Original article
Non-neoplastic pathology results after endoscopic submucosal dissection for gastric epithelial dysplasia or early gastric cancer
Authors
Min Jae Yang1, Sung Jae Shin1, Ki Seong Lee1, Kee Myung Lee1, Sun Gyo Lim1, Joon Koo Kang1, Jae Chul Hwang1, Soon Sun Kim1, Dakeun Lee2, Joo-Sung Kim1, Gil-Ho Lee1, Han Seok Ryu1, Byung Moo Yoo1, Kwang Jae Lee1, Young Bae Kim2, Jin Hong Kim1
Institutions
1 2
submitted 5. August 2014 accepted after revision 27. November 2014
Bibliography DOI http://dx.doi.org/ 10.1055/s-0034-1391375 Published online: 4.2.2015 Endoscopy 2015; 47: 598–604 © Georg Thieme Verlag KG Stuttgart · New York ISSN 0013-726X Corresponding author Sung Jae Shin, MD, PhD Department of Gastroenterology Ajou University School of Medicine 164 Worldcup-ro Yeongtong-gu, Suwon-si Gyeonggi-do, 443–380 South Korea Fax: 82–31–2195999 [email protected]
Department of Gastroenterology, Ajou University School of Medicine, Suwon, South Korea Department of Pathology, Ajou University School of Medicine, Suwon, South Korea
Background and study aims: Endoscopists sometimes face paradoxical cases in which the endoscopic submucosal dissection (ESD) specimen reveals a non-neoplastic pathology result. The aims of the study were to determine the reasons for such results, and to compare the endoscopic characteristics of non-neoplastic and conventional neoplastic pathology groups after ESD. Patients and methods: A total of 1186 gastric ESDs performed between February 2005 and December 2011 were retrospectively reviewed. The ESD specimens included 52 (4.4 %) that were confirmed as negative or indefinite for neoplasia. Patient characteristics and endoscopic and pathological data were reviewed and compared. Results: Non-neoplastic pathology after ESD was due to complete removal of the lesion at biopsy in 45 cases (86.5 %), pathology overestimation in 5 (9.6 %), and incorrect localization of the original tumor with subsequent ESD performed at the
wrong site in 2 (3.8 %). The mean length and surface area of the non-neoplastic lesions were 9.2 ± 2.6 mm and 49.6 ± 23.6 mm 2, respectively. Mean sampling ratios were 3.0 ± 1.5 mm/fragment and 16.3 ± 10.0 mm2/fragment. Compared with 1134 cases confirmed as neoplastic on the final ESD specimen, non-neoplastic cases showed a significantly smaller tumor size and surface area, and lower sampling ratios in a logistic regression analysis adjusted for potential confounders (P < 0.001 for all). Conclusions: Complete lesion removal by biopsy, pathology overestimation, and incorrect localization of the original tumor with subsequent ESD at the wrong site were the main reasons for nonneoplastic results after ESD. Small tumor size and surface area, and low sampling ratios were associated with non-neoplastic pathology results after ESD.
Introduction
therapeutic plan, and worse clinical outcome than expected [3, 7 – 9]. However, endoscopists have faced paradoxical cases in which the endoscopic resection specimen revealed a downgraded final pathology, especially a non-neoplastic result, albeit infrequently, with reported rates of 3.2 % – 7.0 % [5, 7,10]. Until recently, few studies have been conducted on non-neoplastic pathology results after endoscopic resection, especially ESD. The aims of this study were to analyze the reasons for non-neoplastic pathology results on ESD specimens, and to compare the endoscopic characteristics of non-neoplastic and conventional neoplastic pathology groups after ESD.
!
Endoscopic submucosal dissection (ESD) is the current standard of care for selected cases of gastric epithelial dysplasia or early gastric cancer because of its curative en bloc resectability and complete histopathological assessment of the specimen [1, 2]. However, in many cases, the initial pathological diagnosis changes after ESD because the biopsy specimen does not reflect the entire histological background of the lesion. Histological discrepancies between endoscopic forceps biopsy (EFB) and endoscopic resection specimens, including ESD and endoscopic mucosal resection, have been reported at rates of 2.7 % – 49 % [3 – 6]. Most studies on pathological discrepancies after endoscopic resection have concentrated on the difference between concordant and upgraded histological changes that imply initial underestimation of the dysplasia grade, modification of the
Yang Min Jae et al. Non-neoplastic pathology results after endoscopic submucosal dissection … Endoscopy 2015; 47: 598–604
Downloaded by: UPSTATE Medical University. Copyrighted material.
598
Original article
Table 1 [11].
!
Study population Between February 2005 and December 2011, 1186 consecutive patients underwent gastric ESD at Ajou University Hospital in Suwon, South Korea. These cases were retrospectively reviewed from a prospectively designed ESD database system, and included 52 patients (4.4 %) who were confirmed to have negative or indefinite neoplastic results on the ESD specimen (category 1 or 2 according to the revised Vienna classification of gastrointestinal " Table 1). neoplasia [11]) (● The study was approved by the institutional review board (approval no. AJIRB-MED-MDB-13-248), and informed consent was obtained from each patient for the ESD.
Endoscopic protocol before ESD Diagnostic endoscopy was performed in patients who were referred to the Ajou University Hospital for evaluation of upper gastrointestinal symptoms without a preceding endoscopy. If epithelial dysplasia or cancer was suspected, neoplasia-targeted EFB was performed using a narrow-band imaging system and chromoendoscopy with indigo carmine dye spray. Before EFB, the size of the lesion was estimated macroscopically using standard biopsy forceps with a 6-mm opening diameter (FB-21K-1; Olympus, Tokyo, Japan) and 6-mm circular pieces of punched out paper. The number of biopsy fragments was determined at the discretion of the endoscopist, after considering the size and surface configuration of the lesion. Patients who were pathologically diagnosed at private clinics and then referred to the University Hospital with a pathology slide did not undergo routine re-endoscopy with neoplasia-targeted biopsy because the additional cost and potential risk of biopsyinduced fibrosis, which might compromise the safe completion of ESD, outweighed the potential benefits of additional pathological information, which only rarely modifies the treatment strategy. As a result, all lesions in the study population were pathologically confirmed as gastric epithelial dysplasia or cancer via preESD EFB performed either at the University Hospital or at private clinics.
ESD procedure Indications for ESD included lesions with pathological findings of either gastric epithelial dysplasia based on the EFB, irrespective of size or surface configuration, or early gastric cancer that met the expanded criteria of the Japanese Gastric Cancer Association [12]. All ESDs were performed using a standard endoscope with a single channel (GIF-H260 or GIF-Q260J; Olympus) under conscious sedation with both midazolam and propofol. Standard ESD techniques was performed with an insulated-tip knife (Olympus) and/or a FlexKnife (Olympus) in the dry cut or swift coagulation mode (Erbotom ICC200 or VIO 300D; ERBE, Tübingen, Germany). Periodic surveillance endoscopies were scheduled for 3, 6, 12, 18, and 24 months after ESD, and annually thereafter. EFB was carried out in the area surrounding the post-ESD scar and at other sites if needed at every follow-up endoscopy.
Revised Vienna classification of gastrointestinal epithelial neoplasia
Category
Diagnosis
1
Negative for neoplasia
2
Indefinite for neoplasia
3
Mucosal low grade neoplasia Low grade adenoma or dysplasia
4
Mucosal high grade neoplasia 4 – 1 High grade adenoma or dysplasia 4 – 2 Noninvasive carcinoma (carcinoma in situ) 4 – 3 Suspicious for invasive carcinoma 4 – 4 Intramucosal carcinoma (lamina propria invasion)
5
Submucosal invasion by carcinoma
number of biopsy fragments, is a reflection of the extent of biopsy sampling, with a lower ratio reflecting more extensive sampling [13].
