Reports

This report is based on medical evidence presented at sanctioned medical congress, from peer reviewed literature or opinion provided by a qualified healthcare practitioner. The consumption of the information contained within this report is intended for qualified Canadian healthcare practitioners only.

FRONTLINE - Based on presentations from the Second Eastern Ontario Gastrointestinal Stromal Tumours (GIST) Consensus Meeting

Ottawa, Ontario / November 6, 2009

Two years ago, the first eastern Ontario consensus conference on the detection, staging and treatment of gastrointestinal stromal tumours (GISTs) provided a summary of current management in the context of regional resources. A second conference was recently held to accommodate several advances important to the definition of optimal care that have occurred since that initial meeting. This summary, like the summary of the previous conference, is expected to be relevant to other regions of Canada as well as regions outside of Canada where rapidly evolving concepts of care must be translated into practical management strategies. Not all of the most advanced initiatives in the identification, evaluation and treatment of GIST, particularly those that remain experimental, are necessarily applicable at centres where these tumours are first encountered or followed. It was the design of this conference, summarized in this publication, to update standards of care for GIST in practical terms.

Conference Objectives

• Summarize recent advances in the diagnosis and molecular classification of GISTs • Assess application of adjuvant therapy of GIST including selection of patients and duration of treatment • Consider opportunities for management of progressive GIST with novel systemic therapy or surgery • Determine direction of research and educational needs

Conference Outcomes

• Patients with GISTs require referral and management by expert multidisciplinary teams consisting of surgeons, pathologists, diagnostic radiologists and oncologists familiar with this entity. • All patients with GISTs in eastern Ontario should be registered in the GIST database. • Pathology should be reported as per a newly devised synoptic report. Standard immunohistochemistry should be employed but should also include DOG1 (Discovered on GIST 1). • Mutational testing should be undertaken at a central laboratory as part of an ongoing research and quality assurance process. • Patients with early-stage, resected GISTs should be assessed for relapse risk using Miettinen criteria and if risk is >20%, should be offered adjuvant imatinib 400 mg/day for at least one year. • Patients with initially unresectable GISTs or in whom a resection would result in critical loss of functionality should be considered for neoadjuvant therapy with imatinib for three to 12 months with careful follow-up to determine optimal surgical timing. • Patients with advanced GISTs should be offered imatinib 400 mg/day with dose escalation to 800 mg/day or switch to sunitinib upon progression. Therapy for such patients should be considered for an indefinite period of time as rapid progression occurs upon discontinuation of tyrosine kinase inhibitor (TKI) therapies. • Consideration should be given to using a higher dose of imatinib (800 mg/day) in patients with a known exon 9 mutation. • Imatinib levels should be considered in patients experiencing undue toxicity or those in whom predicted response is suboptimal.

Introduction

In this second eastern Ontario consensus regional conference on the diagnosis, evaluation and treatment of gastrointestinal stromal tumours (GISTs), the focus was again placed on identifying practical opportunities to improve care of this malignancy. These deliberations, like those previously published, were conducted to guide care consistent with commonly available clinical tools, including imaging modalities, laboratory tests and therapies. As opposed to emerging treatment strategies and promising initiatives pursued in the research setting, practical guidelines are essential when boundaries on resource use and reimbursement require a consistent and evidence-based approach.

At the last consensus conference, the agenda included a series of initial presentations on the epidemiology and natural history of GIST followed by a summary of where current research has defined opportunities for appropriate care. While a historical background on GIST provided common ground on which to establish the status of current strategies, this more recent conference did not revisit such areas as disease incidence, basic pathophysiology, or other aspects of GIST that have remained essentially unchanged since the previous conference. Although a very brief summary of this information begins this document, more complete information regarding these subjects can be drawn from the previous summary or from one of the many recently published review articles.^1,2,3

This conference summary, like the meeting itself, has been designed as an update of clinically relevant information that has become available since the initial consensus conference. The most important sections of this and the previous consensus documents involve recommendations for the diagnosis and treatment of early-stage, resected GIST. Of the clinically relevant advances introduced since the last consensus conference, a phase III, multicentre trial of adjuvant use of imatinib mesylate for resected GIST is considered the most significant. The relevance of this study to current practice is reviewed first. This document will also cover new information on the classification and treatment of GIST resistant to imatinib, including the use of molecular markers to guide therapy, the approach to GISTs of borderline resectability and the options for treatment of metastatic GIST.

