Optimal Use of Targeted Agents for Advanced Gastrointestinal Stromal Tumours

Abstract

Imatinib is the recommended 1st-line treatment for a KIT-positive unresectable and/or metastatic gastrointestinal stromal tumour (GIST). However, some patients experience intolerance to imatinib and most patients will eventually experience disease progression while on imatinib treatment. Sunitinib is approved for treatment of a GIST after disease progression on, or intolerance to, imatinib therapy. Progression may occur early or later on, in treatment and is determined by factors including initial GIST genotype and mutational status. GISTs with KIT exon 11 mutations appear to be sensitive to standard dose imatinib, and patients with GISTs exhibiting KIT exon 9 mutations whose disease has progressed on imatinib 400 mg/day have been shown to respond to imatinib 800 mg/day, albeit with a higher incidence of adverse events. Sunitinib has shown clinical benefit in all major GIST mutational subtypes, particularly in patients with wild-type or KIT exon 9 genotype and against GISTs with secondary KIT exon 13 or 14 mutations. The choice between higher-dose imatinib and sunitinib after progression on standard dose imatinib is unclear, and apart from the GIST primary resistance genotype and mutational status, individual patient factors such as tumour characteristics, drug pharmacokinetics, and other clinical factors may affect response to treatment. Individualisation of therapy may help to maximise clinical benefit of therapy in these patients.

Introduction

Gastrointestinal stromal tumour (GIST) is the most common mesenchymal tumour of the gastrointestinal tract, with an estimated annual incidence of between 11.0 and 14.5 per million [1,2,3]. Imatinib mesylate, a receptor tyrosine kinase (RTK) inhibitor of stem-cell factor receptor (KIT) and platelet-derived growth factor receptor-α, is the 1st-line treatment of choice for KIT-positive unresectable and/or metastatic GIST [4,5]. However, 4% of patients with GIST are likely to be intolerant to imatinib [5], and patients can experience disease progression at various different stages of treatment [6,7,8]. Imatinib has recently been approved in Europe for the adjuvant treatment of adult patients who are at significant risk of relapse following resection of KIT-positive GIST [5].

Sunitinib malate is an oral, multitargeted RTK inhibitor of KIT, vascular endothelial growth factor receptor (VEGFR-1, −2, and −3), platelet-derived growth factor receptor -α and -β (PDGFR-α and -β), glial cell line-derived neurotrophic factor receptor (REarranged during Transfection; RET), receptor for macrophage colony-stimulating factor 1 (CSF-1R) and FMS-like tyrosine kinase 3 receptor (FLT3) [9,10,11,12,13,14]. Sunitinib has received multinational approval for the treatment of unresectable and/or metastatic GISTs after disease progression on, or intolerance to, imatinib therapy [15].

Some patients with GIST who experience disease progression on imatinib 400 mg/day may respond to an increase of the imatinib dose to 800 mg/day. Alternatively, patients may be switched to sunitinib as a 2nd-line therapy. The ideal choice of treatment after progression on imatinib 400 mg/day depends on factors such as primary resistance genotype, mutational status and individual patient factors like tolerance of therapy, drug pharmacokinetics and other clinical factors.

This review is intended to evaluate the evidence available to guide the pharmacological management of advanced (unresectable and/or metastatic) GIST in the 1st-line setting and after initial imatinib failure, and explore the factors that may influence the choice between higher-dose imatinib and sunitinib as 2nd-line treatments. This review focuses on the approved agents imatinib and sunitinib, but will also briefly consider pharmacological agents under investigation for the treatment of GIST. The potential influence of recently approved imatinib adjuvant therapy on the long-term management of patients with GIST will also be examined.

Imatinib Efficacy and Tolerability in the 1st-Line Setting

The efficacy and tolerability of imatinib 400 and 800 mg/day in the 1st-line setting has been demonstrated in phase II/III clinical trials.

Phase II Trials

In a phase II multicentre, randomised trial of imatinib at a dose of 400 or 600 mg/day in patients with an advanced GIST (n = 147) [6], the primary end-point was a response rate based on the Southwest Oncology Group criteria [16]. A total of 79 patients (53.7%) had a partial response, 41 (27.9%) had stable disease and no patients had a complete response. The trial was not powered to detect differences between the two dosage groups.

