• gastrointestinal stromal tumor;
  • gene expression;
  • imatinib mesylate;
  • insulin-like growth factor binding protein-3;
  • positron emission tomography


  1. Top of page
  2. Abstract
  6. Acknowledgements


Imatinib has demonstrated marked clinical efficacy against gastrointestinal stromal tumor (GIST). Microarray technology, real-time polymerase chain reaction (PCR) validation, and fluorodeoxyglucose-positron emission tomography (FDG-PET) imaging were used to study the early molecular effects of imatinib antitumor activity in GIST.


After exposure of sensitive and resistant sarcoma cell lines to imatinib for 24 to 48 hours, the changes in gene expression were evaluated using a 1146 unique pathway array with Western blot validation. Real-time PCR was used to confirm changes in gene expression in human GIST samples (preimatinib biopsy and postimatinib surgical specimen after 3–7 days of therapy). FDG-PET was performed to correlate radiographic findings with the effects of imatinib on gene expression in GIST.


In all, 55 genes demonstrated a ≥ 2-fold change after imatinib treatment of the GIST882 cells. Among these genes there was up-regulation of insulin-like growth factor binding protein-3 (IGFBP-3), a protein that modulates proliferation and apoptosis. Western blot analysis confirmed the increase of IGFBP-3 only in imatinib-sensitive GIST882 cells. Up to a 7-fold induction (49% mean increase; P = .08) of IGFBP-3 mRNA was found in tumor samples from patients with low residual FDG uptake, whereas there was an up to 12-fold reduction (−102% mean decrease; P = .03) in IGFBP-3 in those patients with high residual FDG uptake after imatinib therapy.


In the current study, imatinib appears to regulate numerous genes and specifically induces IGFBP-3 in GIST cells and tumor samples. IGFBP-3 levels also were found to be inversely correlated with residual FDG uptake in GIST patients early in imatinib therapy. These initial observations suggest that IGFBP-3 is an important early marker of antitumor activity of imatinib in GIST. Cancer 2006. © 2006 American Cancer Society.

Malignant gastrointestinal stromal tumor (GIST) is a type of sarcoma that occurs primarily in the gastrointestinal (GI) tract and abdomen. Although GIST represents < 1% of all tumors of the GI tract, it is the most common mesenchymal malignancy of this site.1 Approximately 4000 new cases of GIST are diagnosed annually in the U.S., predominantly in middle-aged or older persons.2, 3 GISTs are often clinically silent until they reach a large size, may be associated with life-threatening hemorrhage, and exhibit a high rate of subsequent recurrence and metastatic spread. Consequently, GIST is often not diagnosed until it is at an incurable stage.

Before the availability of tyrosine kinase inhibitors, there was no effective systemic therapy for advanced GIST.4 The median survival of patients with advanced GIST was 2 years after diagnosis.2 Chemotherapy regimens such as doxorubicin, ifosfamide, or other agents used in the treatment of soft-tissue sarcomas are reported to have limited activity in patients with GIST.4–6

The use of 1 selective tyrosine kinase inhibitor, imatinib mesylate (also known as STI571; Gleevec, Novartis Pharmaceuticals, East Hanover, NJ), indicates that inhibition of the kinase function of KIT serves as an effective anticancer therapy for patients with GIST.7–9 A decrease in the standardized uptake value noted on positron emission tomography (PET) scanning provides a sensitive early indication of response that predicts long-term clinical benefit,10 correlates with response by computed tomograph (CT) size-density assessment,11, 12 and has been recommended by a consensus panel of GIST experts.13 Patients who were evaluated < 3 days after initiation of imatinib were reported to have residual standardized uptake values approaching baseline.10

Although imatinib has had marked clinical success, the early molecular events within tumor cells after exposure to imatinib are under investigation. We approached this task by studying gene expression profiles of imatinib-sensitive and imatinib-resistant sarcoma cell lines in vitro and found up-regulation of insulin-like growth factor binding protein 3 (IGFBP3) in imatinib-sensitive GIST cells. IGFBP3, when secreted, is a major protein involved in the sequestration of insulin-like growth factor (IGF) from the IGF receptor (IGFR).14 In addition, IGFBP3 is induced by tumor suppressor gene p53 and localizes to the nucleus,15 exerting IGFR-independent effects such as inhibition of cell proliferation and induction of apoptosis.16 To determine whether IGFBP3 was up-regulated in tumors from imatinib-treated patients with GIST, we studied tumor samples before and after imatinib treatment and correlated IGFBP-3 expression with residual FDG uptake by PET imaging. To our knowledge, this is the first study to demonstrate a correlation between early radiographic events and molecular changes in tumor samples from patients with GIST on imatinib therapy. This implicates the increase in IGFBP3 as a potential molecule involved in the radiographic findings and antitumor activity of imatinib in GIST.