Brief Background

GISTs are the most common mesenchymal tumours of the GI tract, representing 1% to 3% of all GI malignancies. Approximately 60% of GISTs occur in the stomach, but GISTs have been observed throughout the GI tract.⁴ Approximately 80% of GISTs express c-KIT (CD117), a 145-kDa transmembrane glycoprotein receptor of the TK subclass III, which also includes platelet-derived growth factor (PDGF). Although the absence of CD117 expression does not rule out GIST, mutations in the KIT gene play an important role in the dysregulated growth that characterizes most GISTs.⁵ Mutations in the PDGF-receptor-alpha (PDGFRA) gene appear to account for dysregulated cell growth in a substantial proportion of GISTs that are CD117-negative. In tumours negative for mutations in both KIT and PDGFRA genes, loss of tumour suppressor genes in chromosomes 14q and 22q appear to be the most important pathogenic mechanism. Such GISTs, which represent fewer than 5% of cases, often have other atypical features, including clinical indolence.

Like most malignancies, early detection of GIST, which is palpable in less than 40% of cases, correlates with likelihood of successful therapy.⁶ Size, mitotic count and location are among factors that predict outcome. In a study comparing outcome based on 12 characteristics, including perivascular growth pattern, nuclear atypia and muscularis propria infiltration, five-year disease-free rates ranged from 99.3% in patients with primary tumours exhibiting the most favourable features to 43.9% in those with tumours exhibiting the least favourable set of characteristics.⁷ The ability of imatinib and other tyrosine kinase inhibitors (TKIs) to block activation of c-KIT and PDGFRA is a frequently cited demonstration of the potency of targeted therapy when key molecular steps are known and suppressed. However, surgery is the only potentially curative therapy for GIST, making diagnosis of tumours during the time of resectability a critical step toward an optimal outcome.

Therapy Update: Role of Adjuvant Imatinib

Surgery is the only curative treatment for GIST. In GIST, as in other cancers, the goal of resection is disease-free (R0) margins. However, R0 margins provide relatively modest protection from relapse in GIST with unfavourable characteristics, such as large size or high mitotic rate. In some series, 50% of tumours with unfavourable characteristics recur within five years despite R0 margins at the time of resection.⁸ For this reason, adjuvant therapies have been explored in an effort to reduce relapse rates in high-risk primary tumours. However, as in metastatic GIST, neither conventional cytotoxic agents nor radiotherapy has proven effective at reducing the risk of relapse or prolonging survival.⁹

In the now published double-blind, phase III study (ACOSOG Z9001) of adjuvant imatinib, 713 adults with a KIT-positive primary GIST of at least 3 cm in size were randomized to receive 400 mg imatinib daily or placebo for one year after surgical resection.¹⁰ The primary end point was recurrence-free survival. Accrual in the study was stopped early when the predefined boundary for efficacy in favour of imatinib was crossed during an interim analysis. One year after surgery, the hazard ratio (HR) for a recurrence was 65% lower in the group receiving adjuvant imatinib relative to those randomized to placebo (HR 0.35; 95% CI, 0.22-0.53; P<0.0001). The rate of recurrence or death after a median of 19.7 months of follow-up was less than half as great in the imatinib arm (8% vs. 20%). Survival rates have not yet differed significantly, but follow-up is ongoing. The results of the trial led the US Food and Drug Administration to issue an accelerated conditional approval for adjuvant use of imatinib for GIST with the provision that overall survival data be submitted for further review. Health Canada has recently issued a notice of compliance indicating imatinib for the adjuvant treatment of adults at intermediate to high risk of relapse following complete resection of KIT-positive GIST.

