You are on page 1of 7

Original article

Diffusion-weighted magnetic resonance imaging in

metastatic gastrointestinal stromal tumor (GIST): a pilot study
on the assessment of treatment response in comparison
with 18F-FDG PET/CT

Sabine Schmidt1, Vincent Dunet2, Melanie Koehli1, Michael Montemurro3, Reto Meuli1
and John O Prior2
Department of Radiology; 2Department of Nuclear Medicine; 3Department of Oncology, Centre Hospitalier Universitaire Vaudois (CHUV),
Lausanne, University of Lausanne, Lausanne, Switzerland
Correspondence to: Sabine Schmidt. Email:

Background: Diffusion-weighted magnetic resonance imaging (MRI) is increasingly being used for
assessing the treatment succes in oncology, but the real clinical value needs to evaluated by comparison
with other, already established, metabolic imaging techniques.
Purpose: To prospectively evaluate the clinical potential of diffusion-weighted MRI with apparent diffusion
coefficient (ADC) mapping for gastrointestinal stromal tumor (GIST) response to targeted therapy compared
with 18F-fluorodeoxyglucose positron emission tomography/computed tomography (18F-FDG PET/CT).
Material and Methods: Eight patients (mean age, 56 + 11 years) known to have metastatic GIST underwent
18F-FDG PET/CT and MRI (T1Gd, DWI [b ¼ 50,300,600], ADC mapping) simultaneously, before and after
change in targeted therapy. MR and PET/CT examinations were first analyzed blindly. Second, PET/CT
images were co-registered with T1Gd-MR images for lesion detection. Only 18F-FDG avid lesions were
considered. Maximum standardized uptake value (SUVmax) and the corresponding minimum ADCmin were
measured for the six largest lesions per patient, if any, on baseline and follow-up examinations. The
relationship between changes in SUVmax and ADCmin was analyzed (Spearman’s correlation).
Results: Twenty-four metastases (12 hepatic, 12 extra-hepatic) were compared on PET/CT and MR
images. SUVmax decreased from 7.7 + 8.1 g/mL to 5.5 + 5.4 g/mL (P ¼ 0.20), while ADCmin increased
from 1.2 + 0.3  1023mm2/s to 1.5 + 0.3  1023mm2/s (P ¼ 0.0002). There was a significant association
between changes in SUVmax and ADCmin (rho ¼ 2 0.62, P ¼ 0.0014), but not between changes in lesions
size (P ¼ 0.40).
Conclusion: Changes in ADCmin correlated with the response of 18F-FDG avid GIST to targeted therapy.
Thus, diffusion-weighted MRI may represent a radiation-free alternative for follow-up treatment for
metastatic GIST patients.

Keywords: Gastrointestinal stromal tumor (GIST), MR diffusion imaging, 18F-FDG PET/CT, apparent diffusion
coefficient (ADC), standard uptake value (SUV)

Submitted September 19, 2012; accepted for publication March 3, 2013

Gastrointestinal stromal tumors (GISTs) are the most factor receptor, permits their distinction from other
common mesenchymal neoplasms of the intestine and mesenchymal neoplasms, such as leiomyomas and leio-
may be localized anywhere throughout the digestive tract, myosarcomas. This allows for targeted therapy with
mesentery, omentum, and retroperitoneum (1). In approxi- KIT-inhibitor agents (2, 3), such as imatinib, sunitinib, or
mately 95% of patients, their expression of the protein sorafenib, which have a significant impact on the outcome.
KIT (CD117), a transmembraneous tyrosine kinase growth After initial surgery for the localized disease, GISTs have