Histological features Two experienced gastrointestinal pathologists (Y.B.K., D.L.) confirmed the histological diagnoses of the pre-ESD biopsies and ESD specimens. The histological diagnosis was made according to the revised Vienna classification of gastrointestinal neoplasia [11]. When no tumor was found in the ESD specimen, all slides from pre-ESD biopsies and ESD specimens were peer evaluated or re-evaluated by the pathologist who provided the original pre-ESD diagnosis in order to confirm that the entire lesion had been removed during biopsy or to confirm initial pathological misdiagnosis. In patients who were referred from private clinics with a pathology slide, one of the two pathologists re-evaluated the pre-ESD slide and the ESD specimen. Involvement of atrophic gastritis and intestinal metaplasia was assessed histologically, and the presence of Helicobacter pylori infection was confirmed on the basis of the results of the EFB and rapid urease test or urea breath test.
Procedure-related features Procedure time, procedure-related adverse events (perforation or delayed bleeding), and the presence of submucosal fibrosis were assessed. Submucosal fibrosis was diagnosed when the submucosal layer could not be easily separated from the whitish muscular area, which appears without a blue transparent layer, during lesion dissection (consistent with F1 or F2 fibrosis, as suggested by Higashimaya et al. [14]).
Statistical analyses Statistical analyses were performed using SPSS software version 18.0 (SPSS Inc., Chicago, Illinois, USA). The significance of differences for quantitative data was determined using unpaired Student’s t test and was presented as mean ± SD. Categorical data were compared using the chi-squared test and Fisher’s exact test. Binary logistic regression analyses were applied to identify significant endoscopic predictors of a non-neoplastic pathology result after ESD. All statistical tests were conducted with a 2-sided significance level of 0.05.
Endoscopic features The following endoscopic variables were analyzed and compared: tumor size and location, dominant morphology type, number of biopsy fragments, and sampling ratios. Sampling ratio, calculated by dividing the length of the tumor or surface area by the
Yang Min Jae et al. Non-neoplastic pathology results after endoscopic submucosal dissection … Endoscopy 2015; 47: 598–604
Downloaded by: UPSTATE Medical University. Copyrighted material.
Patients and methods
599
Original article
Table 2 Distribution of the 1186 patients according to the endoscopic forceps biopsy and endoscopic submucosal dissection pathology.
ESD pathology
EFB pathology LGD
Total
HGD
Cancer
Category 1
28
2
8
38
Category 2
11
1
2
14
LGD
468
25
4
497
HGD
61
51
2
114
Cancer
135
92
296
523
Total
703
171
312
1186
ESD (n = 1186)
Concordant (n = 815) 68.7 %
Discordant (n = 371) 31.3 %
Down-graded (n = 83) 7%
Up-graded (n = 288) 24.3 %
Non-neoplastic result (n = 52) 4.4 %
Fig. 1
Flow chart of the enrolled study population.
EFB, endoscopic forceps biopsy; ESD, endoscopic submucosal dissection; LGD, low grade dysplasia; HGD, high grade dysplasia.
Table 3 Clinical characteristics of patients with non-neoplastic pathology results after endoscopic submucosal dissection. Non-neoplasia cases after ESD, n/N (%)
52/1186 (4.4)
Sex, n (%) Male
33 (63.5)
Female
19 (36.5)
Age, mean ± SD (range), years
59.6 ± 12.2 (35 – 81)
Follow-up period, mean ± SD (range), months
20.3 ± 18.7 (1 – 75)
Forceps biopsy diagnosis 1, n (%) Low grade dysplasia High grade dysplasia Carcinoma
39 (75.0) 3 (5.8) 10 (19.2)
Post-ESD diagnosis 1, n (%) Category 1
38 (73.1)
Category 2
14 (26.9)
Coexistent pathology on ESD specimen, n (%) Helicobacter gastritis
28 (53.8)
Atrophic gastritis
43 (82.7)
Intestinal metaplasia
48 (92.3)
Procedure-related factors ESD time, mean ± SD (range), minutes
47.5 ± 27.8 (12 – 129)
Perforation, n (%)
2 (3.8)
Delayed bleeding, n (%)
3 (5.8)
Submucosal fibrosis during ESD, n (%)
8 (15.4)
ESD, endoscopic submucosal dissection. 1 Revised Vienna classification [11].
Table 4 Reasons for non-neoplastic pathology results after endoscopic submucosal dissection (n = 52). Complete biopsy removal, n (%)
45 (86.5)
Pathology overestimation, n (%)
5 (9.6)
Incorrect localization of ESD site, n (%)
2 (3.8)
ESD, endoscopic submucosal dissection.
Results !
The distribution of the 1186 patients according to EFB and ESD " Table 2. The overall histological dispathology is presented in ● crepancy rate between EFB and ESD specimens was 31.3 % (371/ 1186). A downgrade of the pathological diagnosis was found in 83 cases (7.0 %). In 52 patients (4.4 %), a non-neoplastic result on " Fig. 1). the ESD specimen was found (● The clinical characteristics of the patients with non-neoplastic " Table 3. The initial pathoresults after ESD are summarized in ● logical diagnoses based on EFB samplings were 39 low grade dysplasias (LGDs) (75.0 %), 3 high grade dysplasias (5.8 %), and 10 adenocarcinomas (19.2 %). The final histological diagnoses based
on ESD specimens were 38 category 1 cases (negative for neoplasia, 73.1 %) and 14 category 2 cases (indefinite for neoplasia, 26.9 %). Intestinal metaplasia, atrophic gastritis, and H. pylori infection were observed in the surrounding gastric mucosa in 48 (92.3 %), 43 (82.7 %), and 28 (53.8 %) cases, respectively. Eight patients (15.4 %) with non-neoplastic pathology results after ESD showed submucosal fibrosis during ESD, one of whom experienced ESD-induced perforation. After pathological re-evaluation, the EFB-based initial pathological diagnosis in five cases from private clinics was changed from category 3.1 (LGD) to category 2 (indefinite for neoplasia). The final ESD diagnoses of these cases were all consistent with category 2. In one patient with adenocarcinoma on EFB samples and a nonneoplastic result on the ESD specimen, cancerous tissue was found in the follow-up endoscopic biopsy performed 3 months after ESD. This indicated incorrect localization of the original tumor and subsequent ESD at the wrong site as a result of confusing nodular changes in the surrounding mucosa, which were due to " Fig. 2). This patient subsequently underintestinal metaplasia (● went total gastrectomy, as it was not possible to perform complete re-ESD because of severe fibrosis and bleeding. In another patient with LGD on EFB specimens and a non-neoplastic pathology on the ESD specimen, a polypoid lesion was found close to the margin of the post-ESD scar at the follow-up endoscopy performed 12 months after ESD. This also indicated initial incorrect localization of the original tumor and subsequent ESD at the wrong site, this time as a result of morphological alteration fol" Fig. 3). In summary, the lowing pre-ESD multiple biopsies (● non-neoplastic ESD pathology was considered to be due to complete removal of the lesion by EFB in 45 patients (86.5 %), pathology overestimation in 5 (9.6 %), and incorrect localization of the original tumor with subsequent ESD at the wrong site in 2 (3.8 %) " Table 4). A representative case of complete biopsy removal is (● " Fig. 4. shown in● The baseline clinicopathological and endoscopic characteristics of the non-neoplastic (n = 52) and conventional neoplastic (n = " Table 5. The 1134) pathology groups after ESD are shown in ● mean length, width, and surface area of the non-neoplastic lesions were 9.2 ± 2.6 mm, 6.4 ± 1.8 mm, and 49.6 ± 23.6 mm2, respectively. The mean number of EFB fragments was 3.5 ± 1.5, and the mean sampling ratios were 3.0 ± 1.5 mm/fragment and 16.3 ± 10.0 mm2/fragment. Compared with patients with conventional neoplastic pathology after ESD, patients with non-neoplastic pathology showed a significantly smaller tumor size and " Table 5). These differsurface area, and lower sampling ratios (● ences were still statistically significant in logistic regression analyses after adjusting for clinical and endoscopic confounders (all " Table 6). Although the univariate analyses showed P < 0.001) (●
Yang Min Jae et al. Non-neoplastic pathology results after endoscopic submucosal dissection … Endoscopy 2015; 47: 598–604
Downloaded by: UPSTATE Medical University. Copyrighted material.
