map the absolute value of the complexshear modulus G� in terms of its two components, the elasticitymodulus Gd (a measure of the ability of an object to resume anormal shape after being stretched or compressed) and theviscosity modulus Gl (a measure of resistance to gradual defor-mation by shear or tensile stress). We and others have demon-strated, in preclinical tumormodels and cancer patients in vivo, thepotential of the MR elastography-derived mechanical phenotypeto inform on the underlying tumormicrostructure and treatment-induced changes to its integrity (18–25).

Early imaging biomarker development demands closeimaging-pathology correlation to understand the biologicalprocesses underpinning the imaging measurement, which canbe meaningfully studied using animal models (26–28). Thesystematic evaluation of tumor stromal components and theircontribution to tissue stiffness as measured by MR elasto-graphy is in its infancy. This study describes the use of MRelastography, coupled with computational histopathology, tointerrogate the contribution of the collagen network to thebiomechanical phenotype imaged in vivo in a wide range oforthotopically-propagated and spontaneously arising trans-genic models with disparate pathologies. The sensitivity ofMR elastography to monitor stromal modulation in vivofollowing administration of collagenase is also demonstrated.

Materials and MethodsCells

All cell lines used in this study tested negative for Mycoplasmainfection at the time of tumor implantation and human cells wereauthenticated by short tandem repeat (STR) profiling. Theirorigins, provenance, and culture conditions are summarized inSupplementary Tables S1 and S2.

Animals and tumor modelsAll animal experiments were approved by the Institute of

Cancer Research Animal Welfare and Ethical Review Body, andperformed in accordance with the UK Home Office Animals(Scientific Procedures) Act 1986, the United Kingdom NationalCancer Research Institute guidelines for the welfare of animals incancer research (29), and reported according to the AnimalResearch: Reporting In Vivo Experiments (ARRIVE) guide-lines (30). Mice were housed in specific pathogen-free rooms inautoclaved, aseptic microisolator cages with a maximum of5 animals per cage. Mice were allowed access to sterile food andwater ad libitum. A total of 84 mice were enrolled and a summaryof the in vivo models used in this study is presented in Table 1.

Intracranial tumor propagation.Human luc-U-87 MG glioblasto-ma (5 � 104), rat luc-RG2 glioma (5 � 103), human luc-D-212MG pediatric hemispheric giant-cell glioblastoma (5 � 103), orhuman metastatic luc-MDA-MB-231 LM2-4 breast cancer (5 �103) cells were implanted supratentorially in the brains of adultfemale athymic NCr-Foxn1nu mice (Charles River) as describedpreviously (19). Animals were anesthetized using 1% to 2%isoflurane in oxygen (1 l/min). A �1 cm incision was made inthe skin on the top of the head, and a 1 mm hole drilled using asurgical bone microdrill (Harvard Apparatus). Cell suspension(5mL)was then injected at a depth of 3mm from the dura, at a rateof 2 mL/minute, using a 10 mL syringe (VWR International) and ananomite syringe pump (Harvard Apparatus). The needle wasremoved 3 minutes after completion of the injection and theskin repaired with Vetbond Tissue Adhesive (3M Animal CareProducts).

Tumor establishment and growth were monitored with biolu-minescence imaging (BLI) using a Xenogen IVIS 200 system

Table 1. Summary of the in vivo models

Cell line Characteristics StudyInjectionroute Cells injected n

Tumor volume(mm3)a

luc-MDA-MB-231 LM2–4 Highly malignant human triple-negativebreast cancer cells isolated from a lungmetastasis

MRE-histology correlation i.c. 5 � 103 5 29 � 8MRE-histology correlation o.t. 2 � 106 6 406 � 47

luc-RG2 Rat glioma cells MRE-histology correlation o.t. 5 � 103 5 30 � 4luc-U-87 MG Human glioblastoma cells MRE-histology correlation i.c. 5 � 104 8 26 � 4

MRE-histology correlation s.c 3 � 106 6 616 � 114luc-D-212 MG Derived from a pediatric hemispheric

giant-cell glioblastomaMRE-histology correlation i.c. 1.5 � 105 7 12 � 3

MDA-MB-231 Highly malignant human triple negativebreast cancer cells

Collagenase treatment o.t. 2 � 106 14 385 � 49

BT-474 Invasive ductal breast carcinoma cells MRE-histology correlation o.t. 5 � 106 7 438 � 36Collagenase treatment o.t. 5 � 106 7 287 � 51

luc-PANC-1 Pancreatic epithelioid carcinoma cells MRE-histology correlation o.t. 1 � 107 7 1402 � 280Transgenic mouse modelsTh-MYCN High-risk neuroblastoma MRE-histology correlation Spontaneously arising tumors 5 1369 � 254GTML/Trp53KI/i Medulloblastoma MRE-histology correlation Spontaneously arising tumors 7 27 � 4

Abbreviations: i.c., intracranial; o.t., orthotopic; s.c., subcutaneous.aTumor volumes at the time of the MR elastography experiment and determined using segmentation from regions of interest drawn on each tumor-containingT2-weighted MRI slice.

Imaging Tumor Viscoelasticity with MR Elastography

www.aacrjournals.org Cancer Res; 79(22) November 15, 2019 5875




coupled with LivingImage software (Caliper Life Sciences). Lucif-erin (150 mg/kg; Caliper Life Sciences) was administered intra-peritoneally 10 minutes before imaging. MR elastography wasperformed when the BLI photon flux reached a threshold valuepreviously determined to represent a tumor of approximately 30to 40 mm3, a volume considered of sufficient size to acquire MRelastography data but not large enough to cause neurologic effectsin the mice. The average time from implantation to imaging was18 days for the luc-U-87 MG and luc-MDA-MB-231 LM2-4tumors, 22 days for the luc-RG2 tumors, and 45 days for theluc-D-212 MG glioblastomas.

GTML/Trp53KI/KI transgenic model of medulloblastoma. The gen-eration of the GTML/Trp53KI/KI mice has been reported previous-ly (31). Mice were genotyped to detect the presence of humanMYCN and Trp53 transgenes. Male and female mice were mon-itored twice weekly for the development of a BLI signal from themidbrain. MR elastography was performed when the photon fluxreached a threshold value previously determined to represent atumor of approximately 20 to 30 mm3.