Long-term data from this trial [17] showed that at a median follow-up of up to 71 months, 1% of the patients (n = 2) achieved complete response, 67% (n = 98) had a partial response and 16% (n = 23) had stable disease. The objective response rate (ORR) was 68.1% (95% CI: 59.8–75.5). The median time to tumour progression (TTP) was 24 months overall, 20 months in the low-dose group and 26 months in the high-dose group. The median duration of response was 29 months and median overall survival was 57 months.

In a smaller phase II Japanese study of similar design (n = 74) [18], the overall response rate was 55.4%; 41% of the patients exhibited a partial response, 30% had stable disease and 1% had progressive disease. The percentage of patients with a partial response was higher in the 600-mg dose group than in the 400-mg dose group (60.9 vs. 46.4%); however, this was not statistically significant.

In a phase II study of imatinib 800 mg/day, the endpoints were anti-tumour response and TTP for patients with soft tissue sarcomas including patients with advanced and/or metastatic GIST (n = 51 for patients with soft tissue sarcomas, n = 27 for GIST patients) [19]. In GIST patients, 4% had a complete response, 67% had a partial response, 18% had stable disease, and 11% had disease progression, with 73% of patients free from progression at 1 year.

In this study, adverse events with the 800-mg/day dose were generally mild to moderate and there were no treatment discontinuations due to adverse events. The most frequently reported adverse events were anaemia (92%), periorbital oedema (84%), skin rash (69%), fatigue (76%), nausea (57%), granulocytopenia (47%), and diarrhoea (47%).

Phase III Trials

The efficacy of imatinib 800 mg/day has been investigated in two phase III clinical trials [8,20,21]. Both trials compared standard (400 mg/day) versus higher-dose imatinib (800 mg/day) in patients with advanced or metastatic GIST with or without prior chemotherapy [8,21]. Patients were randomly allocated imatinib 400 mg once or twice daily. Patients with disease progression on the once daily regimen were offered the option of crossover to the twice daily arm. The best ORRs for the two trials are shown in table 1.

Table 1.

Phase III trials of imatinib and sunitinib in GIST–best overall objective tumour response for the ITT population

Response, n	Imatinib phase III trials				Sunitinib phase III trial [45]
	European-Australasian trial [8]		Intergroup S0033 trial [21]		Sunitinib 50 mg/day, Schedule 4/2 (n = 207)	Placebo (n = 105)
	Imatinib 400 mg/day (n = 473)	Imatinib 800 mg/day (n = 473)	Imatinib 400 mg/day (n = 345)	Imatinib 800 mg/day (n = 349)
Complete response	24 (5)	28 (6)	17 (5)	12 (3)	−	−
Partial response	213 (45)	229 (48)	137 (40)	148 (42)	14 (7)a	0 (0)a
Stable disease	150 (32)	150 (32)	85 (25)	76 (22)	120 (58)	50 (48)
Progressive disease	61 (13)	42 (9)	42 (12)	37 (10)	38 (19)	39 (37)
Not assessable/assessment inadequate	25 (5)	24 (5)	34 (10)	52 (15)	−	−

Values in parentheses represent percentages.

^a95% CI: 3.7–11.1%; p = 0.006.

In the European-Australasian trial (n = 946) [8], there was a significant difference between the imatinib 800-mg/day and 400-mg/day groups in the proportion of patients who were progression free (50 vs. 44%, respectively; p = 0.026) after a median follow-up of 760 days.

The North American trial (n = 746) [21] found no significant difference in median progression-free survival (PFS) between patients receiving imatinib 800 mg/day and those receiving imatinib 400 mg/day (20 vs. 18 months, respectively) after a median follow-up of 4.5 years. The 2-year PFS rates were 46% (imatinib 800 mg/day) and 41% (imatinib 400 mg/day). Median overall survival was 51 and 55 months for imatinib 800 mg/day and 400 mg/day, respectively. Estimated overall survival rates at 2 years in the two trials were 72–74% in patients treated with imatinib 800 mg/day and 69–76% among patients receiving imatinib 400 mg/day. There was no advantage in terms of overall survival for imatinib 800 mg/day versus imatinib 400 mg/day.

In the European-Australasian trial [8], the tolerability of imatinib (assessed in terms of adverse events) differed in the two treatment arms, with the 800-mg/day dose resulting in significantly increased grade 3–4 adverse events compared with the 400-mg/day dose (table 2). The proportion of patients experiencing grade 3–5 adverse events was higher in the group treated with imatinib 800 mg/day as compared with those receiving the 400 mg/day dose in the North American trial (63 vs. 43%, respectively) [21].

