LEARNING CURVE FOR ENDOSCOPIC SUBMUCOSAL DISSECTION OF EARLY GASTRIC CANCER BASED ON TRAINEE EXPERIENCE
Ichiro Oda, Endoscopy Division, National Cancer Center Hospital, 5-1-1 Tsukiji, Chuo-ku, Tokyo 104-0045, Japan. Email: email@example.com
Background and Aim: There have been few previous reports on endoscopic submucosal dissection (ESD) learning curve for early gastric cancer (EGC) so we retrospectively assessed this subject based on experience of our trainees.
Methods: Trainees in our center start performing ESDs for lesions in lower third of stomach with hands-on support by experts during first 10 cases and then perform ESDs by themselves primarily with verbal guidance from experts. They are gradually assigned to perform ESDs in middle and upper thirds of stomach. From January 1999 to December 2008, 464 EGC patients, who underwent ESD performed by 13 trainees, were assessed by dividing ESD cases into five training periods (A, 1-10; B, 11-20; C, 21-30; D, 31-40; and E, 41-50). We compared data from B to C, D and E.
Results: Lesions in lower third were A/59%, B/57%, C/55%, D/36% and E/40% with B significantly higher than D (p<0.01) and E (p<0.05). Mean tumor sizes were A/13.9±7.5mm, B/18.3±11.4mm, C/19.0±12.5mm, D/19.3±11.7mm and E/16.8±10.3mm. En-bloc resection rate was 100% in every period. Delayed bleeding / perforation rates were A/0%/1.8%, B/2.8%/1.9%, C/1.9%/2.9%, D/1.1%/0% and E/2.1%/2.1%, respectively. Lower third procedure times were A/76±39, B/90±61, C/70±48, D/60±50 and E/55±26 minutes with B significantly longer than D and E (p<0.05). Middle and upper third procedure times were A/104±80, B/115±68, C/106±67, D/134±86 and E/96±55 minutes.
Conclusion: Step-by-step training was highly effective with 100% en-bloc resection rate and few complications. Learning curve point for our trainees to acquire performing ESD in lower third of stomach was 30 cases.
Endoscopic submucosal dissection (ESD) is widely accepted for treating early gastric cancer (EGC) with a negligible risk of lymph node metastasis.1–3 ESD requires a high level of technical expertise, however, so it is technically challenging especially for less experienced endoscopists. ESD trainees at the National Cancer Center Hospital (NCCH) in Tokyo, Japan, follow a step-by-step process for learning ESD techniques as shown in Table 1. The first step entails acquiring a basic knowledge and understanding of EGC and ESD in particular diagnosis of EGC and the indications for ESD. The next step is for trainees to observe expert endoscopists in action as they perform various ESD procedures. The third step involves trainees acquiring first-hand experience by assisting during actual ESD followed by ESD training using animal models as the fourth step. In the final step, it is important for trainees to start by performing ESD on lesions easier to treat including those that are located in the lower third of the stomach, smaller in size and without ulcer fibrosis. Trainees perform ESD with direct hands-on support from highly qualified endoscopists for the first 10 ESD procedures and then start to perform ESD by themselves with mostly verbal guidance from the expert endoscopists. As their ESD techniques improve, trainees are gradually assigned to perform ESD on lesions located in the middle and upper thirds of the stomach and larger in size.
Table 1. Phased learning of gastric ESD techniques
|1. Acquire basic knowledge about EGC and ESD |
|2. Observe experts in action|
|3. Train using animal model|
|4. Perform ESD procedures†|
Few studies have previously reported on the subject so we decided to assess the ESD learning curve for EGC based on the experience of our ESD trainees.4,5
PATIENTS AND METHODS
From January 1999 to December 2008, 13 trainees performed their initial gastric ESD and conducted more than 30 subsequent ESD at NCCH. The previous mean ± standard deviation endoscopy experience of the 13 trainees prior to performing their first ESD was 5.2 ± 2.2 years. During this period, 630 patients with solitary EGC clinically diagnosed before treatment as intestinal type EGC with a negligible risk of lymph node metastasis underwent ESD performed by the 13 trainees. After excluding 104 cases with pathological ulcer findings and 62 other cases involving some trainees who had performed more than 50 ESD, we retrospectively assessed 464 EGC cases by dividing the ESD into five equal training periods: A, 1–10 cases; B, 11–20 cases; C, 21–30 cases; D, 31–40 cases; and E, 41–50 cases. Clinicopathogical findings including age, sex, tumor location and tumor size were investigated as well as technical results regarding the en bloc resection rate, procedure time, delayed bleeding rate and perforation rate.
