If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
The Netherlands Cancer Institute Antoni van Leeuwenhoek, Plesmanlaan 121, 1066 CX, Amsterdam, the NetherlandsErasmus MC Cancer Institute, Erasmus University Medical Center, Dr. Molewaterplein 40, 3015 GD, Rotterdam, the Netherlands
Primary surgically treated GISTs have only a limited risk of local recurrence (LR).
•
Patients who developed LR showed more aggressive and ruptured primary GIST.
•
In patients with LR only promising treatment strategies are still remaining.
•
Less intense follow-up could be considered during adjuvant treatment.
•
Less intense follow-up could also be considered after treatment with surgery only.
ABSTRACT
Background and objectives
Previous literature showed a high risk of recurrence following surgical treatment in patients with gastro-intestinal stromal tumors (GIST). However, little is known about the patient- and treatment characteristics of local recurrences (LR) in GIST patients. Therefore, this study aimed to better understand patterns of LR in surgically treated localized GIST and to describe treatment options based on our Dutch GIST Registry (DGR).
Methods
Data of primary surgically treated localized GIST between January 2009 until July 2021 were retrospectively retrieved from the DGR.
Results
Of 1452 patients registered in the DGR, 912 patients were included in this study. Only 3.8% (35/912) of patients developed LR, including 20 patients with LR only and 15 patients with simultaneous LR and distant metastases (DM)). Median time to LR was 30 (IQR 8-53) months from date of surgery. Eleven percent (100/912) of patients developed only DM. A total of 2.3% (6/259) of patients treated with adjuvant treatment developed a LR during adjuvant therapy. Seventy percent of patients with LR only (14/20) were treated with surgery (85.7% R0), which was mostly combined with systemic treatment.
Conclusions
Patients with primary surgically treated localized GIST have a limited risk of developing recurrence. Fifteen percent developed recurrence, of which one quarter developed a LR. Therefore, less intensified follow-up schedules could be considered, especially during treatment with adjuvant imatinib. In patients with LR only, potentially curative treatment strategies, including surgical (re-)resection, are often possible as treatment for LR.
Gastro-intestinal stromal tumors (GIST) are mesenchymal tumors originating from interstitial cells of Cajal arising predominantly in the gastro-intestinal tract [
]. Mutations in GIST are most often observed in the proto-oncogene receptor tyrosine kinase (KIT) gene and less frequently in the platelet-derived growth factor receptor α (PDGFRA). KIT exon 11 is the most common mutation, occurring in 70% of all GISTs [
Gastrointestinal stromal tumors: the incidence, prevalence, clinical course, and prognostication in the preimatinib mesylate era--a population-based study in western Sweden.
]. Since the implementation of tyrosine kinase inhibitors (TKIs) such as imatinib, the overall survival (OS) of metastatic GIST or locally advanced GIST dramatically increased [
Correlation of Long-term Results of Imatinib in Advanced Gastrointestinal Stromal Tumors With Next-Generation Sequencing Results: Analysis of Phase 3 SWOG Intergroup Trial S0033.
Discontinuation of imatinib in patients with advanced gastrointestinal stromal tumours after 3 years of treatment: an open-label multicentre randomised phase 3 trial.
Follow-up results after 9 years (yrs) of the ongoing, phase II B2222 trial of imatinib mesylate (IM) in patients (pts) with metastatic or unresectable KIT+ gastrointestinal stromal tumors (GIST).
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.
For localized disease, after surgery, GIST are usually classified as low-, intermediate-, and high-risk tumors according to the Miettinen’s 2006 criteria to predict the risk of recurrence. The first accepted classification to define risk of aggressive behavior in GISTs was the U.S. National Institutes of Health (NIH) Consensus Criteria (2001), which was based on tumor size and mitotic index [
]. The Miettinen’s criteria, also known as the staging system of the American Joint Committee on Cancer (AJCC), is based on tumor size, location and mitotic index [
]. Approximately half of the patients treated with curative intent develop recurrent disease. In 5-35% of patients this is a local recurrence (LR) meaning recurrence at or near the primary tumor location [
]. Little is known about the patient- and treatment characteristics as well as the oncological outcome of LR in GIST patients.
Therefore, the main aim of this study is to better understand patterns of local recurrence in surgically treated localized GIST and to describe treatment options based on our large series of data of Dutch GIST patients treated in GIST reference centers.