Table 2.

Comparison between toxic effects in the imatinib 400 and 800 mg/day groups [8]

Adverse event, n	Imatinib 400 mg/day (n = 470)	Imatinib 800 mg/day (n = 472)	pa
Oedema	336 (71)	412 (87)	<0.0001
Anaemia	418 (89)	468 (98)	<0.0001
Rash	125 (27)	220 (47)	<0.0001
Nausea	229 (49)	286 (61)	<0.0001
Bleeding	51 (11)	105 (22)	<0.0001
Diarrhoea	226 (48)	268 (57)	0.0026
Dyspnoea	54 (11)	83 (18)	0.036
Pleuritic pain	240 (51)	160 (55)	0.053

Values in parentheses represent percentages.

^aAdjusted for repetitive testing (Hommel step-up procedure).

Mutations in GIST

Most GISTs express activating mutations of the RTKs KIT (approximately 75–85%) or PDGFR (approximately 5%) [22,23]. However, a small proportion of GISTs (10–15%) exhibit wild-type mutations, expressing neither the KIT nor PDGFR genotype [22,24].

Mutations in KIT most commonly involve the exon 11 (57–70%) and exon 9 (5–18%) domains and less commonly the exon 13 and 17 domains (0.06–1.6%) [22,24,25]. PDGFR mutations involve exons 12, 14, and 18 [22], with the most common mutation being the D842V mutation in exon 18 (62.6%) [26].

Imatinib Resistance and the Role of Mutational Status

It is widely documented that the initial mutational status of the KIT and PDGFR mutants will affect response to imatinib [24,27,28]. Phase II and III imatinib trials [21,24,27] in patients with advanced GIST reported higher partial response rates in patients with mutations in KIT exon 11 compared to patients with a mutation in KIT exon 9 or no detectable mutation. In addition, patients with exon 11 mutations had longer median overall survival than those with exon 9 or other KIT mutations/no detectable mutation (63 months vs. 44 and 26 months, respectively; p = 0.005) [17].

More than one third of PDGFR mutants [26] have exhibited a response to imatinib [24], particularly in exon 12 and 14, as well as some mutations in exon 18 [26]. These observations suggest that imatinib-sensitive PDGFR mutants may account for responses in patients with KIT-negative GISTs, and warrants further investigation.

Furthermore, almost all patients who experience early disease progression on imatinib treatment (400 mg/day) have mutations inKIT exon 9 or PDGFRA, or have wild-type mutations [24,27,28]. In particular, patients with GISTs expressing the PDGFR D842V (exon 18) mutation (the majority of PDGFR mutations) may show no clinical response to imatinib [24]. A similar lack of response to imatinib has been demonstrated with certain KIT mutants such as D816H and D816V (exon 17) [23]. However, patients with KIT exon 9 mutations have been shown to respond to higher doses (800 mg/day) of imatinib [7,27], but with a higher incidence of adverse events (table 2) [8]. Analysis of the phase III European-Australasian trial [8] found that in patients with tumours expressing a primary KIT exon 9 mutation, imatinib 800 mg/day improved median PFS compared with imatinib 400 mg/day (hazard ratio 0.392; 95% CI: 0.218–0.706; p = 0.0013) [27].

On the basis of these and additional emerging data showing that patients with KIT exon 9 mutations may achieve longer PFS on higher-dose imatinib, guidelines support the use of imatinib 800 mg/day as the standard treatment of choice in this subgroup [4,29].

Secondary (acquired) mutations in imatinib-treated patients are more common in KIT exons 13, 14, 17, or 18 and develop more often in tumours with primary exon 11 mutations than in those with exon 9 mutations [30]. Approximately 50–70% of the patients showing late progression will have secondary KIT mutations in exons 13, 14, 17, or 18 [31]. In tumours that exhibit primary imatinib resistance in KIT exon 9 or wild-type mutants, secondary mutations are rare.

Disease Progression in Patients Receiving 1st-Line Imatinib for GIST

Disease progression can occur at any time during imatinib treatment. Early progression in patients who have never shown any response to imatinib generally occurs within 3–6 months of imatinib initiation [31]. In clinical trials, 12–14% of patients showed evidence of disease progression on imatinib within 3 months of starting treatment [6,7].