Tumor locations were divided into the lower third of the stomach and the middle and upper thirds of the stomach combined based on the Japanese classification of gastric carcinoma.6 Tumor sizes were estimated endoscopically. An en bloc resection was defined as a one-piece resection while delayed bleeding was defined as a clinical indication of bleeding evidenced by hematemesis or melena requiring endoscopic treatment and occurring within 30 days of an ESD. We also recorded the incidence of perforations as observed during the procedure or evidenced clinically following an ESD.
We compared data from period B to periods C, D and E because highly qualified endoscopists provided hands-on support to trainees through most of the process of performing the various ESD procedures during period A. Data were analyzed using the χ2-test, Fisher's exact test or Student's t-test as appropriate (STATA version 10.0; StataCorp, College Station, TX, USA) with value differences of P < 0.05 considered statistically significant.
Clinicopathogical findings of EGC resected by 13 trainees are shown in Table 2. The proportion of lesions located in the lower third of the stomach was 59%, 57%, 55%, 36% and 40% in periods A, B, C, D and E, respectively. Lesions located in the lower third of the stomach were significantly less frequent in periods D (P < 0.01) and E (P < 0.05) than in period B. Table 3 indicates the technical results of the ESD performed by the 13 trainees for all 464 cases. The en bloc resection rate was 100% in all five periods and there were no significant differences in any of the technical results between period B and periods C, D and E. When procedure times for the 13 trainees were analyzed by tumor location (Table 4), the procedure times for lesions located in the lower third of the stomach were 76 ± 39, 90 ± 61, 70 ± 48, 60 ± 50 and 55 ± 26 min in periods A, B, C, D and E, respectively, with the procedure times significantly shorter in periods D (P < 0.05) and period E (P < 0.05) than in period B. Procedure times for lesions located in the middle and upper thirds of the stomach were 104 ± 80, 115 ± 68, 106 ± 67, 134 ± 86 and 96 ± 55 in periods A, B, C, D and E, respectively.
Table 2. Clinicopathogical EGC findings
|Age, mean ± SD (years)||65.7 ± 9.6||68.1 ± 8.7||66.8 ± 9.3||66.4 ± 8.7||66.8 ± 7.9||67.0 ± 9.0|
| Male||91 (80)||84 (78)||84 (80)||67 (75)||35 (73)||361 (78)|
| Female||23 (20)||24 (22)||21 (20)||22 (25)||13 (27)||103 (22)|
|Tumor location (%)|
| Lower third||67 (59)||62 (57)||58 (55)||32 (36)†||19 (40)†||238 (51)|
| Middle and upper thirds||47 (41)||46 (43)||47 (45)||57 (64)||29 (60)||226 (49)|
| Tumor size, mean ± SD (mm)||13.9 ± 7.5||18.3 ± 11.4||19.0 ± 12.5||19.3 ± 11.7||16.8 ± 10.3||17.4 ± 11.0|
Table 3. ESD technical results
|En bloc resection rate (%)||114 (100)||108 (100)||105 (100)||89 (100)||48 (100)||464 (100)|
|Procedure time, mean ± SD (min)||88 ± 61||101 ± 65||86 ± 60||108 ± 83||80 ± 50||93 ± 66|
|Delayed bleeding rate (%)||0 (0)||3 (2.8)||2 (1.9)||1 (1.1)||1 (2.1)||7 (1.5)|
|Perforation rate (%)||2 (1.8)||2 (1.9)||3 (2.9)||0 (0)||1 (2.1)||8 (1.7)|
Table 4. ESD procedure times by tumor location
|Procedure time, mean ± SD (min)|
|Lower third (n = 238)||76 ± 39||90 ± 61||70 ± 48||60 ± 50†||55 ± 26†||75 ± 50|
|Middle and upper thirds (n = 226)||104 ± 80||115 ± 68||106 ± 67||134 ± 86||96 ± 55||113 ± 75|
