The Lessons of GIST PET and PET/CT: A New Paradigm for Imaging

The Lessons of GIST PET and PET/CT: A New Paradigm for Imaging Annick D. Van den Abbeele Dana-Farber Cancer Institute, Boston, Massachusetts, USA Key Words. Positron emission tomography Computed tomography Fluorodeoxyglucose Gastrointestinal stromal tumor Imatinib mesylate Disclosure: No potential conflicts of interest were reported by the authors, planners, reviewers, or staff managers of this article. Abstract Traditional anatomic tumor response criteria are based on uni- or bidimensional changes in tumor size, and do not take into account changes in tumor metabolism, tumor density, or decrease in the number of intratumoral vessels. These changes are, however, all indicative of response to imatinib therapy in patients with gastrointestinal stromal tumor (GIST). In these patients, metabolic responses seen on positron emission tomography (PET) using fluorine-18-fluorodeoxyglucose ( 18 FDG) have been shown to be closely related to clinical benefit. Furthermore, these metabolic changes precede by weeks or months significant decrease in tumor size on computed tomography (CT). Conversely, lack of metabolic response on FDG-PET indicates primary resistance to the drug and may help identify patients who would benefit from another therapy, while re-emergence of metabolic activity within tumor sites following a period of therapeutic response indicates secondary resistance to the drug. Newly proposed CT criteria using either no growth in tumor size or a combination of tumor density and size criteria have shown a close correlation with the predictive value results of FDG-PET. Thus, the integration of FDG-PET and CT, as in the combined hybrid PET/CT scanners now available, will not only optimize the evaluation of patients with GIST treated with molecularly targeted drugs, but may ultimately help shorten clinical trials, and accelerate drug development in this disease and other cancers as well. The Oncologist 2008;13(suppl 2):8 13 Introduction Traditional anatomic tumor response criteria, such as the World Health Organization (WHO) criteria and the Response Evaluation Criteria in Solid Tumors (RECIST), are not as useful for sarcomas as they are for other tumor types. This is because sarcomas, including gastrointestinal stromal tumors (GISTs), may not change size in response to therapy [1 4]. Because sarcomas often show high metabolic activity related to intense glycolysis, positron emission tomography (PET) using fluorine-18-fluorodeoxyglucose ( 18 FDG) can be used to evaluate these tumors and the response to therapy. In patients with GIST treated with imatinib mesylate, responses seen on 18 FDG-PET are closely related to clinical benefit while conventional objective response criteria Correspondence: Annick D. Van den Abbeele, M.D., Dana-Farber Cancer Institute, Boston, Massachusetts 02115, USA. Telephone: 617-632-2595; Fax: 617-582-8574; e-mail: abbeele@dfci.harvard.edu Received August 20, 2007; accepted for publication September 25, 2007. AlphaMed Press 1083-7159/2008/$30.00/0 doi: 10.1634/theoncologist.13-S2-8 The Oncologist 2008;13(suppl 2):8 13 www.theoncologist.com based on tumor size, as measured by computed tomography (CT), lag weeks and months behind the 18 FDG-PET imaging results [5, 6]. 18 FDG-PET response is characterized by a marked decrease in the glycolytic metabolism of GISTs in all patients who respond to imatinib. This response can be seen 1 month after initiating therapy, and as early as 24 hours after treatment begins [5]. Tumor response to therapy in sarcomas, including GISTs, is best monitored by using both PET and CT. This dual monitoring can now be performed in one setting on combined hybrid PET/CT scanners, which facilitate both anatomic and functional tumor evaluation. 18 FDG-PET scanning is effective in staging and restaging GISTs, and for evaluating therapeutic response to a variety of treatments. This paper summarizes some of the lessons we have learned about the

Van den Abbeele use of 18FDG-PET (or 18FDG-PET/CT) in GIST and reports on data that have either been published or presented at various national and international meetings since 2001. Relevance of Baseline Evaluation Because 18FDG-PET imaging is a functional imaging study that can evaluate tumor metabolism over time, a baseline scan should always be obtained prior to initiating treatment. Conducting a baseline evaluation allows one to establish a denominator against which future studies or quantitative measurements, such as the standardized uptake value (SUV) or maximum SUV (SUVmax), can be compared. This denominator is essential for characterizing the metabolic response when using the European Organization for Research and Treatment of Cancer (EORTC) criteria, which are based on the magnitude of the change in SUV relative to baseline [7]. 