Mutational characteristics of gastrointestinal stromal tumors: A single‑center analysis of 302 patients
- Authors:
- Published online on: January 4, 2021 https://doi.org/10.3892/ol.2021.12435
- Article Number: 174
-
Copyright: © Liang et al. This is an open access article distributed under the terms of Creative Commons Attribution License.
Abstract
Introduction
Gastrointestinal stromal tumors (GISTs) are the most frequently occurring mesenchymal tumors of the digestive tract and are characterized by differentiation from the interstitial cells of Cajal (1). GISTs predominantly (60-70%) arise in the stomach, followed by the small bowel (20-30%) (2). The pathological diagnosis of GISTs is based on various morphological manifestations together with sensitive and specific markers, including CD117, discovered on GIST-1 (DOG1, also known as anoctamin 1) and CD34, identified by immunostaining (3). GISTs represent a wide spectrum of tumors, with variable disease behaviors associated with tumor size, mitotic activity and anatomical origin (3–5). According to these three clinicopathological features, several recurrence risk assessment systems have been developed and used for primary GISTs. The Chinese Society of Clinical Oncology Expert Committee recommends the modified US National Institutes of Health (NIH) classification, which they consider to be particularly suitable for Asian populations (6,7). In addition, the World Health Organization (WHO) classification recommends the US Armed Forced Institute of Pathology criteria, which are classified into eight grades (grades 1, 2, 3a, 3b, 4, 5, 6a and 6b) (8,9).
It has been reported that 82–87% of GISTs harbor gain-of-function mutations in the KIT proto-oncogene, receptor tyrosine kinase (KIT) or platelet derived growth factor receptor α (PDGFRA) oncogenes encoding type III receptor tyrosine kinases (10–12). KIT mutations have been reported to occur in 69–83% of all GISTs and PDGFRA mutations have been detected in 12.9-14.0% of primary GISTs (10,12,13). In adults, ~15% of GISTs without detectable mutations in KIT or PDGFRA are considered wild-type (WT) GISTs (14). The most frequent site of KIT mutation is at the 5′ end of exon 11; this encodes the JM domain which has an autoinhibitory function under ligand-free conditions. Mutations of exon 11 disrupt this autoinhibitory function and thereby result in ligand-independent receptor activation (15). Several studies have demonstrated that exon 11 deletions of KIT are associated with a high risk of relapse and metastasis. In particular, GISTs with deletions affecting codons 557/558 exhibit a higher risk of progression (10,11,16,17). However, in comparison with KIT deletions, KIT exon 11 substitutions indicate an improved patient outcome (10,18). Furthermore, GISTs with PDGFRA exon 18 mutations have a lower risk of relapse than those with KIT deletions (11). However, although the roles of KIT and PDGFRA mutations in the assessment of the response to imatinib therapy are well documented (3,7,9), the prognostic significance of these mutations in the Chinese population has yet to be defined. Furthermore, whether genotype analysis should be considered as an additional prognostic approach is currently unclear.
Therefore, the aim of the present study was to analyze the clinicopathological and mutational characteristics of GISTs and evaluate the prognostic significance of these parameters in a large cohort of 302 cases in North China. The findings may be helpful for risk assessment and personalized targeted therapy.
Materials and methods
Patients and samples
A series of 302 GIST cases was retrospectively collected from records archived in the Department of Pathology, Peking University First Hospital (Beijing, China). The cases were diagnosed between May 2009 and June 2019, and represent ~2.1% of all gastrointestinal malignancies at this center. Inclusion and exclusion criteria were as follows. Patients with primary GIST diagnosis and primary therapy-naive tumors with curative resection were eligible (R0 or R1). The information on clinicopathological factors, follow-up data and the mutational status of KIT and PDGFRA were available. Patients undergoing neoadjuvant imatinib or chemoradiotherapy for GIST before surgery were excluded. GIST was diagnosed based on histopathological features, immunochemical findings and genotype according to the 5th edition of the WHO classification (9). The study was approved by the ethics committee of Peking University First Hospital and was conducted in compliance with the Declaration of Helsinki. Written patient consent for use of their tissues in research was obtained.
For each case, the histological assessment included location, tumor size, mitotic count per 50 high-power fields (HPF; equivalent to 5 mm2), cell type and the presence or absence of rupture. Risk was stratified and prognostic grades were evaluated according to the modified NIH consensus and WHO classification, respectively (6,9).
Immunohistochemical staining was carried out on 4-µm thick sections form a total of 302 cases of paraffin-embedded tissue blocks which were fixed using 10% formalin at room temperature for 24 h. Briefly, tissue sections were incubated at 65°C for 10 min, followed by two 10-min cycles of deparaffinization using xylene and then hydration in a graded ethanol series (100, 100, 95, 80 and 70% for 2 min, respectively). They were pretreated to promote antigen retrieval in EDTA-Tris (pH 9.0) at 95°C for 20 min (PT Link; Dako; Agilent Technologies, Inc.) and were treated with 3% hydrogen peroxide for 10 min to block endogenous peroxidase. Tissues were subsequently incubated with primary antibodies for 50 min at room temperature. The panel of protein immunostained for GIST diagnosis was as follows: CD117 (working solution, Maixin), DOG1 (1:100; ZSGB-BIO), CD34 (1:200; ZSGB-BIO), Ki67 (1:100; ZSGB-BIO), S-100 (1:200; ZSGB-BIO), SDHB (1:100; ZSGB-BIO), smooth muscle actin (1:200; ZSGB-BIO) and desmin (1:100; ZSGB-BIO). Sections were incubated with secondary antibody using a EnVision™, FLEX+, High pH kit (cat. no. K8002; Dako; Agilent Technologies, Inc.) according to the manufacturer's instructions. Strong positive expression of CD117, DOG1, CD34 or SDHB in >50% of the tumor tissue was defined as cytoplasmic immunopositivity by eye under a light microscope.