While early progression during imatinib treatment is principally thought to be due to intrinsic mutational status, secondary KIT mutations may contribute to early progression in approximately 10% of patients [32]. Clinical factors may also be involved in the development of early progression. In a phase III European-Australasian study of imatinib 400 mg/day and 800 mg/day in patients with advanced GIST, the presence of lung and absence of liver metastases, low haemoglobin level, and high granulocyte count were independently predictive of early resistance [7].

Late progression is defined as progression occurring in patients who initially had a tumour response or a PFS interval of more than 3–6 months after starting imatinib treatment [31]. Late progression is thought to result from secondary or acquired resistance to imatinib, largely due to the development of secondary KIT mutations. However, other mechanisms related to pharmacokinetics may be involved in late progression, for example changes in systemic drug availability over time due to increased clearance [33] and/or increased binding of imatinib to α₁-acid glycoprotein [34]. The observed decrease in the area under the curve of imatinib may be due to increased expression of drug efflux pumps resulting in lower absorption or increased clearance of imatinib [35].

The effect of pharmacokinetics on treatment response is further illustrated by observations made by Demetri et al. [36], who noted interpatient variability in imatinib pharmacokinetic profiles and plasma drug exposure in GIST patients receiving imatinib, and that trough drug plasma levels correlated to treatment response.

Further, the effect of renal dysfunction on imatinib pharmacokinetics and toxicity has been investigated in a phase I trial of imatinib in patients with advanced solid tumours and varying renal function ranging from normal to severe dysfunction [37]. The study concluded that patients with mild or moderate renal dysfunction could tolerate imatinib doses up to 800 or 600 mg/day and did not require dose modification in spite of increased drug exposure. However, caution is required in patients with severe renal dysfunction. These observations suggest that drug-blood level monitoring may be necessary to ensure adequate exposure to treatment to maintain response and prevent disease progression.

Other clinical factors that may play a role in the development of late resistance during imatinib treatment include a high baseline granulocyte count, nongastric primary tumour, large tumour size and initial imatinib dose of 400 mg/day [7]. In addition, as patients may skip doses for a variety of reasons (adverse events, cost concerns, etc.), clinicians should bear this in mind when reviewing the patient's response to treatment.

Evaluating Response to Treatment

Response to treatment with RTKs in patients with GIST has been defined as the absence of progression at the time of the first formal evaluation of the disease [7], which is usually performed 2–3 months after starting treatment. However, the onset of response to treatment with RTKs is variable and some clinicians define the initial period of assessment of response or progression as up to 6 months after the start of therapy.

Accurately assessing response to treatment is essential in order to determine whether particular treatments should continue or be altered to potentially improve outcomes.

The current accepted standard for objective tumour response measurement is RECIST (Response Evaluation Criteria in Solid Tumors), which measures tumour size (table 3) [38]. However, imatinib treatment may not cause immediate tumour regression, but rather, may result in an initial inhibition of growth [7], tumour necrosis, and/or cystic or myxoid degeneration. Alternatively, the tumour may increase in size despite response according to clinical assessment or positron emission tomography (PET) [39]. Therefore, as RECIST measures only size, response may not be detected with this method.

Table 3.

Definition of best response according to RECIST criteria [38]

Best response	RECIST change in sum of longest diameter of all target lesions
CR	Disappearance; confirmed at 4 weeks
PR	30% decrease; confirmed at 4 weeks
SD	Neither PR nor PD criteria met
PD	20% increase; no CR, PR or SD documented before increased disease

CR = Complete response; PR = partial response; SD = stable disease; PD = progressive disease.

PET using [¹⁸F] fluorodeoxyglucose (FDG) has been shown to be sensitive in detecting early response to imatinib in patients with GIST and has shown its usefulness in predicting long-term response in these patients [40]. However, some lesions are not detectable by PET before the start of treatment. Recently, this technique has also been prospectively evaluated in sunitinib-treated patients with advanced GIST following failure of 1st-line imatinib treatment [41]. A single FDG-PET performed after 4 weeks of treatment appeared to predict long-term response to sunitinib treatment. PFS was found to significantly correlate to response by FDG-PET.

Currently, PET is not as widely available as computed tomography (CT) and is not the current standard of care.

