- Email: [email protected]
Endoscopic submucosal dissection of early gastric neoplasia with a water jet–assisted knife: a Western, single-center experience
ORIGINAL ARTICLE: Clinical Endoscopy
Endoscopic submucosal dissection of early gastric neoplasia with a water jet–assisted knife: a Western, single-center experience Brigitte Schumacher, MD,1 Jean-Pierre Charton, MD,1 Thomas Nordmann, MD,1 Michael Vieth, MD,2 Markus Enderle,3 Horst Neuhaus, MD1 Düsseldorf, Germany
Background: Endoscopic submucosal dissection (ESD) of early gastric neoplasia has not yet been established in Western countries because of a lack of data and the difficult, time-consuming, and hazardous nature of the method. Some of the technical limitations may be overcome by use of a water jet–assisted knife, which allows a combination of a high-pressure water jet and electrosurgical interventions. Objective: To evaluate the efficacy and safety of water jet–assisted ESD (WESD) with a water jet–assisted knife in selected patients with early gastric neoplasia. Design: Single-center, prospective study. Patients: This study involved 29 consecutive patients (13 female; median age 61 years; age range 35-93 years) with early gastric neoplasia that met the expanded criteria of the Japanese Gastric Cancer Association. Histology of biopsies had shown gastric adenocarcinoma in 21 cases, adenoma in 8 case, and suspicion of a GI stromal tumor in 1 case. The median maximal diameter of the lesions was 20 mm (range 10-40 mm). Intervention: All procedures were done with patients under sedation with propofol. The water jet–assisted knife was used for setting coagulation markers around the neoplastic lesions, then for circumferential incision and dissection in combination with repeated submucosal injection of saline solution with a water jet system. Bleeding was treated with diathermia by use of the water jet–assisted knife or hemostatic forceps in case of failure or larger vessels. Clips were used for closure of perforations. Main Outcome Measurements: Complete resection of neoplasia, procedure time, complication and recurrence rates. Results: According to endoscopic criteria, complete resection of the targeted area could be achieved in all cases, with an en bloc resection rate of 90%. The median procedure duration was 74 minutes (range 15-402 minutes). Exchange of the device was needed in only 10 cases because of severe bleeding from larger vessels, which could be managed by use of hemostatic forceps. The 30-day morbidity rate was 4 of 30 (13.8%) because of postprocedure pain in 3 cases and delayed bleeding in 1 case. A 93-year-old patient died the night after WESD without evidence of a procedure-related complication. Histology of the resected specimens showed adenocarcinoma in 20 cases, adenoma in 7, no neoplasia in 2, and a plasmacytoma in 1. Complete resection (R0) was histologically confirmed in 18 of 28 patients (64.3%) with resected neoplastic specimens. A horizontal or vertical neoplasia-free margin could not be confirmed in 9 cases and 1 case, respectively. Complete local remission of neoplasia was achieved in 25 of 28 patients (89.3%) who were followed over a median period of 22 months (range 6-44 months). In 1 patient, a metachronous gastric adenocarcinoma was identified 54 weeks after initial WESD. Limitations: Noncontrolled study with a limited number of patients. Conclusion: The use of a water jet–assisted knife simplifies ESD because exchange of devices is rarely needed. WESD promises to be effective and safe. The study demonstrates that the high rates of en bloc resection of early gastric neoplasia reported in Asia can be reproduced in Western referral centers. However, histology may not always confirm complete resection of horizontal tumor margins. In spite of the unfavorable histology results, the high rate of complete local remission of neoplasia promises that surgical treatment of early gastric neoplasia can be avoided in the majority of cases. (Gastrointest Endosc 2012;75:1166-74.) (footnotes appear on last page of article)
1166 GASTROINTESTINAL ENDOSCOPY Volume 75, No. 6 : 2012
www.giejournal.org
Schumacher et al
Conventional techniques of EMR by piecemeal resection are frequently performed but often do not allow histologic confirmation of complete resection. In addition, EMR can be associated with a high recurrence rate, particularly with larger lesions that lead to incomplete resection.1,2 In contrast to EMR, the technique of endoscopic submucosal dissection (ESD) allows en bloc resection of even large lesions (⬎2 cm). ESD of well-defined selected cases of early gastric cancer was recently accepted as an alternative to surgery in Japan.3-5 A few small series also were reported from Western countries.6-11 However, ESD has not yet been standardized, and it is technically difficult, time consuming, and potentially hazardous.7 The aim of this study was to evaluate the efficacy and safety of ESD by use of a new water jet–assisted ESD (WESD) system in patients with early gastric neoplastic lesions. This technology allows pressure controlled injection of fluids through the tip of a HybridKnife (HK; Erbe Elektromedizin GmbH, Tuebingen-Germany). Submucosal injection and circumferential cutting and dissection of lesions as well as coagulation of bleeding can be performed with the same device without need for changing the instrument. These options should accelerate the procedure and may increase its safety and efficacy. Experimental trials in ex vivo and in vivo pig models showed that gastric mucosa can be lifted effectively and safely by placement of the HK on the mucosa and injection of saline solution with pressures between 30 and 70 bar.12-14 Our group recently reported on an effective en bloc resection of esophageal areas in 14 of 14 cases of an Erlanger porcine model.15 The results demonstrated that complete resection, with a significantly lower number of specimens in comparison with EMR, was achieved more frequently with WESD.16 Another application of the HK performed in a porcine colon confirmed the efficacy and safety of the HK in a very thin-walled organ like the colon.17
PATIENTS AND METHODS Patients Between January 2008 and April 2010, 30 patients with early gastric neoplasia were consecutively enrolled in this prospective, single-center, noncontrolled study. Inclusion criteria were as follows: adults (ⱖ18 years) with a histopathologic diagnosis of gastric adenoma or early gastric adenocarcinoma that met the expanded criteria for local resection proposed by Gotoda5: (1) differentiated mucosal cancers of any size without ulceration or scarring, (2) differentiated mucosal cancers ⱕ30 mm with ulceration or scarring, or (3) differentiated cancers invading the submucosa ⱕ30 mm. The diameter of mucosal adenocarcinoma without ulcer findings was limited to a maximum of 60 mm in diameter. The maximum size of those with undifferentiated histology was limited to 20 mm. One patient with undifferentiated histology was included because of the small size of the lesion and the patient’s multiple medical www.giejournal.org
ESD with a water jet–assisted knife
Take-home Message ●
●
Endoscopic submucosal dissection of early gastric neoplasia has not yet been established in Western countries. Some of the technical limitations may be overcome by use of a new device with a combination of a high-pressure water jet and electrosurgical intervention. This study demonstrated safety, efficacy, and a high rate of en bloc resection.
comorbidities. Two additional patients showed undifferentiated histology in the preinterventional biopsies of small focal lesions (⬍20 mm), but final histology showed no malignancy. Exclusion criteria included pregnancy, coagulopathy (international normalized ratio ⬎2.0, platelet count ⬍ 70 mm3), evidence of local or distant metastases, deeper tumor infiltration according to EUS and/or CT scan. The study was performed in compliance with the Helsinki Declaration and good clinical practice. The protocol was approved by the International Medical and Dental Ethics Commission GmbH, Freiburg, Germany. All patients provided written informed consent.
Objectives and criteria of efficacy and safety The primary objective of the study was to achieve complete resection of early gastric neoplasia by using the new ESD-HK. Secondary objectives were to minimize the number of resected specimens required to achieve resection and the rate of recurrence of neoplasia during follow-up. En bloc resection was defined as single-piece resection of the neoplastic lesion according to endoscopic evaluation with or without histologic confirmation of tumor-free margins. Complete resection was defined as en bloc resection with histologic confirmation of tumor-free margins (R0). Piecemeal resection was endoscopically complete removal of the lesion in ⱖ2 pieces, which does not allow a histologic evaluation of the lateral tumor margins (R1). Cases in which the neoplastic lesion could not be completely resected according to endoscopic aspects (R2) were considered failures. An endoscopically visible perforation during the procedure was closed with hemoclips. It was considered a complication if it could not be managed during the same procedure or caused clinical symptoms or required additional interventions. Major bleeding was defined as ongoing bleeding, clinical and/or laboratory test (drop of hemoglobin level ⬎2 g/dL), or signs of bleeding that required endoscopic reintervention and/or blood transfusion.
Procedure Technical devices and parameters of setting. The selection of diagnostic or therapeutic high-definition gastroscopes (EXERA H180 series; Olympus Europe, Hamburg, Germany) was based on the location of the lesion. A Volume 75, No. 6 : 2012 GASTROINTESTINAL ENDOSCOPY 1167
ESD with a water jet–assisted knife
Figure 1. A, HybridKnife, no needle technique with water-jet nozzle: Above, I-type version. Below, T-type version. B, Submucosal injection by the HybridKnife with preselected pressure through an axial water jet channel. Injection and dissection can be performed with the same device.