Th-MYCN transgenic model of neuroblastoma. Transgenic Th-MYCNmice were genotyped to detect the presence of the humanMYCN transgene (27). Both male and female hemizygous micewere used, which spontaneously developed palpable abdominaltumors between 50 and 130 days with a 25% penetrance. Tumorprogression was monitored weekly by palpation by an experi-enced technician until the tumor reached a diameter greater than�5 mm, at which point they underwent MR elastography.

SubcutaneousU87-MGxenografts.Adult femaleNCr-Foxn1numicewere injected subcutaneously in the flank with 2 � 106 luc-U-87MG cells. Tumor development was monitored weekly by calipermeasurements, and MR elastography performed when tumorsreached a diameter of �7 mm.

Orthotopic models of breast and pancreatic cancer. Human luc-MDA-MB-231 LM2-4 or BT-474 breast cancer cells (5� 106) wereinjected into the third abdominal fat pad of adult female NCr-Foxn1numice [100mL cell suspension in PBS andMatrigel (1:1)]. A17b-estradiol pellet (60-day release; Innovative Research of Amer-ica) was implanted in the neck nape 1 day before implantation ofBT-474 cells. Tumor development was monitored weekly bycaliper measurements, and MR elastography performed whentumors reached a diameter of �7 mm.

For the orthotopic propagation of pancreatic cancer xeno-grafts, a small incision was made on the left flank of adultfemale athymic CD1-Foxn1nu mice (Charles River) through theskin and peritoneum, the pancreas exteriorized, and humanluc-Panc-1 [1 � 107 cells in suspension in PBS and Matrigel(1:1)] injected using a Hamilton syringe. The pancreas was thenreturned into the abdominal cavity and the incision sutured.Successful engraftment and tumor progression were confirmedusing BLI, and MR elastography performed �90 dayspostimplantation.

Response to collagenaseOrthotopic parental MDA-MB-231 and BT-474 breast cancer

xenografts were propagated as described previously. MR elasto-graphy was performed on established tumors 24 hours prior toand 5 hours after intravenous treatment with either collagenase

(62 U/kg, bacterial collagenase from Clostridium histolyticum;Sigma-Aldrich) or saline.

MRI and MR elastography data acquisition and analysisMRI was performed on a 7T horizontal bore MicroImaging

system (Bruker) using a 3 cm birdcage volume coil. Tumor-bearing mice were anesthetized with a 10 mL/kg intraperitonealinjection of fentanyl citrate (0.315 mg/mL) plus fluanisone(10mg/mL; Hypnorm; Janssen Pharmaceutical Ltd.), midazolam(5mg/mL; Hypnovel; Roche), and sterile water (used at a ratio of1:1:2). Themouse core temperature wasmaintained at 37�Cwithwarm air blown through the magnet bore.

Anatomical T2-weighted images [using a rapid acquisitionwith refocused echoes (RARE) sequence, with TE ¼ 36 millise-conds, TR ¼ 4.5 seconds, RARE factor ¼ 8, 40 contiguous 1-mm-thick transverse slices, 1 average, matrix size 128� 128 over a 3�3 cm field of view (FOV)] were used to localize and determine thetumor volume, plan the MR elastography acquisition, and opti-mize the local field homogeneity over the region of interest (ROI)using the FASTMAP algorithm.

MR elastography was performed as described previous-ly (18, 19). The mechanical vibrations, generated by an electro-magnetic shaker (Br€uel & Kjaer), were transmitted through aflexible nylon rod to either (i) a square piston with a concavecurved face positionedon themouse head for intracranial tumors,or directly on the skin over subcutaneous luc-U87MG tumors andorthotopically propagated breast cancer xenografts, or (ii) a roundflat-faced piston placed on the abdomen above a palpated tumor(Th-MYCNor orthotopic pancreatic tumor), all positionedwithinthe volume coil at the isocenter of the magnetic field. MRelastography was performed using mechanical excitations at avibration frequency of 1,000Hz, exciting the shakerwith a voltagethat generatedmechanicalwaves inside the tumorwith amplitudegreater than 0.5 mm. A 2D spin-echo sequence incorporatingsinusoidal motion-sensitizing gradients synchronized to themechanical excitation was used. Data were acquired in threeorthogonal directions from 10 contiguous transverse slices(300 mm thick), using 2 averages of 64 phase encoding steps overa 1.92 � 1.92 cm FOV, with TE ¼ 27 milliseconds, TR ¼ 1001milliseconds, and 8 time sampling steps, giving an isotropicspatial sampling of 300� 300� 300 mm of the mechanical wavepropagation displacement inside the tumor. The total acquisitiontime was � 51 minutes. Finally, high-resolution T2-weightedRARE images were acquired from the same 10 contiguous trans-verse slices (TE¼ 36milliseconds, TR¼ 4.5 seconds, RARE factor¼ 8, 300 mm thick, 10 averages, matrix size 128� 128 over a 1.92� 1.92 cm FOV).

Image reconstruction and analysis. Parametric maps of the abso-lute value of the complex shear modulus |G�|, elasticity Gd, andviscosity Gl (where |G�| ¼ Gd þ iGl) were reconstructed using in-house software from the 3D displacement vector measured asdescribed previously, and using the following equation (32):

�rv2~q ¼ G�r2~q; ~q ¼ ~r�~u 2 C3;

where~q is the complex-valued curl of themeasured displacementfield ~u, r is the density of the material, and v is the angularfrequency. For each slice, Gd and Gl (kPa) were determinedpixelwise from an ROI covering the whole tumor delineated fromthe high resolution T2-weighted images.

Li et al.

Cancer Res; 79(22) November 15, 2019 Cancer Research5876




Computational histopathologyTissue preparation.Guided by the T2-weightedMR images, tumorswere carefully excised and orientated for subsequent histopath-ologic processing. Adjacent formalin-fixed paraffin embeddedsections (3 mm) were cut and tinctorially stained with picrosiriusred (for collagen I and III), hematoxylin and eosin (H&E, forcellularity), or immunohistochemically processed for detectionof the murine vascular endothelial marker Cd31 (rabbit EP3095;Millipore), using diaminobenzidine (DAB) as the chromogen.

Digitized histology. Whole-slide images were digitized using aNanoZoomer XR scanner (�20 magnification, 0.46 mm resolu-tion; Hamamatsu). ROIs of viable tumor and necrosis for eachsample were independently provided. Histology images weresubsequently split into tiles of 2,000 � 2,000 pixels (jpeg).

