Gastric submucosal tumors (SMTs), a rare disease, are often found incidentally during gastric surgery [
Although biopsies generally provide a histological diagnosis and thus facilitate planning of the treatment strategy, it is sometimes difficult to make a pathological diagnosis of gastric SMTs because they are covered by normal gastric mucosa. In addition, biopsies of gastric GISTs may cause tumor rupture and dissemination [
Some new biopsy techniques [
The aim of this retrospective study was to investigate the efficacy and safety of biopsies for gastric GIST.
All patients who had undergone surgical resection and pathologically diagnosed gastric GIST after surgery, at the Digestive Disease Center, Showa University Northern Yokohama Hospital between April 2001 and March 2012 were retrospectively studied.
The inclusion criteria were: (i) solitary submucosal tumor, (ii) pretreatment endoscopic biopsy performed, (iii) no prior treatment by endoscopic resection, surgery, chemotherapy, or radiation therapy. The exclusion criterion was synchronous malignancy.
The accuracy of the preoperative histological diagnoses and any complications of the pretreatment biopsies were assessed. Patient clinical records and pathology reports were reviewed to identify the clinical characteristics of the tumors as assessed by endoscopy and CT, biopsy techniques, and pathological diagnoses of the biopsy and surgical specimens.
The formalin-fixed biopsy and surgical specimens were prepared for diagnosis as follows. Hematoxylin and eosin, and immunohistochemical staining using antibodies of CD117 (Dako-Japan, Tokyo, Japan), CD34 (QBEnd10, Dako-Japan), alpha smooth muscle actin (
The
The study was approved by the Institutional Review Board of the Showa University, Northern Yokohama Hospital (no. 1203-01). The research reported in this paper was in compliance with the Helsinki Declaration. This study was registered with the University Hospital Medical Information Network in Japan (no. UMIN000007428).
Altogether, 23 patients (12 males and 11 females, mean age 60.0 years) were eligible (Figure
Patients’ clinicopathological characteristics (
Variables | Number of subjects |
---|---|
Sex | |
Male | 12 |
Female | 11 |
Age (years, [mean, range]) | 60.0, 26–92 |
Main tumor site | |
Upper third of stomach | 16 |
Middle third of stomach | 6 |
Lower third of stomach | 1 |
Clinically determined tumor diameter |
36.6, 18–60 |
Ulceration | |
Present | 12 |
Absent | 11 |
Number of endoscopies (mean, range) | 2.7, 1–8 |
Endoscopic ultrasound | |
Yes | 15 |
No | 8 |
Number of biopsies/patient (mean, range) | 1.4, 1–5 |
Fine-needle aspiration biopsy | |
Yes | 12 |
No | 11 |
Preoperative pathological diagnosis | |
Gastrointestinal stromal tumor | 18 |
Leiomyoma | 1 |
Normal gastric mucosa | 4 |
Pathologically determined tumor diameter |
45.4, 16–110 |
Gastrectomy | |
Partial | 20 |
Proximal | 2 |
Distal | 1 |
GIST risk group | |
Low | 13 |
Intermediate | 5 |
High | 5 |
Univariate analysis (34 biopsies).
Variables | Accuracy of pathological diagnosis (correct/total) |
|
---|---|---|
Sex | 0.510 | |
Male | 9/15 (60.0%) | |
Female | 9/19 (47.3%) | |
Age (years) | 0.510 | |
≤60 | 9/15 (60.0%) | |
>60 | 9/19 (47.3%) | |
Clinically determined tumor diameter (mm) | 1.000 | |
≤30 | 7/14 (50.0%) | |
>30 | 11/20 (55.0%) | |
Ulceration | 0.327 | |
Present | 10/16 (62.5%) | |
Absent | 8/18 (44.4%) | |
Number of endoscopy | 1.000 | |
First | 8/16 (50.0%) | |
Second or later | 10/18 (55.6%) | |
Endoscopic ultrasound | 0.001** | |
Yes | 11/12 (91.7%) | |
No | 7/22 (31.8%) | |
Number of biopsy | 0.274 | |
First | 14/23 (60.9%) | |
Second or later | 4/11 (36.4%) | |
Biopsy procedure | 0.005** | |
Normal | 7/21 (33.3%) | |
Fine-needle aspiration biopsy | 11/13 (84.6%) |
**
Enrollment of patients included in the study. Total 47 patients had undergone surgical resection for gastric submucosal tumors at the Digestive Disease Center, Showa University Northern Yokohama Hospital between April 2001 and March 2012 were retrospectively studied. The 23 patients pathologically diagnosed GIST after surgery were eligible.
The pathological diagnoses on biopsy and surgical specimens were the same for 18 (52.9%) of the 34 biopsies. Univariate analysis showed that agreement between pathological diagnoses on biopsy and surgical specimens was statistically significant for EUS and FNA biopsies (
In a multivariate analysis, only EUS was a significant factor (odds ratio, 11.884; 95% confidence interval, 1.204–289.230;
Multivariate analysis.