Acta Radiologica 2013; 54: 837–742. DOI: 10.1177/0284185113485732

838 S Schmidt et al.

a high risk of metastatic relapse, in particular in the liver informed consent form prior to study enrolment. Inclusion
and peritoneum (2, 4). criteria were: (i) patients previously treated for a GIST;
Traditionally, intravenous contrast-enhanced computed (ii) known to have hypermetabolic hepatic and/or extra-
tomography (CT) has been used for the detection, evalu- hepatic metastases on 18F-FDG PET/CT; and (iii) currently
ation, and follow-up of the primary tumor, including progressing clinically or radiologically despite therapy and
metastases (5). It may be complemented by magnetic reson- for whom a therapy change was considered necessary.
ance imaging (MRI), known for its higher sensitivity in Eight consecutive patients (5 women, 3 men; mean age,
detecting hepatic metastases (4). CT and MRI accurately 56 + 11 years) known for metastatic GIST underwent one
demonstrate the morphological tumor features (2, 4– 7), paired 18F-FDG PET/CT and MRI examination before and
but assessment of the metabolic activity remains difficult. one paired 18F-FDG PET/CT and MRI examination after
Therefore, for the clinical follow-up of these patients, change in targeted treatment to evaluate the response of
18F-fluorodeoxyglucose positron emission tomography/ the secondary lesions to KIT-inhibitor therapy.
computed tomography (18F-FDG PET/CT) is increasingly Mean elapsed time between the baseline and follow-up
being used (8, 9). At present, 18F-FDG PET/CT yields the examinations (MRI and 18F-FDG PET/CT) was 4.1 + 2.8
most accurate evaluation of GISTs’s response to targeted months (range, 2 – 6). Mean delay between each pair of
medical treatment, more so than CT. The results of the MRI and 18F-FDG PET/CT was 2.6 + 3.6 days (range, 0– 9).
latter are merely based on the measurement of the lesion’s Primary tumors had the following locations: stomach
largest diameter, according to RECIST (Response (n ¼ 3), extraintestinal (mesentery n ¼ 1 and mesocolon
Evaluation Criteria In Solid Tumors) (10, 11), thus also n ¼ 1), small bowel (n ¼ 2), and unknown (n ¼ 1).
including necrotic areas of the tumor (12). Furthermore, Metastases were present in the liver only (n ¼ 3), peritoneal
hepatic and peritoneal metastases resulting from GISTs only (n ¼ 1), or at an intra- and an extrahepatic, essentially
typically undergo cystic transformation following the peritoneal location (n ¼ 4). Five patients had received first-
response to targeted therapy. This cystic transformation is line therapy (imatinib), which, after the first MRI and
usually associated with a decrease in CT attenuation 18F-FDG PET/CT examinations, was changed to second-
values, but often without significant decrease or even with line (sunitinib) (n ¼ 3) or to fourth-line (sorafenib) (n ¼ 1)
a slight increase in size (13). In 2007, Choi et al., taking treatment. In three patients, the change in therapy consisted
into account the limited value of the size changes, intro- in an increased dose of the KIT-inhibitor agent (imitanib
