Carbon dioxide insufflation during endoscopic resection of large colorectal polyps can reduce post-procedure abdominal pain: A prospective, double-blind, randomized controlled trial

Abstract

Background

Studies of the use of CO₂ insufflation during endoscopic resection of large colorectal polyps (LCPs) are lacking.

Objective

We evaluated the effect of CO₂ insufflation on pain after endoscopic resection of LCPs.

Methods

In a prospective randomized controlled trial (RCT), 132 patients were randomly assigned to groups who underwent endoscopic resection with CO₂ insufflation (CO₂ group, n = 66) or air insufflation (air group, n = 66). The primary outcome was abdominal pain post-procedure (PP). The secondary outcomes were abdominal distension, rates of technical success, amounts of sedatives prescribed, use of analgesics, and adverse events.

Results

Baseline patient characteristics were similar between the groups. The mean abdominal pain score was 12.3 in the CO₂ group vs. 17.5 in the air group at 1 h PP (p  =  0.047). Also, the proportion of patients without pain was significantly higher in the CO₂ group at 1 h PP (p = 0.008). The pain score differed more in the endoscopic submucosal dissection group and long-time group. The secondary outcomes were not significantly different between the two groups.

Conclusions

The results of this RCT demonstrate the superiority of CO₂ insufflation for endoscopic resection of LCPs in terms of decreasing PP abdominal pain (KCT0001636).

Key summary

• 1. Summarize the established knowledge on this subject

•  • Compared with air insufflation, carbon dioxide (CO₂) insufflation reduces abdominal pain after colonoscopy.

•  • Understanding of the effects of CO₂ insufflation during endoscopic resection of large colorectal polyps (LCPs) is lacking.

• 2. What are the significant and/or new findings of this study?

•  • CO₂ insufflation during endoscopic resection of LCPs leads to a significant reduction in post-procedure abdominal pain.

•  • We found that the abdominal pain score differed more in the endoscopic submucosal dissection group and long-time group compared with the whole group.

•  • We suggest that CO₂ insufflation should be used routinely during endoscopic resection of LCPs.

Introduction

Endoscopic resection of colorectal polyps reduces the risk and mortality rate of colorectal cancer.¹ For large colorectal polyps (LCPs; diameter, ≥20 mm), endoscopic resection is more important because the risk of malignancy increases with increasing polyp size.^2,3 Insufflation is needed during endoscopic resection to facilitate visualization and make the technique more exact. Air insufflation is the standard method; however, excessive air insufflation results in greater expansion of the bowel lumen, causing abdominal pain or discomfort.^4–7 Generally, endoscopic resection takes longer to perform for LCPs than for small colorectal polyps; thus, a considerable amount of air enters the intestinal lumen with the former.^6,8–12 Compared with air, carbon dioxide (CO₂) is readily absorbed by the intestinal mucosa and eliminated via the respiratory tract. Therefore, CO₂ insufflation can reduce post-procedure (PP) intestinal distension and abdominal pain.^4–7

CO₂ insufflation reportedly decreases colonoscopy-related abdominal fullness and pain compared with air insufflation.^4,6,13–17 However, most previous studies have focused on the effects of pain relief by CO₂ insufflation during diagnostic or routine colonoscopy, and so data on therapeutic colonoscopy are limited.^6,16,18 Furthermore, clinical studies on the efficacy of CO₂ insufflation for endoscopic resection (endoscopic submucosal dissection [ESD] or endoscopic mucosal resection [EMR]) of LCPs have focused on safety rather than pain.^8,9,19,20 A PubMed search yielded only one observational cohort study of pain following endoscopic resection of LCPs.⁹ Therefore, we performed a prospective, double-blind, randomized controlled trial (RCT) to assess the effects of CO₂ insufflation on the PP pain of patients undergoing endoscopic resection of LCPs.

Materials and methods

This was a prospective, double-blind RCT conducted to assess the efficacy of CO₂ insufflation during endoscopic resection of LCPs.