Variable | Odds ratio | 95% confidence interval |
|
---|---|---|---|
Endoscopic ultrasound | |||
No | 1.000 | 1.204–289.230 | 0.034* |
Yes | 11.884 | ||
Biopsy procedure | |||
Forceps biopsy | 1.000 | 0.316–30.964 | 0.312 |
Fine-needle aspiration biopsy | 3.102 |
*
To assess the relationship between EUS and FNA biopsies, all 34 biopsies were categorized by the presence of EUS and FNAB as follows: EUS-guided FNA biopsy (EUS+/FNAB, 10 biopsies), forceps biopsy with EUS (EUS+/forceps, 2 biopsies), FNA biopsy without EUS (EUS−/FNAB, 3 biopsies), and forceps biopsy without EUS (EUS−/forceps, 19 biopsies). The accuracy of pathological diagnoses of biopsy specimens by EUS+/FNAB, EUS+/forceps, EUS−/FNAB, and EUS−/forceps was 100%, 50.0%, 33.3%, and 31.6%, respectively. There were significant differences between these four groups according to Pearson’s test (
Of the 23 patients, two in whom forceps biopsies had been performed developed recurrent disease. One of these patient developed liver metastases 15 months after surgery, the other developed liver metastases and peritoneal dissemination 22 months after surgery. No patients in whom FNA biopsies had been performed had disease recurrences.
Gastrointestinal stromal tumors are the commonest mesenchymal subepithelial tumors of the gastrointestinal tract. They originate from the interstitial cells of Cajal. Thus, they are mainly located in the submucosal to muscular layer of the digestive tract and present as submucosal tumors. Because GISTs smaller than 2 cm are rarely associated with symptoms, small ones are usually detected incidentally during abdominal surgery or radiological examinations [
The treatment strategies for neoplastic tumors are generally based on their histological characteristics. Of the gastric submucosal tumors, leiomyomas, neurilemmomas (schwannomas), and other benign tumors do not generally require any treatment; potentially malignant tumors such as GISTs, leiomyosarcomas, and carcinomas do require treatment because of the possibility that they will invade and metastasize. Because submucosal tumors are covered by normal mucosa, obtaining a specimen of tumor tissue by endoscopic biopsy is often difficult. Accordingly, various biopsy techniques have been developed [
Accordingly, diagnosis and treatment strategies for GIST have been mainly determined based on the findings of radiological investigations such as CT [
Endoscopic ultrasound-guided FNA biopsy is reportedly useful for diagnosing submucosal tumors [
In this study, univariate analysis showed that EUS and FNA biopsies were predictors for accurate pretreatment diagnosis of gastric GISTs. Multivariate analysis showed that EUS was the only significant factor. Although this difference was not statistically significant, FNA biopsies were more accurate than forceps biopsies. Tumor-related characteristics such as tumor size and presence of ulceration did not correlate significantly with accurate pretreatment diagnoses. These results suggest that EUS and FNA biopsy procedures are capable of obtaining adequate samples of tumor tissue for pretreatment pathological diagnosis regardless of the tumor characteristics.
We attempted to determine whether EUS or FNA biopsies are preferable for the diagnosis of GISTs. Of 13 FNA biopsies, 10 were performed using EUS (EUS-guided FNA biopsy) and three without EUS. All 10 EUS-guided FNA biopsy specimens were accurately diagnosed as GIST. In contrast, the accuracy of diagnostics samples obtained by forceps and FNA biopsies without using EUS was 31.6% and 33.3% (Figure
Consistency of pathological diagnosis between biopsy and surgical specimens. All 34 biopsies were categorized by presence of EUS and FNAB as EUS-guided FNA biopsy (EUS+/FNAB, 10 biopsies), forceps biopsy with EUS (EUS+/forceps, 2 biopsies), FNA biopsy without EUS (EUS−/FNAB, 3 biopsies), and forceps biopsy without EUS (EUS−/forceps, 19 biopsies). The EUS+/FNAB group had significantly better diagnostic accuracy than the EUS−/FNAB and EUS−/forceps groups (
We experienced no major complications in 34 biopsies, including FNA biopsies. Additionally, no patient who underwent FNA biopsy has developed disease recurrence.
Accordingly, we have concluded that EUS-guided FNA biopsy is useful for pretreatment pathological diagnosis of gastric GIST and prevention both of early complications and late idiopathic disease recurrence. We are now working on a study with a larger sample size to confirm the efficacy of EUS-guided FNA biopsy for diagnosis of submucosal tumors of the digestive tract.
In this study, EUS-guided FNA biopsies provided extremely accurate pathological diagnoses and were associated with no major complications or disease recurrence. The diagnostic accuracy of FNA biopsies without EUS is not good. The results of this small sample size study suggest that EUS-guided FNA biopsy is safe and useful for obtaining an accurate pretreatment pathological diagnosis of gastric GISTs.
Alpha smooth muscle actin.
Cluster of differentiation.
Computed tomography.
Endoscopic ultrasound.
Fine-needle aspiration.
Fine-needle aspiration biopsy.
Magnetic resonance imaging.
Positron emission tomography.
Smooth muscle actin.
Submucosal tumor.
The authors declare they have no conflict of interests.
H. Ito conceived and designed the study, collected clinical and pathological data, and performed the statistical analysis and interpretation of data. H. Inoue performed medical examinations, collected biopsy samples, performed surgical operation and performed interpretation of data. S. Ryozawa, H. Ikeda, N. Odaka performed medical examinations and surgical operation. N. Eleftheriadis, R. Maselli, and N. Sando participated in the study design and performed interpretation of data. S. Kimura delivered the patients’ clinical data. S. Kudo participated in the study design and coordination. All authors have read and approved the final paper.
The authors are extremely grateful to all the patients and to the clinical staff who cared for these patients. they also are thankful Dr. Shigeharu Hamatani for his reliable pathological diagnoses.