duced new criteria, including the measurement of the CT in first-line [n ¼ 2] and sorafenib in fourth-line [n ¼ 1]
attenuation coefficient (in Hounsfield units [HU]), that treatment).
was better correlated with treatment response at 3 months Exclusion criteria were known contraindications to MRI
(as measured by PET) than the purely morphological ( pacemaker, claustrophoby), renal failure, and having a
RECIST criteria (12, 14). PET-negative GIST during the baseline examination.
However, the regular availability of 18F-FDG PET/CT
may still be limited and the high costs as well as the radi-
ation exposure invariably associated with this modality Technical parameters of MRI and 18F-FDG PET/CT
need to be taken into account. Diffusion-weighted MR MR data were acquired on a 3.0 Tesla MR scanner (TRIO
imaging (DW-MRI) of the abdomen is widely used today, or VERIO, Siemens Healthcare, Erlangen, Germany) with
and is being considered for oncologic imaging because a maximum gradient strength of 40 mT/m. To include
of its high detection sensitivity for small lesions of high the whole abdomen in our examination protocol we com-
cellular density, as well as because of the biological in- bined two 12-channel phased-array body coils simul-
formation obtained by measuring the apparent diffusion taneously fixed around the upper and lower abdomen,
coefficient (ADC) (15). respectively.
To our knowledge, the role of DW-MRI in the metabolic After fasting for at least 6 h prior to MRI all patients were
treatment evolution of GIST in direct comparison with intravenously (i.v.) injected with 20 mg of scopolaminbutyl-
18F-FDG PET/CT has not yet been investigated. bromide (Buscopanw, Boehringer Ingelheim, Basel,
The purpose of our study was, therefore, to evaluate Switzerland) or, if that was contraindicated, 1 mg of gluca-
DW-MRI with apparent diffusion coefficient (ADC) map- gon (GlucaGenw, Novo Nordisk, Bagsvaerd, Denmark), to
ping as an assessment tool to measure the response reduce imaging artifacts caused by peristalsis. The MR
of metastatic GIST to a change in targeted therapy, as acquisition protocol, included the whole abdomen. Axial
compared to 18F-FDG PET/CT. three-D-VIBE (three-dimensional volumetrical interpolation
breath-hold examination) MR-sequences (TR, 3.62 ms;
TE, 1.32 ms; flip angle, 108; pixel size, 1.8  1.5 mm; NEX 1;
Material and Methods slice thickness, 2 mm; parallel imaging [iPAD (integrated
parallel acquisition technique)] GRAPPA [generalized auto-
Patients calibrating partially parallel acquisitons], 2) were performed
This pilot study complied with local and national ethical during breath-hold before and after an i.v. gadolinium
guidelines and was approved in its present form by our (Gd-)-DOTA injection, acquired in a dynamic (arterial,
institutional ethical board. Thus, it has been performed portal, and venous) phase centered on the upper abdomen
in accordance with the ethical standards laid down in (Dotaremw, 0.2 mmol/kg of body weight), followed by
the 1964 Declaration of Helsinki. Each patient signed an a 40 mL flush of 0.9% NaCl saline.
Diffusion-weighted magnetic resonance imaging in metastatic gastrointestinal stromal tumor (GIST) 839