9 other hand, even when a patient benefits from treatment, it may take weeks, months, or even years for these tumors to shrink on CT (Fig. 2). Importantly, 18FDG-PET scans performed soon after treatment begins may identify patients with primary resistance to the drug who may not benefit from this therapy, and for whom alternative treatment should be considered. Response Evaluation 18 FDG-PET may be used to detect both short-term and long-term tumor responses that may not be apparent with CT. As mentioned earlier, 18FDG-PET responses can be observed as early as 24 hours after a single dose of imatinib. A significant change in SUVmax to <2.5 and/or a >25% decrease in SUVmax relative to baseline can be seen within 1 month of starting imatinib therapy in all GIST patients responding to the drug (Figs. 1 and 2). On the Figure 1. FDG-PET scans of a patient with metastatic GIST prior to and 1 month after imatinib therapy. (A): Coronal fluorine-18-fluorodeoxyglucose positron emission tomography (18FDG-PET) scan in a patient with a gastrointestinal stromal tumor metastatic to the abdomen, mesentery, peritoneum, and liver shows intense glycolytic activity in all tumor sites at baseline prior to initiating imatinib. (B): A restaging 18FDG-PET scan 1 month after initiating imatinib therapy shows complete resolution of abnormal 18FDG uptake in all tumor sites and no evidence of metabolically active disease, consistent with response to therapy. www.theoncologist.com Figure 2. FDG-PET and correlating CT scans of a patient prior to and 1 month after imatinib therapy. (A): Coronal fluorine-18-fluorodeoxyglucose positron emission tomography (18FDG-PET) and axial 18FDG-PET scan slice, with corresponding axial computed tomography (CT) slice through the upper abdomen, in a patient with metastatic GIST prior to initiating imatinib shows intense 18FDG uptake in a large left upper quadrant mass arising from the stomach, a liver metastasis, and a small mass in the left lower quadrant. (B): After 1 month of imatinib therapy, there is interval resolution of abnormal glycolytic activity in all disease sites, consistent with response to imatinib therapy despite the fact that the gastric and hepatic lesions are stable by Southwest Oncology Group criteria (i.e., show a <50% bidimensional decrease in size). Note, however, that the densities of the gastric mass and the hepatic lesion have significantly decreased on CT in the interval, and that a small lesion in the posterior hepatic lobe is now much more apparent on CT, but should not be characterized as progressive disease.

10 The Lessons of GIST: A New Paradigm for Imaging Metabolic Responses Are Predictive of Outcome In the pivotal trial of imatinib in metastatic GIST, we reported that patients who achieved an absolute value for SUV max of 2.5 in their tumors 1 month after starting treatment fared much better than patients who did not [5]. The EORTC guidelines for the use of 18 FDG-PET as a biomarker of response suggest that a 25% reduction in SUV max should be considered as the threshold for partial metabolic response (PMR) [7]. We observed a similar prognostic value in this patient population when using that criterion [8, 9]. Conversely, response at 1 month by standardized CT criteria, using the Southwest Oncology Group (SWOG) criteria, was not predictive of outcome [10]. The results in a comparable group of patients from a different center who were imaged later, 2 months after beginning imatinib therapy, indicated that PMR was also predictive of outcome at that time point [11]. We conducted a pilot study in the patient population that participated in the pivotal imatinib trial and was evaluated 1 month following initiation of treatment aimed at optimizing response criteria and assessing their predictive values using outcome analysis [12]. We found that an absolute value for SUV max of 3.4 and a 40% reduction in SUV max were more predictive of time-to-treatment failure (TTF) than the 2.5 SUV max threshold and the 25% reduction (PR) EORTC criteria reported in the original pivotal trial (p =.00002 versus p =.04 and p =.002 versus p =.004, respectively) [5]. While conducting the pivotal trial, we observed that patients clinically benefited from imatinib therapy, even though their tumor masses showed no significant change in size by standard objective CT criteria. We performed bidimensional measurement (BDM) of up to 12 lesions per patient in 52 patients restaged at 1 month, and correlated the measurements with TTF to determine prospective values. As mentioned earlier, we found that SWOG criteria for partial response (PR) (i.e., a 50% reduction in CT-BDM) were not predictive of TTF (log-rank, p =.63). However, patients who had stable disease (SD), as defined by a <50% reduction in CT-BDM, had significantly better outcomes than patients who experienced growth in their tumor masses (p <.00005). Interestingly, the followup data from