600
Original article
601
Fig. 3 Incorrect localization of the original tumor and subsequent endoscopic submucosal dissection (ESD) at the wrong site as a result of morphological alteration following pre-ESD forceps biopsy. a A slightly elevated round lesion of the lesser curvature of the proximal antrum (black arrows) was confirmed as low grade dysplasia on the pre-ESD forceps biopsy. b The dysplastic lesion was not clearly defined at the time of ESD. c The post-ESD scarring found at 3 months after ESD. d Endoscopy 12 months later, showing a polypoid lesion at the margin of the post-ESD scar (arrows).
Yang Min Jae et al. Non-neoplastic pathology results after endoscopic submucosal dissection … Endoscopy 2015; 47: 598–604
Downloaded by: UPSTATE Medical University. Copyrighted material.
Fig. 2 Incorrect localization of the original tumor and therefore subsequent endoscopic submucosal dissection (ESD) at the wrong site as a result of confusing nodular changes in the surrounding mucosa. a, b Endoscopic images acquired during endoscopic forceps biopsy, showing a slightly elevated nodular lesion with central depression of the mid body posterior wall. c Endoscopic image acquired during ESD, showing a cancerous lesion (black arrow), and surrounding intestinal metaplasia that was mistaken for a cancerous lesion (red circle). d ESD of the incorrect site.
Original article
Fig. 4 Complete removal of a small neoplasia during endoscopic forceps biopsy (EFB). a Arrows indicate a 4-mm superficial elevated round lesion on the posterior wall of the lower body. b EFB. c PostEFB image acquired during endoscopic submucosal dissection (ESD). d Submucosal fibrosis confirmed during ESD.
Age, mean ± SD, years
Non-neoplasia group
Conventional group
(n = 52)
(n = 1134)
59.6 ± 12.2
62.4 ± 10.3
Male
33 (63.5)
827 (72.9)
Female
19 (36.5)
307 (27.1)
Upper
4 (7.7)
64 (5.6)
Middle
22 (42.3)
377 (33.2)
Lower
26 (50.0)
693 (61.1)
Protruded
31 (59.6)
786 (69.3)
Flat
15 (28.8)
172 (15.2)
6 (11.5)
176 (15.5)
39 (75.0)
664 (58.6)
Longitudinal location 2, n (%)
0.274
0.029
Dominant morphology, n (%)
0.047
EFB pathology, n (%) Low grade dysplasia High grade dysplasia Cancer
3 (5.8)
168 (14.8)
10 (19.2)
302 (26.6)
Number of biopsy fragments, mean ± SD (range)
3.5 ± 1.5 (1 – 6)
Tumor size, mean length ± SD (range), mm
9.2 ± 2.6 (4 – 16)
Surface area, mean ± SD (range), mm 2
0.119 0.135
Sex, n (%)
Depressed
P value1
49.6 ± 23.6 (9.4 – 109.9)
3.6 ± 1.9 (1 – 9)
0.622
19.4 ± 9.1 (5 – 48)
< 0.001
281.5 ± 265.2 (7.9 – 1413)
< 0.001
Sampling ratio 3 mm/fragment
3.0 ± 1.5
6.7 ± 4.1
< 0.001
mm 2/fragment
16.3 ± 10.0
89.7 ± 94.3
< 0.001
EFB, endoscopic forceps biopsy. 1 Univariate analysis, not adjusted by other variables. 2 Upper (cardia, fundus, upper body), middle (mid body, lower body, angle), and lower (antrum, prepylorus). 3 Calculated by dividing the length of the tumor or surface area by the number of biopsy fragments. It reflects the extent of biopsy sampling, with a lower ratio reflecting more extensive sampling.
Yang Min Jae et al. Non-neoplastic pathology results after endoscopic submucosal dissection … Endoscopy 2015; 47: 598–604
Table 5 Comparison of baseline clinicopathological and endoscopic features between non-neoplastic and conventional neoplastic pathology groups after endoscopic submucosal dissection.
Downloaded by: UPSTATE Medical University. Copyrighted material.
602
Table 6 Logistic regression analyses of endoscopic predictors for non-neoplastic pathology results after endoscopic submucosal dissection, adjusted for clinical and endoscopic factors.
Adjusted OR [95 %CI]
P value
Tumor size (length) , mm
0.75 [0.69 – 0.82]
< 0.001
Surface area1, mm 2
0.97 [0.96 – 0.98]
< 0.001
Sampling ratio 2, mm/fragment
0.52 [0.42 – 0.64]
< 0.001
Sampling ratio 2, mm 2/fragment
0.92 [0.90 – 0.95]
< 0.001
1
OR, odds ratio; CI, confidence interval. 1 Adjusted for patient age, sex, longitudinal location, dominant morphology type, endoscopic forceps biopsy pathology before endoscopic submucosal dissection, and number of biopsy fragments. 2 Adjusted for above confounders, except for number of biopsy fragment, which is a constitutive factor of sampling ratio.
that the dominant morphology type and EFB pathology were sta" Table 5), tistically different between the two patient groups (● the statistical difference was not demonstrated in the logistic regression analyses after adjusting for confounders.
Discussion !
Currently, there is only one published report analyzing the clinicopathological and endoscopic features of non-neoplastic pathology results on endoscopic resection specimens [10]. It included 20 non-neoplastic patients (3.2 %) who underwent mainly endoscopic mucosal resection, and reported small tumor sizes (mean 6.40 ± 2.19 mm, range 3 – 10 mm) and low sampling ratios (2.10 ± 1.12 mm/fragment, 11.2 ± 7.85 mm2/fragment), indicating extensive biopsy sampling. The authors suggested that these findings provided evidence to support the EFB removal of small neoplastic foci. Although there was no case of pathology overestimation or incorrect localization of the original tumor, the authors mentioned the chance of pathology misdiagnosis associated with interobserver differences in the LGD diagnosis and ESD at the wrong site as a result of incorrect localization of the original tumor. In the current study, complete removal of small neoplastic foci via relatively extensive EFB was judged as the main reason for " Table 4, " Fig. 4), as in the non-neoplastic results after ESD (● ● study by Kim et al. [10], although the current data showed a slightly larger mean maximum tumor size and higher sampling ratios. However, there were statistically significant differences in tumor size, surface area, and sampling ratios between non-neoplastic and conventional neoplastic pathology groups after ESD " Table 5). These variables were still significant in logistic re(● gression analyses after adjusting for clinical and endoscopic con" Table 6). These findings support the founders (all P < 0.001; ● usefulness of small tumor size and surface area, and low sampling ratios as significant endoscopic predictors of non-neoplastic pathology results after ESD. However, further study is warranted to determine a reasonable cutoff value for these variables. The incidence of submucosal fibrosis during ESD has been reported to be 7.6 % – 8.3 % [14, 15]. In the current study, eight cases (15.4 %) of non-neoplastic pathology after ESD showed submucosal fibrosis during ESD, according to the classification of Higashimaya et al. [14]. It seems likely that the relatively extensive biopsy sampling reflected by the low sampling ratio may increase the mechanical stresses applied per unit area of tumor via higher instant axial force and negative pressure. Subsequently, it aug-
ments reactive proinflammatory processes and regenerative hyperplasia of fibrous tissue. Overestimation of initial pathology according to interobserver differences in the diagnosis of LGD is another important reason for non-neoplastic pathology results after ESD. In the current study, five cases indefinite for neoplasia were overestimated as LGD at private clinics. Epithelial reactive and regenerative changes usually include vascular congestion, and a gradual transition between the atypical and surrounding normal cells [16]. However, the differentiation between reactive/regenerative atypia and LGD is often difficult and complex because the two entities commonly overlap in clinical practice [16 – 18]. Therefore, pathological re-evaluation of EFB slides from referred clinics is needed in order to exclude pathology overestimation in the LGD diagnosis if no tumor is found in the ESD specimen. Incorrect localization of the original tumor with subsequent ESD of the wrong site is rarely the reason for non-neoplastic pathology results after ESD but could have a significant clinical impact on the prognosis for the patient. Correct localization of the ESD site is often difficult in cases with confusing nodular changes of the surrounding mucosa resulting from intestinal metaplasia, " Fig. 2). Furthermore, atrophic gastritis, or H. pylori infection (● subtle or significant morphological changes after EFB may exert an additional confusing influence on tumor localization during " Fig. 3). ESD (● The strengths of the current study are that it exclusively analyzed gastric ESD cases in a large consecutive population (> 1000 cases), determined the practical reasons for non-neoplastic results after ESD, and found statistically significant endoscopic predictors of a non-neoplastic pathology result after ESD via logistic regression analyses adjusting for potential confounders. Nevertheless, the study also has several limitations. The main limitation is potential bias related to its retrospective design such as patient selection, incompleteness of information, and the presence of clinical or endoscopic confounders. To minimize bias, a large number of consecutive patients were included in the study, and a prospectively designed database with organized completion was used to record the demographic characteristics and pre-ESD and ESD endoscopic features of each patient for comprehensive data collection. In addition, logistic regression analyses were performed to control for potential clinical and endoscopic confounders. Another limitation is the relatively short follow-up period for some patients, which may have led to the misinterpretation of results; some patients with lesions assumed to have been removed completely by EFB may in fact have undergone ESD at the wrong site. Among the 45 patients with lesions judged to have been completely removed by EFB, 29 (64.4 %) were followed for more than 2 years, 7 (15.6 %) were followed for between 1 and 2 years, and 9 (20.0 %) were followed for less than 1 year. Six of the nine patients followed for less than 1 year underwent two separate surveillance endoscopies with multiple biopsies, whereas the remaining three patients received only one follow-up endoscopy with biopsy. As a final limitation, there may be an error in the gross measurement of tumor size. In the referred patients who had a pathology slide and no diagnostic endoscopy at the University Hospital, the tumor size tended to be estimated based on the pre-ESD endoscopic images from other hospitals and the ESD images from the University Hospital. However, for patients who showed extreme morphological alterations at the time of ESD, tumor size was de-
Yang Min Jae et al. Non-neoplastic pathology results after endoscopic submucosal dissection … Endoscopy 2015; 47: 598–604
603
Downloaded by: UPSTATE Medical University. Copyrighted material.
