cancer metabolism and tumor heterogeneity: …downloads.hindawi.com › journals › cmmi › 2017...

Review ArticleCancer Metabolism and Tumor HeterogeneityImaging Perspectives Using MR Imaging and Spectroscopy

Gigin Lin12 Kayvan R Keshari345 and Jae Mo Park678

1Department of Medical Imaging and Intervention Imaging Core Laboratory Institute for Radiological ResearchChang Gung Memorial Hospital at Linkou and Chang Gung University Taoyuan Taiwan2Clinical Phenome Center Chang Gung Memorial Hospital at Linkou Taoyuan Taiwan3Department of Radiology Memorial Sloan Kettering Cancer Center New York NY USA4Molecular Pharmacology Program Memorial Sloan Kettering Cancer Center New York NY USA5Weill Cornell Medical College New York NY USA6Advanced Imaging Research Center University of Texas Southwestern Medical Center Dallas TX USA7Department of Radiology University of Texas Southwestern Medical Center Dallas TX USA8Department of Electrical and Computer Engineering University of Texas at Dallas Richardson TX USA

Correspondence should be addressed to Jae Mo Park jaemoparkutsouthwesternedu

Received 15 April 2017 Revised 31 July 2017 Accepted 27 August 2017 Published 9 October 2017

Academic Editor Gaurav Malviya

Copyright copy 2017 Gigin Lin et alThis is an open access article distributed under theCreative CommonsAttribution License whichpermits unrestricted use distribution and reproduction in any medium provided the original work is properly cited

Cancer cells reprogram their metabolism to maintain viability via genetic mutations and epigenetic alterations expressing overalldynamic heterogeneity The complex relaxation mechanisms of nuclear spins provide unique and convertible tissue contrastsmaking magnetic resonance imaging (MRI) and magnetic resonance spectroscopy (MRS) pertinent imaging tools in both clinicsand research In this review we summarized MR methods that visualize tumor characteristics and its metabolic phenotypes onan anatomical microvascular microstructural microenvironmental and metabolomics scale The review will progress from theutilities of basic spin-relaxation contrasts in cancer imaging to more advanced imaging methods that measure tumor-distinctiveparameters such as perfusion water diffusion magnetic susceptibility oxygenation acidosis redox state and cell death Analyticalmethods to assess tumor heterogeneity are also reviewed in brief Although the clinical utility of tumor heterogeneity from imagingis debatable the quantification of tumor heterogeneity using functional and metabolic MR images with development of robustanalytical methods and improved MR methods may offer more critical roles of tumor heterogeneity data in clinics MRIMRScan also provide insightful information on pharmacometabolomics biomarker discovery disease diagnosis and prognosis andtreatment response With these future directions in mind we anticipate the widespread utilization of these MR-based techniquesin studying in vivo cancer biology to better address significant clinical needs

1 Cancer Metabolism and MR

Cancer cells by definition are highly proliferative and growrapidly Tumors adapt their metabolism to maintain viabilitywhich is one of the emerging hallmarks of cancer [1] Thecommon metabolic alterations include increased glucoseuptake and lactate production decreased mitochondrialactivity modulated bioenergetic status and aberrant phos-pholipid metabolism accompanied by significant changes inthe tumor microenvironment and structural malformation

in the tumor mass cellular microstructure and surroundingvascular networks

Knowledge of metabolic patterns in cancer can be imple-mented not only for early detection and diagnosis of cancerbut also in the evaluation of tumor response to medicalinterventions and therapies [2] Many targeted therapiesalter cancer metabolism and the changes in endogenousmetabolites in cancer cells may be detectable before changesin tumor sizes [3ndash5] The noninvasive nature of imagingmethods is ideal for detecting early metabolic changes in

HindawiContrast Media amp Molecular ImagingVolume 2017 Article ID 6053879 18 pageshttpsdoiorg10115520176053879

2 Contrast Media amp Molecular Imaging

Microvasculatureperfusion permeability

Gross anatomy amptissue characteristics

morphology spin relaxation

Microstructurediusion elasticity

Blood vessel

（+

（+

（+

2

2

2

Microenvironmenthypoxia acidosis redox

Metabolismbioenergetics redox

Glucose

LactateGlycolysis

Pyruvate

TCAGlutamine

Figure 1MRI andMRS in imaging cancerMRI andMRS provide useful information of cancermetabolism and tumor heterogeneity rangingfrom anatomical change tomicrovascular development biophysical characteristicsmicrostructural deformation altered cellularmetabolismand tumor microenvironment

cancer following treatment which could be useful readoutsfor monitoring response to therapies [6 7]

Ideal utilization of molecular imaging is to ldquodose paintrdquothe radiotherapy dose administered to each tumor withreference to positron emission tomography (PET) [8] andto identify the geographic subregions that drive response totherapy subsequent resistance and relapse during treatmentfailure [9 10] However further work has shown that theinterplay between abnormal metabolism vascularizationand hypoxia expression in tumors may lead to different mapsof abnormality depending on the functional pathophysio-logical readout (ie perfusion hypoxia glucose metabolismetc) [11] In order to select optimal imaging paradigms toguide treatment a deeper understanding of the underlyingbiological mechanisms is critical There is a strong rationalefor investigating whether hypoxic regions should be treatedwith differing radiation doses to well-oxygenated tumorsas well as investigating regional variation based on func-tional and molecular imaging This idea represents a majorparadigm shift where images will be composed of arrays ofdata arranged spatially in individual voxels [10] Each voxel is

a cube of data which summarizes a particular morphologicmetabolic or physiologic signal over a volume of around025ndash5mm [12] depending on modality and subject (animalor human)

Complex cancer metabolism and associated characteris-tics have been extensively explored by magnetic resonanceimaging (MRI) and spectroscopy (MRS) using the versatilerelaxation mechanisms of nuclear spins that provide uniqueand convertible tissue contrasts Advances in MR techniqueshave enabled noninvasive access to significant amounts ofuseful information on cancer metabolism and tumor hetero-geneity ranging in spatial scales from gross anatomy bio-physical characteristics and functional or metabolic imaging(Figure 1) It is important to appreciate that these abundantparameters can be extracted from a single acquisition toprovide general structural data (eg size) functional patho-physiological data (eg average bloodflowandpermeability)and various heterogeneity-based metrics in the tumor In thefollowing sections imaging techniques for evaluating cancermetabolism and tumor heterogeneity will be reviewed on avariety of scales

Contrast Media amp Molecular Imaging 3

2 Imaging Morphological Changes andTissue Characteristics

Themorphological and tissue characteristics of conventionalanatomic MRI are based on mixtures of two distinct contrastmechanisms1198791 and1198792 Longitudinal relaxation (1198791) relies ona dipole-dipole interaction of adjacent spins specifying howfast the longitudinal magnetization is recovered Transverserelaxation (1198792) refers to the decay rate of transverse magneti-zation due to the progressive dephasing of excited spins [20]While initial efforts to quantify tissue 1198791 and 1198792 were done byDamadian et al in the early 1970s [21 22] current clinicalcancer diagnosis and monitoring are heavily reliant onqualitative measurements (eg 1198791- or 1198792-weighted images)

Although longer 1198791 values were reported in varioustumors as compared to normal tissues [22] tumor lesionswith high-fat content (eg lipomas) or a high fibrous content(eg breast cancer) have shorter 1198791 values [23] Due to thecomplication and insufficient intrinsic contrast 1198791-weightedimaging is mostly used with 1198791-shortening macromoleculessuch as gadolinium (Gd) chelates that can be delivered intotumor stroma through the surrounding expanded vessels andcapillaries [24]This contrast-enhanced1198791-weighted imagingunderpins much of the clinically relevant MRI

In contrast most cancerous tissues typically have signif-icantly longer transverse relaxation rates (1198792) than normalsoft tissues without having any exogenous contrast agent 1198792-weighted imaging therefore offers a powerful method fordelineation of the tumor Moreover the difference in mag-netic susceptibility (120594) of tumors and normal tissues accel-erates intravoxel dephasing of transverse magnetization intumor and creates off-resonance effects or119879lowast2 contrast a com-bination of spin-spin relaxation (1198792) and 1198610 magnetic fieldinhomogeneity [25]

3 Imaging Microvasculature

Angiogenesis is a key element in the progression of cancerfor both proliferation and metastasis by adequately sup-plying oxygen and nutrients to the tumor sites [26 27]As mentioned earlier 1198791-weighted imaging with exogenouscontrast agents can demonstrate the relative vascularity oftumor masses as the bolus of contrast agent passes throughthe microvasculature Dynamic contrast-enhanced- (DCE-)MRI is a 1198791-weighted sequence that detects an increasein signal intensity proportional to contrast concentrationand can measure perfusion in the tissue microstructure bytracking the first pass of the injected contrast agent witha kinetic tracer model [28 29] Several pharmacokineticmodels have been proposed to extract kinetic parameters [30]such as 119870trans (volume transfer coefficient) and V119890 (extracel-lular volume ratio) that describe tissue vasculature perfusionand permeability On the other hand dynamic susceptibilitycontrast- (DSC-) MRI exploits the changes in local suscep-tibility (119879lowast2 ) of the injected contrast agents resulting in adecrease of signal intensity in areas of higher contrast con-centration

Arterial spin labeling (ASL) technique offers sim-ilar information as conventional dynamic susceptibility

sequences without having any contrast agent by introducingan endogenous tracer in the form of proximally saturatedspins [31] Tumor angiogenesis and the tumor grade canbe measured by kinetic analysis of perfusion imaging usingparameters such as tumor blood flow and tumor blood vol-ume as well as mean transit time [32] Moreover tumor vas-cular permeability and perfusion are reported as biomarkersof understanding pharmacokinetics and assessing treatmentresponse to several anti-cancer treatments including anti-vascular endothelial growth factor (VEGF) treatment (Fig-ure 2) and radiotherapy [13 33ndash36]

4 Imaging Microstructure

Rapid proliferation and change in the morphology of tumorsresults in a transformation of endogenous cell-architecturesuch as cell density membranes sizes and fluid poolsleading to altered molecular water diffusion Diffusion-weighted MR imaging (DWI) is a noninvasive measurementof water diffusivity that reflects the cell architecture Withincreasing cell density the confining effect of membranesincreases and thus tumors typically have lower signal onapparent diffusion constant (ADC) maps than healthy cellsdue to restricted water diffusion ADC captures fluid volumechanges in the intra- and extracellular compartments andthe literature reports an inverse relationship between ADCvalues and tumor grade [37] Intravoxel incoherent motion(IVIM) analysis allows for the separation of diffusion andperfusion parameters from diffusion weighted imaging withmulti 119887-values by compartmentalizing fast and slow movingspins [38] Although the efficacy of IVIM in cancer imagingstill needs further verification recent imaging studies havereported promising utilities of IVIM in characterizing varioustumor types [39 40] and assessing therapeutic effects [41ndash43] As compared to the Gaussian diffusion model that relieson monoexponential analysis (eg ADC of DWI) diffusionkurtosis imaging (DKI) captures non-Gaussian factor ofwater diffusion which becomes prominent with larger 119887-values by including an excess kurtosis term (119870) in additionto the Gaussian apparent diffusion coefficient (119863) [44 45]Studies have reported that DKI can assess tumor grade andtreatment response [46 47] as well as improving diagnosticaccuracy [14 48ndash50] Tamura et al for example showed sig-nificantly higher119870 and lower119863 in prostate cancer than non-stromal benign prostatic hyperplasia (BPH) implying moreimpediments to normal diffusion and greater complexity intissue microstructure in the tumors (Figure 3) [14]

Diffusion tensor imaging (DTI) assesses changes inmicrostructural anisotropy from water diffusion by applyingseveral directional diffusion gradients Measuring fractionalanisotropy (FA) from DTI can detect blockage of the ductsand lobules by cancer cells in the breast which increasesthe extracellular tortuosity and restriction of the watermovement causing a reduction of the diffusion coefficients inall directions and consequently also decreasing the diffusionanisotropy [51] DTI also predicts tumor infiltration andanisotropic pathways of cancer invasion [52 53] and FAmaps are associated with the diagnostic utility in glioma [54]pancreatic cancer [55] breast cancer [56] prostate cancer


Data

0

50

100

150

200

250

300

350

400Si

gnal

inte

nsity

0 2 3 4 5 61

Time (min)

(a)

DataModel fit

0

055

010

015

020

025

030

035

040

Gad

olin

ium

(nM

olk

g)

1 2 3 4 5 60

Time (min)

(b)

1 2 3 4 5 60

Time (min)

0

1

2

3

4

5

6

7

Cp(t)

(nM

ol)

(c)

08

0

(ＧＣＨ

minus1)

(d)08

0

(ＧＣＨ

minus1)

(e)minus50

minus40

minus30

minus20

minus10

0

10

Chan

ge in

KＮＬＨ

Ｍ(

)

2 4 6 8 10minus2

Time (days)

(f)

Figure 2 Perfusion monitoring using dynamic contrast-enhanced (DCE)MRI in liver metastasis after anti-VEGF treatment (bevacizumab)DCE-MRI requires the acquisition of (a) time series of signal intensity data converted into (b) a gadolinium contrast agent concentrationndashtimecurve and (c) an arterial input function 119862119901(119905) (d) Model-fitting enables calculation of bulk transfer coefficient (119870trans) and a 119870trans map ofextensive liver metastasis overlaid on a slice from 1198792-weighted MRI (e) 119870trans reduced three days after treatment with bevacizumab (f)Proportion of decrease in 119870trans over time (adapted from [13])