4-mm transparent distance cap was mounted to the tip of the endoscope for WESD to facilitate positioning of accessories and compression of bleeding sites. The VIO mode ENDO CUT Q 2-3-3 (Erbe Elektromedizin GmbH, Tuebingen, Germany) was used for circumferential cutting as well as for submucosal dissection. The argon plasma coagulation mode forced APC 25 W was used for marking of lesions, and forced COAG E2 60 W was used for coagulation of vessels with hot biopsy forceps or a coagulation grasper. An isotonic saline solution with diluted epinephrine (1:250,000) slightly stained with indigo carmine was used for submucosal injection. The water jet system (ERBEJET 2; Erbe Elektromedizin, Tübingen, Germany) with 120 m inner diameter was used for WESD, with a preselected effect setting of 30 according to results of previous experimental trials in animals.16,17 Circumferential cutting and dissection for ESD were exclusively done with the HK (Fig. 1A and B). This flexible device with an outer diameter of 2.1 mm allows injection or hydrodissection without a needle with a preselected effect setting through a standard working channel of a flexible endoscope. Without changing the instrument, the operator can use the tool alternatively for marking of the targeted lesion, circumferential cutting, dissection, and coagulation by radiofrequency application. Sedation. All endoscopic procedures were done with patients under deep sedation by intravenous application of midazolam (2.5 mg) and propofol (bolus 1 mg/kg followed by 300-350 mg/kg/h continuous infusion). Patients were monitored by a physician trained in intensive care medicine including expertise in resuscitation and assisted or controlled respiration according to German guidelines for sedation in endoscopy.20 1168 GASTROINTESTINAL ENDOSCOPY Volume 75, No. 6 : 2012
Schumacher et al
WESD. The lateral margin of the targeted neoplasm was determined by white light, narrow-band imaging, and chromoendoscopy by use of indigo carmine. All lesions were characterized according to the updated Paris classification.18,19 The tip of the HK was then used for setting coagulation markers at 4-mm distances around the targeted area, with a safety margin of at least 3 mm to the lateral tumor margin. Submucosal injection was done by use of the HK. Submucosal injection/hydrodissection was accomplished without a needle by use of the HK to obtain a clearly visible bulging. Visible vessels were coagulated with the HK, and larger vessels were coagulated with diathermic forceps. Hemoclips were used only in case of failure of these techniques. The interim time of the procedure was taken at the end of complete circumferential cutting of the targeted area. For dissection, submucosal injection was done with the same technique as for circumferential cutting. There was no limitation of the number and registered volume of injections during the procedure. Resection of a partially dissected specimen with a snare was done only in case of inability to achieve complete dissection with a knife. After endoscopic removal of the specimen, the resected area was reinvestigated for residual superficial vessels, which were then coagulated with the HK or if necessary with coagulation forceps (Fig. 2A-F). Preparation and histopathologic evaluation of resected specimens. Each resected specimen was gently stretched and fixed on cork by needles immediately after removal. Completeness of the specimen, including identification of the coagulation marks, was assured. The specimen was sent in 4% neutral buffered formalin for subsequent histopathologic evaluation. The specimens were sectioned completely perpendicularly into 1.5-mm slices for evaluation of type and grade of neoplasia, vertical depth of tumor invasion, and completeness of vertical and lateral resection of neoplasia, with differentiation of tumor-free margins (R0), tumor-infiltrated margins (R1), or horizontally (HM1), basally (VM1), and undetermined margins due to coagulation artifacts or piecemeal resection (Rx, HMx, VMx). In addition, permeation of lymphatic or blood vessels was investigated (Fig. 3A and B). Postprocedural measures. Patients were hospitalized for at least 48 hours after WESD because of the potential risk of delayed complications. Proton pump inhibitors were administered in double standard doses for 6 weeks. In case of Helicobacter pylori infection, eradication was performed. In all cases, gastroscopy was done on the day after the procedure to exclude delayed bleeding and to investigate the resected area for residual superficial vessels, which were treated with coagulation forceps. Follow-up endoscopy was performed according to the protocol: 30 days after the procedure, patients were interviewed via a telephone call for evaluation of complaints or www.giejournal.org
Schumacher et al
ESD with a water jet–assisted knife
Figure 2. A, Suspicious lesion localized at the lesser curvature of the gastric antrum; preoperative histology: adenoma with high-grade dysplasia. B, Marking the margins of the lesion by using the HybridKnife. C, Submucosal injection and beginning of circumferential incision. D, Complete circumferential incision. E, Submucosal dissection. F, Resected area after complete dissection (en bloc).
delayed complications. Follow-up endoscopy was performed after approximately 2 months as part of the study. Further endoscopic surveillance was scheduled at 6 and 12 months after ESD. The aim was to reevaluate the WESD scar for residual or recurrent neoplasia by visual appearance and histology of at least 5 biopsy specimens taken from the scar and the surrounding tissue. In addition, all surveillance endoscopies were done for evaluation of metachronous neoplastic lesions. Number of cases and number of operators. The trial was limited to 30 patients. All procedures were exclusively done by two operators (H.N., B.S.) who www.giejournal.org
have extensive experience in EMR and conventional ESD and performed all cases of previous experimental trials with WESD.6,7,15-17 The number of cases and the limitation of operators were considered to be appropriate to study the objectives of this trial before further evaluation at the level of a multicenter trial or a randomized controlled trial.
Statistical analysis Data were collected and analyzed by means of descriptive statistics (mean and standard deviation as well as median and range). Volume 75, No. 6 : 2012 GASTROINTESTINAL ENDOSCOPY 1169
ESD with a water jet–assisted knife
Schumacher et al
TABLE 1. Baseline characteristics of patients N ⴝ 30 Male/female
17/13
Age, median (range), y
61 (35-93)
ASA score I
12
II
15
III
3
Characteristics of neoplasia Macroscopic type 0-Is
1
0-IIa
10
0-IIb
7
0-IIa and 0-IIc
12
Location of gastric neoplasia* Cardia
6
Body
10
Antrum
14
Lesser curvature
14
Greater curvature
10
Anterior wall
10
Posterior wall
14
Histology of biopsies Figure 3. A, Resected specimen fixed onto cork. B, Histopathologic image showing adenoma with focal transition to well-differentiated adenocarcinoma with lateral expansion (intertubular connection) in the center of the figure confined to the mucosa layer (m-type) (H&E, orig. mag. x 40).
RESULTS
Suspicion of GI stromal tumor
1
Adenoma
8
High-grade intraepithelial neoplasia
1
Low-grade intraepithelial neoplasia
7
Adenocarcinoma
21
Characteristics of patients
Differentiated
18
Clinical pathologic characteristics of the included 30 patients are summarized in Table 1. Tumor location toward the pylorus was separated into gastric cardia, body, and antrum, depending on the most prominent part of the lesion. Determination of the lateral extension of neoplasia was considered to be easy in 25 cases. Delineation of the margins was difficult to determine in 5 cases in spite of use of chromoendoscopy and magnifying endoscopy. This may have been caused by chronic gastritis or intestinal metaplasia of the surrounding mucosa. Safety margins for resection were increased in these cases to reduce the risk of incomplete resection. One patient had preprocedure histology suspicious of a GI stromal tumor. However, the lesion was endoscopically classified as a type 0-IIa lesion with no endosonographic cri-
Undifferentiated
3
1170 GASTROINTESTINAL ENDOSCOPY Volume 75, No. 6 : 2012
Median lesion diameter (range), mm
20 (10-40)
ASA, American Society of Anesthesiologists Physical Status Classification System. *May be more than one per case depending on the extension of the lesion.
teria for a GI stromal tumor. This patient was included so that we could achieve a complete resection of this indeterminate lesion.
Procedural characteristics and outcome at 30 days The procedural characteristics are summarized in Table 2. All lesions could be marked with spotty cautery by using www.giejournal.org
Schumacher et al
ESD with a water jet–assisted knife
TABLE 2. Procedural characteristics and outcome at 30 days N ⴝ 30 Procedure En bloc resection, no. (%) Maximum diameter of specimen, median (range), mm Complete resection in 2 pieces, no. Procedure duration, median (range), min Related to circumferential incision, median (range) Related to dissection, median (range) Amount of fluid injection, median (range), mL
27 (90) 25 (20-70) 3 74 (15-402) 23 (3-95) 52 (11-370) 188 (20-490)
Large vessels or extensive bleeding Treated with hemostatic forceps, no.
10
Treated with hemoclips, no.
1
Intraprocedural complications Visible perforation, no.
3
Closure with hemoclips, no.
