The occurrence rate of PPS varies widely from 0%-7.6% in previous reports; however, most studies report a rate around 1%. It is considered some reports had high percentages of occurrence due to small patient populations[18-21,23-26,28-33].

Some previous reports have investigated the risk factors of PPS Nivatvongs[18] showed that 83% of PPS patients had polyps in the right side of the colon, and all were sessile polyps. Choo et al[24] also showed that right-colon polypectomies had a statistically significantly higher tendency for developing PPS[24]. Lee et al[20] reported that a polyp size > 2 cm (OR = 1.08) and hypertension (OR = 14.40) were associated with a significantly increased risk of PPS. The most recent report showed that hypertension, a large lesion size, and non-polypoid configuration of the lesion were independently associated with PPS according to multivariate analysis[19].

PPS develops when the electrical current applied during colonoscopic polypectomy extends past the mucosa into the muscularis propria and serosa, resulting in a transmural burn without perforation[17-33]. Therefore, larger lesions and non-polypoid configuration are logical risk factors, as they usually require a large amount of thermal energy for a longer duration. However, the mechanism of hypertension to promote PPS is unclear. Patients with hypertension are more likely to have endothelial dysfunction[43] and atherosclerosis[44,45], which may be contributing factors.

However, with thinness of the wall, there is also concern regarding the frequency of PPS. The right colon wall is thin, and a large study that addressed major post-polypectomy complications reported barotraumatic perforations, and all of them were caused by cecal blow-out[34,46-48]. Regarding colonic perforation, it has been suggested that air insufflation during colonoscopy generates a higher pressure in the cecum than in the rest of the colon, increasing vulnerability to injury. In addition, Rutter et al[49] hypothesized that a more perpendicular approach to polypectomy in the cecum may increase the risk of complications. However, scientific evidence in support of these theories is lacking.

Loffeld et al[47] also reported that barotrauma caused by insufflated air occurs more often than therapeutic perforation due to polypectomy or coagulation.

Theoretically, submucosal saline injections of large, non-polypoid lesions prior to EMR may reduce the risk of PPS. The rationale for this is that a submucosal saline injection may increase the thickness of the submucosal layer and consequently reduce the risk of PPS[32]. However, no studies have supported this assumption. Sethi et al[17] hypothesized that submucosal injection itself leads to serosal irritation and localized peritonitis, and then patients present with PPS symptoms. Therefore, the protective role of the saline “cushion” for PPS should be considered in future studies.

The improvement of devices would likely reduce PPS. Galloro et al[50] reported that steel snares induced significantly deeper tissue injury than tungsten snares in the pure cut mode; therefore, tungsten snares may reduce the risk of PPS[32]. Another way to reduce the risk of PPS is dependent on skill. Using lower risk procedures when clinically appropriate or referring patients to high-volume endoscopists can reduce the complication rates[51].

PPS is considered different from infection from a local mucosal defect. Min et al[52] reported that blood cultures at baseline and 5 min after the procedure were all negative, and a blood culture at 30 min after the procedure showed a positive result in only 1 of 40 patients (2.5%). However, this one positive sample was considered contamination. None of the 40 patients showed any signs or symptoms associated with infection. Therefore, the prior administration of antibiotics is considered controversial for preventing PPS.

Most cases of PPS have an excellent prognosis, and they are managed conservatively with medical therapy. In some reports, all patients were admitted to the hospital, while in other reports, some cases underwent outpatient observation[19,21,23,26,30,31]. Treatment of PPS requires bowel rest and the administration of intravenous fluids and broad-spectrum parenteral antibiotics to cover the colonic bacterial flora. Nothing is taken by mouth until the symptoms subside. Patients with mild symptoms and adequate outpatient follow-up can be managed with oral antibiotics and a clear liquid diet for 1-2 d.

In contrast, to diffuse peritoneal signs, there is an indication for immediate surgical intervention. Within the spectrum of post-polypectomy cautery injury, “mini-perforation” falls between a “serosal burn” and frank perforation (with diffuse peritonitis). It is a minimal defect that can be quickly covered by peri-intestinal fat and omentum[16]. Its clinical features include pneumoperitoneum without signs and symptoms of diffuse or spreading peritonitis, and with local tenderness that is characteristic of a full-thickness burn. The patient usually improves within 24 h, and the symptoms should resolve within 96 h with conservative treatment. The dilemma as to whether the conservative or surgical approach is more appropriate for managing this kind of perforation still exists[19-22,26,30,31].

Although conservative treatment can generally be performed in most patients, it is important to adopt careful measures such as prolonging the fasting period and considering the possibility of delayed perforation[26,35-39].

