APotentCombinationoftheNovelPI3KInhibitor,GDC …...Committee of KU Leuven (Leuven, Belgium). A...

Cancer Therapy: Preclinical

A Potent Combination of the Novel PI3K Inhibitor, GDC-0941,with Imatinib in Gastrointestinal Stromal Tumor Xenografts:Long-Lasting Responses after Treatment Withdrawal

Giuseppe Floris1, Agnieszka Wozniak4, Raf Sciot1, Haifu Li4, Lori Friedman5, Thomas Van Looy4,JasmienWellens4, Peter Vermaelen2, Christophe M. Deroose2, Jonathan A. Fletcher6, Maria Debiec-Rychter3,and Patrick Sch€offski4

AbstractPurpose: Oncogenic signaling in gastrointestinal stromal tumors (GIST) is sustained via PI3K/AKT

pathway. We used a panel of six GIST xenograft models to assess efficacy of GDC-0941 as single agent or in

combination with imatinib (IMA).

Experimental Design: Nude mice (n ¼ 136) were grafted bilaterally with human GIST carrying diverse

KITmutations. Mice were orally dosed over four weeks, grouped as follows: (A) control; (B)GDC-0941; (C)

imatinib, and (D)GDCþIMA treatments. Xenografts regrowth after treatment discontinuationwas assessed

in groups C and D for an additional four weeks. Tumor response was assessed by volume measurements,

micro-PET imaging, histopathology, and immunoblotting. Moreover, genomic alterations in PTEN/PI3K/

AKT pathway were evaluated.

Results: In allmodels, GDC-0941 caused tumor growth stabilization, inhibiting tumor cell proliferation,

but did not induce apoptosis. Under GDCþIMA, profound tumor regression, superior to either treatment

alone, was observed. This effect was associated with the best histologic response, a nearly complete

proliferation arrest and increased apoptosis. Tumor regrowth assays confirmed superior activity of

GDCþIMA over imatinib; in three of six models, tumor volume remained reduced and stable even after

treatment discontinuation. A positive correlation between response to GDCþIMA and PTEN loss, both on

gene and protein levels, was found.

Conclusion:GDCþIMAhas significant antitumor efficacy inGIST xenografts, inducingmore substantial

tumor regression, apoptosis, and durable effects than imatinib. Notably, after treatment withdrawal, tumor

regression was sustained in tumors exposed to GDCþIMA, which was not observed under imatinib.

Assessment of PTEN statusmay represent a useful predictive biomarker for patient selection.ClinCancer Res;

19(3); 620–30. �2012 AACR.

IntroductionGastrointestinal stromal tumors (GIST) are the most

common mesenchymal tumors of the digestive system.About 95% of GISTs express the receptor tyrosine kinaseKIT. About 80% to 85% of GIST carry somatic activatingmutations either in the KIT or PDGFRA gene, being thecausative events in GIST development (1). With the adventof imatinib (IMA), the clinical course of advanced, meta-static inoperable GIST has dramatically changed from analmost incurable disease with a dismal prognosis to a verytreatable condition. This paradigmatic change is attribut-able not only to imatinib but to other tyrosine kinaseinhibitors (TKI), such as sunitinib, which collectivelyachieved significant clinical benefit in the vast majority ofpatients with GIST with advanced disease (2, 3). KIT/PDGFRAmutation type is a major determinant of responseto imatinib treatment (4, 5). Despite the spectacular long-lasting responses to imatinib, the majority of patients withGIST develops resistance during the therapy and is then

Authors' Affiliations: Departments of 1Pathology, 2Nuclear Medicine andMolecular Imaging, and 3Human Genetics; 4Laboratory of ExperimentalOncology, Department of Oncology and Department of General MedicalOncology, KU Leuven and University Hospitals Leuven, Leuven CancerInstitute, Leuven, Belgium; 5Genentech, San Francisco, California; and6Department of Pathology, Brigham and Women's Hospital, Boston,Massachusetts

Note: Supplementary data for this article are available at Clinical CancerResearch Online (http://clincancerres.aacrjournals.org/).

G. Floris, A. Wozniak, M. Debiec-Rychter, and P. Sch€offski contributedequally to this work.

Prior presentation: Part of the results of the study have been presented atthe Annual Meeting of the American Society of Clinical Oncology (Chicago,IL; 2010 Jun 4–9). J Clin Oncol 2010;28(7s):10020.

Corresponding Author: Giuseppe Floris, Department of Pathology, Uni-versity Hospitals Leuven,Minderbroedersstraat 12, 3000 Leuven, Belgium.Phone: 32-16-336584; Fax: 32-16-336548; E-mail:[email protected].

doi: 10.1158/1078-0432.CCR-12-2853

�2012 American Association for Cancer Research.

ClinicalCancer

Research

Clin Cancer Res; 19(3) February 1, 2013620

on November 1, 2020. © 2013 American Association for Cancer Research. clincancerres.aacrjournals.org Downloaded from

Published OnlineFirst December 11, 2012; DOI: 10.1158/1078-0432.CCR-12-2853

http://clincancerres.aacrjournals.org/

treated with sunitinib. Eventually, sunitinib also ceases tobe effective, leaving patients with GIST without an alterna-tive approved treatment option (6, 7).Regardless of the type of KIT/PDGFRA mutation, the

PI3K/AKT pathway is crucial for tumor cell survival of bothimatinib-sensitive and -resistant GIST (8, 9). However, littleis knownabout themodulationof thePI3K/AKTpathway inrelation to genomic changes occurring during GIST pro-gression (or selected during TKI therapy). Of note, progres-sion toward malignancy in GIST is characterized by accu-mulation of secondary genetic events involving numericaland structural changes of chromosomes (2). In particular,the tumor suppressor gene PTEN (10q23.31) encodes for aprotein known to be a negative regulator of the PI3K/AKTpathway. PTEN activities are not limited to the PI3K/AKTpathway but also seem to be involved in the RAS/MAPKpathway and in the activity of the focal adhesion kinase(FAK), a crucial kinase involved in cell migration. Notsurprisingly, somatic PTEN inactivatingmutations or PTENdeletions are among the most common cancer-relatedmolecular changes in humans (10).Deregulation of the PTEN/PI3K/AKT pathway results in

uncontrolled proliferation and cell survival of tumor cells(10). Clinically applicable approaches to counteract theeffects of the deregulated pathway include a number ofphosphoinositide 3-kinase (PI3K), AKT, and mTOR inhi-bitors that are currently studied in early-phase clinical trials(11). Previous in vitro evidence shows that inhibition of thePTEN/PI3K/AKT pathway by pan-PI3K inhibitor LY294002translates into arrest of cell proliferation and induction ofconsistent tumor cells death in diverse imatinib-sensitiveand -resistant GIST cell lines (8). GDC-0941 is an orallybioavailable, potent and selective pan-inhibitor of class IPI3Ks and inhibits common mutant forms of the PI3Kp110a subunit as effectively as wild-type PI3K. In addition,GDC-0941 is a weak inhibitor of classes II, III, and IV PI3Kfamilymembers (includingDNA-dependent protein kinaseandmTOR).GDC-0941 is currently in clinical development

in anumberof solid tumors, showingpromising results (12,13). In the present study, we evaluated the efficacy of GDC-0941 in vivo, as single agent or in combination with ima-tinib, using a panel of 6 GIST xenograft models carryingdiverseKITmutations and PTEN/PI3K/AKT pathway hyper-activation through different mechanisms.

Materials and MethodsCell lines, biopsy, and generation of mouse GISTxenografts

GIST882 and GIST882Ly cell lines carrying the homozy-gous KIT exon 13 p.K642E mutation and the GIST48 cellline carrying double KIT mutations (KIT exon11 p.V560Dand KIT exon17 p.D820A) were from Dr. J.A. Fletcher(Brigham andWomen’s Hospital, Boston, MA). GIST882Lyis a subline of GIST882 established by selective pressure,through continuous exposure to LY294002. GIST biopsieswith either KIT exon 9 (GIST-BOE) or KIT exon 11 (GIST-PSW and GIST-DFR) mutations were obtained frompatients radiologically progressing under imatinib, andtreated in the Department of General Medical Oncology,University Hospitals Leuven (Leuven, Belgium).