Follow-up were performed for 259 GIST cases and patient information was obtained by regular outpatient visits or by telephone. Relapse-free survival (RFS) was defined as the duration from surgery to relapse (local recurrence or metastasis). Relapse was identified based on biopsy and/or imaging assessment.
Gene mutation analysis
Genomic DNA was extracted from formalin-fixed, paraffin-embedded tissue using a DNeasy Blood & Tissue Kit (Qiagen, Inc.). Prior to extraction, histological assessment was performed to ensure that the percentage of tumor in the specimens by area was >80%. Mutational analysis of KIT and PDGFRA was carried out by PCR amplification, followed by Sanger sequencing of the amplified products. Briefly, initial amplification was performed using Takara LA Taq polymerase (cat. no. RR02MA; Takara Bio, Inc.). The PCR amplification program was as follows: Denaturation at 94°C for 5 min, 45 cycles of denaturation at 94°C for 30 sec, annealing at 56°C for 45 sec, extension at 72°C for 20 sec, and finally, incubation at 72°C for 10 min. The sequencing reaction products were electrophoresed on an ABI3700 genetic analyzer (Applied Biosystems; Thermo Fisher Scientific, Inc.). KIT exons 9, 11, 13 and 17, and PDGFRA exons 12, 14 and 18 were analyzed. The primer sequences used are shown in Table SI.
Statistical analysis
For univariate analysis, the χ2 test or Fisher's exact test was used to compare categorical variables. Survival analysis was carried out using the Kaplan-Meier method, and statistical significance was determined using the log-rank test. Univariate and multivariate Cox proportional hazard models were used to determine the prognostic impact of variables on RFS. The strengths of the associations are shown as hazard ratios (HRs) and corresponding 95% confidence intervals (CIs). A 2-sided P<0.05 was considered to indicate a statistically significant result. Analyses were performed using IBM SPSS Statistics V23.0 software (IBM Corp.).
Results
Clinicopathological characteristics
The detailed clinical and pathological characteristics of the 302 patients with GISTs are provided in Table I. There were 160 males (53.0%) and 142 females (47.0%), who ranged in age from 13 to 84 years (median, 61.8 years), with patients >60 years old accounting for 54.3% of cases. The majority of tumors were located in the stomach (66.9%), and the small intestine was the second most common location (25.5%). Only 14 cases (4.6%) had tumors in the colon or rectum, and the remaining locations (3.0%) involved the prostate (1 case), retroperitoneum (2 cases), abdominal cavity (5 cases) and pelvic cavity (1 case). The tumor size ranged from 0.3 to 26.0 cm (median, 4.2 cm) at initial diagnosis, and was ≤5 cm in most cases (60.6%) and >10 cm in 41 cases (13.6%). However, the sizes of the tumors in non-gastric sites were larger than those in the stomach. Most GISTs (75.8%) exhibited low mitotic activity (≤5/50 HPF), with no marked difference between gastric and non-gastric sites. Tumor necrosis was observed in 67 cases (22.2%), and was predominant in non-gastric sites. Only 2 cases (0.7%) had a ruptured tumor. The predominant morphology was spindle variant (87.4%), and only 9 cases (3.0%) presented epithelioid histology, which was more frequently seen in tumors in the stomach. The remaining 29 cases (9.6%) featured a combination of spindle and epithelioid morphology (mixed variant).
Strong expression of CD117, DOG1 and CD34 was detected in the majority of GIST cases (95.0, 93.4 and 81.1%, respectively). Among these cases, triple expression was detected in 219 cases (72.5%), double expression in 74 cases (24.5%) and single expression in 9 cases (3.0%). No triple-negative cases were observed in the study. Notably, CD34 expression was more frequently detected in tumors located in the stomach than in other sites (78.4 vs. 21.6%), and the loss of CD34 expression was predominant in specimens with epithelioid and mixed variant morphology (33.3 and 34.5%, respectively).
Based on the modified NIH consensus (6), as shown in Table I, more than half of cases were assessed as low or very low risk (162/302, 53.6%). There was very low risk of relapse in 19.2% of cases, in which the tumors were predominantly located in gastric sites. There was a high risk of relapse in 88 cases (29.1%), and this assessment was more frequent for non-gastric GISTs than for gastric GISTs. Correspondingly, based on the prognostic assessment recommended by the WHO classification (9), grade 6b tumors were predominantly located outside of the stomach, while grade 1 tumors were more frequently detected in the stomach (Table I).