  1. Top of page
  2. Abstract
  6. Acknowledgements

Cell Lines and GIST Tumor Tissue

Eight human sarcoma cell lines were used to determine the gene expression profile in imatinib-sensitive and imatinib-resistant sarcoma cells. The imatinib-sensitive GIST cell line GIST882 (kindly provided by Dr. Jonathan Fletcher, Dana-Farber Cancer Institute, Boston, MA) encodes an exon 13 K642E mutation. TM03 is a primary GIST culture that has a wild-type pdgfr-α sequence and a kit exon 9 AY insertion at codon 502 displaying in vitro resistance to imatinib (Fig. 1). The remaining cell lines were purchased from the American TypeCulture Collection (ATCC; Manassas, VA): A204 (unclassified sarcoma with wild-type kit and pdgfr-α), RD (rhabdomyosarcoma), SW872 (fibrosarcoma), SW684 (liposarcoma), SKLMS-1 (leiomyosarcoma), and SaOS-2 (osteosarcoma). Cells were maintained under the conditions recommended by the ATCC and GIST882 cells were grown as previously described.17

thumbnail image

Figure 1. (A) Viability of sarcoma cells in the presence of imatinib mesylate in vitro. The MTT assay was performed and the viability curve plotted for imatinib-treated and untreated sarcoma cell lines. All cells were treated for 72 hours with imatinib at the specified concentrations or with vehicle only. (B) Induction of apoptosis after exposure of sarcoma cells to imatinib. A flow-cytometric terminal deoxynucleotidyl transferase (TDT)-mediated deoxy-uridine triphosphate (dUTP) biotin nick-end labeling (TUNEL) assay was performed after sarcoma cell lines were exposed to imatinib or the vehicle for 24 hours.

Download figure to PowerPoint

Eleven patients with KIT-expressing GIST who were undergoing surgical resection were enrolled in an Institutional Review Board-approved protocol at the University of Texas, M. D. Anderson Cancer Center. After informed consent was obtained, patients underwent a core needle biopsy of the primary tumor. Patients were subsequently randomized to 3, 5, or 7 days of imatinib mesylate at a dose of 600 mg. At the completion of preoperative imatinib mesylate, patients proceeded with their surgical procedure, which occurred within 6 hours of the last dose of imatinib. Biopsy and surgical tumor tissue were immediately processed.

Imatinib Solution

Imatinib (provided by Novartis Oncology), was dissolved in water to a concentration of 10 mM and then filtered (0.22 micron Millex low protein binding sterile filter; Millipore, Bedford, MA). The solution was stored in aliquots at −80°C.

Cell Viability Assay

The effect of imatinib on the viability of the sarcoma cells at incrementally increasing concentrations was assessed using the thiazolyl blue tetrazolium dye (MTT; 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-tetrazolium bromide) assay (Sigma-Aldrich Chemical Company, St. Louis, MO). Exponentially growing cell suspensions (5 × 103 cells/mL) were seeded onto 96-well microtiter plates (100 μL per well). After 24 hours of incubation at 37°C, 100 μL of imatinib solution was added to bring the final concentration to 0-, 0.5-, 1.0-, or 5.0 μM. After incubation for 72 hours, the MTT assay was performed.

The optical density of the samples was measured at 570 nanometers (nm) using the KC4 analysis program (Bio-Tek Instruments, Winooski, VT) for a Microsoft Windows-based computer interfaced with a microplate reader (Bio-Tek Instruments). Cell viability was calculated as the (mean absorbance of 3 replicate wells containing imatinib minus the mean absorbance of 3 replicate background wells)/(the mean absorbance of 3 replicate imatinib-free wells minus the mean absorbance of 3 replicate background wells) × 100 and conducted independently in triplicate.

Assessment of Apoptosis after Exposure to Imatinib

Cells in log-phase growth in 100-mm tissue culture dishes at 80% confluence were treated with medium with or without 5-μM imatinib for 24 hours, harvested, washed with phosphate-buffered saline (PBS), and fixed for 15 minutes with 1% formaldehyde, then 70% ethanol. Cells were stored at −20°C for at least 24 hours.

For the terminal deoxynucleotidyl transferase (TDT)-mediated deoxy-uridine triphosphate (dUTP) biotin nick-end labeling (TUNEL) assay, the pellets were washed with 1 mL PBS, then resuspended overnight in 50 μL of a solution containing TDT buffer and bromo-dUTP (AU1001; Phoenix Flow Systems, San Diego, CA). HL60 cells treated with camptothecin served as a control for apoptosis. Cells were stained with a solution of avidin plus fluorescein isothiocyanate (FITC) in the dark at room temperature. For cell cycle analysis, 500 μL of propidium iodide plus RNase (Phoenix Flow Systems) was added for 30 minutes on ice. The samples were analyzed using a flow cytometer (Epics XL-MCL; Beckman Coulter, Fullerton, CA) with System II software for 2-color detection. Finally, using Multicycle software (Phoenix Flow Systems), we calculated the percentages of cells in the G1, S-, and G2 phases to determine the effect of imatinib on the cell cycle and in the sub-G1 phase to corroborate its apoptotic effect. The percentage of cells that were positive for FITC was calculated with the same software to confirm the effect of imatinib on apoptosis.

Microarray Analysis

Microarray experiments were performed as described previously18, 19 on slides spotted with oligonucleotide sequences (70 nucleotides in length) that were custom-produced in the Cancer Genomics Core facility at M. D. Anderson Cancer Center (available at URL:˜genomics [accessed January 17, 2006]). The array contains 1146 genes that are functionally well characterized and spotted in duplicate. Many of these genes are involved in cell cycle control, apoptosis, DNA repair, metabolism, cell-cell adhesion, invasion, and metastasis.

