EffectofAngiosonographytoMonitorResponseDuring ...disease, the imatinib dose was increased to 800 mg/d, if possible, according to the tolerance and compliance of patients. In cases

Effect of Angiosonography toMonitor Response DuringImatinib Treatment in PatientswithMetastaticGastrointestinal Stromal TumorsUgoDe Giorgi,1Camillo Aliberti,3 Giorgio Benea,3Matteo Conti,2 andMaurizioMarangolo1

Abstract Purpose: Gastrointestinal stromal tumor (GIST) metastases are typically intra-abdominal andhypervascular.We assessed the effect of angiosonography with a second-generation contrastagent to monitor response during imatinib treatment in patients withmetastatic KIT+GIST.Experimental Design:Ten consecutive patients with known advanced KIT+ GISTwere investi-gated with angiosonography and computerized tomography (CT).We also monitored the serumlevels of the major angiogenic growth factor, vascular endothelial growth factor.Results: Angiosonography showed a reduction in tumor vascularization of liver metastasesduring imatinib treatment in all cases.We observed a reduction in tumor vascularization before areduction in tumor size. The tumor perfusion appeared reduced in the central part of the livermetastases.With amedian follow-up of18months (range 3-33), a reduction in tumor vasculariza-tionwas initially observed in all patients, but progressive diseasewas documented in four patientsfollowing imatinib treatment. CTdocumented tumor response according to standardized criteria insix patients, stable disease in four, and progressive disease according to angiosonography. Thereduction of tumor perfusion at angiosonography correlated with the pseudocystic appearanceat CT. The ‘‘nodule(s) within amass’’pattern of recurrence occurred in twopatients withno differ-ence observed between angiosonography and CT. Early decreasing serum vascular endothelialgrowth factor levels were observed in the two cases withhigher pretreatment levels.Conclusions: Imatinib could induce antiangiogenic and/or antivascular effects in GIST, and thiseffect couldbe easilymonitoredwith angiosonography. Angiosonographymight be auseful com-plement for monitoring the therapeutic effect of imatinib in these patients.

The use of imatinib in gastrointestinal stromal tumors (GIST)represents the first effective development of a targeted therapyfor a solid neoplasm. Imatinib produces objective responses innearly 50% to 55% of patients (median time to response, 3-4months; range 1-12), and stable disease in 30% to 35% asmeasured with computerized tomography (CT) or magneticresonance by standardized criteria (Southwest Oncology Groupcriteria, and Response Evaluation Criteria in Solid Tumors; refs.1–3). These results correlated closely with findings of responsebased on metabolic functional imaging with fluorine-18

fluorodeoxyglucose positron emission tomography (PET) in80% to 90% of patients, before any measurable changes wereseen on CT (3, 4).

GIST metastases are typically intra-abdominal and hyper-vascular (5). The accuracy of ultrasound in the detection ofintra-abdominal metastases from GIST is lower than CT ormagnetic resonance, but the introduction of contrast agents ledto new possibilities for functional imaging studies. BR1(Sonovue, Bracco, Milan, Italy) is a new blood pool ultrasoundsecond-generation contrast agent, which consists of stabilizedmicrobubbles of a totally innocuous sulfur hexafluoride gas,which is of low solubility. It is isotonic to human plasma anddevoid of antigenic potential, as it does not contain anyproteineous material (6). BR1 allows angiosonography throughcontinuous real-time examination during different vascularphases of contrast enhancement using low transmission power,expressed as mechanical index. We evaluated the effect ofangiosonography with BR1 to monitor response duringimatinib treatment (400 mg oral administration, once daily)in patients with metastatic GIST.