Genotype analysis
Analyses of KIT and PDGFRA mutations were performed for all 302 GIST specimens in the present study. In total, KIT and PDGFRA mutations were identified in 233 (77.2%) and 30 (9.9%) cases, respectively, and WT KIT and PDGFRA were found in 39 specimens (12.9%). The mutational landscape is presented in Tables I and II. The predominant genotype was KIT exon 11 deletion (36.4%), followed by KIT exon 11 substitution (26.2%), PDGFRA exon 18 substitution (7.3%) and KIT exon 11 duplication (7.0%).
 | Table II.Associations between KIT mutations and clinicopathological characteristics of patients with GISTs (n=302). |
As shown in Fig. 1B, the spectrum of the exon 11 mutations involved codons ranging from codon 548 to 589. Among these codons, the more frequent mutations were found in codons 557–560, which harbored deletion and/or substitution subtypes, and the less frequent mutations were observed in codons 572–580, which exhibited duplication mutations. The length of the in-frame deletion ranged from 3 to 15 bp. The most prevalent type of deletion, involving codons 557/558/559, was identified in 86 cases (28.5%). Substitutions mostly involved codons 559 (n=36; p.V559D, p.V559A, p.V559G), 560 (n=20; p.V560D, p.V560E, p.V560G), 576 (n=9; p.L576P) and 557 (n=8; p.W557R, p.W557G).
In-frame duplication of exon 9 was detected in 16 cases (5.3%), and comprised p.A502_Y503dup (n=15) and p.F506_F508dup (n=1). There were 4 cases with exon 13 substitution (all p.K642E) and 3 cases with exon 17 substitutions (p.N822K, p.N822Y). Notably, in one tumor, the coexistence of exon 11 p.Y553_K558del and exon 13 duplication (one base pair insertion between codons 642 and 643) at a hotspot mutation site was observed. This tumor, whose size was >10 cm and mitotic count was >5/50 HPF, presented spindle and epithelioid morphology, and was classified as NIH high risk and high WHO grade 6b (Fig. 1C-F). Whether this patient received adjuvant therapy after resection is unknown.
The most frequent type of PDGFRA mutation comprised exon 18 substitutions involving codons 842 (n=21; p.D842V, p.D842T, p.D842Y) and 839(n=1;p.L839Q). Exon 18 deletions were also found in 5 cases (p.D842_H845del, p.M844_S847del, p.M844_D846del). In addition, PDGFRA exon 12 (p.Y555C) and exon 14 (p.N659K) substitutions were detected in 1 and 2 cases, respectively.
Genotype and clinicopathological characteristics
The associations between genotype and clinicopathological characteristics are presented in Tables I and II. The KIT exon 11 substitution subtype was commonly detected in tumors at gastric sites (58/79, 73.4%), while KIT exon 9 duplication was more frequently found in tumors at non-gastric sites (12/16, 75%). In addition, PDGFRA exon 18 substitutions and deletions were more frequent in tumors of the stomach (24/27, 88.9%) than in those of other sites. Larger (>10 cm) tumors were more likely to bear GISTs with KIT exon 11 deletions than exon 11 substitutions. KIT exon 11 deletions were also more frequent than KIT exon 11 substitutions in tumors with high mitotic counts (>5/50 HPF), a high risk of relapse and high (3b/5/6a/6b) WHO grades (Table II; all P<0.05). Deletions involving KIT codons 557/558/559 were more frequently identified in gastric GISTs than in non-gastric GISTS (62.8 vs. 37.2%, respectively). Furthermore, when compared with cases with other KIT mutations, cases with exon 11 deletions involving codons 557/558/559 were significantly associated with worse clinicopathological features: High mitotic activity, high risk according to the NIH classification and high WHO grade (Table II; all P<0.05). These findings suggested that KIT exon 11 deletions, particularly those involving codons 557/558/559, may contribute to poor prognosis.
In the present study, 6 cases had two nodular masses that were located in different sites of the abdominal cavity at the initial diagnosis of the primary tumor. Their clinicopathological features and genotypes are provided in Table III. Notably, in 4 cases (cases 1, 2, 5 and 6), specimens from the two different locations presented distinct morphological appearances and genetic alterations. However, in the other two cases (cases 3 and 4), the same morphological manifestations and genotypes were observed in specimens from both locations. In case 1, the primary tumor in the greater curvature of the stomach had spindle morphology, strong CD117 expression and KIT p.V560D mutation, whereas the other mass in the lesser curvature of the stomach was an epithelioid variant with loss of CD117 expression and WT KIT (Fig. 2). In case 6, as depicted in Fig. 3, most of the specimens of the tumor in the stomach (case 6A) exhibited anaplastic/pleomorphic morphology with multinucleated giant cells, high mitotic activity and necrosis, but the tumor located between the stomach and spleen (case 6B) predominantly exhibited spindle histology. The deletion of KIT codon 579 occurred in the former, while KIT substitution (p.W557G) occurred in the latter. Furthermore, CD34 expression was lost in the specimen of case 6A but not in that of case 6B. Thus, GISTs presented heterogeneous histopathological, immunostaining and gene mutation features, suggesting the need for comprehensive mutation detection and highlighting new challenges for diagnosis and therapy.
To evaluate the occurrence of morphological or genotypic heterogeneity in relapsed or metastatic GIST cases, histological and mutational analyses were performed for 9 patients with progression after resection (8 cases with recurrence and 1 case with hepatic metastasis). Of these cases, 3 were receiving imatinib treatment prior to progression. Notably, the morphological, immunohistochemical and mutational phenotypes of the primary tumors were observed to be the same as those of the recurrent and/or metastatic tumors (data not shown), suggesting the existence of homogenous features in recurrent and/or metastatic tumors, in contrast to the aforementioned primary tumors with two masses.