Matched untreated and treated (24- and 48-hour) sarcoma cell lines were processed for total RNA extraction. Biopsy and surgical tissues were stored in RNAlater. The cells and tissue were lysed (TRI; Molecular Research Center, Cincinnati, OH), their total RNA extracted, and tested for quality using an Agilent (Palo Alto, CA) Bioanalyzer.

Microarray experiments were performed as described previously.19 The hybridized slides were scanned with a microarray scanner (GeneTAC LS IV; Genomic Solutions, Ann Arbor, MI), quantified with ArrayVision II software (Imaging Research, St. Catherines, ON, Canada), and values were recorded for spot intensity, local background intensity, and signal-to-noise ratio. The image data were imported into S-plus software (Insightful, Seattle, WA) for background-correction, log-transformation (base 2) and smooth t-statistics computation. The genes whose t-scores exceeded |3| were selected as the ones that were differentially expressed.20 The analyzed data were imported into GeneSpring 6 software (Silicon Genetics/Agilent Technologies, Redwood City, CA) and were visualized as tree clusters using a similarity measure of standard correlation.

Western Blot Analysis

Cells were treated with 5-μM imatinib for 24 or 48 hours and then washed twice with PBS containing 5 mM ethylenediamine tetraacetic acid (EDTA) and 1 mM sodium orthovanadate (Calbiochem, San Diego, CA). We next lysed the cells in phospho-lysis buffer with 10 μg/mL phenylmethylsulfonyl fluoride, 1% Nonidet p-40, 50-mM Tris-HCl (pH 8.0), 5-mM EDTA, 30-mM sodium pyrophosphate, 2-mM sodium molybdate, 100-mM sodium fluoride, 2-mM sodium orthovanadate, and 1× protease inhibitor cocktail (Calbiochem). The cell lysates were collected and the protein concentration of the supernatant fluid was measured using the BCA protein assay (Pierce Biotechnology, Rockford, IL).

We performed gel electrophoresis on 8% SDS polyacrylamide gels (Mini-Protean II electrophoresis cell; Bio-Rad Laboratories, Hercules, CA) as previously described.21 The membranes were probed with primary antibodies: anti-KIT (Santa Cruz Biotechnology, Santa Cruz, CA), anti-PDGFRα (Upstate Biotechnology, Lake Placid, NY), anti-MAPK (Santa Cruz Biotechnology), anti-phospho-MAPK (Santa Cruz Biotechnology), and anti-IGFBP3 (Cell Signaling Technology, Beverly, MA) per the manufacturers' recommendations. Membranes were washed with PBS plus 0.05% Tween-20 (PBS-T) 4 times for 5 minutes each time before we incubated the membrane with secondary antibodies (horseradish peroxidase-conjugated antigoat [Dako Cytomation, Carpinteria, CA], antirabbit [R&D Systems, Minneapolis, MN], and antimouse [R&D Systems] for an hour at room temperature at a dilution of 1:2000). Membranes were washed with PBS-T, incubated for 1 minute in enhanced chemiluminescence solution (Amersham Life Science, Piscataway, NJ), and subjected to autoradiography.

Positron Emission Tomography Imaging

FDG-PET scans were performed by the Division of Radiology at the M. D. Anderson Cancer Center. FDG-PET was performed using a CTI ECAT HR+ PET scanner (Siemens, Erlangen, Germany) after administration of 10–15 mCi (370–555 MBq) of FDG over 1 minute. All patients had nothing by mouth at least 6 hours before scanning. After a 60-minute uptake phase, patients were scanned from the neck to the pelvis; 5- and 3-minute transmission scans (for attenuation correction) were obtained for each field of view in 2-dimensional mode. The CT used for attenuation correction was acquired in helical mode (speed, 13.5 mm/rotation) from the base of the skull to mid-thigh during suspended mid-expiration at a 3.75-mm slice thickness, 140 kVp, and 120 mA. The images were interpreted using volumetric projection and multiple orthogonal projection analysis on a Xeleris workstation. Vendor-specific software was used to measure the postimatinib residual maximum standardized uptake value (SUVmax) for each patient with GIST. Residual SUVmax is an early marker of efficacy in several tumor types including lymphoma, esophageal carcinoma, and non-small-cell lung cancer.22–24 Our experience with GIST, lymphoma, thyroid cancer, and esophageal cancer suggests that those patients with an SUVmax < 3.9 should be considered responders, whereas patients with an SUVmax ≥ 3.9 should be considered nonresponders.24–26

Quantitative Real-Time Reverse Transcriptase-PCR for IGFBP3 in GIST

The ABI Prism 7900HT Sequence Detection System (Applied Biosystems, Foster City, CA) was used to assess transcript levels of IGFBP3. IGFBP3 primers were ordered by using the Applied Biosystems Assay-on-Demand system (assay ID, Hs00181211_m1). Housekeeping gene primer mix cyclophilin A (assay ID, Hs99999904_m1) was used as the endogenous control. All primer-probe mixes were preoptimized by Applied Biosystems. Reverse transcriptase (RT)- and quantitative real-time PCR were performed in 25-μL reactions per the manufacturer's protocol. Cycling conditions were 30 minutes at 48°C and 10 minutes at 95°C, followed by 45 cycles of 15 seconds at 95°C and 1 minute at 60°C. Amplicons were quantified on a cycle-by-cycle basis by the emission intensities. All experiments were performed in triplicate with samples obtained from the patients before and after imatinib treatment. Transcript levels in the various samples were calculated after normalization to the internal reference standard, cyclophilin-A.