Patients

From October 2015 to April 2017, we prospectively recruited consecutive patients scheduled for endoscopic resection of ≥20 mm colorectal polyps at the Gil Medical Center. Polyp size was assessed visually during colonoscopy. All procedures were performed by five expert endoscopists. The exclusion criteria were as follows: chronic obstructive pulmonary disease with retention of CO₂, moderate or severe heart failure (New York Heart Association classification III or IV), long-term (>3 months) analgesic use in the past 6 months, and refusal to participate. In total, 132 patients were enrolled in the study, and all underwent bowel preparation with 2 l polyethylene glycol solution (Coolprep; Taejoon Pharm Inc, Seoul, South Korea) or 4 l polyethylene glycol solution (Colyte; Taejoon Pharm Inc.).

Written informed consent was obtained from all enrolled patients before the procedure. We informed participants of the study aims, methods, and possible adverse events. The study protocol conforms to the ethical guidelines of the 1975 Declaration of Helsinki as reflected in a prior approval by the institution’s human research committee. The Institutional Review Board of Gil Medical Center approved this study (GCIRB2015-211, approved date: 4 August 2015) and this study was registered in the Clinical Research Information Service (KCT0001636).

Blinding and endoscopic resection procedure

Participants were randomly assigned to the CO₂ insufflation group (CO₂ group) or air insufflation group (air group) by a clinical research coordinator who was not associated with the study. The randomization was performed using computer-generated random numbers. All endoscopists, recovery-room nurses, and study subjects were blinded to the type of insufflation used. The endoscopy nurse switched the CO₂ equipment “on” and “off” as dictated by the randomization. To maintain blinding of the endoscopists, the CO₂ equipment was hidden under a fabric cover at all times.

All endoscopic procedures were performed using a CF-Q260AL or CF-H260AL colonoscope (Olympus Medical Systems Corp, Tokyo, Japan). The methods of endoscopic resection (ESD, EMR, endoscopic piecemeal mucosal resection [EPMR], and EMR after precutting [EMR-P]) were standardized.^21–25 For ESD procedures, sodium hyaluronate (Endo-MucoUp 20, BMI Korea Corp., Uiwang, South Korea) was locally injected into the submucosa. After injection, a circumferential incision was made in the mucosa, and the submucosal layer was dissected using a Dual Knife (KD-650U, Olympus Medical Systems Corp., Tokyo, Japan). The electrocautery unit (VIO 300 D; ERBE, Tübingen, Germany) was operated in endo-cut mode (effect 2, cut duration 3, cut interval 2) to make the incision and in the 40 W forced-coagulation mode to perform dissection. For EMR procedures, a solution of glycerol, epinephrine, and methylene blue was used to generate a submucosal layer cushion. After injection, a snare was used to resect the lifted lesion. During EPMR procedures, the number of resections was minimized. The most suspicious portion of malignant and submucosal invasion was resected preferentially, and the remainder was resected so that no residual tissue remained. In the case of EMR-P, an incision was made around the lesion using a knife before resecting using a snare. All procedures were performed on an inpatient basis and the durations of all procedures were recorded.

Intra- and post-procedure management

In the CO₂ group, CO₂ was administered using a CO₂-efficient endoscopic insufflator (Colosence Pro-500; Miraemedics Inc, Sung Nam, South Korea) connected to a CO₂ bottle. In the air group, air was administered using an ordinary air insufflation system.

The patients received intravenous sedation with midazolam and propofol. The target sedation level was moderate sedation as defined by the practice guidelines of the American Society of Anesthesiologists Task Force.²⁶ In the following cases, unsedated colonoscopy was performed: the patient refused sedated colonoscopy or a considerable number of posture changes was required due to polypectomy of difficult or defiant polyps. Pulse oxygen saturation, respiration rate, heart rate, and blood pressure were monitored during the procedure.