Succinate dehydrogenase (SDH)-deficient GIST accounts for approximately half of all WT GISTs. Previous studies have demonstrated that all SDH mutations are reliably detected by the loss of SDH subunit B (SDHB) expression using immunostaining (19,20). Thus, SDHB-immunohistochemistry (IHC) is an efficient method for the detection of SDH deficiency. In the present study, the results of SDHB-IHC were obtained for 139 GIST cases. Loss of SDHB expression was detected in 5 cases (3.6%), suggesting the presence of SDH-deficient GIST. These SDH-deficient GIST tumors, which were predominantly found in women (4 cases), occurred in the stomach and in older adults (age ≥48 years). The tumor cells of SDH-deficient GIST exhibited spindle morphology, which is not in accordance with previous observations of the predominance of epithelioid or mixed types.
Survival analysis
Follow-up data were available for 259 GIST cases. In this group, the median follow-up time ranged from 12 to 129 months (median, 45 months), and the median RFS ranged from 1 to 96 months (median, 41 months). Of these cases, regular imatinib therapy following surgery was received by 59 cases (22.8%), including 40 cases classified as high risk, 17 cases classified as intermediate risk, and 2 cases classified as low risk. Disease progression was observed in 33 cases (12.7%) and hepatic metastasis occurred in 11 cases (4.2%).
Kaplan-Meier survival analysis and log-rank tests indicated that male sex, a larger tumor size (>5 cm), a high mitotic index (>5/50 HPF), necrosis and epithelioid morphology were associated with an inferior RFS (Fig. 4; P=0.017, P<0.001, P<0.001, P<0.001, respectively). In addition, patients with non-gastric GISTs had poorer outcomes than those with gastric tumors (Fig. 4B; P<0.001), and patients classified as high risk and high WHO grade also exhibited significantly lower RFS rate (Fig. 4G and H; all P<0.001). Furthermore, KIT deletions involving codons 557/558/559 were significantly associated with a lower RFS rate than other genotypes (Fig. 4I; P=0.020). This adverse effect of KIT deletions affecting codons 557/558/559 on RFS was also observed among patients with gastric GISTs (Fig. 4J; P=0.001) but not among those with non-gastric tumors. KIT exon 11 deletion was significantly associated with a lower RFS rate, whereas exon 11 substitution was associated with a favorable outcome (Fig. 4K; P=0.025). Compared with codon 557/558/559 deletion, PDGFRA exon 18 mutation was associated with a favorable prognosis, although the association was not statistically significant (Fig. 4L, P=0.504).
Univariate analysis revealed that patients who were female or had tumors with spindle morphology had a lower risk of relapse (Table IV; P=0.021 and P=0.024, respectively). Conversely, non-gastric localization, high NIH risk stratification and high WHO grade were significantly associated with a higher risk of relapse [HR (95% CI)=3.383 (1.629-7.025), P=0.001; HR (95% CI)=10.359 (3.147-34.101), P<0.001; HR (95% CI)=9.125 (3.930-21.187), P<0.001, respectively; Table IV]. Additionally, regarding cases with KIT deletions involving codons 557/558/559, the risk of tumor progression was almost 3-fold higher that of cases with other KIT mutations [HR (95% CI)=2.722 (1.235-6.003), P=0.013; Table IV]. Multivariate Cox regression analysis also indicated that patients bearing KIT deletions involving codons 557/558/559 had an almost 3-fold higher risk of relapse than those with other KIT mutations [HR (95% CI)=2.794 (1.204-6.482), P=0.017; Table IV], suggesting that this genotype is an independent predictor of RFS. According to this multivariate analysis, tumor location and WHO grade also markedly influenced the risk of tumor progression [HR (95% CI)=2.420 (1.116-5.251), P=0.025; HR (95% CI)=3.166 (1.033-9.708), P=0.044; Table IV). In addition, based on multivariate Cox regression analysis, KIT codon 557/558/559 deletion was significantly associated with a high risk of recurrence in the gastric location [HR (95% CI)=9.820 (1.142-84.484), P=0.037; Table V], but not in non-gastric sites (data not shown).
| Table IV.Univariate and multivariate analysis of clinicopathological features and mutational status affecting the RFS of patients with GISTs (n=259). |
| Table V.Univariate and multivariate analysis of clinicopathological features and mutational status affecting the RFS of patients with gastric GISTs (n=168). |
Discussion
The identification of pathological and molecular subtypes is essential in patients with GISTs for prognostic and therapeutic purposes, particularly in the adjuvant and/or advanced disease setting. Previous studies have revealed the prognostic significance of KIT and PDGFRA mutational alterations; however, studies of GISTs in the Chinese population have only investigated small groups of patients (21,22), and the data remain limited. In the present study, the profile and prognostic value of pathological variables and gene mutations were assessed in a large cohort comprising 302 GIST cases in a single center, with a detailed description of KIT and PDGFRA mutations and the prognostic value of specific gene alterations.