DW-MRI was acquired before the i.v. gadolinium injec- metastatic lesions with increased 18F-FDG uptake, both in
tion while the patients breathed freely but shallowly (15). the baseline and the follow-up examinations. Then, 18F-
The DW-MRI included transverse single-shot spin-echo FDG PET/CT images were co-registered with Gd-enhanced
echo-planar sequences (TR, 7700 ms; TE, 66 ms; flip angle, VIBE-MR images for lesion detection (PFUS, PMOD Tech-
908; pixel size, 3.5  2.8 mm; NEX 6; slice thickness, 6 mm; nologies, Zürich, Switzerland).
parallel imaging [iPAD] GRAPPA 2) in three orthogonal In consensus, the radiologists and nuclear physicians
directions (frequence-encoding, phase-encoding, and slice- identified up to six mestastatic lesions in each patient,
selection) with three b-values (50, 300, 600 s/mm2). These including the largest ones visible on MRI, provided that
b-values were chosen in order to obtain images with a suffi- they showed pathological SUVmax on the baseline 18F-
cient contrast-to-noise ratio. Six averages were performed. FDG PET/CT. They carefully verified that the measure-
Fat suppression was implemented by the spectral adiabatic ments of the ADCmin and SUVmax in each lesion had been
inversion recovery (SPAIR) technique. By inverting the centered on the same intratumoral location.
fat spins in the imaging volume SPAIR allows for more
profound and homogenous fat saturation compared to
conventional fat suppression techniques. Statistical analysis
Trace images were synthesized for each b-value, and an
ADC map was then automatically computed (with a The relationship between the SUVmax and ADCmin of each
vendor-provided software) from all diffusion weightings lesion as well as their changes over time were assessed
and directions. Window width and level were set to ade- with non-parametric Spearman’s rank correlation.
quately visualize the whole abdomen. We compared the lesion size, and assessed the relation-
All patients were refrained from food 6 h and had ship between SUVmax and ADCmin before and after targeted
a glucose plasma level 6.1 mmol/L before the 18F-FDG therapy using the Wilcoxon rank sum test.
intravenous injection. PET/CT (Discovery LS, GE Data were expressed as mean + SD. Statistical analyses
Healthcare Milwaukee, WI, USA) included a whole-body were performed by using Stata 11.1 and all differences
acquisition (from the skull base to mid-thighs) performed were considered significant for a P value ,0.05.
60 min after an i.v. injection of 5.5 MBq/kg of 18F-FDG.
PET acquisition was preceded by a craniocaudal unen-
hanced acquisition of MDCT (4-row-detector) used for
attenuation correction and localization (140 kV, 80 mA; Results
pitch, 1.5; 0.5 s/rotation; 5-mm slice thickness). PET data Thirty metastatic lesions (17 hepatic and 13 extra-hepatic)
were subsequently reconstructed using an ordered-subset were analyzed on both modalities. Two PET/CT lesions
expectation maximization method with eight subsets and could not be evaluated on ADC maps, because of their
two iterations. infracentimetric size, and four lesions were excluded
because they were too close to other areas of increased phys-
iological uptake (bile ducts or renal parenchyma) seen on
Image analysis the follow-up PET/CT. Therefore, 24 lesions (12 hepatic
MRI and 18F-FDG PET/CT studies were first analyzed and 12 extrahepatic) were analyzed on both imaging modal-
separately. Using an Advantage Window workstation ities and compared before and after targeted therapy.
with version 4.3 software (GE Healthcare, Buc, France) Their largest lesion diameter ranged from 12 to 178 mm
two radiologists read the baseline and follow-up MR (mean, 49 + 42 mm) before, and from 8 to 103 mm (mean,
images in consensus. Lesions with evident artifacts on 40 + 27 mm) after treatment, which was not a statistically
DW images and infracentimetric metastases were excluded significant change (P ¼ 0.60).
from the analysis (n ¼ 5). The largest pre- and post- On MRI, active GIST metastases exhibited high signal
treatment size of each lesion was measured on axial intensity on DW images and low signal intensity on ADC
Gd-enhanced VIBE MR-images. After a circular ROI was maps (Fig. 1), except for areas with necrosis or with cystic
manually drawn within the most restricted area visible on or myxoid transformation.
the highest b-value DW-image, the ROI was then automati- Taking the change of SUVmax as the accepted imaging
cally transferred onto the ADC map, and the average value parameter for the follow-up of GIST metastases (9, 12, 13, 16),
of ADCmin indicated within this ROI was registred for each the response to therapy was classified as partial response
metastasis. Furthermore, the standard deviation (SD) inside (significant activity decrease 25%) in 12 (50%) out of the
this ROI was assessed, as a surrogate measure for the 24 measured lesions, stable disease (no significant change)
heterogenicity of the lesion. By constantly comparing the in 11 (46%), and progressive disease in one lesion (4%)
intralesional position of the ROI with the correspondent showing an increase in activity of SUVmax.
T2-weighted and Gd-T1-weighted MR images, we were Overall SUVmax was 7.7 + 8.1 g/mL before and 5.5 +
certain that we were measuring the tumoral tissue. This 5.4 g/mL after the therapy change, while the overall
method avoided the inclusion of cystic and necrotic areas ADCmin was 1.2 + 0.3  1023mm2/s before and 1.5 +
in our measurements (15). 0.3  1023mm2/s after the change in treatment.
18F-FDG PET/CT was read by two nuclear physicians in The evolution of these 24 lesions showed significant cor-
consensus and the maximum standardized uptake value relation between SUVmax decrease and ADCmin increase
(SUVmax) corrected for body mass was measured in (rho ¼ 2 0.62, P ¼ 0.0014) (Fig. 2).
840 S Schmidt et al.