Picrosirius red staining segmentation. A macro was written in Fiji(https://fiji.sc/) to segment picrosirius red staining from each tileusing ImageJ/Fiji plugins (Java 8). Images were first convertedfrom the RGB color space into the green-red (a�) color channel ofthe CIELAB color space (lightness, green–red, and blue–yellow),and subsequently thresholded to segment the picrosirius redstaining from the background. Following appraisal of automaticandmanual thresholding, twomanual thresholds were chosen toachieve the optimal picrosirius red staining segmentation acrossall samples, while compensating for variations in staining inten-sity and complex background values associated with the differentcancer pathologies. A threshold value of above þ17 was used inthe a� color channel for the majority of tumor types, increased toþ23 for medulloblastomas and neuroblastomas arising in theGTML/Trp53KI/KI and Th-MYCN transgenic mice respectively, andsubcutaneous luc-U-87 MG tumors. The segmentation algorithmwas tested using independent annotation, from a single observerblinded to the algorithm's result, of stained/nonstained points(3920/3478) on 17 different samples across all cancer types,giving an accuracy of 95% with 91% sensitivity.

Cell segmentation from H&E-stained sections. Images were pro-cessed using the EBImage Bioconductor package (33). Cell nucleiwere extracted from each image tile using the Otsu thresholdingalgorithm, followed by morphologic opening to delete the noisystructures and the Watershed algorithm to separate clusterednuclei (34).

Cd31 segmentation from Cd31-stained IHC sections. A macro waswritten in Fiji to extract DAB staining from each tile by applyingcolor unmixing to extract the brown color channel, followed byapplication of the maximum entropy threshold detection meth-od, both using ImageJ/Fiji plugins (Java 8) as describedpreviously (27).

Generation of collagen fraction, cellularity, and vascular densityparametric maps. Whole-slide images of picrosirius red staining,segmented cells, and Cd31 staining were converted into binaryand processed to match MR elastography resolution (300 � 300mm), with the fraction of pixels occupied by the center of each cellnucleus, picrosirius red, and Cd31 staining within 664 � 664pixel-regions representing a single pixel in the final calculatedmaps. Quantitative analysis of each stain was performed from 1histologic tumor section aligned with the central slice of theelastogram. Necrotic areas visible on both T2-weighted images

and corresponding histopathology slides were subtracted fromthe viable tumor by manual regional segmentation and excludedfrom the quantitative analysis.

Texture analysis. We evaluated the 2D heterogeneity of collagendistribution by quantifying entropy and fractal dimension (FD),as described by Nieskoski and colleagues (35), on the extractedcollagen parametric maps. The texture analysis was implementedin Matlab (R2018b; Mathworks). Necrotic areas were subtractedfrom the viable tumor and excluded from the analysis. Entropywas quantified tomeasure the irregularity of collagen distributionby applying the "entropy" function of Matlab, which uses theequation:

Entropy ¼ �XN

i¼1p � log2 p;

where p represents the normalized histogram count of the colla-gen fraction. FD defines the complexity (textural roughness) ofcollagen distribution within the tumor samples. The Hausdorff(box-counting) method was applied in the histology imageswithin the 664 � 664 pixel-regions described earlier, using theequation:

FD ¼ log N eð Þð Þ= log 1e

� �;

where e represents thebox size set to the size of the image andN(e)corresponds to the number of boxes of size e, which containcollagen. The median value of the generated FDmaps (excludingnecrotic areas) was used as the tumor's FD value.

Statistical analysisStatistical analysis was performed using GraphPad Prism 6

(GraphPad Software Inc.). Unless stated otherwise, data are pre-sented as mean � 1 SEM. Significant differences in quantitativeMR elastography parameters between tumor types, and in relativetreatment-induced changes, were identified using the non-parametric Mann–Whitney U test with a 5% level of significance,while significant changes in MR elastography parameters withtreatment were identified using the Wilcoxon matched-pairssigned ranks test with a 5% level of significance. Significantcorrelationswere determined using linear regression analysiswitha 5% level of significance using the robust regression and outlierremoval approach (36).

ResultsMR elastography was successfully performed in all mice, yield-

ing an intertumoral coefficient of variation (CoV) of 13.0% and15.4% for repeated measurements of Gd and Gl, respectively(Supplementary Materials and Methods). MR elastographyrevealed a heterogeneous distribution of elasticity and viscosityacross the 9 orthotopic and transgenic models of cancer investi-gated (Fig. 1A). Pronounced contrast between the establishedtumor and the surrounding brain was clearly evident in theintracranial models, with the lesion boundaries aligning withthose seen in the high-resolution T2-weighted images. Quantita-tive analysis of the MR elastography data demonstrated a widerange in Gd and Gl values across the models, from 3.5 � 0.2 and2.2 � 0.2 kPa respectively, in the GTML/Trp53KI/KI transgenicmousemodel ofmedulloblastoma, to 7.9� 0.4 and 6.0� 0.2 kPain the orthotopic luc-MDA-MB-231 LM2-4mammary carcinomas





https://fiji.sc/


(Fig. 1B). Collectively, elasticity and viscosity were significantly(P < 0.0001) greater for the orthotopically-propagated breast andpancreas models, compared with the tumors associated with thecentral or peripheral nervous system.

Intravenous injection of collagenase resulted in a clear overallreduction in the elasticity and viscosity of orthotopic MDA-MB-231 and BT-474 mammary tumors, as measured by MR elasto-graphy, 5 hours after administration (Fig. 2A). Tumor regionsexhibiting relatively high Gd and Gl pretreatment were typicallyreduced following challenge with collagenase. No similarresponse was evident in the vehicle-treated mice. Collectively,collagenase resulted in a significant reduction in both Gd (6.0 �0.4 kPa to 4.9� 0.4 kPa, P¼ 0.001) andGl (3.8� 0.6 kPa to 3.0�0.5 kPa, P¼ 0.001), whichwas not observed in the vehicle-treatedcohort (Gd: 5.1 � 0.4 kPa to 5.4 � 0.3 kPa, P ¼ 0.3; Gl: 3.0 � 0.6kPa to 3.4� 0.5 kPa, P¼ 0.13). Relative changes in bothGd andGl

were significantly different between the collagenase and vehicle-treated cohorts (P ¼ 0.0008 and 0.0004, Fig. 2B).