3
Complications and outcome at 30 d Morbidity, no. (%)
5/30 (16.6)
Delayed bleeding
2
Postprocedure pain
3
Mortality (not procedure related), no. (%)
1/30 (3.3)
the HK. The median procedural duration for circumferential incision was 23 minutes (range 3-95 minutes) and was 52 minutes (range 11-370 minutes) for dissection. A median amount of 188 (20-490) mL of the prepared saline solution was applied by using the water jet for sequential submucosal injection and cutting. The device had to be changed to coagulation forceps in 10 patients for treatment of larger vessels or more extensive bleeding. These measurements failed in only one case, which required application of hemoclips. Hemoclips also were used for successful closure of visible or suspected small sites of perforation in 3 patients at the beginning of the study. En bloc resection could be achieved in 27 of the 30 patients (90%). Submucosal dissection in a single-piece fashion was discontinued in 3 cases (10%) because of a difficult location at the cardia. The remaining targeted area was resected with a snare, which allowed complete removal in two pieces. www.giejournal.org
Postprocedure complications and outcome at 30 days are shown in Table 2. A 93-year-old patient unexpectedly and suddenly died the night after WESD. The procedure had been uneventful, and routinely determined laboratory tests as well as clinical parameters were normal before the patient went to sleep. The patient had not reported any postprocedure complaints. A sudden cardiac arrest is the most likely cause of death. Autopsy was refused by relatives of the patient. Routine gastroscopy was performed in all other patients on the day after the procedure. There was no evidence of bleeding or visible vessels so that no interventions were required. In spite of normal findings, delayed bleeding occurred in 2 patients 2 and 3 days after ESD, respectively. Blood transfusions were administered, and definitive hemostasis could be achieved by endoscopic identification of the bleeding sites at the resection area and placement of hemoclips. Plain abdominal radiographs or CT scans revealed free air in the abdominal cavity in the 3 patients, with intraoperative diagnosis and treatment of perforation. Two of these cases remained asymptomatic so that these events were not considered as complications. There was no evidence of peritonitis and no need for any additional interventions in the 3 patients with pain after WESD, and the subsequent clinical course was uneventful.
Histopathology Histopathology of the specimens showed adenoma in 7 cases and adenocarcinoma in 20 (Table 3). No neoplasia could be detected in 2 cases. Negative follow-up examinations suggested that the previously diagnosed adenocarcinoma had been either completely removed by biopsies or ablated by WESD-related diathermia or the lesion was too small for histologic identification. In 1 patient, extramedullary plasmacytoma was diagnosed. Histology confirmed complete resection (R0) in 18 of the 28 patients with neoplastic lesions (64.3%). In 1 patient, a welldifferentiated adenocarcinoma infiltrated the submucosa deeper than 0.5 mm (sm 2). It was histologically described as an R1 situation, with histologic incomplete horizontal (HM1) and vertical (VM1) resection. In spite of the potential risk of lymph node metastases, surgery was not considered because of severe comorbidities. Complete resection could not be confirmed in the remaining 9 patients. Resection had been performed in two pieces in 3 of these cases, and in another 6 cases parts of lateral resection margins were not free (R1: HM1) or were indeterminate of neoplasia (Rx). Risk of lymph node metastasis seemed to be low because vertical tumor infiltration was limited to the mucosal layer and showed free margins in all specimens (VM0). In view of the low risk, all 10 patients with histologic incomplete resection refused surgery and were scheduled for follow-up examinations to exclude or treat residual neoplasia at the resection scar. Volume 75, No. 6 : 2012 GASTROINTESTINAL ENDOSCOPY 1171
ESD with a water jet–assisted knife
Schumacher et al
TABLE 3. Histology of resected specimens Chronic gastritis, no neoplasia, no.
2
Plasmacytoma, no.
1
Adenoma with low-grade intraepithelial neoplasia, no.
7
Adenocarcinoma, no.
20
Grading, no. Differentiated (G1, G2)
15
Undifferentiated (G3)
1
Signet-ring-cell carcinoma
4
Deepest layer of infiltration, no. Mucosa
17
Submucosa (sm1)
2
Submucosa (sm2)
1
Complete resection (R0) of neoplasia, no. (%)
18/28 (64.3)
complete local remission of neoplasia. Three patients underwent attempts at endoscopic resection of a local recurrence. One of them was treated successful by a second ESD during a follow-up period of 19 months, which achieved complete remission. The histology of the first specimen after ESD had shown a complete resection of a signet-ring-cell carcinoma. The retreatment remained incomplete in the 2 other patients with local recurrence, and gastrectomy was performed in both of them. In one of these cases, initial histology of WESD had demonstrated incomplete resection of an adenocarcinoma infiltrating the superficial layer of the submucosa (sm1). Lack of a lifting sign caused failure of endoscopic retreatment in the other case. A third patient underwent elective gastrectomy because of a metachronous lesion that was identified 54 weeks after initial ESD with an R0 resection. Histology showed adenocarcinoma, and surgery was performed at the request of the patient. The two patients with no neoplasia in the initial ESD procedure were followed for up to 49 months with no evidence of a recurrent or metachronous lesion.
Incomplete resection of neoplasia At lateral tumor margins, no.
9
At lateral and vertical tumor margins, no.
1
G1, differentiated; G2, moderate differentiated; G3, undifferentiated; sm1, infiltration of submucosal layer ⬍_500m; R0, tumorfree margins lateral and vertical.
TABLE 4. Follow-up* of more than 30 days after WESD Patients alive at 30 d, no.
29
Lost to follow-up, no.
1
Complete local remission of neoplasia, no. (%) Recurrence of neoplasia, no.
25/28 (89.3) 3
Endoscopic retreatment, no.
1
Gastrectomy, no.
2
Metachronous gastric adenocarcinoma, no.
1
Gastrectomy, no.
1
WESD, Water jet–assisted endoscopic submucosal dissection. *Median period of 22 months (range 6-44 months).
Medium-term outcome The medium-term outcome of the 29 patients who were alive at 30 days is summarized in Table 4. One of these patients was lost to follow-up. The remaining 28 patients were followed over a median period of 22 months. Twenty-five (89.3%) patients had a 1172 GASTROINTESTINAL ENDOSCOPY Volume 75, No. 6 : 2012
DISCUSSION ESD acceptance in Western countries is low because it is a more time-consuming, difficult procedure, which is correlated with a higher rate of complication. To simplify the technique, a German company (ERBE Elektromedizin GmbH) developed a new device called the HybridKnife. This technique was first evaluated in a porcine model followed by a randomized study comparing EMR and ESD in pigs.15,16 These two experimental studies showed that the use of the new water jet dissector in combination with the HK accelerates the ESD procedure and increases the safety and efficacy. Our aim in this prospective study was to evaluate the efficacy and safety of ESD in patients with early gastric neoplastic lesions. The results of the present study suggest that the technique using the HK for ESD achieves high rates of endoscopic complete resection (100%) and en bloc resection (90%). Twenty-nine of 30 patients met the expanded Japanese criteria for ESD based on the ESD specimen. One patient with undifferentiated histology on the ESD specimen was included even though his condition was outside of the expanded criteria because of the small size of the lesion and the patient’s multiple medical comorbidities. In 3 patients, en bloc resection was not possible because the lesion was located in the upper part of the stomach. EMR was done in these patients, with complete resection in two pieces. Histologic incomplete resection due to laterally indeterminate or infiltrated tumor margins was diagnosed in 10 patients. Our complication rate was low because of the ESD procedure. We had delayed bleeding in one case and 3 visible perforations during the procedures. These could be www.giejournal.org
Schumacher et al
sufficiently closed by hemoclips without need for surgery. The low complication rate can be attributed to the easily repeatable hydrodissection, which maintains the tissue elevation like a cushion before cutting. We needed a median amount of fluid injection of 188 mL for a lesion dissection. The potential risk of perforation may be lower with a larger amount of saline solution injection, as previously reported.16 The relevant bleeding rate also was low. The HK was used for short bursts of coagulation of visible vessels or minor bleeding. The median total procedure time was 74 minutes, which is comparable to other published data with different devices.2-4,23,24 Compared with Western studies using ESD, our procedure time was less than previously reported in these studies: the median procedure times were 85 minutes,22 108 minutes,11 and 161 minutes.10 Compared with EMR, the ESD procedure time is significantly longer.16 As we showed in our animal trial, a tendency of time shortening is obvious with increasing experience, which means a learning curve is also a significant factor for the procedure time. Early gastric cancer is defined as mucosal or submucosal invasive cancer (T1 cancer) irrespective of the presence of lymph node metastasis. Endoscopic treatment should be performed only in patients with early gastric cancer who have no risk of lymph node metastasis, and the resection has to be carried out en bloc. Endoscopic mucosal resection is potentially curative for treatment of early neoplasia, but histopathology of the resected specimen must show no tumor infiltration in the submucosa deeper than 500 m. When lesions are larger than 20 mm in diameter, EMR usually must be performed in a piecemeal fashion that is associated with a high rate of local recurrence.21,22 ESD may overcome the well-known limitations of EMR. ESD allows precise resection in an en bloc procedure.2,25,26 A variety of prospective and retrospective trials from Japan showed that ESD achieves higher rates of curative resection of early gastric cancer than EMR.23-25,27-29 The guidelines for ESD are constantly being updated and newly discussed. The acceptable indications for ESD of early gastric cancer are based on the guidelines of the Japanese Gastric Cancer Association.5,21,29,30 After evaluation of these criteria, new criteria to expand the indications for endoscopic treatment were recommended.5,21,29,30 These studies showed that en bloc, complete, and curative resection rates were significantly higher in the guideline group compared with the expanded group, but there was no difference in overall survival.30,31 Compared with the Japanese data, European experience with ESD is rare. In a recently published German study, R0 en bloc resection for all lesions (early gastric cancers and adenomas) was reported in 74% of cases. Resection rates were lower in patients with early gastric cancer with expanded criteria (68, 6%).10 In our study, 30 patients met the criteria for ESD. The en bloc resection rate was 27 of 30 (90%). The R0 rate of www.giejournal.org
ESD with a water jet–assisted knife
resected neoplasia was 18 of 28 (64.3%). Compared with the Japanese reports, there is less published European experience with ESD.10,32 The discrepancy of a high en bloc resection rate but low rates of histologically confirmed R0 margins is likely caused by inappropriate delineation of lateral tumor margins or too small distances of coagulation markers at the tumor periphery. The majority of the lesions were classified morphologically as type IIa and type IIa⫹c. Even using high-resolution endoscopy with narrow-band imaging may underestimate lateral tumor extension. In spite of these unfavorable histologic results, only 3 recurrences and 1 metachronous cancer were observed during a median follow-up period of 22 months (range 6-44 months). All recurrences developed after 3 months; one metachronous cancer was detected after 13 months. One of the patients with a recurrent tumor was treated by an endoscopic procedure and is still in remission. Two other patients underwent surgery because of deeper infiltration and failure with endoscopic retreatment. One patient developed a metachronous lesion with histologically confirmed early cancer. At the request of the patient, surgery was performed. In summary, this first prospective study using a new device for ESD in patients with early gastric neoplastic lesions demonstrated a high rate of en bloc resection. The procedure times were acceptable (mean 74 minutes), and all complications could be endoscopically managed. The R0 rate of resected neoplasia was 64.3%, which is lower than expected. Complete local remission of neoplasia was achieved in 25 of 28 patients (89.3%), who were followed over a median period of 22 months (range 6-44 months). Three patients underwent gastrectomy because of endoscopically inaccessible recurrences or a metachronous lesion. To establish clear indications for ESD and efficient ESD procedures, additional prospective studies are necessary, with more patients included. Further improvement of the technique seems to be essential.