Fig. 1 A 52-year-old man with metastatic GIST before and after targeted therapy. The upper row images show a left pararenal metastasis (arrows) before
therapy, while the lower row images demonstrate the evolution 3 months after treatment (arrows): the first column demonstrates axial Gd-enhanced VIBE MR
images that reveal metastatic maximal size reduction of 20%. However, according to RECIST, this does not even indicate a partial response, since a decrease
of at least 30% is required [10, 11]. In the second column, axial 18F-FDG PET/CT-images also show tumor regression because of decreasing SUVmax from 6 g/mL
to 1.9 g/mL. This corresponds to an increase of ADC (third column) from 0.89  1023mm2/s before (upper row) to 1.41  1023mm2/s after (lower row) treatment.
The forth column demonstrates the co-registration of the 18F-FDG PET/CT images to Gd-enhanced VIBE MR images to enable accurate lesion detection

Traditionally, tumor response to treatment has been

evaluated on the basis of morphologic features established
by the RECIST group (10, 11). Cystic lesions are considered
non-measurable lesions, and thus excluded from this evalu-
ation (14). However, some targeted therapy agents have
an antiangiogenic effect primarily leading to intratumoral
changes, such as decrease of lesion-to-vessel and lesion-
to-muscle ratio, radiologically characterized as cystic trans-
formation, rather than as a shrinkage of the lesion size
(13). Thus, unchanged lesion size after treatment with
targeted agents does not automatically indicate stable
disease and metabolic imaging modalities are needed,
since pure morphological criteria underestimate tumor
response to treatment.
The uptake of 18F-FDG detected on PET/CT represents
tumor cell viability, and after effective treatment, the
reduction of 18F-FDG uptake reflects the tumor cell killing
rate (16). At present, dual modality 18F – FDG PET/CT has
become a standard modality in the assessment of treatment
response for metastatic GIST (8, 9).
However, PET alone has been shown to be less sensitive
Fig. 2 Correlation diagram. The association between the difference of for lesion detection in metastatic GIST than dual modalitiy
SUVmax value (x-axis) and the difference of ADC value (y-axis) before and
after targeted treatment is shown for each of the 24 intra- and extrahepatic
18F-FDG PET/CT (17, 18). Furthermore, patients with no
metastases we included in our evaluation appreciable 18F-FDG uptake in pre-treatment examination
continue to be observed on 18F-FDG PET/CT (19), since
in up to 20% of GIST patients the lesions do not significantly
There was no statistically significant correlation between
accumulate the 18F-FDG tracer higher than the background
the evolution of the lesions’ size and the ADCmin values
on baseline images (the so-called “18F-FDG-non-avid
(rho ¼ 2 0.16, P ¼ 0.40)
GIST”) (13).
Thanks to the development of diffusion-weighted
Discussion sequences, abdominal MRI today not only provides mor-
In patients with metastatic GIST, the accurate assessment phological information, but also enables the detection of
of response to targeted therapy is crucial, because of the metabolic tumor changes. Classically, restricted diffusion
high cost, selective character, and systemic toxic secondary has been linked with high cellular density, presented by
effects of the KIT-inhibitor agents. active tumors, whereas free molecular diffusion occurs in
Diffusion-weighted magnetic resonance imaging in metastatic gastrointestinal stromal tumor (GIST) 841