Chest and abdominal radiographs were obtained immediately after completion of the procedure. If adverse events such as colon perforation occurred, abdominal–pelvic computed tomography (APCT) was performed. In addition, laboratory tests were conducted pre-procedure and on day 1 PP. Follow-up colonoscopy was not performed PP; however, emergency colonoscopy was performed if a large amount of hematochezia was present. All patients were fasted on the day of the procedure and started a liquid diet on day 1 PP if there were no adverse events. After the procedure, if the patient complained of severe abdominal pain, an analgesic (tramadol, Tridol; Yuhan Corporation, Seoul, South Korea or propacetamol hydrochloride, Denogan; Yungjin Pharmaceutical Corporation, Seoul, South Korea) was prescribed.

Study outcomes

The primary outcome was the degree of abdominal pain at 1 h PP. Pain was recorded on a 100 mm visual analog scale (VAS). The VAS ranged from 0 (left) to 100 (right), where 0 designated no pain and 100 designated the worst pain imaginable. Abdominal pain was evaluated at 1, 3, 6, and 24 h PP. Also, subgroup analyses were performed for treatment method and procedure time.

The secondary outcomes were the extent of abdominal distension (waist circumference at the level of the umbilicus was estimated immediately before and after the procedure using a tape measure), rates of technical success (en bloc resection rate, complete resection rate [en bloc resection with histologically confirmed tumor-free lateral and vertical resection margins], and successful EMR rate [complete endoscopic lesion removal with no remnant lesion visible]), amounts of sedative drugs, frequency of analgesic prescription PP, and rates of adverse events.

Sample size calculation and statistical analysis

We carried out a prospective power analysis to determine the required sample size. The sample size calculation was based on differences in the VAS scores of the CO₂ and air groups at 1 h PP. Based on previous studies, we hypothesized that the mean VAS score of patients in the CO₂ group would be 10 mm and that that of patients in the air group would be 30 mm.^13,27,28 A sample size of 59 patients per group was required for a significance level of 5% and power of 80%. Thus, assuming a 10% dropout rate, 132 patients were required.

Statistical analyses were performed using SPSS software (ver. 23.0; IBM, Armonk, New York, USA). None of the variables exhibited deviation from a normal distribution. Categorical variables were compared using a chi square test (or Fisher’s exact test) and are expressed as percentages and frequencies. Continuous variables were compared using a t-test and are presented as means ± standard deviation. A p-value <0.05 was considered indicative of statistical significance.

Results

Patients

During the study period, 140 patients were screened for eligibility. Of them, 8 patients were excluded, and thus 132 patients were assigned randomly to receive insufflation with CO₂ (n = 66) or air (n = 66). Patients who did not complete endoscopic resection (n = 9) were excluded from the final analyses: four patients had non-lifting colorectal lesions, four had huge colorectal lesions (involving more than three-quarters of the circumference of the colorectal lumen), and one patient had a colorectal lesion with extensive fibrosis. Finally, data from 123 patients (CO₂ group, 62; air group, 61) were analyzed. Group allocations and reasons for exclusion of patients are shown in Figure 1.

Figure 1.

Patients’ flow-chart. COPD: chronic obstructive pulmonary disease; CO₂: carbon dioxide.

There were no significant differences in age, sex ratio, body mass index, medical history, or history of abdominal surgery between the two groups. The most common lesion location was the right colon in both groups (p = 0.441). The groups were similar in terms of mean tumor size (26.1 mm in the CO₂ group; 28.2 mm in the air group, p = 0.217) and mean procedure time (38.1 min in the CO₂ group; 34.4 min in the air group, p = 0.354). Bowel preparation quality, laterally spreading tumor (LST), type of endoscopic resection, histology, and frequency of sedation did not differ significantly between the two groups (Table 1).

Table 1.

Clinical characteristics of patients.