REFERENCES 1. Korenaga D, Orita H, Maekawa S, et al. Pathological appearance of the stomach after endoscopic mucosal resection for early gastric cancer. Br J Surg 1997;84:1563-6. 2. Ono H, Kondo H, Gotoda T, et al. Endoscopic mucosal resection for treatment of early gastric cancer. Gut 2001;48:225-9. 3. Miyamoto S, Muto M, Hamamoto Y, et al. A new technique for endoscopic mucosal resection with an insulated-tip electrosurgical knife improves the completeness of resection of intramucosal gastric neoplasms. Gastrointest Endosc 2002;55:576-81. 4. Yamamoto H, Kawata H, Sunada K, et al. Success rate of curative endoscopic mucosal resection with circumferential mucosa incision assisted by submucosal injection of sodium hyaluronate. Gastrointest Endosc 2002;56:507-12. 5. Gotoda T. Endoscopic resection of early gastric cancer: the Japanese perspective. Curr Opin Gastroenterol 2006;22:561-9. 6. Rösch T, Sarbia M, Schumacher B, et al. Attempted endoscopic en bloc resection of mucosal and submucosal tumors using insulated-tip knives: a pilot series. Endoscopy 2004;36:788-801.
Volume 75, No. 6 : 2012 GASTROINTESTINAL ENDOSCOPY 1173
ESD with a water jet–assisted knife
7. Neuhaus H, Costamagna G, Devière J, et al. Endoscopic submucosal resection (ESD) of early neoplastic gastric lesions using a new double channel endoscope (the “R-scope”). Endoscopy 2006;38:1016-23. 8. Coda S, Trentino P, Antonellis F, et al. A Western single-center experience with endoscopic submucosal dissection for early gastrointestinal cancers. Gastric Cancer 2010;13:258-63. 9. Probst A, Golger D, Arnholdt H et al. Endoscopic submucosal dissection of early cancers, flat adenomas and submucosal tumors in the gastrointestinal tract. Clin Gastroenterol Hepatol 2009;7:149-55. 10. Probst A, Pommer B, Golger D, et al. Endoscopic submucosal dissection in gastric neoplasia– experience from a European center. Endoscopy 2010;42:1037-44. 11. Riberiro-Mourao F, Pimentel-Nunes P, Dinis-Ribeiro, et al. Endoscopic submucosal dissection for gastric lesions: results of a European inquiry. Endoscopy 2010;42:814-9. 12. Kaehler GFBA, Sold MG, Fischer K, et al. Selective fluid cushion in the submucosal layer by water jet: advantage for endoscopic mucosal resection. Eur Surg Res 2007;39:93-7. 13. Kaehler GFBA, Colet PH, Sold MG, et al. Waterjet for mucosal elevation in the GI tract–first clinical trial [abstract]. Gastrointest Endosc 2007;65: AB288. 14. Kaehler GFBA, Sold MG, Lingenfelder T, et al. Needle knife and water jet combined as a hybrid knife–a comparison with standard knife in an in vivo animal study [abstract]. Gastrointest Endosc 2007;65:AB285. 15. Schumacher B, Neuhaus H, Enderle MD. Experimental use of a new device for mucosectomy. Acta Endoscopica 2007;5:673-8. 16. Neuhaus H, Wirths K, Enderle MD, et al. Randomized controlled study of EMR versus endoscopic submucosal dissection with a water-jet hybridknife of esophageal lesions in a porcine model. Gastrointest Endosc 2009;70:112-20. 17. Yahagi N, Neuhaus H, Schumacher B, et al. Comparison of standard endoscopic submucosal dissection (ESD) versus an optimized ESD technique for colon: an animal study. Endoscopy 2009;41:340-5. 18. Participants in the Paris workshop. The Paris endoscopic classification of superficial neoplastic lesions: esophagus, stomach and colon: November 30 to December 1, 2002. Gastrointest Endosc 2003;58:S3-43. 19. Endoscopic Classification Review Group. Update on the Paris classification of superficial neoplastic lesions in the digestive tract. Endoscopy 2005;37:570-8. 20. Riphaus A, Wehrmann T, Weber B, et al. S3-guidelines sedation in gastrointestinal endoscopy. Z Gastroenterol 2008;46:1298-330. 21. Soetikno R, Kaltenbach T, Yeh R, et al. Endoscopic mucosal resection for early cancers of the upper gastrointestinal tract. J Clin Oncol 2005;23: 4490-8. 22. Hondo FY, Maluf-Filho F, Kishi HS, et al. Predictive factors for local recurrence and incomplete resection of early gastric cancer treated by endoscopic resection: a Western experience. Can J Gastroenterol 2009;23: 357-63.
1174 GASTROINTESTINAL ENDOSCOPY Volume 75, No. 6 : 2012
Schumacher et al
23. Oka S, Tanaka S, Kaneko I, et al. Advantage of endoscopic submucosal dissection compared with EMR for early gastric cancer. Gastrointest Endosc 2006;64:877-83. 24. Oda I, Saito D, Tada M, et al. A multicenter retrospective study of endoscopic resection for early gastric cancer. Gastric Cancer 2006;9:262-70. 25. Cao Y, Liao C, Tan A, et al. Meta-analysis of endoscopic submucosal dissection versus endoscopic mucosal resection for tumors of the gastrointestinal tract. Endoscopy 2009;41:751-7. 26. Yamamoto H, Yahagi N, Oyama T, et al. Usefulness and safety of 0.4% sodium hyaluronate solution as a submucosal fluid “cushion” in endoscopic resection for gastric neoplasms: a prospective multicenter trial. Gastrointest Endosc 2008;67:830-9. 27. Chaves DM, Maluf Filho F, de Moura EG, et al. Endoscopic submucosal dissection for the treatment of early esophageal and gastric cancer–initial experience of a Western center. Clinics (Sao Paulo) 2010;65:377-82. 28. Hamanaka H, Gotoda T. Endoscopic resection for early cancer and future expectations. Dig Endosc 2005;17:275-85. 29. 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;3:219-25. 30. Yamaguchi N, Isomoto H, Fukuda E, et al. Clinical outcomes of endoscopic submucosal dissection for early cancer by indication criteria. Digestion 2009;80:173-81. 31. Isomoto H, Shikuwa S, Yamaguchi N, et al. Endoscopic submucosal dissection for early gastric cancer: a large-scale feasibility study. Gut 2009; 58:331-6. 32. Farhat S, Chaussade S, Ponchon T, et al. Endoscopic submucosal dissection in a European setting: a multi-institutional report of a technique in development. Endoscopy 2011;43:664-70.