necrotic or cystic lesions (15). Second, this diffusion- only, possibly merely describing a “trend”. Therefore, the
weighted signal is quantified by calculating the apparent latter now needs to be confirmed in a larger trial, a multi-
diffusion coefficient (ADC) that serves as a surrogate centric one.
marker of tissular cellularity. A low ADC value means a Second, our population was composed of patients with
tumor of dense cellularity, unlike most necrotic or cystic long-standing GIST, already treated prior to inclusion in
lesions with high ADC values (15). An increase in ADC our study. Thus, our lesions were quite heterogeneous.
after successful treatment has been reported for several However, we took great care to adequately place the ROI
abdominopelvic neoplasias (20 –22), which has been for ADC measurements outside these degenerative areas
attributed to cell apoptosis (15). and inside the active tumor tissue, always in perfect agree-
Tang et al. have observed that the ADC of GIST lesions ment with the conventional T1-weighted Gd-enhanced
significantly increases after successful molecularly targeted and T2-weighted MR images. Also, we used automatic
treatment (23), however, without comparing these results co-registration of MR images with 18F –FDG PET/CT
with SUVmax values obtained by 18F-FDG PET/CT. images in order to be sure that the ROI placements for
Furthermore, the main goal of Tang et al. was to differen- ADC and for SUVmax measurements were identical (Fig. 1).
tiate between responders and non-responders to targeted Finally, our mean interval between the pre- and post-
therapy by comparing pre- and post-treatment MRI. The treatment change MRI/18F –FDG PET/CT was 4 months,
authors focussed on patients with newly diagnosed GIST which is quite long compared to previous similar trials
(23). The tumor stage of our study population with meta- (20, 21, 23). However, Tang et al. showed that the well-
static GIST was more advanced and more heterogeneous. responding GIST lesions display a steady increase of ADC
Because of exhibited resistance to the individual KIT- values in cases of effective treatment, irrespective of the
inhibitor treatment detected in our baseline MRI and time delay (23).
18F-FDG PET/CT, each of our patients needed a change Thus, MRI with diffusion-weighted sequences represents
of therapy, which was our main inclusion criterion. This a radiation-free alternative for controling GIST metastases
advanced tumor stage present in our patients may also in patients needing multiple follow-up examinations. The
explain, why in only 12 (50%) of our 24 included lesions proven association between SUVmax and ADCmin could
did the metabolic activity significantly decrease, whereas be particularly useful for assessing treatment success in
in 11 (48%) it remained unchanged. patients with 18F –FDG-non-avid GIST. The exact diagnostic
Several authors have correlated DW-MRI with 18F-FDG value of ADC needs to be investigated further, as does the
PET/CT at initial tumor stage, prior to treatment. Various effect of lesion size and time under therapy before imaging.
newly diagnosed and pathologically confirmed malignant In conclusion, changes in ADCmin derived from DW-MRI
tumors were included in these studies (24– 29). The two seem to reflect the response of 18F-FDG-avid GIST lesions to
techniques were correlated either visually by simple image a change in targeted therapy. These results may demon-
fusion without any measurements (24, 28), or by measuring strate the excellent potential of ADCmin as a metabolic
the mean or the maximal SUV, respectively, and comparing imaging modality. However, larger studies are needed to
it with the mean or minimal ADC value or even with an find out the exact role of DW-MRI in the management of
ADC ratio, respectively (25– 27, 29). Most of these results patients with metastatic GIST treated with targeted agents.
show a definite correlation between SUV and ADC,
however, unlike in our study, no measurements over time Conflict of interest: None.
that is, during or after tumor treatment, were performed
in these studies.
To the best of our knowledge, there has only been one REFERENCES
direct comparison between the change in MR-DWI and
1 Levy AD, Remotti HE, Thompson W, et al. Gastrointestinal stromal
18F – FDG PET/CT parameters for tumoral treatment
tumors: radiologic features with pathologic correlation. Radiographics
response. Evaluating the success of chemotherapy in eight 2003;23:283 –304
patients with diffuse large B-cell lymphoma Wu et al. 2 Blay JY, Bonvalot S, Casali P, et al. Consensus meeting for the
found a significant inverse correlation between central management of gastrointestinal stromal tumors. Report of the GIST
and peripheral ADC values and SUVmax in baseline examin- consensus conference of 20 –21 march 2003, under the auspices of ESMO.
Ann Oncol 2005;16:566 – 78
ations (P ,0.05), as well as in imaging performed 1 week
3 Miettinen M, Lasota J. Gastrointestinal stromal tumours – definition,
after starting treatment (P , 0.05) (30). However, instead of clinical, histological, immunohistochemical, and molecular genetic
evaluating several ADC values for each lesion, our tech- features and differential diagnosis. Virchows Arch 2001;438:1 –12
nique consisted in measuring one ADCmin value through 4 Sandrasegaran K, Rajesh A, Rushing DA, et al. Gastrointestinal stromal
a manually contoured ROI centered within the most tumors : CT and MRI findings. Eur Radiol 2005;15:1407 – 14
5 Tateishi U, Hasegawa T, Satake M, et al. Gastrointestinal stromal tumor.
restricted area visible on the DWI image, which, after Correlation of CT findings with tumor grade and mortality. J Comput
fusion of 18F-FDG PET/CT images with MR images, Assist Tomogr 2003;27:792 –8
corresponded to the area of the SUVmax evaluation. Thus, 6 Caramella T, Schmidt S, Chevallier P, et al. MR features of
we obtained the closest resemblance between the two gastrointestinal stromal tumors. J Clin Imaging 2005;29:251 – 4
measurements. 7 Stroszczynski C, Jost D, Reichardt P, et al. Follow-up of gastro-intestinal
stromal tumours (GIST) during treatment with imatinib mesylate
There are some limitations in our study. First, since by abdominal MRI. Eur Radiol 2005;15:2448 – 56
we conducted a pilot study, our population is quite small 8 Goerre GW, Stupp R, Barghouth G, et al. The value of PET, CT
and our results show a statistically moderate correlation and in-line PET/CT in patients with gastrointestinal stromal
842 S Schmidt et al.

