Abstract

Background

Imatinib has been found to be effective in the treatment of patients with gastrointestinal stromal tumors (GIST). We sought to evaluate the clinical outcome of imatinib interruption in GIST patients who had achieved stable disease (SD) or showed better response to imatinib therapy.

Methods

From July 2001 to December 2004, we prospectively collected clinical data from 62 consecutive patients with advanced GIST, of whom 58 (93.5%) achieved SD or better response to imatinib therapy and were included in this study. Imatinib therapy was interrupted in 14 of the 58 patients (interruption group, INT), after a median time of 11.9 months. Progression-free survival (PFS) after imatinib interruption was calculated and imatinib-refractory PFS and overall survival (OS) were compared between the INT group and the 44 patients who continued imatinib treatment (continuation group, CONT).

Results

After a median follow-up of 17.9 months following imatinib interruption, nine patients (64%) had progressive disease (PD) with a median PFS from the date of imatinib interruption of 10.0 months. Median PFS dated from the time of imatinib initiation in the INT group was 21.8 months (95% CI, 17.3–26.3 months), but was not reached in the CONT group (P=0.029). Following imatinib reintroduction in the INT group, 88% of patients achieved disease control. There were no statistically significant differences in imatinib-refractory PFS (P=0.405) and OS (P=0.498) between the groups.

Conclusion

In GIST patients controlled with imatinib, treatment might be interrupted, at least temporarily, when clinically warranted.

INTRODUCTION

Imatinib has revolutionized the care of patients with gastrointestinal stromal tumor (GIST) and has become standard therapy for patients with advanced GIST. In phase II and III clinical trials in metastatic GIST, up to 85% of patients were shown to benefit from treatment with imatinib mesylate, with 4–5% of patients achieving complete response (CR), 45–67% achieving partial response (PR), and 18–32% achieving stable disease (SD), after follow-up of 9–24 months (1–4). In addition, treatment with imatinib mesylate increased overall survival (OS) compared to historical controls (1–5). In these studies, patients were treated with imatinib until disease progression or the development of intolerable toxicities (3,4).

In clinical settings, a significant proportion of patients choose to temporarily discontinue therapy, with or without supervision by their physicians, because of recurrent toxicities, economic constraints, concomitant co-morbidities, or because of the desire to have a break from therapy. Little is known, however, about the clinical outcome of imatinib interruption in patients with GIST who are free from disease progression. A recent randomized trial in France reported that imatinib interruption after 1 year was associated with a high risk of disease progression, indicating that imatinib should be given until disease progression or intolerance (6). The primary end-point of this trial, however, was progression-free survival (PFS) and it was halted earlier than intended after meeting step 1 early stopping rule requirements, six or more progressive diseases (PD) in the first 14 patients who interrupted imatinib at 3 months. Interestingly, the majority of patients who resumed imatinib therapy after documentation of PD had a favorable response to retreatment with imatinib, indicating that these patients had imatinib-sensitive PD, not imatinib-refractory PD. It is therefore not clear if such patients require continuous treatment with imatinib, or whether imatinib could be safely discontinued, at least temporarily, without the impairments on the incidence of imatinib-refractory PD and OS.

To add more information on this issue, we have retrospectively investigated the clinical outcome of imatinib interruption in GIST patients with an SD or better response to imatinib treatment and we have compared these outcomes with those of GIST patients who received continuous imatinib therapy.

PATIENTS AND METHODS

Patients and Treatment

From July 2001 to December 2004, we prospectively gathered clinical data from 62 consecutive patients with metastatic or unresectable GIST diagnosed and treated at our institution; 12 of these patients were also enrolled in the recently closed Korean multi-center phase II clinical trial of imatinib mesylate. All patients had histological confirmation of GIST, with positive KIT expression as shown by CD117 immunohistochemical staining (A4502, DAKO), and at least one measurable lesion. The protocol of this study was approved by the institutional review board of the Asan Medical Center.