Abbreviations: ESD, endoscopic submucosal dissection; HK, HybridKnife; WESD, water jet–assisted ESD. DISCLOSURE: M. Enderle is an employee of ERBE Elektromedizin. No other financial relationships relevant to this publication were disclosed. Copyright © 2012 by the American Society for Gastrointestinal Endoscopy 0016-5107/$36.00 doi:10.1016/j.gie.2012.02.027 Received December 16, 2011. Accepted February 13, 2012. Current affiliations: Department of Gastroenterology (1), EVK Evangelisches Krankenhaus Düsseldorf, Institute of Pathology (2), Klinikum Bayreuth, GmbH, ERBE Elektromedizin GmbH (3), Tübingen, Germany. Reprint requests: Brigitte Schumacher, MD, Department of Internal Medicine, Evangelisches Krankenhaus Düsseldorf, Kirchfeldstr 40, 40217 Düsseldorf, Germany.
www.giejournal.org
Endoscopic submucosal dissection of early gastric neoplasia with a water jet–assisted knife: a Western, single-center experience Brigitte Schumacher, MD,1 Jean-Pierre Charton, MD,1 Thomas Nordmann, MD,1 Michael Vieth, MD,2 Markus Enderle,3 Horst Neuhaus, MD1 Düsseldorf, Germany
Background: Endoscopic submucosal dissection (ESD) of early gastric neoplasia has not yet been established in Western countries because of a lack of data and the difficult, time-consuming, and hazardous nature of the method. Some of the technical limitations may be overcome by use of a water jet–assisted knife, which allows a combination of a high-pressure water jet and electrosurgical interventions. Objective: To evaluate the efficacy and safety of water jet–assisted ESD (WESD) with a water jet–assisted knife in selected patients with early gastric neoplasia. Design: Single-center, prospective study. Patients: This study involved 29 consecutive patients (13 female; median age 61 years; age range 35-93 years) with early gastric neoplasia that met the expanded criteria of the Japanese Gastric Cancer Association. Histology of biopsies had shown gastric adenocarcinoma in 21 cases, adenoma in 8 case, and suspicion of a GI stromal tumor in 1 case. The median maximal diameter of the lesions was 20 mm (range 10-40 mm). Intervention: All procedures were done with patients under sedation with propofol. The water jet–assisted knife was used for setting coagulation markers around the neoplastic lesions, then for circumferential incision and dissection in combination with repeated submucosal injection of saline solution with a water jet system. Bleeding was treated with diathermia by use of the water jet–assisted knife or hemostatic forceps in case of failure or larger vessels. Clips were used for closure of perforations. Main Outcome Measurements: Complete resection of neoplasia, procedure time, complication and recurrence rates. Results: According to endoscopic criteria, complete resection of the targeted area could be achieved in all cases, with an en bloc resection rate of 90%. The median procedure duration was 74 minutes (range 15-402 minutes). Exchange of the device was needed in only 10 cases because of severe bleeding from larger vessels, which could be managed by use of hemostatic forceps. The 30-day morbidity rate was 4 of 30 (13.8%) because of postprocedure pain in 3 cases and delayed bleeding in 1 case. A 93-year-old patient died the night after WESD without evidence of a procedure-related complication. Histology of the resected specimens showed adenocarcinoma in 20 cases, adenoma in 7, no neoplasia in 2, and a plasmacytoma in 1. Complete resection (R0) was histologically confirmed in 18 of 28 patients (64.3%) with resected neoplastic specimens. A horizontal or vertical neoplasia-free margin could not be confirmed in 9 cases and 1 case, respectively. Complete local remission of neoplasia was achieved in 25 of 28 patients (89.3%) who were followed over a median period of 22 months (range 6-44 months). In 1 patient, a metachronous gastric adenocarcinoma was identified 54 weeks after initial WESD. Limitations: Noncontrolled study with a limited number of patients. Conclusion: The use of a water jet–assisted knife simplifies ESD because exchange of devices is rarely needed. WESD promises to be effective and safe. The study demonstrates that the high rates of en bloc resection of early gastric neoplasia reported in Asia can be reproduced in Western referral centers. However, histology may not always confirm complete resection of horizontal tumor margins. In spite of the unfavorable histology results, the high rate of complete local remission of neoplasia promises that surgical treatment of early gastric neoplasia can be avoided in the majority of cases. (Gastrointest Endosc 2012;75:1166-74.) (footnotes appear on last page of article)
1166 GASTROINTESTINAL ENDOSCOPY Volume 75, No. 6 : 2012
www.giejournal.org
Schumacher et al
Conventional techniques of EMR by piecemeal resection are frequently performed but often do not allow histologic confirmation of complete resection. In addition, EMR can be associated with a high recurrence rate, particularly with larger lesions that lead to incomplete resection.1,2 In contrast to EMR, the technique of endoscopic submucosal dissection (ESD) allows en bloc resection of even large lesions (⬎2 cm). ESD of well-defined selected cases of early gastric cancer was recently accepted as an alternative to surgery in Japan.3-5 A few small series also were reported from Western countries.6-11 However, ESD has not yet been standardized, and it is technically difficult, time consuming, and potentially hazardous.7 The aim of this study was to evaluate the efficacy and safety of ESD by use of a new water jet–assisted ESD (WESD) system in patients with early gastric neoplastic lesions. This technology allows pressure controlled injection of fluids through the tip of a HybridKnife (HK; Erbe Elektromedizin GmbH, Tuebingen-Germany). Submucosal injection and circumferential cutting and dissection of lesions as well as coagulation of bleeding can be performed with the same device without need for changing the instrument. These options should accelerate the procedure and may increase its safety and efficacy. Experimental trials in ex vivo and in vivo pig models showed that gastric mucosa can be lifted effectively and safely by placement of the HK on the mucosa and injection of saline solution with pressures between 30 and 70 bar.12-14 Our group recently reported on an effective en bloc resection of esophageal areas in 14 of 14 cases of an Erlanger porcine model.15 The results demonstrated that complete resection, with a significantly lower number of specimens in comparison with EMR, was achieved more frequently with WESD.16 Another application of the HK performed in a porcine colon confirmed the efficacy and safety of the HK in a very thin-walled organ like the colon.17
PATIENTS AND METHODS Patients Between January 2008 and April 2010, 30 patients with early gastric neoplasia were consecutively enrolled in this prospective, single-center, noncontrolled study. Inclusion criteria were as follows: adults (ⱖ18 years) with a histopathologic diagnosis of gastric adenoma or early gastric adenocarcinoma that met the expanded criteria for local resection proposed by Gotoda5: (1) differentiated mucosal cancers of any size without ulceration or scarring, (2) differentiated mucosal cancers ⱕ30 mm with ulceration or scarring, or (3) differentiated cancers invading the submucosa ⱕ30 mm. The diameter of mucosal adenocarcinoma without ulcer findings was limited to a maximum of 60 mm in diameter. The maximum size of those with undifferentiated histology was limited to 20 mm. One patient with undifferentiated histology was included because of the small size of the lesion and the patient’s multiple medical www.giejournal.org
ESD with a water jet–assisted knife
Take-home Message ●
●
Endoscopic submucosal dissection of early gastric neoplasia has not yet been established in Western countries. Some of the technical limitations may be overcome by use of a new device with a combination of a high-pressure water jet and electrosurgical intervention. This study demonstrated safety, efficacy, and a high rate of en bloc resection.
comorbidities. Two additional patients showed undifferentiated histology in the preinterventional biopsies of small focal lesions (⬍20 mm), but final histology showed no malignancy. Exclusion criteria included pregnancy, coagulopathy (international normalized ratio ⬎2.0, platelet count ⬍ 70 mm3), evidence of local or distant metastases, deeper tumor infiltration according to EUS and/or CT scan. The study was performed in compliance with the Helsinki Declaration and good clinical practice. The protocol was approved by the International Medical and Dental Ethics Commission GmbH, Freiburg, Germany. All patients provided written informed consent.
Objectives and criteria of efficacy and safety The primary objective of the study was to achieve complete resection of early gastric neoplasia by using the new ESD-HK. Secondary objectives were to minimize the number of resected specimens required to achieve resection and the rate of recurrence of neoplasia during follow-up. En bloc resection was defined as single-piece resection of the neoplastic lesion according to endoscopic evaluation with or without histologic confirmation of tumor-free margins. Complete resection was defined as en bloc resection with histologic confirmation of tumor-free margins (R0). Piecemeal resection was endoscopically complete removal of the lesion in ⱖ2 pieces, which does not allow a histologic evaluation of the lateral tumor margins (R1). Cases in which the neoplastic lesion could not be completely resected according to endoscopic aspects (R2) were considered failures. An endoscopically visible perforation during the procedure was closed with hemoclips. It was considered a complication if it could not be managed during the same procedure or caused clinical symptoms or required additional interventions. Major bleeding was defined as ongoing bleeding, clinical and/or laboratory test (drop of hemoglobin level ⬎2 g/dL), or signs of bleeding that required endoscopic reintervention and/or blood transfusion.
Procedure Technical devices and parameters of setting. The selection of diagnostic or therapeutic high-definition gastroscopes (EXERA H180 series; Olympus Europe, Hamburg, Germany) was based on the location of the lesion. A Volume 75, No. 6 : 2012 GASTROINTESTINAL ENDOSCOPY 1167
ESD with a water jet–assisted knife
Figure 1. A, HybridKnife, no needle technique with water-jet nozzle: Above, I-type version. Below, T-type version. B, Submucosal injection by the HybridKnife with preselected pressure through an axial water jet channel. Injection and dissection can be performed with the same device.