Female adult athymic NMRI nude mice were obtainedfrom Janvier Laboratories. Heterotopic GIST xenograftswere generated by subcutaneous bilateral injection ofGIST882, GIST882Ly, or GIST48 cells (5 � 106 cells persite, in culture medium), or by bilateral subcutaneoustransplantation of fresh GIST biopsies, as previouslydescribed (14). The detailed characteristics of these modelsare presented in Supplementary Table S1. The number ofmice used and xenograft’s passage information are pre-sented in Supplementary Table S2.

Drugs, experimental design, and evaluation ofresponse to treatment

Imatinib mesylate was purchased from Sequoia ResearchProducts and was dissolved in sterile water. GDC-0941 waskindly provided by Genentech and was dissolved in 0.5%methylcellulose/0.2%Tween 80 in awater bath (40�C). Thesolutionwas sonicated for 10minutes prior administration.

All animal experiments were approved by the EthicsCommittee of KU Leuven (Leuven, Belgium). A total of136 mice bearing bilateral GIST tumors (average �800mm3) were grouped as follows: group A—control mice(treated with sterile water per os); group B—GDC-0941(75mg/kg daily per os); group C—imatinib (50mg/kg twicedaily per os); and groupD—GDCþIMA combination (samedoses and schedules as for single treatments). Detailedinformation on number of animals assigned to treatmentgroups is presented in Supplementary Table S2.

Treatment lasted 4 weeks, when animals were sacrificedand tumor samples were preserved for further analysis. Inaddition, for the imatinib and GDCþIMA treatmentgroups, the tumor regrowth assessment was conducted,that is, 4 weeks of treatment was followed by another 4weeks of observation after treatment discontinuation. In theaforementioned groups, half of animals were sacrificed at

Translational RelevanceRegardless of the type of KITmutation, the PI3K/AKT/

PTEN pathway is substantial for tumor cell survival inimatinib-sensitive and -resistant gastrointestinal stromaltumors (GIST). The pharmacologic inhibition of thephosphoinositide 3-kinase (PI3K) pathway in combina-tion with imatinib constitutes a promising strategy toovercome or delay resistance to standard treatment inGIST. In vivodata presented in this study show thatGDC-0941, an orally bioavailable PI3K inhibitor, in combi-nation with imatinib shows superior antitumor efficacyin comparison with standard treatment, resulting insustained effects even after treatment withdrawal. Sec-ond, PTEN status emerges as a possible surrogate pre-dictive biomarker of response in GIST treated with thecombination of GDC-0941 and imatinib.

In Vivo Efficacy of PI3K Inhibitor in GIST

www.aacrjournals.org Clin Cancer Res; 19(3) February 1, 2013 621




week 4 for efficacy assessment and the remaining half wereeuthanized at week 8. For GIST-PSW, 2 separate experi-ments were carried out; in the first, the treatment lasted 19days (groups A–D), whereas in the second experimentregrowth in groups C and D was conducted as describedearlier. Tumor volume, body weight, histologic assessment,and Western blot analysis were implemented as previouslydescribed (14). The histologic response was evaluated usingAgaram and colleagues criteria, by assessing the magnitudeof necrosis, myxoid degeneration, and/or fibrosis on hema-toxylin and eosin (H&E) staining, using the followinggrading system: grade 1 (0%–10%), grade 2 (>10% and�50%), grade 3 (>50% and �90%), and grade 4 (>90%;ref. 15).

Micro positron emission tomography studiesIn GIST-PSW and GIST-DFR models, the evaluation of res-

ponse to treatment was complemented with 2[18F]fluoro-2-deoxy-D-glucose (FDG) micro positron emission tomogra-phy (micro-PET) scans, to study the glucose metabolism ofthe tumor under therapy, as previously described (16).Scans were conducted at day 0 (baseline) and then at week4, 6, and 8. Images from PET scans were evaluated using themedical image data examiner Amide (Sourceforge, ver.0.9.1). Values of FDG uptake were standardized accordingto the dose of FDG injected, body weight, and specific scalefactor to obtain the standardized uptake values (SUV) andthe tumor/lesion glycolysis (TLG) values. Themean and themaximal SUV values were studied for each time point. TLGparameter describes the glucose uptake of a specific tumorin relation to volumetric changes; hence, it defines themeta-bolically active tumor mass. TLG was calculated using thefollowing formula: SUVmean�micro-PET tumor volume.

Mutational analysis, methylation status, and FISHFor molecular assessment DNA was isolated from

frozen tumor fragments using QIAamp DNA Mini Kit(Qiagen). KIT (ENSG00000157404) genotypes in ourxenograft models were determined by mutationalanalysis as previously described (17). Mutation analysisfor PTEN (ENST00000371953; exons 1–9), PI3KCA(ENST00000263967; exons 5, 6, 10 and 21), and AKT(ENSG00000142208; exon 4) genes were also conducted.The primer sequences are available upon request. More-over, DNA methylation status of PTEN promoter (SABIOCpG island ID:28526) was evaluated using EpiTect Meth-yl Profiler qPCR Assay (Qiagen) according to manufac-turer’s protocol.

To assess KIT and PTEN copy number, a dual-colorinterphase FISHwas conducted on paraffin sections. Digox-igenin-labeled bacterial artificial chromosome’s RP11-568A2 DNA for KIT (4q12) and SpectrumGreen-labeledchromosome 4 centromeric probe (CEP4-SG, Abbott) orLSI PTEN(10q23)/CEP10 Dual Color Probe (Abbott) wereused, respectively, in 2 independent experiments. Probelabeling, hybridization, and detection were carried out aspreviously described (17, 18). The number of hybridizationsignals representing investigated genes and chromosome

centromeres were individually recorded for at least 100nuclei. From 0 to 1 CEP signals per nucleus in more than60% of cells was defined as a whole chromosome loss. Inaddition, ratios of KIT/CEP4 or PTEN/CEP10 were calcu-lated. A ratio of 2 or more was defined as specific geneamplification. Heterozygous/homozygous loss was consid-ered if ratio was less than 0.6.

StatisticsThe comparison between tumor volumes on day 0 (base-

line) versus last day of experiment was conducted usingWilcoxon’s matched paired test. The comparison betweendifferent treatment groups (histologic assessments andmicro-PET studies) was done by the Mann–Whitney U test(MW-U). Bonferroni’s correction was used for multipletesting. Statistically significant differences were defined asP < 0.05. The STATISTICA software (Stat Soft, version 9.0)was used for all calculations.

ResultsXenografts characterization

The morphology of untreated GIST-PSW, GIST-BOE,GIST882, and GIST48 was identical to that previouslydescribed (14, 18). GIST-DFR xenografts were composedof spindle cells; in GIST882Ly a mixed population ofmonotonous spindle shaped and highly atypical and pleo-morphic tumor cells was observed. By immunohistochem-istry (IHC), GIST-related biomarkers KIT, ANO1, CD44,and ETV1were expressed at variable degrees of intensity inall xenografts; CD34 was expressed in all models but one(GIST-BOE; Supplementary Fig. S1).

Mutational analysis of the KIT gene confirmed the pres-ence of mutations previously described in cell lines used fortumor induction (8). All xenografts derived from GISTbiopsies showed the same KIT mutations as present in theoriginal sample obtained from patients (SupplementaryTable S1). Moreover, FISH analysis confirmed our previousKIT copy number results obtained from GIST-PSW,GIST882, GIST48, and GIST-BOE (14, 18). In addition, inGIST-DFR, we observed trisomy of chromosome 4 (asjudged by �60% of nuclei with 3 KIT and CEP4 signals),whereas a subpopulation of tumor cells (�25%) ofGIST882Ly xenograft disclosed up to 4 copies of the chro-mosome 4. Nevertheless, KIT amplification was found innone of the xenografts.