To investigate the pathologic determinants of the regionalvariations in tumor viscoelasticity seen in vivo, parametric mapsof Gd and Gl were compared with maps of picrosirius red

(collagen I and III), H&E (cellularity), and Cd31 (vasculardensity) staining, automatically segmented from high-resolution images of aligned tissue sections from the sametumor (Fig. 3A). In the GTML/Trp53KI/KI transgenic mousemodel, which exhibited the lowest mean Gd and Gl, of all themodels investigated, tumors presented with a thin layer-likeregion of elevated Gd and Gl spatially associated with strongpicrosirius red staining adjacent to the skull, consistent withtumor invasion into the collagen-rich meninges (Fig. 3B). Ingeneral, but especially in breast and pancreatic tumors, regionsdemonstrating high values of Gd and Gl spatially correspondedto cellular regions with higher deposition of collagen. In sometumors, histologically-defined regions of extensive tissue dam-age (e.g., necrosis) were spatially associated with areas ofmarkedly lower elasticity and viscosity. It is important to notethe relative softness of these necrotic regions irrespective oftheir also having high collagen content. These regions wereexcluded from the subsequent quantitative analysis. Quantita-tive analysis identified statistically significant positive intertu-mor correlations of both tumor-mean elasticity Gd (r ¼ 0.72,P < 0.0001) and viscosity Gl (r ¼ 0.78, P < 0.0001) with

Figure 1.

MR elastography reveals marked differences in viscoelasticity measured in vivo in orthotopic and transgenic models of cancer. A, Representative T2-weightedMRI and associated parametric maps of elasticity Gd and viscosity Gl obtained from amedulloblastoma spontaneously arising in the brain of a GTML/Trp53KI/KI

transgenic mouse (n¼ 7), intracranially (i.c.) propagated luc-MDA-MB-231 LM2-4 breast cancer (n¼ 5) and luc-D-212-MG glioblastoma xenografts (n¼ 7), aneuroblastoma spontaneously arising in a Th-MYCN transgenic mouse (n¼ 5), intracranially propagated luc-RG2 (n¼ 5) and luc-U-87 MG glioblastomaxenografts (n¼ 8), and orthotopic BT-474 breast cancer (n¼ 7), luc-PANC1 pancreatic cancer (n¼ 7) and luc-MDA-MB-231 LM2-4 breast cancer (n¼ 6)xenografts. The dashed line indicates the tumor boundaries defined on T2-weighted images. B,Quantitative summary of mean of the mean tumor Gd and Gl (�1SEM) for each model. The insets show that the collective average of Gd and Gl determined in the breast and pancreatic tumors was significantly higher than intumors arising in the nervous system. �� , P < 0.0001.

Li et al.





tumor-mean collagen fraction, but not with tumor-mean cel-lularity and vascular density (Fig. 3C).

The irregularity of collagen distribution and deposition, and itsrelationship to tumor viscoelasticity in vivo, was further investi-gated using texture analysis of the extracted images of picrosiriusred staining. Significant negative correlations of entropy with Gd

(r ¼ �0.69, P < 0.0001) and Gl (r ¼ �0.76, P < 0.0001), andpositive correlations of FD with Gd (r ¼ 0.75, P < 0.0001) and Gl

(r¼ 0.78, P < 0.0001) were found (Fig. 4A and B). Entropy valuesclose to 0 and relatively high values of FD determined in the BT-474, luc-PANC-1, and luc-MDA-MB-231 LM2-4 tumors are con-sistent with the presence of a homogeneous and dense collagennetwork.Note that in themodels investigated, increasing collagencontent was associated with both increasing density and unifor-mity of its distribution, as shown by the mono-exponentialrelationship of entropy (y ¼ 3.86e�0.24x, r2 ¼ 0.76) and logarith-mic relationship of FD (y ¼ 0.15 lnx þ 1.16, r2 ¼ 0.98) withcollagen fraction, respectively (Fig. 4C).

DiscussionECM stiffening is increasingly recognized as a major mechan-

ical signal, which alters cell behavior and in part confers cancercell hallmark capabilities including sustained growth, invasion,and metastasis (6, 7, 37–42). ECM stiffening is also associatedwith increased solid stress and IFP, two other hallmarks oftumor mechanobiology that induce blood and lymphatic vesselcompression and reduce transcapillary transport, respectively,and that impair effective drug delivery (3). Disrupting thecross-talk between cancer cells and the ECM, as well as reversingECM stiffness, solid stress or IFP, thus represents a promisingtherapeutic strategy (11). The clinical development of stromalmodulating therapies would be facilitated and accelerated bynoninvasive imagingmethods to longitudinally image and quan-tify tumor mechanical properties in vivo (15–17).

In this preclinical study, our quantitative MR elastographytumor data, combined with aligned computational histopathol-ogy, showed that elevated tumor Gd and Gl correlated withincreased collagen deposition across a wide range of clinically-relevant tumor models with disparate pathologies. Our dataparticularly highlight the relative softness of tumors arising inthe nervous system, and the predicted elevated stiffness of ortho-topic breast and pancreatic models. Furthermore, modulation ofthe collagen network with bacterial collagenase in two relativelystiff orthotopic models of breast cancer revealed a marked reduc-tion in both elasticity and viscosity. Collectively our resultsdemonstrate that the deposition and density of the collagen fibernetwork is a major determinant of the tumor biomechanicalphenotype, for which MR elastography provides sensitivein vivo imaging biomarkers, including biomarkers of its pharma-cological modulation.

OurMR elastography study also aligns with several establishedconcepts in tissue mechanics. The elastic properties of tissues aredefined by the mechanical characteristics of their cells and theirsurrounding ECM, yet tissue macroscopic viscoelastic propertiesare not greatly affected by differences in cell morphology anddensity, but are largely governed by their ECM composition (43).Herein, histologically-confirmed regions of tissue damage (necro-sis) were associatedwithmarkedly reduced elasticity and viscosityvisualized in vivo. This relation is irrespective of the collagenfraction within the necrotic regions and explains the heteroge-neous appearance of the viscoelastic maps, as well as the absenceof spatial correspondence with the homogenous network ofcollagen in the luc-PANC-1 tumors, which exhibit marked andwidespread regions of multifocal discrete necrosis. This is alsoconsistent with previous MR elastography studies reporting areduction in viscoelastic properties following vascular targeted-therapy induced cell death and/or reduction in vascular densi-ty (18, 20, 21). Although this demonstrates the importance ofboth a viable cellular and vascular network to tissue integrity and

Figure 2.