Patients andMethods

Patients. Between May 2002 and February 2004, 10 consecutivepatients (8 males and 2 females) with known advanced KIT+ GIST wereinvestigated with angiosonography and CT. Median age was 66 years(range, 35-85). A pathologic review in other centers was done in all

Imaging, Diagnosis, Prognosis

Authors’Affiliations: 1Istituto Oncologico Romagnolo, Department of Oncologyand 2Laboratory of Clinical Pharmacology andToxicology, Santa Maria delle CrociHospital, Ravenna; and 3Department of Imaging, Delta Hospital, Lagosanto, Ferrara,ItalyReceived10/6/04; revised 5/25/05; accepted 6/3/05.Grant support: Istituto Oncologico Romagnolo, ForlI' , Italy.The costs of publication of this article were defrayed in part by the payment of pagecharges.This article must therefore be hereby marked advertisement in accordancewith18 U.S.C. Section1734 solely to indicate this fact.Note: Presented at the 16th European Organization for Research andTreatment ofCancer-National Cancer Institute-AACR Symposium on ‘‘Molecular Targets andCancer Therapeutics’’, Geneva, Switzerland, September 28 to October1, 2004.Requests for reprints: Ugo De Giorgi, Istituto Toscano Tumori-Departmentof Oncology, San Giuseppe Hospital,Via Paladini 40, I-50053 Empoli, Florence,Italy. Phone: 39-349-222-1235; Fax: 39-0571-702671; E-mail: [email protected].

F2005 American Association for Cancer Research.doi:10.1158/1078-0432.CCR-04-2046

www.aacrjournals.org Clin Cancer Res 2005;11(17) September1, 20056171

Cancer Research. on October 17, 2020. © 2005 American Association forclincancerres.aacrjournals.org Downloaded from

http://clincancerres.aacrjournals.org/

cases. No patient was previously treated with imatinib. Patient

characteristics before imatinib treatment are listed in Table 1. Inaddition, nine more patients were studied with both techniques in the

follow-up after complete resection of GIST.Imatinib treatment. All patients were treated with imatinib at the

starting dose of 400 mg/d. The imatinib dose was modified during the

treatment in patients with severe toxicities. In these cases, the treatmentwas briefly interrupted until toxicity recovery and then was restarted

with the imatinib dose of 300 mg/d. In patients with progressive

disease, the imatinib dose was increased to 800 mg/d, if possible,according to the tolerance and compliance of patients. In cases with

minimal progressive disease with unresectable disease, continuingimatinib treatment at 400 mg/d was considered. In patients with

massive progressive disease, the imatinib treatment was stopped.Imaging techniques. The ultrasound examinations were done by

means of a Sequoia digital ultrasound scanner (Acuson, Mountain

View, CA) equipped with a wideband C5-2 MHz convex probe. Before

contrast agent injection, the scanning plane of the abdomen thatincluded the nodule(s) was determined. After the baseline ultrasound

examination, all patients underwent angiosonography using BR1.Nonlinear imaging was done starting with high mechanical index and

then reducing to a continuous low acoustic power (mechanical index =

0.20) for each patient. Twelve milliliters of a 5 mg/mL solution of BR1were administered as an i.v. bolus. The whole vascular phase was

studied, consisting of the arterial phase (15-30 seconds after theinjection), the portal phase (30-60 seconds), and the sinusoidal phase

(60-240 seconds). A second injection of BR1 was needed in some cases.

Moreover, patients underwent multi-phase helical CT (Mx8000,Philips Medical Systems, Eindhoven, the Netherlands). Each patient

received a bolus of 120 mL of low-osmolar iodinated contrast medium(Iomeron 400, Iomeprol, Bracco, Milan, Italy).

PET scan was not regularly done. However, PET was done in cases

with changes in tumor features at angiosonography and/or CT as in thecase of progressive disease. PET scans were carried out using a PET

tomograph (Advance scanner, GE Medical System, Waukesha, WI). PET

was carried out after i.v. injection of 370 MBq of fluorodeoxyglucose;images were recorded after 60 to 90 minutes.