the pivotal B2222 trial demonstrated that longterm survival (up to 152 weeks) did not differ significantly between GIST patients treated with imatinib who achieved a PR and those who had SD [13]. The results of this pilot study strongly suggest that patients with SD may fare as well as those who have traditional responses, and that new anatomic criteria are needed to assess response to molecular-targeted therapies such as imatinib. This finding may be particularly relevant for patients who may not have access to PET technology and must therefore rely on CT for detection and assessment of GIST. The pilot study results also suggest that, in this patient population, which was naive to the drug, optimized metabolic and anatomic criteria may prove to be superior to traditional anatomic response criteria such as SWOG or RECIST. These data have been peer-reviewed, and the manuscript is currently in press. 18 FDG-PET has also been very helpful in resolving ambiguous findings on CT. For example, patients with a history of GIST metastatic to the liver may present with new lesions in the liver during imatinib treatment. Traditional response criteria, such as SWOG, WHO, or RECIST, would label the appearance of these lesions as disease progression rather than as response to therapy. However, these patients are most likely responding to the treatment. This phenomenon was observed by Joensuu et al. [14] in the very first GIST patient treated with imatinib, when the metastatic hepatic lesions in the index patient became hypodense and showed a decrease in contrast enhancement on magnetic resonance imaging. The same phenomenon has since been reported on CT [15]. It is important to recognize this pattern of response, because it may potentially lead to misinterpretation of progressive disease on CT in a patient who is actually responding to the treatment. These lesions can be isodense to the hepatic parenchyma prior to therapy and may not be seen on CT at that time. As a response to imatinib, there is a change in the density of these tumor masses, which become more cystic and therefore more visible against the surrounding dense hepatic parenchyma. In our experience, 18 FDG-PET can be of great help in resolving these ambiguous CT results. In the previous scenario, 18 FDG-PET will be negative, which confirms the patient s response to treatment (Fig. 2B). 18 FDG-PET is also helpful when CT findings may suggest tumor growth while the increase in size is actually related to intratumoral bleeding or to tumor swelling unrelated to progressive disease. In both cases, 18 FDG-PET will be negative. These observations also underscore the importance of defining updated response criteria for GISTs and other sarcomas. In the absence of 18 FDG-PET or 18 FDG-PET/CT scans, response evaluation can be performed with CT to assess tumor size, detect qualitative changes in tumor density, and provide quantitative information about these density changes in terms of Hounsfield units [3]. For CT technology to be useful, contrast-enhanced and unenhanced images are required. The newly proposed CT response criteria use either no growth in tumor size [12] or a combination of tumor density and size criteria [15] to assess the response of GIST patients to imatinib, and have shown predictive values regarding TTF, as well as a close correlation with the results and pre- The Oncologist

Van den Abbeele 11 dictive values demonstrated by 18 FDG-PET. However, these new CT anatomic criteria are not yet universally accepted, and have been studied only in a patient population that was naive to any treatment with tyrosine kinase inhibitors. These findings require further validation beyond the singlecenter setting, and possibly in a different patient population. Use of 18 FDG-PET to Detect Secondary Resistance As mentioned earlier, 18 FDG-PET can detect primary resistance to imatinib, is useful in the follow-up of patients treated with this drug, and can also aid in the identification of secondary resistance. We have observed that GIST patients who originally respond to imatinib, as demonstrated by the inhibition of glycolytic activity soon after starting the drug, may exhibit re-emergence of glycolytic activity consistent with clonal dedifferentiation and secondary resistance within the site of the original mass, as shown by re-emergence of 18 FDG uptake, months to years afterward (Fig. 3A). Interestingly, termination of imatinib therapy in patients with imatinib-refractory GIST results in a flare phenomenon, in which a significant and rapid upregulation of glucose metabolism occurs in most of the GIST lesions within days after stopping imatinib treatment [16]. This phenomenon was observed in the context of a phase I/II clinical trial investigating the multitargeted receptor tyrosine kinase inhibitor sunitinib malate (Sutent, formerly SU11248; Pfizer Inc., New York). During