The study confirmed an adjuvant benefit of imatinib in the population selected for which the primary entry criterion was a KIT-positive tumour >3 cm. Patients were not stratified for mitotic rate. The patient characteristics were typical of a GIST population, including a median age near 60 and a relatively even distribution of men and women. The performance status was 0 in approximately 75% of patients, 1 in 23% of patients and 2 in 2% of patients. Approximately 40% of the tumours measured 3 to 6 cm (25% >10 cm) and 60% were in the stomach. R0 margins were achieved in 92% of the patients; the remainder, except for one unknown, were R1. Imatinib was characterized as well tolerated, producing adverse events consistent with previous studies.

By tumour size, the advantage of imatinib was consistent but not necessarily greater for larger vs. smaller tumours. The authors, while acknowledging that mitotic count has been previously associated with prognostic value, explained that mitotic rate was not an entry criterion in this study due to the absence of standardized measurements, potentially complicating interpretation of this variable in a multicentre study. These results provide an important validation of adjuvant imatinib and define a population in which benefit can be anticipated. However, they also create numerous questions, including whether treatment might be beneficial in patients defined as high risk by criteria other than those employed in this study.

Many of the most widely cited methodologies for determining prognosis in GIST include criteria other than tumour size, especially mitotic rate. In a study that compared major methods of prognostication, including the one created by the Armed Forces Institute of Pathology (AFIP), which is among the most influential, mitotic count >5 per high powered field (HPF), tumour necrosis and serosal involvement were all found to have predictive value independent of size.¹¹ Importantly, risk of recurrence occurs on a continuum. In individuals with tumour size <2 cm and a mitotic rate of <5 HPF risk of relapse may be extremely low, but it climbs incrementally with larger size and increasing mitotic rate (Table 1). Tumour location is a significant factor in determining risk of relapse.

Table 1. Risk Stratification (AFIP criteria)

ACOSOG Z9001 has established criteria for benefit from imatinib as an adjuvant therapy, but it does not rule out important benefits in patients who did not meet study criteria, such as those with tumours <3 cm but >2 cm with a high mitotic rate. While the European Medicines Agency (EMEA) conferred imatinib with an indication for adjuvant use, it does not recommend its use in patients with a low risk of recurrence. Duration of therapy is another important question not answered in the current study. In ACOSOG Z9001, the protection against recurrence diminished rapidly after imatinib was discontinued (Figure 1).

Figure 1. ACOSOG Z9001: Recur
ing Adjuvant Therapy

<img3782|center>

It is unclear whether a longer course of treatment would eventually produce additional advantages. Several studies evaluating imatinib with different entry criteria than those employed in ACOSOG Z9001 may provide greater insight on patient selection and duration

Table 2. Imatinib Studies in Adjuvant and Neoadjuvant Settings

<img3783|center>

Attendees here at the conference were unanimous that adjuvant imatinib should be discussed with and offered to high-risk patients (>20% risk of relapse) in Ontario. Based on multiple prognostic algorithms, it is appropriate to consider both mitotic rate and tumour size in selecting patients for this treatment. The AFIP risk stratification criteria provide reasonable guidance for patient selection at the current time. The recommended dose of imatinib in this setting is 400 mg/day for 12 months.