Study protocol. In patients with advanced GIST treated withimatinib, CT was done at baseline, at 1, 2, 4, and 6 months, and thenevery 3 months. Angiosonography was done before treatment, at6 months, and then every 6 months in the first five cases, but at thesame time points of CT in the next five. Two radiologists evaluated thepercentage of sonographic contrast uptake during different vascularphases at each tumor site and compared results obtained byangiosonography and CT. In a single case, the last one (Table 1,patient 10), we did a strict evaluation with angiosonography at 1, 2, 4,6, and 8 weeks in order to define the first time point of reduction intumor vascularization during imatinib treatment. The study protocolwas approved by the Institutional Review Board and patients provided

written consent. In nine additional patients, in the follow-up after GISTcomplete resection, CT and angiosonography were done at the sametime points every 3 to 6 months.

Vascular endothelial growth factor. We monitored the serum levelsof the major angiogenic growth factor, vascular endothelial growthfactor (VEGF), in all patients at baseline, at 1, 2, and 4 weeks, and thenevery month until treatment failure. The serum levels of VEGF weredetermined by ELISA. The analyses were carried out according to themanufacturer’s recommendations (Human VEGF, Pierce Endogen,Rockford, IL).

Results

Overall, five patients had z10 liver metastases, four patientshad <10 liver lesions. Liver metastases appeared hypervascu-larized before imatinib treatment. Instead, only five peritonealmetastases in four patients were detected and appeared lessvascularized than liver lesions.

Angiosonography showed a reduction in tumor vasculariza-tion of the liver metastases in all cases (Fig. 1). The tumorperfusion appeared reduced in the central part of the livermetastases. With a median follow-up of 18 months (range3-33), a reduction in tumor vascularization was initiallyobserved in all patients with liver metastases and in the patientwith a retroperitoneal mass only. In four patients, angiosonog-raphy documented progressive disease after 12, 21, 24, and 27months of imatinib treatment.

CT showed an initial tumor response according to standard-ized criteria in six patients (median time to response, 4 months;range 1-9), whereas in the other four patients, stable diseasewas documented lasting 12, 21, 24, and 24+ months,respectively. In these four patients, achieving initial stabledisease based on CT, the first angiosonography done at6 months, documented a reduction in tumor vascularizationof 80%, 60%, >90%, and 70%, respectively. The reduction oftumor perfusion of the liver metastases reported at angioso-nography correlated with the pseudocystic appearance whichcan be seen on post-contrast CT following imatinib treatment(Fig. 2). In four cases, during imatinib treatment, CT showedprogressive disease according to angiosonography. In two caseswith progressive disease after 21 and 24 months of treatment,the imatinib dose was increased to 800 mg/d, but thesetwo patients experienced further progressive disease after 3 and9 months, respectively.

Of the four patients with recurrent disease, two presenteda ‘‘classical’’ progressive disease with both regrowth of

Table1. Patient characteristics before imatinib treatment

Patient no. Age (y) Sex Histology Primary site Primary treatment Sites of disease before imatinib

1 80 M KIT+GIST stomach radical resection liver, peritoneum2 73 M KIT+GIST ileum radical resection liver, peritoneum3 66 M KIT+GIST colon palliative surgery liver4 46 F KIT+ GIST duodenum radical resection liver5 63 F KIT+ GIST retroperitoneum radical resection retroperitoneum6 66 M KIT+GIST ileum radical resection liver, peritoneum7 85 M KIT+GIST stomach radical resection liver8 35 M KIT+GIST peritoneum/retroperitoneum radical resection liver9 69 M KIT+GIST stomach radical resection liver10 53 M KIT+GIST stomach no surgery liver, peritoneum, stomach


www.aacrjournals.orgClin Cancer Res 2005;11(17) September1, 2005 6172



Fig. 1. Angiosonography with BR1of the right lobe of the liver in a patient withliver metastasis fromGIST before and after 2 months of imatinib treatment.A, unenhanced sonography with detection of one lesionwith small centralhypoechoic area and hypoechoic halo around metastasis (arrow); B, following BR1administration, in the same scan of (A), liver metastasis appears isoechoic withrespect to intensely bright normal parenchyma in the portal phase (arrow);C, unenhanced sonography after 2 months of imatinib treatment shows one lesionwith the same characteristics as unenhanced sonography before treatment(A; arrow);D, following BR1administration, in the same scan of (C) after 2 monthsof imatinib treatment, liver metastasis appears hypoechoic and completelysurrounded by intensely bright normal parenchyma in the portal phase (arrow).