	CO2 group	Air group	p-value
Patients, no.	62	61	N/A
Lesions, no.	69	67	N/A
Age, mean ± SD, years	63.9 ± 8.9	61.2 ± 11.0	0.141
Sex (male/female)	41/21	35/26	0.318
BMI, mean ± SD, kg/m2	23.9 ± 3.3	24.4 ± 3.3	0.382
History of smoking	17 (27.4%)	13 (21.3%)	0.430
History of alcohol	26 (41.9%)	25 (41.0%)	0.915
Hypertension	25 (40.3%)	22 (36.1%)	0.627
Diabetes mellitus	15 (24.2%)	10 (16.4%)	0.282
Pulmonary disease historya	1 (1.6%)	2 (3.3%)	0.619
Cardiovascular disease historyb	4 (6.5%)	1 (1.6%)	0.365
Abdominal surgery historyc	16 (25.8%)	14 (23.0%)	0.712
Laparoscopic/open	10/6	7/7	0.737
Bowel preparationd			0.759
Excellent	47	47
Good	12	10
Fair	3	3
Poor	0	1
Location of lesion			0.441
Right colon	34	27
Left colon	23	23
Rectum	12	17
Laterally spreading tumor	40 (58.0%)	39 (58.2%)	0.978
Tumor size, mean ± SD, mm	26.1 ± 7.5	28.2 ± 11.2	0.217
Type of endoscopic resection			0.849
ESD	22	17
EPMR	23	23
EMR-P	2	2
Conventional EMR	22	25
Histology			0.252
Adenoma	46	48
Adenocarcinoma (well-differentiated/ moderately-differentiated)	20 (12/8)	13 (10/3)
Hyperplastic polyp	3	3
Otherse	0	3
Procedure time, mean ± SD, minutes	38.1 ± 21.9	34.4 ± 22.2	0.354
Sedated/unsedated	51/11	53/8	0.478

SD: standard deviation; BMI: body mass index; ESD: endoscopic submucosal dissection; EPMR: endoscopic piecemeal mucosal resection; EMR-P: endoscopic mucosal resection after precutting; EMR: endoscopic mucosal resection.

^aOf the three patients with a history of pulmonary disease, all had an old tuberculosis scar.

^bOf the patients with a history of cardiovascular disease, three had coronary artery obstructive disease, one had angina pectoris, and one patient had atrial fibrillation.

^cAbdominal surgeries included cholecystectomy, caesarean section, hysterectomy, appendectomy, radical subtotal gastrectomy, and ovarian cystectomy.

^dThe Aronchick bowel preparation scale (ABPS) was used to evaluate bowel preparation quality.

^eOther histology findings included lipoma, juvenile polyp, and lymphangioma.

Post-procedure pain measurements

Figure 2 shows mean PP pain scores. Pain scores in the CO₂ group were lower than those in the air group at all time points. The mean pain score at 1 h was significantly lower in the CO₂ group than in the air group (scores at 1 h PP: 12.3 ± 13.0 vs. 17.5 ± 16.1, p = 0.047). Figure 3 shows the proportion of patients who did not experience pain (VAS score of 0). Compared with the CO₂ group, the air group had a lower frequency of no pain at all time points. The difference was significant at 1 h PP (CO₂ vs. air group: 41.9% vs. 19.7%, p = 0.008).

Figure 2.

Visual analog scale scores for abdominal pain of the CO₂ and air groups after endoscopic resection of large colorectal polyps. Patients in the CO₂ group had significantly lower pain scores 1 hour after the procedure.

Figure 3.

Proportion of patients free of pain. The proportion of patients free of pain was higher in the CO₂ group than in the air group at all time points. The difference was significant 1 hour after the procedure.

Subgroup analyses were performed for treatment method and procedure time. Table 2 shows the mean abdominal pain scores within each subgroup (ESD group, EMR group, short-time group, and long-time group). We note that the effect of CO₂ insufflation on abdominal pain at 1 h PP was statistically significant in the ESD group. Subgroup analysis on the long-time group also revealed pain scores that were significantly lower in the CO₂ group than in the air group at 1, 6, and 24 h PP.

Table 2.

Subgroup analysis of the treatment method and procedure time.