The median age of the patients was 57 years (range, 33–73 years), with a male to female ratio of 1.6:1. The primary tumor sites were the stomach (44%) and small bowel (47%). At enrollment, 12 patients (19%) had metastatic disease, 45 (73%) had recurrent disease and 5 (8%) had unresectable locally advanced disease. Patients were supposed to receive 400 mg imatinib mesylate p.o. per day until disease progression, unacceptable toxicity or patient refusal. Initial responses to imatinib mesylate treatment were CR in 4 patients (6.5%), PR in 35 (56.5%), SD in 19 (30.6%), and PD in 4 (6.5%) patients. Among the 58 patients who achieved SD or better response to imatinib, 14 interrupted imatinib treatment for various reasons. These 14 patients were grouped into the INT (interruption) group, whereas the remaining 44 patients, who continued imatinib treatment after achieving SD or better response, were grouped into the CONT (continuation) group. Patients who interrupted imatinib briefly because of transient toxicities or reversible concomitant illnesses but restarted within 1 month without tumor reassessment were not defined as having interrupted imatinib treatment and were categorized into the CONT group. Imatinib at a dose of 400 mg/day was reintroduced for patients in the INT group who had PD during the period of imatinib interruption. In cases of imatinib-refractory PD, the dose of imatinib mesylate was escalated to 600 or 800 mg/day.

Assessment of Response and Progression

The objective tumor response was measured using contrast-enhanced spiral computed tomography (CT) scans, using one of three commercially available helical CT scanners (Somatom Plus-S, Siemens, Erlangen, Germany; Somatom Plus 4, Siemens; Hispeed, GE Medical Systems, Milwaukee, WI, USA). Target lesions were prospectively evaluated 1–2 months after the start of treatment and every 2–3 months thereafter. CT scans were performed earlier if tumor progression was suspected. Hardcopy archived CT images were evaluated by three non-blinded oncologists (J-LL and M-HR separately, but in consensus with Y-KK) for confirmation of disease progression according to RECIST (7). Consistent with Southwest Oncology Group (SWOG) criteria, an increase in the sum of the longest diameters of the target lesions was not regarded as disease progression if accompanied by definite cystic change in the tumor suggesting necrosis (8). In addition, the development of new small cystic lesions in the liver during earlier phases of imatinib treatment was not regarded as disease progression (9). Patients who interrupted imatinib treatment continued to have tumor assessment every 2–3 months.

Statistical Analysis

The difference in baseline characteristics was assessed using the chi-squared test or Fisher's exact test for categorical variables or the Mann–Whitney U test for continuous variables. Survival distributions were estimated using the Kaplan–Meier method (10), and comparisons between groups were made using two-tailed log-rank tests (11). PFS in the INT group was measured from the date of imatinib interruption to the date of PD or death. PFS was also measured from the date of imatinib treatment to the date of first documentation of PD. Imatinib-refractory PD was defined as any PD in the CONT group, PD that developed in the INT group during imatinib interruption and could not be controlled with imatinib retreatment, or secondary PD that developed after a period of control with imatinib retreatment. Imatinib-refractory PFS was defined from the date of imatinib treatment to the date of imatinib-refractory PD or death. OS was calculated from the date of imatinib treatment to the date of death from any cause. P-values <0.05 were defined as statistically significant.

RESULTS

Patient Characteristics and Clinical Course Following Imatinib Interruption

The INT group consisted of 14 patients with a median age of 51 years (range, 33–73). Table 1 shows the clinical characteristics and disease course following imatinib interruption in these patients. The median time from imatinib initiation to interruption was 11.9 months (range, 3.3–27.0 months). In seven patients, imatinib treatment was interrupted by the physician because of the absence of residual disease, with one of these patients also having recurrent grade 3 thrombocytopenia. Four patients interrupted treatment due to economic constraints, one because of nausea, and two for unknown reasons.

Table 1.