Fig. 2. Angiosonography with BR1and CTof the right lobe in a patient withliver metastases fromGIST after 12 months of imatinib treatment showingcorrespondence between the reduction of tumor perfusion at angiosonographywith pseudocystic appearance at CT. A, angiosonography reveals adishomogeneous lesion (1), whereas lesions 2 and 3 appear hypoechoic duringthe late arterial phase with evident reduction in vascularization; B, CTshowsa lesion with a mixed appearance (1), whereas lesions 2 and 3 show apseudocystic appearance.

Angiosonography During Imatinib in GIST




preexisting lesions and new sites of disease was revealed atboth angiosonography and CT with an increase of fluoro-deoxyglucose uptake on PET scan. In two other cases, a newpattern of progressive disease with ‘‘nodule(s) within a mass’’only was revealed at both angiosonography and CT, but PETscan appeared negative (Fig. 3) or with a very small positivity(Fig. 4), at first recurrence.

Serum VEGF levels behaved in a heterogeneous manner, buttwo cases with higher pretreatment serum VEGF levels showedearly decreasing levels. Moreover, the classical pattern ofrecurrence-only was correspondent to increased serum VEGFlevels (Fig. 5).

In the single case receiving strict angiosonographic evalua-tion, a reduction in tumor vascularization was observed as earlyas 2 weeks, but standardized tumor response based on CT wasreported only after 9 months. Liver subcentimetric metastaseswere better depicted with angiosonography.

In the nine additional patients in follow-up after GISTcomplete resection, 21 CT and angiosonography were done atthe same time points. No difference was observed in thedetection of disease recurrence in two patients.

Discussion

The early experience with imatinib in GIST patients involvedclassical anatomic imaging methods for tumor responseassessment. Validated response criteria (WHO, SouthwestOncology Group, and Response Evaluation Criteria in SolidTumors) were used, but these methods are validated for otherdiseases with classical antiproliferative agents. Angiosonogra-phy with a second-generation contrast agent (BR1) is a simple,noninvasive, and reproducible imaging technique, whichrepresents a new method of assessing changes in tumorvascularity.

The results of this preliminary study show that angiosonog-raphy improved the display of tumor vascularity as well as theaccuracy in identifying reduction in tumor vascularization.Because the number of the examined cases was relatively low,general predictions of therapeutic effects must be discussedwith caution. It is very likely that an initial high vascularizationand a strong reduction of tumor perfusion under treatment arecorrelated with the subsequent imatinib-related tumor sizereduction. However, angiosonography showed an evidentreduction in tumor vascularization in the cases achievinglong-term stable disease based on CT according to standardizedcriteria. The reduction of tumor perfusion reported at angioso-nography correlated with the pseudocystic appearance whichcan be seen on post-contrast CT during imatinib treatment. The

Fig. 3. Metastatic GIST in a patient with initial positive response to imatinibtreatment showing the ‘‘nodule(s) within a mass’’pattern of progression duringimatinib treatment (case1). A, after 24 months of imatinib treatment,angiosonography reveals two small contrast-enhanced nodules (arrows) within alesionhypoechoic during the late arterial phase with evident reduction invascularization; B, CTobtained at the same time point of (A) shows two enhancednodules (arrows) within the matrix of a low-attenuationmass; C, PETscanobtained at the same time point of (A) and (B) shows lack of fluorodeoxyglucoseuptake in the treated lesion;D, after 33 months of imatinib treatment, despite theincrease of the dose of imatinib to 800 mg/d, angiosonography reveals the increaseof the size of the two contrast-enhanced nodules (arrows); E, CTobtained at thesame time point of (D) shows the increase of the size of the two enhanced nodules(arrows); F, PETscan obtained at the same time point of (D) and (E) shows a focalarea of increased fluorodeoxyglucose uptake in the treated lesion (arrow).





effective overall response rate to imatinib treatment in GISTsmight be higher than reported with standard anatomicimaging, according to standardized criteria.