Based on their morphological manifestations, GISTs are classified as three variants: Spindle, epithelioid and mixed. The present study identified spindle morphology as the predominant variant (87.4%), which is consistent with previous reports. However, the frequency of epithelioid morphology (3.0%), which was mainly observed in GISTs located in the stomach, was lower than that in previous studies (5-20%) (7,23,24). This discrepancy may be due to selection bias. In addition, tumors characterized by epithelioid and mixed histology showed a higher risk of progression than those of the spindle subtype, which may be due to the manifestation of pleomorphic/anaplastic features and high mitotic activity observed in epithelioid and mixed-histology tumors. In the present study, high mitotic activity and necrosis were also associated with a lower RFS rate, which is consistent with the study by Liu et al (25). The majority of specimens in the present study were positive for CD117, DOG1 and CD34 (95.0, 93.4 and 81.1%, respectively). Strong triple and double expression of these markers were detected in 72.5 and 24.5% of specimens, respectively. CD34 negativity was more common in tumors in non-gastric locations and with epithelioid and mixed morphology. In a study by Hashmi et al (23), CD117 and CD34 were found to be expressed in 46/48 (95.8%) and 34/46 (73.9%) of GISTs, respectively. The authors also observed that CD34 negativity was associated with epithelioid type. In a study by Liu et al (26), the CD34 positivity rate of primary GISTs was 92.3%. Therefore, the combined IHC detection of CD117, DOG1 and CD34 is helpful for GIST diagnosis. In the North American Intergroup Phase III Trial of imatinib mesylate, patients whose tumors were CD117-negative by immunostaining were found to have an inferior prognosis (13). However, no prognostic significance of CD117 expression was observed in the present study.
The frequency of gene mutations in the present study cohort was 87.1%, including 77.2% KIT mutations and 9.9% PDGFRA mutations. This frequency is in accordance with those detected in two phase III clinical trials of imatinib, conducted by the American College of Surgeons Oncology Group (ACOSOG) (87.4%) (12) and the European Organization for Research and Treatment of Cancer (86.2%) (27). The mutation rates in these trials are higher than those in a study based on the Polish Clinical GIST Registry (82.2%) (10) and a European multicenter analysis (85.1%) (11), but lower than that in a study of Chinese cases (93.8% overall, 89.1% for KIT and 4.7% for PDGFRA) conducted by Wang et al (28). KIT mutations have been reported to be associated with a high risk of progression (18,24,27). The proportion of high-risk subtypes in the single-center study conducted by Wang et al (28) was higher than that in the present series (42.5 vs. 29.1%, respectively), which explains the relatively lower frequency of KIT mutations in the present data. In one study of patients with GISTs in China, KIT mutations were identified in 76.1% of CD117-positive GIST cases (21). In another Chinese study, Du et al (22) found the frequency of KIT and PDGFRA mutations in GISTs was 76.6 and 2.8%, respectively. Thus, the frequency of PDGFRA mutations in the present study was higher than that in previous studies. In addition, the percentages of patients with tumors >10 cm in diameter and with a mitotic index >5/50 HPF were lower in the present study than in previous studies (10,12), further suggesting that relatively few cases were in a high-risk prognostic group, in contrast with the observations in the aforementioned phase III clinical trials. The spectrum of KIT and PDGFRA mutations was similar to that described in the literature (10,28). For example, the predominant mutated area of KIT exon 11 involved codons 557–560 with deletions and/or substitutions, followed by codons 572–580 with duplication. In addition to the common duplication of codons 502 and 503 of KIT exon 9, a rare duplication of codons 506–508 was detected in one case. A less common genotype with PDGFRA exon 18 deletion was observed in 5 cases (1.7%), but its significance remains unclear.
Importantly, the findings of the present study demonstrated that KIT deletions affecting codons 557/558/559 are independent adverse predictors of RFS in patients with GISTs. Patients with deletions involving codons 557/558/559 had a significantly shorter RFS, higher mitotic activity, high risk of relapse (according to the NIH classification) and high WHO grade compared with those of patients with other KIT mutations. Previous studies have reported that deletions affecting codons 557/558 are associated with metastasis and poor patient outcomes (17,29). In the Polish Clinical GIST Registry study, patients whose tumors had KIT deletions encompassing codons 557/558 had a lower 5-year RFS rate than those with other exon 11 mutations or exon 11 deletions not involving codons 557/558 (10). The study also found that deletions involving codons 557/558 were more frequently present in tumors of larger size, with a higher mitotic count and high risk of relapse. Regarding gastric GISTs, the patients with KIT deletions in codons 557/558/559 had an almost 6-fold higher risk of relapse than patients with WT KIT (30). In a European multicenter study, KIT del-inc557/558 was a predictor of inferior outcomes inpatients with gastric GISTs but not those with non-gastric GISTs (11). However, in the ACOSOG Z9001 trial, deletion of codons 557 and/or 558 did not independently affect RFS in either the placebo or imatinib arm (12). In addition, Wang et al (28) demonstrated that codon 557/558 deletion was associated with a high mitotic rate but not with 5-year RFS. In the present study, KIT deletions involving codons 557/558/559 were significantly associated with a high risk of recurrence in the stomach but not in non-gastric locations. Furthermore, this type of deletion was more frequently identified in gastric GISTs than in non-gastric cases and had an adverse effect on RFS in patients with gastric GISTs. Both the present study and previous studies identified non-gastric origin as an adverse indicator of GIST progression. Thus, we hypothesize that the adverse effect of deletions involving codons 557/558/559 may be associated with malignant features of the tumors, such as a higher mitotic count, which may result from the robust activation of KIT signaling. The loss of the side chains of amino acids encoded by codons 553, 557, 559 or 560 may be associated with increased phosphorylation. Deletions of these codons, which encode a juxta-membrane residue, may disrupt the conformation of the KIT protein and induce a loss of inhibitory control of the kinase activity in the KIT receptor. However, the detailed molecular mechanism requires further study.