The Choi criteria (a 10% reduction in unidimensional tumour size or a 15% decrease in tumour density on contrast-enhanced CT) have been proposed as an alternative, more accurate measure of response to GIST treatment [39]. Choi criteria measure tumour density in addition to tumour size (used in RECIST) and have been shown to correlate better with TTP and survival than response by RECIST (table 4) [39,42].

Table 4.

Modified CT response evaluation criteria (Choi criteria) [39]

Response	Definition
CR	Disappearance of all lesions No new lesions
PR	A decrease in size1 of ≥10% or a decrease in tumour density (HU) ≥15% on CT No new lesions No obvious progression of non-measurable disease
SD	Does not meet the criteria for CR, PR or PD No symptomatic deterioration attributed to tumour progression
PD	An increase in tumour size of ≥10% and does not meet criteria of PR by tumour density (HU) on CT New lesions New intra-tumoural nodules or increase in the size of the existing intra-tumoural nodules

CR = Complete response; PR = partial response; HU = Houns-field unit; SD = stable disease; PD = progression of disease.

¹The sum of longest diameters of target lesions as defined in RECIST [38].

Efficacy of Targeted Agents for 2nd-Line Treatment of GIST

Among patients with GIST who experienced progression on 1st-line imatinib 400 mg/day, clinical data suggests that 2nd-line treatment with either imatinib 800 mg/day or sunitinib may be considered for subsequent treatment, and those who progress on 1st-line imatinib 800 mg/day may be switched to sunitinib [4,29]. To date, no published trials have compared the efficacy and toxicity of sunitinib and high-dose imatinib in patients with a GIST who showed disease progression while receiving imatinib 400 mg/day. However, this is the focus of an ongoing phase III study, which is currently enrolling patients [43]. The choice of 2nd-line therapy may be influenced by whether imatinib failure results from primary or secondary mutations, and by the differing sensitivities of certain KIT mutants to imatinib (at different dosages) and to sunitinib.

Here we consider the relative efficacy of 2nd-line treatments for advanced GIST, including a discussion of specific subgroups.

Imatinib 800 mg/Day Efficacy as a 2nd-Line Treatment

The efficacy of imatinib 800 mg/day after disease progression has been investigated in the two phase III clinical trials discussed above [8,20,21]. Patients with disease progression on the once daily (400 mg/day) imatinib regimen were offered the option of crossover to the twice daily (800 mg/day) arm.

In the European-Australasian trial, patients who crossed over from imatinib 400 mg/day to imatinib 800 mg/day obtained further benefit in terms of PFS. ORR was 29%, PFS was 81 days and 18.1% were progression free after 1 year [20]. In the North American trial, in the patients who crossed over to 800 mg/day, ORR was 3% (95% CI, 1–7), 28% experienced stable disease (95% CI, 20–37) and median PFS was 5 months (95% CI, 2–10) [21].

Sunitinib Efficacy and Tolerability

Sunitinib is administered at 50 mg/day in 6-week cycles comprising 4 weeks on treatment followed by 2 weeks off treatment (Schedule 4/2). Two clinical studies have investigated the efficacy of sunitinib following the failure of initial imatinib therapy [44,45].

In a phase III double-blind trial, patients with an imatinib-resistant/-intolerant GIST (n = 312) were randomised 2:1 to receive sunitinib 50 mg/day on Schedule 4/2 or placebo [45]. Median TTP was 27.3 weeks in patients receiving sunitinib and 6.4 weeks in patients receiving placebo (p < 0.0001). The median PFS was 24.1 versus 6.0 weeks for sunitinib and placebo, respectively (p < 0.0001). The best overall objective tumour responses in this study are shown in table 1.

The trial was unblinded early after a planned interim analysis and all patients receiving placebo were allowed to crossover to receive sunitinib. Of the patients who crossed over to sunitinib from the placebo group, 10% (6/59) had confirmed partial responses and 7% (4/59) had stable disease for at least 26 weeks after crossover [45].

Analysis of long-term survival data for this trial [41] revealed no statistically significant differences in overall survival between patients receiving sunitinib (73.9 weeks) and those receiving placebo (64.9 weeks, p = 0.161). This was due to bias introduced by the inclusion of patients who had crossed over to active treatment from the placebo group. To overcome this bias, overall survival was analysed using a novel statistical method, the rank-preserving structural failure time method, which found that overall survival in the sunitinib group (73.9 weeks) was twice that of the placebo group (35.7 weeks, p < 0.001) [46].