Primary (Preoperative) Therapy with Imatinib

Neoadjuvant therapy for reducing tumour bulk and surgical resection of metastatic GIST is also a rapidly evolving area. In GIST of borderline resectability, the advantages of an initial course of a TKI include a potential to reduce morbidity associated with surgery while lowering the risk of post-operative recurrence. However, unresectability is often poorly defined and there is a risk of progression or metastatic spread during the period of neoadjuvant therapy. Moreover, the optimal duration of such preoperative therapy has not been defined in any well-conducted prospective trials. While clinical trials with imatinib suggest that the time to optimal response is approximately three months,12 the median time to progression (TTP) is 12 months,13 suggesting a window of preoperative therapy in the range of three to five months. In the only published study of neoadjuvant treatment for GIST, the RTOG 0132/ACRIN 6665 phase II study, 30 patients with a locally advanced tumour, as defined by the investigator, received daily imatinib (600 mg) for eight to 12 weeks followed by surgical resection. The results indicated that neoadjuvant treatment was safe, in that no patient progressed on preoperative treatment, there was no increase in surgical complications, and that an R0 resection was achieved in 77%. Further studies will be required to define the role of neoadjuvant imatinib in GIST.

Strategies for Advanced Disease (Unresectable or Metastatic)

The activity of the TKI imatinib in unresectable or metastatic GIST is well established,¹² but many questions persist, including those regarding dose, duration and course of action in the event of primary resistance or treatment failure. In resistant patients, doubling the dose of imatinib from 400 mg to 800 mg has been associated with an improvement in response rates,¹⁴ but a study comparing 400 mg to 600 mg in previously untreated advanced GIST patients found no significant difference in time to progression or overall survival.¹⁵ When patients who progressed on the lower dose were switched to the higher dose, 33% achieved an objective response or stable disease.

The absence of a significant benefit argues against a higher dose in previously untreated advanced GIST patients, particularly as adverse events were both more frequent and more severe on 800 mg of imatinib vs. 400 mg. However, there may be a benefit from a higher dose in patients with primary resistance, which occurs in about 15% of patients with GIST, or those who acquire resistance on prolonged therapy. Primary resistance has been found to be most common with KIT exon 9 mutations, PDGFRA D842V mutations and the wild-type KIT or PDGFRA genotypes. In patients with primary resistance mutations, it may be appropriate to consider a higher dose of imatinib or an alternative TKI. An 800-mg dose of imatinib has been associated with a longer median progression-free survival (PFS) than a 400-mg dose in patients with KIT exon 9 mutation (19 vs. 6 months; P=0.017) in a small and unpublished study,¹⁶ leading to speculation that routine mutational analysis should be performed when this is suspected (e.g. small bowel primary GIST).

Despite the evidence that an initial imatinib dose of 800 mg is not beneficial as a routine strategy in unselected patients, the dose-response relationship is important in patients with resistance. In addition to the greater PFS observed in patients with KIT exon 9 mutation when treated with imatinib 800 mg, plasma concentrations predicted response for GIST with either KIT exon 9 or KIT exon 11 mutations in a study that stratified patients by trough imatinib plasma quartile.¹⁷ Compared by time to TTP, those in the three highest quartiles experienced significantly greater disease control than those in the lowest quartile (Figure 2). Furthermore, imatinib pharmacokinetics (PK) are known to vary significantly by a variety of clinically relevant fac
eight, granulocyte count and hemoglobin.¹⁸

Figure 2. TTP According to Cmin Quartiles

<img3784|center>

To the extent that the PK of imatinib might predict for response and PFS, steady-state plasma drug concentration analyses might be useful for identifying patients who are not achieving therapeutic levels, but it is not yet clear whether this should be performed in all patients. There is increasing consensus that plasma levels should be carried out in patients at special risk for abnormal drug metabolism, in those experiencing undue toxicity or in patients who do not respond after a reasonable trial of therapy at the 400-mg daily dose. Testing of drug levels may also be appropriate when non-compliance with therapy is suspected. The overall role of plasma levels in PK-driven therapy requires confirmation of the preceding observations in a large-scale prospective study.