Subsequently, we characterized the PTEN/PI3K/AKT sig-naling pathway. By FISH, we identified a heterogeneousspectrum of alterations affecting the PTEN gene. Namely,homozygous loss was observed in the GIST-PSW andGIST882Ly xenografts, whereas GIST48 was characterizedby heterozygous PTEN loss (PTEN/CEP10¼ 0.52). In GIST-DFR, polysomic chromosome 10 was found (�55% ofnuclei with 3–4 CEP10 signals), which was associated withthe loss of 1 PTEN allele in a subpopulation of cells (PTEN/CEP10¼ 0.73), resulting in classification of this model intothe group without PTEN loss. In the remaining models, nocopy number changes in the PTEN locus were observed by

Floris et al.

Clin Cancer Res; 19(3) February 1, 2013 Clinical Cancer Research622




FISH. Mutational analysis of PTEN and methylation anal-ysis of the promoter did not reveal further changes in thePTEN gene. Importantly, we confirmed the lack of PTENexpression in the GIST-PSW and GIST882Ly models at theprotein level (Supplementary Fig. S2).Of note, an additional mutation in exon 6 of PI3KCA

gene (c.1093 G>A; p.E365K) was detected in GIST-PSWby direct sequencing. We confirmed the presence of thismutation in GIST-PSW tumors retrieved from differenttreatment groups and in the earlier grafts passages. Thismutation has been described in endometrial carcinoma(COSM86044; ref. 19). In the remaining models, we didnot find additional mutations in PIK3CA or AKT genes inany of exons tested.

Tumor volume assessmentWefirst treated 12GIST-PSWmice, over 19days. Imatinib

led to objective responses [tumor regression to 17% (P <0.01, vs. baseline)], confirming the high sensitivity of GIST-PSW to imatinib. In contrast, GDC-0941 treatment resultedin a steady increase in tumor burden. Yet, the most remark-able tumor regression was observed under GDCþIMAtreatment, which led to a decrease in tumor size to 6% ofbaseline volume (P < 0.05). Importantly, statistical analysisconfirmed that the GDCþIMA regimen was better thaneither treatment alone [P < 0.01 in both comparisons withimatinib and GDC-0941 (MW-U)] suggesting an additiveeffect (Table 1). On the basis of these results, we decided toassess the duration of response also after treatment discon-tinuation in imatinib andGDCþIMA groups by conducting

a tumor regrowth assay in GIST-PSW and in 5 additionalmodels (Supplementary Fig. S3A).

Overall, regardless of KIT genotype, at week 4 a 3.2-foldincrease in tumor volume in control mice was recorded.Furthermore, treatment with GDC-0941 resulted in tumorburden stabilization in 3 of 5 xenografts (namely GIST-BOE, GIST-882, and GIST-48), resulting overall in tumorgrowth delay rather than tumor regression (Table 1).

After 4 weeks under treatment with imatinib, the tumorburden was reduced to 78% (P < 0.001 vs. baseline) whenall models were considered together. Not surprisingly theKIT genotype had an impact on the response to imatinib;namely, the best tumor regressions were observed in KITexon 11 mutants (Table 1; Supplementary Fig. S3A).

With regard to combination treatment, we noted additiveeffect of the GDCþIMA in all models. Overall, we observedtumor burden reduction to 36% (P < 0.001, vs. baseline)after 4 weeks of GDCþIMA treatment. This response wassignificantly better than in any other group (P < 0.01 vs.imatinib and GDC-0941; Supplementary Fig. S3A).

No treatment-related side effects were observed in anytreatment group, in any models used.

After treatment discontinuation, we observed immediatetumor regrowth of imatinib-treated tumors to levels abovebaseline values (152% of baseline at week 8). The superiorantitumor activity of GDCþIMA treatment was confirmedduring treatment discontinuation. Even after 28 days oftreatment cessation, the tumor burden was yet 73% of thebaseline values (P < 0.05), suggesting long-lasting volumet-ric effects of the combination regimen. A detailed

Table 1. Relative tumor volume assessment in GIST models after treatment (at week 4) and after regrowth(at week 8)

Relative tumor volume [mean% (95% CI)]

At week 4 At week 8

Xenograft model Control Imatinib GDC-0941 GDCþIMA Imatinib GDCþIMA

All models " 324.0 (240–409) # 78 (62–94)�� " 207 (139–275)

�� # 36 (26–45)�� " 152 (107–197) # 73 (53–93)�

GIST-DFR " 494 (193–794) # 44 (35–53)�� " 528 (149–908) # 73 (30–115)

�� # 76 (9–142) ¼ 101 (41–160)GIST-BOE " 173 (137–209) " 146 (121–170) ¼ 108 (90–126)

�� # 77 (58–96)�� " 335 (192–477) " 147 (128–166)

��

GIST882 " 232 (99–364) # 86 (47–125)�� ¼ 102 (37–167) # 16 (11–21)

�� # 151 (95–208) ¼ 90 (53–127)�

GIST48 " 140 (76–205) # 32 (25–38)�� # 81 (51–111) # 9 (5–14)

�� # 82 (50–113) # 13 (�4–30)��

GIST-PSW " 235 (246–451) # 17 (10–24)�� " 257 (168–346)a # 6 (�2–15)

�� ¼ 98 (70–125) # 15 (�4–33)��

GIST882Ly " 680 (137–1,223) " 145 (75–216)�� " 224 (120–328) # 22 (19–25)

�� ¼ 95 (�97–287) # 25 (3–48)PTEN (IHC status)PTEN� " 473 (281–664) # 72 (32–112)

�� " 241 (184–297)� # 14 (8–19)�� ¼ 97 (63–130) # 21 (9–33)

��

PTENþ " 247 (170–325) # 81 (64–97)�� " 192 (96–289)� # 43 (30–55)

�� " 171 (113–229) ¼ 90 (66–113)�

PTEN status by FISHPTENloss " 362 (221–503) # 56 (32–80)

�� " 177 (126–227) # 12 (8–15)�� # 90 (69–111) # 17 (9–26)

��

PTENno loss " 285 (185–384) ¼ 99 (79–119) " 239 (105–374) # 55 (39–70) " 208 (131–286) " 114 (93–135)

NOTE: Arrows indicate whether relative tumor volume was less (arrow down) or more (arrow up) than the starting volume. Statisticalsignificance was calculated using Mann–Whitney U test and related to the control treatment group (at week 4) or to imatinib-treatedgroup (at week 8). �, P < 0.05; ��, P < 0.005.aIn GIST-PWS GDC-0941 efficacy was assessed only after 19 days.






description of the results obtained in each model is pro-vided in Table 1 and Supplementary Fig. S3A.

As indicatedby themolecular characterization,PTEN genelosswas frequentlyobserved inourpanel of xenografts. Thus,we used PTEN status both at the protein and at the genomiclevels as a potential surrogate predictive marker of response.At the protein level, we identified 2 groups: xenograftsexpressing PTEN (PTENþ: GIST-DFR, GIST-BOE, GIST882,and GIST48) and xenografts not expressing PTEN (PTEN�:GIST-PSW and GIST882Ly). Similarly, we used PTEN geno-mic status to divide mice into 2 groups, that is, xenograftswith PTEN retained (PTENno loss: GIST-DFR, GIST-BOE, andGIST882) and those with homozygous/heterozygous PTENloss (PTENloss: GIST-PSW, GIST882Ly, and GIST48).

PTEN status affected substantially tumor volumeresponses to GDC-0941 single treatment. In particular,when compared with control tumors, GDC-0941 alone ledto remarkable tumor growth delay only in PTEN� orPTENloss mice. In contrast, under GDCþIMA, an additiveeffect of the 2 drugs was observed in PTEN�, PTENþ, orPTENloss mice, but was not observed in PTENno loss animals.Most importantly, the tumor regression induced byGDCþIMA in PTEN� or PTENloss groups was better thanthat recorded in the corespective PTENþ or PTENno lossmice(14% or 12% of starting volume vs. 43% or 55%, respec-tively). This remarkable result was further confirmed duringthe tumor regrowth. Upon GDCþIMA discontinuation, weobserved a steady regrowth of tumors only in PTENþ orPTENno loss mice, that is, to 90% and 114% of baseline,respectively. Notably, in PTEN� or PTENloss mice treatedwith GDCþIMA the tumor volume remained substantiallysmall, even in the absence of treatment, being atweek821%(PTEN�) and 17% (PTENloss) of baseline values (Supple-mentary Fig. S3B and S3C).