  1. Top of page
  2. Abstract
  6. Acknowledgements

Imatinib Differentially Inhibits Viability of Sarcoma Cell Lines by Induction of Apoptosis

A cell viability assay (MTT) was performed to determine the effects of imatinib on sarcoma cell lines (Fig. 1). Compared with other sarcoma cell lines, the GIST cell line GIST882 was found to undergo the greatest reduction in cell viability (Fig. 1A) and those cells underwent a 40% absolute reduction in the number of viable cells at imatinib concentrations as low as 0.1 μM. Although A204 cells underwent a 20% absolute reduction in cell viability, these cells required a 4 μM imatinib concentration before any effect was observed. The modest imatinib sensitivity of A204 may be related to platelet-derived growth factor receptor (PDGFR) expression but does not appear to be due to mutation because these cells encode both the wildtype PDGFR and kit genes (data not shown). In comparison, the remaining sarcoma cell lines were resistant to the inhibitory effects of imatinib (Fig. 1A).

To determine whether the decreased viability of sarcoma cells treated with imatinib, as indicated by the MTT assay, was due to inhibition of cell cycle progression or to induction of apoptosis, TUNEL and flow cytometric analyses were performed. The TUNEL assay revealed that the rate of apoptosis was 24.9% for GIST882 cells and 15.4% for A204 cells treated with 5 μM imatinib for 24 hours. Conversely, untreated GIST882 and A204 cells were found to have a rate of apoptosis of 2.1% and 1.3%, respectively (Fig. 1B). Cell cycle analysis indicated a similar increase in the proportion of treated cells in the sub-G1 phase (data not shown). This finding suggested that the reduction in GIST882 and, to a lesser degree, A204 cell viability was due to apoptosis.

Imatinib Inhibits Signaling Pathways in Imatinib-Sensitive Sarcoma Cell Lines

We then evaluated the sarcoma cell lines for expression of KIT and PDGFRα as well as the ability of imatinib to inhibit the downstream signaling molecule, mitogen-activated protein kinase (MAPK), in those cell lines that expressed the tyrosine kinase receptor targets of imatinib (Fig. 2A). GIST882 cells were found to express both Kit and PDGFRα, whereas TM03 and A204 cells were found to express only PDGFRα. The remaining sarcoma cell lines expressed neither Kit nor PDGFRα.

thumbnail image

Figure 2. (A) Expression of Kit and platelet-derived growth factor receptor-α (PDGFRα) in sarcoma cells treated with imatinib mesylate. Western blot analysis was performed to determine protein expression of targets of imatinib in sarcoma cell lines. (B) Inhibition of signal-transduction pathways by imatinib in human sarcoma cell lines. Gastrointestinal stromal tumor (GIST) cells GIST882 and TM03 were exposed to stem cell factor or vehicle in the presence or absence of 5 μM imatinib. Phosphoprotein analysis was performed by Western blot analysis. Blots were probed for mitogen-activated protein kinase (MAPK) and phospho-MAPK.

Download figure to PowerPoint

Next we evaluated the ligand independence and inhibitory effects of imatinib on the phosphorylation of MAPK in the GIST cell lines (Fig. 2B). GIST882 and TM03 cells were grown in culture with stem cell factor (SCF) or with imatinib followed by stem cell factor or were left untreated. Incubation of the GIST cell lines with SCF did not increase the phosphorylation of MAPK. Imatinib was able to abrogate phosphorylation of MAPK in GIST882 but not in TM03 cells. The total levels of MAPK were found to remain constant. Thus, imatinib effectively inhibited signaling through Kit in GIST882 cells.

Identification of Imatinib-Altered Genes in Sarcoma Cell Lines by Microarray Analysis

To identify genes that are involved in imatinib response, we performed microarray analysis. Control sarcoma cell lines were treated with only vehicle, whereas the experimental group was treated with 5 μM imatinib for 24 or 48 hours. RNA isolated from the imatinib-treated cell lines was labeled with Cy3 and the untreated RNA was labeled with Cy5. Cohybridization with oligonucleotides on microarray slides allowed identification of genes that were either up- or down-regulated after imatinib exposure for each sarcoma cell line; thus, each sarcoma cell line served as its own control. Our analysis identified 55 genes that were altered ≥ 2-fold after exposure to imatinib at both 24 and 48 hours in GIST882 but not the other sarcoma cell lines. Thus, genes were only selected if their expression was altered by imatinib in only the imatinib-sensitive GIST cells but not the other sarcoma cell lines and not the imatinib-resistant GIST cells. This enrichment process was undertaken to select for GIST-specific genes and against the nonspecific effects of imatinib. Among the induced genes is the known apoptosis-promoting IGFBP3 (Table 1). We found no changes in the GLUT-1 or hexokinase mRNA expression levels.

Table 1. Selected Genes Altered by Imatinib in Sarcoma Cell Lines
Genes affectedLog ratio
  1. TNF indicates tumor necrosis factor.