ESD has been a reliable method for en bloc resection of colorectal tumors regardless of the lesion size for years. Although colorectal ESD has been established as a procedure with reproducible safety and efficacy, complications such as intestinal perforation and delayed bleeding remain to be problematic. Similarly, few studies have reported CS after ESD[11-14,26].

Hong et al[48] reported that 8.6% showed CS after colorectal ESD. There were no differences in the demographic and endoscopic characteristics (age, sex, underlying disease, procedure time, tumor size, macroscopic type, location, and pathologic findings) between patients with CS and those without CS. The mean hospitalization stay was statically significantly longer in the CS group than that in the non-CS group. All patients with CS were treated with conservative (non-surgical) management (e.g., fasting and intravenous antibiotics). CS showed a favorable progression even after ESD, and delayed perforation was not reported. See comment in pubmed commons below.

CS is reported even after ESD, and its frequency is clearly higher than polypectomy or EMR[26,34,48]. The procedure time and ulcer bed to energization that largely affects the characteristics of ESD procedures is evident theoretically. Delayed perforation in ESD is also a great concern. The indications for ESD are markedly different from those for conventional EMR, and the overall perforation rate is higher compared to conventional EMR[55]. Delayed perforation after ESD reportedly ranges from about 0.1%-0.4%; however, this may be because of the small number of reports[26,40-42,55].

Saito et al[41] reported that delayed perforations occurred in another 4 patients (0.4%) after ESD. Two of the 4 patients with delayed perforations were successfully treated conservatively, because the abdominal findings and inflammatory changes based on laboratory data were slight. However, other patients with delayed perforation required emergency surgery because of the risk of peritonitis. Saito et al[41] also reported that 0.11% (1/900) showed delayed perforation that required emergency surgery. Previous studies have cautioned that clinicians must carefully follow patients with delayed perforation, and continually close communication with consulting surgeons is essential since the number of such cases has been quite limited to date.

Few studies have reported on delayed perforation after ESD. While previous reports have shown the success of endoscopic clip closure with over-tube[42], the treatment and prognosis often require emergent surgery.

A 44-year-old woman underwent colonoscopy for surveillance of ulcerative colitis, and a 30 mm cecal sessile polyp was revealed (Figure 1). We diagnosed this tumor as a sessile serrated adenoma/polyp using the pit and narrow-band imaging patterns. Because of the size of the tumor and the tumor morphology, we chose ESD in order to perform en-bloc resection. ESD was performed safely without any perioperative complications (Figures 2 and 3), and she reported no symptoms. However, 24 h after ESD, she had a high fever (38.6 °C) with slight abdominal pain and leukocytosis. Subsequently, she was diagnosed with CS after ESD. She fasted and received antibiotics (cefmetazole). CT was obtained immediately, but no findings were suggestive of perforation (i.e., free air and ascites were not present) (Figure 4). Thirty hours after ESD, severe abdominal pain developed, and 36 h after ESD, free air appeared on radiography and CT (Figures 5 and 6). At this point, we diagnosed the patient with delayed perforation that developed after CS. Emergent laparoscopic surgery was performed, and a perforation site was found in the ESD ulcer at the bottom of the cecum (Figure 7). Partial cecum resection was performed, and the patient’s condition improved rapidly.

Open in New Tab   Full Size Figure   Download Figure

Figure 1 Chromoendoscopy showing a 30 mm laterally spreading tumor (non-granular type) located at the bottom of the cecum bottom.

Open in New Tab   Full Size Figure   Download Figure

Figure 2 Endoscopic submucosal dissection is performed safely without intraoperative perforation. The procedure time is 47 min.

Open in New Tab   Full Size Figure   Download Figure

Figure 3 The tumor is resected in an en bloc fashion, and the specimen measures 51 mm × 40 mm.

Open in New Tab   Full Size Figure   Download Figure

Figure 4 Abdominal contrast enhanced computed tomography 24 h after endoscopic submucosal dissection. Free air is not recognized.

Open in New Tab   Full Size Figure   Download Figure

Figure 5 Abdominal radiography 36 h after endoscopic submucosal dissection because of a complaint of severe abdominal pain. A large amount of free air is observed under the diaphragm.

Open in New Tab   Full Size Figure   Download Figure

Figure 6 Abdominal computed tomography 36 h after endoscopic submucosal dissection. A large amount of free air is observed on the surface of the liver and the intraperitoneal cavity.

Open in New Tab   Full Size Figure   Download Figure

Figure 7 A small perforation site is present in the surgical specimen.