Original article
Original article
termined retrospectively based only on pre-ESD images from other hospitals. On the basis of the results and limitations of this study, our recommendations for general ESD practice are as follows. Before ESD, the number of biopsy fragments for a suspected dysplastic lesion during diagnostic endoscopy should be minimized, considering sampling ratio. Although multiple biopsies increase the chance of complete biopsy removal of small neoplastic foci, adverse fibrotic effects imposed by extensive biopsy sampling can make it difficult to perform safe and complete ESD, and morphological alteration of a small tumor after extensive EFB may hinder its correct localization during ESD. In addition, pathological reevaluation of EFB slides from referred clinics is needed to exclude pathology overestimation, especially in patients with LGD. At the time of ESD, repetitive and careful gross inspection using chromoendoscopy or a narrow-band imaging system is essential in order to accurately determine the ESD site and prevent incorrect ESD. Identifying post-EFB scarring can also help to localize the tumor if morphological alteration is extreme after EFB. After ESD, if no tumor is found on the ESD specimen, sequential and comprehensive approaches are needed to confirm the complete removal of the tumor via EFB and exclude other reasons for the non-neoplastic pathology result after ESD. First, re-evaluation of the EFB slides is required to verify the presence of neoplastic tissue before ESD. Second, careful retrospective review of the images from diagnostic endoscopy and ESD is needed to ensure the appropriate localization of the ESD site and the presence of biopsy scars on the ESD specimen. Third, and most importantly, a follow-up period of at least 2 years is needed to clearly distinguish between complete biopsy removal and incorrect ESD site, especially in patients with LGD. In conclusion, complete biopsy removal of small neoplastic foci, initial pathology overestimation, and incorrect localization of the original tumor with subsequent ESD at the wrong site are the main reasons for non-neoplastic pathology results after ESD. Small tumor size and surface area, and low sampling ratios are significant endoscopic predictors of a non-neoplastic pathology result after ESD. A comprehensive review of all pre-ESD data and sufficient follow-up time are needed to determine the cause of a non-neoplastic pathology result after ESD. Competing interests: None
References 1 Gotoda T, Yamamoto H, Soetikno RM. Endoscopic submucosal dissection of early gastric cancer. J Gastroenterol 2006; 41: 929 – 942 2 Tanaka M, Ono H, Hasuike N et al. Endoscopic submucosal dissection of early gastric cancer. Digestion 2008; 77 Suppl 1: 23 – 28 3 Yoon WJ, Lee DH, Jung YJ et al. Histologic characteristics of gastric polyps in Korea: emphasis on discrepancy between endoscopic forceps biopsy and endoscopic mucosal resection specimen. World J Gastroenterol 2006; 12: 4029 – 4032 4 Sung HY, Cheung DY, Cho SH et al. Polyps in the gastrointestinal tract: discrepancy between endoscopic forceps biopsies and resected specimens. Eur J Gastroenterol Hepatol 2009; 21: 190 – 195 5 Lee CK, Chung IK, Lee SH et al. Is endoscopic forceps biopsy enough for a definitive diagnosis of gastric epithelial neoplasia? J Gastroenterol Hepatol 2010; 25: 1507 – 1513 6 Takao M, Kakushima N, Takizawa K et al. Discrepancies in histologic diagnoses of early gastric cancer between biopsy and endoscopic mucosal resection specimens. Gastric Cancer 2012; 15: 91 – 96 7 Kim YJ, Park JC, Kim JH et al. Histologic diagnosis based on forceps biopsy is not adequate for determining endoscopic treatment of gastric adenomatous lesions. Endoscopy 2010; 42: 620 – 626 8 Won CS, Cho MY, Kim HS et al. Upgrade of lesions initially diagnosed as low-grade gastric dysplasia upon forceps biopsy following endoscopic resection. Gut Liver 2011; 5: 187 – 193 9 Cho SJ, Choi IJ, Kim CG et al. Risk of high-grade dysplasia or carcinoma in gastric biopsy-proven low-grade dysplasia: an analysis using the Vienna classification. Endoscopy 2011; 43: 465 – 471 10 Kim ES, Jeon SW, Park SY et al. Where has the tumor gone? The characteristics of cases of negative pathologic diagnosis after endoscopic mucosal resection. Endoscopy 2009; 41: 739 – 745 11 Dixon MF. Gastrointestinal epithelial neoplasia: Vienna revisited. Gut 2002; 51: 130 – 131 12 Gotoda T, Yanagisawa A, Sasako M et al. Incidence of lymph node metastasis from early gastric cancer: estimation with a large number of cases at two large centers. Gastric Cancer 2000; 3: 219 – 225 13 Hull MJ, Mino-Kenudson M, Nishioka NS et al. Endoscopic mucosal resection: an improved diagnostic procedure for early gastroesophageal epithelial neoplasms. Am J Surg Pathol 2006; 30: 114 – 118 14 Higashimaya M, Oka S, Tanaka S et al. Outcome of endoscopic submucosal dissection for gastric neoplasm in relationship to endoscopic classification of submucosal fibrosis. Gastric Cancer 2013; 16: 404 – 410 15 Jeong JY, Oh YH, Yu YH et al. Does submucosal fibrosis affect the results of endoscopic submucosal dissection of early gastric tumors? Gastrointest Endosc 2012; 76: 59 – 66 16 Ming SC, Bajtai A, Correa P et al. Gastric dysplasia. Significance and pathologic criteria. Cancer 1984; 54: 1794 – 1801 17 Alfaro EE, Lauwers GY. Early gastric neoplasia: diagnosis and implications. Adv Anat Pathol 2011; 18: 268 – 280 18 Lauwers GY, Riddell RH. Gastric epithelial dysplasia. Gut 1999; 45: 784 – 790
Yang Min Jae et al. Non-neoplastic pathology results after endoscopic submucosal dissection … Endoscopy 2015; 47: 598–604
Downloaded by: UPSTATE Medical University. Copyrighted material.