Up-regulated24 h48 h
 Insulin-like growth factor binding protein-31.062.88
 Protein tyrosine phosphatase receptor type K1.641.51
 TNF-α-induced protein 31.061.4
 Cathepsin L−2.86−3.07
 Cyclin D3−1.27−2.04
 Tissue inhibitor of metalloproteinase 3−1.49−1.59

To determine whether gene expression alteration is indicative of cellular response to imatinib mesylate, we performed a hierarchical cluster analysis of all 1146 genes with the sarcoma cell lines in this study. We found that imatinib-treated GIST882 and A204 cells clustered together at both timepoints (Fig. 3). The imatinib-resistant sarcoma cell lines, including the GIST cell line TM03, did not cluster with imatinib-sensitive GIST882 cells on hierarchical analysis of all genes in the dataset.

thumbnail image

Figure 3. Hierarchical clustering of the differential gene expression values for sarcoma cell lines at 2 time points. Each cell line was treated with 5 μM imatinib mesylate or vehicle for 24 or 48 hours. This dendrogram shows the hierarchical cluster tree corresponding to the effect of imatinib on each sarcoma cell line. The hierarchical cluster tree was constructed using software analysis. The metric used was the incremental sum of squares within each cell line group as a result of joining groups.

Download figure to PowerPoint

Validation of Up-Regulation of IGFBP3 in GIST Cells and Human Tissue

To our knowledge, this is the first study whereby IGFBP3 was found to be induced by imatinib. We were intrigued by this observation because IGFBP3 is known to be an apoptosis-inducing protein. To gain insight into the potential involvement of IGFBP3 in response to imatinib, we first evaluated IGFBP3 protein in sarcoma cells lines before and after exposure. Protein was isolated from the sarcoma cells with and without treatment and examined for IGFBP3 expression by Western blot analysis. In GIST882 cells the level of IGFBP3 underwent up-regulation after 24 and 48 hours (Fig. 4).

thumbnail image

Figure 4. Induction of insulin-like growth factor binding protein-3 (IGFBP-3) expression by imatinib mesylate in human gastrointestinal stromal tumor (GIST) cells but not other sarcoma cell lines by Western blot analysis. IGFBP-3 is expressed in the kit-expressing imatinib-treated GIST882 cells but not other sarcoma cell lines not in the Kit-negative GIST cell line. β-Actin serves as a control for loading.

Download figure to PowerPoint

Effects of Imatinib on IGFBP-3 Expression in Human GIST Samples

Because imatinib blocked Kit signaling and induced IGFBP3 expression in our cell lines, we next sought to determine whether there was any relation between IGFBP3 expression and FDG-PET imaging after treatment of patients with imatinib. FDG-PET imaging is an early indicator of clinical benefit when GIST patients are treated with imatinib.10, 11 GIST patients in our clinical trial underwent a preimatinib core biopsy, imatinib therapy (3, 5, or 7 days), postimatinib FDG PET scan, and surgical resection of the tumor. Patients were observed to undergo rapid and significant responses to imatinib therapy as assessed by PET imaging. In a typical case a patient underwent preimatinib imaging that consisted of noncontrast CT (Fig. 5A), FDG-PET (Fig. 5B), and fusion PET-CT (Fig. 5C). After 5 days of imatinib therapy the patient had a significant reduction in FDG-PET activity (Fig. 5E and 5F) with no change in tumor size (Fig. 5D). We next sought to determine whether this indicator of imatinib activity correlated with alteration of IGFBP-3 expression.

thumbnail image

Figure 5. Typical positron emission tomography (PET) imaging findings when a patient with gastrointestinal stromal tumor (GIST) is treated with imatinib. The GIST of the patient is visualized in the lower abdominal cavity and indicated by the white arrow. Noncontrast (A and D) computed tomography (CT), (B and E) PET imaging, and (C and F) fusion PET-CT were obtained simultaneously before imatinib therapy (A-C) and after 5 days of therapy with imatinib given at a dose of 300 mg, orally, twice a day (D-F). Although there was no change in tumor size (A and D), there was a marked decrease in the maximum standardized uptake value (SUVmax) observed by PET (B-E) and by the overlay of PET with noncontrast CT (C-F).

Download figure to PowerPoint

We isolated RNA from patient samples before and after imatinib treatment and analyzed IGFBP3 expression using RT-PCR. The results demonstrated that there were 2 distinct patterns of response of IGFBP3 among the patients (Fig. 6). IGFBP3 was induced in 6 patients' samples after imatinib therapy (log ratio of [postimatinib/preimatinib]). Among patients whose tumors had increased IGFBP-3 expression, the expression increased up to 7-fold. In contrast, IGFBP3 was reduced in 3 patients (up to 12-fold) after treatment with imatinib. Interestingly, the patients' samples that were found to have undergone an up-regulation in IGFBP-3 were more likely to have kit mutation in exon 11 (4/6 cases), whereas those samples with a reduction in IGFBP-3 were found to have no mutations in kit exon 11.

thumbnail image

Figure 6. Patients on a prospective trial underwent baseline core needle biopsy and positron emission tomography (PET) imaging, imatinib therapy, repeat PET imaging, and subsequent resection of their tumor. Expression of insulin-like growth factor binding protein-3 (IGFBP-3) mRNA was evaluated by quantitative reverse transcriptase-polymerase chain reaction (RT-PCR) in patients before and after imatinib and subjected to Student t test for paired data. (A) Patients whose tumor had no residual fluorodeoxyglucose (FDG) uptake by PET imaging were found to have an increase (49%; P = .08) in the level of IGFBP-3 by RT-PCR. (B) Those patients whose tumors had residual FDG uptake were found to have a decrease in IGFBP-3 levels (102%; P = .03).