In a phase II, open-label trial of patients (n = 97) with advanced GIST who experienced progression on imatinib [44], the estimated median TTP was 34 weeks (95% CI, 22.0–46.0 weeks). At the time of data analysis, 7% (n = 7) of the patients showed evidence of a partial response and 29% (n = 28) had stable disease for ≥6 months.

Sunitinib is also being assessed in an ongoing treatment-use study which was initiated to provide access to sunitinib for patients with imatinib-resistant/intolerant advanced GIST not otherwise eligible to participate in clinical trials. In this study (n = 1,126), a median TTP of 41.0 weeks was reported and the median overall survival was 75.0 weeks [47].

In an ongoing phase II trial [48], continuous daily dosing (CDD) of sunitinib (37.5 mg/day) is being evaluated in patients (n = 61) with imatinib-resistant/intolerant GIST. At the last analysis, the estimated median PFS was 34.0 weeks (95% CI, 25–59), and 12% of the patients (n = 7) had a partial response while 42% (n = 25) had stable disease ≥24 weeks. Median overall survival was 107 weeks (95% CI: 72–not yet calculable). These results suggest that CDD is an effective alternative dosing schedule for sunitinib.

The dosing schedule used (i.e. intermittent vs. CDD) could potentially affect patient drug exposure and, consequently, clinical outcomes in GIST patients treated with sunitinib. Some patients receiving sunitinib may experience clinical disease progression during the 2-week off-treatment period of the standard intermittent Schedule 4/2 resulting in an increase or recurrence of disease-related symptoms. The CDD schedule may help to prevent this occurrence by maintaining continuous drug exposure. In addition, for patients whose adverse events are not alleviated by dose reductions on Schedule 4/2, restarting sunitinib treatment on the CDD schedule may prove to be a more effective alternative strategy, helping to maintain patients on treatment for longer and delaying disease progression.

There is some evidence to suggest that tailoring imatinib dosing schedules in tandem with blood-drug level monitoring may be of benefit in some individuals [49]. This may be true of other targeted agents, such as sunitinib, that have similar metabolic pathways, but this remains to be proven definitively by further research. However, it is worth noting that recommended target plasma levels for optimal sunitinib RTK inhibition have been defined based on predictions from clinical studies [15].

Blood-drug level monitoring of targeted agents is a tool used mainly in the clinical trial setting. Research is ongoing to develop blood plasma assays for sunitinib (and other targeted agents) that can be used in the clinical setting [50,51]. While currently available in only a few centres, these assays may become widely used in oncology clinical practice in the future.

Sunitinib was generally well tolerated in both the phase III and treatment-use studies. The most common treatment-related grade 3–4 adverse events reported in the phase III trial were fatigue (10%), hypertension (7%), hand-foot syndrome (6%), diarrhoea (5%) and asthenia (5%). There were five grade 5 (2%) adverse events reported: hepatic failure, left ventricular failure, cardiac arrest, cerebral ischaemia and multi-organ failure. Similarly, common grade 3–4 treatment-related adverse events occurring in the treatment-use study included fatigue (8%), hand-foot syndrome (8%), hypertension (5%) and diarrhoea (5%). Twenty-three (2%) grade 5 treatment-related adverse events were reported including one case of diarrhoea and one of nausea.

Effect of a Prior Dose of Imatinib

In patients with imatinib-resistant GIST, sunitinib appears to be effective regardless of prior imatinib dose [45]. However, patients receiving a lower prior imatinib dose appear to achieve the most benefit from sunitinib. Subgroup analysis from the treatment-use study in a broad patient population showed that patients receiving prior imatinib treatment at a lower dose (≤400 mg/day) had a median overall survival of 90 weeks (95% CI: 73–106; n = 351) compared with a median overall survival of 70 weeks (95% CI: 63–76; n = 763) for those who previously received a higher dose of imatinib (>400 mg/day) [47]. The reason for this difference in response may be due to the emergence of secondary mutations on prolonged exposure to imatinib. Patients on prior high-dose imatinib are likely to have been exposed to the drug for longer as most clinicians will start patients on standard dose imatinib 400 mg/day before escalating to 800 mg/day on evidence of disease progression.