604
Non-neoplastic pathology results after endoscopic submucosal dissection for gastric epithelial dysplasia or early gastric cancer
Authors
Min Jae Yang1, Sung Jae Shin1, Ki Seong Lee1, Kee Myung Lee1, Sun Gyo Lim1, Joon Koo Kang1, Jae Chul Hwang1, Soon Sun Kim1, Dakeun Lee2, Joo-Sung Kim1, Gil-Ho Lee1, Han Seok Ryu1, Byung Moo Yoo1, Kwang Jae Lee1, Young Bae Kim2, Jin Hong Kim1
Institutions
1 2
submitted 5. August 2014 accepted after revision 27. November 2014
Bibliography DOI http://dx.doi.org/ 10.1055/s-0034-1391375 Published online: 4.2.2015 Endoscopy 2015; 47: 598–604 © Georg Thieme Verlag KG Stuttgart · New York ISSN 0013-726X Corresponding author Sung Jae Shin, MD, PhD Department of Gastroenterology Ajou University School of Medicine 164 Worldcup-ro Yeongtong-gu, Suwon-si Gyeonggi-do, 443–380 South Korea Fax: 82–31–2195999 [email protected]
Department of Gastroenterology, Ajou University School of Medicine, Suwon, South Korea Department of Pathology, Ajou University School of Medicine, Suwon, South Korea
Background and study aims: Endoscopists sometimes face paradoxical cases in which the endoscopic submucosal dissection (ESD) specimen reveals a non-neoplastic pathology result. The aims of the study were to determine the reasons for such results, and to compare the endoscopic characteristics of non-neoplastic and conventional neoplastic pathology groups after ESD. Patients and methods: A total of 1186 gastric ESDs performed between February 2005 and December 2011 were retrospectively reviewed. The ESD specimens included 52 (4.4 %) that were confirmed as negative or indefinite for neoplasia. Patient characteristics and endoscopic and pathological data were reviewed and compared. Results: Non-neoplastic pathology after ESD was due to complete removal of the lesion at biopsy in 45 cases (86.5 %), pathology overestimation in 5 (9.6 %), and incorrect localization of the original tumor with subsequent ESD performed at the
wrong site in 2 (3.8 %). The mean length and surface area of the non-neoplastic lesions were 9.2 ± 2.6 mm and 49.6 ± 23.6 mm 2, respectively. Mean sampling ratios were 3.0 ± 1.5 mm/fragment and 16.3 ± 10.0 mm2/fragment. Compared with 1134 cases confirmed as neoplastic on the final ESD specimen, non-neoplastic cases showed a significantly smaller tumor size and surface area, and lower sampling ratios in a logistic regression analysis adjusted for potential confounders (P < 0.001 for all). Conclusions: Complete lesion removal by biopsy, pathology overestimation, and incorrect localization of the original tumor with subsequent ESD at the wrong site were the main reasons for nonneoplastic results after ESD. Small tumor size and surface area, and low sampling ratios were associated with non-neoplastic pathology results after ESD.
Introduction
therapeutic plan, and worse clinical outcome than expected [3, 7 – 9]. However, endoscopists have faced paradoxical cases in which the endoscopic resection specimen revealed a downgraded final pathology, especially a non-neoplastic result, albeit infrequently, with reported rates of 3.2 % – 7.0 % [5, 7,10]. Until recently, few studies have been conducted on non-neoplastic pathology results after endoscopic resection, especially ESD. The aims of this study were to analyze the reasons for non-neoplastic pathology results on ESD specimens, and to compare the endoscopic characteristics of non-neoplastic and conventional neoplastic pathology groups after ESD.
!
Endoscopic submucosal dissection (ESD) is the current standard of care for selected cases of gastric epithelial dysplasia or early gastric cancer because of its curative en bloc resectability and complete histopathological assessment of the specimen [1, 2]. However, in many cases, the initial pathological diagnosis changes after ESD because the biopsy specimen does not reflect the entire histological background of the lesion. Histological discrepancies between endoscopic forceps biopsy (EFB) and endoscopic resection specimens, including ESD and endoscopic mucosal resection, have been reported at rates of 2.7 % – 49 % [3 – 6]. Most studies on pathological discrepancies after endoscopic resection have concentrated on the difference between concordant and upgraded histological changes that imply initial underestimation of the dysplasia grade, modification of the
Yang Min Jae et al. Non-neoplastic pathology results after endoscopic submucosal dissection … Endoscopy 2015; 47: 598–604
Downloaded by: UPSTATE Medical University. Copyrighted material.
598
Original article
Table 1 [11].
!
Study population Between February 2005 and December 2011, 1186 consecutive patients underwent gastric ESD at Ajou University Hospital in Suwon, South Korea. These cases were retrospectively reviewed from a prospectively designed ESD database system, and included 52 patients (4.4 %) who were confirmed to have negative or indefinite neoplastic results on the ESD specimen (category 1 or 2 according to the revised Vienna classification of gastrointestinal " Table 1). neoplasia [11]) (● The study was approved by the institutional review board (approval no. AJIRB-MED-MDB-13-248), and informed consent was obtained from each patient for the ESD.
Endoscopic protocol before ESD Diagnostic endoscopy was performed in patients who were referred to the Ajou University Hospital for evaluation of upper gastrointestinal symptoms without a preceding endoscopy. If epithelial dysplasia or cancer was suspected, neoplasia-targeted EFB was performed using a narrow-band imaging system and chromoendoscopy with indigo carmine dye spray. Before EFB, the size of the lesion was estimated macroscopically using standard biopsy forceps with a 6-mm opening diameter (FB-21K-1; Olympus, Tokyo, Japan) and 6-mm circular pieces of punched out paper. The number of biopsy fragments was determined at the discretion of the endoscopist, after considering the size and surface configuration of the lesion. Patients who were pathologically diagnosed at private clinics and then referred to the University Hospital with a pathology slide did not undergo routine re-endoscopy with neoplasia-targeted biopsy because the additional cost and potential risk of biopsyinduced fibrosis, which might compromise the safe completion of ESD, outweighed the potential benefits of additional pathological information, which only rarely modifies the treatment strategy. As a result, all lesions in the study population were pathologically confirmed as gastric epithelial dysplasia or cancer via preESD EFB performed either at the University Hospital or at private clinics.
ESD procedure Indications for ESD included lesions with pathological findings of either gastric epithelial dysplasia based on the EFB, irrespective of size or surface configuration, or early gastric cancer that met the expanded criteria of the Japanese Gastric Cancer Association [12]. All ESDs were performed using a standard endoscope with a single channel (GIF-H260 or GIF-Q260J; Olympus) under conscious sedation with both midazolam and propofol. Standard ESD techniques was performed with an insulated-tip knife (Olympus) and/or a FlexKnife (Olympus) in the dry cut or swift coagulation mode (Erbotom ICC200 or VIO 300D; ERBE, Tübingen, Germany). Periodic surveillance endoscopies were scheduled for 3, 6, 12, 18, and 24 months after ESD, and annually thereafter. EFB was carried out in the area surrounding the post-ESD scar and at other sites if needed at every follow-up endoscopy.
Revised Vienna classification of gastrointestinal epithelial neoplasia
Category
Diagnosis
1
Negative for neoplasia
2
Indefinite for neoplasia
3
Mucosal low grade neoplasia Low grade adenoma or dysplasia
4
Mucosal high grade neoplasia 4 – 1 High grade adenoma or dysplasia 4 – 2 Noninvasive carcinoma (carcinoma in situ) 4 – 3 Suspicious for invasive carcinoma 4 – 4 Intramucosal carcinoma (lamina propria invasion)
5
Submucosal invasion by carcinoma
number of biopsy fragments, is a reflection of the extent of biopsy sampling, with a lower ratio reflecting more extensive sampling [13].