Summary of the clinical features of 14 patients who interrupted imatinib mesylate treatment

Patient Age Sex Primary site Disease status Sites of active disease c-kit mutation Initial response Time to imatinib interruption Cause of imatinib interruption PFS after imatinib interruption Response to imatinib reintroduction Imatinib-refractory PFS Overall survival 
67 Exon 11, 557del558 CR 10.7 Dr 6.0 SD 21.1 26.0+ 
46 Pe Exon 9, 502dup503 CR 12.2 Dr 8.9+  21.1+ 21.1+ 
73 Pe Exon 11, 553del557 CR 11.7 Dr 10.1 PR 31.0+ 31.0+ 
63 Pe Exon 11, 556del560 CR 11.8 Dr 6.2+  18.0+ 18.0+ 
60 SB H, Pe Exon 11, 551del556 PR 3.3 1.8 NE 13.6 13.6 
46 Pe, adrenal NA PR 20.3 4.9 SD 41.3+ 41.3+ 
48 H, Pe Exon 11, 556del576 PR 13.8 10.9 PR 41.0+ 41.0+ 
47 Pe Exon 11, 557del558 PR 14.2 7.4 PR 39.8+ 39.8+ 
54 SB H, Pe Exon 11, 558ins PR 7.6 Pt 2.4 PR 41.6+ 41.6+ 
10 33 SB Exon 11, 552del557 PR 12.0 Pt 19.9 SD 40.9+ 40.9+ 
11 65 LA  Exon 11, 561del577 PR* 27.0 Dr 7.1+  34.2+ 34.2+ 
12 47 LA  NA PR* 11.1 Dr 3.8+  14.9+ 14.9+ 
13 59 SB Pe Exon 11, 576del SD* 13.0 Dr 15.6+  29.6+ 29.6+ 
14 41 Exon 11, 582del SD 6.9 Pt 12.9 PR 33.9+ 33.9+ 
Patient Age Sex Primary site Disease status Sites of active disease c-kit mutation Initial response Time to imatinib interruption Cause of imatinib interruption PFS after imatinib interruption Response to imatinib reintroduction Imatinib-refractory PFS Overall survival 
67 Exon 11, 557del558 CR 10.7 Dr 6.0 SD 21.1 26.0+ 
46 Pe Exon 9, 502dup503 CR 12.2 Dr 8.9+  21.1+ 21.1+ 
73 Pe Exon 11, 553del557 CR 11.7 Dr 10.1 PR 31.0+ 31.0+ 
63 Pe Exon 11, 556del560 CR 11.8 Dr 6.2+  18.0+ 18.0+ 
60 SB H, Pe Exon 11, 551del556 PR 3.3 1.8 NE 13.6 13.6 
46 Pe, adrenal NA PR 20.3 4.9 SD 41.3+ 41.3+ 
48 H, Pe Exon 11, 556del576 PR 13.8 10.9 PR 41.0+ 41.0+ 
47 Pe Exon 11, 557del558 PR 14.2 7.4 PR 39.8+ 39.8+ 
54 SB H, Pe Exon 11, 558ins PR 7.6 Pt 2.4 PR 41.6+ 41.6+ 
10 33 SB Exon 11, 552del557 PR 12.0 Pt 19.9 SD 40.9+ 40.9+ 
11 65 LA  Exon 11, 561del577 PR* 27.0 Dr 7.1+  34.2+ 34.2+ 
12 47 LA  NA PR* 11.1 Dr 3.8+  14.9+ 14.9+ 
13 59 SB Pe Exon 11, 576del SD* 13.0 Dr 15.6+  29.6+ 29.6+ 
14 41 Exon 11, 582del SD 6.9 Pt 12.9 PR 33.9+ 33.9+ 

S, stomach; SB, small bowel; M, metastatic; LA, locally advanced and unresectable; L, lung; Pe, peritoneum; H, hepatic; NA, not available; CR, complete response; PR, partial response; SD, stable disease; Dr, treating physician's clinical decision; E, economic constraint; Pt, patient will; NE, not evaluable; PFS, progression-free survival.

*Surgery was undertaken to eradicate residual or non-responding disease sites.