Sunitinib Response and GIST Genotype

In a phase I/II trial of sunitinib in patients with imatinib-resistant GIST [30], all major GIST mutational subtypes showed clinical benefit with sunitinib (KIT exon 11: 34%; KIT exon 9: 58%; PDGFRA/wild-type KIT: 56%). The partial response rate for GISTs with primary KIT exon 9 mutations compared with those exhibiting KIT exon 11 mutations was 37 versus 5%, respectively (p = 0.002). Additionally, median PFS was significantly longer in patients with a primary KIT exon 9 genotype (19.4 months, p = 0.0005) or wild-type KIT genotype (19.0 months, p = 0.0356) compared with those with an exon 11 genotype (5.1 months). Similarly, median overall survival was significantly longer for patients with an exon 9 mutation (26.9 months, p = 0.012) or wild-type KIT mutation (30.5 months, p = 0.0132) than for those with an exon 11 mutation (12.3 months) [30].

In vitro, secondary kinase mutations of KIT exon 13 and 14 were sensitive to sunitinib, while secondary KIT exon 17 and 18 mutations were resistant to sunitinib. These findings were consistent with the clinical results: clinical benefit was observed in 61% of the patients with secondary KIT mutations in exon 13 or 14 compared with 15% of patients with secondary KIT exon 17 or 18 mutations (p = 0.011) [30]. In addition, certain imatinib-resistant exon 17 mutants have been shown to be sunitinib-resistant as well. These include D816H and D816V [23].

Disease Progression on Sunitinib

As with imatinib, development of resistant genotypes will eventually lead to disease progression in patients treated with sunitinib. Pharmacokinetic factors may play a role in disease progression. A correlation between exposure and clinical response has been noted in patients with a GIST treated with sunitinib [52,53]. This correlation has also been observed in patients with advanced renal cell carcinoma treated with sunitinib [53].

While no data are currently available to suggest that sunitinib's systemic availability decreases over time, and no studies have been conducted in patients with renal dysfunction receiving sunitinib, it is possible that the mechanisms demonstrated with imatinib will affect other RTK inhibitors. Further research in this area is warranted.

Therapy Choices after Failure of 1st- and 2nd-Line Targeted Agents

While prolonged response to treatment is achievable with the targeted agents imatinib and sunitinib as discussed above, the majority of patients with advanced GIST will develop resistance to either treatment (or both treatments) at some stage. For patients whose disease progresses after 2nd-line treatment, there is currently no approved 3rd-line agent for the management of advanced GIST. The only available options are palliative care or enrolment in a clinical trial.

Several agents are under investigation for the management of advanced GIST, including the tyrosine kinase inhibitors sorafenib, nilotinib and dasatinib (table 5) [54,55,56,57,58].

Table 5.

Targeted agents under investigation for the management of advanced GIST [54, 55, 56, 57, 58]

Agent	Targets
Sorafenib	KIT, PDGFRA/B, VEGFR2–3, Raf, FLT3, RET
Nilotinib (AMN107)	KIT, PDGFRA/B, BCR-Abl
Dasatinib (BMS-354825)	KIT, PDGFRB, BCR-Abl, SRC
Masatinib (AB1010)	KIT, PDGFRA/B, FGFR3
Motesanib (AMG 706)	KIT, PDGFRA/B, VEGFR1–3, FLT3 or RET
Vatalanib (PTK787/ZK222584)	KIT, PDGFRA/B, VEGFR1–3
Cediranib (AZD2171)	KIT, PDGFRA/B, VEGFR1–3, FLT3, RET
Panobinostat	Histone deacetylase
PKC412	KIT, PDGFRA/B, VEGFR2, protein kinase C
Everolimus	mTOR
IPI-504	HSP-90

PDGFRA/B = Platelet-derived growth factor receptor α/β; FLT3 = FMS-like tyrosine kinase 3; RET = glial cell-line derived neurotrophic factor receptor (REarranged during Transfection); BCR-ABL = breakpoint cluster region-Abelson gene; SRC = proto-oncogenic tyrosine kinases; FGFR3 = fibroblast growth factor receptor 3; HSP-90 = heat shock protein 90; mTOR = mammalian target of rapamycin.

Potential Impact of Adjuvant Therapy in Primary GIST on Advanced Disease

Targeted agents for the prevention of recurrence in patients with resectable GIST are under investigation for potential use following surgery. Recently, imatinib received approval in the USA and Europe for the adjuvant treatment of adult patients who are at significant risk of relapse following resection of KIT-positive GIST.