Gastric ESD has a number of advantages including higher rates of en bloc resection and a lower incidence of local recurrence in comparison to endoscopic mucosal resection (EMR).7–13 When compared to surgery, ESD preserves the stomach and, therefore, improves a patient's overall quality of life. Although many endoscopists are interested in learning how to perform ESD, the procedure requires a high level of expertise and is technically challenging especially for less experienced endoscopists. There have been few reports regarding the learning curve for ESD4,5 and they have focused on determining the learning curve point for gastric ESD based on all data from the time this procedure was first introduced at their hospital. Choi et al. examined the learning curve for the circumferential incision and snare technique reporting an increase in the en bloc resection rate from 45% to 85% after 40 cases. There were three perforations in the first 20 procedures (15%), but only one in the following 60 procedures. They concluded that 20–40 procedures was the learning curve point for this particular technique.4 Kakushima et al. demonstrated that a reduction in procedure time was an indicator for becoming proficient in this skill.5
There is no other published report of a similar size series to the best of our knowledge on establishing the learning curve point based on the experience of trainee endoscopists under the direction of highly skilled endoscopists. We developed and introduced the ESD procedure for clinical use on patients in the mid-1990s and have gained substantial experience and expertise in performing ESD over the years. As a result, we have had several expert endoscopists specializing in ESD at NCCH throughout the study period. We assessed the ESD learning curve for EGC based on the experience of our ESD trainees. The step-by-step training program at NCCH has been very effective as evidenced by the trainees achieving an en bloc resection rate of 100% with comparatively few complications. It should be noted that a basic level of diagnostic endoscopy with detection and characterization of lesions as well as being able to perform targeting biopsy and experience in therapeutic endoscopic procedures including hemostasis, polypectomy and EMR are required before starting to learn how to perform ESD. Actually, the 13 trainees in this study had a mean of 5.2 ± 2.2 years of basic diagnostic and therapeutic endoscopic experiences prior to performing their first ESD. As Kakushima et al. reported, we used procedure time as an indicator of ESD proficiency and determined that 30 cases was the learning curve point to acquire the basic technical skills for successfully performing ESD in the lower third of the stomach. The procedure time for lesions located in the middle and upper thirds of the stomach was 96 ± 55 min in period E compared to 115 ± 68 min in period B, although there was no significant difference probably due to the relatively small number of cases. In our estimation, performing at least 40 ESD would be the minimum learning curve point for performing ESD in the middle and upper thirds of the stomach. It should be noted that we excluded 104 EGC with ulcer fibrosis because of increased technical difficulty and longer procedure time in performing ESD and such cases were too few in number to accurately assess the learning curve for EGC with ulcer fibrosis.
Limitations included this study being a retrospective investigation from a single center between January 1999 and December 2008. There have been various improvements over the years in endoscopes, cutting instruments, injection solutions, electrosurgical units, training animal models and other equipment which probably have affected the ESD learning curve point.14–22 We assume that such improvements have contributed to improving the learning curve process and further advances should shorten the learning curve point in the future.
In conclusion, the step-by-step training system in our center has been highly effective with an en bloc resection rate of 100% and a low complication rate. As a result of this program, the learning curve point for our trainees to acquire the basic technical skills for successfully performing ESD in the lower third of the stomach was 30 cases.
The authors thank Drs Seiichiro Abe, Taku Sakamoto and Yutaka Saito (National Cancer Center Hospital) for their efforts and advice concerning this study.
CONFLICT OF INTEREST