ESD groupa	CO2 group (n = 22)	Air group (n = 17)	p-value
Abdominal pain, mean ± SD
Before procedure	1.8 ± 5.0	1.8 ± 5.3	0.974
1 h after	13.6 ± 11.8	26.5 ± 22.9	0.029
3 h after	11.8 ± 15.6	17.1 ± 23.4	0.407
6 h after	5.9 ± 15.0	10.6 ± 18.5	0.389
24 h after	2.7 ± 8.8	5.9 ± 14.2	0.399
EMR groupb	CO2 group (n = 40)	Air group (n = 44)	p-value
Abdominal pain, mean ± SD
Before procedure	0.8 ± 4.7	0.2 ± 1.5	0.490
1 h after	11.5 ± 13.7	14.1 ± 11.1	0.341
3 h after	7.5 ± 12.8	8.9 ± 12.2	0.619
6 h after	3.3 ± 10.5	5.5 ± 15.8	0.457
24 h after	1.0 ± 5.0	2.1 ± 5.9	0.386
Short-time group ( ≤ 30 minutes)c	CO2 group (n = 24)	Air group (n = 31)	p-value
Abdominal pain, mean ± SD
Before procedure	0.8 ± 2.8	0.3 ± 1.8	0.418
1 h after	9.6 ± 12.7	13.2 ± 11.7	0.274
3 h after	6.7 ± 13.7	6.1 ± 8.4	0.859
6 h after	5.0 ± 13.5	1.3 ± 4.3	0.206
24 h after	2.5 ± 8.5	0.3 ± 1.8	0.227
Long-time group ( > 30 minutes)d	CO2 group (n = 38)	Air group (n = 30)	p-value
Abdominal pain, mean ± SD
Before procedure	1.3 ± 5.8	1.0 ± 4.0	0.800
1 h after	14.0 ± 13.1	22.0 ± 18.8	0.041
3 h after	10.5 ± 13.9	16.3 ± 20.6	0.192
6 h after	3.7 ± 11.5	12.7 ± 22.0	0.049
24 h after	1.1 ± 5.1	6.0 ± 12.2	0.044

ESD: endoscopic submucosal dissection; SD: standard deviation; EMR: endoscopic mucosal resection.

^aThe median procedure time of ESD group was 50 minutes.

^bThe median procedure time of EMR group was 26 minutes.

^cThe median procedure time of short-time group was 18 minutes.

^dThe median procedure time of long-time group was 46 minutes.

Secondary outcomes

Table 3 shows the secondary outcomes. There were no significant differences in the amount of sedative administered, en bloc resection rate, complete resection rate, or successful EMR rate between the two groups. The mean abdominal circumference before and after the procedure was 85.4 ± 10.5 and 85.9 ± 9.9 cm in the CO₂ group and 85.3 ± 7.8 and 86.6 ± 8.3 cm in the air group, respectively. The mean increase in abdominal circumference was greater in the air group than in the CO₂ group; however, there were no significant differences between the two groups (CO₂ vs. air group; 0.5 ± 2.6 vs. 1.3 ± 2.5; p = 0.196). The frequency of need for analgesics PP was also not significantly different between the groups (CO₂ vs. air group; 9.7% [6/62] vs. 11.5% [7/61], p = 0.746).

Table 3.

Secondary outcomes measurement in the study.

	CO2 group (n = 62)	Air group (n = 61)	p-value
Sedative drugs, mean ± SD, mg
Midazolam	2.6 ± 1.2	2.8 ± 1.0	0.444
Propofol	210.6 ± 148.2	181.9 ± 101.4	0.250
Procedure data
En bloc resection rate	19/22 (86.4%)	14/17 (82.4%)	0.731
Complete resection rate	10/22 (45.5%)	7/17 (41.2%)	0.789
Successful EMR	41/47 (87.2%)	46/50 (92.0%)	0.440
Abdominal circumference, mean ± SD, cm
Before procedure	85.4 ± 10.5	85.3 ± 7.8	0.977
Post-procedure	85.9 ± 9.9	86.6 ± 8.3	0.747
Change	0.5 ± 2.6	1.3 ± 2.5	0.196
Number of patients receiving analgesic Adverse events	5 (8.1%)	8 (13.1%)	0.362
Perforation	0 (0.0%)	3 (4.9%)	0.119
Post-procedure bleeding	0 (0.0%)	1 (1.6%)	0.496

SD: standard deviation; EMR: endoscopic mucosal resection.