The results of the present study suggest that another common molecular alteration, KIT exon 11 substitution, was a clinical predictor of an indolent tumor, namely, a tumor of small size, with a low mitotic count, low risk of relapse and low WHO grade. In the survival analysis, patients with exon 11 substitution had a significantly longer RFS than those with exon 11 deletions or WTKIT. In a Norwegian population-based study, GISTs with KIT substitutions exhibited low mitotic activity (18). In the Polish Clinical GIST Registry study, the 5-year RFS rate of patients with KIT exon 11 point mutations was improved compared with that of patients with other KIT exon 11 mutations; the study also documented that tumors with PDGFRA mutations had a lower risk of recurrence (10). In the present study, the presence of a PDGFRA exon 18 mutation appeared to have a favorable influence on GIST relapse, although the effect was not found to be statistically significant. Thus, KIT exon 11 substitution and PDGFRA exon 18 mutation may be positive prognostic indicators for disease progression.
Notably, the coexistence of exon 11 deletion and exon 13 duplication was observed in the GIST of one female patient and was characterized by adverse prognostic indicators: larger tumor size, high mitotic activity and high risk of recurrence/relapse. The treatment and prognosis of this patient after resection is unknown as she was lost to follow-up. This molecular alteration is rarely seen, and has been reported in one imatinib-resistant tumor also featuring factors indicative of an inferior prognosis (31). Moreover, the cases in the present study and previous report exhibited a similar morphology (spindle and epithelioid histology). Activating mutations of KIT exon 11, particularly deletions, result in ligand-independent activation, ultimately increasing cell proliferation and inhibiting apoptosis, and mutations in KIT exon 13, which encodes the ATP-binding region of the protein, are considered to be associated with the autoinhibitory function of the JM domain (15). Thus, we hypothesize that the combination of exon 11 deletion and exon 13 duplication may increase the malignant behavior of GIST and adversely influence patient prognosis.
The existence of tumor heterogeneity with regard to morphology, immunostaining and genotype was observed in the present study. Four cases had tumors with distinct gene mutations and histology at two different sites, and three of these cases also had distinct immunostaining phenotypes. All these distinct gene mutations occurred in KIT exon 11, including KIT exon 11 p.V560D vs. WT, p.W557R vs. p.W557_K558 del and p.D579 del vs. p.W557G. A previous report emphasized the heterogeneity of clinical resistance to tyrosine kinase inhibitors in GIST (32). Recently, Serrano et al (33) detected the presence of secondary KIT mutations in patients whose GISTs were resistant to imatinib, and suggested that these distinct mutations may have resulted from tumor subclones that emerged after imatinib therapy. In the present study, the four cases with tumor heterogeneity did not receive any therapy prior to surgery. Furthermore, no heterogeneity was observed between the primary and recurrent and/or metastatic tumors of 9 patients, 3 of whom received imatinib therapy before progression. The observation of heterogeneous morphology, immunostaining and genotype in GISTs may be due to the presence of different tumor subclones, providing further challenges for diagnosis and therapy.
In summary, the present study found that KIT and PDGFRA mutations were frequent in GISTs. KIT exon 11 deletions, particularly deletions affecting codons 557/558/559, were genotypes indicative of more aggressive tumors and were associated with a higher risk of relapse. Furthermore, the emergence of specific genotypes, such as multiple coexisting mutations and heterogeneous genetic alterations, is challenging for individual targeted therapy. Based on different pathological features and KIT/PDGFRA mutations, the results contribute to the identification of patients with different subtypes of GIST for tailored adjuvant treatments, and support the notion that specific molecular phenotypes should be included in the present risk classification system.
Supplementary Material
Supporting Data
Acknowledgements
Not applicable.
Funding
The present study was supported by a research grant (grant no. 2018SF089) from the Scientific Research Seed Fund of Peking University First Hospital.
Availability of data and materials
The datasets used and/or analyzed during the present study are available from the corresponding author on reasonable request.
Authors' contributions
TL designed the study. LL, LN and YD collected and analyzed the patient data. LL, PL and TL evaluated and interpreted the clinicopathological data and mutation phenotypes. XL was responsible for the technical operation of PCR and Sanger sequencing. DL performed the immunohistochemical staining. JL and SH performed the statistical analysis and interpreted the results. LL and TL wrote the manuscript. TL revised the manuscript. All authors have read and approved the final manuscript.
Ethics approval and consent to participate
The present study was approved by the ethics committee of Peking University First Hospital and in compliance with the Declaration of Helsinki. Written patient consent for use of their tissues in research was obtained.
Patient consent for publication
Not applicable.
Competing interests
The authors declare that they have no competing interests.