METHODS
This multicenter study was approved by the local review boards of centers participating in the national database of patients with GIST in the Netherlands: the Dutch GIST Registry (DGR).
2.1. Study population
Patients of 18 years or older with primary GIST, surgically treated between January 2009 until July 2021 were selected from the DGR. Primary GIST was defined as local or locally advanced GIST, without distant metastases. The data were collected from the DGR database containing data of all adult GIST patients treated in one of the five Dutch GIST centers (UMC Groningen, LUMC Leiden, Erasmus MC Cancer Institute Rotterdam, RadboudUMC Nijmegen and the Netherlands Cancer Institute Amsterdam). Patients with metastases at entry, multiple primary tumors or patients that did not underwent a surgical resection were excluded.
2.2. Variables
Patient demographics and tumor characteristics including molecular pathology reports, treatment and follow-up data were collected. Locally advanced disease was defined as patients with non-metastatic disease and an indication for neo-adjuvant treatment. Local recurrence (LR) was defined as the first clinical, radiological (CT- and/or PET-scan and/or MRI-scan) or pathological manifestation of GIST within or contiguous to the previously treated tumor bed, three or more months after surgery with or without (neo-)adjuvant treatment. LR only was defined as local recurrence without distant metastases (DM) at time of presentation of LR. Resection margins were defined as follow: R0, R1 and R2 corresponding to no residual tumor at the resection margin, microscopic residual tumor and intralesional macroscopic residual tumor, respectively. Tumor rupture was defined as (microscopically) disrupted serosa, as described by an experienced pathologist, tumor spillage or gastro-intestinal perforation at tumor site in the abdominal cavity. Mitotic rate was defined as high or low that is respectively > 5 mitoses per 5 mm2 or 50 High Power Fields (HPF) or ≤ 5 mitoses per 5 mm2 or 50 HPF. Patients were classified into risk groups based on the Miettinen’s criteria [
]. Mutational status was assessed by sequencing on formalin fixed paraffin (FFPE). Different hotspots of both KIT and PDGFRA genes were tested. In case of a wildtype GIST, there was no mutation found in the both KIT and PDGFRA genes. Time of follow-up (FU) was defined as date of surgery until date of last contact or dead.
2.3. Statistical analysis
Descriptive statistics were used to summarize patient- and tumor characteristics. For categorical or dichotomous variables absolute numbers and percentages within each group were reported. For continuous variables mean and standard deviation (SD) were reported in case of normally distributed data and median and interquartile range (IQR) were reported for non-normally distributed data. Baseline patient- and tumor characteristics were compared using the Chi-squared test or the Kruskal-Wallis test. All statistical analyses were performed using IBM SPSS Statistics 24.0 [
]. Values of P < 0.05 were considered as statistically significant.
RESULTS
3.1. Study population
3.1.1. Baseline characteristics
From the 1452 patients registered in the Dutch GIST Registry, a total of 912 GIST patients were included in this study cohort. Median FU in the entire study cohort was 30 months (IQR 8-62) from date of surgery with a range of 0-140 months. Five-hundred-forty patients were excluded due to diagnosis before 2009, metastasis at time of registry in the DGR database, multiple primary tumors, no surgical treatment or surgical treatment and follow-up in center of referral. Thirty-five (3.8%) primary surgically treated GIST patients developed a LR, 100 (11%) developed distant metastasis (DM) only (Figure 1.).
Regarding baseline patient characteristics, patients who developed LR had significantly more often larger tumors of more than 10 cm, higher mitotic rate and tumor rupture. Patients who developed LR had more often an indication for neo-adjuvant therapy at presentation and a R1- or R2 resection. As expected, patients more often developed LR after surgery in higher risk tumors according to the Miettinen’s criteria compared with patients who did not develop LR. Location, mutation status and histology did not significantly differ between both groups. However, patients who developed LR showed a tendency to present more often with primary non-gastric GIST (Table 1).
Table 1Baseline patient characteristics in primary surgically treated GIST patients.