A new ‘‘nodule(s) within a mass’’ pattern of recurrence afterresponse to imatinib has been recently described in theliterature (7). In our series, two patients presented this patternof recurrence, with no difference observed in the detectionbetween angiosonography and CT.

Importantly, a reduction in tumor vascularization observedbefore a reduction in tumor size, as assessed throughstandardized criteria, coupled with the observation that theperfusion is mainly reduced in the central part of the treatedtumors is in line with recent studies monitoring antiangiogenicand antivascular therapy with functional imaging (8). Areduction in serum VEGF levels was observed as early as1 week in the two cases with higher pretreatment serum VEGFlevels (Fig. 5). Imatinib-mediated antiangiogenic and/orantivascular properties have been shown in experimentalmodels and in vivo in chronic myelogenous leukemia,neuroblastoma, and prostate cancer (9–14). Imatinib couldinduce antiangiogenic and/or antivascular effects in GIST,possibly through an angiogenesis-dependent mechanism in-volving the downstream cascade of KIT in the tumor cells and/or platelet-derived growth factor receptors in the endothelialtumor cells and/or an antimigratory and antiproliferative effectupon smooth muscle cells needed to surround and stabilize a

Fig. 4. Metastatic GIST in a patient with initial positive response to imatinibtreatment showing the ‘‘nodule(s) within a mass’’pattern of progression duringimatinib treatment (case 2). A, after 27 months of imatinib treatment,angiosonography reveals a hypoechoic contrast-enhanced nodule (arrow) withina mass; B, CTobtained at the same time point of (A) shows an enhanced nodule(arrow) within a mass; C, PET scan obtained at the same time point of (A) and(B) shows a small focal area of very low increased fluorodeoxyglucose uptakein the treated lesion (arrow).

Fig. 5. Serum levels ofVEGF in10 GIST patients with longitudinally taken serumsamples.Two patients achieved a classical progressive diseasewith both regrowthof preexisting lesions and new sites of disease (continuous black line): an increaseof serum VEGF level was evident in both cases at progression.Two other patientsshowed a new pattern of progressive disease with a nodule within a mass only(intermittent black line): a clear increase of serum VEGF level was not evident. In thefigure, when progressive disease was clinically evident is shownwith a bold point.After first progression, in two cases, the dose of imatinibwas increased to 800mg/dwith further progressionwith the same radiological pattern of the first. Arrowindicates the serum VEGF level spike coupled with classical progressive diseasefollowing a 3-week interruption of imatinib treatment because of severe asthenia inan 86-year-old patient. In two patients with higher pretreatment serum VEGFlevels, an early reduction in serum VEGF levels was observed as early as1week:continuous (n = 1) and intermittent line (n = 1).

Angiosonography During Imatinib in GIST




nascent tumor vessel (11, 15). This effect could be easilymonitored with angiosonography.

A translation research interaction between radiologicalfindings and biological/clinical studies could be suggested.The close relation between clinical outcome and the findingson angiosonography indicates that such scanning might be auseful complement to standard anatomic imaging for moni-toring the therapeutic effect of imatinib in patients withmetastatic GIST. Moreover, patterns of tumor response to

imatinib in GIST should be described through new criteriawhich should in part differ from those conventionally used forcytotoxic drugs.

Large studies are warranted to compare angiosonographywith standard anatomic imaging (CT or magnetic resonance)and metabolic functional imaging (fluorodeoxyglucose-PET)for monitoring the therapeutic effect of imatinib in patientswith metastatic GIST and the follow-up of patients after GISTcomplete resection.