Procedure-related perforation occurred in three patients (CO₂ vs. air group; 0.0% [0/62] vs. 4.9% [3/61], p = 0.119) and PP bleeding occurred in one patient (CO₂ vs. air group; 0.0% [0/62] vs. 1.6% [1/61], p = 0.496). All of the patients with perforation had LST; of these, two patients underwent ESD and one patient underwent EPMR. All patients had the perforation discerned at the time of endoscopic resection, and endoscopic clipping was attempted immediately. However, extensive pneumoperitoneum findings on APCT continued and perforation could not be resolved. Therefore, we performed the laparotomy with surgical suturing in all patients with perforation, and as a result, we could treat the perforation. The patient with PP bleeding was treated by endoscopic hemostasis. No deaths or other serious cardiopulmonary adverse events occurred.

Discussion

This is the first RCT to report that CO₂ insufflation during endoscopic resection of LCPs leads to a significant reduction in PP abdominal pain based on a VAS. The pain level at 1 h PP was significantly lower in the CO₂ group than in the air group. In addition, the frequency of no pain at 1 h PP was significantly higher in the CO₂ group than in the air group. Since the introduction of CO₂ insufflation during gastrointestinal endoscopy, its efficacy during colonoscopy has been demonstrated.^{5,6,13,16,18,27} Colonoscopy usually requires at least 20 min, so a relatively large amount of air is insufflated. In addition, the entire colon is in contact with the insufflated air, which occupies a large portion of the abdomen. Therefore, CO₂ insufflation during colonoscopy has a marked effect on the abdomen. Endoscopic resection of LCPs is technically more challenging and requires more time; thus, CO₂ insufflation may have a greater impact on patients undergoing such procedures. However, few studies have assessed CO₂ insufflation during endoscopic resection of LCPs,^8,9,19,20 and most have focused on safety rather than pain. To the best of our knowledge, only one observational study has evaluated the effects of CO₂ insufflation on procedure-related pain after endoscopic resection of LCPs.⁹

Few studies have addressed the reduction in abdominal pain after CO₂ insufflation in patients undergoing colorectal polypectomy. To date, two studies of colorectal polypectomy have been published,^9,29 only one of which involved LCPs.⁹ Bassan et al. reported that CO₂ insufflation significantly reduces the rate of admission due to PP abdominal pain; our findings are in concordance, but our data differ for several reasons.⁹ First, the study design was different. Bassan and colleagues employed a prospective observational cohort design, whereas we conducted a prospective double-blind RCT; the latter is the optimum design for comparing interventions.³⁰ Therefore, our findings not only validate the results of Bassan et al. but also provide more reliable data. Second, Bassan et al. analyzed only the proportion of patients with abdominal pain sufficient to result in hospitalization, whereas we quantified changes in abdominal pain using a VAS. This method could show more clearly the effect of CO₂ insufflation on abdominal pain. Finally, they evaluated only EMR, where we compared various methods, including ESD.

We also performed subgroup analyses for treatment method and procedure time. We found that the abdominal pain scores differed more in the ESD group and the long-time group compared with the whole group. In the ESD group, the effect of CO₂ insufflation on abdominal pain at 1 h PP was greater than that in the whole group. Similarly, in the long-time group, pain scores were significantly lower in the CO₂ group than in the air group, not only at 1 h PP but also at 6 and 24 h PP. The total volume of gas used to complete a colonoscopy may be influenced by the type of procedure or procedure duration, and this volume of insufflated gas is one of the important factors affecting abdominal pain. Therefore, the pain-reduction effect of CO₂ insufflation is more remarkable when the procedure time is prolonged or when complicated procedures such as ESD are performed.