Histological features Two experienced gastrointestinal pathologists (Y.B.K., D.L.) confirmed the histological diagnoses of the pre-ESD biopsies and ESD specimens. The histological diagnosis was made according to the revised Vienna classification of gastrointestinal neoplasia [11]. When no tumor was found in the ESD specimen, all slides from pre-ESD biopsies and ESD specimens were peer evaluated or re-evaluated by the pathologist who provided the original pre-ESD diagnosis in order to confirm that the entire lesion had been removed during biopsy or to confirm initial pathological misdiagnosis. In patients who were referred from private clinics with a pathology slide, one of the two pathologists re-evaluated the pre-ESD slide and the ESD specimen. Involvement of atrophic gastritis and intestinal metaplasia was assessed histologically, and the presence of Helicobacter pylori infection was confirmed on the basis of the results of the EFB and rapid urease test or urea breath test.
Procedure-related features Procedure time, procedure-related adverse events (perforation or delayed bleeding), and the presence of submucosal fibrosis were assessed. Submucosal fibrosis was diagnosed when the submucosal layer could not be easily separated from the whitish muscular area, which appears without a blue transparent layer, during lesion dissection (consistent with F1 or F2 fibrosis, as suggested by Higashimaya et al. [14]).
Statistical analyses Statistical analyses were performed using SPSS software version 18.0 (SPSS Inc., Chicago, Illinois, USA). The significance of differences for quantitative data was determined using unpaired Student’s t test and was presented as mean ± SD. Categorical data were compared using the chi-squared test and Fisher’s exact test. Binary logistic regression analyses were applied to identify significant endoscopic predictors of a non-neoplastic pathology result after ESD. All statistical tests were conducted with a 2-sided significance level of 0.05.
Endoscopic features The following endoscopic variables were analyzed and compared: tumor size and location, dominant morphology type, number of biopsy fragments, and sampling ratios. Sampling ratio, calculated by dividing the length of the tumor or surface area by the
Yang Min Jae et al. Non-neoplastic pathology results after endoscopic submucosal dissection … Endoscopy 2015; 47: 598–604
Downloaded by: UPSTATE Medical University. Copyrighted material.
Patients and methods
599
Original article
Table 2 Distribution of the 1186 patients according to the endoscopic forceps biopsy and endoscopic submucosal dissection pathology.
ESD pathology
EFB pathology LGD
Total
HGD
Cancer
Category 1
28
2
8
38
Category 2
11
1
2
14
LGD
468
25
4
497
HGD
61
51
2
114
Cancer
135
92
296
523
Total
703
171
312
1186
ESD (n = 1186)
Concordant (n = 815) 68.7 %
Discordant (n = 371) 31.3 %
Down-graded (n = 83) 7%
Up-graded (n = 288) 24.3 %
Non-neoplastic result (n = 52) 4.4 %
Fig. 1
Flow chart of the enrolled study population.
EFB, endoscopic forceps biopsy; ESD, endoscopic submucosal dissection; LGD, low grade dysplasia; HGD, high grade dysplasia.
Table 3 Clinical characteristics of patients with non-neoplastic pathology results after endoscopic submucosal dissection. Non-neoplasia cases after ESD, n/N (%)
52/1186 (4.4)
Sex, n (%) Male
33 (63.5)
Female
19 (36.5)
Age, mean ± SD (range), years
59.6 ± 12.2 (35 – 81)
Follow-up period, mean ± SD (range), months
20.3 ± 18.7 (1 – 75)
Forceps biopsy diagnosis 1, n (%) Low grade dysplasia High grade dysplasia Carcinoma
39 (75.0) 3 (5.8) 10 (19.2)
Post-ESD diagnosis 1, n (%) Category 1
38 (73.1)
Category 2
14 (26.9)
Coexistent pathology on ESD specimen, n (%) Helicobacter gastritis
28 (53.8)
Atrophic gastritis
43 (82.7)
Intestinal metaplasia
48 (92.3)
Procedure-related factors ESD time, mean ± SD (range), minutes
47.5 ± 27.8 (12 – 129)
Perforation, n (%)
2 (3.8)
Delayed bleeding, n (%)
3 (5.8)
Submucosal fibrosis during ESD, n (%)
8 (15.4)
ESD, endoscopic submucosal dissection. 1 Revised Vienna classification [11].
Table 4 Reasons for non-neoplastic pathology results after endoscopic submucosal dissection (n = 52). Complete biopsy removal, n (%)
45 (86.5)
Pathology overestimation, n (%)
5 (9.6)
Incorrect localization of ESD site, n (%)
2 (3.8)
ESD, endoscopic submucosal dissection.
Results !
The distribution of the 1186 patients according to EFB and ESD " Table 2. The overall histological dispathology is presented in ● crepancy rate between EFB and ESD specimens was 31.3 % (371/ 1186). A downgrade of the pathological diagnosis was found in 83 cases (7.0 %). In 52 patients (4.4 %), a non-neoplastic result on " Fig. 1). the ESD specimen was found (● The clinical characteristics of the patients with non-neoplastic " Table 3. The initial pathoresults after ESD are summarized in ● logical diagnoses based on EFB samplings were 39 low grade dysplasias (LGDs) (75.0 %), 3 high grade dysplasias (5.8 %), and 10 adenocarcinomas (19.2 %). The final histological diagnoses based
on ESD specimens were 38 category 1 cases (negative for neoplasia, 73.1 %) and 14 category 2 cases (indefinite for neoplasia, 26.9 %). Intestinal metaplasia, atrophic gastritis, and H. pylori infection were observed in the surrounding gastric mucosa in 48 (92.3 %), 43 (82.7 %), and 28 (53.8 %) cases, respectively. Eight patients (15.4 %) with non-neoplastic pathology results after ESD showed submucosal fibrosis during ESD, one of whom experienced ESD-induced perforation. After pathological re-evaluation, the EFB-based initial pathological diagnosis in five cases from private clinics was changed from category 3.1 (LGD) to category 2 (indefinite for neoplasia). The final ESD diagnoses of these cases were all consistent with category 2. In one patient with adenocarcinoma on EFB samples and a nonneoplastic result on the ESD specimen, cancerous tissue was found in the follow-up endoscopic biopsy performed 3 months after ESD. This indicated incorrect localization of the original tumor and subsequent ESD at the wrong site as a result of confusing nodular changes in the surrounding mucosa, which were due to " Fig. 2). This patient subsequently underintestinal metaplasia (● went total gastrectomy, as it was not possible to perform complete re-ESD because of severe fibrosis and bleeding. In another patient with LGD on EFB specimens and a non-neoplastic pathology on the ESD specimen, a polypoid lesion was found close to the margin of the post-ESD scar at the follow-up endoscopy performed 12 months after ESD. This also indicated initial incorrect localization of the original tumor and subsequent ESD at the wrong site, this time as a result of morphological alteration fol" Fig. 3). In summary, the lowing pre-ESD multiple biopsies (● non-neoplastic ESD pathology was considered to be due to complete removal of the lesion by EFB in 45 patients (86.5 %), pathology overestimation in 5 (9.6 %), and incorrect localization of the original tumor with subsequent ESD at the wrong site in 2 (3.8 %) " Table 4). A representative case of complete biopsy removal is (● " Fig. 4. shown in● The baseline clinicopathological and endoscopic characteristics of the non-neoplastic (n = 52) and conventional neoplastic (n = " Table 5. The 1134) pathology groups after ESD are shown in ● mean length, width, and surface area of the non-neoplastic lesions were 9.2 ± 2.6 mm, 6.4 ± 1.8 mm, and 49.6 ± 23.6 mm2, respectively. The mean number of EFB fragments was 3.5 ± 1.5, and the mean sampling ratios were 3.0 ± 1.5 mm/fragment and 16.3 ± 10.0 mm2/fragment. Compared with patients with conventional neoplastic pathology after ESD, patients with non-neoplastic pathology showed a significantly smaller tumor size and " Table 5). These differsurface area, and lower sampling ratios (● ences were still statistically significant in logistic regression analyses after adjusting for clinical and endoscopic confounders (all " Table 6). Although the univariate analyses showed P < 0.001) (●
Yang Min Jae et al. Non-neoplastic pathology results after endoscopic submucosal dissection … Endoscopy 2015; 47: 598–604
Downloaded by: UPSTATE Medical University. Copyrighted material.