References1. De Giorgi U,Verweij J. Imatinib and gastrointestinalstromal tumors: where do we go from here? MolCancer Ther 2005;4:495^501.2.Verweij J, Casali PG, Zalcberg J, et al. Progression-free survival in gastrointestinal stromal tumours withhigh-dose imatinib: randomised trial. Lancet 2004;364:1127^34.3. Demetri GD, von Mehren M, Blanke CD, et al.Efficacy and safety of imatinib mesylate in advancedgastrointestinal stromal tumors. N Engl JMed 2002;347:472^80.4. van Oosterom AT, Judson IR,Verweij J, et al. Updateof a phase I study of imatinib (STI571) in advancedsoft tissue sarcomas and gastrointestinal stromaltumors: a report of the EORTC Soft Tissue and BoneSarcoma Group. EurJCancer 2002;38:83^7.5. Joensuu H, Fletcher C, Dimitrijevic S, Silberman S,Roberts P, Demetri G. Management of malignant gas-trointestinal stromal tumours. Lancet Oncol 2002;3:655^64.

6. Schneider M, Arditi M, Barrau M, et al. BR1: a newultrasonographic contrast agent based on sulfur hexa-fluoride-filled microbubbles. Invest Radiol 1995;30:451^7.7. Shankar S, vanSonnenberg E, DesaiJ, et al. Gastroin-testinal stromal tumor: new nodule-within-a-masspattern of recurrence after partial response to imatinibmesylate. Radiology, Published online before printApril 15, 2005; DOI10.1148/radiol.2353040332.8. Krix M, Kiessling F, Vosseler S, et al. Sensitivenoninvasive monitoring of tumor perfusion duringantiangiogenic therapy by intermittent bolus-contrastpower Doppler sonography. Cancer Res 2003;63:8264^70.9. Buchdunger E, O’Reilly T,Wood J. Pharmacology ofimatinib (STI571). EurJCancer 2002;38:S28^36.10. EbosJML,Tran J, Master Z, et al. Imatinib mesylate(STI-571) reduces Bcr-Abl-mediated vascular endo-thelial growth factor secretion in chronicmyelogenousleukemia. Mol Cancer Res 2002;1:89^95.

11. Dudley A, Gilbert RE, Thomas D, et al. STI-571inhibits in vitro angiogenesis. Biochem Biophys ResCommun 2003;310:135^42.12. Legros L, Bourcier C, Jacquel A, et al. Imatinibmesylate (STI571) decreases the vascular endothelialgrowth factor plasma level in patients with chronicmyeloid leukemia. Blood 2004;104:495^501.13. Beppu K, Jaboine J, Merchant MS, Mackall CL,Thiele CJ. Effect of imatinib mesylate on neuroblasto-ma tumorigenesis and vascular endothelial growthfactor expression. JNatl Cancer Inst 2004;96:46^55.14. Uehara H, Kim SJ, Karashima T, et al. Effects ofblocking platelet-derived growth factor-receptor sig-naling in amouse model of experimental prostate can-cer bone metastases. J Natl Cancer Inst 2003;95:458^70.15. Bono P, KrauseA, von Mehren M, et al. Serum KITand KIT ligand in gastrointestinal stromal tumorpatients treated with imatinib. Blood 2004;103:2929^35.





2005;11:6171-6176. Clin Cancer Res Ugo De Giorgi, Camillo Aliberti, Giorgio Benea, et al. Gastrointestinal Stromal TumorsImatinib Treatment in Patients with Metastatic Effect of Angiosonography to Monitor Response During

Updated version

http://clincancerres.aacrjournals.org/content/11/17/6171

Access the most recent version of this article at:

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

Subscriptions

Reprints and

[email protected] at

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

Permissions

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

.http://clincancerres.aacrjournals.org/content/11/17/6171To request permission to re-use all or part of this article, use this link



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

mailto:[email protected]



Documents

EffectofAngiosonographytoMonitorResponseDuring ...disease, the imatinib dose was increased to 800 mg/d, if possible, according to the tolerance and compliance of patients. In cases