[57] and hepatocellular carcinoma [58] Susceptibility tensorimaging (STI) is a new imaging technique that measuresstructural anisotropy using directional field perturbationbetween tissues with magnetic susceptibility difference andapplied magnetic field [50] However repeated 119879lowast2 -weightedMR acquisitions with various subject orientations along the1198610magnetic field are required for the directional informationhaving STI impractical for clinical use

Mechanical properties related to the structures are alsosignificantly altered in cancer The transformation of cellarchitecture in malignant tumors changes their mechanicalproperties as a much stiffer structure during the pathophys-iological processes of malignancy in aggressive cancer [59]Tumor-associated fibroblasts are one of the most abundantstromal cell types in different carcinomas and are comprisedof a heterogeneous cell population [60] MR elastography


(a) (b) (c)

28

0

(d)

30

0

(e)

Figure 3 Non-Gaussian water diffusion analysis using diffusion kurtosis imaging (DKI) in prostate cancer A 73-year-old man (prostate-specific antigen level 121 ngmL) with prostate cancer (arrows) (a) 1198792-weighted image (b) diffusion-weighted image (119887 = 1500 smm2)(c) apparent diffusion coefficient (ADC) map (d) diffusivity map and (e) kurtosis map Compared with healthy tissue prostate cancerin left peripheral zone (indicated by an arrow) showed hypointensity on 1198792-weighted image hyperintensity on diffusion-weighted imagehypointensity on ADC lower diffusivity and higher kurtosis (adapted from [14])

(MRE biomarker stiffness) measures the viscoelasticity ofsoft tissues in vivo by introducing shear waves and imagingtheir propagation using MRI [15 61] The role of MRE in theevaluation of malignant tumors has been tested in variouscancer types including breast cancer (Figure 4) [62ndash64]brain tumor [65] hepatocellular carcinoma [66] and prostatecancer [67]

5 Imaging Tumor Microenvironment CellularFunction and Metabolism

Pathological tumormicroenvironment represented by insuf-ficient oxygenation (hypoxia) [68] and tissue acidosis [69]is known to contribute to tumor progression and treat-ment resistance Hypoxia and acidosis affect the balance ofreducingoxidizing species These changes in the aberranttissue redox state can impact biological cellular statuses suchas cell proliferationdifferentiation and necrosisapoptosis[70 71] Normalization of the tumor microenvironment isconsidered a therapeutic strategy [72] and a series of imagingtechnologies have been developed to unravel the hostiletumor microenvironment

51 Oxygenation Oxygen is often a limiting resource in thetumor microenvironment Since the tissue oxygen level isdependent on the transportation of red blood cells cellsthat are distant from well-perfused capillaries will be under

hypoxic conditions despite still being supplied with glucose[73]This gradient can cause a hypoxic environment at almost60 of the cancer cells [74] limiting oxidative phosphoryla-tion and promoting the growth of cancer cells Therefore asignificant number of studies have been done to image tumorperfusion and tumor hypoxia yielding information about thephysiological status of the tumormicroenvironment Cancer-associated fibroblasts on the other hand suffer from hypoxiato a less severe extent [75]

Thehypoxicmicroenvironment of tumors can be assessedusing several MR methods that exploit either endogenouscontrast mechanisms or exogenous contrast agents [7677] Due to the differential magnetic susceptibility betweendeoxy-hemoglobin and oxy-hemoglobin 119879lowast2 and 119879

10158402 (1119879

10158402 =

1119879lowast2 minus 11198792) can be used to estimate blood and tissueoxygenation [78 79] The methods however are heavilydependent on magnetic field inhomogeneities and suscep-tible to possible errors in the correction of macroscopicinhomogeneities of the static field (1198610) Quantitative suscep-tibility mapping (QSM) is a more advanced post-processingtechnique that calculates quantitative susceptibility (120594) fromthe perturbed magnetic field map and has been shown tomeasure oxygen saturation (SvO2) along cerebral venousvasculature [80 81] Although these susceptibility-weightedimaging techniques demonstrate the unique potential formapping blood depositions and tumoral neovascularity inbrain tumors [82 83] the utility is so far focused on


(a) (b)

(c) (d)

50

25

(kPa

)

0

minus20

0

20

(m

)Figure 4 Shear stiffness assessment of breast cancer using MR elastography (MRE) (a) An axial MR magnitude image of the right breast ofa patient volunteer A large adenocarcinoma is shown as the outlined mildly hyperintense region on the lateral side of the breast (b) A singlewave image from MRE performed at 100Hz is shown along with (c) the corresponding elastogram (d) An overlay image of the elastogramand the magnitude image shows good correlation between the tumor and the stiff region detected by MRE (adapted from [15])

venous oxygenation and thus limited by spatial resolutionSimilarly 119879lowast2 -based blood oxygen level dependent (BOLD)functional MRI can detect changes in oxygenation in thevascular compartment but has limitations in quantitativerelationships between response signal intensity and changesin tumor tissue pO2 [84 85] Oxygen-enhanced MRI isa recently proposed imaging method that detects the 1198791-shortening as a function of tissue oxygen concentrations [86ndash88] The oxygen-enhanced MRI has the potential to providenoninvasive measurements of changes in the oxygen levelof tissue as an addition to BOLD imaging The techniquehowever is often hampered by insufficient sensitivity and the1198791 contrast may be affected by other factors in the tissuesuch as an alteration in blood flow and the H2O content ofthe tissue [88] A more quantitative tumor pO2 can be mea-sured by MR methods that use exogenous contrast agentselectron paramagnetic resonance imaging (EPRI) [89 90]

and Overhauser-enhanced MRI (OMRI) [16 91] Howeveran injection of free radical (trityl OX63) is required beforethe imaging (Figure 5)

52 Acidosis Glycolytic metabolism and the hypoxicmicroenvironment lead to extracellular acidosis in solidtumors [69] The acidification typically starts from thecenter of the tumor mass where vascular perfusion is poorThe cells at the center of the tumor mass adapt to the newacidic environment which can then stimulate invasion andmetastasis [92] Therefore the acidic pH distribution is oftenused to describe the tumor progression and the hostilityof tumor microenvironment Dissolution dynamic nuclearpolarization (DNP) provides an unprecedented opportunityto investigate cellular metabolism in vivo by polarizing MRdetectable substrates (eg 13C-labeled substrates) achievinga dramatic signal enhancement which facilitates in vivo


SCC tumor

Anatomy KＮＬＨＭ map (Gd-DTPA)

KＮＬＨＭ map (OX63) p2 map (mmHg)

(a) (b)

(c) (d)

(e) (f)

04

03

02

01

00

40

30

20

10

0

012

009

006

003

000

Figure 5 Tumor oxygenation andmicrovascular permeability using Overhauser-enhancedMRI (OMRI) Comparison of119870trans maps of Gd-DTPA and OX63 (radical) in a squamous cell carcinoma (SCC) (a) SCC tumor region can be detected in a 1198792-weighted image by using 7-TMRI (b)119870trans map ofGd-DTPA (c)119870trans map ofOX63 usingOMRI of the same SCC tumor NoteOMRIOX63 images were obtained beforethe 7-TMRIGd-DTPA study (d) Corresponding pO2 map computed from the same OMRI images for119870trans OX63 map ((e) (f)) Based on theanatomical image ROI of SCC tumor was selected and enlarged Tumor region with low 119870trans OX63 values (ROI 1) was relatively oxygenatedand normal muscle tissue and the region with high 119870trans OX63 values (ROI 2) coincided with hypoxia in pO2 image (adapted from [16])


pH80

70

60

(a)

pH80

70

60

(b)

pH80

70

60

(c)

Figure 6 Multi-slice assessment of extracellular pH (pHe) of the tumor kidney and bladder using acidoCESTMRI and exogenous contrastagent iopromide in a mice model of MDA-MB-231 humanmammary carcinoma (a)The tumor showed an average pHe of 674 (b)The pHeincreased from the renal pelvis (654) to the cortex (684) (c) The bladder had a pHe of 63 The region into which the bladder swells afterinjection during infusion could not be fit indicating that the fitting method used is robust against overfitting (adapted from [17])

metabolic imaging [93] Gallagher et al imaged acidicextracellular pH in vivo by measuring the balance ofhyperpolarized 13C-labeled bicarbonate and 13CO2 [94]Following studies have continued to measure extracellularpH via 13C-bicarbonate as well as novel alternative strategieshowever they are limited by short in vivo 1198791 relaxation timesand low polarization [95ndash97] More recently substrates thatchange their chemical shifts with proton binding such as15N-imidazole or 13C-zymonic acid a pyruvate derivativeare suggested as alternative pH sensors that overcome theproblems of short 1198791 and low polarization [98 99] Variouschemical exchange saturation transfer (CEST) approaches arealso available for in vivo pHmapping (Figure 6) [17 100ndash103]

53 Redox-State The hypoxic and acidic tumor microen-vironment affects redox status by elevating reactive oxygenspecies (ROS) production [70 104] In a biological systemthe redox-state can be represented by a multitude of ratiosincluding the [NAD+][NADH] and [NADP+][NADPH]Hyperpolarized 13C-dehydroascorbate and 13C-ascorbic acidare suggested as biomarkers to interrogate the in vivo[NADP+][NADPH] balance as the ratio reflects extra-cellular oxidation of 13C-ascorbic acid and intracellularreduction of 13C-dehydroascorbate [105 106] Intracellular[NAD+][NADH] redox-state can also be estimated from[13C-lactate][13C-pyruvate] ratio after an injection of hyper-polarized 13C-glucose [107 108] but the clinical utilityrequires further investigation due to the fast 1198791 decay of

13C-glucose

54 Bioenergetics Reprogramming of energy metabolismis a fundamental characteristic of cancer The first discov-ered metabolic phenotype was aberrant glycolysis (Warburgeffect) by which energy generation shifts from oxidativephosphorylation to anaerobic glycolysis even under normaloxygen concentrations [109] Anaerobic glycolysis producestwo ATPs per glucose molecule which is less efficient incomparison to the 36 ATPs generated by oxidative phos-phorylation [69 110] Recent evidence shows that when

glucose is limited cancer cells may recapture lactate andconvert it into pyruvate to fuel the tricarboxylic acid (TCA)cycle [111] This change in glycolysis involves alterations inthe regulation of glucose transporters (GLUT) glycolyticenzymes such as hexokinase 2 (HK2) and pyruvate kinaseisozyme M2 (PKM2) lactate dehydrogenases (LDH) andtransporters of lactate (MCT) as well as the downregula-tion or inactivation of pyruvate dehydrogenase (PDH) [112113] Several key oncogenes which drive the developmentand progression of common human cancers are known toregulate glycolysis For example the deregulated activity ofthe serine-threonine kinase Akt has been shown to increaseglucose uptake by cancer cells [114ndash116]The oncogene c-myca transcription factor controls numerous glycolytic genes(eg HK2 enolase and LDH-A) [117 118] Oncogenic Rasan essential protein that controls signaling pathways thatregulate normal cell growth and malignant transformation[119] increases the concentration of an allosteric activatorof phosphofructo-1-kinase fructose-26-bisphosphate thatcatalyzes the phosphorylation of fructose-6-phosphate tofructose-16-bisphosphate [120] Therefore altered glucoseutilization and the associated enzymaticoncogenic activitiescan serve as surrogates for the development of imaging bio-markers and anti-cancer treatments

Upregulated glucose uptake and altered glycolysis havebeen known for decades [109] but non-invasive imagingmethods that provide a true assessment of in vivo bioen-ergetics are still lacking Hyperpolarized [1-13C]pyruvatehas demonstrated increased lactate labeling in tumors [121]and decreasing metabolism to bicarbonate [122] indicatingsuppressed pyruvate flux into mitochondria (biomarkersmetabolite ratios [lactate][pyruvate] or apparent conversionrate 119896pyr-lac [123 124]) The metabolic fate of pyruvate inthe mitochondria has also been explored with hyperpolar-ized [2-13C]pyruvate in a preclinical glioma model as thelabeled carbon is retained in acetyl-CoA and enters the TCAcycle [18] In particular decreased [5-13C]glutamate produc-tion in tumors implies that the metabolic pathway frompyruvate to the TCA cycle is suppressed in the tumor


Pre-DCA Post-DCA Pre-DCA Post-DCA10

0

[2-13]Pyruvate 009

0

[2-13]Lactate

001

0

003

0

[5-13]Glutamate 1（ MRI

Tumor

Surface coil

1 cm

(a) (b)

(c) (d)

Figure 7 Glioma metabolism using hyperpolarized [2-13C]pyruvate before and after dichloroacetate (DCA) administration 125-mMhyperpolarized [2-13C]pyruvate was injected intravenously into a rat with C6 glioma cells Metabolite maps of (a) [2-13C]pyruvate (b) [2-13C]lactate and (c) [5-13C]glutamate from a tumor slice of a representative glioma-implanted rat brain measured pre- and post-DCA (d)Contrast-enhanced 1198791-weighted