MR elastography can inform on tumor stromal modulation induced by collagenase.A, Representative T2-weighted anatomical MRI images and parametric mapsof Gd and Gl acquired frommice bearing orthotopic BT-474 breast cancer xenografts 24 hours prior to and 5 hours after intravenous injection of either vehicle orcollagenase. B, Relative changes (%) in tumor median Gd andGl measured in orthotopic MDA-MB-231 (blue symbols) and BT-474 (red symbols) breast cancerxenografts measured 5 hours after administration of either vehicle (*) or collagenase (D). Data are the individual changes from each tumor and the combinedcohort mean� 1 SEM. Collagenase induced a significant reduction in both Gd and Gl compared with vehicle control.






mechanopathology, our data show they provide a very moderatecontribution to the observed range of mechanical phenotypes, incontrast to differences in ECM characteristics.

Intracranial tumors were typically at the softer end of thespectrum of viscoelastic properties measured in this study (Gd

�4 kPa and Gl �2 kPa; refs. 19, 22). The high compliance(inversely related to elastic modulus) of human brain tumorsrelative to the surrounding brain parenchyma has been reported,and in the case of glioblastoma, stiffness has been shown todecrease with tumor grade, measured as part of clinical MR

elastography-embedded prospective studies (44, 45). Braintumors share the unique composition of the healthy brain,characterized by the absence of a fibrillar network (which resistsshear deformation) and a reliance on hygroscopic hyaluronic acid(themainmechanical support against compressive forces that actto cause a change in volume but offer little resistance to shear) formechanical support. The lack of collagen, a major facilitator oftumor cell intravasation, one of the earliest stages of metastasis, isconsistentwith the fact that intracranial tumors rarely disseminateoutside the brain (46). Note also the relatively low viscoelastic

Figure 3.

MR elastography-derived elasticityGd and viscosity Gl correlate withtumor collagen fraction. A and B,Representative anatomical T2-weighted MRI images and parametricmaps of elasticity (Gd) and viscosity(Gl), and the correspondingcomputed maps of picrosirius redstaining (collagen I and III),hematoxylin and eosin staining(cellularity), and IHC detection ofCd31 (vascular density) extractedfrom high-resolution images of tissuesections from: a medulloblastomaspontaneously arising in the brain ofa GTML/Trp53KI/KI transgenic mouse,a neuroblastoma spontaneouslyarising in a Th-MYCN transgenicmouse, and an intracraniallypropagated luc-U87 MGglioblastoma xenograft (A), andfrom orthotopic BT-474, luc-PANC-1,and luc-MDA-MB-231 LM2-4 breastand pancreatic cancer xenografts(B). Note that the areas above thebrain tumors showing both highviscoelastic properties and highpicrosirius red staining correspond toregions where the tumor has invadedthe collagen-rich meninges. Dottedlines, regions of necrosis. Note that innecrotic areas, high collagen densityis not associated with high value ofGd and Gl. C, Scatter graphs ofindividual tumor mean Gd and Gl

plotted against mean collagenfraction (%), cellularity and vesseldensity. Linear regression analysisand 95% confidence intervals forsignificant correlations are shown.

Li et al.





properties of the intracranially grown luc-MDA-MB-231 LM2-4tumors comparedwith thosepropagated in themammary fat pad,highlighting the contribution of implantation site to the resultingbiomechanical phenotype. Abdominal neuroblastomas sponta-neously arising in Th-MYCN transgenic mice, and which arederived from the sympathetic nervous system, were also relativelysoft with little collagen, consistent with their general clinicalpresentation (47). Interestingly, increased collagen III (reticulin)deposition helps to define an ultra-high-risk group of patients inwhich increased stiffness relates tometastatic potential, themajorcause of mortality for children with neuroblastoma (48, 49).

MR elastography revealed an acute reduction in breast tumorelasticity and viscosity following systemic administration of col-lagenase. A similar biomechanical response was recently reportedfollowing intratumoral injection of collagenase, measured usingUS-based elastography (16). Collagenase has also been shown torapidly decrease tumor IFP in breast cancer xenografts (50, 51).Cleavage of collagen with collagenase produces large peptidefragments, which remain trapped within the ECM, making theacute effects of collagenase challenging to detect on conventionalhistology at such an acute time point (12). Having establishedtheir dependence on relative tumor collagen fraction, quantita-tion of tumorGd andGl can also provide early MRI biomarkers of

response to collagen degradation in vivo. In this way, MR elasto-graphy could thus be exploited for monitoring direct enzymaticdegradation of the ECM and/or targeted inhibition of collagensynthesis, potent strategies being actively investigated to improvedrug penetration in solid tumors, and showing promising resultsin clinical trials (10, 12, 52).

Solid stress has been shown to correlate with collagen depo-sition and be the major contributor to total tumor pressure inmodels of pancreatic cancer (53). ECM stiffening is a marker ofpoor prognosis in breast cancer and pancreatic ductal adenocar-cinoma (6, 37, 38, 40). Given its sensitivity to collagen deposi-tion, MR elastography may thus potentially provide prognosticinformation, and through its sensitivity to collagen modulation,provide biomarkers of response to collagen-targeted approachesdesigned to alleviate solid stress for improved drug delivery.Tumor shear modulus, measured by US elastography, has beenshown to positively correlate with collagen deposition andinversely correlate with functional vasculature and drug deliv-ery (17). However, MR elastography cannot measure pressuredirectly, and as suchmay not be directly informative for strategiesdesigned to decrease fluid stress (e.g., hyaluronidase), for whichdirectmeasurements of IFP are required. Reconstruction of tumorviscoelastic parameters from the properties of propagating waves

Figure 4.

The relationship between MR elastography-derived tumor viscoelastic properties and the spatial distribution and deposition of collagen evaluated by textureanalysis. A and B, Scatter graphs of individual mean tumor elasticity Gd and viscosity Gl plotted against entropy (A) and FD (B). Linear regression analysis and95% confidence intervals for significant correlations are shown. Note that both Gd and Gl are negatively correlated with entropy and positively correlated withFD, indicating that increased tumor stiffness is associated with more complex andmore homogenously distributed collagen. Entropy values close to zero andrelatively high values of FD determined in the BT-474, luc-PANC-1, and luc-MDA-MB-231 LM2-4 tumors are consistent with the presence of such a homogeneousand dense collagen network. C, Scatter graphs showing the mono-exponential relationship between entropy and collagen fraction (y¼ 3.86e�0.24x, r2¼ 0.76),and logarithmic relationship between FD and collagen fraction (y¼ 0.15 lnxþ 1.16, r2¼ 0.98).






or strain visualized by MRI or US has often used a simplemonophasic linear viscoelastic mechanical model, which doesnot take into account any mobile fluid component (54). Mea-suring IFP and solid stress in itself represents a major clinicalchallenge, as current approaches for measuring both are invasive.Clinical measurement of IFP with invasive wick-in-needles onlypermits very discrete sampling of the tissue and is unreliable, andinnovative preclinical MRI methods shown to correlate with IFPin vivo would be difficult to routinely implement in a clinicalsetting (51, 55).