the trial, patients with imatinib-refractory GIST were taken off imatinib therapy prior to starting sunitinib. Several of these patients had a repeat 18 FDG-PET scan performed days following cessation of imatinib therapy and showed a marked increase in glycolytic activity extending through most of the tumor masses (Fig. 3B). This finding implies that tumor cell populations responsive to imatinib likely remain in these patients, and that secondary resistance and disease progression occur in a subpopulation of cells that most likely develop another mutation that is resistant to imatinib. These results are consistent with the observation that GIST patients who are no longer naive to tyrosine kinase inhibitors may show differential response by 18 FDG-PET, suggesting more than one mutation. The results also raise the question of the potential usefulness of combining tyrosine kinase inhibitors. In that scenario, patients could potentially remain on imatinib therapy if it still inhibits metabolic activity in portions of their tumor mass, and receive a new molecularly targeted drug directed toward the resistant clones. The results also suggest that 18 FDG-PET could become an extremely useful functional noninvasive test to tailor treatment by defining whether a new drug is able to inhibit the reemerging glycolytic activity seen in the resistant clones. This rebound of tumor metabolic activity was also observed in imatinib-resistant or intolerant patients enrolled in the phase I/II trial investigating sunitinib malate [17]. These patients received various schedules of sunitinib malate, which involved alternate periods on and off therapy. 18 FDG-PET was performed at baseline, after the first 7 days on therapy, after the first period off therapy, and while on therapy in subsequent cycles. Glycolytic activity within the tumor sites decreased during the first cycle on sunitinib malate, increased when treatment was stopped, and decreased again during the next treatment Figure 3. Patient with metastatic GIST showing re-emergence of glycolytic activity on FDG-PET while on imatinib therapy indicative of secondary resistance, and metabolic flare after cessation of imatinib. (A): Coronal, sagittal, and axial fluorine-18-fluorodeoxyglucose positron emission tomography ( 18 FDG-PET) scan slices in a patient with GIST who had been treated with imatinib and who developed secondary resistance 14 months after initiation of treatment, as demonstrated by abnormal 18 FDG uptake seen in known metastatic lesions to the liver. (B): Corresponding coronal, sagittal, and axial 18 FDG-PET scan slices in the same patient obtained 18 days after stopping imatinib therapy show marked metabolic rebound of glycolytic activity, or flare. www.theoncologist.com

12 The Lessons of GIST: A New Paradigm for Imaging cycle(s). These results confirmed the efficacy of sunitinib malate in these patients and highlight how 18 FDG-PET can be used as a pharmacodynamic assay to assess target hit and confirm subsequent metabolic tumor response to treatment [17]. Integration of PET and CT ( 18 FDG-PET/CT) Contrast-enhanced CT has conventionally been the method of choice and will continue to play an important role in the management of patients with GIST. However, it is also clear that 18 FDG-PET has been and continues to be instrumental to the success story of the molecularly targeted drugs that are now being used to treat this disease. The integration of anatomic and functional imaging in the combined hybrid PET/CT scanner setting maximizes the benefits of each system and demonstrates their complementarities. This combined imaging approach optimizes the diagnostic and treatment plans for patients with GIST and other sarcomas and has shown greater sensitivity, specificity, and accuracy in the staging and restaging of patients with GIST [18]. The hybrid scanner offers a unique opportunity to test and validate both the new anatomic and functional response criteria in one setting, using the patient as his/her own control. This one-stop-shopping approach may ultimately, and cost-effectively, shorten clinical trials and accelerate drug development. Conclusion Metabolic response by 18 FDG-PET clearly precedes anatomic response in patients whose metastatic GIST is treated with molecularly targeted drugs, and is predictive of outcome. There is therefore a compelling need to integrate functional imaging such as 18 FDG-PET into the routine diagnosis and evaluation of patients with GIST and other sarcomas and to use 18 FDG-PET in the design, testing, and implementation of new drugs and clinical trials in GIST and other cancers. It is also clear that new optimized criteria must be defined to assess response to therapy for both anatomic and functional imaging modalities. Because these modalities are now available in a combined setting, it makes sense to use the one-stop-shopping