Very similar issues are raised when one considers the role of mutational analysis and choice of initial TKI or dose of imatinib. Although molecular analysis may also play a role in the diagnosis of GIST, the specific molecular characteristics of GIST do have prognostic value. For example, KIT exon 11 mutations project better outcomes than exon 9 mutations in gastric tumours and have the potential for defining groups who might benefit from higher doses of imatinib or an alternative TKI (Table 3).¹⁹ Indeed, specific mutational characteristics have been employed to predict respon
uch as enhanced sensitivity to sunitinib in the presence of mutations in exon 13 or 14.²⁰

Table 3. GIST Mutation Type Is Prognostic

<img3785|center>

While mutational analysis at the time of recurrence on imatinib may be helpful to characterize the cause of progression, the rationale for performing mutational analysis at the time of diagnosis in unresectable GIST is increasing because of the potential to immediately optimize patient treatment options. Eventually such molecular characterization may be undertaken routinely in all GIST, whether resectable or unresectable, because of the importance of mutation status for subclassifying GIST with distinct biologic and clinical behaviours. The rationale on which to support mutational analysis
d by data confirming that these steps can alter decisions or outcomes, but conference participants considered some preliminary recommendations (Table 4).

Table 4. GIST Mutational Analysis at Diagnosis

<img3786|center>

Information about the clinical significance of mutations in GIST is expected to expand substantially by an ambitious ongoing effort at the Université de Montréal to perform genotyping on a large series of GIST tissue samples collected in eastern Ontario. The molecular features, including mutations, will be correlated with a variety of pathologic features in an effort to correlate these characteristics to prognosis and response to therapies. The analyses of the first 238 cases are near completion.

There is less controversy regarding duration of therapy for patients with advanced disease. Supportive data include the French BFR14 study which clearly demonstrated tha
matinib in responding patients led to rapid progression (Figure 3). Hence, there is now consensus that treatment in patients with advanced disease should be continued indefinitely or until progression.

Figure 3. BFR14 Three-year Randomization PFS

<img3788|center>

Surgery for Metastatic Disease

Surgery for advanced GIST after the tumour has been controlled with a TKI appears to be of benefit if a complete resection can be performed before the tumour develops resistance, usually within one year of initiating the TKI. Progression on TKI can be separated into two types: focal progression and multifocal progression. Focal progression occurs within 12 months and may represent an acquired resistance to TKI. This scenario can sometimes be salvaged with surgical resection. Multifocal progression, on the other hand, occurs within three months and portends a poor prognosis, despite surgery. In a study of PFS following surgical resection, patients with stable disease on imatinib prior to surgery had the most favourable outcome, while patients with generalized progression during imatinib treatment
prognosis (Figure 4). The differences in PFS between patients with stable disease and focal progression (P=0.002) and between patients with focal progression and generalized disease progression (P<0.0001) were significant.

Figure 4. PFS According to Preoperative Disease Classification

<img3787|center>

In another single-centre study, there was no evidence of disease after surgery in 78% of patients who were stable on TKI therapy at the time of surgery, 25% of those with limited progression and 7% of those with generalized progression.¹³ Based on these results, surgery has been recommended in select patients with advanced disease who have stable disease or focal progression during TKI treatment, in order to reduce tumours from which drug-resistant clones might develop and to delay disease progression.²¹

Systemic Therapy: Options for Progressive Disease

The efficacy of sunitinib in imatinib failures has been supported by a phase III, placebo-controlled trial that enrolled 312 patients with unresectable GIST who had progressed on imatinib.22 The median time to tumour progression was 27.3 weeks on sunitinib vs. 6.4 weeks on placebo (P<0.0001) (Figure 5). Sunitinib was well tolerated in this study, but it is important to emphasize that the types of side effects on different TKIs vary and that the phase III study employed an intermittent dosing schedule that may be better tolerated
inib. In general, sunitinib tends to be less well tolerated than imatinib and may require dose reductions for mucositis or hand-foot syndrome. Patients on sunitinib should be monitored for blood cell counts, thyroid function, blood pressure and cardiac function.