4-mm transparent distance cap was mounted to the tip of the endoscope for WESD to facilitate positioning of accessories and compression of bleeding sites. The VIO mode ENDO CUT Q 2-3-3 (Erbe Elektromedizin GmbH, Tuebingen, Germany) was used for circumferential cutting as well as for submucosal dissection. The argon plasma coagulation mode forced APC 25 W was used for marking of lesions, and forced COAG E2 60 W was used for coagulation of vessels with hot biopsy forceps or a coagulation grasper. An isotonic saline solution with diluted epinephrine (1:250,000) slightly stained with indigo carmine was used for submucosal injection. The water jet system (ERBEJET 2; Erbe Elektromedizin, Tübingen, Germany) with 120 m inner diameter was used for WESD, with a preselected effect setting of 30 according to results of previous experimental trials in animals.16,17 Circumferential cutting and dissection for ESD were exclusively done with the HK (Fig. 1A and B). This flexible device with an outer diameter of 2.1 mm allows injection or hydrodissection without a needle with a preselected effect setting through a standard working channel of a flexible endoscope. Without changing the instrument, the operator can use the tool alternatively for marking of the targeted lesion, circumferential cutting, dissection, and coagulation by radiofrequency application. Sedation. All endoscopic procedures were done with patients under deep sedation by intravenous application of midazolam (2.5 mg) and propofol (bolus 1 mg/kg followed by 300-350 mg/kg/h continuous infusion). Patients were monitored by a physician trained in intensive care medicine including expertise in resuscitation and assisted or controlled respiration according to German guidelines for sedation in endoscopy.20 1168 GASTROINTESTINAL ENDOSCOPY Volume 75, No. 6 : 2012
Schumacher et al
WESD. The lateral margin of the targeted neoplasm was determined by white light, narrow-band imaging, and chromoendoscopy by use of indigo carmine. All lesions were characterized according to the updated Paris classification.18,19 The tip of the HK was then used for setting coagulation markers at 4-mm distances around the targeted area, with a safety margin of at least 3 mm to the lateral tumor margin. Submucosal injection was done by use of the HK. Submucosal injection/hydrodissection was accomplished without a needle by use of the HK to obtain a clearly visible bulging. Visible vessels were coagulated with the HK, and larger vessels were coagulated with diathermic forceps. Hemoclips were used only in case of failure of these techniques. The interim time of the procedure was taken at the end of complete circumferential cutting of the targeted area. For dissection, submucosal injection was done with the same technique as for circumferential cutting. There was no limitation of the number and registered volume of injections during the procedure. Resection of a partially dissected specimen with a snare was done only in case of inability to achieve complete dissection with a knife. After endoscopic removal of the specimen, the resected area was reinvestigated for residual superficial vessels, which were then coagulated with the HK or if necessary with coagulation forceps (Fig. 2A-F). Preparation and histopathologic evaluation of resected specimens. Each resected specimen was gently stretched and fixed on cork by needles immediately after removal. Completeness of the specimen, including identification of the coagulation marks, was assured. The specimen was sent in 4% neutral buffered formalin for subsequent histopathologic evaluation. The specimens were sectioned completely perpendicularly into 1.5-mm slices for evaluation of type and grade of neoplasia, vertical depth of tumor invasion, and completeness of vertical and lateral resection of neoplasia, with differentiation of tumor-free margins (R0), tumor-infiltrated margins (R1), or horizontally (HM1), basally (VM1), and undetermined margins due to coagulation artifacts or piecemeal resection (Rx, HMx, VMx). In addition, permeation of lymphatic or blood vessels was investigated (Fig. 3A and B). Postprocedural measures. Patients were hospitalized for at least 48 hours after WESD because of the potential risk of delayed complications. Proton pump inhibitors were administered in double standard doses for 6 weeks. In case of Helicobacter pylori infection, eradication was performed. In all cases, gastroscopy was done on the day after the procedure to exclude delayed bleeding and to investigate the resected area for residual superficial vessels, which were treated with coagulation forceps. Follow-up endoscopy was performed according to the protocol: 30 days after the procedure, patients were interviewed via a telephone call for evaluation of complaints or www.giejournal.org
Schumacher et al
ESD with a water jet–assisted knife
Figure 2. A, Suspicious lesion localized at the lesser curvature of the gastric antrum; preoperative histology: adenoma with high-grade dysplasia. B, Marking the margins of the lesion by using the HybridKnife. C, Submucosal injection and beginning of circumferential incision. D, Complete circumferential incision. E, Submucosal dissection. F, Resected area after complete dissection (en bloc).
delayed complications. Follow-up endoscopy was performed after approximately 2 months as part of the study. Further endoscopic surveillance was scheduled at 6 and 12 months after ESD. The aim was to reevaluate the WESD scar for residual or recurrent neoplasia by visual appearance and histology of at least 5 biopsy specimens taken from the scar and the surrounding tissue. In addition, all surveillance endoscopies were done for evaluation of metachronous neoplastic lesions. Number of cases and number of operators. The trial was limited to 30 patients. All procedures were exclusively done by two operators (H.N., B.S.) who www.giejournal.org
have extensive experience in EMR and conventional ESD and performed all cases of previous experimental trials with WESD.6,7,15-17 The number of cases and the limitation of operators were considered to be appropriate to study the objectives of this trial before further evaluation at the level of a multicenter trial or a randomized controlled trial.
Statistical analysis Data were collected and analyzed by means of descriptive statistics (mean and standard deviation as well as median and range). Volume 75, No. 6 : 2012 GASTROINTESTINAL ENDOSCOPY 1169
ESD with a water jet–assisted knife
Schumacher et al
TABLE 1. Baseline characteristics of patients N ⴝ 30 Male/female
17/13
Age, median (range), y
61 (35-93)
ASA score I
12
II
15
III
3
Characteristics of neoplasia Macroscopic type 0-Is
1
0-IIa
10
0-IIb
7
0-IIa and 0-IIc
12
Location of gastric neoplasia* Cardia
6
Body
10
Antrum
14
Lesser curvature
14
Greater curvature
10
Anterior wall
10
Posterior wall
14
Histology of biopsies Figure 3. A, Resected specimen fixed onto cork. B, Histopathologic image showing adenoma with focal transition to well-differentiated adenocarcinoma with lateral expansion (intertubular connection) in the center of the figure confined to the mucosa layer (m-type) (H&E, orig. mag. x 40).
RESULTS
Suspicion of GI stromal tumor
1
Adenoma
8
High-grade intraepithelial neoplasia
1
Low-grade intraepithelial neoplasia
7
Adenocarcinoma
21
Characteristics of patients
Differentiated
18
Clinical pathologic characteristics of the included 30 patients are summarized in Table 1. Tumor location toward the pylorus was separated into gastric cardia, body, and antrum, depending on the most prominent part of the lesion. Determination of the lateral extension of neoplasia was considered to be easy in 25 cases. Delineation of the margins was difficult to determine in 5 cases in spite of use of chromoendoscopy and magnifying endoscopy. This may have been caused by chronic gastritis or intestinal metaplasia of the surrounding mucosa. Safety margins for resection were increased in these cases to reduce the risk of incomplete resection. One patient had preprocedure histology suspicious of a GI stromal tumor. However, the lesion was endoscopically classified as a type 0-IIa lesion with no endosonographic cri-
Undifferentiated
3
1170 GASTROINTESTINAL ENDOSCOPY Volume 75, No. 6 : 2012
Median lesion diameter (range), mm
20 (10-40)
ASA, American Society of Anesthesiologists Physical Status Classification System. *May be more than one per case depending on the extension of the lesion.
teria for a GI stromal tumor. This patient was included so that we could achieve a complete resection of this indeterminate lesion.
Procedural characteristics and outcome at 30 days The procedural characteristics are summarized in Table 2. All lesions could be marked with spotty cautery by using www.giejournal.org
Schumacher et al
ESD with a water jet–assisted knife
TABLE 2. Procedural characteristics and outcome at 30 days N ⴝ 30 Procedure En bloc resection, no. (%) Maximum diameter of specimen, median (range), mm Complete resection in 2 pieces, no. Procedure duration, median (range), min Related to circumferential incision, median (range) Related to dissection, median (range) Amount of fluid injection, median (range), mL
27 (90) 25 (20-70) 3 74 (15-402) 23 (3-95) 52 (11-370) 188 (20-490)
Large vessels or extensive bleeding Treated with hemostatic forceps, no.
10
Treated with hemoclips, no.
1
Intraprocedural complications Visible perforation, no.
3
Closure with hemoclips, no.