1HMRI of the corresponding slice (adapted from [18])

as compared to the contralateral normal-appearing braintissue while glycolytic characteristics of tumors could bestill assessed by increased [2-13C]lactate Using hyperpo-larized [2-13C]pyruvate it was further demonstrated thatdysregulatedmitochondrial metabolism (eg the TCA cycle)is potentially recoverable in glioma by inhibiting pyruvatedehydrogenase kinase (PDK) with dichloroacetate (DCA)(Figure 7) The utility of this technique is further verified formonitoring anti-cancer treatment responses [125ndash129] Glu-tamine addiction another phenotype in bioenergetics oftenfound in multiple cancer models that are less glycolytic [130131] can also be assessed by hyperpolarized 13C substrates[132ndash134]

In addition to these approaches proton (1H) magneticresonance spectroscopic imaging (MRSI) has been a powerfultool for characterizing tumor metabolism by quantifyingcancer-related metabolites such as choline creatine N-acetylaspartate (NAA) and 2-hydroxyglutarate (2HG) Increasedlevels of choline specifically are associated with tumorproliferation with recent studies emphasizing the complexinteractions between choline metabolism and oncogenicsignaling [135] NAA is a neuro-specific metabolite thatdecreases in most brain tumors as neurons are destroyedor displaced by proliferating tumors [136 137] While NAAmetabolism has been primarily studied in the central nervoussystems a recent study discovered cancer-specific productionof NAA via overexpressed NAA synthetase (NAT8L) in non-small cell lung cancers [138] A separate study in ovariancancer patients reported that patients with elevated NAAlevels have worse clinical outcomes [139] suggesting thatthe NAA pathway has a prominent role in promoting tumorgrowth Creatine is another cancer-associatedmetabolite that

is reduced by depleted energy stores due to the highmetabolicactivity of malignant tumors [140] Lactate also appearshigh in tumors when hypoxia-induced anaerobic glycolysisdominates mitochondrial oxidative phosphorylation andoraerobic glycolytic rate is increased [141] Moreover detectionof 2HG can differentiate brain tumors with isocitrate dehy-drogenase (IDH)mutation from tumors with wild-type IDHfor use in selecting patients for targeted therapies and devel-opment of novel therapeutic approaches (Figure 8) [19] Con-ventional 13C-MRS with 13C-enrichedmetabolites could alsobe useful to investigate the utilization of specific metabolicpathways but typically it is not used for imaging since thelimited signal intensity inhibits high spatial resolution encod-ing

55 Cell Death and Necrosis High-grade neoplasms arefrequently heterogeneous and may have central necrosisThe destruction of cell membranes in necrotic brain lesionsallows for virtually unhindered diffusion yielding areas ofhigh ADC [142] It has also been suggested that necrosis canbe detected using hyperpolarized 13C-fumarate and its extra-cellular conversion to 13C-malate via fumarase an enzymethat presents in cytosol and mitochondria [143]

6 Imaging Tumor Heterogeneity

Common tools of cancer research such as DNA sequencinggene and protein expression and metabolomics are basedon biopsy measurements and the assumption of a homoge-nous cell population within a tumor Tumors progressivelyaccumulate geneticmutations and epigenetic alterations [144145] Genetic mutations of cancer cells lead to diversity and


Cr ChoCr

NAA

In vivoFitResiduals

2HG (0)GABA (11 plusmn 02)

Glu (90 plusmn 02)

Gln (24 plusmn 02)

Normal brain

3 2 14

(ppm)

(a)

Grade 4 glioblastoma(IDH12 wild-type)

ChoCr NAA

Lac Lip


Glu (46 plusmn 02)

Gln (53 plusmn 02)

3 2 14

(ppm)

(b)

Grade 2 oligodendroglioma(IDH1 mutated)

Cho

Cr NAA Lac

2HG (33 plusmn 02)

GABA (04 plusmn 02)Glu (14 plusmn 02)

Gln (29 plusmn 02)

3 2 14

(ppm)

(c)


Cho

Cr NAA Lac

2HG (91 plusmn 03)

GABA (02 plusmn 02)

Glu (05 plusmn 02)

Gln (02 plusmn 02)

3 2 14

(ppm)

(d)


Gly

Cho

CrNAA Lac

2HG (89 plusmn 03)

GABA (0)

Glu (21 plusmn 02)

Gln (23 plusmn 02)

3 2 14

(ppm)

(e)

Grade 3 astrocytoma(IDH1 mutated)

ChoCr NAA

Lip

2HG (46 plusmn 03)

GABA (01 plusmn 01)

Glu (14 plusmn 02)

Gln (26 plusmn 02)

3 2 14

(ppm)

(f)

Figure 8 2-Hydroxyglutarate (2HG) detection by MRS in isocitrate dehydrogenase- (IDH-) mutated glioma patients In vivo single-voxellocalized spectra from normal brain (a) and tumors ((b)ndash(f)) at 3 T are shown together with spectral fits (LCModel) and the componentsof 2HG GABA glutamate and glutamine and voxel positioning (2 times 2 times 2 cm3) Spectra are scaled on the water signal from the voxelVertical lines are drawn at 225 ppm to indicate the H4 multiplet of 2HG Shown in brackets is the estimated metabolite concentration (mM)plusmn standard deviation Cho choline Cr creatine NAA N-acetyl aspartate Glu glutamate Gln glutamine GABA 120574-aminobutyric acid Glyglycine Lac lactate Lip lipids Scale bars 1 cm (adapted from [19])

heterogeneity which may favor cooperation for growth [146]andmetastasis [147] Recently the intratumoral heterogeneityand branched evolution have been investigated in renal cellcarcinomas by genome sequencing of multiple spatially sepa-rated samples fromprimary tumors and associatedmetastaticsites [148]Themetabolic heterogeneity is attributed not onlyto genetic alteration but also to the adaptation to the hypoxictumor microenvironment As glycolysis confers a significantgrowth advantage by producing the requiredmacromoleculesas building blocks lactate can be utilized by oxygenatedcancer cells as oxidative fuel [149] to save the glucose forthe more anoxic cells in the center of the tumor [150] Thiscooperation between hypoxic and normoxic tumor cells opti-mizes energy production and allows cells to adapt efficientlyto their environmental oxygen conditions [151 152] With

this in mind there is a considerable research interest toidentify andmeasure both the overall degree of spatial tumorheterogeneity and pinpointing where subpopulations withintumors are responsive to therapy or resistant [10 153 154]

Tumors are versatile and have been described as evolvingecosystems expressing dynamic heterogeneity [155] Forexample tumor pO2 fluctuates over time with a possibilityof rapid (minutes) adaptation to O2 availability via directand post-translational modulation or slow adaptation withchronic or delayed changes involving transcriptional epi-genetic and genetic mechanisms [156 157] Furthermorethe degree of intratumoral heterogeneity tends to increaseas tumors grow [158 159] Specifically the spatial spread oftumors is dependent on temporally evolving neovascular-ization and tissue perfusion [157 160] Microenvironmental


signals of lactate or pH stimulate adjustments in cell behaviorand protein patterns [161] which promotemechanisms of cellmigration [162 163] angiogenesis and immunosuppression[161ndash164] These consequences which are the results of epi-genetic transcriptional translational or post-translationalmechanisms are more or less reversible [163 165 166]

61 Analytical Methods Imaging genuinely reflects spatialheterogeneity in tumors MRI is one of the leading imagingmodalities for quantifying tumor heterogeneity due to itsability to take advantage of multiple tissue contrasts [167]Many analytical methods are proposed for the quantificationof tumor heterogeneity as an imaging biomarker for cancerstaging tumor classification and assessment of treatmentresponses Nonspatial methods such as histogram analysiscan quantify tumor heterogeneity by analyzing a statisticalmetric such as the frequency distributions variance andpercentile values [168ndash170] Due to the accessibility of thenonspatial analytical tools a rapidly increasing number ofstudies have been performed presenting prognostic poten-tials [171 172] Texture analysis extracts the local or regionalspatial signal distribution or ldquotexture featuresrdquo to evaluatethe intratumoral heterogeneity [173ndash175] Fractal analysis isa mathematical model-based texture-analyzing method thatprovides a statistical measure of geometric pattern changeor recognition as a function of scale [176] The fractalbiomarkers derived from DCE-MRI showed a significantcorrelation with therapeutic outcomes [177ndash179] Transform-based methods are also available for analyzing texture infrequency or a spatial domain (eg Fourier Gabor andwavelet transforms) Currently most image-based analyticalstudies focus on the anatomical microvascular [153 180]and microstructural heterogeneities [181 182] To evaluatethe efficacy of MRIMRS parameters for assessing metabolicheterogeneity of tumors other MR methods that capturefunctional or metabolic information should be exploredusing the analytical methods Moreover integrated investi-gations should be performed between MR-based parametricmaps and genomichistopathological data

7 Feasibility of Clinical Translation

Once technically and biologically validated imagingbiomarkers can serve as useful medical research toolsMany of the MR techniques reviewed here have shownexcellent promises as research tools being highly useful inthe development of therapies but the methods that madeclinical impact are few To cross the translational gaps andbecome a clinical decision-making tool the imaging methodand the corresponding biomarker should satisfy a seriesof criteria with considerations of cost effectiveness anddiagnosticpredictive values in patient care

Compared to other imaging modalities the noninvasiveand nonradioactive nature of MR renders it readily transla-tional from bench to bedside and the spatial and temporalinformation has revolutionized the imaging approaches tocancer diagnosis and treatment In addition to 1198791 and 1198792-weighted imaging that are routinely used in the clinicperfusion and diffusion imaging pulse sequences using DCE

DSC ASL and DWI are included in standard clinical MRprotocols for multiple cancers and frequently used in clinicaltrials to report on therapeutic effects [183] In particulara large number of clinical studies regarding the diagnos-tic values of DCE-MRI have been explored resulting inimproved margins of radiotherapy dose delivery and surgicalmargins [30 177] Imaging methods that depend on basicpulse sequences such as DKI (from DWI) SWI and QSM(from119879lowast2 -weighted imaging) can be available by further post-processing without having separate data acquisitions MRE isalready being used in clinics for assessment of chronic liverdiseases and also available for hepatic tumors and breastcancers

Most functional and metabolic MR methods howeverare rarely used in clinics due to low signal sensitivitiesresulting in poor spatial resolution and reproducibility 1HMRS methods for example often suffer from inconsistentquantification and require a long scan time Nonetheless2HG assessment using 1H MRS is expected to play anemerging role in brain tumor imaging in clinics due to itsuniqueness of identifying IDH mutations in vivo [184 185]CEST MRI has shown encouraging results in tumor patients[186 187] and several early phase clinical trials are being per-formed (for more information refer to ClinicalTrialsgov)EPRI and OMRI that require an injection of trityl are usedfor small animals as research tools and significant technicaladvances and further evaluation are needed prior to humanapplications [91 188] DNP-MRS plays an emerging rolein assessing cancer metabolism and tumor heterogeneitywith an increasing number of cancer-specific molecularprobes Despite the transience of hyperpolarized signals andthe long polarization times hyperpolarized 13C MRS usingdissolution DNP is promising in terms of clinical translationThe completion of the first clinical trial for the assessment ofprostate cancer established the feasibility of human hyper-polarized [1-13C]pyruvate studies and illuminated a cleartranslational path for other additional applications [189]Other hyperpolarized substrates however are still underevaluation for toxicity technical feasibility (eg faster 1198791decays and lower polarizations) and biological validation inanimals

8 Concluding Remarks andFuture Perspectives

The interaction between tumor metabolism and cancer biol-ogy is essential for supporting tumor growth and prolongingsurvival during stress [190] and has important implicationsfor the way tumors respond to therapies Recognizing the sig-nificance contemporary oncologic therapeutics have movedforward from cytotoxic treatment to personalized therapiessuch as targeting specific signaling pathways oncogenes ormetabolic enzymes These therapies will potentially lead toa shift of metabolic signature in tumor tissue that could bemonitored by usingMRI andMRS as described in this reviewarticle Inclusion of noninvasive MR methods for biomarkerdevelopment in conjunction with early drug development is


therefore vital to ensure the progression of imaging use intoclinical practice

Although spatial and metabolic heterogeneity of tumoris an important prognostic factor the current clinical utilityof quantifying tumor heterogeneity from imaging is contro-versial [173 191] and requires more robust and standardizedmethods before it can contribute significantly to clinicalpractice Further details of tumor heterogeneity are availablewith increased accuracy as MRIMRS technologies advanceFor example smart 119896-space sampling schemes and parallelimaging methods can lead to accelerated data acquisitionwith a narrower point spread function achieving higherspatial resolution and improved imaging contrastThis wouldbe beneficial for analyzing tumor heterogeneity particularlyin the setting of dynamic imaging or metabolic imagingMore robust assessments of tissue heterogeneity should beavailable by enhancing image integrity at high-resolutions viaimprovements in MR hardware (eg stronger field strengthhigh-order shim coils and more capable gradient coils)It is also critical to understand each data acquisition andreconstruction scheme for proper image analysis and validassessment of associated metabolic parameters For exam-ple some imaging signals from neighboring voxels are notnecessarily entirely independent as seen in advanced MRItechniques where zero-filling techniques are employed tokeep scan times as fast as allowable [192] and this should becontrolled for when defining subregional analysis Ultimatelythe image-based assessment of tumor heterogeneity willrequire multidimensional approaches and therefore shouldbe done systematically A large database can be built by shar-ing existing MR data and patient information between mul-tiple institutions with proper data conversion This recentlyemerging approach named ldquoradiomicsrdquo however should beaccompanied by standardized data acquisition and analyticalmodels The mineable database will accelerate to identifyimage features that depict intratumoral heterogeneity andeventually provide useful decision support in clinics [10 193]