Baseline demographics were balanced between the INT and CONT groups (Table 2), except that no patients in the CONT group had an initial response of CR to imatinib treatment. After a median of 17.9 months following imatinib interruption, nine of 14 patients (64%) in the INT group had PD. Median PFS since the date of imatinib interruption was 10.0 months (Fig. 1, 95% CI, 5.6–14.5 months). The pattern of progression was general progression according to the previous published criteria in eight patients; only one patient (No. 7) had focal progression in peritoneal lesion while liver lesions were stable without progression (9). The five patients in the INT group who did not have PD after imatinib interruption showed no residual disease at the time of interruption; two of these patients had clinical CR and three with PR or SD underwent surgery for their residual or non-responding lesions.

Figure 1.

Progression-free survival following imatinib interruption (n=14).

Figure 1.

Progression-free survival following imatinib interruption (n=14).

Table 2.

Patient characteristics

 INT group (n=14) CONT group (n=44) P value 
Age, years (median, range) 51.0 (33–73) 57.5 (33–73) 0.429 
Gender    
 Male 6 (43%) 29 (66%) 0.125 
 Female 8 (57%) 15 (34%)  
Primary site   0.162 
 Stomach 10 (71%) 17 (39%)  
 Small bowel 4 (29%) 22 (50%)  
 Rectum 3 (7%)  
 Others 2 (5%)  
Distant metastasis 12 (86%) 39 (89%) 0.770 
Prior surgery 11 (79%) 31 (71%) 0.554 
Initial response to imatinib   0.001 
 CR 4 (29%)  
 PR 8 (57%) 27 (61%)  
 SD 2 (14%) 17 (39%)  
Surgery on residual or non-responding lesion 3 (21%) 11 (26%) 1.0 
c-kit mutation    
 Exon 11 11* (92%) 31(74%) 0.261 
 Exon 9 1* (8%) 4 (10%) 1.0 
 INT group (n=14) CONT group (n=44) P value 
Age, years (median, range) 51.0 (33–73) 57.5 (33–73) 0.429 
Gender    
 Male 6 (43%) 29 (66%) 0.125 
 Female 8 (57%) 15 (34%)  
Primary site   0.162 
 Stomach 10 (71%) 17 (39%)  
 Small bowel 4 (29%) 22 (50%)  
 Rectum 3 (7%)  
 Others 2 (5%)  
Distant metastasis 12 (86%) 39 (89%) 0.770 
Prior surgery 11 (79%) 31 (71%) 0.554 
Initial response to imatinib   0.001 
 CR 4 (29%)  
 PR 8 (57%) 27 (61%)  
 SD 2 (14%) 17 (39%)  
Surgery on residual or non-responding lesion 3 (21%) 11 (26%) 1.0 
c-kit mutation    
 Exon 11 11* (92%) 31(74%) 0.261 
 Exon 9 1* (8%) 4 (10%) 1.0 

CONT, continuation; INT, interruption; CR, complete response; PR, partial response; SD, stable disease.

*Available in 12 patients.

†Available in 42 patients.

The median time after stopping imatinib to retreatment was 10.1 months. Table 3 shows the response to retreatment. Prior to imatinib reintroduction, nine of the patients in the INT group had PD, but it was not associated with rapid deterioration of performance status; with the exception of one patient who had an ECOG performance status of two, all the remaining patients with PD after imatinib interruption had an ECOG performance status of one or better at the time of imatinib retreatment. Following reintroduction of imatinib, eight of the nine patients (88%) achieved disease control, with a 1-year PFS rate of 78%, measured from the date of imatinib reintroduction. One patient (No. 5) in the INT group was dead of disease. He discontinued imatinib after achieving PR observed 3 months after imatinib treatment because of economic constraints and then his disease progressed 1.8 months after discontinuation. Although the treating physician strongly recommended the patient to take imatinib, he did not adhere to medication and died of disease without further tumor assessment. Another patient (No. 1) in the INT group had PD, in the form of a new solid lesion in the lung (azygoesophageal recess), during the period of imatinib interruption. Although he achieved SD to imatinib reintroduction, his disease progressed 6 months after imatinib rechallenge and he underwent metastatectomy. At the time of analysis, he had no evidence of disease and took maintenance imatinib.