600
Original article
601
Fig. 3 Incorrect localization of the original tumor and subsequent endoscopic submucosal dissection (ESD) at the wrong site as a result of morphological alteration following pre-ESD forceps biopsy. a A slightly elevated round lesion of the lesser curvature of the proximal antrum (black arrows) was confirmed as low grade dysplasia on the pre-ESD forceps biopsy. b The dysplastic lesion was not clearly defined at the time of ESD. c The post-ESD scarring found at 3 months after ESD. d Endoscopy 12 months later, showing a polypoid lesion at the margin of the post-ESD scar (arrows).
Yang Min Jae et al. Non-neoplastic pathology results after endoscopic submucosal dissection … Endoscopy 2015; 47: 598–604
Downloaded by: UPSTATE Medical University. Copyrighted material.
Fig. 2 Incorrect localization of the original tumor and therefore subsequent endoscopic submucosal dissection (ESD) at the wrong site as a result of confusing nodular changes in the surrounding mucosa. a, b Endoscopic images acquired during endoscopic forceps biopsy, showing a slightly elevated nodular lesion with central depression of the mid body posterior wall. c Endoscopic image acquired during ESD, showing a cancerous lesion (black arrow), and surrounding intestinal metaplasia that was mistaken for a cancerous lesion (red circle). d ESD of the incorrect site.
Original article
Fig. 4 Complete removal of a small neoplasia during endoscopic forceps biopsy (EFB). a Arrows indicate a 4-mm superficial elevated round lesion on the posterior wall of the lower body. b EFB. c PostEFB image acquired during endoscopic submucosal dissection (ESD). d Submucosal fibrosis confirmed during ESD.
Age, mean ± SD, years
Non-neoplasia group
Conventional group
(n = 52)
(n = 1134)
59.6 ± 12.2
62.4 ± 10.3
Male
33 (63.5)
827 (72.9)
Female
19 (36.5)
307 (27.1)
Upper
4 (7.7)
64 (5.6)
Middle
22 (42.3)
377 (33.2)
Lower
26 (50.0)
693 (61.1)
Protruded
31 (59.6)
786 (69.3)
Flat
15 (28.8)
172 (15.2)
6 (11.5)
176 (15.5)
39 (75.0)
664 (58.6)
Longitudinal location 2, n (%)
0.274
0.029
Dominant morphology, n (%)
0.047
EFB pathology, n (%) Low grade dysplasia High grade dysplasia Cancer
3 (5.8)
168 (14.8)
10 (19.2)
302 (26.6)
Number of biopsy fragments, mean ± SD (range)
3.5 ± 1.5 (1 – 6)
Tumor size, mean length ± SD (range), mm
9.2 ± 2.6 (4 – 16)
Surface area, mean ± SD (range), mm 2
0.119 0.135
Sex, n (%)
Depressed
P value1
49.6 ± 23.6 (9.4 – 109.9)
3.6 ± 1.9 (1 – 9)
0.622
19.4 ± 9.1 (5 – 48)
< 0.001
281.5 ± 265.2 (7.9 – 1413)
< 0.001
Sampling ratio 3 mm/fragment
3.0 ± 1.5
6.7 ± 4.1
< 0.001
mm 2/fragment
16.3 ± 10.0
89.7 ± 94.3
< 0.001
EFB, endoscopic forceps biopsy. 1 Univariate analysis, not adjusted by other variables. 2 Upper (cardia, fundus, upper body), middle (mid body, lower body, angle), and lower (antrum, prepylorus). 3 Calculated by dividing the length of the tumor or surface area by the number of biopsy fragments. It reflects the extent of biopsy sampling, with a lower ratio reflecting more extensive sampling.
Yang Min Jae et al. Non-neoplastic pathology results after endoscopic submucosal dissection … Endoscopy 2015; 47: 598–604
Table 5 Comparison of baseline clinicopathological and endoscopic features between non-neoplastic and conventional neoplastic pathology groups after endoscopic submucosal dissection.
Downloaded by: UPSTATE Medical University. Copyrighted material.
602
Table 6 Logistic regression analyses of endoscopic predictors for non-neoplastic pathology results after endoscopic submucosal dissection, adjusted for clinical and endoscopic factors.
Adjusted OR [95 %CI]
P value
Tumor size (length) , mm
0.75 [0.69 – 0.82]
< 0.001
Surface area1, mm 2
0.97 [0.96 – 0.98]
< 0.001
Sampling ratio 2, mm/fragment
0.52 [0.42 – 0.64]
< 0.001
Sampling ratio 2, mm 2/fragment
0.92 [0.90 – 0.95]
< 0.001
1
OR, odds ratio; CI, confidence interval. 1 Adjusted for patient age, sex, longitudinal location, dominant morphology type, endoscopic forceps biopsy pathology before endoscopic submucosal dissection, and number of biopsy fragments. 2 Adjusted for above confounders, except for number of biopsy fragment, which is a constitutive factor of sampling ratio.
that the dominant morphology type and EFB pathology were sta" Table 5), tistically different between the two patient groups (● the statistical difference was not demonstrated in the logistic regression analyses after adjusting for confounders.
Discussion !
Currently, there is only one published report analyzing the clinicopathological and endoscopic features of non-neoplastic pathology results on endoscopic resection specimens [10]. It included 20 non-neoplastic patients (3.2 %) who underwent mainly endoscopic mucosal resection, and reported small tumor sizes (mean 6.40 ± 2.19 mm, range 3 – 10 mm) and low sampling ratios (2.10 ± 1.12 mm/fragment, 11.2 ± 7.85 mm2/fragment), indicating extensive biopsy sampling. The authors suggested that these findings provided evidence to support the EFB removal of small neoplastic foci. Although there was no case of pathology overestimation or incorrect localization of the original tumor, the authors mentioned the chance of pathology misdiagnosis associated with interobserver differences in the LGD diagnosis and ESD at the wrong site as a result of incorrect localization of the original tumor. In the current study, complete removal of small neoplastic foci via relatively extensive EFB was judged as the main reason for " Table 4, " Fig. 4), as in the non-neoplastic results after ESD (● ● study by Kim et al. [10], although the current data showed a slightly larger mean maximum tumor size and higher sampling ratios. However, there were statistically significant differences in tumor size, surface area, and sampling ratios between non-neoplastic and conventional neoplastic pathology groups after ESD " Table 5). These variables were still significant in logistic re(● gression analyses after adjusting for clinical and endoscopic con" Table 6). These findings support the founders (all P < 0.001; ● usefulness of small tumor size and surface area, and low sampling ratios as significant endoscopic predictors of non-neoplastic pathology results after ESD. However, further study is warranted to determine a reasonable cutoff value for these variables. The incidence of submucosal fibrosis during ESD has been reported to be 7.6 % – 8.3 % [14, 15]. In the current study, eight cases (15.4 %) of non-neoplastic pathology after ESD showed submucosal fibrosis during ESD, according to the classification of Higashimaya et al. [14]. It seems likely that the relatively extensive biopsy sampling reflected by the low sampling ratio may increase the mechanical stresses applied per unit area of tumor via higher instant axial force and negative pressure. Subsequently, it aug-
ments reactive proinflammatory processes and regenerative hyperplasia of fibrous tissue. Overestimation of initial pathology according to interobserver differences in the diagnosis of LGD is another important reason for non-neoplastic pathology results after ESD. In the current study, five cases indefinite for neoplasia were overestimated as LGD at private clinics. Epithelial reactive and regenerative changes usually include vascular congestion, and a gradual transition between the atypical and surrounding normal cells [16]. However, the differentiation between reactive/regenerative atypia and LGD is often difficult and complex because the two entities commonly overlap in clinical practice [16 – 18]. Therefore, pathological re-evaluation of EFB slides from referred clinics is needed in order to exclude pathology overestimation in the LGD diagnosis if no tumor is found in the ESD specimen. Incorrect localization of the original tumor with subsequent ESD of the wrong site is rarely the reason for non-neoplastic pathology results after ESD but could have a significant clinical impact on the prognosis for the patient. Correct localization of the ESD site is often difficult in cases with confusing nodular changes of the surrounding mucosa resulting from intestinal metaplasia, " Fig. 2). Furthermore, atrophic gastritis, or H. pylori infection (● subtle or significant morphological changes after EFB may exert an additional confusing influence on tumor localization during " Fig. 3). ESD (● The strengths of the current study are that it exclusively analyzed gastric ESD cases in a large consecutive population (> 1000 cases), determined the practical reasons for non-neoplastic results after ESD, and found statistically significant endoscopic predictors of a non-neoplastic pathology result after ESD via logistic regression analyses adjusting for potential confounders. Nevertheless, the study also has several limitations. The main limitation is potential bias related to its retrospective design such as patient selection, incompleteness of information, and the presence of clinical or endoscopic confounders. To minimize bias, a large number of consecutive patients were included in the study, and a prospectively designed database with organized completion was used to record the demographic characteristics and pre-ESD and ESD endoscopic features of each patient for comprehensive data collection. In addition, logistic regression analyses were performed to control for potential clinical and endoscopic confounders. Another limitation is the relatively short follow-up period for some patients, which may have led to the misinterpretation of results; some patients with lesions assumed to have been removed completely by EFB may in fact have undergone ESD at the wrong site. Among the 45 patients with lesions judged to have been completely removed by EFB, 29 (64.4 %) were followed for more than 2 years, 7 (15.6 %) were followed for between 1 and 2 years, and 9 (20.0 %) were followed for less than 1 year. Six of the nine patients followed for less than 1 year underwent two separate surveillance endoscopies with multiple biopsies, whereas the remaining three patients received only one follow-up endoscopy with biopsy. As a final limitation, there may be an error in the gross measurement of tumor size. In the referred patients who had a pathology slide and no diagnostic endoscopy at the University Hospital, the tumor size tended to be estimated based on the pre-ESD endoscopic images from other hospitals and the ESD images from the University Hospital. However, for patients who showed extreme morphological alterations at the time of ESD, tumor size was de-
Yang Min Jae et al. Non-neoplastic pathology results after endoscopic submucosal dissection … Endoscopy 2015; 47: 598–604
603
Downloaded by: UPSTATE Medical University. Copyrighted material.