Studies based on a single modality might oversimplifythe dynamics of cancer metabolism into a static descriptionA combination of multi-modal in vivo imaging techniquessuch as the integration of MR and PET (anatomic andfunctional imaging by MRI and metabolic imaging by MRSand PET) would further help in unraveling the molecularcomplexities of cancer metabolism The authors believe thedevelopment of integrated bioinformatics tools would aid inthe handling of spatial temporal and multiparametric datafrom cancer metabolic imaging With these future directionsin mind we anticipate the widespread integration of theseMR-based approaches into the study of cancer biology in vivoto better address significant clinical needs Prospective trialswith well-defined endpoints are encouraged to evaluate thebenefits of these emerging imaging tools in the managementof malignancies

Conflicts of Interest

The authors declare that there are no conflicts of interestregarding the publication of this paper

Acknowledgments

This work was supported by Chang Gung Medical Foun-dation grant CIRPG3E0022 (to Gigin Lin) National Sci-ence Council (Taiwan)MOST 104-2314-B-182A-095-MY3 (toGigin Lin) National Institute of Health (USA) grant P30CA008748 (to Kayvan R Keshari) R01 CA195476 (to KayvanR Keshari) R21 CA212958 (to Kayvan R Keshari) P41EB015908 (to Jae Mo Park) and S10 OD018468 (to Jae MoPark) UT Southwestern Mobility Foundation Center (to JaeMo Park) The Texas Institute of Brain Injury and Repair (toJae Mo Park)

References

[1] D Hanahan and R AWeinberg ldquoHallmarks of cancer the nextgenerationrdquo Cell vol 144 no 5 pp 646ndash674 2011

[2] R Kaddurah-Daouk B S Kristal and R M WeinshilboumldquoMetabolomics a global biochemical approach to drugresponse and diseaserdquo Annual Review of Pharmacology andToxicology vol 48 pp 653ndash683 2008

[3] R L Yauch and J Settleman ldquoRecent advances in pathway-targeted cancer drug therapies emerging from cancer genomeanalysisrdquo Current Opinion in Genetics and Development vol 22no 1 pp 45ndash49 2012

[4] M G Vander Heiden ldquoTargeting cancer metabolism a thera-peutic window opensrdquo Nature Reviews Drug Discovery vol 10no 9 pp 671ndash684 2011

[5] D A Tennant R V Duran and E Gottlieb ldquoTargeting meta-bolic transformation for cancer therapyrdquo Nature Reviews Can-cer vol 10 no 4 pp 267ndash277 2010

[6] J Evelhoch M Garwood D Vigneron et al ldquoExpanding theuse of magnetic resonance in the assessment of tumor responseto therapy workshop reportrdquo Cancer Research vol 65 no 16pp 7041ndash7044 2005

[7] P Workman E O Aboagye Y-L Chung et al ldquoMinimallyinvasive Pharmacokinetic and pharmacodynamic technologiesin hypothesis-testing clinical trials of innovative therapiesrdquoJournal of the National Cancer Institute vol 98 no 9 pp 580ndash598 2006

[8] A J G Even J VanDer Stoep CM L Zegers et al ldquoPET-baseddose painting in non-small cell lung cancer comparing uniformdose escalation with boosting hypoxic and metabolically activesub-volumesrdquo Radiotherapy and Oncology vol 116 no 2 pp281ndash286 2015

[9] R A Gatenby O Grove and R J Gillies ldquoQuantitative imagingin cancer evolution and ecologyrdquo Radiology vol 269 no 1 pp8ndash15 2013

[10] R J Gillies P E Kinahan and H Hricak ldquoRadiomics imagesare more than pictures they are datardquo Radiology vol 278 no 2pp 563ndash577 2016

[11] W van Elmpt CM L Zegers B Reymen et al ldquoMultiparamet-ric imaging of patient and tumour heterogeneity in non-small-cell lung cancer quantification of tumour hypoxia metabolismand perfusionrdquo European Journal of Nuclear Medicine andMolecular Imaging vol 43 no 2 pp 240ndash248 2016

[12] R A Burrell N McGranahan J Bartek and C Swanton ldquoThecauses and consequences of genetic heterogeneity in cancerevolutionrdquo Nature vol 501 no 7467 pp 338ndash345 2013

[13] J P OrsquoConnor A Jackson M-C Asselin D L Buckley G JParker and G C Jayson ldquoQuantitative imaging biomarkers in


the clinical development of targeted therapeutics current andfuture perspectivesrdquoThe Lancet Oncology vol 9 no 8 pp 766ndash776 2008

[14] C Tamura H Shinmoto S Soga et al ldquoDiffusion kurtosisimaging study of prostate cancer preliminary findingsrdquo Journalof Magnetic Resonance Imaging vol 40 no 3 pp 723ndash729 2014

[15] Y K Mariappan K J Glaser and R L Ehman ldquoMagneticresonance elastography a reviewrdquo Clinical Anatomy vol 23 no5 pp 497ndash511 2010

[16] S Matsumoto H Yasui S Batra et al ldquoSimultaneous imag-ing of tumor oxygenation and microvascular permeabilityusing Overhauser enhanced MRIrdquo Proceedings of the NationalAcademy of Sciences of the United States of America vol 106 no42 pp 17898ndash17903 2009

[17] E A Randtke J C Granados C M Howison M D Pageland J Cardenas-Rodrıguez ldquoMultislice CEST MRI improvesthe spatial assessment of tumor pHrdquo Magnetic Resonance inMedicine vol 78 no 1 pp 97ndash106 2017

[18] J M Park S Josan T Jang et al ldquoVolumetric spiral chemicalshift imaging of hyperpolarized [2-13c]pyruvate in a rat c6glioma modelrdquo Magnetic Resonance in Medicine vol 75 no 3pp 973ndash984 2016

[19] C Choi S K Ganji R J DeBerardinis et al ldquo2-Hydroxy-glutarate detection by magnetic resonance spectroscopy inIDH-mutated patients with gliomasrdquo Nature Medicine vol 18no 4 pp 624ndash629 2012

[20] W R Nitz and P Reimer ldquoContrast mechanisms in MRimagingrdquo European Radiology vol 9 no 6 pp 1032ndash1046 1999

[21] R Damadian ldquoTumor detection by nuclear magnetic reso-nancerdquo Science vol 171 no 3976 pp 1151ndash1153 1971

[22] R Damadian K Zaner D Hor and T DiMaio ldquoHuman tumorsdetected by nuclear magnetic resonancerdquo Proceedings of theNational Academy of Sciences of the United States of Americavol 71 no 4 pp 1471ndash1473 1974

[23] C M Gaskin and C A Helms ldquoLipomas lipoma variantsand well-differentiated liposarcomas (atypical lipomas) resultsof MRI evaluations of 126 consecutive fatty massesrdquo AmericanJournal of Roentgenology vol 182 no 3 pp 733ndash739 2004

[24] M J Kransdorf and M D Murphey ldquoRadiologic evaluation ofsoft-tissue masses a current perspectiverdquo American Journal ofRoentgenology vol 175 no 3 pp 575ndash587 2000

[25] G B Chavhan P S Babyn B Thomas M M Shroff and EMark Haacke ldquoPrinciples techniques and applications of T2-based MR imaging and its special applicationsrdquo Radiographicsvol 29 no 5 pp 1433ndash1449 2009

[26] J Folkman ldquoTumor angiogenesis therapeutic implicationsrdquoThe New England Journal of Medicine vol 285 no 21 pp 1182ndash1186 1971

[27] R K Jain E di Tomaso D G Duda J S Loeffler A GSorensen and T T Batchelor ldquoAngiogenesis in brain tumoursrdquoNature Reviews Neuroscience vol 8 no 8 pp 610ndash622 2007

[28] P S Tofts G Brix D L Buckley et al ldquoEstimating kineticparameters from dynamic contrast-enhanced T1-weightedMRIof a diffusable tracer standardized quantities and symbolsrdquoJournal of Magnetic Resonance Imaging vol 10 no 3 pp 223ndash232 1999

[29] B R Rosen J W Belliveau J M Vevea and T J Brady ldquoPerfu-sion imaging with NMR contrast agentsrdquo Magnetic Resonancein Medicine vol 14 no 2 pp 249ndash265 1990

[30] F Khalifa A Soliman A El-Baz et al ldquoModels and methodsfor analyzing DCE-MRI a reviewrdquoMedical Physics vol 41 no12 Article ID 124301 2014

[31] J A Detre J S Leigh D S Williams and A P Koretsky ldquoPer-fusion imagingrdquo Magnetic Resonance in Medicine vol 23 no1 pp 37ndash45 1992

[32] C Warmuth M Gunther and C Zimmer ldquoQuantification ofblood flow in brain tumors Comparison of arterial spin labelingand dynamic susceptibility-weighted contrast-enhanced MRimagingrdquo Radiology vol 228 no 2 pp 523ndash532 2003

[33] S B Wedam J A Low S X Yang et al et al ldquoAntiangiogenicand antitumor effects of bevacizumab in patients with inflam-matory and locally advanced breast cancerrdquo Journal of ClinicalOncology vol 24 pp 769ndash777 2006

[34] B F Jordan M Runquist N Raghunand et al ldquoDynamiccontrast-enhanced and diffusionMRI show rapid and dramaticchanges in tumor microenvironment in response to inhibitionof HIF-1120572 using PX-478rdquo Neoplasia vol 7 no 5 pp 475ndash4852005

[35] M Fuss F Wenz M Essig et al ldquoTumor angiogenesis oflow-grade astrocytomas measured by dynamic susceptibilitycontrast-enhancedMRI (DSC-MRI) is predictive of local tumorcontrol after radiation therapyrdquo International Journal of Radia-tion Oncology Biology Physics vol 51 no 2 pp 478ndash482 2001

[36] S E Heethuis P S N van Rossum I M Lips et al ldquoDynamiccontrast-enhanced MRI for treatment response assessmentin patients with oesophageal cancer receiving neoadjuvantchemoradiotherapyrdquo Radiotherapy and Oncology vol 120 no1 pp 128ndash135 2016

[37] Y Hayashida T Hirai S Morishita et al ldquoDiffusion-weightedimaging ofmetastatic brain tumors comparisonwith histologictype and tumor cellularityrdquoAmerican Journal of Neuroradiologyvol 27 no 7 pp 1419ndash1425 2006

[38] D Le Bihan E Breton D Lallemand M-L Aubin J Vignaudand M Laval-Jeantet ldquoSeparation of diffusion and perfusion inintravoxel incoherent motionMR imagingrdquo Radiology vol 168no 2 pp 497ndash505 1988

[39] Y-D Zhang Q Wang C-J Wu et al ldquoThe histogram analysisof diffusion-weighted intravoxel incoherent motion (IVIM)imaging for differentiating the gleason grade of prostate cancerrdquoEuropean Radiology vol 25 no 4 pp 994ndash1004 2015

[40] M Klauss PMayer KMaier-Hein et al ldquoIVIM-diffusion-MRIfor the differentiation of solid benign andmalign hypervascularliver lesions-evaluation with two different MR scannersrdquo Euro-pean Journal of Radiology vol 85 no 7 pp 1289ndash1294 2016

[41] T Hauser M Essig A Jensen et al ldquoPrediction of treatmentresponse in head and neck carcinomas using IVIM-DWIevaluation of lymph node metastasisrdquo European Journal ofRadiology vol 83 no 5 pp 783ndash787 2014

[42] S Nougaret H A Vargas Y Lakhman et al ldquoIntravoxelincoherentmotion-derived histogrammetrics for assessment ofresponse after combined chemotherapy and radiation therapyin rectal cancer initial experience and comparison betweensingle-section and volumetric analysesrdquo Radiology vol 280 no2 pp 446ndash454 2016

[43] L Zhu L Zhu H Shi et al ldquoEvaluating early response of cervi-cal cancer under concurrent chemo-radiotherapy by intravoxelincoherent motion MR imagingrdquo BMC Cancer vol 16 no 1article 79 2016

[44] J H Jensen J A Helpern A Ramani H Lu and K Kaczyn-ski ldquoDiffusional kurtosis imaging the quantification of non-Gaussian water diffusion by means of magnetic resonanceimagingrdquo Magnetic Resonance in Medicine vol 53 no 6 pp1432ndash1440 2005


[45] A J Steven J Zhuo and E R Melhem ldquoDiffusion kur-tosis imaging an emerging technique for evaluating themicrostructural environment of the brainrdquo American Journal ofRoentgenology vol 202 no 1 pp W26ndashW33 2014

[46] R Jiang J Jiang L Zhao et al ldquoDiffusion kurtosis imaging canefficiently assess the glioma grade and cellular proliferationrdquoOncotarget vol 6 no 39 pp 42380ndash42393 2015

[47] S Goshima M Kanematsu Y Noda H Kondo H Watanabeand K T Bae ldquoDiffusion kurtosis imaging to assess response totreatment in hypervascular hepatocellular carcinomardquo Ameri-can Journal of Roentgenology vol 204 no 5 pp W543ndashW5492015

