Effect of Antifoaming Agent on Colorectal Benign Tumours in Colonoscopy: A Meta-Analysis

Background: Although several trials have showed that addition of antifoaming agent to polyethylene glycol (PEG) can improve bowel preparation, whether PEG plus antifoaming agent has a benecial role in detection of benign tumours during colonoscopy has yet to be conrmed. Our aim was to clarify whether adding simethicone to PEG solution could improve the effect of detection of colorectal benign tumours. Methods: The PubMed, EMBASE, and Cochrane Library databases were searched for articles published prior to September 2019. The outcomes included the detection rates of colorectal adenomas and polyps. Results: Twenty studies were eligible. Although there was no difference in the colorectal adenoma detection rate (ADR), a signicant effect of simethicone for diminutive adenomas (< 10 mm) was revealed in the group with simethicone. We also found that simethicone could signicantly improve the ADR in proximal colon, but did not affect the colorectal polyp detection rate (PDR). Furthermore, the subgroup analyses revealed that the benecial effect of simethicone in the ADR among Asian (P=0.005) and ADR < 25% (P =0.003). Moreover, it was a signicant nding that the low dose simethicone was as effective as the high dose one with respect to the detection of colorectal benign tumours. Conclusions: In summary, the addition of simethicone to PEG might benet to improve diminutive adenomas in the right colon for colonoscopy in Asia. The low-dose simethicone was recommend to the detection of colorectal benign tumours. However, large clinical trials are necessary to validate our results and determine the ideal dose of simethicone.


Background
Colorectal cancer (CRC) is one of the most common cancers worldwide. The incidence and mortality of CRC have been rapidly increasing in Asian countries [1,2]. Early diagnosis is associated with better survival and quality of life. Currently, colonoscopy is a standard rst line tool in screening, surveillance, and prevention of colorectal tumours [3,4]. The colorectal adenoma detection rate (ADR) has been recognized as the most indicator of the quality of colonoscopy. In the consensus statement, polyethylene glycol (PEG) is recommended as the rst choice for bowel preparation [5]. However, up to 24% of patients have been shown to have a poor quality of bowel preparation [6]. Inadequate bowel preparation is associated with an increased risk of missed colorectal benign tumours and greater discomfort for the patient [7][8][9].
Simethicone, preventing bubble formation and gas retention to alleviate bloating, is an effective and safe antifoaming agent during endoscopic procedures. A combination of simethicone and PEG has been shown to improve the visualization of the bowel for colonoscopy. Thus, simethicone may have a theoretical bene t in the detection of benign tumours in colonoscopy, especially diminutive lesions.
A large number of previous studies have evaluated the effect of simethicone in ADR during colonoscopy, but the results are inconsistent. Hence, a recently meta-analysis would be necessary. However, whether simethicone plus PEG has a bene cial role in detection of benign tumours during colonoscopy has yet to be con rmed. Therefore, we performed a meta-analysis to investigate the effect of detection of colorectal benign tumours.

Literature search
The literature was searched in PubMed, EMBASE, and Cochrane Central databases (up to September 1, 2019) using the keywords 'colonoscopy', 'antifoaming agent' or 'simethicone', and 'randomized'. We also performed a manual search of the reference list of published articles.

Data Extraction
The three investigators (HZ, JL and LM) were independently assigned to extract the data from each primary manuscript. Discrepancies or disagreements were resolved by consensus. The extracted data included the rst author, year of publication, country or region, number of patients, study design, detailed information of interventions and controls (ADR and PDR), and adverse events.

Page 4/26
The Cochrane Collaboration's risk of bias tool [10], in which the description of sequence generation, allocation concealment, blinding of participants, personnel and outcome assessors, incomplete outcome data and selective outcome reporting assessed, was used to evaluate the quality of the randomized studies. The quality scale was assessed as 'low risk of bias', 'unclear risk of bias' and 'high risk of bias'.

Data Syntheses and Statistical Analyses
The odds ratio (OR) was used for discrete variables, and the mean difference (MD) and standardized difference in mean (SDM) were used for continuous variables. The pooled OR and 95% con dence intervals (CIs) were calculated from each study using either a xed-effects model (Mantel-Haenszel method) or a random-effects model (Der Simonian and Laird method). When the heterogeneity was signi cant, the random-effects model was used in the pooled data; otherwise, the xed-effects model was used. Heterogeneity among the studies was assessed using I 2 statistic or χ 2 test. I 2 > 50% or P < 0.10 were considered to represent signi cant heterogeneity. The publication bias was evaluated by the Egger's test, where P < 0.1 in a two-tailed test was considered to be positive. In the subgroup analyses, P < 0.05 for the χ 2 test indicated statistically a signi cant heterogeneity. For all the pooled results, a sensitivity analysis was conducted to examine the robustness by excluding one or more studies each time [11]. All the statistical analyses and plots were performed using the Review Manager statistical software version 5.0 (The Cochrane Collaboration, Copenhagen, Denmark) and Stata version 12.0 (Stata Corp LLC, Texas, USA).

Study Selection
The literature search retrieved 169 citations, 96 of which were excluded because they were duplicates. Of the 73 potentially eligible studies, 53 publications were excluded, and 20 studies focused on comparing PEG with and without simethicone evaluated the effect of ADR and PDR. This meta-analysis was conducted in accordance with the standard guidelines of the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) [12] (Fig. 1).

Quality assessment
The Cochrane Collaboration's risk of bias tool was used to evaluate the quality of the randomized studies. Although all studies were single-blind to the examiner, the participants and the blinding of outcome assessment were not in uence. Thereby the risk bias of selective reporting of each trial was considered low risk. The quality assessment of randomized studies is shown in supplementary  (Fig. 3). The Egger's regression test revealed that there was no signi cant effect of publication bias on the overall PDR (P = 0.221).