Finally, the utility of MR elastography is being actively evalu-ated clinically in a wide range of health conditions includingneuro- and cardiovascular pathologies, with applications includ-ing diagnosis, staging, surgical planning, and intraoperative guid-ance. For example, the integration of MR elastography into themanagement of patients with chronic liver disease is now becom-ing well established for its excellent diagnostic accuracy, superiorto that of US-based techniques, and ability to discriminate thedifferent stages of liver fibrosis, characterized by increased depo-sition and crosslinking of collagen (56). Although less widelyavailable and more costly than US-based techniques, MR elasto-graphy does allow the 3D investigation of large tissue areas anddeep-seated organs. MR elastography can uniquely afford thenoninvasive investigation of the mechanical properties of thebrain and its pathology including, as recently demonstrated,neuronal activity (57). Finally, MR elastography can be incorpo-rated into multiparametric MR imaging protocols to enablecomparison with other MRI-derived biomarkers of tumorstructure and function in a single clinical scanning session.

In conclusion, we have shown that quantitation of elasticityGd

and viscosity Gl are sensitive imaging biomarkers of tumorcollagen deposition, and response to direct enzymatic degrada-tion of the collagen network. Given the importance of elevatedECM stiffness in tumor progression, and the continuing need fornew technologies for faster and more accurate detection, diagno-sis and monitoring, MR elastography has the potential to informnoninvasively on prognosis and improve risk stratification ofpatients with cancer with dense stroma, and accelerate the devel-opment of stromal targeting/modulating treatment strategies.

Disclosure of Potential Conflicts of InterestNo potential conflicts of interest were disclosed.

Authors' ContributionsConception and design: J. Li, K. Zormpas-Petridis, C.J. Springer, J.C. Bamber,Y. Jamin, S.P. RobinsonDevelopment of methodology: K. Zormpas-Petridis, M. Vinci, C. Cummings,C.J. Springer, J.C. Bamber, Y. Yuan, R. Sinkus, Y. JaminAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.): J. Li, K. Zormpas-Petridis, J.K.R. Boult, E.L. Reeves,M. Vinci, F. Lopes, C. Cummings, C.J. Springer, L. Chesler, C. Jones, Y. Jamin,S.P. RobinsonAnalysis and interpretation of data (e.g., statistical analysis, biostatistics,computational analysis): J. Li, K. Zormpas-Petridis, A. Heindl, C. Cummings,J.C. Bamber, Y. Yuan, R. Sinkus, Y. JaminWriting, review, and/or revision of the manuscript: J. Li, K. Zormpas-Petridis,J.K.R. Boult, E.L. Reeves,M. Vinci, C.J. Springer, L. Chesler, J.C. Bamber, Y. Yuan,R. Sinkus, Y. Jamin, S.P. RobinsonAdministrative, technical, or material support (i.e., reporting or organizingdata, constructing databases): J. Li, K. Zormpas-Petridis, F. Lopes, L. Chesler,Y. JaminStudy supervision: F. Lopes, J.C. Bamber, Y. Jamin, S.P. Robinson

AcknowledgmentsEuropeanUnionHorizon 2020Research and InnovationProgramme (Grant

No. 668039), Cancer Research UK and EPSRC support to the Cancer ImagingCentre at ICR, in associationwith theMRCandDepartment ofHealth (England;Grant Nos. C1060/A10334 and C1060/A16464), The Rosetrees Trust (M593),support to S.P. Robinson from Cancer Research UK Programme Grant(No. C16412/A27725), support from the Cancer Research UK Centre at theICR, support to C.J. Springer from Wellcome Trust (Grant Nos. WT1005Xand 100282/Z/12/Z) and CTU at ICR (Grant No. C309/A11566), support toL. Chesler from Cancer Research UK Programme Grants (C34648/A18339 andC34648/A14610). Y. Jamin andM. Vinci are Children with Cancer UK ResearchFellows (2014/176 and 2016/234).

The costs of publication of this article were defrayed in part by thepayment of page charges. This article must therefore be hereby markedadvertisement in accordance with 18 U.S.C. Section 1734 solely to indicatethis fact.

Received June 10, 2019; revised August 15, 2019; accepted September 17,2019; published first October 11, 2019.

References1. Northey JJ, Przybyla L, Weaver VM. Tissue force programs cell fate and

tumor aggression. Cancer Discov 2017;7:1224–37.2. Northcott JM, Dean IS, Mouw JK,Weaver VM. Feeling stress: themechanics

of cancer progression and aggression. Front Cell Dev Biol 2018;6.3. Dewhirst MW, Secomb TW. Transport of drugs from blood vessels to

tumour tissue. Nat Rev Cancer 2017;17:738–50.4. Kota J, Hancock J, Kwon J, KorcM. Pancreatic cancer: Stroma and its current

and emerging targeted therapies. Cancer Lett 2017;391:38–49.5. Conklin MW, Keely PJ. Why the stroma matters in breast cancer. Cell Adh

Migr 2012;6:249–60.6. Provenzano PP, InmanDR, Eliceiri KW, Knittel JG, Yan L, Rueden CT, et al.

Collagen density promotes mammary tumor initiation and progression.BMC Med 2008;6:11.

7. Egeblad M, Rasch MG, Weaver VM. Dynamic interplay between thecollagen scaffold and tumor evolution. Curr Opin Cell Biol 2010;22:697–706.

8. Fang M, Yuan J, Peng C, Li Y. Collagen as a double-edged sword in tumorprogression. Tumour Biol 2014;35:2871–82.

9. Levental KR, Yu H, Kass L, Lakins JN, Egeblad M, Erler JT, et al. Matrixcrosslinking forces tumor progression by enhancing integrin signaling. Cell2009;139:891–906.

10. Liu J, Liao S, Diop-Frimpong B, Chen W, Goel S, Naxerova K, et al. TGF-bblockade improves the distribution and efficacy of therapeutics in breastcarcinoma by normalizing the tumor stroma. Proc Natl Acad Sci U S A2012;109:16618–23.