[48] K Sun X ChenWChai et al ldquoBreast cancer diffusion kurtosismr imaging-diagnostic accuracy and correlation with clinical-pathologic factorsrdquo Radiology vol 277 no 1 pp 46ndash55 2015

[49] A B Rosenkrantz E E Sigmund G Johnson et al ldquoProstatecancer feasibility and preliminary experience of a diffusionalkurtosis model for detection and assessment of aggressivenessof peripheral zone cancerrdquo Radiology vol 264 no 1 pp 126ndash135 2012

[50] C Liu ldquoSusceptibility tensor imagingrdquo Magnetic Resonance inMedicine vol 63 no 6 pp 1471ndash1477 2010

[51] S Kakkad J Zhang A Akhbardeh et al ldquoCollagen fibersmediate MRI-detected water diffusion and anisotropy in breastcancersrdquo Neoplasia vol 18 no 10 pp 585ndash593 2016

[52] K J Painter and T Hillen ldquoMathematical modelling of gliomagrowth the use of diffusion tensor imaging (DTI) data topredict the anisotropic pathways of cancer invasionrdquo Journal ofTheoretical Biology vol 323 pp 25ndash39 2013

[53] A R Padhani G Liu D Mu-Koh et al ldquoDiffusion-weightedmagnetic resonance imaging as a cancer biomarker consensusand recommendationsrdquo Neoplasia vol 11 no 2 pp 102ndash1252009

[54] E R Gerstner andA G Sorensen ldquoDiffusion and diffusion ten-sor imaging in brain cancerrdquo Seminars in Radiation Oncologyvol 21 no 2 pp 141ndash146 2011

[55] B Stieltjes M Schluter B Didinger et al ldquoDiffusion tensorimaging in primary brain tumors Reproducible quantita-tive analysis of corpus callosum infiltration and contralateralinvolvement using a probabilistic mixture modelrdquo NeuroImagevol 31 no 2 pp 531ndash542 2006

[56] E Furman-Haran D Grobgeld N Nissan M Shapiro-Feinberg and H Degani ldquoCan diffusion tensor anisotropyindices assist in breast cancer detectionrdquo Journal of MagneticResonance Imaging vol 44 no 6 pp 1624ndash1632 2016

[57] B Gurses N Tasdelen F Yencilek et al ldquoDiagnostic utility ofDTI in prostate cancerrdquo European Journal of Radiology vol 79no 2 pp 172ndash176 2011

[58] X Li Q Liang L Zhuang et al ldquoPreliminary study of MRdiffusion tensor imaging of the liver for the diagnosis ofhepatocellular carcinomardquo PLoS ONE vol 10 no 8 Article IDe0135568 2015

[59] K M Pepin R L Ehman and K P McGee ldquoMagnetic reso-nance elastography (MRE) in cancer technique analysis andapplicationsrdquo Progress in Nuclear Magnetic Resonance Spec-troscopy vol 90-91 pp 32ndash48 2015

[60] MAugsten ldquoCancer-associated fibroblasts as another polarizedcell type of the tumor microenvironmentrdquo Frontiers in Oncol-ogy vol 4 article 62 2014

[61] R Muthupillai D J Lomas P J Rossman J F Greenleaf AManduca and R L Ehman ldquoMagnetic resonance elastography

by direct visualization of propagating acoustic strain wavesrdquoScience vol 269 no 5232 pp 1854ndash1857 1995

[62] A L McKnight J L Kugel P J Rossman A Manduca LC Hartmann and R L Ehman ldquoMR elastography of breastcancer preliminary resultsrdquoAmerican Journal of Roentgenologyvol 178 no 6 pp 1411ndash1417 2002

[63] R Sinkus M Tanter S Catheline et al ldquoImaging anisotropicand viscous properties of breast tissue by magnetic resonance-elastographyrdquoMagnetic Resonance inMedicine vol 53 no 2 pp372ndash387 2005

[64] K M Pepin J Chen K J Glaser et al ldquoMR elastographyderived shear stiffnessmdasha new imaging biomarker for theassessment of early tumor response to chemotherapyrdquoMagneticResonance in Medicine vol 71 no 5 pp 1834ndash1840 2014

[65] M Reiss-Zimmermann K-J Streitberger I Sack et al ldquoHighresolution imaging of viscoelastic properties of intracranialtumours by multi-frequency magnetic resonance elastographyrdquoClinical Neuroradiology vol 25 no 4 pp 371ndash378 2015

[66] U Motosugi T Ichikawa T Koshiishi et al ldquoLiver stiffnessmeasured by magnetic resonance elastography as a risk factorfor hepatocellular carcinoma a preliminary case-control studyrdquoEuropean Radiology vol 23 no 1 pp 156ndash162 2013

[67] R S Sahebjavaher G NirMHonarvar et al ldquoMR elastographyof prostate cancer quantitative comparison with histopathol-ogy and repeatability of methodsrdquoNMR in Biomedicine vol 28no 1 pp 124ndash139 2015

[68] J M Arbeit J M Brown K S C Chao et al ldquoHypoxiaimportance in tumor biology noninvasive measurement byimaging and value of its measurement in the management ofcancer therapyrdquo International Journal of Radiation Biology vol82 no 10 pp 699ndash757 2006

[69] R A Gatenby andR J Gillies ldquoAmicroenvironmentalmodel ofcarcinogenesisrdquo Nature Reviews Cancer vol 8 no 1 pp 56ndash612008

[70] V V Khramtsov and R J Gillies ldquoJanus-faced tumor microen-vironment and redoxrdquo Antioxidants and Redox Signaling vol21 no 5 pp 723ndash729 2014

[71] F Q Schafer and G R Buettner ldquoRedox environment of the cellas viewed through the redox state of the glutathione disulfideglutathione couplerdquo Free Radical Biology and Medicine vol 30no 11 pp 1191ndash1212 2001

[72] R K Jain ldquoNormalizing tumor microenvironment to treat can-cer bench to bedside to biomarkersrdquo Journal of Clinical Oncol-ogy vol 31 no 17 pp 2205ndash2218 2013

[73] F Hyodo R M Davis E Hyodo S Matsumoto M C Krishnaand J BMitchell ldquoThe relationship between tissue oxygenationand redox status using magnetic resonance imagingrdquo Interna-tional Journal of Oncology vol 41 no 6 pp 2103ndash2108 2012

[74] P Vaupel A Mayer and M Hockel ldquoTumor hypoxia andmalignant progressionrdquo Methods in Enzimology vol 381 pp335ndash354 2004

[75] R Sanchez-Alvarez U E Martinez-Outschoorn R Lamb et alldquoMitochondrial dysfunction in breast cancer cells preventstumor growth understanding chemoprevention with met-forminrdquo Cell Cycle vol 12 no 1 pp 172ndash182 2013

[76] MC Krishna SMatsumoto K SaitoMMatsuo J BMitchelland J H Ardenkjaer-Larsen ldquoMagnetic resonance imaging oftumor oxygenation andmetabolic profilerdquoActa Oncologica vol52 no 7 pp 1248ndash1256 2013

[77] J Pacheco-Torres P Lopez-Larrubia P Ballesteros and SCerdan ldquoImaging tumor hypoxia bymagnetic resonancemeth-odsrdquo NMR in Biomedicine vol 24 no 1 pp 1ndash16 2011


[78] T Christen B Lemasson N Pannetier et al ldquoEvaluation ofa quantitative blood oxygenation level-dependent (qBOLD)approach to map local blood oxygen saturationrdquo NMR inBiomedicine vol 24 no 4 pp 393ndash403 2011

[79] X He and D A Yablonskiy ldquoQuantitative BOLD mappingof human cerebral deoxygenated blood volume and oxy-gen extraction fraction default staterdquo Magnetic Resonance inMedicine vol 57 no 1 pp 115ndash126 2007

[80] A P Fan B Bilgic L Gagnon et al ldquoQuantitative oxygenationvenography fromMRI phaserdquoMagnetic Resonance in Medicinevol 72 no 1 pp 149ndash159 2014

[81] B Xu T Liu P Spincemaille M Prince and Y Wang ldquoFlowcompensated quantitative susceptibility mapping for venousoxygenation imagingrdquoMagnetic Resonance in Medicine vol 72no 2 pp 438ndash445 2014

[82] A Deistung F Schweser B Wiestler et al ldquoQuantitativesusceptibilitymapping differentiates between blood depositionsand calcifications in patients with glioblastomardquoPLoSONE vol8 no 3 Article ID e57924 2013

[83] C C-T Hsu T W Watkins G N C Kwan and E M HaackeldquoSusceptibility-weighted imaging of glioma update on currentimaging status and future directionsrdquo Journal of Neuroimagingvol 26 no 4 pp 383ndash390 2016

[84] C Baudelet and B Gallez ldquoHow does blood oxygen level-dependent (BOLD) contrast correlate with oxygen partial pres-sure (pO2) inside tumorsrdquo Magnetic Resonance in Medicinevol 48 no 6 pp 980ndash986 2002

[85] L D McPhail and S P Robinson ldquoIntrinsic susceptibility MRimaging of chemically induced rat mammary tumors relation-ship to histologic assessment of hypoxia andfibrosisrdquoRadiologyvol 254 no 1 pp 110ndash118 2010

[86] J P B OrsquoConnor A Jackson G A Buonaccorsi et al ldquoOrgan-specific effects of oxygen and carbogen gas inhalation on tissuelongitudinal relaxation timesrdquoMagnetic Resonance inMedicinevol 58 no 3 pp 490ndash496 2007

[87] J S Burrell S Walker-Samuel L C J Baker et al ldquoExploringDeltaR(2) and DeltaR(1) as imaging biomarkers of tumoroxygenationrdquo Journal of Magnetic Resonance Imaging vol 38no 2 pp 429ndash434 2013

[88] J P B OrsquoConnor J H Naish G J M Parker et al ldquoPreliminarystudy of oxygen-enhanced longitudinal relaxation in MRI apotential novel biomarker of oxygenation changes in solidtumorsrdquo International Journal of Radiation Oncology BiologyPhysics vol 75 no 4 pp 1209ndash1215 2009

[89] H J Halpern C Yu M Peric E Barth D J Grdina and B ATeicher ldquoOxymetry deep in tissues with low-frequency electronparamagnetic resonancerdquo Proceedings of the National Academyof Sciences of the United States of America vol 91 no 26 pp13047ndash13051 1994

[90] F Colliez M-A Neveu J Magat T T C Pham B Gallezand B F Jordan ldquoQualification of a noninvasive magneticresonance imaging biomarker to assess tumor oxygenationrdquoClinical Cancer Research vol 20 no 21 pp 5403ndash5411 2014

[91] M C Krishna S English K Yamada et al ldquoOverhauserenhanced magnetic resonance imaging for tumor oximetrycoregistration of tumor anatomy and tissue oxygen concentra-tionrdquo Proceedings of the National Academy of Sciences of theUnited States of America vol 99 no 4 pp 2216ndash2221 2002

[92] A I Hashim X Zhang J W Wojtkowiak G V Martinez andR J Gillies ldquoImaging pH andmetastasisrdquoNMR in Biomedicinevol 24 no 6 pp 582ndash591 2011

[93] J H Ardenkjaer-Larsen B Fridlund A Gram et al ldquoIncreasein signal-to-noise ratio of gt 10000 times in liquid-state NMRrdquoProceedings of the National Academy of Sciences of the UnitedStates of America vol 100 no 18 pp 10158ndash10163 2003

[94] F A Gallagher M I Kettunen S E Day et al ldquoMagneticresonance imaging of pH in vivo using hyperpolarized 13C-labelled bicarbonaterdquo Nature vol 453 no 7197 pp 940ndash9432008

[95] R R Flavell C von Morze J E Blecha et al ldquoApplication ofGoodrsquos buffers to pH imaging using hyperpolarizedrdquo ChemicalCommunications vol 51 no 74 pp 14119ndash14122 2015

[96] D M Wilson K R Keshari P E Z Larson et al ldquoMulti-compound polarization by DNP allows simultaneous assess-ment of multiple enzymatic activities in vivordquo Journal ofMagnetic Resonance vol 205 no 1 pp 141ndash147 2010

[97] D J Scholz M A Janich U Kollisch et al ldquoQuantified pHimaging with hyperpolarized 13C-bicarbonaterdquo Magnetic Res-onance in Medicine vol 73 no 6 pp 2274ndash2282 2015

[98] S Duwel C Hundshammer M Gersch et al ldquoImaging of pHin vivo using hyperpolarized 13C-labelled zymonic acidrdquoNatureCommunications vol 8 p 15126 2017

[99] R V Shchepin D A Barskiy A M Coffey et al ldquo15N hyper-polarization of imidazole-15N2 for magnetic resonance pHsensing via SABRE-SHEATHrdquo ACS Sensors vol 1 no 6 pp640ndash644 2016

[100] P Z Sun G Xiao I Y Zhou Y Guo and R Wu ldquoA methodfor accurate pH mapping with chemical exchange saturationtransfer (CEST) MRIrdquo Contrast Media and Molecular Imagingvol 11 no 3 pp 195ndash202 2016

[101] Y Wu S Zhang T C Soesbe et al ldquoPH imaging of mousekidneys in vivo using a frequency-dependent paraCEST agentrdquoMagnetic Resonance in Medicine vol 75 no 6 pp 2432ndash24412016

