Carbon dioxide insufflation reduces number of postprocedure admissions after endoscopic resection of large colonic lesions: a prospective cohort study

ORIGINAL ARTICLE: Clinical Endoscopy

Carbon dioxide insufflation reduces number of postprocedure admissions after endoscopic resection of large colonic lesions: a prospective cohort study Milan S. Bassan, MBBS, FRACP, Bronte Holt, MBBS, FRACP, Alan Moss, MD, MBBS, FRACP, Stephen J. Williams, MD, MBBS, FRACP, Rebecca Sonson, BNurs, Michael J. Bourke, MBBS, FRACP Sydney, New South Wales, Australia

Background: Endoscopic resection (ER) for large colonic lesions is a safe and effective outpatient treatment. Postprocedural pain creates concern for perforation and often results in postprocedure admission (PPA). Carbon dioxide (CO2) insufflation has been shown to reduce pain scores after routine colonoscopy, but an influence on more critical outcomes such as PPA has not been shown. Objective: To assess the outcomes of patients undergoing ER for large colonic lesions, comparing those having air versus those having CO2 insufflation. Design: Prospective, observational, cohort study. Setting: Academic, high-volume, tertiary-care referral center. Patients: Consecutive patients referred for ER of sessile colorectal polyps ⱖ20 mm. Intervention: ER with air or CO2. Main Outcome Measurements: Rates of PPA, technical outcomes, complication rates. Results: ER was performed on 575 lesions ⱖ20 mm, 228 with CO2 insufflation. Mean lesion size was 36.5 mm. Lesion and patient characteristics were similar in both groups. The use of CO2 was associated with a 62% decrease in the PPA rate from 8.9% to 3.4% (P ⫽ .01). This was mainly because of an 82% decrease in PPA for pain from 5.7% to 1.0% (P ⫽ .006). There were no significant difference in the rates of complications. Multiple logistical regression was performed. The adjusted odds ratio (OR) of PPA (OR 0.39; 95% confidence interval [CI], 0.16-0.95; P ⫽ .04) and PPA for pain (OR 0.18; 95% CI, 0.04-0.78; P ⫽ .02) in the CO2 group remained significant. Limitations: Single center, nonrandomized study. Conclusion: CO2 insufflation significantly reduces PPA after ER of large colonic lesions, primarily because of reduced PPA for pain. CO2 insufflation should be routinely used during ER of large colonic lesions. (Gastrointest Endosc 2013;77:90-5.)

Colonoscopy and polypectomy of adenomatous lesions significantly reduces the long-term risk of death from colorectal cancer.1 A subgroup of patients have large (ⱖ20

mm) sessile or laterally spreading flat mucosal neoplasms termed advanced mucosal neoplasia (AMN).2,3 These nonpolypoid lesions have a greater frequency of high-

Abbreviations: AMN, advanced mucosal neoplasia; CO2, carbon dioxide; ER, endoscopic resection; PPA, postprocedure admission; SG, succinylated gelatin.

Copyright © 2013 by the American Society for Gastrointestinal Endoscopy 0016-5107/$36.00 http://dx.doi.org/10.1016/j.gie.2012.06.004

DISCLOSURE: All authors disclosed no financial relationships relevant to this publication.

Received March 31, 2012. Accepted June 6, 2012.

See CME section; p. 108.

Current affiliations: Department of Gastroenterology and Hepatology, Westmead Hospital, Sydney, Australia.

Use your mobile device to scan this QR code and watch the author interview. Download a free QR code scanner by searching ‘QR Scanner’ in your mobile device’s app store. 90 GASTROINTESTINAL ENDOSCOPY Volume 77, No. 1 : 2013

Reprint requests: Dr Michael J. Bourke, Director of Endoscopy, Department of Gastroenterology and Hepatology, Westmead Hospital, c/-Suite 106a, 151-155 Hawkesbury Road, Westmead, Sydney, New South Wales 2143 Australia. If you would like to chat with an author of this article, you may contact Dr Bourke at [email protected].