3
Complications and outcome at 30 d Morbidity, no. (%)
5/30 (16.6)
Delayed bleeding
2
Postprocedure pain
3
Mortality (not procedure related), no. (%)
1/30 (3.3)
the HK. The median procedural duration for circumferential incision was 23 minutes (range 3-95 minutes) and was 52 minutes (range 11-370 minutes) for dissection. A median amount of 188 (20-490) mL of the prepared saline solution was applied by using the water jet for sequential submucosal injection and cutting. The device had to be changed to coagulation forceps in 10 patients for treatment of larger vessels or more extensive bleeding. These measurements failed in only one case, which required application of hemoclips. Hemoclips also were used for successful closure of visible or suspected small sites of perforation in 3 patients at the beginning of the study. En bloc resection could be achieved in 27 of the 30 patients (90%). Submucosal dissection in a single-piece fashion was discontinued in 3 cases (10%) because of a difficult location at the cardia. The remaining targeted area was resected with a snare, which allowed complete removal in two pieces. www.giejournal.org
Postprocedure complications and outcome at 30 days are shown in Table 2. A 93-year-old patient unexpectedly and suddenly died the night after WESD. The procedure had been uneventful, and routinely determined laboratory tests as well as clinical parameters were normal before the patient went to sleep. The patient had not reported any postprocedure complaints. A sudden cardiac arrest is the most likely cause of death. Autopsy was refused by relatives of the patient. Routine gastroscopy was performed in all other patients on the day after the procedure. There was no evidence of bleeding or visible vessels so that no interventions were required. In spite of normal findings, delayed bleeding occurred in 2 patients 2 and 3 days after ESD, respectively. Blood transfusions were administered, and definitive hemostasis could be achieved by endoscopic identification of the bleeding sites at the resection area and placement of hemoclips. Plain abdominal radiographs or CT scans revealed free air in the abdominal cavity in the 3 patients, with intraoperative diagnosis and treatment of perforation. Two of these cases remained asymptomatic so that these events were not considered as complications. There was no evidence of peritonitis and no need for any additional interventions in the 3 patients with pain after WESD, and the subsequent clinical course was uneventful.
Histopathology Histopathology of the specimens showed adenoma in 7 cases and adenocarcinoma in 20 (Table 3). No neoplasia could be detected in 2 cases. Negative follow-up examinations suggested that the previously diagnosed adenocarcinoma had been either completely removed by biopsies or ablated by WESD-related diathermia or the lesion was too small for histologic identification. In 1 patient, extramedullary plasmacytoma was diagnosed. Histology confirmed complete resection (R0) in 18 of the 28 patients with neoplastic lesions (64.3%). In 1 patient, a welldifferentiated adenocarcinoma infiltrated the submucosa deeper than 0.5 mm (sm 2). It was histologically described as an R1 situation, with histologic incomplete horizontal (HM1) and vertical (VM1) resection. In spite of the potential risk of lymph node metastases, surgery was not considered because of severe comorbidities. Complete resection could not be confirmed in the remaining 9 patients. Resection had been performed in two pieces in 3 of these cases, and in another 6 cases parts of lateral resection margins were not free (R1: HM1) or were indeterminate of neoplasia (Rx). Risk of lymph node metastasis seemed to be low because vertical tumor infiltration was limited to the mucosal layer and showed free margins in all specimens (VM0). In view of the low risk, all 10 patients with histologic incomplete resection refused surgery and were scheduled for follow-up examinations to exclude or treat residual neoplasia at the resection scar. Volume 75, No. 6 : 2012 GASTROINTESTINAL ENDOSCOPY 1171
ESD with a water jet–assisted knife
Schumacher et al
TABLE 3. Histology of resected specimens Chronic gastritis, no neoplasia, no.
2
Plasmacytoma, no.
1
Adenoma with low-grade intraepithelial neoplasia, no.
7
Adenocarcinoma, no.
20
Grading, no. Differentiated (G1, G2)
15
Undifferentiated (G3)
1
Signet-ring-cell carcinoma
4
Deepest layer of infiltration, no. Mucosa
17
Submucosa (sm1)
2
Submucosa (sm2)
1
Complete resection (R0) of neoplasia, no. (%)
18/28 (64.3)
complete local remission of neoplasia. Three patients underwent attempts at endoscopic resection of a local recurrence. One of them was treated successful by a second ESD during a follow-up period of 19 months, which achieved complete remission. The histology of the first specimen after ESD had shown a complete resection of a signet-ring-cell carcinoma. The retreatment remained incomplete in the 2 other patients with local recurrence, and gastrectomy was performed in both of them. In one of these cases, initial histology of WESD had demonstrated incomplete resection of an adenocarcinoma infiltrating the superficial layer of the submucosa (sm1). Lack of a lifting sign caused failure of endoscopic retreatment in the other case. A third patient underwent elective gastrectomy because of a metachronous lesion that was identified 54 weeks after initial ESD with an R0 resection. Histology showed adenocarcinoma, and surgery was performed at the request of the patient. The two patients with no neoplasia in the initial ESD procedure were followed for up to 49 months with no evidence of a recurrent or metachronous lesion.
Incomplete resection of neoplasia At lateral tumor margins, no.
9
At lateral and vertical tumor margins, no.
1
G1, differentiated; G2, moderate differentiated; G3, undifferentiated; sm1, infiltration of submucosal layer ⬍_500m; R0, tumorfree margins lateral and vertical.
TABLE 4. Follow-up* of more than 30 days after WESD Patients alive at 30 d, no.
29
Lost to follow-up, no.
1
Complete local remission of neoplasia, no. (%) Recurrence of neoplasia, no.
25/28 (89.3) 3
Endoscopic retreatment, no.
1
Gastrectomy, no.
2
Metachronous gastric adenocarcinoma, no.
1
Gastrectomy, no.
1
WESD, Water jet–assisted endoscopic submucosal dissection. *Median period of 22 months (range 6-44 months).
Medium-term outcome The medium-term outcome of the 29 patients who were alive at 30 days is summarized in Table 4. One of these patients was lost to follow-up. The remaining 28 patients were followed over a median period of 22 months. Twenty-five (89.3%) patients had a 1172 GASTROINTESTINAL ENDOSCOPY Volume 75, No. 6 : 2012
DISCUSSION ESD acceptance in Western countries is low because it is a more time-consuming, difficult procedure, which is correlated with a higher rate of complication. To simplify the technique, a German company (ERBE Elektromedizin GmbH) developed a new device called the HybridKnife. This technique was first evaluated in a porcine model followed by a randomized study comparing EMR and ESD in pigs.15,16 These two experimental studies showed that the use of the new water jet dissector in combination with the HK accelerates the ESD procedure and increases the safety and efficacy. Our aim in this prospective study was to evaluate the efficacy and safety of ESD in patients with early gastric neoplastic lesions. The results of the present study suggest that the technique using the HK for ESD achieves high rates of endoscopic complete resection (100%) and en bloc resection (90%). Twenty-nine of 30 patients met the expanded Japanese criteria for ESD based on the ESD specimen. One patient with undifferentiated histology on the ESD specimen was included even though his condition was outside of the expanded criteria because of the small size of the lesion and the patient’s multiple medical comorbidities. In 3 patients, en bloc resection was not possible because the lesion was located in the upper part of the stomach. EMR was done in these patients, with complete resection in two pieces. Histologic incomplete resection due to laterally indeterminate or infiltrated tumor margins was diagnosed in 10 patients. Our complication rate was low because of the ESD procedure. We had delayed bleeding in one case and 3 visible perforations during the procedures. These could be www.giejournal.org
Schumacher et al
sufficiently closed by hemoclips without need for surgery. The low complication rate can be attributed to the easily repeatable hydrodissection, which maintains the tissue elevation like a cushion before cutting. We needed a median amount of fluid injection of 188 mL for a lesion dissection. The potential risk of perforation may be lower with a larger amount of saline solution injection, as previously reported.16 The relevant bleeding rate also was low. The HK was used for short bursts of coagulation of visible vessels or minor bleeding. The median total procedure time was 74 minutes, which is comparable to other published data with different devices.2-4,23,24 Compared with Western studies using ESD, our procedure time was less than previously reported in these studies: the median procedure times were 85 minutes,22 108 minutes,11 and 161 minutes.10 Compared with EMR, the ESD procedure time is significantly longer.16 As we showed in our animal trial, a tendency of time shortening is obvious with increasing experience, which means a learning curve is also a significant factor for the procedure time. Early gastric cancer is defined as mucosal or submucosal invasive cancer (T1 cancer) irrespective of the presence of lymph node metastasis. Endoscopic treatment should be performed only in patients with early gastric cancer who have no risk of lymph node metastasis, and the resection has to be carried out en bloc. Endoscopic mucosal resection is potentially curative for treatment of early neoplasia, but histopathology of the resected specimen must show no tumor infiltration in the submucosa deeper than 500 m. When lesions are larger than 20 mm in diameter, EMR usually must be performed in a piecemeal fashion that is associated with a high rate of local recurrence.21,22 ESD may overcome the well-known limitations of EMR. ESD allows precise resection in an en bloc procedure.2,25,26 A variety of prospective and retrospective trials from Japan showed that ESD achieves higher rates of curative resection of early gastric cancer than EMR.23-25,27-29 The guidelines for ESD are constantly being updated and newly discussed. The acceptable indications for ESD of early gastric cancer are based on the guidelines of the Japanese Gastric Cancer Association.5,21,29,30 After evaluation of these criteria, new criteria to expand the indications for endoscopic treatment were recommended.5,21,29,30 These studies showed that en bloc, complete, and curative resection rates were significantly higher in the guideline group compared with the expanded group, but there was no difference in overall survival.30,31 Compared with the Japanese data, European experience with ESD is rare. In a recently published German study, R0 en bloc resection for all lesions (early gastric cancers and adenomas) was reported in 74% of cases. Resection rates were lower in patients with early gastric cancer with expanded criteria (68, 6%).10 In our study, 30 patients met the criteria for ESD. The en bloc resection rate was 27 of 30 (90%). The R0 rate of www.giejournal.org
ESD with a water jet–assisted knife
resected neoplasia was 18 of 28 (64.3%). Compared with the Japanese reports, there is less published European experience with ESD.10,32 The discrepancy of a high en bloc resection rate but low rates of histologically confirmed R0 margins is likely caused by inappropriate delineation of lateral tumor margins or too small distances of coagulation markers at the tumor periphery. The majority of the lesions were classified morphologically as type IIa and type IIa⫹c. Even using high-resolution endoscopy with narrow-band imaging may underestimate lateral tumor extension. In spite of these unfavorable histologic results, only 3 recurrences and 1 metachronous cancer were observed during a median follow-up period of 22 months (range 6-44 months). All recurrences developed after 3 months; one metachronous cancer was detected after 13 months. One of the patients with a recurrent tumor was treated by an endoscopic procedure and is still in remission. Two other patients underwent surgery because of deeper infiltration and failure with endoscopic retreatment. One patient developed a metachronous lesion with histologically confirmed early cancer. At the request of the patient, surgery was performed. In summary, this first prospective study using a new device for ESD in patients with early gastric neoplastic lesions demonstrated a high rate of en bloc resection. The procedure times were acceptable (mean 74 minutes), and all complications could be endoscopically managed. The R0 rate of resected neoplasia was 64.3%, which is lower than expected. Complete local remission of neoplasia was achieved in 25 of 28 patients (89.3%), who were followed over a median period of 22 months (range 6-44 months). Three patients underwent gastrectomy because of endoscopically inaccessible recurrences or a metachronous lesion. To establish clear indications for ESD and efficient ESD procedures, additional prospective studies are necessary, with more patients included. Further improvement of the technique seems to be essential.