References
|
Huizinga JD, Thuneberg L, Klüppel M, Malysz J, Mikkelsen HB and Bernstein A: W/kit gene required for interstitial cells of Cajal and for intestinal pacemaker activity. Nature. 373:347–349. 1995. View Article : Google Scholar : PubMed/NCBI | |
|
Joensuu H, Hohenberger P and Corless CL: Gastrointestinal stromal tumour. Lancet. 382:973–983. 2013. View Article : Google Scholar : PubMed/NCBI | |
|
Yamamoto H and Oda Y: Gastrointestinal stromal tumor: Recent advances in pathology and genetics. Pathol Int. 65:9–18. 2015. View Article : Google Scholar : PubMed/NCBI | |
|
DeMatteo RP, Lewis JJ, Leung D, Mudan SS, Woodruff JM and Brennan MF: Two hundred gastrointestinal stromal tumors: Recurrence patterns and prognostic factors for survival. Ann Surg. 231:51–58. 2000. View Article : Google Scholar : PubMed/NCBI | |
|
Rossi S, Miceli R, Messerini L, Bearzi I, Mazzoleni G, Capella C, Arrigoni G, Sonzogni A, Sidoni A, Toffolatti L, et al: Natural history of imatinib-naive GISTs: A retrospective analysis of 929 cases with long-term follow-up and development of a survival nomogram based on mitotic index and size as continuous variables. Am J Surg Pathol. 35:1646–1456. 2011. View Article : Google Scholar : PubMed/NCBI | |
|
Joensuu H: Risk stratification of patients diagnosed with gastrointestinal stromal tumor. Hum Pathol. 39:1411–1419. 2008. View Article : Google Scholar : PubMed/NCBI | |
|
Li J, Ye Y, Wang J, Zhang B, Qin S, Shi Y, He Y, Liang X, Liu X, Zhou Y, et al: Chinese consensus guidelines for diagnosis and management of gastrointestinal stromal tumor. Chin J Cancer Res. 29:281–293. 2017. View Article : Google Scholar : PubMed/NCBI | |
|
Goh BK, Chow PK, Yap WM, Kesavan SM, Song IC, Paul PG, Ooi BS, Chung YF and Wong WK: 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. Ann Surg Oncol. 15:2153–2163. 2008. View Article : Google Scholar : PubMed/NCBI | |
|
Lokuhetty D, White VA and Watanabe R: WHO classfication of digestive system tumours. 5th edition. Lyon: IARC Press; pp. 439–443. 2019 | |
|
Wozniak A, Rutkowski P, Piskorz A, Ciwoniuk M, Osuch C, Bylina E, Sygut J, Chosia M, Rys J, Urbanczyk K, et al: Prognostic value of KIT/PDGFRA mutations in gastrointestinal stromal tumours (GIST): Polish Clinical GIST Registry experience. Ann Oncol. 23:353–360. 2012. View Article : Google Scholar : PubMed/NCBI | |
|
Wozniak A, Rutkowski P, Schöffski P, Ray-Coquard I, Hostein I, Schildhaus HU, Le Cesne A, Bylina E, Limon J, Blay JY, et al: Tumor genotype is an independent prognostic factor in primary gastrointestinal stromal tumors of gastric origin: A european multicenter analysis based on ConticaGIST. Clin Cancer Res. 20:6105–6116. 2014. View Article : Google Scholar : PubMed/NCBI | |
|
Corless CL, Ballman KV, Antonescu CR, Kolesnikova V, Maki RG, Pisters PW, Blackstein ME, Blanke CD, Demetri GD, Heinrich MC, et al: Pathologic and molecular features correlate with long-term outcome after adjuvant therapy of resected primary GI stromal tumor: The ACOSOG Z9001 trial. J Clin Oncol. 32:1563–1570. 2014. View Article : Google Scholar : PubMed/NCBI | |
|
Heinrich MC, Owzar K, Corless CL, Hollis D, Borden EC, Fletcher CD, Ryan CW, von Mehren M, Blanke CD, Rankin C, et al: Correlation of kinase genotype and clinical outcome in the North American intergroup phase III Trial of imatinib mesylate for treatment of advanced gastrointestinal stromal tumor: CALGB 150105 study by cancer and leukemia group B and southwest oncology group. J Clin Oncol. 26:5360–5367. 2008. View Article : Google Scholar : PubMed/NCBI | |
|
Szucs Z, Thway K, Fisher C, Bulusu R, Constantinidou A, Benson C, van der Graaf WT and Jones RL: Molecular subtypes of gastrointestinal stromal tumors and their prognostic and therapeutic implications. Future Oncol. 13:93–107. 2017. View Article : Google Scholar : PubMed/NCBI | |
|
Lennartsson J, Jelacic T, Linnekin D and Shivakrupa R: Normal and oncogenic forms of the receptor tyrosine kinase kit. Stem Cells. 23:16–43. 2005. View Article : Google Scholar : PubMed/NCBI | |
|
Martin-Broto J, Gutierrez A, Garcia-Del-Muro X, Lopez-Guerrero JA, Martinez-Trufero J, de Sande LM, Lainez N, Maurel J, De Juan A, Losa F, et al: Prognostic time dependence of deletions affecting codons 557 and/or 558 of KIT gene for relapse-free survival (RFS) in localized GIST: A Spanish group for sarcoma research (GEIS) study. Ann Oncol. 21:1552–1557. 2010. View Article : Google Scholar : PubMed/NCBI | |
|
Martín J, Poveda A, Llombart-Bosch A, Ramos R, López-Guerrero JA, García del Muro J, Maurel J, Calabuig S, Gutierrez A, González de Sande JL, et al: Deletions affecting codons 557–558 of the c-KIT gene indicate a poor prognosis in patients with completely resected gastrointestinal stromal tumors: A study by the Spanish group for sarcoma research (GEIS). J Clin Oncol. 23:6190–6198. 2005. View Article : Google Scholar : PubMed/NCBI | |