www.giejournal.org

Bassan et al

grade dysplasia and early invasive disease4 and are technically more challenging to remove endoscopically and are thus often treated surgically. However, large prospective multicenter data confirms that wide-field endoscopic resection (ER) is safe and effective therapy even in very large lesions and avoids the need for surgery in the majority of cases.3,5 Moreover prospective data confirm that ER for AMN results in major net health savings of more than $10,000 and 6 bed days per patient when compared with the modeled best possible surgical outcome.6 However, ER is not devoid of complications, and the most feared is colon perforation, occurring with a frequency of 1% to 2%.3,7,8 Abdominal pain after ER is not uncommon and often triggers an investigation and clinical consultation cascade with attendant costs and frequently results in postprocedure admission (PPA). Perforation and serositis should always be considered in the differential diagnosis in patients with pain after ER; however, there are other possible etiologies for pain. Randomized control trials and a systematic review have demonstrated reduced pain scores and reduced intestinal distension on plain abdominal radiographs after routine colonoscopy with insufflation of carbon dioxide (CO2) compared with insufflation of air, reflecting less retained gas after the procedure.9-12 However, PPA is unaltered. Whether CO2 insufflation substantially influences post ER outcomes for AMNs of the colon is currently unknown. This study reports on the outcomes of patients undergoing ER for large colonic lesions with air versus CO2 insufflation.

METHODS Study design, setting, and patients Data were collected as part of a prospective, observational study of all patients referred for ER of sessile colorectal lesions sized ⱖ20 mm to a single, tertiary-care referral center. Ethics Committee approval was obtained (HREC JH/TG 2008/9/6.1 [2858]). All procedures were performed or directly supervised by two senior endoscopists (M.J.B, S.J.W.). Consecutive patients were enrolled from June 2008 to November 2011. The ER technique was standardized and previously has been described in detail.2,6,13 Procedures were performed with air insufflation only from June 2008 to March 2010 and then after a transition period, exclusively with CO2 insufflation from August 2010 to November 2011. Outcomes from these two time periods were compared. Normal saline solution was used as the injection fluid until January 2010, after which it was supplanted by succinylated gelatin (SG) (Gelofusine; B. Braun, Crissier, Switzerland) when the technical superiority of SG became known.13,14 Procedures were performed on an outpatient basis. Patients were observed for 4 hours after ER and, if they were well, were discharged with instructions to maintain a clear-fluid diet overnight, resuming a normal diet the following day. www.giejournal.org

CO2 for endoscopic resection of colon lesions

Take-home Message ●

●

For endoscopic resection of large, sessile colon polyps, the use of CO2 insufflation leads to a significant reduction in the rate of postprocedure admission compared with that of air insufflation, primarily because of reduced rates of admission for pain without perforation. The reduction in pain is likely to be related to the interaction between the large mucosal wound, luminal distension, and the degree of postendoscopic resection serosal inflammation.

Comprehensive data collection included patient demographics (age, sex, presence of major comorbidity), lesion characteristics (size, morphology, location, previous attempt at resection), technical details (duration of ER, type of injection fluid, presence of submucosal fibrosis, involvement of fellow, technical success), immediate postprocedure disposition, and outcomes at 2 weeks, 4 months, and 12 months. Colon perforation was defined as full-thickness resection noted at the time of the procedure or admission to the hospital with clinical and radiologic features of perforation (regardless of the requirement for surgery). The primary outcomes were the rates of PPA and PPA for pain (without evidence of perforation). Secondary outcomes included rates of complications, rates of technical success, and duration of resection.

Statistical analysis If multiple neoplasms were removed at the same procedure, lesion-specific outcomes were analyzed on a per-lesion basis. Other outcomes were analyzed on a per-patient basis, by using the largest lesion or that with the worst outcome. Pearson chi-square or Fisher exact tests were used to test for association between categorical variables and outcome. Odds ratios (OR) and their 95% confidence intervals (CI) were used to quantify the level of association. Two-tailed tests with a significance level of 5% were used throughout. Multiple logistic regression with backward stepwise variable selection was used to identify the independent predictors of PPA. The best-fitting model for PPA with the independent predictors was used. Statistical analyses were performed by using SPSS 17 (SPSS Inc, Chicago, Ill).