Given all earlier, these observations suggest that GISTremain mainly dependent on KIT. The loss of PTEN or thepresence of PI3K mutations (only in GIST-PSW) does notseem to influence response to imatinib.Nevertheless, loss ofPTENmay represent a potential biomarker to select a subsetof GISTs that could respond better to PI3K inhibitors or tothe combination of imatinib with a PI3K inhibitor.

Micro-PET assessmentIn GIST-PSW and GIST-DFR models, we assessed the

dynamics of FDG uptake in relation to treatment, by study-ing the TLG values and the SUV values. In the GIST-PSWmicro-PET studies have been conducted 2 times, first in ashort experiment over 19 days treatment and then in atumor regrowth assay (4 weeks treatment followed by 4weeks treatment discontinuation). Results obtained in ima-tinib and GDCþIMA groups in the short experiment weresimilar to those obtained in the regrowth assay. In bothxenograftmodels, we observed a significant reduction of theTLG values after 4 weeks of imatinib treatment (6.5% of thestarting value in GIST-PSW and 20.1% in GIST-DFR incomparison with baseline, respectively; P < 0.05), confirm-ing their imatinib sensitivity (Supplementary Fig. S4A andS4B). Under GDC-0941, TLG values were higher than

baseline values, reflecting delayed tumor growth whencompared with the TLG values recorded in control tumors(data not shown).

Combination treatment resulted in a significant reduc-tion in TLG values in both xenografts. Interestingly, only inGIST-PSW mice statistical analysis indicated an enhancedtherapeutic effect of the combination treatment in compar-ison with imatinib single agent (2.3% and 6.5% of baselineTLG after 4weeks of treatment, respectively;P<0.05). In theGIST-DFR mice the GDCþIMA treatment yielded the samelevel of reduction as imatinib single agent (20.1% and20.8% of TLG baseline, respectively; Table 2; Supplemen-tary Fig. S4A and S4B).

As judged from micro-PET analysis during treatmentdiscontinuation, TLG values of GIST-DFR xenograftsremained approximately 50% of the starting values in bothimatinib (46.9%) and GDCþIMA (50.3%) groups, indicat-ing an equal durable effect of the 2 treatment options in thismodel. On the other hand, in GIST-PSW in the GDCþIMAgroup the glucose uptake of the tumors remained almostnegligible and stable over additional 4 weeks after thetreatment discontinuation, whereas in the imatinib groupwe observed an immediate increase of TLG values to levelalmost equal to baseline values (73.2% of baseline). Atweek 8, the TLGvalues recorded inGDCþIMA treatedGIST-PSW tumors was still 10% of the baseline values, indicatingdurable effects even after treatment discontinuation. Thechanges in SUV mean values were similar to the above-mentioned TLG changes in both xenografts (Table 2 andSupplementary Fig. S4A and S4B).

HistopathologyWe assessed histologic response by evaluating the mag-

nitude of necrosis, myxoid degeneration, or fibrosisinduced by the treatment in tumor tissues on H&E staining(15).We observedminimal histologic response (grade 1–2)in the vast majority of the tumor under imatinib treatment(36 of 44 specimens) in all but one models tested. In GIST-DFR, the histologic response under imatinib was grade 3 to4 in all tumors analyzed. Interestingly, this result wasconsistent with the changes recorded with the TLG valuesbymicro-PET analysis. Namely, on H&E staining the tumortissues of imatinib-treated GIST-DFR xenografts were sig-nificantly replaced by myxoid degeneration (hallmark ofimatinib response in GIST), which corresponds to an amor-phous matrix almost devoid of cells and hence a metabol-ically inactive tissue.

Overall, the GDC-0941 treatment yielded grade 2 to 3histologic response in about one third of the specimenanalyzed, suggesting some level of antitumor activity. How-ever, the interpretation of these results was uncertain inGIST-DFR, GIST882, and GIST882Ly xenografts, as areas ofspontaneous necrosis of the samemagnitude were encoun-tered also in a minority of control tumors of these models.In contrast, grade 3 to 4histologic responsewas identified inthe vast majority of tumors treated with GDCþIMA (30 of46 specimens). Remarkably, 5 of 6 tumors showed grade 4histologic response in the GIST882Ly model. On the other

Floris et al.





hand, none of the treatments yielded more than grade 2histologic response in GIST-BOE grafts (Fig. 1).

Subsequently, we have evaluated the proliferative andapoptotic activity in tumors collected at the end of treat-ment. Overall, regardless of the model, control tumorsshowed high mitotic activity [mean 31.9 mitoses/10HPF;95% confidence interval (CI), 27.9–35.9].

After 4 weeks under imatinib, significant reduction ofmitotic activity in all but one model (GIST-BOE) wasobserved. Generally, as indicated by mitotic count, theproliferative activity was 3-fold reduced as compared withcontrol tumors. Under imatinib, the apoptotic activity wassignificantly induced only in GIST-PSW and GIST-DFRtumors, whereas in others we recorded either a slightincrease or no significant change in comparison with con-trol tumors.

On the whole, the activity of GDC-0941 was limited to asignificant antiproliferative effect, whereas the apoptoticactivity remained substantially unchanged in all models.The combination GDCþIMA was better than either treat-ment alone. Overall, it yielded 10.1-fold reduction inmitotic activity and 3.5-fold increase in apoptotic activity.Apoptosis was significantly stimulated to higher extent thansingle-agent treatments in 3 models, namely GIST-PSW,GIST-DFR, and GIST48. Ki67- and CC3-immunostainingshowed comparable results with mitotic and apoptoticcount in all models (Table 3).

Of note, the mitotic and apoptotic counts evaluated atweek 8, that is, 4weeks after treatmentwithdrawal, returnedto the level observed in control tumors, suggesting thatdurable stabilization of tumor volume under combinationregimen might be due rather to more efficacious inductionof necrotic and stromal changes than complete eradicationof active tumor cells.

This remarkable result of the combination treatment wasalso confirmed when the xenograft models were subdi-vided into 2 groups according to their PTEN status at theprotein level and at the genomic level. In particular, underGDCþIMA treatment, most of tumors lacking the PTENprotein (PTEN�) or with homozygous/heterozygous lossof PTEN gene (PTENloss) showed grade 3 or 4 histologicresponse. Conversely, grade 1 histologic response waspresent in one third of tumors analyzed in the PTENþ

and PTENno loss groups. The antiproliferative and proapop-totic activities of the GDCþIMA were better in the PTEN�

and PTENloss groups as compared with the PTENþ andPTENno loss groups. However, PTEN status did not correlatewith antiproliferative and/or proapoptotic activities ofimatinib or GDC-0941, when administered as singleagents (Table 3).

Assessment of the oncogenic signaling in response totreatment

As indicated byWestern blot analysis, KIT protein and itsmain intermediates were activated in all GIST xenografts.Activation of KIT was more pronounced in GIST-PSW,GIST-DFR, GIST-BOE, and GIST48, whereas in the other2 models this feature was less apparent, possibly

Tab

le2.