concept to address all of these issues whenever it is available. The lessons we have learned throughout the GIST experience have created many new paradigms with respect to the role of imaging. These new paradigms are helping define and refine the concept of personalized medicine, and already serve as a new basis on which to evaluate new molecularly targeted drugs in other cancers, including breast, lung, colon, prostate, and brain cancer. Acknowledgments The work presented here has been presented in various local, national, and international meetings and published materials, and is a summary of the outstanding collaborative work among many individuals, including (in alphabetical order): Ramsey Badawi, Ph.D.; Jean-Pierre Cliche, M.D.; George Demetri, M.D.; Dan devries, Ph.D.; Jonathan Fletcher, M.D.; Michael Heinrich, M.D.; Clay Holdsworth, Ph.D.; Judi Manola, Ph.D.; Yulia Melenevsky, M.D.; Leonid Syrkin, Rick Tetrault, M.S., CNMT, RT(N), PET; and Jeffrey Yap, Ph.D., as well as the Dana-Farber Cancer Institute radiology/nuclear medicine and sarcoma teams. We also wish to thank our local, national and international collaborators, our industrial partners, and most importantly our patients. References 1 Miller AB, Hoogstraten B, Staquet M et al. Reporting results of cancer treatment. Cancer 1981;47:207 214. 2 Therasse P, Arbuck SG, Eisenhauer EA et al. New guidelines to evaluate the response to treatment in solid tumors. European Organization for Research and Treatment of Cancer, National Cancer Institute of the United States, National Cancer Institute of Canada. J Natl Cancer Inst 2000;92:205 216. 3 Choi H. Response evaluation of gastrointestinal stromal tumors. The Oncologist 2008;13(suppl 2):4 7. 4 Jaffe CC. Response assessment in clinical trials: Implications for sarcoma clinical trial design. The Oncologist 2008;13(suppl 2):14 18. 5 Van den Abbeele AD, for the GIST Collaborative PET Study Group Dana- Farber Cancer Institute, Boston, Massachusetts; OHSU, Portland, Oregon, Helsinki University Central Hospital, Turku University Central Hospital, Finland, Novartis Oncology. F18-FDG-PET provides early evidence of biological response to STI571 in patients with malignant gastrointestinal stromal tumors (GIST). Proc Am Soc Clin Oncol 2001;20:362a. 6 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;347:472 480. 7 Young H, Baum R, Cremerius U et al. Measurement of clinical and subclinical tumour response using [18F]-fluorodeoxyglucose and positron emission tomography: Review and 1999 EORTC recommendations. European Organization for Research and Treatment of Cancer (EORTC) PET Study Group. Eur J Cancer 1999;35:1773 1782. 8 Van den Abbeele AD, Badawi RD, Cliche JP et al. 18 F-FDG-PET predicts response to imatinib mesylate (Gleevec TM ) in patients with advanced gastrointestinal stromal tumors (GIST). Proc Am Soc Clin Oncol 2002;21:403a. 9 Van den Abbeele AD, Badawi RD, Tetrault RJ et al. FDG-PET as a surrogate marker for response to Gleevec (imatinib mesylate) in patients with advanced gastrointestinal stromal tumors (GIST). J Nucl Med 2003;44(suppl):24P. 10 Green S, Weiss GR. Southwest Oncology Group standard response criteria, endpoint definitions and toxicity criteria. Invest New Drugs 1992;10:239 253. The Oncologist

Van den Abbeele 13 11 Gayed I, Vu T, Iyer R et al. The role of 18 F-FDG PET in staging and early prediction of response to therapy of recurrent gastrointestinal stromal tumors. J Nucl Med 2004;45:17 21. 12 Holdsworth CH, Manola J, Badawi RD et al. Use of computerized tomography (CT) as an early prognostic indicator of response to imatinib mesylate (IM) in patients with gastrointestinal stromal tumors (GIST). Proc Am Soc Clin Oncol 2004;22(14 suppl):197. 13 Blanke C, Joensuu H, Demetri G et al. Long-term follow up of advanced gastrointestinal stromal tumor (GIST) patients treated with imatinib mesylate. Proc Gastroint Canc Symp 2004;2:43. 14 Joensuu H, Roberts PJ, Sarlomo-Rikala M et al. Effect of the tyrosine kinase inhibitor STI571 in a patient with a metastatic gastrointestinal stromal tumor. N Engl J Med 2001;344:1052 1056. 15 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;183:1619 1628. 16 Van den Abbeele AD, Badawi RE, Manola J et al. Effects of cessation of imatinib mesylate (IM) therapy in patients (pts) with IM-refractory gastrointestinal stromal tumors (GIST) as visualized by FDG-PET scanning. J Clin Oncol 2004;22(14 suppl):198. 17 Van den Abbeele A, Melenevsky Y, de Vries D et al. Imaging kinase target inhibition with SU11248 by FDG-PET in patients (pts) with imatinibresistant gastrointestinal stromal tumors (I-R GIST). Proc Am Soc Clin Oncol 2005;23:817s. 18 Antoch G, Kanja J, Bauer S et al. Comparison of PET, CT, and dual-modality PET/CT imaging for monitoring of imatinib (ST1571) therapy in patients with gastrointestinal stromal tumors. J Nucl Med 2004;45:357 365. www.theoncologist.com