Original article
Original article
termined retrospectively based only on pre-ESD images from other hospitals. On the basis of the results and limitations of this study, our recommendations for general ESD practice are as follows. Before ESD, the number of biopsy fragments for a suspected dysplastic lesion during diagnostic endoscopy should be minimized, considering sampling ratio. Although multiple biopsies increase the chance of complete biopsy removal of small neoplastic foci, adverse fibrotic effects imposed by extensive biopsy sampling can make it difficult to perform safe and complete ESD, and morphological alteration of a small tumor after extensive EFB may hinder its correct localization during ESD. In addition, pathological reevaluation of EFB slides from referred clinics is needed to exclude pathology overestimation, especially in patients with LGD. At the time of ESD, repetitive and careful gross inspection using chromoendoscopy or a narrow-band imaging system is essential in order to accurately determine the ESD site and prevent incorrect ESD. Identifying post-EFB scarring can also help to localize the tumor if morphological alteration is extreme after EFB. After ESD, if no tumor is found on the ESD specimen, sequential and comprehensive approaches are needed to confirm the complete removal of the tumor via EFB and exclude other reasons for the non-neoplastic pathology result after ESD. First, re-evaluation of the EFB slides is required to verify the presence of neoplastic tissue before ESD. Second, careful retrospective review of the images from diagnostic endoscopy and ESD is needed to ensure the appropriate localization of the ESD site and the presence of biopsy scars on the ESD specimen. Third, and most importantly, a follow-up period of at least 2 years is needed to clearly distinguish between complete biopsy removal and incorrect ESD site, especially in patients with LGD. In conclusion, complete biopsy removal of small neoplastic foci, initial pathology overestimation, and incorrect localization of the original tumor with subsequent ESD at the wrong site are the main reasons for non-neoplastic pathology results after ESD. Small tumor size and surface area, and low sampling ratios are significant endoscopic predictors of a non-neoplastic pathology result after ESD. A comprehensive review of all pre-ESD data and sufficient follow-up time are needed to determine the cause of a non-neoplastic pathology result after ESD. Competing interests: None
References 1 Gotoda T, Yamamoto H, Soetikno RM. Endoscopic submucosal dissection of early gastric cancer. J Gastroenterol 2006; 41: 929 – 942 2 Tanaka M, Ono H, Hasuike N et al. Endoscopic submucosal dissection of early gastric cancer. Digestion 2008; 77 Suppl 1: 23 – 28 3 Yoon WJ, Lee DH, Jung YJ et al. Histologic characteristics of gastric polyps in Korea: emphasis on discrepancy between endoscopic forceps biopsy and endoscopic mucosal resection specimen. World J Gastroenterol 2006; 12: 4029 – 4032 4 Sung HY, Cheung DY, Cho SH et al. Polyps in the gastrointestinal tract: discrepancy between endoscopic forceps biopsies and resected specimens. Eur J Gastroenterol Hepatol 2009; 21: 190 – 195 5 Lee CK, Chung IK, Lee SH et al. Is endoscopic forceps biopsy enough for a definitive diagnosis of gastric epithelial neoplasia? J Gastroenterol Hepatol 2010; 25: 1507 – 1513 6 Takao M, Kakushima N, Takizawa K et al. Discrepancies in histologic diagnoses of early gastric cancer between biopsy and endoscopic mucosal resection specimens. Gastric Cancer 2012; 15: 91 – 96 7 Kim YJ, Park JC, Kim JH et al. Histologic diagnosis based on forceps biopsy is not adequate for determining endoscopic treatment of gastric adenomatous lesions. Endoscopy 2010; 42: 620 – 626 8 Won CS, Cho MY, Kim HS et al. Upgrade of lesions initially diagnosed as low-grade gastric dysplasia upon forceps biopsy following endoscopic resection. Gut Liver 2011; 5: 187 – 193 9 Cho SJ, Choi IJ, Kim CG et al. Risk of high-grade dysplasia or carcinoma in gastric biopsy-proven low-grade dysplasia: an analysis using the Vienna classification. Endoscopy 2011; 43: 465 – 471 10 Kim ES, Jeon SW, Park SY et al. Where has the tumor gone? The characteristics of cases of negative pathologic diagnosis after endoscopic mucosal resection. Endoscopy 2009; 41: 739 – 745 11 Dixon MF. Gastrointestinal epithelial neoplasia: Vienna revisited. Gut 2002; 51: 130 – 131 12 Gotoda T, Yanagisawa A, Sasako M et al. Incidence of lymph node metastasis from early gastric cancer: estimation with a large number of cases at two large centers. Gastric Cancer 2000; 3: 219 – 225 13 Hull MJ, Mino-Kenudson M, Nishioka NS et al. Endoscopic mucosal resection: an improved diagnostic procedure for early gastroesophageal epithelial neoplasms. Am J Surg Pathol 2006; 30: 114 – 118 14 Higashimaya M, Oka S, Tanaka S et al. Outcome of endoscopic submucosal dissection for gastric neoplasm in relationship to endoscopic classification of submucosal fibrosis. Gastric Cancer 2013; 16: 404 – 410 15 Jeong JY, Oh YH, Yu YH et al. Does submucosal fibrosis affect the results of endoscopic submucosal dissection of early gastric tumors? Gastrointest Endosc 2012; 76: 59 – 66 16 Ming SC, Bajtai A, Correa P et al. Gastric dysplasia. Significance and pathologic criteria. Cancer 1984; 54: 1794 – 1801 17 Alfaro EE, Lauwers GY. Early gastric neoplasia: diagnosis and implications. Adv Anat Pathol 2011; 18: 268 – 280 18 Lauwers GY, Riddell RH. Gastric epithelial dysplasia. Gut 1999; 45: 784 – 790
Yang Min Jae et al. Non-neoplastic pathology results after endoscopic submucosal dissection … Endoscopy 2015; 47: 598–604
Downloaded by: UPSTATE Medical University. Copyrighted material.
604