Micro-P

ETresu

ltsin

GIST-PSW

andGIST-DFRmod

els

Relativemicro-P

ETva

lues

[%(95%

CI)]

Atwee

k4

Atwee

k6

Atwee

k8

Treatmen

tgroup

s/models

m-PETvo

lSUVmean

SUVmax

TLG

m-PETvo

lSUVmean

SUVmax

TLG

m-PETvo

lSUVmean

SUVmax

TLG

Imatinib

GIST-PSW

12.6

(7.5–17

.7)�

52.8

(45.0–

60.7)

53.5

(45.2–

59.9)�

6.5(4.3–8.7)

�33

.5(12.9–

54.1)�71

.1(61.2–

81.0)

72.5

(60.6–

84.3)�

23.4

(11.4–

35.5)�79

.8(36.7–

122.9)

93.3

(78.6–

108.0)

79.1

(68.9–

89.4)�

73.2

(39.3–

107.0)

GIST-DFR

25.0

(14.6–

35.4)�84

.3(57.8–

110.7)

77.0

(56.8–

97.1)�

20.1

(12.7–

27.5)�17

.8(10.4–

25.2)�97

.7(75.2–

120.3)

81.4

(71.0–

91.8)�

17.0

(10.3–

23.6)�58

.8(20.8–

96.9)�

87.0

(61.4–

112.6)

76.3

(59.2–

93.3)�

46.9

(24.8–

69.0)�

GDCþI

MA

GIST-PSW

6.1(�

0.1–

12.3)

34.7

(22.7–

46.7)

42.1

(19.2–

65.0)

2.3(�

0.3–

4.9)

�6.2(0.3–12

.0)

43.9

(26.7–

61.2)

50.5

(33.6–

67.4)

2.8(0.2–5.3)

11.8

(�0.1–

23.7)54

.9(23.5–

86.3)

48.2

(26.1–

70.3)

7.3(�

0.7–

15.3)

GIST-DFR

20.1

(15.9–

24.2)�10

5.5(89.8–

121.2)

86.5

(69.1–

103.9)

20.8

(17.6–

24.0)�19

.9(13.4–

26.3)�16

4.3(131

.8–19

6.7)

�11

9.9(90.7–

149.1)

31.5

(23.7–

39.4)�47

.3(28.3–

66.3)�

103.2(87.6–

118.8)

92.8

(70.9–

114.6)

50.3

(22.0–

78.5)�

NOTE

:SUVan

dTL

Gva

lues

wereob

tained

afterm

icro-P

ETsc

ansco

nduc

tedat

wee

ks4,6,an

d8.Th

emea

nSUVva

lues

(SUVmean)and

themax

imalSUVva

lues

(SUVmax)of

thetumor

mas

seswereev

alua

ted.T

LGwas

calculated

asfollo

ws:

SUVmea

n�

tumor

volumeas

mea

suredbymicro-P

ETsc

an(mPETvo

l).Res

ults

aresh

ownas

relativ

eva

lues

,related

tobas

eline.

Statis

tical

sign

ifica

ncewas

calculated

versus

day

0(bas

eline)

with

Wilcox

on's

match

edpairedtest.�,P

<0.05

.






suggesting a lower dependency on KIT. The PI3K 110 kDsubunit was equally expressed in all models (Supple-mentary Fig. S5).

Consistent inhibition of both p-KITY703 and p-KITY719

isoforms under imatinib in theGIST xenografts carryingKITexon 11 mutations (at least 60% reduction for both iso-forms) was observed (Fig. 1 and Supplementary Fig. S5).However, while p-AKT and p-ERK were consistently inhib-

ited in GIST-DFR, the activation of these KIT intermediateswas still detectable inGIST-PSWandGIST48. InKIT exon13mutated GIST882 and GIST882Ly models, minimal KITinhibition under imatinib treatment (�25% inhibition)resulted in a strong inhibition of p-ERK in both models(80% and 70% inhibition, respectively). In contrast, whilep-AKT was strongly inhibited in GIST882, only 15% inhi-bition was observed in GIST882Ly. These findings most

Table 3. Evaluation of proliferative and apoptotic activity

PTEN (IHCstatus) PTEN (status by FISH)

All groups GIST-DFR GIST-BOE GIST882 GIST48 GIST-PSW GIST882Ly PTEN� PTENþ PTENloss PTENno loss

Mitosis (H&E)Imatinib # 3.0�� # 89.0�� ¼ 1.0 # 3.2�� # 7.0�� # 36.8�� # 2.1�� # 5.5�� # 1.9�� # 5.1� # 2.1��

GDC-0941 # 1.7�� # 2.2� # 1.7� # 1.9�� # 4.0�� # 1.6 # 1.2 # 1.3�� # 2.0�� # 1.6�� # 1.9��

GDCþIMA # 10.1�� # 59.3�� # 2.9�� # 16.8�� # 11.2�� # 331.0�� # 19.9�� # 47.8�� # 5.6�� # 40.3�� # 6.8��

Apoptosis (H&E)Imatinib " 1.8�� " 3.5�� " 1.1 " 1.1 " 1.1 " 3.4�� " 1.1 " 1.9�� " 1.7� " 1.8�� " 1.7��

GDC-0941 " 1.4 ¼ 1.0 # 1.3 " 2.0� ¼ 1.0 " 1.5 ¼ 1.0 " 1.7�� " 1.2 " 1.5 " 1.2GDCþIMA " 3.5�� " 6.1�� # 1.2 " 1.6� " 4.3�� " 11.9�� " 1.4 " 4.4�� " 2.8�� " 5.0�� " 2.4�

Ki67Imatinib # 2.1�� " 1.2 # 1.3 # 2.9�� # 11.9�� # 45.2�� # 2.3�� # 4.1�� # 1.5 # 4.7�� # 1.3��

GDC-0941 # 1.6�� # 1.4 # 2.2� # 1.9�� # 5.9�� # 1.1 # 1.6�� ¼ 1.0 # 2.1�� # 1.5 # 1.8GDCþIMA # 9.0�� # 5.8� # 4.0�� # 260.6�� # >500�� # 28.1�� # 16.5�� # 20.2�� # 5.7�� # 18.5�� # 6.2��

Cleaved caspase-3Imatinib " 1.4�� " 1.5 " 1.2� ¼ 1.0 " 1.6� " 1.8� ¼ 1.0 " 1.5 " 1.4� " 1.5� " 1.2�

GDC-0941 " 1.1 # 1.5 # 1.5 " 1.3 " 1.4� # 0.9 " 1.2 " 1.6 ¼ 1.0 " 1.4 " 0.8GDCþIMA " 2.2�� " 2.6�� # 1.7 " 1.5 " 4.5�� " 4.1�� " 1.2 " 2.6�� " 1.7� " 3.1�� " 1.4

NOTE: Results are shown as fold changes in comparison with control. Arrows indicate whether proliferation or apoptosis were less(arrowdown) ormore (arrowup) than respectivecontrol in differentmodels.Statistical significancewascalculatedusingMann–WhitneyU test. �, P < 0.05; ��, P < 0.005.aIn GIST-PWS GDC-0941 efficacy was assessed only after 19 days.

Histologic response (HR)

Grade 1 Grade 2 Grade 3

IMA GDC-0941GIST-DFR GIST-BOE GIST822 GIST48 GIST-PSW GIST822Ly

GDC+IMAGrade 4 Grade 1 Grade 2 Grade 3 Grade 4 Grade 1 Grade 2 Grade 3 Grade 4

35

30

25

20

15

10

5

0

Num

ber

of tu

mor

s

Figure 1. Histologic responseevaluated in 6 models tested.Histologic response was scoredby assessing the magnitude ofnecrosis, myxoid degeneration,and/or fibrosis on H&E stainingusing the following gradingsystem: grade 1 (0%–10%), grade2 (>10% and �50%), grade 3(>50% and �90%), and grade 4(>90%).