[102] X Yang X Song S Ray Banerjee et al ldquoDeveloping imidazolesas CEST MRI pH sensorsrdquo Contrast Media and MolecularImaging vol 11 no 4 pp 304ndash312 2016

[103] L Q Chen C M Howison J J Jeffery I F Robey P H Kuoand M D Pagel ldquoEvaluations of extracellular pH within invivo tumors using acidoCEST MRIrdquo Magnetic Resonance inMedicine vol 72 no 5 pp 1408ndash1417 2014

[104] A Acharya I Das D Chandhok and T Saha ldquoRedox reg-ulation in cancer a double-edged sword with therapeuticpotentialrdquoOxidative Medicine and Cellular Longevity vol 3 no1 pp 23ndash34 2010

[105] S E Bohndiek M I Kettunen D-E Hu et al ldquoHyperpolarized[1-13C]-ascorbic and dehydroascorbic acid vitaminC as a probefor imaging redox status in vivordquo Journal of the AmericanChemical Society vol 133 no 30 pp 11795ndash11801 2011

[106] K R Keshari J Kurhanewicz R Bok P E Z Larson DB Vigneron and D M Wilson ldquoHyperpolarized 13C dehy-droascorbate as an endogenous redox sensor for in vivo meta-bolic imagingrdquo Proceedings of the National Academy of Sciencesof the United States of America vol 108 no 46 pp 18606ndash186112011

[107] C E Christensen M Karlsson J R Winther P R Jensenand M H Lerche ldquoNon-invasive in-cell determination of freecytosolic [NAD+] [NADH] ratios using hyperpolarized glu-cose show large variations in metabolic phenotypesrdquo Journal ofBiological Chemistry vol 289 no 4 pp 2344ndash2352 2014

[108] T B Rodrigues E M Serrao B W C Kennedy D-E Hu MI Kettunen and K M Brindle ldquoMagnetic resonance imaging


of tumor glycolysis using hyperpolarized 13 C-labeled glucoserdquoNature Medicine vol 20 no 1 pp 93ndash97 2014

[109] OWarburg ldquoOn the origin of cancer cellsrdquo Science vol 123 no3191 pp 309ndash314 1956

[110] P R Rich ldquoThe molecular machinery of Keilinrsquos respiratorychainrdquo Biochemical Society Transactions vol 31 no 6 pp 1095ndash1105 2003

[111] R Boidot F Vegran A Meulle et al ldquoRegulation of monocar-boxylate transporter MCT1 expression by p53 mediates inwardand outward lactate fluxes in tumorsrdquo Cancer Research vol 72no 4 pp 939ndash948 2012

[112] J L Griffin and R A Kauppinen ldquoTumour metabolomics inanimal models of human cancerrdquo Journal of Proteome Researchvol 6 no 2 pp 498ndash505 2007

[113] L C Costello and R B Franklin ldquorsquoWhy do tumour cellsglycolysersquo from glycolysis through citrate to lipogenesisrdquoMolecular and Cellular Biochemistry vol 280 no 1-2 pp 1ndash82005

[114] R L Elstrom D E Bauer and M Buzzai ldquoAkt stimulates aer-obic glycolysis in cancer cellsrdquo Cancer Research vol 64 no 11pp 3892ndash3899 2004

[115] D R Plas andC BThompson ldquoAkt-dependent transformationthere is more to growth than just survivingrdquo Oncogene vol 24no 50 pp 7435ndash7442 2005

[116] T Porstmann C R Santos B Griffiths et al ldquoSREBP activityis regulated by mTORC1 and contributes to Akt-dependent cellgrowthrdquo Cell Metabolism vol 8 no 3 pp 224ndash236 2008

[117] J-W Kim and C V Dang ldquoCancerrsquos molecular sweet tooth andthe warburg effectrdquo Cancer Research vol 66 no 18 pp 8927ndash8930 2006

[118] H Shim C Dolde B C Lewis et al ldquoc-Myc transactivationof LDH-A implications for tumor metabolism and growthrdquoProceedings of the National Academy of Sciences of the UnitedStates of America vol 94 no 13 pp 6658ndash6663 1997

[119] J Downward ldquoTargeting RAS signalling pathways in cancertherapyrdquo Nature Reviews Cancer vol 3 no 1 pp 11ndash22 2003

[120] S Telang A Yalcin A L Clem et al ldquoRas transforma-tion requires metabolic control by 6-phosphofructo-2-kinaserdquoOncogene vol 25 no 55 pp 7225ndash7234 2006

[121] M J Albers R Bok A P Chen et al ldquoHyperpolarized13C lactate pyruvate and alanine noninvasive biomarkers forprostate cancer detection and gradingrdquoCancer Research vol 68no 20 pp 8607ndash8615 2008

[122] J M Park L D Recht S Josan et al ldquoMetabolic responseof glioma to dichloroacetate measured in vivo by hyperpolar-ized 13C magnetic resonance spectroscopic imagingrdquo Neuro-Oncology vol 15 no 4 pp 433ndash441 2013

[123] J A Bankson C M Walker M S Ramirez et al ldquoKineticmodeling and constrained reconstruction of hyperpolarized [1-13C]-pyruvate offers improved metabolic imaging of tumorsrdquoCancer Research vol 75 no 22 pp 4708ndash4717 2015

[124] C Harrison C Yang and A Jindal ldquoComparison of kineticmodels for analysis of pyruvate-to-lactate exchange by hyper-polarized 13C NMRrdquo NMR in Biomedicine vol 25 no 11 pp1286ndash1294 2012

[125] I Park R Bok T Ozawa et al ldquoDetection of early response totemozolomide treatment in brain tumors using hyperpolarized13C MR metabolic imagingrdquo Journal of Magnetic ResonanceImaging vol 33 no 6 pp 1284ndash1290 2011

[126] S E Day M I Kettunen and F A Gallagher ldquoDetecting tumorresponse to treatment using hyperpolarized 13C magnetic

resonance imaging and spectroscopyrdquo Nature Medicine vol 13no 11 pp 1382ndash1387 2007

[127] T H Witney M I Kettunen S E Day et al ldquoA com-parison between radiolabeled fluorodeoxyglucose uptake andhyperpolarized13c-labeled pyruvate utilization as methods fordetecting tumor response to treatmentrdquoNeoplasia vol 11 no 6pp 574ndash582 2009

[128] M M Chaumeil T Ozawa I Park et al ldquoHyperpolarized 13CMR spectroscopic imaging can be used to monitor Everolimustreatment in vivo in an orthotopic rodent model of glioblas-tomardquo NeuroImage vol 59 no 1 pp 193ndash201 2012

[129] J M Park D M Spielman S Josan et al ldquoHyperpolarized 13C-lactate to 13C-bicarbonate ratio as a biomarker for monitoringthe acute response of anti-vascular endothelial growth factor(anti-VEGF) treatmentrdquoNMR in Biomedicine vol 29 no 5 pp650ndash659 2016

[130] R J DeBerardinis J J Lum G Hatzivassiliou and C BThompson ldquoThe biology of cancer metabolic reprogrammingfuels cell growth and proliferationrdquo Cell Metabolism vol 7 no1 pp 11ndash20 2008

[131] D R Wise R J Deberardinis A Mancuso et al ldquoMyc regu-lates a transcriptional program that stimulates mitochondrialglutaminolysis and leads to glutamine addictionrdquo Proceedings ofthe National Academy of Sciences of the United States of Americavol 105 no 48 pp 18782ndash18787 2008

[132] C Cabella M Karlsson C Canape et al ldquoIn vivo and invitro liver cancer metabolism observed with hyperpolarized [5-13C]glutaminerdquo Journal ofMagnetic Resonance vol 232 pp 45ndash52 2013

[133] C Canape G Catanzaro E TerrenoMKarlssonMH Lercheand P R Jensen ldquoProbing treatment response of glutaminolyticprostate cancer cells to natural drugs with hyperpolarized [5-13C]glutaminerdquoMagnetic Resonance in Medicine vol 73 no 6pp 2296ndash2305 2015

[134] F A Gallagher M I Kettunen S E Day et al ldquoDetectionof tumor glutamate metabolism in vivo using 13C magneticresonance spectroscopy and hyperpolarized [1- 13C]glutamaterdquoMagnetic Resonance in Medicine vol 66 no 1 pp 18ndash23 2011

[135] K Glunde Z M Bhujwalla and S M Ronen ldquoCholinemetabolism inmalignant transformationrdquoNature Reviews Can-cer vol 11 no 12 pp 835ndash848 2011

[136] P B Barker ldquoN-acetyl aspartatemdasha neuronal markerrdquo Annalsof Neurology vol 49 no 4 pp 423-424 2001

[137] A Horska and P B Barker ldquoImaging of brain tumors MRspectroscopy and metabolic imagingrdquo Neuroimaging Clinics ofNorth America vol 20 no 3 pp 293ndash310 2010

[138] T-F Lou D Sethuraman P Dospoy et al ldquoCancer-specificproduction of n-acetylaspartate via nat8l overexpression innon-small cell lung cancer and its potential as a circulatingbiomarkerrdquo Cancer Prevention Research vol 9 no 1 pp 43ndash522016

[139] B Zand R A Previs N M Zacharias et al ldquoRole of Increasedn-acetylaspartate Levels in Cancerrdquo Journal of the NationalCancer Institute vol 108 no 6 p djv426 2016

[140] S Patra A Ghosh S S Roy et al ldquoA short review oncreatine-creatine kinase system in relation to cancer and someexperimental results on creatine as adjuvant in cancer therapyrdquoAmino Acids vol 42 no 6 pp 2319ndash2330 2012

[141] M G V Heiden L C Cantley and C B Thompson ldquoUnder-standing the warburg effect the metabolic requirements of cellproliferationrdquo Science vol 324 no 5930 pp 1029ndash1033 2009


[142] H Lyng OHaraldseth and E K Rofstad ldquoMeasurement of celldensity and necrotic fraction in human melanoma xenograftsby diffusion weighted magnetic resonance imagingrdquo MagneticResonance in Medicine vol 43 no 6 pp 828ndash836 2000

[143] M R Clatworthy M I Kettunen D-E Hu et al ldquoMagneticresonance imaging with hyperpolarized [14- 13C 2]fumarateallows detection of early renal acute tubular necrosisrdquo Proceed-ings of the National Academy of Sciences of the United States ofAmerica vol 109 no 33 pp 13374ndash13379 2012

[144] N Andor T A Graham M Jansen et al ldquoPan-cancer analysisof the extent and consequences of intratumor heterogeneityrdquoNature Medicine vol 22 no 1 pp 105ndash113 2016

[145] X Zhang S L Marjani Z Hu S M Weissman X Panand S Wu ldquoSingle-cell sequencing for precise cancer researchprogress and prospectsrdquo Cancer Research vol 76 no 6 pp1305ndash1312 2016

[146] A S Cleary ldquoTeamwork the tumor cell editionrdquo Science vol350 no 6265 pp 1174-1175 2015

[147] J Calbo E van Montfort N Proost et al ldquoA functional role fortumor cell heterogeneity in a mouse model of small cell lungcancerrdquo Cancer Cell vol 19 no 2 pp 244ndash256 2011

[148] M Gerlinger A J Rowan S Horswell et al et al ldquoIntratumorheterogeneity and branched evolution revealed by multiregionsequencingrdquoTheNew Englang Journal of Medicine vol 366 pp883ndash892 2012

[149] O Feron ldquoPyruvate into lactate and back from the warburgeffect to symbiotic energy fuel exchange in cancer cellsrdquoRadiotherapy and Oncology vol 92 no 3 pp 329ndash333 2009

[150] P Icard and H Lincet ldquoA global view of the biochemicalpathways involved in the regulation of themetabolism of cancercellsrdquo Biochimica et Biophysica Acta vol 1826 no 2 pp 423ndash433 2012

[151] G L Semenza ldquoTumor metabolism cancer cells give and takelactaterdquo Journal of Clinical Investigation vol 118 no 12 pp3835ndash3837 2008

[152] P Sonveaux F Vegran T Schroeder et al ldquoTargeting lactate-fueled respiration selectively kills hypoxic tumor cells in micerdquoThe Journal of Clinical Investigation vol 118 pp 3930ndash39422008

[153] J P B OrsquoConnor C J Rose J CWaterton R A D Carano G JM Parker and A Jackson ldquoImaging intratumor heterogeneityrole in therapy response resistance and clinical outcomerdquoClinical Cancer Research vol 21 no 2 pp 249ndash257 2015

[154] A A Alizadeh V Aranda A Bardelli et al ldquoToward under-standing and exploiting tumor heterogeneityrdquoNature Medicinevol 21 no 8 pp 846ndash853 2015

[155] D P TabassumandK Polyak ldquoTumorigenesis it takes a villagerdquoNature Reviews Cancer vol 15 no 8 pp 473ndash483 2015

[156] I Serganova M Doubrovin J Vider et al ldquoMolecular imagingof temporal dynamics and spatial heterogeneity of hypoxia-inducible factor-1 signal transduction activity in tumors inliving micerdquo Cancer Research vol 64 no 17 pp 6101ndash61082004

[157] L I Cardenas-Navia D Mace R A Richardson D F WilsonS Shan and M W Dewhirst ldquoThe pervasive presence offluctuating oxygenation in tumorsrdquoCancer Research vol 68 no14 pp 5812ndash5819 2008