The approvals were based on results from the phase III, multicentre, randomised placebo-controlled trial [59], in which patients with KIT-positive GIST (n = 713) received either imatinib 400 mg (n = 359) or placebo (n = 354) daily for 1 year after the surgical resection of a primary GIST. Recurrence-free survival improved significantly with imatinib compared to placebo at 1 year; 98% (95% CI: 96–100) versus 83% (78–88), respectively [hazard ratio 0.35 (0.22–0.53); p < 0.0001]. In patients who have sensitive GIST mutants, this approach may delay the development of unresectable and/or metastatic disease and prolong patients' lives, but this remains to be seen.

However, adjuvant imatinib may not be suitable for patients with mutants that are not responsive to imatinib. Hou et al. [60] showed that disease-free survival after complete surgical resection of a primary GIST differed in patients with different GIST genotypes. There were also differences within the same genotype, e.g. 5-year disease-free survival in patients with specific subsets of KIT exon 11 mutations differed. In addition, Hou et al. noted that patients classified as low risk by the National Institute of Health criteria developed tumour recurrence. These observations raise questions about whether all patients with KIT-positive GIST will achieve similar benefit from adjuvant therapy and also whether subclassification of the KIT and PDGFR genotypes could be used to identify those that would gain most.

Another consideration with adjuvant therapy is that early and prolonged exposure to imatinib may encourage the development of resistant mutants and therefore reduce the choice of treatments available for advanced disease. Further investigation into the impact of adjuvant therapy on resistance to imatinib is warranted.

Discussion

Imatinib and sunitinib have led to dramatic improvements in the pharmacological management of advanced GIST. As our experience with these agents increases, their optimal use in clinical practice can be better elucidated.

The mutational status of GISTs pre- and post-imatinib treatment has been shown to influence clinical outcomes, affecting whether progression occurs early or later in treatment. Other factors may also affect the efficacy of imatinib in individual patients with GIST and should be considered. These include the metastatic site, baseline haemoglobin, baseline granulocyte count, site of the primary tumour, tumour size, initial imatinib dose and patient adherence to treatment.

Pharmacokinetic-related factors such as systemic availability and renal function may also affect efficacy and outcomes in patients treated with imatinib or sunitinib. In addition, tailoring the dosing schedule to improve drug exposure in individual patients treated with sunitinib may improve response to treatment. Data for imatinib suggests that patients with mild to moderate renal impairment may tolerate higher doses, while caution is required in those with severe renal failure. Research is required to determine the potential influence of renal dysfunction on sunitinib pharmacokinetics.

Imatinib with a starting dose of 400 mg/day is the standard therapy for 1st-line use, after which treatment options may vary according to individual patient profiles. Imatinib 800 mg/day has been shown to improve PFS in particular subgroups of patients after disease progression on 400 mg/day. Imatinib 800 mg/day is considered the standard of care for patients with GISTs exhibiting KIT exon 9 mutations, but the higher incidence of adverse events relative to 400 mg/day should be considered [4]. In contrast, GISTs with KIT exon 11 mutations appear to be sensitive to standard dose imatinib in most cases.

Sunitinib has demonstrated efficacy in patients who have progressed on standard dose imatinib (400 mg/day) and appears to be particularly effective for treatment of GISTs with a wild-type genotype or KIT exon 9 mutations and in patients with secondary KIT mutations in exon 13 or 14.

Evidence of the differing efficacy of sunitinib and higher-dose imatinib in patients with GIST opens up the possibility of tailoring 2nd-line therapy to obtain maximum benefit for individual patients. Ongoing studies will assess directly the efficacy and tolerability of sunitinib compared with imatinib 800 mg/day and will provide valuable information to assist in determining the choice of 2nd-line therapy. Individualisation of therapy by considering tumour genotype as well as other individual clinical factors may help to maximise clinical benefit of RTK inhibitors in patients with GIST. Mutational analysis is recommended in current treatment guidelines for the management of GIST. However, this is not routinely available and is currently possible only in specialist centres. With the availability of oral therapies allowing patients to be treated away from specialist centres, services will need to be developed to ensure patients receive optimum/tailored treatment.

Conflicts of Interest

Dr. Reichardt has received honoraria from Novartis and Pfizer, is a member of Advisory Boards for Novartis and Pfizer, and has received research grants from Novartis.