Floris et al.





likely result from the hyperactivation of PI3K/AKT signalingpathway in the tumors carrying PTEN/PI3K alterations(GIST-PSW, GIST48, and GIST882Ly), whichmight requirea higher imatinib dose for equivalent inhibition. In GIST-BOE, imatinib treatment caused only a minimal p-KIT andp-AKT inhibition (�40% and 10% reduction, respectively)as expected for KIT exon 9 mutant, whereas a more consis-tent inhibition of p-ERK was observed (90% reduction; Fig.2; Supplementary Fig. S5).The activity of KIT was unaffected in all models by GDC-

0941 alone. Notably, a remarkable inhibition of p-AKT wasobserved in all but the GIST-PSW model (about 70%reduction in 5 of 6 xenografts). In parallel, some degree ofp-ERK inhibition in GIST-DFR, GIST48, GIST-BOE andGIST882 xenografts were observed, whereas in GIST882Lyand GIST-PSW the p-ERK was upregulated or unchanged,respectively.Importantly, GDCþIMA yielded a more consistent inhi-

bition of KIT signaling pathway, especially in models car-rying mutations that are less sensitive to imatinib. InGIST48, the combination treatment was the only one toresult in complete p-AKT inhibition (96% reduction). Sim-ilarly in GIST-BOE, the combination arm caused a moreconsistent inhibition of p-AKT than any single-agent treat-ment (83% reduction). Of note, GDCþIMA resulted inextracellular signal–regulated kinase (ERK) inactivation inall xenograft models except GIST-PSW, which featuredp-ERK hyperactivation (more than 150%). The latter couldbe related to the accumulation of activated macrophagesthat were observed on the histologic level exclusively inGIST-PSW tumors treated with GDCþIMA, or to the releaseof a negative signaling loop involving the RAS/MAPK path-way (Fig. 2; Supplementary Fig. S5; refs. 20, 21).As expected, 4 weeks after treatment discontinuation the

activation levels for KIT and its signaling intermediatesreverted to levels found in tumors (data not shown).

Importantly, based on IHC and immunoblotting PTENexpression did not change after 4 and 8weeks of experimentin allmodels. InGIST48,we also conducted FISHanalysis ofPTEN, which confirmed the unchanged copy number statusthroughout whole experiment.

DiscussionThe PI3K/AKT pathway is a major contributor to prolif-

eration and survival in imatinib-sensitive and imatinib-resistantGIST (8, 9).Using a panel of diverseGIST xenograftmodels carrying common KIT mutations, we showed forthe first time the in vivo efficacy of the PI3K inhibitorGDC-0941, used as a single agent or in combination withimatinib.

In the present study, GDC-0941 caused either tumorgrowth delay or tumor burden stabilization, resulting inconsistent reduction of the proliferative activity. However,despite a significant inhibition of the AKT activation, we didnot observe substantial in vivo proapoptotic activity in thesemodels. Bauer and colleagues described similar results inGIST882 cells treated with the PI3K inhibitor, LY294002, invitro (8). Moreover, in genetically manipulated imatinib-sensitive GIST cell lines in which AKT is constitutivelyactivated, the survival of GIST cells does not depend onlyon AKT activity (22). Thus, the absence of consistent apo-ptotic activity in GIST treated with GDC-0941 is not sur-prising and underscores the dominant role of KIT in regu-lating GIST proliferation and survival (22).

The combinationGDCþIMA treatment proved to be veryeffective in all xenografts tested here. This regimen reducedtumor burden better than imatinib or GDC-0941 alone inthe majority of models and dramatically enhanced anti-proliferative activity in all xenografts. Furthermore, in 65%of tumors treatedwith a combination regimen, we observedhigh degree of histologic response (grade 3–4), and in 30%

Figure 2. Densitometricassessment of active (phospho-)protein forms in KIT signalingpathway.

IMA

pKIT

(Y70

3)

pKIT

(Y71

9)

pAK

T

pER

K

pKIT

(Y70

3)

pKIT

(Y71

9)

pAK

T

pER

K

pKIT

(Y70

3)

pKIT

(Y71

9)

pAK

T

pER

K

GDC-0941GIST-DFR GIST-BOE GIST882 GIST48 GIST-PSW GIST822Ly

GDC+IMA

250%

200%

150%

100%

50%

0%

–50%

–100%

–150%

Rel

ativ

e in

tens

ity v

alue

s






of the tumors the fraction of vital tumor cells was reduced toless than10%. These astonishing effectsmay account for theoutstanding long-lasting effect observed in the tumorregrowth experiments. Most remarkably, in 3 of the xeno-grafts tested (GIST-PSW, GIST48, and GIST882Ly) thetumor volume remained stable and considerably small evenafter 28 days of treatment discontinuation. Our results areconsistent with previous in vitro results, which showed anadditive antiproliferative effect with the combination ofLY294002 and low doses of imatinib, in imatinib-sensitiveand -resistant GIST cell lines (8).

Generally, the proapoptotic activity of the GDCþIMAtreatment exceeded that for either treatment alone, mainlyin GIST xenografts carrying KIT exon 11 mutations. Twoscenarios may explain this finding. First, KIT-independentmechanisms may influence proapoptotic responses to ther-apies. For example, the differential expression of antiapop-totic proteins such asBCL2 inGISTof distinct anatomic sitescould hamper proapoptotic activity of antitumor-targetedagents (23). For example, gastric GIST have less BCL2expression compared with small intestine GIST. Second,KITmay have a dominant role in controlling cell survival inthis setting. Thus, if the KIT receptor is not completelyinhibited because of an oncogenic mutation with reducedsensitivity to imatinib (e.g.,KIT exon9orKIT exon13), thenapoptosis might not be substantially induced (22, 24).Therefore, the additive proapoptotic effect of GDCþIMAobserved in theGIST xenograftmodels carryingKIT exon 11mutations may be related to their intrinsically higher ima-tinib sensitivity. These observations suggest that the com-bination of PI3K inhibitors with higher doses of imatinibmay further stimulate apoptosis in GIST with non-exon 11primary KIT mutations.

An additional observationof our studywas the possibilityfor a molecular stratification according to PTEN/PI3K/AKTpathway alterations in advanced GIST. The GIST-PSWmodel is derived from a patient with advanced disease,who developed resistance to standard treatments. While wewere not able to identify secondary KIT mutations in theoriginal tumor, we found the coexistence of a point muta-tion in the PI3K gene and homozygous deletion of theregion encoding PTEN in chromosome 10 in the xenograftderived from that patient. Importantly, we were able toconfirm the presence of these molecular events also in earlypassages of the xenograft suggesting their presence in theoriginal tumor. A point mutation in the PI3K gene wasrecently described in one clinical case of a high risk, untreat-ed primary GIST carrying a KIT exon 11 mutation (25). ThePTEN and PI3K copy number changes/mutations are notpresent or rare in primary GIST, as suggested by recent study(26), but the incidence of these events in GIST under TKItherapy might be significant. The coexistence of geneticevents other than KIT/PDGFRA mutations occurring in thePI3Kpathway inGISTmay correlatewith tumorprogressionrather than primary resistance to therapy, as the GIST-PSWxenograftmodel preserved high sensitivity to imatinib treat-ment. Interestingly, also in other cancer types the presenceof mutations in the PTEN/PI3K/AKT pathway has been

associatedwithworse prognosis and enhanced invasiveness(10, 14, 27, 28). Nevertheless, this observation should beconfirmed in a larger cohort of tumors to better understandthe incidence and the role of PTEN/PI3Kmutations in GIST.

Genomic losses affecting the PTEN locus (10q23.31)were identified in 3 of 6 xenograft models by FISH. Wealso confirmed the lack of PTEN protein expression inxenograft showing homozygous loss of PTEN gene (GIST-PSW and GIST882Ly). Chromosomal aberrations are fre-quently observed in GIST and are implicated in progressiontoward malignancy (2). Genome profiling studies carriedout on primary GISTs reveal chromosome 10q loss in about20% to 38% of patients (29, 30). However, homozygousloss of PTEN was not observed, at least in imatinib-na€�vetumors. According to one study, about 40%of primaryGISThave reduced levels of PTEN protein by IHC, which wasimplicated as an independent prognosticmarker (31). As ofyet, the biologic significance of aberrations involving thePTEN locus in GIST is largely unknown. It is not clearwhether a dose-dependent effect in the levels of PTENexpression may contribute to malignancy also in GIST, asit is the case in transgenicmodels of breast carcinomas (28).Genomic alterations targeting the PTEN locus inGIST couldbe a consequence of the continuous treatment with TKI,resulting in selection of subclones of tumor cells with apreferential loss of PTEN. As observed in breast carcinomasand glioblastomas, PTEN loss-of-function could reduce theefficacy of TKI treatment in GIST (11, 32, 33). However,based on our results, we cannot confirm this hypothesis,which would require a larger cohort of tumors.