In our study, the difference in pain scores between CO₂ group and air group was modest and limited to 1 h after the procedure. There are two reasons for this. First, endoscopic resection of LCPs in this study was performed only by selected expert endoscopists (no involvement of fellow or less-experienced endoscopists). Therefore, endoscopists conducted only the necessary amount of insufflations for patients, and did not cause excessive pain. Both groups had relatively low levels of abdominal pain, which may have led to the modest difference in pain levels between the two groups. Second, when the subgroup analysis was carried out, the effect of reducing CO₂ insufflation in the ESD group was greater compared with the EMR group. Although this study only targeted the endoscopic resection for LCPs, the treatment methods were heterogeneous. This heterogeneity may have also affected the modest difference in pain scores in the whole group.

Residual gas in the bowel lumen is a major cause of abdominal pain after colonoscopy. To evaluate bowel gas retention, we assessed the extent of abdominal distension by measuring the change in waist circumference. The increase in waist circumference PP was lower in the CO₂ group than in the air group, albeit not significantly so. Previous studies have evaluated the volume of gas insufflated in the intestinal lumen by abdominal radiography or abdominal circumference measurements.^14,31,32 In two previous studies, the amount of residual bowel gas in the colon and small bowel was significantly less with CO₂ insufflation.^14,31 However, another study reported that there were no significant differences in waist circumference between CO₂ and air insufflation, with which our findings are consistent.³² The discrepancies among studies are likely due to the use of different measurement methods. Abdominal circumference measurement is an indirect indicator of residual gas; therefore, it is less accurate than abdominal radiography. We elected to measure waist circumference because this method is simple, low cost, and does not involve exposure to radiation.

There were no significant differences between the groups in the rates of en bloc resection, complete resection, and successful EMR. The effects of type of insufflation on the success rate of LCP removal are unclear. Bassan et al. reported that there were no significant differences in the overall EMR success rate between air and CO₂ insufflation.⁹ Moreover, Saito et al. demonstrated that there were no differences in the en bloc resection rate between air and CO₂ insufflation; however, the free margin rate was significantly higher in the air group.⁸ The type of insufflation had little impact on the procedure success rate, likely because while polyp size, shape, and location are determinants of successful endoscopic resection, insufflation is not.^12,33,34

No severe cardiopulmonary adverse events occurred. The safety of CO₂ insufflation during colonoscopy has already been confirmed in many previous studies.^6,18,35 In our study, all adverse events occurred in the air group; however, there were no significant differences in the rate of perforation or bleeding. These adverse events are not thought to be directly related to excessive insufflation, that is, barotrauma. In our opinion, most perforations were due to direct thermal or mechanical injury to the muscular layer by knives and snares; therefore, perforations were recognized during the endoscopic resection procedures. Nevertheless, we believe that CO₂ insufflation during endoscopic resection of LCPs affords a clear advantage. Colorectal polypectomy is associated with a risk of perforation, which may cause pneumoperitoneum, severe abdominal pain, or abdominal compartment syndrome.^3,6,10,33,34 Escape of insufflated gas through a perforation into the abdominal cavity can create cardiopulmonary distress associated with a degenerated tension pneumoperitoneum. CO₂ is absorbed quickly, thus reducing any secondary detrimental effects of perforation.⁶ Therefore, we recommend that CO₂ insufflation should be routine during endoscopic procedures associated with a high perforation risk such as endoscopic resection of LCPs.

Several limitations of the present study need to be considered. First, it involved only a single center. Second, sedation was not performed in some patients (15.4%). However, there were no significant differences in the proportion of sedated patients between the two groups. The results demonstrate that CO₂ insufflation during endoscopic resection of LCPs significantly reduces PP abdominal pain. Moreover, CO₂ insufflation is simple and did not adversely affect outcomes. Our results and other recent studies suggest that CO₂ insufflation should be used routinely during endoscopic resection of LCPs.

Conflict of interest

The authors declare no conflict of interest.

Funding

This research was supported by a grant of the Korea Health Technology R&D Project through the Korea Health Industry Development Institute (KHIDI), funded by the Ministry of Health & Welfare, Republic of Korea (grant number : HI16C2319).

Ethics approval

The study was registered in the Clinical Research Information Service (KCT0001636).

Informed consent

Written, informed consent was obtained from each patient