No recurrence
DM only
LR
p-value*
(n = 777)
(n = 100)
(n = 35)
LR only (n = 20)
LR + DM (n = 15)
Gender
0.865
Male Female Missing
387 (49.8) 390(50.2) 0
54 (54.0) 46 (46.0) 0
8 (40.0) 12 (60.0) 0
10 (66.7) 5 (33.3) 0
Age**(median, IQR)
64 (55-72)
60 (51-68)
62 (52-70)
62 (59-70)
0.675
Location
0.094
Gastric Small bowel Duodenal Rectum Esophagus Colon Other Missing
Thirty-five (3.8%) patients developed LR, with a median time to LR of 30 (IQR 8-53) months from date of surgery. Patients with a LR had a median FU of 65 (IQR 36-92.5) months from date of surgery. Median tumor size of all LR at presentation was 36 (IQR 20-102) mm. One patient was previously treated in a foreign medical center and had LR at presentation, data about primary surgery was missing. Fifteen (42.9%) of 35 patients presented with both LR and DM. Patients who presented with DM at time of LR had metastases mostly located in the liver followed by intra-abdominal lesions (Table 2.).
Table 2. Characteristics of local recurrence in primary surgically treated GIST patients.
Total (n = 35)
LR only (n = 20)
LR + DM (n = 15)
Time to LR in months (median, IQR)*
30 (8-53)
19 (7-50)
41 (20-59)
Tumor size LR in mm (median, IQR)
36 (20-102)
35 (20-118)
36 (18-68)
Location DM**
NA
NA
Liver Intra-abdominal Liver + intra-abdominal
4 (26.7) 7 (46.7) 4 (26.7)
Treatment***
Only surgery Systemic treatment Surgery + systemic treatment
3 (8.57) 21 (60.0) 11 (31.4)
3 (15.0) 6 (30.0) 11 (55.0)
0 (0.00) 15 (100) 0 (0.00)
LR = local recurrence. DM = distant metastasis.
* From time of surgery (n = 1 missing).
** DM at presentation of LR.
*** Surgery for local recurrence (n = 1 patient with metastasectomy without resection of LR not included)
Of all patients treated with neo-adjuvant or adjuvant imatinib only, respectively 5.2% (8/154) and 5.5% of patients (8/145) developed a LR. Patients treated with neo-adjuvant therapy only had in 32% a high risk GIST, in patients treated with adjuvant imatinib only this was 69%. Of all patients treated with both neo-adjuvant and adjuvant imatinib, 9.7% of patients (11/114) developed a LR. Among these patients, 87.9% had high risk GISTs.
Of all patients only treated with surgery, 1.6% (8/499) patients developed a LR and 6.2% (31/499) patients developed DM. In this group, patients had a median FU of 22 (IQR 4-56) months. Among these patients, 10.6% had high risk GIST. In the subgroup of intermediate- and high risk patients (20.4%, 102/499), 2.9% (3/102) and 16.7% (17/102) developed LR or DM, respectively.
In patients treated with adjuvant treatment, 2.3% (6/259) developed a LR during adjuvant therapy, of whom 4 patients were classified as high-risk GIST. Four patients had a KIT exon 11 mutation, 1 patient a PDFGRA mutation and 1 patient a wildtype GIST. The median time of developing LR during adjuvant therapy from start of adjuvant therapy was 7 (IQR 6.5-29) months. In addition, 5.0% (13/259) of patients developed DM during adjuvant therapy.
Of all known high-risk GIST, 37.2% (61/164) of the patients was not treated with adjuvant imatinib. Only 4.9% (3/61) of these patients developed a LR, however, 31.1% (19/61) patients developed DM. Reasons for no adjuvant treatment in this group were comorbidity, older age at time of surgery, preference of patient or treatment before the implementation of adjuvant imatinib.
3.2.3. Treatment of LR
Seventy percent (14/20) of patients with LR only were treated by surgical resection. Of all patients of whom LR was surgically resected, 10 (66.7%) had a KIT exon 11 mutated GIST, 3 wildtype GIST and 1 PDGFRA (D842V) mutated GIST (1 missing data). In 85.7% (12/14, 1 missing data) of patients this was a radical (R0) resection. In 78.6% (11/14) surgery was combined with systemic treatment (Table 2.). Thirty percent (6/20) of the patients with LR only was not treated with surgery due to irresectable tumors, preference of patient, stable disease during treatment with imatinib due to increase dose or continuing ongoing treatment with imatinib. One patient with LR only at presentation developed DM and underwent a metastasectomy without resection of LR in addition to systemic therapy. All patients who developed LR with DM at presentation of LR (15/35) were treated with systemic therapy only (Figure 2.).
Fig. 2. Treatment and follow-up in primary surgically treated GIST patients with local recurrence. LR = local recurrence. DM = distant metastasis.