RESULTS During the study period, 524 patients with 575 AMNs ⬎20 mm underwent ER. Of these, ERs on 334 patients (347 lesions) were performed with air insufflation, and ERs on 190 patients (228 lesions) were performed with CO2. Sex, age, lesion characteristics, and mean ER duration were similar between the two groups (Table 1). There were 28 (8.9%) postprocedure admissions in the air insufflation group and 7 (3.4%) in the CO2 insufflation Volume 77, No. 1 : 2013 GASTROINTESTINAL ENDOSCOPY 91

CO2 for endoscopic resection of colon lesions

Bassan et al

group (P ⫽ .01) (Fig. 1). The difference between groups was primarily related to significantly fewer admissions for pain without evidence of perforation in the CO2 group. With air insufflation, there were 18 (5.7%) admissions for pain compared with 2 (1.0%) in the CO2 group (P ⫽ .006) (Fig. 2). Other causes for admission are detailed in Table 2. Of those patients who required admission, the mean length of stay was 1.9 nights in the air group and 1.1 nights in the CO2 group (P ⫽ .20).

Although there was a significant difference between the air and CO2 groups in the proportion of cases performed with SG as the primary constituent of the ER injection fluid (Table 1), there was no significant difference in the overall rates of PPA (9.1% vs 5.9%; P ⫽ .18) or PPA for pain (4.3% vs 3.8%; P ⫽ .82) between those with saline solution or SG as the primary constituent of ER fluid. On further analysis of the air group, there was no significant difference in the rates of PPA (9.0% vs 7.9%; P ⫽ .72) or PPA for pain (4.0% vs 7.4%; P ⫽ .21) between those with saline solution or SG as the primary constituent of ER fluid. Multiple logistical regression was performed to identify independent predictors of PPA and PPA for pain. Variables included were age, lesion size, ER duration, involvement of fellow, the presence of a major comorbidity, location, main constituent of injection solution, the presence of submucosal fibrosis, the presence of invasive carcinoma, and the use of air or CO2 for insufflation. Apart from the use of air compared with CO2 insufflation, increasing age (OR 1.04; 95% CI, 1.01-1.08; P ⫽ .05) and ER duration (OR 1.04; 95% CI, 1.02-1.05; P ⬍ .01) were identified as independent predictors of PPA, and ER duration was identified as an independent predictor of PPA for pain (OR 1.04; 95% CI, 1.02-1.05; P ⬍ .01). After we adjusted for these independent predictors, the OR for CO2 versus air insufflation remained significant both for PPA (OR 0.39; 95% CI, 0.160.95; P ⫽ .04) and PPA for pain (OR 0.18; 95% CI, 0.040.78; P ⫽ .02). The use of SG as the primary constituent of the ER fluid was not identified as an independent predictor of PPA or PPA for pain on the multivariate analysis. Similarly high success rates for ER were achieved in both the air and CO2 groups (Table 1). The presence of submucosal fibrosis (21.4% vs 7.3%; P ⬍ .01) and invasive carcinoma (21.6% vs 9.3%; P ⫽ .04) were associated with failure of successful single-session ER. A similar number of snare resections were required to resect the lesions in both groups (4.3 air vs 4.6 CO2; P ⫽ .20). Fellows were involved in the actual ER in 39% of procedures, with no difference between groups. ER complication rates were comparable between groups. There was no difference in the rates of perforation. In the air group, there were 4 (1.1%) perforations compared with 1 (0.5%) in the CO2 group (P ⫽ .65). Of these, in the air group, all patients had the perforation recognized at the time of ER, and clip closure was attempted. In two patients there was no pain, and patients were able to be discharged the following day with conservative management. One patient had persistent pain after ER and proceeded to laparoscopy where the defect was noted to be completely closed without evidence of peritoneal soiling. The final patient had two areas of possible deep (focal muscularis propria) resection noted, and these were clipped. The patient was initially pain-free but then developed features of perforation in the recovery area and required surgical management. The patient with perforation in the CO2 group had a full-thickness resection

92 GASTROINTESTINAL ENDOSCOPY Volume 77, No. 1 : 2013

www.giejournal.org

TABLE 1. Patient and lesion characteristics

Air

CO2

P value

Patients, no.