Download figure to PowerPoint

We found a 49.2% mean reduction in IGFBP-3 mRNA levels in those patients with a poor response to imatinib as evidenced by high residual FDG uptake after therapy (P = .03), whereas those patients with minimal residual uptake by FDG-PET scan had a mean 102% increase in IGFBP-3 mRNA levels (P = .08) by quantitative RT-PCR (Fig. 6).


  1. Top of page
  2. Abstract
  6. Acknowledgements

The purpose of the current study was to determine early molecular events that occur with imatinib in GIST and whether these events correlate with radiographic findings. We used a microarray approach to determine which genes were altered after imatinib with subsequent analysis performed in tissue obtained prospectively on a clinical trial. One gene, IGFBP3, is up-regulated at both the RNA and protein levels after exposure of GIST cells and human tumor tissue to imatinib. Moreover, decreases in IGFBP3 levels were detected in patients whose GIST exhibited residual FDG uptake by PET imaging. There was a trend for those patients without residual FDG uptake to have an increase in IGFBP3 levels after imatinib.

Although other studies have identified genes that are differentially regulated in GIST cells, to our knowledge this is the first study that has identified genes that are regulated in GIST cells after 24 hours of imatinib and confirmed during the first few days of imatinib therapy of GIST patients. Subramanian et al.27 analyzed 26 frozen GIST tumors and described differential gene expression between those with a kit rather than PDGFRα mutation. The effect of imatinib on these genes was not described.

Frolov et al.28 found that 148 of 10,367 genes were found to be differentially expressed in GIST cells after exposure to imatinib. The GIST cells were treated with imatinib for 6 to 48 hours and were not reported to undergo apoptosis, even when exposed to supratherapeutic concentrations of imatinib. The gene expression observed in tissue samples was that of tumor tissue before and after up to 18 months of imatinib therapy.28 One could speculate that the residual GIST cells in a tumor after 18 months of imatinib therapy express RNA that encodes genes for imatinib resistance rather than sensitivity. This absence of imatinib-induced apoptosis in that study may contribute to the difference in gene expression between their study and ours; we and Tuveson et al.17 both found that apoptosis occurred after GIST cells were exposed to imatinib.

Our microarray analysis identified several differentially expressed genes after imatinib exposure. Genes were selected if they were differentially expressed after imatinib in the imatinib-sensitive GIST cell line GIST882 but not in the cell lines that were not GIST or not sensitive to imatinib. We believe this approach enriches for genes of biologic and therapeutic importance. One of these genes, IGFBP3, is an interesting molecule in the setting of GIST because it has the ability to induce apoptosis and regulate glucose metabolism, an important feature of FDG-PET imaging.29 The IGF signaling pathway appears to be important in multiple aspects of tumor phenotype such as survival, invasion, and metastasis. Moreover, blockage of IGF signaling leads to apoptosis, prevention of metastasis, and reversal of the tumor phenotype.30–32 Our findings demonstrated that IGFBP3 is up-regulated at the RNA and protein levels after exposure of imatinib-sensitive GIST cells to imatinib for only 24 hours and after as few as 3 days of imatinib therapy of patients with GIST. This temporal relation and correlation with PET imaging suggests that up-regulation of IGFBP3 may be involved in the early events leading to imatinib's antitumor activity. This would not be surprising because IGFBP3 has been shown to inhibit proliferation and induce apoptosis in other types of cancer.33, 34

In addition to inhibiting proliferation and survival of tumor cells, IGFBP3 plays a role in glucose metabolism, a property fundamental to FDG-PET imaging. IGFBP3 binds to IGFs with an affinity 10 to 100 times higher than that of its affinity for the IGF-1 receptor.35 Because hexokinase is the first and rate-limiting enzyme in the glycolytic pathway, this enzyme plays a critical role in the cellular uptake of FDG in metabolically active tumors. Although hexokinase RNA transcription and translation in glial cells occurs within 4 hours of their exposure to IGF,36 we did not observe down-regulation of hexokinase or glucose transporter mRNA in cell lines or patient samples exposed to imatinib. This does not exclude the possibility that hexokinase or glucose transporter activity may be posttranscriptionally or posttranslationally inhibited by imatinib. Those patients in our study who were poor responders to imatinib were the same patients whose tumor displayed residual glucose metabolism by FDG uptake and decreased IGFBP-3 mRNA. One could speculate that perhaps the decrease in IGFBP-3 resulted in the availability of IGFs to bind IGFRs and permit continued residual glucose metabolism and tumor cell survival.