[158] C J Eskey and A P Koretsky ldquo2H-nuclear magnetic resonanceimaging of tumor blood flow spatial and temporal hetero-geneity in a tissue-isolatedmammary adenocarcinomardquoCancerResearch vol 52 no 21 pp 6010ndash6019 1992

[159] K G Brurberg J-V Gaustad C S Mollatt and E K RofstadldquoTemporal heterogeneity in blood supply in human tumorxenograftsrdquo Neoplasia vol 10 no 7 pp 727ndash735 2008

[160] M W Dewhirst ldquoRelationships between cycling hypoxia HIF-1 angiogenesis and oxidative stressrdquo Radiation Research vol172 no 6 pp 653ndash665 2009

[161] I Marchiq and J Pouyssegur ldquoHypoxia cancer metabolismand the therapeutic benefit of targeting lactateH+ symportersrdquoJournal of Molecular Medicine vol 94 no 2 pp 155ndash171 2016

[162] P Friedl ldquoDynamic imaging of the immune systemrdquo CurrentOpinion in Immunology vol 16 pp 389ndash393 2004

[163] P Friedl and S Alexander ldquoCancer invasion and the microen-vironment plasticity and reciprocityrdquo Cell vol 147 no 5 pp992ndash1009 2011

[164] F Hirschhaeuser U G A Sattler and W Mueller-KlieserldquoLactate ametabolic key player in cancerrdquoCancer Research vol71 no 22 pp 6921ndash6925 2011

[165] P Dalerba T Kalisky D Sahoo et al ldquoSingle-cell dissection oftranscriptional heterogeneity in human colon tumorsrdquo NatureBiotechnology vol 29 no 12 pp 1120ndash1127 2011

[166] S V Avery ldquoMicrobial cell individuality and the underlyingsources of heterogeneityrdquo Nature Reviews Microbiology vol 4no 8 pp 577ndash587 2006

[167] L Alic W J Niessen and J F Veenland ldquoQuantification ofheterogeneity as a biomarker in tumor imaging a systematicreviewrdquo PLoS ONE vol 9 no 10 Article ID e110300 2014

[168] N Just ldquoImproving tumour heterogeneityMRI assessment withhistogramsrdquo British Journal of Cancer vol 111 pp 2205ndash22132014

[169] S-L Peng C-F Chen H-L Liu et al ldquoAnalysis of parametrichistogram from dynamic contrast-enhanced MRI applicationin evaluating brain tumor response to radiotherapyrdquo NMR inBiomedicine vol 26 no 4 pp 443ndash450 2013

[170] K Downey S F Riches V A Morgan et al ldquoRelationshipbetween imaging biomarkers of stage I cervical cancer andpoor-prognosis histologic features quantitative histogramanal-ysis of diffusion-weighted MR imagesrdquo American Journal ofRoentgenology vol 200 no 2 pp 314ndash320 2013

[171] P S Tofts C E Benton R SWeil et al ldquoQuantitative analysis ofwhole-tumor Gd enhancement histograms predicts malignanttransformation in low-grade gliomasrdquo Journal of MagneticResonance Imaging vol 25 no 1 pp 208ndash214 2007

[172] E A Kidd and P W Grigsby ldquoIntratumoral metabolic hetero-geneity of cervical cancerrdquo Clinical Cancer Research vol 14 no16 pp 5236ndash5241 2008

[173] F Davnall C S P Yip G Ljungqvist et al ldquoAssessment of tumorheterogeneity an emerging imaging tool for clinical practicerdquoInsights into Imaging vol 3 no 6 pp 573ndash589 2012

[174] G Castellano L Bonilha L M Li and F Cendes ldquoTextureanalysis of medical imagesrdquo Clinical Radiology vol 59 no 12pp 1061ndash1069 2004

[175] E Scalco and G Rizzo ldquoTexture analysis of medical images forradiotherapy applicationsrdquo British Journal of Radiology vol 90no 1070 Article ID 20160642 2017

[176] F Michallek and M Dewey ldquoFractal analysis in radiologicaland nuclearmedicine perfusion imaging a systematic reviewrdquoEuropean Radiology vol 24 no 1 pp 60ndash69 2014

[177] J P B OrsquoConnor A Jackson G J M Parker C Roberts and GC Jayson ldquoDynamic contrast-enhancedMRI in clinical trials ofantivascular therapiesrdquoNature Reviews Clinical Oncology vol 9no 3 pp 167ndash177 2012


[178] L Alic M Van Vliet C F Van Dijke A M M Eggermont J FVeenland andW J Niessen ldquoHeterogeneity in DCE-MRI para-metric maps a biomarker for treatment responserdquo Physics inMedicine and Biology vol 56 no 6 pp 1601ndash1616 2011

[179] D L Longo W Dastru L Consolino et al ldquoCluster analysisof quantitative parametric maps fromDCE-MRI application inevaluating heterogeneity of tumor response to antiangiogenictreatmentrdquoMagnetic Resonance Imaging vol 33 no 6 pp 725ndash736 2015

[180] A Jackson J P B OrsquoConnor G J M Parker and G C JaysonldquoImaging tumor vascular heterogeneity and angiogenesis usingdynamic contrast-enhanced magnetic resonance imagingrdquoClinical Cancer Research vol 13 no 12 pp 3449ndash34592007

[181] F Szczepankiewicz D van Westen E Englund et al ldquoThe linkbetween diffusion MRI and tumor heterogeneity mapping celleccentricity and density by diffusional variance decomposition(DIVIDE)rdquo NeuroImage vol 142 pp 522ndash532 2016

[182] M D Jenkinson D G Du Plessis T S Smith A R BrodbeltK A Joyce and C Walker ldquoCellularity and apparent diffusioncoefficient in oligodendroglial tumours characterized by geno-typerdquo Journal of Neuro-Oncology vol 96 no 3 pp 385ndash3922010

[183] R G Abramson L R Arlinghaus A N Dula et al ldquoMR imag-ing biomarkers in oncology clinical trialsrdquoMagnetic ResonanceImaging Clinics of North America vol 24 no 1 pp 11ndash29 2016

[184] C Choi J M Raisanen S K Ganji et al ldquoProspective longitu-dinal analysis of 2-hydroxyglutarate magnetic resonance spec-troscopy identifies broad clinical utility for the management ofpatients with IDH-mutant gliomardquo Journal of Clinical Oncologyvol 34 no 33 pp 4030ndash4039 2016

[185] A Tietze C Choi B Mickey et al ldquoNoninvasive assessment ofisocitrate dehydrogenase mutation status in cerebral gliomas bymagnetic resonance spectroscopy in a clinical settingrdquo Journalof Neurosurgery pp 1ndash8 2017

[186] C K Jones M J Schlosser P C M Van Zijl M G Pomper XGolay and J Zhou ldquoAmide proton transfer imaging of humanbrain tumors at 3TrdquoMagnetic Resonance inMedicine vol 56 no3 pp 585ndash592 2006

[187] J Zhou J O Blakeley J Hua et al ldquoPractical data acquisitionmethod for human brain tumor amide proton transfer (APT)imagingrdquo Magnetic Resonance in Medicine vol 60 no 4 pp842ndash849 2008

[188] B Epel G Redler V Tormyshev and H J Halpern ldquoTowardshuman oxygen images with electron paramagnetic resonanceimagingrdquo Advances in Experimental Medicine and Biology vol876 pp 363ndash369 2016

[189] S J Nelson J Kurhanewicz and D B Vigneron ldquoMetabolicimaging of patients with prostate cancer using hyperpolarized[1-13C]pyruvaterdquo Science Translational Medicine vol 5 no 198Article ID 198ra108 2013

[190] M G Vander Heiden and R J DeBerardinis ldquoUnderstandingthe Intersections betweenmetabolism and cancer biologyrdquoCellvol 168 no 4 pp 657ndash669 2017

[191] F J Brooks and P W Grigsby ldquoCurrent measures of metabolicheterogeneity within cervical cancer do not predict diseaseoutcomerdquo Radiation Oncology vol 6 no 1 article 69 2011

[192] A Eklund T E Nichols andH Knutsson ldquoCluster failure whyfMRI inferences for spatial extent have inflated false-positiveratesrdquo Proceedings of the National Academy of Sciences of theUnited States of America vol 113 no 28 pp 7900ndash7905 2016

[193] H J Aerts E R Velazquez R T Leijenaar et al ldquoDecodingtumour phenotype by noninvasive imaging using a quantitativeradiomics approachrdquo Nature Communications vol 5 article4006 2014

Submit your manuscripts athttpswwwhindawicom

Stem CellsInternational

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014


MEDIATORSINFLAMMATION

of


Behavioural Neurology

EndocrinologyInternational Journal of



Disease Markers


BioMed Research International

OncologyJournal of



Oxidative Medicine and Cellular Longevity


PPAR Research

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

Immunology ResearchHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Journal of

ObesityJournal of



Computational and Mathematical Methods in Medicine

OphthalmologyJournal of


Diabetes ResearchJournal of



Research and TreatmentAIDS


Gastroenterology Research and Practice


Parkinsonrsquos Disease

Evidence-Based Complementary and Alternative Medicine

Volume 2014Hindawi Publishing Corporationhttpwwwhindawicom


Microvasculatureperfusion permeability

Gross anatomy amptissue characteristics

morphology spin relaxation

Microstructurediusion elasticity

Blood vessel

（+

（+

（+

2

2

2

Microenvironmenthypoxia acidosis redox

Metabolismbioenergetics redox

Glucose

LactateGlycolysis

Pyruvate

TCAGlutamine

Figure 1MRI andMRS in imaging cancerMRI andMRS provide useful information of cancermetabolism and tumor heterogeneity rangingfrom anatomical change tomicrovascular development biophysical characteristicsmicrostructural deformation altered cellularmetabolismand tumor microenvironment

cancer following treatment which could be useful readoutsfor monitoring response to therapies [6 7]

Ideal utilization of molecular imaging is to ldquodose paintrdquothe radiotherapy dose administered to each tumor withreference to positron emission tomography (PET) [8] andto identify the geographic subregions that drive response totherapy subsequent resistance and relapse during treatmentfailure [9 10] However further work has shown that theinterplay between abnormal metabolism vascularizationand hypoxia expression in tumors may lead to different mapsof abnormality depending on the functional pathophysio-logical readout (ie perfusion hypoxia glucose metabolismetc) [11] In order to select optimal imaging paradigms toguide treatment a deeper understanding of the underlyingbiological mechanisms is critical There is a strong rationalefor investigating whether hypoxic regions should be treatedwith differing radiation doses to well-oxygenated tumorsas well as investigating regional variation based on func-tional and molecular imaging This idea represents a majorparadigm shift where images will be composed of arrays ofdata arranged spatially in individual voxels [10] Each voxel is

a cube of data which summarizes a particular morphologicmetabolic or physiologic signal over a volume of around025ndash5mm [12] depending on modality and subject (animalor human)

Complex cancer metabolism and associated characteris-tics have been extensively explored by magnetic resonanceimaging (MRI) and spectroscopy (MRS) using the versatilerelaxation mechanisms of nuclear spins that provide uniqueand convertible tissue contrasts Advances in MR techniqueshave enabled noninvasive access to significant amounts ofuseful information on cancer metabolism and tumor hetero-geneity ranging in spatial scales from gross anatomy bio-physical characteristics and functional or metabolic imaging(Figure 1) It is important to appreciate that these abundantparameters can be extracted from a single acquisition toprovide general structural data (eg size) functional patho-physiological data (eg average bloodflowandpermeability)and various heterogeneity-based metrics in the tumor In thefollowing sections imaging techniques for evaluating cancermetabolism and tumor heterogeneity will be reviewed on avariety of scales














Data

0

50

100

150

200

250

300

350

400Si

gnal

inte

nsity

0 2 3 4 5 61

Time (min)

(a)

DataModel fit

0

055

010

015

020

025

030

035

040

Gad

olin

ium

(nM

olk

g)

1 2 3 4 5 60

Time (min)

(b)

1 2 3 4 5 60

Time (min)

0

1

2

3

4

5

6

7

Cp(t)

(nM

ol)

(c)

08

0

(ＧＣＨ

minus1)

(d)08

0

(ＧＣＨ

minus1)

(e)minus50

minus40

minus30

minus20

minus10

0

10

Chan

ge in

KＮＬＨ

Ｍ(

)

2 4 6 8 10minus2

Time (days)

(f)





(a) (b) (c)

28

0

(d)

30

0

(e)








10158402 (1119879

10158402 =



(a) (b)

(c) (d)

50

25

(kPa

)

0

minus20

0

20

(m






SCC tumor



(a) (b)

(c) (d)

(e) (f)

04

03

02

01

00

40

30

20

10

0

012

009

006

003

000



pH80

70

60

(a)

pH80

70

60

(b)

pH80

70

60

(c)




13C-glucose






0

[2-13]Pyruvate 009

0

[2-13]Lactate

001

0

003

0


Tumor

Surface coil

1 cm

(a) (b)

(c) (d)










Cr ChoCr

NAA

In vivoFitResiduals


Glu (90 plusmn 02)

Gln (24 plusmn 02)

Normal brain

3 2 14

(ppm)

(a)


ChoCr NAA

Lac Lip


Glu (46 plusmn 02)