11. Valkenburg KC, de Groot AE, Pienta KJ. Targeting the tumour stroma toimprove cancer therapy. Nat Rev Clin Oncol 2018;15:366–81.

12. Dolor A, Szoka FC Jr. Digesting a path forward: the utility of collagenasetumor treatment for improved drug delivery. Mol Pharm 2018;15:2069–83.

13. TangH, Leung L, SaturnoG, Viros A, SmithD,Di LevaG, et al. Lysyl oxidasedrives tumour progression by trapping EGF receptors at the cell surface.Nat Commun 2017;8:14909.

14. Mitchell MJ, Jain RK, Langer R. Engineering and physical sciences inoncology: challenges and opportunities. Nat Rev Cancer 2017;17:659–75.

15. Narunsky L, Oren R, Bochner F, Neeman M. Imaging aspects of the tumorstroma with therapeutic implications. Pharmacol Ther 2014;141:192–208.

16. Riegler J, Labyed Y, Rosenzweig S, Javinal V, Castiglioni A, Dominguez CX,et al. Tumor elastography and its association with collagen and the tumormicroenvironment. Clin Cancer Res 2018;24:4455–67.

Li et al.





17. Wang H, Mislati R, Ahmed R, Vincent P, Nwabunwanne SF, Gunn JR, et al.Elastography can map the local inverse relationship between shear mod-ulus and drug delivery within the pancreatic ductal adenocarcinomamicroenvironment. Clin Cancer Res 2019;5:2136–43.

18. Li J, Jamin Y, Boult JKR, Cummings C, Waterton JC, Ulloa J, et al. Tumourbiomechanical response to the vascular disrupting agent ZD6126 in vivoassessed by magnetic resonance elastography. Br J Cancer 2014;110:1727–32.

19. Jamin Y, Boult JKR, Li J, Popov S, Garteiser P, Ulloa JL, et al. Exploring thebiomechanical properties of brain malignancies and their pathologicdeterminants in vivo with magnetic resonance elastography. Cancer Res2015;75:1216–24.

20. Jug�e L, Doan B-T, Seguin J, AlbuquerqueM, Larrat B,Mignet N, et al. Colontumor growth and antivascular treatment in mice: complementary assess-ment with MR elastography and diffusion-weighted MR imaging. Radiol-ogy 2012;264:436–44.

21. PepinKM,Chen J, Glaser KJ,Mariappan YK, Reuland B, Ziesmer S, et al.MRelastography derived shear stiffness—a new imaging biomarker for theassessment of early tumor response to chemotherapy. Magn Reson Med2014;71:1834–40.

22. Feng Y, Clayton EH, Okamoto RJ, Engelbach J, Bayly PV, GarbowJR. A longitudinal magnetic resonance elastography study ofmurine brain tumors following radiation therapy. Phys Med Biol2016;61:6121.

23. Garteiser P, Doblas S, Daire J-L, Wagner M, Leitao H, Vilgrain V, et al. MRelastography of liver tumours: value of viscoelastic properties for tumourcharacterisation. Eur Radiol 2012;22:2169–77.

24. Reiss-ZimmermannM, Streitberger K-J, Sack I, Braun J, Arlt F, Fritzsch D,et al. High resolution imaging of viscoelastic properties of intracranialtumours by multi-frequency magnetic resonance elastography.Clin Neuroradiol 2015;25:371–8.

25. Bohte AE, Nelissen JL, Runge JH, Holub O, Lambert SA, de Graaf L, et al.Breast magnetic resonance elastography: a review of clinical work andfuture perspectives. NMR Biomed 2018;31:e3932.

26. O'Connor JPB, Aboagye EO, Adams JE, Aerts HJWL, Barrington SF, Beer AJ,et al. Imaging biomarker roadmap for cancer studies. Nat Rev Clin Oncol2017;14:169–86.

27. Zormpas-Petridis K, Jerome NP, Blackledge M, Carceller F, Poon E, ClarkeM, et al. MRI imaging of the hemodynamic vasculature of neuroblastomapredicts response to antiangiogenic treatment. Cancer Res 2019;79:2978–91.

28. Boult JKR, Jamin Y, Jacobs V, Gilmour LD, Walker-Samuel S, Halliday J,et al. False-negativeMRI biomarkers of tumour response to targeted cancertherapeutics. Br J Cancer 2012;106:1960–6.

29. Workman P, Aboagye EO, Balkwill F, Balmain A, Bruder G, Chaplin DJ,et al. Guidelines for the welfare and use of animals in cancer research. Br JCancer 2010;102:1555–77.

30. Kilkenny C, Browne WJ, Cuthill IC, Emerson M, Altman DG. Improvingbioscience research reporting: the ARRIVE guidelines for reporting animalresearch. PLoS Biol 2010;8:e1000412.

31. Hill RM, Kuijper S, Lindsey JC, Petrie K, Schwalbe EC, Barker K, et al.Combined MYC and P53 defects emerge at medulloblastoma relapse anddefine rapidly progressive, therapeutically targetable disease. Cancer Cell2015;27:72–84.

32. Sinkus R, Siegmann K, Xydeas T, Tanter M, Claussen C, Fink M. MRelastography of breast lesions: understanding the solid/liquid duality canimprove the specificity of contrast-enhanced MR mammography.Magn Reson Med 2007;58:1135–44.

33. PauG, Fuchs F, Sklyar O, BoutrosM,HuberW. EBImage—an R package forimage processing with applications to cellular phenotypes. Bioinformatics2010;26:979–81.

34. Yuan Y, Failmezger H, Rueda OM, Ali HR, Gr€af S, Chin S-F, et al. Quan-titative image analysis of cellular heterogeneity in breast tumors comple-ments genomic profiling. Sci Transl Med 2012;4:157ra43–ra43.

35. Nieskoski MD, Marra K, Gunn JR, Hoopes PJ, Doyley MM, Hasan T, et al.Collagen complexity spatially definesmicroregions of total tissue pressurein pancreatic cancer. Sci Rep 2017;7:10093.

36. Motulsky HJ, Brown RE. Detecting outliers when fitting data with nonlin-ear regression - anewmethodbasedon robust nonlinear regression and thefalse discovery rate. BMC Bioinformatics 2006;7:123.

37. Acerbi I, Cassereau L,Dean I, ShiQ,AuA, ParkC, et al.Humanbreast cancerinvasion and aggression correlates with ECM stiffening and immune cellinfiltration. Integr Biol 2015;7:1120–34.