334

190

N/A

Lesions

347

228

N/A

Age, mean, y

69.1

67.8

.15

Male, no. (%) Lesion size, mean, mm

179 (53.6) 113 (59.4) 35.7

.63

37.5

.17

53 (25.6)

.99

Right side of colon, no. (%) 203 (58.5) 143 (63)

.28

Previous attempt at resection, no. (%)

46 (13.4)

26 (11.4)

.49

Submucosal fibrosis, no. (%)

61 (18.6)

51 (23.3)

.18

Presence of invasive carcinoma, no. (%)

29 (8.4)

13 (6.3)

.37

EMR time, mean, min

24.9

21.8

.12

292 (88)

203 (93)

.07

Fellow involved, no. (%)

117 (38.4) 79 (39.1)

.87

Succinylated gelatin as main constituent of injection solution, no. (%)

153 (46.4) 220 (99.5) ⬍.001

Major comorbidity, no. (%) 81 (25.6)

Successful EMR, no. (%)

CO2, Carbon dioxide; N/A, not applicable; EMR, endoscopic mucosal resection.

Figure 1. Postprocedure admission rates for all admissions and admissions for abdominal pain without perforation. PPA, postprocedure admission.

Bassan et al

CO2 for endoscopic resection of colon lesions

Figure 2. Postulated mechanism for the reduction in pain with CO2 insufflation.

With the introduction of colorectal cancer screening programs, more advanced mucosal neoplasms are being identified. Since the description of submucosal preinjec-

tion by Deyhle et al in 1973,15 there have been significant advances in the technique of ER, allowing complete endoscopic removal of extremely large laterally spreading lesions.3,13 The safety and efficacy of ER techniques has been demonstrated in multicenter studies and systematic review.3,16,17 The majority of patients with AMN can be managed endoscopically,18 with a substantial cost saving6 compared with that of surgery. Careful inspection of the post-ER mucosal defect may identify risk factors for ER complications and consequently mitigate against the frequency and severity of their sequelae, particularly perforation.2,19 The target sign has been described as a marker of focal muscularis propria resection,20 and its presence should alert the endoscopist to deep resection, and endoscopic closure of the site of injury should be performed and the patient closely observed after the procedure before discharge. Even in the absence of this sign, patients require close assessment after ER because of the small but significant risk of perforation. Abdominal pain is regarded as the cardinal symptom of perforation, and the presence of significant abdominal pain after the procedure usually requires further investigation and admission to the hospital. However, abdominal pain after the procedure also may be due to serositis, excessive transmural injection, or gaseous distension.2 We have shown that CO2 insufflation during ER of AMNs of the colon significantly reduces PPA compared with air insufflation. This is primarily related to a reduced rate of PPA due to abdominal pain without perforation. The postulated mechanism is presented in Figure 2. CO2 insufflation causes less postprocedure luminal distension than does air because CO2 is rapidly absorbed from the intestinal lumen.21 In contrast to conventional colonoscopic polypectomy, after ER for AMN there is a large

www.giejournal.org

Volume 77, No. 1 : 2013 GASTROINTESTINAL ENDOSCOPY 93

TABLE 2. Admission by cause Air CO2 (% of total (% of total P cohort) cohort) value Total admissions, no. (%)

28 (8.9)

7 (3.4)

.01

Pain (without perforation), no. (%)

18 (5.7)

2 (1.0)

.006

Perforation, no. (%)

4 (1.1)

1 (0.5)

.36

Bleeding, no. (%)

1 (0.3)

0

.42

Comorbidities requiring in-hospital observation after sedation, no. (%)

4 (1.1)

3 (1.6)

.86

Other clinical concern (no pain or perforation), no. (%)

1 (0.3)

1 (0.5)

.76

that was successfully closed with endoscopic clips, and the patient was discharged the following day. One patient in the air group had postprocedure bleeding. This patient was well after the procedure but developed major bleeding in the context of anticoagulation for a mechanical aortic valve prosthesis and developed a perforation requiring surgery as a result of prolonged attempts at endoscopic hemostasis. There were no respiratory complications in either group.