Intriguingly, we observed a correlation between the geno-mic status of PTEN and response to treatment in the com-bination treatment group (i.e., GDCþIMA regimen). Thesestudies suggest that patients whose GIST reveals at least aheterozygous deletion of the PTEN locus may have thehighest benefit from combination of PI3K inhibitors withimatinib. Thereby, we propose that PTEN status and pos-sibly PI3K mutations could serve as a predictive biomarkerfor the selection of patients with GIST who could respondthe best to the combination of imatinib and PI3K-inhibitors.

The dynamic evaluation of glucose metabolism in GISTby means of PET scans is regarded as a tool that relates wellto imatinib responses or resistance in patients with GIST(34). Consistent with the observation in the clinic, we showthat FDG uptake measured in GIST-xenograft bearing micecorrelated well with response to imatinib treatment. How-ever, as observed in the GIST-DFR model the evaluation ofthe FDG uptake should not be limited to the study of SUVvalues but should be complemented by the assessment ofparameters that combine changes in the tumor tissue withthe level of glycolysis, such as TLG values. Indeed, TLGvalues correlated well with histologic changes observed inthe tumor tissue and provided further information aboutthe activity of the drugs over time. The systematic study ofthese parameters by PET scan could improve the radiologicclassification of responding or nonresponding GIST undertreatment with targeted agents in the clinic. Nevertheless, it

Floris et al.





must be acknowledged that FDG changes studied by PET donotmeasure directly the antitumoral effects of the drugs butmost likely reflect a relocation of glucose transporters in thecellmembrane as response to treatmentwith targeted agents(19, 35).In conclusion, we provide the first evidence for the in vivo

activity of the PI3K inhibitor GDC-0941 in GIST. GDC-0941 shows a significant antiproliferative effect in highlyproliferating GIST, and induces variable degrees of necrosisin tumor tissue. In combination with imatinib, the thera-peutic activity of GDC-0941 is dramatically enhanced asshown by more prominent histologic response, inductionof apoptosis, and even more profound inhibition of cellproliferation. The therapeutic effects observed in the com-bination treatment provide more durable effects after treat-ment discontinuation. Assessment of the PTEN status atgenomic and protein levels in GISTs and its relevance to TKItreatment (including PI3K inhibitors) warrants furtherstudy. However, the lack of PTEN expression is unlikelyresponsible for decreased sensitivity to standard treatmentwith imatinib in GIST. Our results provide a very strongpreclinical rationale for the use of the IMAþPI3K inhibitorscombination inpatientswithGIST in the clinic, and identifyPTEN status as a potential predictive biomarker to select asubset of patients with GIST with higher sensitivity towardthis type of therapeutic strategy. Of note, a first dose-findingclinical GIST trial, that is based on these and related obser-vations and combines imatinib with another PI3K inhibi-tor, has started recruiting patients (36).

Disclosure of Potential Conflicts of InterestA.Wozniak has honoraria from Speakers Bureau of Novartis. J.A. Fletcher

has a commercial research grant fromDeciphera and is a consultant/advisory

board member of Novartis, Ariad, and Deciphera. No potential conflicts ofinterest were disclosed by the other authors.

Authors' ContributionsConception and design:G. Floris, A.Wozniak, R. Sciot, M. Debiec-Rychter,P. Sch€offskiDevelopment of methodology: G. Floris, J.A. Fletcher, M. Debiec-Rychter,P. Sch€offskiAcquisitionofdata (provided animals, acquired andmanagedpatients,provided facilities, etc.): G. Floris, A. Wozniak, H. Li, L. Friedman, T. VanLooy, J. Wellens, P. Vermaelen, M. Debiec-Rychter, P. Sch€offskiAnalysis and interpretation of data (e.g., statistical analysis, biosta-tistics, computational analysis): G. Floris, A. Wozniak, R. Sciot, C.M.Deroose, M. Debiec-Rychter, P. Sch€offskiWriting, review, and/or revision of the manuscript: G. Floris, A. Woz-niak, R. Sciot, H. Li, L. Friedman, T. Van Looy, J. Wellens, J.A. Fletcher, M.Debiec-Rychter, P. Sch€offskiAdministrative, technical, or material support (i.e., reporting or orga-nizing data, constructing databases):G. Floris, A. Wozniak, J. Wellens, P.Vermaelen, M. Debiec-Rychter, P. Sch€offskiStudy supervision: G. Floris, A. Wozniak, R. Sciot, M. Debiec-Rychter, P.Sch€offski

AcknowledgmentsThe authors thank Dr. Erna Dewil for the logistic support offered for the

animal facility. The authors also thank LieveOphalvens, Ulla Vanleeuw, AnnVan Santvoort, and Vanessa Vanspauwen for their excellent technicalassistance.

Grant SupportThis work is supported by research grants from the Fonds voor

Wetenschappelijk Onderzoek Vlanderen (FWO grant # G. 0510.06 to P.Sch€offski), NIH GI SPORE (1P50CA12703-05 to J.A. Fletcher), and Life RaftGroup (toM.Debiec-Rychter and J.A. Fletcher). P. Sch€offski received researchgrants fromGenentech. L. Friedman is employed by Genentech, whose drugwas herein tested.

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

Received September 4, 2012; revised November 6, 2012; acceptedNovember 26, 2012; published OnlineFirst December 11, 2012.

References1. Patil DT, Rubin BP. Gastrointestinal stromal tumor, advances in diag-

nosis and management. Arch Pathol Lab Med 2011;135:1298–310.2. Rubin BP, Heinrich MC, Corless CL. Gastrointestinal stromal tumour.

Lancet 2007;369:1731–41.3. CorlessCL,BarnettCM,HeinrichMC.Gastrointestinal stromal tumors:

origin and molecular oncology. Nat Rev Cancer 2011;11:865–78.4. Debiec-Rychter M, Sciot R, Le Cesne A, Schlemmer M, Hohemberger

P, van Oosterom AP, et al. KIT mutations and dose selection forimatinib in patients with advanced gastrointestinal stromal tumours.Eur J Cancer 2006;42:1093–103.

5. Gastrointestinal stromal tumormeta-analysis group (MetaGIST).Com-parison of two doses of imatinib for the treatment of unresectable ormetastatic gastrointestinal stromal tumors: a meta-analysis of 1640patients. J Clin Oncol 2010;28:1247–53.

6. Le Cesne A, Ray-Coquard I, Bui BN, Adenis A, Rios M, Bertucci F,et al. Discontinuation of imatinib in patients with advanced gas-trointestinal stromal tumours after 3 years of treatment: an open-label multicentre randomised phase 3 trial. Lancet Oncol 2010;11:942–9.

7. Gramza AW, Corless CL, Heinrich MC. Resistance to tyrosine kinaseinhibitors in gastrointestinal stromal tumors. Clin Cancer Res 2009;15:7510–8.

8. Bauer S, Duensing A, Demetri GD, Fletcher JA. KIT oncogenic sig-naling mechanisms in imatinib-resistant gastrointestinal stromaltumor: PI3-kinase/AKT is a crucial survival pathway. Oncogene2007;26:7560–8.

9. Duensing A, Medeiros F, McConarty B, Joseph NE, Panigrahy D,Singer S, et al. Mechanism of oncogenic KIT signal transduction inprimary gastrointestinal stromal tumors (GISTs). Oncogene 2004;23:3999–4006.

10. Song MS, Salmena L, Pandolfi PP. The functions and regulation of thePTEN tumor suppressor. Nat Rev Mol Cell Biol 2012;13:283–96.

11. Engelman JA. Targeting PI3K signaling in cancer: opportunities, chal-lenges and limitations. Nat Rev Cancer 2009;9:550–62.

12. Sheppard K, Kinross KM, Solomon B, Pearson RB, Phillips BA.Targeting PI3kinase/AKT/mTOR signaling in cancer. Crit Rev Oncog2012;17:69–95.