|
Steigen SE, Eide TJ, Wasag B, Lasota J, Lasota J and Miettinen M: Mutations in gastrointestinal stromal tumors-a population-based study from Northern Norway. APMIS. 115:289–298. 2007. View Article : Google Scholar : PubMed/NCBI | |
|
Gill AJ: Succinate dehydrogenase (SDH) and mitochondrial driven neoplasia. Pathology. 44:285–292. 2012. View Article : Google Scholar : PubMed/NCBI | |
|
Miettinen M, Wang ZF, Sarlomo-Rikala M, Osuch C, Rutkowski P and Lasota J: Succinate dehydrogenase-deficient GISTs: A clinicopathologic, immunohistochemical, and molecular genetic study of 66 gastric GISTs with predilection to young age. Am J Surg Pathol. 35:1712–1721. 2011. View Article : Google Scholar : PubMed/NCBI | |
|
He HY, Fang WG, Zhong HH, Li Y, Zheng J, Du J, Heng WJ and Wu BQ: Status and clinical implication of c-kit and PDGFRA mutations in 165 cases of gastrointestinal stromal tumor (GIST). Zhonghua Bing Li Xue Za Zhi. 35:262–266. 2006.(In Chinese). PubMed/NCBI | |
|
Du CY, Shi YQ, Zhou Y, Fu H and Zhao G: The analysis of status and clinical implication of KIT and PDGFRA mutations in gastrointestinal stromal tumor (GIST). J Surg Oncol. 98:175–178. 2008. View Article : Google Scholar : PubMed/NCBI | |
|
Hashmi AA, Faraz M, Nauman Z, Qureshi MU, Hashmi SK, Waseem HF, Edhi MM, Faridi N and Khan A: Clinicopathologic features and prognostic grouping of gastrointestinal stromal tumors (GISTs) in Pakistani patients: An institutional perspective. BMC Res Notes. 11:4572018. View Article : Google Scholar : PubMed/NCBI | |
|
Mayr P, Märkl B, Agaimy A, Kriening B, Dintner S, Schenkirsch G and Schneider-Stock R: Malignancies associated with GIST: A retrospective study with molecular analysis of KIT and PDGFRA. Langenbecks Arch Surg. 404:605–613. 2019. View Article : Google Scholar : PubMed/NCBI | |
|
Liu X, Qiu H, Zhang P, Feng X, Chen T, Li Y, Tao K, Li G, Sun X and Zhou Z; China Gastrointestinal Stromal Tumor Study Group (CN-GIST), : Prognostic role of tumor necrosis in patients undergoing curative resection for gastric gastrointestinal stromal tumor: A multicenter analysis of 740 cases in China. Cancer Med. 6:2796–2803. 2017. View Article : Google Scholar : PubMed/NCBI | |
|
Liu FY, Qi JP, Xu FL and Wu AP: Clinicopathological and immunohistochemical analysis of gastrointestinal stromal tumor. World J Gastroenterol. 12:4161–4165. 2006. View Article : Google Scholar : PubMed/NCBI | |
|
Debiec-Rychter M, Sciot R, Le Cesne A, Schlemmer M, Hohenberger P, van Oosterom AT, Blay JY, Leyvraz S, Stul M, Casali PG, et al: KIT mutations and dose selection for imatinib in patients with advanced gastrointestinal stromal tumours. Eur J Cancer. 42:1093–1103. 2006. View Article : Google Scholar : PubMed/NCBI | |
|
Wang M, Xu J, Zhao W, Tu L, Qiu W, Wang C, Shen Y, Liu Q and Cao H: Prognostic value of mutational characteristics in gastrointestinal stromal tumors: A single-center experience in 275 cases. Med Oncol. 31:8192014. View Article : Google Scholar : PubMed/NCBI | |
|
Wardelmann E, Losen I, Hans V, Neidt I, Speidel N, Bierhoff E, Heinicke T, Pietsch T, Büttner R and Merkelbach-Bruse S: Deletion of Trp-557 and Lys-558 in the juxtamembrane domain of the c-kit protooncogene is associated with metastatic behavior of gastrointestinal stromal tumors. Int J Cancer. 106:887–895. 2003. View Article : Google Scholar : PubMed/NCBI | |
|
Capelli L, Petracci E, Quagliuolo V, Saragoni L, Colombo P, Morgagni P, Calistri D, Tomezzoli A, Di Cosmo M, Roviello F, et al: Gastric GISTs: Analysis of c-Kit, PDGFRA and BRAF mutations in relation to prognosis and clinical pathological characteristics of patients-A GIRCG study. Eur J Surg Oncol. 42:1206–1214. 2016. View Article : Google Scholar : PubMed/NCBI | |
|
Antonescu CR, Romeo S, Zhang L, Nafa K, Hornick JL, Nielsen GP, Mino-Kenudson M, Huang HY, Mosquera JM, Dei Tos PA and Fletcher CD: Dedifferentiation in gastrointestinal stromal tumor to an anaplastic KIT-negative phenotype: A diagnostic pitfall: Morphologic and molecular characterization of 8 cases occurring either de novo or after imatinib therapy. Am J Surg Pathol. 37:385–392. 2013. View Article : Google Scholar : PubMed/NCBI | |
|
Liegl B, Kepten I, Le C, Zhu M, Demetri GD, Heinrich MC, Fletcher CD, Corless CL and Fletcher JA: Heterogeneity of kinase inhibitor resistance mechanisms in GIST. J Pathol. 216:64–74. 2008. View Article : Google Scholar : PubMed/NCBI | |
|
Serrano C, Mariño-Enríquez A, Tao DL, Ketzer J, Eilers G, Zhu M, Yu C, Mannan AM, Rubin BP, Demetri GD, et al: Complementary activity of tyrosine kinase inhibitors against secondary kit mutations in imatinib-resistant gastrointestinal stromal tumours. Br J Cancer. 120:612–620. 2019. View Article : Google Scholar : PubMed/NCBI |