DISCUSSION

CO2 for endoscopic resection of colon lesions

Bassan et al

Figure 3. Examples of large sessile (A) and laterally spreading (C, E) lesions and the corresponding vast mucosal wound following endoscopic resection (B, D, F). All procedures were done on an outpatient basis, with patients being discharged home the same day even with (B) visible but uninjured muscle evident.

mucosal wound of 20 mm and often substantially greater (Fig. 3), and we postulate that the reduction in tension on this large defect leads to a significant reduction in postprocedure pain. In addition, the serosa contains afferent sensory innervation,22 and it is likely that after ER for AMN there is a significant subgroup of patients who have a spectrum of serosal inflammation from subclinical to frank clinical serositis. Nociception of this injury is dependent on several factors, including the amount of tension on the bowel wall. We do not believe that the benefit in reduced PPA simply reflects less luminal gas distension in isolation, but rather the interaction between the mucosal wound, the severity of post-ER serosal inflammation, and the degree of tension on the colon wall. Compared with PPA after ER,

PPA after routine outpatient colonoscopy is relatively uncommon, at a rate of approximately 0.16%.23 This is the first study demonstrating a reduction in the rates of PPA by the use of CO2 insufflation after advanced ER. Reduced PPA translates to decreased use of resources, including acute inpatient hospital beds, medical imaging, and cross-team consultation, leading to a significant cost saving. When weighed against the relatively modest capital and consumable cost of CO2 insufflation systems and their excellent safety profiles, we believe that CO2 should be used routinely for gas insufflation during ER for AMN. There are several limitations to this study that need to be considered. The patients were not randomized to air or CO2. However, the two groups contain large numbers and

94 GASTROINTESTINAL ENDOSCOPY Volume 77, No. 1 : 2013

www.giejournal.org

Bassan et al

are well-matched in terms of patient and lesion characteristics apart from the use of SG as the primary constituent of the ER injection fluid. In comparison to saline solution, the use of SG was not significant on univariate or multivariate analysis for PPA or PPA for pain. SG previously has been shown to reduce the duration of ER and reduce the number of resections13 and thus could potentially impact distension and serosal inflammation and a beneficial effect in this regard upon the CO2 cohort cannot be completely excluded. This was a single-center study, conducted at an academic tertiary-care center with a strong interest in ER and a large referral base. It is possible that the observed outcomes therefore may not be generalizable. However, one would expect the benefits of CO2 insufflation to be greater in a less-experienced institution where procedure times are likely to be longer. ER times were relatively short at just over 20 minutes and not significantly different between the two groups. To validate the results of this study, a randomized controlled trial would be appropriate. Given the high-level evidence supporting the superiority of CO2 insufflation for routine colonoscopy and the strength of our findings, we suggest that CO2 insufflation be adopted as the standard insufflation gas for ER until such a trial is performed. In summary, based on these data, we believe that CO2 should be the accepted standard for gas insufflation during ER of AMNs of the colon. Its use leads to a significant reduction in PPA, which likely confers substantial net health care savings. ACKNOWLEDGMENT The authors thank Dr Karen Byth for assistance with statistical analysis of the data for this study. REFERENCES 1. Zauber AG, Winawer SJ, O’Brien MJ, et al. Colonoscopic polypectomy and long-term prevention of colorectal-cancer deaths. N Engl J Med 2012;366:687-96. 2. Bourke M. Endoscopic mucosal resection in the colon: a practical guide. Tech Gastrointest Endosc 2011;13:35-49. 3. Moss A, Bourke MJ, Williams SJ, et al. Endoscopic mucosal resection outcomes and prediction of submucosal cancer from advanced colonic mucosal neoplasia. Gastroenterology 2011;140:1909-18. 4. Soetikno RM, Kaltenbach T, Rouse RV, et al. Prevalence of nonpolypoid (flat and depressed) colorectal neoplasms in asymptomatic and symptomatic adults. JAMA 2008;299:1027-35. 5. Saito Y, Matsuda T, Fujii T. Endoscopic submucosal dissection of nonpolypoid colorectal neoplasms. Gastrointest Endosc Clin N Am 2010;20: 515-24.