REFERENCES 1. Korenaga D, Orita H, Maekawa S, et al. Pathological appearance of the stomach after endoscopic mucosal resection for early gastric cancer. Br J Surg 1997;84:1563-6. 2. Ono H, Kondo H, Gotoda T, et al. Endoscopic mucosal resection for treatment of early gastric cancer. Gut 2001;48:225-9. 3. Miyamoto S, Muto M, Hamamoto Y, et al. A new technique for endoscopic mucosal resection with an insulated-tip electrosurgical knife improves the completeness of resection of intramucosal gastric neoplasms. Gastrointest Endosc 2002;55:576-81. 4. Yamamoto H, Kawata H, Sunada K, et al. Success rate of curative endoscopic mucosal resection with circumferential mucosa incision assisted by submucosal injection of sodium hyaluronate. Gastrointest Endosc 2002;56:507-12. 5. Gotoda T. Endoscopic resection of early gastric cancer: the Japanese perspective. Curr Opin Gastroenterol 2006;22:561-9. 6. Rösch T, Sarbia M, Schumacher B, et al. Attempted endoscopic en bloc resection of mucosal and submucosal tumors using insulated-tip knives: a pilot series. Endoscopy 2004;36:788-801.
Volume 75, No. 6 : 2012 GASTROINTESTINAL ENDOSCOPY 1173
ESD with a water jet–assisted knife
7. Neuhaus H, Costamagna G, Devière J, et al. Endoscopic submucosal resection (ESD) of early neoplastic gastric lesions using a new double channel endoscope (the “R-scope”). Endoscopy 2006;38:1016-23. 8. Coda S, Trentino P, Antonellis F, et al. A Western single-center experience with endoscopic submucosal dissection for early gastrointestinal cancers. Gastric Cancer 2010;13:258-63. 9. Probst A, Golger D, Arnholdt H et al. Endoscopic submucosal dissection of early cancers, flat adenomas and submucosal tumors in the gastrointestinal tract. Clin Gastroenterol Hepatol 2009;7:149-55. 10. Probst A, Pommer B, Golger D, et al. Endoscopic submucosal dissection in gastric neoplasia– experience from a European center. Endoscopy 2010;42:1037-44. 11. Riberiro-Mourao F, Pimentel-Nunes P, Dinis-Ribeiro, et al. Endoscopic submucosal dissection for gastric lesions: results of a European inquiry. Endoscopy 2010;42:814-9. 12. Kaehler GFBA, Sold MG, Fischer K, et al. Selective fluid cushion in the submucosal layer by water jet: advantage for endoscopic mucosal resection. Eur Surg Res 2007;39:93-7. 13. Kaehler GFBA, Colet PH, Sold MG, et al. Waterjet for mucosal elevation in the GI tract–first clinical trial [abstract]. Gastrointest Endosc 2007;65: AB288. 14. Kaehler GFBA, Sold MG, Lingenfelder T, et al. Needle knife and water jet combined as a hybrid knife–a comparison with standard knife in an in vivo animal study [abstract]. Gastrointest Endosc 2007;65:AB285. 15. Schumacher B, Neuhaus H, Enderle MD. Experimental use of a new device for mucosectomy. Acta Endoscopica 2007;5:673-8. 16. Neuhaus H, Wirths K, Enderle MD, et al. Randomized controlled study of EMR versus endoscopic submucosal dissection with a water-jet hybridknife of esophageal lesions in a porcine model. Gastrointest Endosc 2009;70:112-20. 17. Yahagi N, Neuhaus H, Schumacher B, et al. Comparison of standard endoscopic submucosal dissection (ESD) versus an optimized ESD technique for colon: an animal study. Endoscopy 2009;41:340-5. 18. Participants in the Paris workshop. The Paris endoscopic classification of superficial neoplastic lesions: esophagus, stomach and colon: November 30 to December 1, 2002. Gastrointest Endosc 2003;58:S3-43. 19. Endoscopic Classification Review Group. Update on the Paris classification of superficial neoplastic lesions in the digestive tract. Endoscopy 2005;37:570-8. 20. Riphaus A, Wehrmann T, Weber B, et al. S3-guidelines sedation in gastrointestinal endoscopy. Z Gastroenterol 2008;46:1298-330. 21. Soetikno R, Kaltenbach T, Yeh R, et al. Endoscopic mucosal resection for early cancers of the upper gastrointestinal tract. J Clin Oncol 2005;23: 4490-8. 22. Hondo FY, Maluf-Filho F, Kishi HS, et al. Predictive factors for local recurrence and incomplete resection of early gastric cancer treated by endoscopic resection: a Western experience. Can J Gastroenterol 2009;23: 357-63.
1174 GASTROINTESTINAL ENDOSCOPY Volume 75, No. 6 : 2012
Schumacher et al
23. Oka S, Tanaka S, Kaneko I, et al. Advantage of endoscopic submucosal dissection compared with EMR for early gastric cancer. Gastrointest Endosc 2006;64:877-83. 24. Oda I, Saito D, Tada M, et al. A multicenter retrospective study of endoscopic resection for early gastric cancer. Gastric Cancer 2006;9:262-70. 25. Cao Y, Liao C, Tan A, et al. Meta-analysis of endoscopic submucosal dissection versus endoscopic mucosal resection for tumors of the gastrointestinal tract. Endoscopy 2009;41:751-7. 26. Yamamoto H, Yahagi N, Oyama T, et al. Usefulness and safety of 0.4% sodium hyaluronate solution as a submucosal fluid “cushion” in endoscopic resection for gastric neoplasms: a prospective multicenter trial. Gastrointest Endosc 2008;67:830-9. 27. Chaves DM, Maluf Filho F, de Moura EG, et al. Endoscopic submucosal dissection for the treatment of early esophageal and gastric cancer–initial experience of a Western center. Clinics (Sao Paulo) 2010;65:377-82. 28. Hamanaka H, Gotoda T. Endoscopic resection for early cancer and future expectations. Dig Endosc 2005;17:275-85. 29. 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;3:219-25. 30. Yamaguchi N, Isomoto H, Fukuda E, et al. Clinical outcomes of endoscopic submucosal dissection for early cancer by indication criteria. Digestion 2009;80:173-81. 31. Isomoto H, Shikuwa S, Yamaguchi N, et al. Endoscopic submucosal dissection for early gastric cancer: a large-scale feasibility study. Gut 2009; 58:331-6. 32. Farhat S, Chaussade S, Ponchon T, et al. Endoscopic submucosal dissection in a European setting: a multi-institutional report of a technique in development. Endoscopy 2011;43:664-70.
Abbreviations: ESD, endoscopic submucosal dissection; HK, HybridKnife; WESD, water jet–assisted ESD. DISCLOSURE: M. Enderle is an employee of ERBE Elektromedizin. No other financial relationships relevant to this publication were disclosed. Copyright © 2012 by the American Society for Gastrointestinal Endoscopy 0016-5107/$36.00 doi:10.1016/j.gie.2012.02.027 Received December 16, 2011. Accepted February 13, 2012. Current affiliations: Department of Gastroenterology (1), EVK Evangelisches Krankenhaus Düsseldorf, Institute of Pathology (2), Klinikum Bayreuth, GmbH, ERBE Elektromedizin GmbH (3), Tübingen, Germany. Reprint requests: Brigitte Schumacher, MD, Department of Internal Medicine, Evangelisches Krankenhaus Düsseldorf, Kirchfeldstr 40, 40217 Düsseldorf, Germany.
www.giejournal.org