13. Bartholomeusz C, Gonzalez-Angulo AM. Targeting the PI3K signalingpathway in cancer therapy. Expert Opin Ther Targets 2012;16:121–30.

14. Floris G, Sciot R, Wozniak A, Van Looy T, Wellens J, Faa G, et al. TheNovel HSP90 inhibitor, IPI-493, is highly effective in human gastro-strointestinal stromal tumor xenografts carrying heterogeneous KITmutations. Clin Cancer Res 2011;17:5604–14.

15. Agaram NP, Besmer P, Wong GC, Guo T, Socci ND, Maki RG, et al.Pathologic and molecular heterogeneity in imatinib-stable or imatinib-responsive gastrointestinal stromal tumors. Clin Cancer Res 2007;13:170–81.

16. Prenen H, Deroose C, Vermaelen P, Sciot R, Debiec-Rychter M,Stroobants S, et al. Establishment of a mouse gastrointestinalstromal tumor model and evaluation of response to imatinib by smallanimal positron emission tomography. Anticancer Res 2006;26:1247–52.






17. Debiec-RychterM,Cools J,DumezH,Sciot R, StulM,MentensN, et al.Mechanisms of resistance to imatinib mesylate in gastrointestinalstromal tumors and activity of the PKC412 inhibitor against imati-nib-resistant mutant. Gastroenterology 2005;128:270–9.

18. Floris G, Debiec-Rychter M, Sciot R, Stefan C, Fieuws S, Machiels K,et al. High efficacy of the (Histone)DeACetylase-inhibitor panobinostat(LBH589) towards human gastrointestinal stromal tumors in a xeno-graft mouse model. Clin Cancer Res 2009;15:4066–76.

19. Rudd ML, Price JC, Fogoros S, Godwin AK, Sgroi DC, Merino MJ,et al. Unique spectrum of somatic PIK3CA (p110alpha) mutationswithin primary endometrial carcinomas. Clin Cancer Res 2011;17:1331–40.

20. Efeyan A, Sabatini DM.mTORand cancer:many loops in onepathway.Curr Opin Cel Biol 2010;22:169–176.

21. Zhang S, Yu D. PI(3)king apart PTEN's role in cancer. Clin Cancer Res2010;16:4325–30.

22. Tarn C, Skorobogatko YV, Taguchi T, Eisenberg B, von Mehren M,Godwin AK. Therapeutic effect of imatinib in gastrointestinal stromaltumors: AKT signaling dependent and independent mechanisms.Cancer Res 2006;66:5477–86.

23. Romeo S, Debiec-Rychter M, Van Glabbeke M, Van Paassen H,Comite P, Van Eijk R, et al. Cell cycle/apoptosis molecule expressioncorrelates with imatinib response in patients with advanced gastroin-testinal stromal tumor. Clin Cancer Res 2009;15:4191–8.

24. HeinrichMC, Corless CL, BlankeCD, Demetri GD, JoensuuH, RobertsPJ, et al. Molecular correlates of imatinib resistance in gastrointestinalstromal tumors. J Clin Oncol 2006;24:4764–74.

25. DanielsM, Lurkin I, Pauli R, Erbst€ober E,Hildebrandt U,HellwigK, et al.Spectrum of KIT/PDGFRA/BRAFmutations and Phosphatidylinositol-3-kinase pathway gene alterations in gastrointestinal stromal tumors(GIST). Cancer Lett 2011;312:43–54.

26. Barretina J, Taylor BS, Banerji S, Ramos AH, Lagos-Quintana M,Decarolis PL, et al. Subtype-specific genomic alterations definenew targets for soft-tissue sarcoma therapy. Nat Genet 2010;42:715–21.

27. Alimonti A, CarracedoA, Clohessy JG, Trotman LC,Nardella C, Egia A,et al. Subtle variations in PTEN dose determine cancer susceptibility.Nat Genet 2010;42:454–9.

28. Berger AH, Pandolfi PP. Haplo-insufficiency: a driving force in cancer.J Pathol 2011;223:137–46.

29. Gunawan B, von Heydebreck A, Sander B, Schulten HJ, Haller F,Langer C, et al. An oncogenetic tree model in gastrointestinal stromaltumours (GISTs) identifies different pathways of cytogenetic evolutionwith prognostic implications. J Pathol 2007;211:463–70.

30. Wozniak A, Sciot R, Guillou L, Pauwels P, Wasag B, Stul M, et al. ArrayCGH analysis in primary gastrointestinal stromal tumours: cytogeneticprofile correlates with anatomic site and tumo aggressiveness irrespec-tiveofmutationalstatus.GenesChromosomesCancer2007;46:261–76.

31. Ricci R,MaggianoN, Castri F, Rinelli A,MurazioM, Pacelli F, et al. Roleof PTEN ingastrointestinal stromal tumor progression.ArchPathol LabMed 2004;128:421–5.

32. AbounaderR. Interactions betweenPTENand receptor tyrosine kinasepathways and their implications for glioma therapy. Expert Rev Anti-cancer Ther 2009;9:235–45.

33. Echeverria CG, Sellers WR. Drug discovery approaches targeting thePI3K/AKT pathway in cancer. Oncogene 2008;27:5511–26.

34. Choi H, Charnsangavej C, Faria SC, Macapinlac HA, Burgess MA,Patel SR, et al. Correlation of computed tomography and positronemission tomography in patients with metastatic gastrointestinal stro-mal tumor treated at a single institution with imatinib mesylate: pro-posal of new computed tomography response criteria. J Clin Oncol2007;25:1753–9.

35. Prenen H, Stefan C, Landuyt B, Vermaelen P, Debiec-Rychter M,Bollen M, et al. Imatinib mesylate inhibits glucose uptake in gastro-intestinal stromal tumor cells by downregulation of the glucosetransporters recruitment to the plasma membrane. Am J BiochemBiotechnol 2005;1:95–102.

36. EUDRACT number 2011-002938-39; Clinical Trials.gov identifierNCT01468688. Available from: http://www.clinicaltrials.gov/ct2/show/NCT01468688?term=NCT01468688&rank=1.

Floris et al.





2013;19:620-630. Published OnlineFirst December 11, 2012.Clin Cancer Res Giuseppe Floris, Agnieszka Wozniak, Raf Sciot, et al. Long-Lasting Responses after Treatment WithdrawalImatinib in Gastrointestinal Stromal Tumor Xenografts: A Potent Combination of the Novel PI3K Inhibitor, GDC-0941, with

Updated version

10.1158/1078-0432.CCR-12-2853doi:

Access the most recent version of this article at:

Material

Supplementary

http://clincancerres.aacrjournals.org/content/suppl/2012/12/11/1078-0432.CCR-12-2853.DC1

Access the most recent supplemental material at:

Cited articles

http://clincancerres.aacrjournals.org/content/19/3/620.full#ref-list-1

This article cites 35 articles, 12 of which you can access for free at:

Citing articles

http://clincancerres.aacrjournals.org/content/19/3/620.full#related-urls

This article has been cited by 11 HighWire-hosted articles. Access the articles at:

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

Subscriptions

Reprints and

[email protected]

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

Permissions

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

.http://clincancerres.aacrjournals.org/content/19/3/620To request permission to re-use all or part of this article, use this link



http://clincancerres.aacrjournals.org/lookup/doi/10.1158/1078-0432.CCR-12-2853

http://clincancerres.aacrjournals.org/content/suppl/2012/12/11/1078-0432.CCR-12-2853.DC1

http://clincancerres.aacrjournals.org/content/19/3/620.full#ref-list-1

http://clincancerres.aacrjournals.org/content/19/3/620.full#related-urls

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

mailto:[email protected]

http://clincancerres.aacrjournals.org/content/19/3/620


APotentCombinationoftheNovelPI3KInhibitor,GDC …...Committee of KU Leuven (Leuven, Belgium). A...

Documents

Transcript of APotentCombinationoftheNovelPI3KInhibitor,GDC …...Committee of KU Leuven (Leuven, Belgium). A...