Table 3.

Response to imatinib: initial and reintroduction

Initial response No PD PD Response to imatinib reintroduction 
PR SD NE 
CR (n=4) 
PR (n=8) 1* 
SD (n=2) 
Initial response No PD PD Response to imatinib reintroduction 
PR SD NE 
CR (n=4) 
PR (n=8) 1* 
SD (n=2) 

CR, complete response; PR, partial response; SD, stable disease; PD, progressive disease during the period of imatinib interruption; NE, not evaluable.

*Poor adherence to regular imatinib medication and death without evaluation of response.

PFS, Imatinib-refractory PFS and OS

After a median follow-up of 25.6 months (range, 8.8–42.6 months), median PFS from the date of initial imatinib treatment was 21.8 months (95% CI, 17.3–26.3 months) for the INT group, but was not reached in the CONT group [Fig. 2; OR=0.37 (favoring continuous imatinib therapy), 95% CI, 0.15–0.94, P=0.029]. However, there was no significant difference in imatinib-refractory PFS between the groups (Fig. 3, OR=1.90, 95% CI, 0.41–8.83, P=0.404). Median OS was not reached in both groups and the two-year survival rates were 82.0% for the CONT group and 92.3% for the INT group (Fig. 4, OR=2.07, 95% CI, 0.24–17.8, P=0.498).

Figure 2.

Progression-free survival from start of imatinib treatment to disease progression. [P=0.0285; solid line, CONT (continuation) group; dotted line, INT (interruption) group].

Figure 2.

Progression-free survival from start of imatinib treatment to disease progression. [P=0.0285; solid line, CONT (continuation) group; dotted line, INT (interruption) group].

Figure 3.

Imatinib-refractory progression-free survival from the start of imatinib treatment to disease progression, excluding the first occurrence of disease progression during the period of imatinib interruption and response to imatinib retreatment [P=0.4048; solid line, CONT (continuation) group; dotted line, INT (interruption) group].

Figure 3.

Imatinib-refractory progression-free survival from the start of imatinib treatment to disease progression, excluding the first occurrence of disease progression during the period of imatinib interruption and response to imatinib retreatment [P=0.4048; solid line, CONT (continuation) group; dotted line, INT (interruption) group].

Figure 4.

Overall survival [P=0.498; solid line, CONT (continuation) group; dotted line, INT (interruption) group].

Figure 4.

Overall survival [P=0.498; solid line, CONT (continuation) group; dotted line, INT (interruption) group].

When we compared the time-related outcome between the groups, after excluding patients who achieved CR in the INT group to remove the potential effects of initial response on imatinib-refractory PFS and OS, we observed no significant between-group differences in imatinib-refractory PFS (OR=2.7, 95% CI, 0.34–21.34, P=0.328) and OS (OR=1.57, 95% CI, 0.18–13.55, P=0.679).

DISCUSSION

Interruption of imatinib treatment, whether because of the patient's wish or the treating physician's judgment, is not infrequent in clinical practice. Patients with GIST who achieve SD or better response without further decrease in tumor size despite continued treatment may ask their treating physicians what would happen if they were to discontinue taking imatinib or whether they must continue to take imatinib throughout their lives. This is especially true when patients experience adverse events, even if mild or moderate, when long-term side effects are not known, or when patients want to have children and are taking this potentially teratogenic medication. Although physicians confronted with this situation do not encourage treatment interruption, they have no firm, convincing evidence to discourage it. In addition, continuous imatinib treatment may contribute to the development or selection of drug-resistant clones (12). Thus, more information on the impact of imatinib interruption on the clinical course of GIST is needed.