IGFBP3 is markedly up-regulated after imatinib therapy in imatinib-sensitive GIST cells and human tumor tissue according to our cDNA microarray analysis and confirmatory studies. Moreover, the multiple biologic functions of IGFBP3 make it feasible that this is the molecule that carries out both the antitumor effects of imatinib and inhibition of FDG uptake. Our observation that IGFBP-3 up-regulation tended to occur in those patients whose tumors encoded kit exon 11 mutation and were imatinib-sensitive by PET imaging supports the importance of this molecule. IGFBP3 was neither expressed nor up-regulated in the imatinib-sensitive, PDGFR-expressing A204 cell line, suggesting that it may be a gene specific for GIST cells or the Kit pathway. Moreover, IGFBP3 was not up-regulated in imatinib-resistant sarcoma cell lines nor in the tumor tissue. The importance of IGFBP3 is well known for other tumor types, but its expression has not previously been described in GISTs.

We found that imatinib up-regulates and down-regulates genes in GIST. We found that imatinib induces IGFBP3 expression specifically in GIST cell lines and human tumor tissue obtained early in the patient's treatment with imatinib. The increase we found in IGFBP3 in imatinib-sensitive GIST cells and tumor tissue suggests that this protein may be an important early marker of FDG-PET imaging and the antitumor activity of imatinib. Investigation of this protein in a larger number of human GIST tumor samples as well as other tumor cell lines and tissue is warranted.


  1. Top of page
  2. Abstract
  6. Acknowledgements

We thank Novartis Oncology, Basel, Switzerland, for providing imatinib mesylate.