Forty-three percent (15/35) of all GIST patients who developed LR presented with DM at time of LR. Seventeen percent (6/35) of patients developed DM several months after LR. Median time from LR to DM in these patients was 13 (IQR 6-37) months. One patient was treated for DM before presentation of LR (Figure 2.).
3.2.5. Follow-up in expertise center
A total of 31 patients with LR were followed in an expertise center. Follow-up schedules are based mainly on regular CT scans with intervals depending on risk classification ranging from every 4 to every 12 months. The other 11.4% (4/35) patients were followed in the center of referral, and therefore the number of scans was unknown in this group. Median FU time of the patients with LR who were followed in an expertise center was 64 (IQR 36-90) months after surgery of primary GIST. Within this timeframe, a median of 6 (IQR 2-10) scans were made during follow-up. Median time to LR was 34 (IQR 16-53) months in patients followed in a expertise center. However, a wide range from 1 to 20 scans was performed in a time frame of 3 to 114 months before detection of the LR.
Twenty-eight patients (80.0%) survived after treatment of LR, with a median follow-up of 29 (IQR 14 – 49.5) months after presentation of LR. Nine patients (32.1%) had no evidence of disease while 19 (67.9%) patients were alive with evidence of disease.
Six patients (17.1%) died after developing LR due to GIST, while 1 patient died due to another cause (cerebrovascular accident). Almost all patients who died (86%) had DM at presentation of LR. The median time of survival after presentation of LR of patients who died due to disease was 33 (IQR 17-39.5) months.
DISCUSSION
In this study, primary surgically treated GIST showed a limited risk of recurrence. Fifteen percent developed a recurrence, of which only one quarter developed a LR. Patients who developed LR had more aggressive GIST with larger tumor sizes, higher mitotic count and therefore were most often classified as high-risk GIST according to the Miettinen’s criteria.
According to the ESMO guidelines an R1 margin is acceptable after treatment with neo-adjuvant therapy when R0 otherwise implicates a multivisceral resection [
Quality of Surgery and Outcome in Localized Gastrointestinal Stromal Tumors Treated Within an International Intergroup Randomized Clinical Trial of Adjuvant Imatinib.
]. In this study patients who developed LR showed significant more often R1 resections compared with patients who did not develop a LR. However, a multivariate regression analysis adjusting for important prognostic factors such as Miettinen’s criteria, margin and tumor rupture could not be performed due to a small sample size of patients with a LR.
The ESMO guidelines recommend treatment with adjuvant imatinib for high-risk GIST [
]. In our study, a third of the patients with a high-risk GIST according to Miettinen’s criteria did not receive adjuvant therapy after surgery. The main reason for this could be that adjuvant treatment with imatinib was implemented after 2011, other reasons were comorbidity or wish of patient. As expected, this group showed a relatively higher incidence of recurrences (4.91% LR and 31.3% DM) compared with the total group. However, a longer FU time is also observed in these patients.
No data are available in known literature for an optimal follow-up strategy in GIST patients and follow-up strategies differ across institutions. ESMO guidelines are therefore expert opinion based. Because of this lack of data, there seems to be a conservative tentative trend for intensive follow-up schedules in GIST patients. This could also be explained by previous literature describing a high risk of recurrence (both LR and DM) of 50% in primary surgically treated GIST [
]. However, our current study showed a significantly lower risk of recurrence of only 15%, which could be explained by the implementation of adjuvant imatinib in 2011. Patients who developed LR and were followed in a expertise center underwent median 6 follow-up scans with a wide range of 1 to 20 scans. In addition, in all patients with LR, the median time to LR was 30 (IQR 8-53) months from date of surgery of primary GIST. Therefore, less intensified follow-up schedules should be considered, especially in patients who were treated with surgery only. These patients also showed less high-risk GIST compared with patients treated with (neo-)adjuvant imatinib. However, further research is needed to assess long-term data of both DM and LR rates.