Figure 5. Phase III Trial: Sunitinib in Advanced GIST After Imatinib Failure

<img3789|center>

The unique activity profiles of TKIs suggest that it may be possible to treat beyond first- and second-line agents. In an uncontrolled study of third-line nilotinib, PFS was longer than would be anticipated in an untreated patient, and approximately half of patients remained alive one year after starting therapy.²³ An ongoing phase II study is randomizing patients who have progressed on both imatinib and sunitinib to nilotinib or best supportive care (that may include re-challenge with imatinib or sunitinib). Other TKIs, such as masitinib, which was recently evaluated in a phase I GIST study,²⁴ raise the potential for further expansion of therapeutic options.

Emerging Areas: Progress in Classifying and Treating GIST

The potential for rapid evolution in the diagnosis, classification and management of GIST is substantial. Relatively recently, a gene overexpressed in 94.4% of all GIST, called DOG1, has shown substantial promise as a target of more rapid and specific diagnostic tests.²⁵ Testing for DOG1 now appears likely to be incorporated into routine diagnostic strategies.

Similarly, the progress in characterizing GIST and predictors of outcome has promulgated a variety of criteria for risk stratification (Table 5) that are likely to further evolve. Relatively basic components of how information is captured, such as the definition of an HPF for quantifying mitotic activity, have yet to be standardized. The rapid expansion of characteristics with prognostic importance has created the potential for confusion in the absence of standards for how and which information shoul
ptic pathology report developed at the University of Ottawa to ensure that relevant information is captured and to direct attention to those areas most important for patient management may be widely applicable (Table 6). The goal of this report is to standardize the evaluation of GIST and to provide guidance on evolving standards of therapy.

Table 5. Evolution of Risk Stratification

<img3790|center>

Therapeutic strategies are also expected to evolve not only for resistant and advanced GIST but in front-line treatment. Although imatinib is the established first-line therapy in advanced or metastatic G
studies, it has not been directly compared to alternative TKIs as a front-line treatment. A multinational phase III study that included Canadian centres is now assessing outcome in a study that directly compared imatinib to nilotinib in previously untreated advanced GIST (Figure 6). A retrospective analysis of mutational status may provide insight about relative activity of the two agents.

Table 6. GIST Patient Pathology Report

<img3791|center>

The rapid progress in classifying and understanding the molecular ch
a cancer not differentiated from other mesenchymal tumours until the early 1990s, has been relatively unique among life-threatening malignancies. Although early resection remains the only effective avenue to cure, TKIs not only provide significant survival advantages in patients with advanced, unresectable disease but also have a proven benefit in the adjuvant therapy of early-stage resected disease.

Figure 6. Nilotinib First-line Registration Study CAMN107G2301 (ENESTg1)

<img3792|center>

^Summary

Consensus conferences to define current standards in the management of cancer are rarely required at two-year intervals, but the progress made in the diagnosis and treatment of GIST since 2007 has been substantial. Although confirmation of a benefit from adjuvant imatinib in patients with resectable GIST is among the most important advances that have taken place in the interval since the last meeting, the most recent conference was witness to expanding areas of information in a broad array of topics. It provided increased opportunity for collaboration and consensus building in order to optimize outcomes for patients with GISTs.

References 1. Biasco et al. Gastrointestinal stromal tumours: report of an audit and review of the literature. Eur J Cancer Prev 2009;18:106-16.

2. Reichardt et al. Gastrointestinal stromal tumours I: pathology, pathobiology, primary therapy and surgical issues. Semin Oncol 2009;36:290-301.

3. Benjamin et al. Gastrointestinal stromal tumours II : medical oncology and tumour response assessment. Semin Oncol 2009;36:302-11.

4. Corless et al. Biology of gastrointestinal stromal tumors. J Clin Oncol 2004;22:3813-25.

5. Hirota et al. Gain-of-function mutations of c-kit in human gastrointestinal stromal tumors. Science 1998;279:577-80.

6. Nowain et al. Gastrointestinal stromal tumors: Clinical profile, pathogenesis, treatment strategies and prognosis. J Gastroenterol Hepatol 2005;20:818-24.