  1. Top of page
  2. Abstract
  6. Acknowledgements
  • 1
    Fletcher CD, Berman JJ, Corless C, et al. Diagnosis of gastrointestinal stromal tumors: a consensus approach. Int J Surg Pathol. 2002; 10: 8189.
  • 2
    DeMatteo RP, Lewis JJ, Leung D, Mudan SS, Woodruff JM, Brennan MF. Two hundred gastrointestinal stromal tumors: recurrence patterns and prognostic factors for survival. Ann Surg. 2000; 231: 5158.
  • 3
    Nilsson B, Beumming P, Meis-Kindblom JM, et al. Gastrointestinal stromal tumors: the incidence, prevalence, clinical course, and prognostication in the preimatinib mesylate era—a population-based study in western Sweden. Cancer. 2005; 103: 821829.
  • 4
    Trent JC, Beach J, Burgess MA, et al. A two-arm phase II study of temozolomide in patients with advanced gastrointestinal stromal tumors and other soft tissue sarcomas. Cancer. 2003; 98: 26932699.
  • 5
    Patel S, Vadhan-Raj S, Papadopoulos N, et al. High-dose Ifosfamide in bone and soft-tissue sarcomas — results of phase II and pilot studies — dose response and schedule dependence. J Clin Oncol. 1997; 15: 23782384.
  • 6
    Patel S, Vadhan-Raj S, Burgess M, et al. Results of two consecutive trials of dose-intensive chemotherapy with doxorubicin and ifosfamide in patients with sarcomas. Am J Clin Oncol. 1998; 21: 317321.
  • 7
    Demetri GD, von Mehren M, Blanke CD, et al. Efficacy and safety of imatinib mesylate in advanced gastrointestinal stromal tumors. N Engl J Med. 2002; 347: 472480.
  • 8
    Rankin C, von Mehren M, Blanke CD, et al. Dose effect of imatinib (IM) in patients (pts) with metastatic GIST — Phase III Sarcoma Group Study S0033. Am Soc Clin Oncol. 2004; 23: 815. Abstract 9005.
  • 9
    Verweij J, Casali PG, Zalcberg J, et al. Progression-free survival in gastrointestinal stromal tumours with high-dose imatinib: randomised trial. Lancet. 2004; 364: 11271134.
  • 10
    Van den Abbeele A, Badawi R, Cliche J-P, et al. 18F-FDG PET predicts response to imatinib mesylate (Gleevec) in patients with advanced gastrointestinal stromal tumors (GIST). Proc Am Soc Clin Oncol. 2002; 21: 403a. Abstract 1610.
  • 11
    Choi H, Charnsangavej C, de Castro Faria S, et al. CT evaluation of the response of gastrointestinal stromal tumors after imatinib mesylate treatment: a quantitative analysis correlated with FDG PET findings. AJR Am J Roentgenol. 2004; 183: 16191628.
  • 12
    Choi H, Charnsangavej C, Macapinlac H, et al. Correlation of computerized tomography (CT) and proton emission tomography (PET) in patients with metastatic GIST treated at a single institution with imatinib mesylate. Am Soc Clin Oncol. 2003; 22: 819. Abstract 3290.
  • 13
    Blay JY, Bonvalot S, Casali P, et al. Consensus meeting for the management of gastrointestinal stromal tumors. Report of the GIST Consensus Conference of 20–21 March 2004, under the auspices of ESMO. Ann Oncol. 2005; 16: 566578.
  • 14
    Renehan AG, Zwahlen M, Minder C, O'Dwyer ST, Shalet SM, Egger M. Insulin-like growth factor (IGF)-I, IGF binding protein-3, and cancer risk: systematic review and meta-regression analysis. Lancet. 2004; 363: 13461353.
  • 15
    Schedlich LJ, Graham LD. Role of insulin-like growth factor binding protein-3 in breast cancer cell growth. Microsc Res Tech. 2002; 59: 1222.
  • 16
    Schedlich LJ, O'Han MK, Leong GM, Baxter RC. Insulin-like growth factor binding protein-3 prevents retinoid receptor heterodimerization: implications for retinoic acid-sensitivity in human breast cancer cells. Biochem Biophys Res Commun. 2004; 314: 8388.
  • 17
    Tuveson DA, Willis NA, Jacks T, et al. STI571 inactivation of the gastrointestinal stromal tumor c-KIT oncoprotein: biological and clinical implications. Oncogene. 2001; 20: 50545058.
  • 18
    Wang G, Kong H, Sun Y, et al. Genome-wide analysis of the cyclin family in Arabidopsis and comparative phylogenetic analysis of plant cyclin-like proteins. Plant Physiol. 2004; 135: 10841099.
  • 19
    Shmulevich I, Hunt K, El-Naggar A, et al. Tumor specific gene expression profiles in human leiomyosarcoma: an evaluation of intratumor heterogeneity. Cancer. 2002; 94: 20692075.
  • 20
    Baggerly KA, Coombes KR, Hess KR, et al. Identifying differentially expressed genes in cDNA microarray experiments. J Comput Biol. 2001; 8: 639659.
  • 21
    Trent JC2nd, McConkey DJ, Loughlin SM, Harbison MT, Fernandez A, Ananthaswamy HN. Ras signaling in tumor necrosis factor-induced apoptosis. EMBO J. 1996; 15: 44974505.
  • 22
    Hutchings M, Mikhaeel NG, Fields PA, Nunan T, Timothy AR. Prognostic value of interim FDG-PET after two or three cycles of chemotherapy in Hodgkin lymphoma. Ann Oncol. 2005; 16: 11601168.
  • 23
    Naumann R, Vaic A, Beuthien-Baumann B, et al. Prognostic value of positron emission tomography in the evaluation of post-treatment residual mass in patients with Hodgkin's disease and non-Hodgkin's lymphoma. Br J Haematol. 2001; 115: 793800.
  • 24
    Swisher SG, Maish M, Erasmus JJ, et al. Utility of PET, CT, and EUS to identify pathologic responders in esophageal cancer. Ann Thoracic Surg. 2004; 78: 11521160; discussion 1152–1160.
  • 25
    Kazama T, Faria SC, Varavithya V, Phongkitkarun S, Ito H, Macapinlac HA. FDG PET in the evaluation of treatment for lymphoma: clinical usefulness and pitfalls. Radiographics. 2005; 25: 191207.
  • 26
    Wang W, Macapinlac H, Larson SM, et al. [18F]-2-fluoro-2-deoxy-D-glucose positron emission tomography localizes residual thyroid cancer in patients with negative diagnostic (131)I whole body scans and elevated serum thyroglobulin levels. J Clin Endocrinol Metab. 1999; 84: 22912302.
  • 27
    Subramanian S, West RB, Corless CL, et al. Gastrointestinal stromal tumors (GISTs) with KIT and PDGFRA mutations have distinct gene expression profiles. Oncogene. 2004; 23: 77807790.
  • 28
    Frolov A, Chahwan S, Ochs M, et al. Response markers and the molecular mechanisms of action of Gleevec in gastrointestinal stromal tumors. Mol Cancer Ther. 2003; 2: 699709.
  • 29
    Baxter RC, Butt AJ, Schedlich LJ, Martin JL. Antiproliferative and pro-apoptotic activities of insulin-like growth factor-binding protein-3. Growth Horm IGF Res. 2000; 10: S1011.
  • 30
    Brodt P, Samani A, Navab R. Inhibition of the type I insulin-like growth factor receptor expression and signaling: novel strategies for antimetastatic therapy. Biochem Pharmacol. 2000; 60: 11011107.
  • 31
    Mauro L, Salerno M, Morelli C, Boterberg T, Bracke ME, Surmacz E. Role of the IGF-I receptor in the regulation of cell-cell adhesion: implications in cancer development and progression. J Cell Physiol. 2003; 194: 108116.
  • 32
    Zhang D, Brodt P. Type 1 insulin-like growth factor regulates MT1-MMP synthesis and tumor invasion via PI 3-kinase/Akt signaling. Oncogene. 2003; 22: 974982.
  • 33
    Camirand A, Pollak M. Co-targeting IGF-1R and c-kit: synergistic inhibition of proliferation and induction of apoptosis in H 209 small cell lung cancer cells. Br J Cancer. 2004; 90: 18251829.
  • 34
    Lee HY, Moon H, Chun KH, et al. Effects of insulin-like growth factor binding protein-3 and farnesyltransferase inhibitor SCH66336 on Akt expression and apoptosis in non-small-cell lung cancer cells. J Natl Cancer Inst. 2004; 96: 15361548.
  • 35
    Oesterreicher S, Blum WF, Schmidt B, Braulke T, Keubler B. Interaction of insulin-like growth factor II (IGF-II) with multiple plasma proteins: high affinity binding of plasminogen to IGF-II and IGF-binding protein-3. J Biol Chem. 2005; 280: 999410000.
  • 36
    Sebastian S, Kenkare UW. Stimulation of brain hexokinase gene expression by recombinant brain insulin-like growth factor in C6 glial cells. Exp Cell Res. 1999; 246: 243247.