Adjuvant imatinib was implemented to reduce the risk of recurrence in primary surgically treated (high-risk) GIST. Among patients who develop recurrences despite adjuvant treatment, patients with high-risk GIST most often present with recurrences 1-3 years after cessation of adjuvant treatment [
]. Therefore patients with high-risk GIST are more intensified followed after cessation of adjuvant treatment with imatinib is completed. Also in this study, irrespectively of the risk classification, most patients developed recurrence after cessation of adjuvant treatment with imatinib. According to national Dutch guidelines, patients with an indication for adjuvant therapy undergo an intensive follow-up schedule with every 6 months a follow-up CT-scan during adjuvant therapy and every 4 months a CT-scan within the first 2 years after end of adjuvant treatment. Our study showed that only 2.3% of the patients treated with adjuvant therapy developed LR during adjuvant therapy. A total of 5.0% of patients developed DM only during adjuvant therapy. According to this finding, a less intensive follow-up schedule of a CT-scan once a year could be considered during treatment with adjuvant imatinib despite the relatively high proportion of high-risk GIST in this group. After completion of adjuvant therapy, more intensive follow-up schedules are needed, especially in patients with high-risk GIST.
Most (70%) patients with LR only could be treated with surgical resection. This was most often combined with systemic therapy. All patients with both LR and DM at presentation were treated with systemic treatment only.
Before the implementation of imatinib, poor survival with a median of 5 months was reported in GIST patients who developed LR [
]. However, in our study only 17% (6/35) of all patients who developed LR died due to disease with a median time of survival after presentation of LR of 33 (IQR 17-39.5) months. GIST patients who developed LR who were still alive (with or without disease) in this cohort had a median time of FU 29 (IQR 14-49.5) months after development of LR. Patients who showed DM at presentation of LR appeared to have worse survival, however, due to the small sample size a statistical analysis could not be performed.
Another limitation of this study is the retrospectively collected data, resulting in missing data which could induce selection bias. Moreover, the cohort design resulted in a wide range of 0 to 140 months of FU after surgery. This resulted in a median FU of 30 months and therefore no long-term conclusions could be drawn. Due to a low number of LR, a required landmark analysis for overall survival in GIST patients according to LR status and/or according to treatment of LR (surgical versus systemic treatment) could not be performed. Further research is needed to better understand long-term patterns of recurrences (both LR and DM) in patients with primary surgically treated localized GIST. Nevertheless, this national multicenter study is the largest series examining LR in primary surgically treated GIST patients.
CONCLUSION
After the introduction of adjuvant imatinib patients with primary surgically treated localized GIST generally have a limited risk of recurrent disease and only have a small risk to develop LR. In patients with LR without DM, potentially curative treatment strategies, including surgical (re-) resection, are still possible. This study shows that less intensified follow-up schedules could be considered, especially during treatment with adjuvant imatinib.
FUNDING
This work was supported by a research grant for the Dutch GIST Registry, which was received from Novartis (3017/13), Pfizer (WI189378), Bayer (2013-MED-12005) and Deciphera (4EE9EEC-7F19-484D-86A4-646CFE0950A5).
The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: This work was supported by a research grant for the Dutch GIST Registry, which was received from Novartis (3017/13), Pfizer (WI189378), Bayer (2013-MED-12005) and Deciphera (4EE9EEC-7F19-484D-86A4-646CFE0950A5). These funding sources did not have any involvement in the conduction of this research.
ACKNOWLEDGEMENTS
The authors would like to thank Novartis, Pfizer, Deciphera and Bayer for grants received for the infrastructure of the Dutch GIST Registry.
DECLARATION OF INTEREST STATEMENT
The funding sources did not have any involvement in the conduction of this research.
REFERENCES
DeMatteo R.P.
et al.
Two hundred gastrointestinal stromal tumors: recurrence patterns and prognostic factors for survival.
Gastrointestinal stromal tumors: the incidence, prevalence, clinical course, and prognostication in the preimatinib mesylate era--a population-based study in western Sweden.
Correlation of Long-term Results of Imatinib in Advanced Gastrointestinal Stromal Tumors With Next-Generation Sequencing Results: Analysis of Phase 3 SWOG Intergroup Trial S0033.
Discontinuation of imatinib in patients with advanced gastrointestinal stromal tumours after 3 years of treatment: an open-label multicentre randomised phase 3 trial.
Follow-up results after 9 years (yrs) of the ongoing, phase II B2222 trial of imatinib mesylate (IM) in patients (pts) with metastatic or unresectable KIT+ gastrointestinal stromal tumors (GIST).
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.
Quality of Surgery and Outcome in Localized Gastrointestinal Stromal Tumors Treated Within an International Intergroup Randomized Clinical Trial of Adjuvant Imatinib.
00142-9&id=gr1.jpg){kind=link}
00142-9&id=gr2.jpg){kind=link}