38. Armstrong T, PackhamG,Murphy LB, BatemanAC,Conti JA, FineDR, et al.Type I collagen promotes the malignant phenotype of pancreatic ductaladenocarcinoma. Clin Cancer Res 2004;10:7427–37.

39. Schedin P, Keely PJ. Mammary gland ECM remodeling, stiffness, andmechanosignaling in normal development and tumor progression.Cold Spring Harb Perspect Biol 2011;3:a003228.

40. Whatcott CJ, Diep CH, Jiang P, Watanabe A, LoBello J, Sima C, et al.Desmoplasia in primary tumors and metastatic lesions of pancreaticcancer. Clin Cancer Res 2015;21:3561–8.

41. Yu H, Mouw JK, Weaver VM. Forcing form and function: biomechanicalregulation of tumor evolution. Trends Cell Biol 2011;21:47–56.

42. Tadeo I, Berbegall AP, Castel V, Garcia-Miguel P, Callaghan R, Pahlman S,et al. Extracellular matrix composition defines an ultra-high-risk group ofneuroblastoma within the high-risk patient cohort. Br J Cancer 2016;115:480–9.

43. de Schellenberger AA, Bergs J, Sack I, Taupitz M. The extracellular matrixas a target for biophysical and molecular magnetic resonance imaging.Quantification of biophysical parameters in medical imaging. Cham,Switzerland: Springer, 2018. p123–50.

44. Pepin K, McGee K, Arani A, Lake D, Glaser K, Manduca A, et al. MRelastography analysis of glioma stiffness and IDH1-mutation status. Am JNeuroradiol 2018;39:31–6.

45. Simon M, Guo J, Papazoglou S, Scholand-Engler H, Erdmann C, MelchertU, et al. Non-invasive characterization of intracranial tumors by magneticresonance elastography. New J Phys 2013;15:085024.

46. Han W, Chen S, Yuan W, Fan Q, Tian J, Wang X, et al. Oriented collagenfibers direct tumor cell intravasation. Proc Natl Acad Sci U S A 2016;113:11208–13.

47. Braekeveldt N, Wigerup C, Tadeo I, Beckman S, Sanden C, Jonsson J, et al.Neuroblastoma patient-derived orthotopic xenografts reflect the microen-vironmental hallmarks of aggressive patient tumours. Cancer Lett 2016;375:384–9.

48. Tadeo I, PiquerasM,MontanerD, Villamon E, Berbegall AP, Canete A, et al.Quantitative modeling of clinical, cellular, and extracellular matrix vari-ables suggest prognostic indicators in cancer: a model in neuroblastoma.Pediatr Res 2014;75:302–14.

49. Zhu S, ZhangX,Weichert-LeaheyN,DongZ, ZhangC, LopezG, et al. LMO1synergizes with MYCN to promote neuroblastoma initiation and metas-tasis. Cancer Cell 2017;32:310–23.e5.

50. Eikenes L, Bruland OS, Brekken C, Davies Cde L. Collagenase increases thetranscapillary pressure gradient and improves the uptake and distributionof monoclonal antibodies in human osteosarcoma xenografts. Cancer Res2004;64:4768–73.

51. Hassid Y, Eyal E, Margalit R, Furman-Haran E, Degani H. Non-invasiveimaging of barriers to drug delivery in tumors. Microvasc Res 2008;76:94–103.

52. Murphy JE,Wo JY-L, RyanDP, JiangW,YeapBY,DudaGD, et al. Potentiallycurative combination of TGF-b1 inhibitor losartan and FOLFIRINOX(FFX) for locally advanced pancreatic cancer (LAPC): R0 resection ratesand preliminary survival data from a prospective phase II study. J ClinOncol 2018;36:4116.

53. Nieskoski MD, Marra K, Gunn JR, Kanick SC, Doyley MM, Hasan T, et al.Separation of solid stress from interstitial fluid pressure in pancreas cancercorrelates with collagen area fraction. J Biomech Eng 2017;139. doi:10.1115/1.4036392.

54. Nia HT, Liu H, Seano G, Datta M, Jones D, Rahbari N, et al. Solidstress and elastic energy as measures of tumour mechanopathology.Nat Biomed Eng 2016;1. doi: 10.1038/s41551-016-0004.

55. Walker-Samuel S, Roberts TA, Ramasawmy R, Burrell JS, Johnson SP, SiowBM, et al. Investigating low-velocity fluid flow in tumors with convection-MRI. Cancer Res 2018;78:1859–72.

56. Kennedy P, Wagner M, Castera L, Hong CW, Johnson CL, Sirlin CB, et al.Quantitative elastography methods in liver disease: current evidence andfuture directions. Radiology 2018;286:738–63.

57. Patz S, Fovargue D, Schregel K, Nazari N, Palotai M, Barbone PE, et al.Imaging localized neuronal activity at fast time scales through biome-chanics. Sci Adv 2019;5:eaav3816.






2019;79:5874-5883. Published OnlineFirst October 11, 2019.Cancer Res Jin Li, Konstantinos Zormpas-Petridis, Jessica K.R. Boult, et al. ElastographyBiomechanical Phenotype with Noninvasive Magnetic Resonance Investigating the Contribution of Collagen to the Tumor

Updated version

10.1158/0008-5472.CAN-19-1595doi:

Access the most recent version of this article at:

Material

Supplementary

http://cancerres.aacrjournals.org/content/suppl/2019/10/10/0008-5472.CAN-19-1595.DC1

Access the most recent supplemental material at:

Cited articles

http://cancerres.aacrjournals.org/content/79/22/5874.full#ref-list-1

This article cites 51 articles, 12 of which you can access for free at:

E-mail alerts related to this article or journal.Sign up to receive free email-alerts

Subscriptions

Reprints and

[email protected]

To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at

Permissions

Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/79/22/5874To request permission to re-use all or part of this article, use this link



http://cancerres.aacrjournals.org/lookup/doi/10.1158/0008-5472.CAN-19-1595

http://cancerres.aacrjournals.org/content/suppl/2019/10/10/0008-5472.CAN-19-1595.DC1

http://cancerres.aacrjournals.org/content/79/22/5874.full#ref-list-1

http://cancerres.aacrjournals.org/cgi/alerts

mailto:[email protected]

http://cancerres.aacrjournals.org/content/79/22/5874


investigating the contribution of collagen to the tumor ... · jin li1, konstantinos zormpas ......

Documents