7. Hou et al. Predictive values of clinical and pathological parameters for malignancy of gastrointestinal stromal tumors. Histol Histopathol 2009;24:737-47.

8. DeMatteo et al. Two hundred gastrointestinal stromal tumors: recurrence patterns and prognostic factors for survival. Ann Surg 2000;231:51-8.

9. DeMatteo et al. Clinical management of gastrointestinal stromal tumors: before and after STI-571. Hum Pathol 2002;33:466-77.

10. DeMatteo et al. Adjuvant imatinib mesylate after resection of localised, primary gastrointestinal stromal tumour: a randomised, double-blind placebo-controlled trial. Lancet 2009;373:1097-104.

11. Goh et al. Which is the optimal risk stratification system for surgically treated localized primary GIST? Comparison of three contemporary prognostic criteria in 171 tumors and a proposal for a modified Armed Forces Institute of Pathology risk criteria for gastrointestinal stromal tumors (GISTs). Ann Surg Oncol 2008;15(8):2153-63.

12. Demetri et al. Efficacy and safety of imatinib mesylate in advanced gastrointestinal stromal tumors. N Engl J Med 2002;347:472-80.

13. Raut et al. Surgical management of advanced gastrointestinal stromal tumors after treatment with targeted systemic therapy using kinase inhibitors. J Clin Oncol 2006;24:2325-31.

14. Verweij et al. Progression-free survival in gastrointestinal stromal tumours with high-dose imatinib: randomised trial. Lancet 2004;364:1127-34.

15. Blanke et al. Long-term results from a randomized phase II trial of standard-versus higher-dose imatinib mesylate for patients with unresectable or metastatic gastrointestinal stromal tumors expressing KIT. J Clin Oncol 2008;26:620-5.

16. Van Glabbeke et al. Comparison of two doses of imatinib for the treatment of unresectable or metastatic gastrointestinal stromal tumors (GIST): A meta-analysis based on 1640 patients (pts). ASCO 2007, Abstract 10004.

17. Demetri et al. Imatinib plasma levels are correlated with clinical benefit in patients with unresectable/metastatic gastrointestinal stromal tumors. J Clin Oncol 2009;27:3141-7.

18. Judson et al. Imatinib pharmacokinetics in patients with gastrointestinal stromal tumour: a retrospective population pharmacokinetic study over time. EORTC Soft Tissue and Bone Sarcoma Group. Cancer Chemother Pharmacol 2005;55:379-86.

19. Miettinen et al. Gastrointestinal stromal tumors of the stomach: A clinicopathologic, immunohistochemical, and molecular genetic study of 1765 cases with long-term follow-up. Am J Surg Pathol 2005;29:52-68. 20. Lasota J, Miettinen M. Clinical significance of oncogenic KIT and PDGFRA mutations in gastrointestinal stromal tumours.Histopathology 2008;53:245-66. 21. Gold JS, DeMatteo RP. Combined surgical and molecular therapy: The gastrointestinal stromal tumor model. Ann Surg 2006;244:176-84.

22. Demetri et al. Efficacy and safety of sunitinib in patients with advanced gastrointestinal stromal tumour after failure of imatinib: a randomised controlled trial. Lancet 2006;368:1329-38.

23. Montemurro et al. Nilotinib in the treatment of advanced gastrointestinal stromal tumours resistant to both imatinib and sunitinib. Eur J Cancer 2009;45:2293-7.

24. Soria et al. Phase 1 dose-escalation study of oral tyrosine kinase inhibitors masitinib in advanced and/or metastatic solid cancers. Eur J Cancer 2009;45:2333-41.

25. Miettinen et al. DOG1 antibody in the differential diagnosis of gastrointestinal stromal tumors: a study of 1840 cases. Am J Surg Pathol 2009;33:1401-8.