Our analysis has shown that most interruptions of imatinib treatment were due to physicians' decisions, economic constraints or patients' wishes. Imatinib-associated toxicities did not have a significant impact on this study population, in that only two patients experienced drug-related toxicities, nausea and thrombocytopenia. More importantly, our results have shown that, in patients with metastatic GIST controlled with imatinib, imatinib interruption is associated with a high risk of disease progression within 1 year. Sixty-four percent of patients in the INT group had PD with a median PFS of 10 months, but all could resume treatment because disease progression was not associated with rapid deterioration in performance status. Moreover, the patients who restarted therapy responded favorably, suggesting that acquired resistance to imatinib did not develop during the period of treatment interruption. Furthermore, the survival of patients with imatinib interruption was similar to that of patients who received continuous imatinib treatment. Although this may reflect a more favorable overall prognosis in the INT group than in the CONT group, in that all the patients who achieved CR in response to initial imatinib treatment were in the INT group, survival outcome was not altered when these patients were excluded.

A recent randomized trial on 58 French patients with GIST who were free from progression after 1 year of treatment with imatinib found that imatinib interruption was associated with a high risk of relapse (6). These findings suggested that, to maintain a continued response, patients should be treated with imatinib until disease progression or intolerance. Further analysis showed, however, that imatinib interruption was safe, in that 79% of patients with disease progression after interruption achieved further tumor control when imatinib treatment was resumed. In addition, imatinib interruption had no impact on OS – 1 year survival rates were 89% in the imatinib interruption group and 87% in the continuous treatment group (P=0.46) – a finding largely consistent with ours (6). Furthermore, imatinib interruption did not have a detrimental effect on quality of life (6).

In the current study, two patients in the INT group initially had locally advanced unresectable gastric GIST. After achievement of PR in response to imatinib, both patients received surgery on their residual disease with curative intent, and both have had no evidence of disease progression in the absence of imatinib. These findings suggest that imatinib can be effective as a neoadjuvant treatment in locally advanced GIST. Surgery was also performed on 11 patients in the CONT group who achieved SD or better response on imatinib and, following surgery, none of these patients had gross evidence of residual disease. At the time of the present analysis, seven of these patients have achieved CR without any evidence of DP, whereas four had PD, despite continuous imatinib therapy (data not shown). However, all of these 11 patients in the CONT group underwent metastatectomy, with none having locally advanced tumors, making it difficult to compare the outcome of surgery between the groups.

Our results also showed divergent clinical outcomes in patients with GIST who discontinued imatinib therapy after obtaining radiological CR. Two patients maintained CR throughout the period of interruption, whereas others showed overt progression, highlighting the occurrence of invisible residual disease after achievement of radiological CR and the uncertain impact of radiological CR on further progression after the discontinuation of imatinib.

Although most patients with advanced GIST benefit from imatinib, some may subsequently develop resistance to it, with a median time of about 24 months (4). Acquired imatinib resistance during treatment has been found to occur through the development of secondary c-kit mutations, which were not present in primary resistant tumors (12). It is important to determine whether continuous imatinib treatment may contribute to the development of acquired resistance through secondary c-kit mutations and/or the selection of resistant clones. The possibility that imatinib interruption may reduce the emergence of resistance should be evaluated carefully.

The present study has several limitations inherent to retrospective design and analysis, calling for cautious interpretation of the results. First, imatinib treatment was interrupted at different times in the INT group. Second, the two arms are not well balanced in the number of patients included in each group and in the initial responses to imatinib. Last, a relatively small number of patients were enrolled and the follow-ups were relatively short, which makes it hard to address the impact of imatinib interruption on imatinib-refractory PFS and OS. However, to our knowledge, this is the first study on the clinical outcome of imatinib interruption in clinical practice setting, which could give us hints to confronting these problems.

In conclusion, in the patients with advanced GIST controlled with imatinib, imatinib interruption resulted in the high risk of PD within 1 year. However, the majority of the disease was controlled with imatinib re-challenge on PD and there were no statistically significant differences in imatinib-refractory PFS and OS between groups. Imatinib may be interrupted, at least temporarily, in patients with GIST controlled with it when various clinical situations constrain continuous treatment. Well-designed randomized controlled trials with OS as a primary endpoint are needed to make a firm conclusion.