www.giejournal.org

CO2 for endoscopic resection of colon lesions 6. Swan MP, Bourke MJ, Alexander S, et al. Large refractory colonic polyps: Is it time to change our practice? A prospective study of the clinical and economic impact of a tertiary referral colonic mucosal resection and polypectomy service (with videos). Gastrointest Endosc 2009;70:1128-36. 7. Tanaka S, Haruma K, Oka S, et al. Clinicopathologic features and endoscopic treatment of superficially spreading colorectal neoplasms larger than 20 mm. Gastrointest Endosc 2001;54:62-6. 8. Ferrara F, Luigiano C, Ghersi S, et al. Efficacy, safety and outcomes of ’inject and cut’ endoscopic mucosal resection for large sessile and flat colorectal polyps. Digestion 2010;82:213-20. 9. Bretthauer M, Thiis-Evensen E, Huppertz-Hauss G, et al. NORCCAP (Norwegian colorectal cancer prevention): a randomised trial to assess the safety and efficacy of carbon dioxide versus air insufflation in colonoscopy. Gut 2002;50:604-7. 10. Sumanac K, Zealley I, Fox BM, et al. Minimizing postcolonoscopy abdominal pain by using CO(2) insufflation: a prospective, randomized, double blind, controlled trial evaluating a new commercially available CO(2) delivery system. Gastrointest Endosc 2002;56:190-4. 11. Stevenson GW, Wilson JA, Wilkinson J, et al. Pain following colonoscopy: elimination with carbon dioxide. Gastrointest Endosc 1992;38:564-7. 12. Dellon ES, Hawk JS, Grimm IS, et al. The use of carbon dioxide for insufflation during GI endoscopy: a systematic review. Gastrointest Endosc 2009;69:843-9. 13. Moss A, Bourke MJ, Metz AJ. A randomized, double-blind trial of succinylated gelatin submucosal injection for endoscopic resection of large sessile polyps of the colon. Am J Gastroenterol 2010;105:2375-82. 14. Moss A, Bourke MJ, Kwan V, et al. Succinylated gelatin substantially increases en bloc resection size in colonic EMR: a randomized, blinded trial in a porcine model. Gastrointest Endosc 2010;71:589-95. 15. Deyhle P LF, Jenny S, Fumagalli I. A method for endoscopic electroresection of sessile colonic polyps. Endoscopy 1973;5:38-40. 16. Repici A, Hassan C, De Paula Pessoa D, et al. Efficacy and safety of endoscopic submucosal dissection for colorectal neoplasia: a systematic review. Endoscopy 2012;44:137-50. 17. Puli SR, Kakugawa Y, Gotoda T, et al. Meta-analysis and systematic review of colorectal endoscopic mucosal resection. World J Gastroenterol 2009;15:4273-7. 18. Doniec JM, Mathias SL, Bodo S, et al. Endoscopic removal of large colorectal polyps: prevention of unnecessary surgery? Dis Colon Rectum 2003;46:340-8. 19. Holt B, Bassan M, Trivedi S, et al. The endoscopic mucosal resection defect—a prospective interrogation of 188 consecutive post resection defects for colonic lesions 20 mm or greater. Endoscopy 2011;43:A95. 20. Swan MP, Bourke MJ, Moss A, et al. The target sign: an endoscopic marker for the resection of the muscularis propria and potential perforation during colonic endoscopic mucosal resection. Gastrointest Endosc 2011;73:79-85. 21. Yasumasa K, Nakajima K, Endo S, et al. Carbon dioxide insufflation attenuates parietal blood flow obstruction in distended colon: potential advantages of carbon dioxide insufflated colonoscopy. Surg Endosc 2006; 20:587-94. 22. Peiris M, Bulmer DC, Baker MD, et al. Human visceral afferent recordings: preliminary report. Gut 2011;60:204-8. 23. Leffler DA, Kheraj R, Garud S, et al. The incidence and cost of unexpected hospital use after scheduled outpatient endoscopy. Arch Intern Med 2010;170:1752-7.

Volume 77, No. 1 : 2013 GASTROINTESTINAL ENDOSCOPY 95