tumours: long-term outcome of treatment with imatinib mesylate. 20 Kamel IR, Liapi E, Reyer DK, et al. Unresectable hepatocellular
Eur J Nucl Med Mol Imaging 2005;32:153 –62 carcinoma. Serial early vascular and cellular changes after transarterial
9 Holdsworth CH, Badawi RD, Manola JB, et al. CT and PET: Early chemoembolization as detected with MR Imaging. Radiology
prognostic indicators of response to imatinib mesylate in patients with 2009;250:466 – 73
gastrointestinal stromal tumor. Am J Roentgenol 2007;189:W324– 30 21 Cui Y, Zhang XP, Sun YS, et al. Apparent diffusion coefficient:
10 Eisenhauer EA, Therasse P, Bogaerts J, et al. New response evaluation potential imaging biomarker for prediction and early detection of
criteria in solid tumours: revised RECIST guideline (version 1.1.). response to chemotherapy in hepatic metastases. Radiology 2008;
Eur J Cancer 2009;45:228 – 47 248:894 – 900
11 Therasse P, Arbuck SG, Eisenhauer E, et al. New guidelines to evaluate 22 Kim SH, Lee JM, Hong SH, et al. Locally advanced rectal cancer: added
the response to treatment in solid tumors. European Organization for value of diffusion-weighted MR imaging in the evaluation of tumor
Research and Treatment of Cancer, National Cancer Institute of the response to neoadjuvant chemo-and radiation therapy. Radiology
United States, National Cancer Institute of Canada. J Nat Cancer Inst 2009;253:116 – 25
2000;92:205 –16 23 Tang L, Zhang X-P, Sun Y-S, et al. Gastrointestinal stromal tumors treated
12 Choi H, Charnsangavej C, Faria SC, et al. Correlation of computed with imatinib mesylate: apparent diffusion coefficient in the evaluation
tomography and positron emission tomography in patients with of therapy response in patients. Radiology 2011;258:729 –38
metastatic gastrointestinal stromal tumor treated at a single institutuion 24 Reiner CS, Fischer MA, Hany T, et al. Molecular imaging of malignant
with imatinib mesylate : proposal of new computed tomography tumor metabolism whole-body image fusion of DWI/CT vs. PET/CT.
response criteria. J Clin Oncol 2007;25:1753 –9 Acad Radiol 2011;18:940 – 6
13 Choi H, Charnsangavej C, de Castro Fario S, et al. CT evaluation of 25 Choi SH, Paeng JC, Sohn CH, et al. Correlation of 18F-FDG uptake
the response of gastrointestinal stromal tumors after imatinib mesylate with apparent diffusion coefficient ratio measured on standard and
treatment: a quantitative analysis correlated with FDG PET findings. high b value diffusion MRI in head and neck cancer. J Nucl Med
Am J Roentgenol 2004;183:1619 – 28 2011;52:1056 – 62
14 Benjamin RS, Choi H, Macapinla HA, et al. We should desist using 26 Gu J, Khong P-L, Wang S, et al. Quantitative assessment of
RECIST, at least in GIST. J Clin Oncol 2007;25:1760 –4 diffusion-weighted MR imaging in patients with primary rectal cancer:
15 Padhani AR, Liu G, Mu-Koh D, et al. Diffusion-weighted magnetic correlation with FDG-PET/CT. Mol Imaging Biol 2011;13:1020 – 8
resonance imaging as a cancer biomarker: consensus and 27 Ho KC, Lin G, Wan J-J, et al. Correlation of apparen diffusion
recommendations. Neoplasia 2009;11:102 –25 coefficients measured by 3T diffusion-weighted MRI and SUV from
16 Young H, Baum R, Cremerius U, et al. Measurement of clinical and FDG PET/CT in primary cervical cancer. Eur J Nucl Med Mol Imaging
subclinical tumour response using [18F]-fluorodeoxyglucose and 2009;36:200 –8
positron emission tomography: review and 1999 EORTC 28 Abdulqadhr G, Molin D, Aström G, et al. Whole-body
recommendations. Eur J Cancer 1999;13:1773 –82 diffusion-weighted imaging compared with FDG-PET/CT instaging
17 Antoch G, Kanja J, Bauer S, et al. Comparison of PET, CT and dual of lymphoma patients. Acta Radiol 2011;52:173 – 80
modality PET/CT imaging for monitoring of imatinib (STI571) therapy 29 Wong CS, Gong N, Chu Y-C, et al. Correlation of measurements from
in patients with gastrointestinal stromal tumors. J Nucl Med diffusion weighted MR imaging and FDG PET/CT in GIST patients:
2004;45:357 –65 ADC versus SUV. Eur J Radiol 2012;81:2122 –6
18 Otomi Y, Otsuka H, Morita N, et al. Relationship between FDG uptake 30 Wu X, Kellokumpu-Lehtinen P.K, Pertovaara H, et al.
and the pathological risk category in gastrointestinal stromal tumors. Diffusion-weighted MRI in early chemotherapy response evaluation of
J Med Intest 2010;57:270 – 4 patients with diffuse large B-cell lymphoma – a pilot study: comparison
19 Van den Abbeele AD. The lessons of GIST-PET and PET/CT: a new with 2-deoxy-2-fluoro-D-glucose-positron emission tomography/
paradigm for imaging. Oncologist 2008;13:8– 13 computed tomography. NMR Biomed 2011;24:1181 – 90
Copyright of Acta Radiologica is the property of Sage Publications, Ltd. and its content may
not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's
express written permission. However, users may print, download, or email articles for
individual use.