References

1
Demetri
GD
von Mehren
M
Blanke
CD
Van den Abbeele
AD
Eisenberg
B
Roberts
PJ
, et al.  . 
Efficacy and safety of imatinib mesylate in advanced gastrointestinal stromal tumors
N Engl J Med
 , 
2002
, vol. 
347
 (pg. 
472
-
80
)
2
van Oosterom
AT
Judson
IR
Verweij
J
Stroobants
S
Dumez
H
Donato di Paola
E
, et al.  . 
Update of phase I study of imatinib (STI571) in advanced soft tissue sarcomas and gastrointestinal stromal tumors: a report of the EORTC Soft Tissue and Bone Sarcoma Group
Eur J Cancer
 , 
2002
, vol. 
38
 
Suppl 5
(pg. 
S83
-
7
)
3
Verweij
J
van Oosterom
A
Blay
JY
Judson
I
Rodenhuis
S
van der Graaf
W
, et al.  . 
Imatinib mesylate (STI-571 Glivec, Gleevec) is an active agent for gastrointestinal stromal tumours, but does not yield responses in other soft-tissue sarcomas that are unselected for a molecular target. Results from an EORTC Soft Tissue and Bone Sarcoma Group phase II study
Eur J Cancer
 , 
2003
, vol. 
39
 (pg. 
2006
-
11
)
4
Verweij
J
Casali
PG
Zalcberg
PJ
LeCesne
A
Reichardt
P
Blay
J-Y
, et al.  . 
Progression-free survival in gastrointestinal stromal tumours with high-dose imatinib: randomised trial
Lancet
 , 
2004
, vol. 
364
 (pg. 
1127
-
34
)
5
Wilson
J
Connock
M
Song
F
Yao
G
Fry-Smith
A
Raftery
J
, et al.  . 
Imatinib for the treatment of patients with unresectable and/or metastatic gastrointestinal stromal tumours: systematic review and economic evaluation
Health Technol Assess
 , 
2005
, vol. 
9
 (pg. 
1
-
142
)
6
Le Cesne
A
Perol
E
Ray-Coquard
I
Bui
B
Duffaud
F
Rios
M
, et al.  . 
Interruption of imatinib (IM) in GIST patients with advanced disease: updated results of the prospective French Sarcoma Group randomized phase III trial on survival and quality of life
J Clin Oncol
 , 
2005
, vol. 
23
 
16: suppl
pg. 
9031
  
(Abstracts)
7
Therasse
P
Arbuck
SG
Eisenhauer
EA
Wanders
J
Kaplan
RS
Rubinstein
L
, et al.  . 
New guidelines to evaluate the response to treatment in solid tumors
J Natl Cancer Inst
 , 
2000
, vol. 
92
 (pg. 
205
-
16
)
8
Green
S
Weiss
GR
Southwest Oncology Group standard response criteria, endpoint definitions and toxicity criteria
Invest New Drugs
 , 
1992
, vol. 
10
 (pg. 
239
-
53
)
9
Ryu
M-H
Lee
J-L
Chang
HM
Kim
TW
Kang
HJ
Sohn
HJ
, et al.  . 
Patterns of progression in gastrointestinal stromal tumor treated with imatinib mesylate
Jpn J Clin Oncol
 , 
2006
, vol. 
36
 (pg. 
17
-
24
)
10
Kaplan
ES
Meier
PM
Nonparametric estimation from incomplete observation
J Am Stat Assoc
 , 
1958
, vol. 
53
 (pg. 
457
-
81
)
11
Mantel
N
Evaluation of survival and tow new rank order statistics arising in its consideration
Cancer Chemother Rep
 , 
1966
, vol. 
50
 (pg. 
163
-
70
)
12
Antonescu
CR
Besmer
P
Guo
T
Arkun
K
Hom
G
Koryotowski
B
, et al.  . 
Acquired Resistance to imatinib in gastrointestinal stromal tumor occurs through secondary gene mutation
Clin Cancer Res
 , 
2005
, vol. 
11
 (pg. 
4182
-
90
)
Presented orally in part at the European Cancer Conference-ECCO13, Paris, France, 30 October–3 November 2005.

Author notes

The first two authors contributed equally to this work.