Gln (53 plusmn 02)

3 2 14

(ppm)

(b)


Cho

Cr NAA Lac

2HG (33 plusmn 02)


Gln (29 plusmn 02)

3 2 14

(ppm)

(c)


Cho

Cr NAA Lac

2HG (91 plusmn 03)

GABA (02 plusmn 02)

Glu (05 plusmn 02)

Gln (02 plusmn 02)

3 2 14

(ppm)

(d)


Gly

Cho

CrNAA Lac

2HG (89 plusmn 03)

GABA (0)

Glu (21 plusmn 02)

Gln (23 plusmn 02)

3 2 14

(ppm)

(e)


ChoCr NAA

Lip

2HG (46 plusmn 03)

GABA (01 plusmn 01)

Glu (14 plusmn 02)

Gln (26 plusmn 02)

3 2 14

(ppm)

(f)





















Acknowledgments


References

















































































































































































































of






Disease Markers



OncologyJournal of





PPAR Research



Journal of

ObesityJournal of





























Data

0

50

100

150

200

250

300

350

400Si

gnal

inte

nsity

0 2 3 4 5 61

Time (min)

(a)

DataModel fit

0

055

010

015

020

025

030

035

040

Gad

olin

ium

(nM

olk

g)

1 2 3 4 5 60

Time (min)

(b)

1 2 3 4 5 60

Time (min)

0

1

2

3

4

5

6

7

Cp(t)

(nM

ol)

(c)

08

0

(ＧＣＨ

minus1)

(d)08

0

(ＧＣＨ

minus1)

(e)minus50

minus40

minus30

minus20

minus10

0

10

Chan

ge in

KＮＬＨ

Ｍ(

)

2 4 6 8 10minus2

Time (days)

(f)





(a) (b) (c)

28

0

(d)

30

0

(e)








10158402 (1119879

10158402 =



(a) (b)

(c) (d)

50

25

(kPa

)

0

minus20

0

20

(m






SCC tumor



(a) (b)

(c) (d)

(e) (f)

04

03

02

01

00

40

30

20

10

0

012

009

006

003

000



pH80

70

60

(a)

pH80

70

60

(b)

pH80

70

60

(c)




13C-glucose






0

[2-13]Pyruvate 009

0

[2-13]Lactate

001

0

003

0


Tumor

Surface coil

1 cm

(a) (b)

(c) (d)










Cr ChoCr

NAA

In vivoFitResiduals


Glu (90 plusmn 02)

Gln (24 plusmn 02)

Normal brain

3 2 14

(ppm)

(a)


ChoCr NAA

Lac Lip


Glu (46 plusmn 02)

Gln (53 plusmn 02)

3 2 14

(ppm)

(b)


Cho

Cr NAA Lac

2HG (33 plusmn 02)


Gln (29 plusmn 02)

3 2 14

(ppm)

(c)


Cho

Cr NAA Lac

2HG (91 plusmn 03)

GABA (02 plusmn 02)

Glu (05 plusmn 02)

Gln (02 plusmn 02)

3 2 14

(ppm)

(d)


Gly

Cho

CrNAA Lac

2HG (89 plusmn 03)

GABA (0)

Glu (21 plusmn 02)

Gln (23 plusmn 02)

3 2 14

(ppm)

(e)


ChoCr NAA

Lip

2HG (46 plusmn 03)

GABA (01 plusmn 01)

Glu (14 plusmn 02)

Gln (26 plusmn 02)

3 2 14

(ppm)

(f)





















Acknowledgments


References

















































































































































































































of






Disease Markers



OncologyJournal of





PPAR Research



Journal of

ObesityJournal of

















(a) (b) (c)

28

0

(d)

30

0

(e)








10158402 (1119879

10158402 =



(a) (b)

(c) (d)

50

25

(kPa

)

0

minus20

0

20

(m






SCC tumor



(a) (b)

(c) (d)

(e) (f)

04

03

02

01

00

40

30

20

10

0

012

009

006

003

000



pH80

70

60

(a)

pH80

70

60

(b)

pH80

70

60

(c)




13C-glucose






0

[2-13]Pyruvate 009

0

[2-13]Lactate

001

0

003

0


Tumor

Surface coil

1 cm

(a) (b)

(c) (d)










Cr ChoCr

NAA

In vivoFitResiduals


Glu (90 plusmn 02)

Gln (24 plusmn 02)

Normal brain

3 2 14

(ppm)

(a)


ChoCr NAA

Lac Lip


Glu (46 plusmn 02)

Gln (53 plusmn 02)

3 2 14

(ppm)

(b)


Cho

Cr NAA Lac

2HG (33 plusmn 02)


Gln (29 plusmn 02)

3 2 14

(ppm)

(c)


Cho

Cr NAA Lac

2HG (91 plusmn 03)

GABA (02 plusmn 02)

Glu (05 plusmn 02)

Gln (02 plusmn 02)

3 2 14

(ppm)

(d)


Gly

Cho

CrNAA Lac

2HG (89 plusmn 03)

GABA (0)

Glu (21 plusmn 02)

Gln (23 plusmn 02)

3 2 14

(ppm)

(e)


ChoCr NAA

Lip

2HG (46 plusmn 03)

GABA (01 plusmn 01)

Glu (14 plusmn 02)

Gln (26 plusmn 02)

3 2 14

(ppm)

(f)





















Acknowledgments


References

















































































































































































































of






Disease Markers



OncologyJournal of





PPAR Research



Journal of

ObesityJournal of

















(a) (b)

(c) (d)

50

25

(kPa

)

0

minus20

0

20

(m






SCC tumor



(a) (b)

(c) (d)

(e) (f)

04

03

02

01

00

40

30

20

10

0

012

009

006

003

000



pH80

70

60

(a)

pH80

70

60

(b)

pH80

70

60

(c)




13C-glucose






0

[2-13]Pyruvate 009

0

[2-13]Lactate

001

0

003

0


Tumor

Surface coil

1 cm

(a) (b)

(c) (d)










Cr ChoCr

NAA

In vivoFitResiduals


Glu (90 plusmn 02)

Gln (24 plusmn 02)

Normal brain

3 2 14

(ppm)

(a)


ChoCr NAA

Lac Lip


Glu (46 plusmn 02)

Gln (53 plusmn 02)

3 2 14

(ppm)

(b)


Cho

Cr NAA Lac

2HG (33 plusmn 02)


Gln (29 plusmn 02)

3 2 14

(ppm)

(c)


Cho

Cr NAA Lac

2HG (91 plusmn 03)

GABA (02 plusmn 02)

Glu (05 plusmn 02)

Gln (02 plusmn 02)

3 2 14

(ppm)

(d)


Gly

Cho

CrNAA Lac

2HG (89 plusmn 03)

GABA (0)

Glu (21 plusmn 02)

Gln (23 plusmn 02)

3 2 14

(ppm)

(e)


ChoCr NAA

Lip

2HG (46 plusmn 03)

GABA (01 plusmn 01)

Glu (14 plusmn 02)

Gln (26 plusmn 02)

3 2 14

(ppm)

(f)





















Acknowledgments


References

















































































































































































































of






Disease Markers



OncologyJournal of





PPAR Research



Journal of

ObesityJournal of

















SCC tumor



(a) (b)

(c) (d)

(e) (f)

04

03

02

01

00

40

30

20

10

0

012

009

006

003

000



pH80

70

60

(a)

pH80

70

60

(b)

pH80

70

60

(c)




13C-glucose






0

[2-13]Pyruvate 009

0

[2-13]Lactate

001

0

003

0


Tumor

Surface coil

1 cm

(a) (b)

(c) (d)










Cr ChoCr

NAA

In vivoFitResiduals


Glu (90 plusmn 02)

Gln (24 plusmn 02)

Normal brain

3 2 14

(ppm)

(a)


ChoCr NAA

Lac Lip


Glu (46 plusmn 02)

Gln (53 plusmn 02)

3 2 14

(ppm)

(b)


Cho

Cr NAA Lac

2HG (33 plusmn 02)


Gln (29 plusmn 02)

3 2 14

(ppm)

(c)


Cho

Cr NAA Lac

2HG (91 plusmn 03)

GABA (02 plusmn 02)

Glu (05 plusmn 02)

Gln (02 plusmn 02)

3 2 14

(ppm)

(d)


Gly

Cho

CrNAA Lac

2HG (89 plusmn 03)

GABA (0)

Glu (21 plusmn 02)

Gln (23 plusmn 02)

3 2 14

(ppm)

(e)


ChoCr NAA

Lip

2HG (46 plusmn 03)

GABA (01 plusmn 01)

Glu (14 plusmn 02)

Gln (26 plusmn 02)

3 2 14

(ppm)

(f)





















Acknowledgments


References

















































































































































































































of






Disease Markers



OncologyJournal of





PPAR Research



Journal of

ObesityJournal of

















pH80

70

60

(a)

pH80

70

60

(b)

pH80

70

60

(c)




13C-glucose






0

[2-13]Pyruvate 009

0

[2-13]Lactate

001

0

003

0


Tumor

Surface coil

1 cm

(a) (b)

(c) (d)










Cr ChoCr

NAA

In vivoFitResiduals


Glu (90 plusmn 02)

Gln (24 plusmn 02)

Normal brain

3 2 14

(ppm)

(a)


ChoCr NAA

Lac Lip


Glu (46 plusmn 02)

Gln (53 plusmn 02)

3 2 14

(ppm)

(b)


Cho

Cr NAA Lac

2HG (33 plusmn 02)


Gln (29 plusmn 02)

3 2 14

(ppm)

(c)


Cho

Cr NAA Lac

2HG (91 plusmn 03)

GABA (02 plusmn 02)

Glu (05 plusmn 02)

Gln (02 plusmn 02)

3 2 14

(ppm)

(d)


Gly

Cho

CrNAA Lac

2HG (89 plusmn 03)

GABA (0)

Glu (21 plusmn 02)

Gln (23 plusmn 02)

3 2 14

(ppm)

(e)


ChoCr NAA

Lip

2HG (46 plusmn 03)

GABA (01 plusmn 01)

Glu (14 plusmn 02)

Gln (26 plusmn 02)

3 2 14

(ppm)

(f)





















Acknowledgments


References

















































































































































































































of






Disease Markers



OncologyJournal of





PPAR Research



Journal of

ObesityJournal of


















0

[2-13]Pyruvate 009

0

[2-13]Lactate

001

0

003

0


Tumor

Surface coil

1 cm

(a) (b)

(c) (d)










Cr ChoCr

NAA

In vivoFitResiduals


Glu (90 plusmn 02)

Gln (24 plusmn 02)

Normal brain

3 2 14

(ppm)

(a)


ChoCr NAA

Lac Lip


Glu (46 plusmn 02)

Gln (53 plusmn 02)

3 2 14

(ppm)

(b)


Cho

Cr NAA Lac

2HG (33 plusmn 02)


Gln (29 plusmn 02)

3 2 14

(ppm)

(c)


Cho

Cr NAA Lac

2HG (91 plusmn 03)

GABA (02 plusmn 02)

Glu (05 plusmn 02)

Gln (02 plusmn 02)

3 2 14

(ppm)

(d)


Gly

Cho

CrNAA Lac

2HG (89 plusmn 03)

GABA (0)

Glu (21 plusmn 02)

Gln (23 plusmn 02)

3 2 14

(ppm)

(e)


ChoCr NAA

Lip

2HG (46 plusmn 03)

GABA (01 plusmn 01)

Glu (14 plusmn 02)

Gln (26 plusmn 02)

3 2 14

(ppm)

(f)





















Acknowledgments


References

















































































































































































































of






Disease Markers



OncologyJournal of





PPAR Research



Journal of

ObesityJournal of

















Cr ChoCr

NAA

In vivoFitResiduals


Glu (90 plusmn 02)

Gln (24 plusmn 02)

Normal brain

3 2 14

(ppm)

(a)


ChoCr NAA

Lac Lip


Glu (46 plusmn 02)

Gln (53 plusmn 02)

3 2 14

(ppm)

(b)


Cho

Cr NAA Lac

2HG (33 plusmn 02)


Gln (29 plusmn 02)

3 2 14

(ppm)

(c)


Cho

Cr NAA Lac

2HG (91 plusmn 03)

GABA (02 plusmn 02)

Glu (05 plusmn 02)

Gln (02 plusmn 02)

3 2 14

(ppm)

(d)


Gly

Cho

CrNAA Lac

2HG (89 plusmn 03)

GABA (0)

Glu (21 plusmn 02)

Gln (23 plusmn 02)

3 2 14

(ppm)

(e)


ChoCr NAA

Lip

2HG (46 plusmn 03)

GABA (01 plusmn 01)

Glu (14 plusmn 02)

Gln (26 plusmn 02)

3 2 14

(ppm)

(f)





















Acknowledgments


References

















































































































































































































of






Disease Markers



OncologyJournal of





PPAR Research



Journal of

ObesityJournal of
































Acknowledgments


References

















































































































































































































of






Disease Markers



OncologyJournal of





PPAR Research



Journal of

ObesityJournal of



















































































































































































































of






Disease Markers



OncologyJournal of





PPAR Research



Journal of

ObesityJournal of
















cancer metabolism and tumor heterogeneity: …downloads.hindawi.com › journals › cmmi › 2017...

Documents