Secondary endpoints
Adverse events 16

Sensitivity analyses
We performed further sensitivity analyses to assess the impact on the heterogeneity by the exclusion of one or more studies at a time. There was statistically signi cant heterogeneity for ADR in the right colon (heterogeneity P = 0.09, I 2 = 58%). When Bai's study was excluded, it no longer showed heterogeneity of ADR (heterogeneity P = 0.18, I 2 = 45%). The other two outcomes had signi cant heterogeneity, included PDR and the adverse event of bloating. When Valiante's study was excluded, it no longer showed heterogeneity of PDR. The studies associated with the heterogeneity of each outcome were listed in Table 3.

Subgroup analyses
The results of the subgroup analyses for the ADR and PDR in relation to sites of colorectal adenomas or polyps (right or left colon), sizes of adenomas or polyps (≥ 10 mm or < 10 mm), populations (Asia or non-Asia), dose of simethicone (≥ 400 mg or < 400 mg and NR), and proportion of ADR (≥ 25% or < 25%) are shown in Table 3.
When analyzed separately, a signi cantly higher proportion of ADR in the right colon was present in the PEG plus simethicone group (21.5% vs 9.7%, OR = 2.61, 95% CI 1.43-4.76, P = 0.002) (Fig. 4b). In addition, the ADR in the left colon was also higher than that in the PEG group with borderline statistical signi cance (13.8% vs 10.0%, P = 0.04).
The subgroup analysis revealed a signi cative increase of ADR in the studies from Asian in the PEG with simethicone group (26.8% vs. 20.7%, OR = 1.45, 95% CI 1.12-1.87, P = 0.005) (Fig. 4c), and the baseline ADR < 25% of the studies included was associated with the signi cant bene t of simethicone (OR = 1.55, 95% CI 1.16-2.07, P = 0.003) (Fig. 4d). In addition, our analysis revealed that there was no signi cant difference of ADR between the two groups with respect to the dose of simethicone, suggesting that the low-dose simethicone was as effective as the high-dose one with respect to the detection of colorectal benign tumours.
The comparison of PDR between the two groups showed no differences for proportion of PDR, dose of simethicone, size of polyps, and the populations, when simethicone was added.

Conclusions
Simethicone is an effective antifoaming agent during endoscopic procedures. The Gastroenterological Society of Australia consensus panel found that the current evidence supported the use of simethicone for improving visibility and likely facilitates adenoma detection at colonoscopy [33]. In addition, a recently study reported that a 10% increase in the detection rate of colorectal polyp when simethicone is added to the water pump during colonoscopy [34].
Although simethicone adding to PEG solution could improve bowel cleanness and mucosal visibility [35], out results found that simethicone did not the affect the total ADR and PDR. It may be related to the possible explanation that solid stool was unlikely to be cleaned, despite simethicone could improve the overall bowel cleanness. ADR has been recognized as the most indicator for quality of the colonoscopy. The current international guidelines have recommended the ADR should be ≥ 25% overall as the minimal requirement of the surveillance colonoscopy [36] In our subgroup analysis, we revealed a positive effect of simethicone with statistical signi cance in the low ADR group (< 25%) and the studies from Asia. An interesting nding in our study was that the population of the low ADR group were all from Asian. This phenomenon might be related to the gene, diet and microbiota of Asian.
Furthermore, our subgroup analysis revealed that simethicone could signi cantly improve the small ADR (< 10 mm) of the proximal colon. The main reason is that simethicone can improve the bowel preparation, especially in the right colon [37]. As bubbles are more commonly present in the proximal colon, especially the ascending colon, the bubble elimination could enhance the ability to detect smaller proximal adenomas. Previous study revealed that missed cancers of the proximal colon were more often found in the poor bowel preparation [38]. But simethicone did not affect the ADR in the left colon signi cantly, which may be associated with the small sample of studies included. Therefore, further large clinical trials are necessary to con rm our results.
The addition of 2-3 mL of 120 mg/mL simethicone to lavage uid was recommend [33]. In the subgroup analysis, we compared the effect of the low dose simethicone (< 400 mg) versus that of high dose simethicone (≥ 400 mg) for ADR and PDR. Our results also revealed that simethicone with high or low dose had no signi cant difference of ADR and PDR, suggesting that the low-dose simethicone was not inferior to the high-dose one in terms of ADR and PDR, which is similar to the study of Li et al [39]. Although residual simethicone in biopsy channels could promote bio lm formation [40], simethicone adding to PEG solution could decrease the infection risk of endoscope-transmission [31].However, the optimal dose of SIM is yet to be ascertained [41]. Further research is required to determine the optimal dose of simethicone in clinical practice.
The strengths of our study were as follows. First, the subgroup analyses and sensitivity analyses were performed to explore the potential causes. To minimize potential bias, we performed sensitivity analyses to assess the impact on the heterogeneity by excluding one or more studies at a time and performed subgroup analyses according to the site and size of colorectal benign tumours, the population included, and the proportion of ADR. There was no signi cant heterogeneity found in the meta-analysis of ADR, except for the right-side ADR. When Valiante's study was excluded, it no longer showed heterogeneity of PDR. Second, our results of the subgroup analyses for ADR and PDR included the population included and the dose of simethicone before colonoscopy. Thirdly, 20 studies were eligible for our meta-analysis. A large number of included studies allowed for rm conclusions and adequate subgroup analyses.
Therefore, the results of our study are convincing.
There are several limitations to our meta-analysis. First, our meta-analysis was restricted to publications written in English, which might have produced a potential selection bias. Second, all included studies were single-blinded for outcome assessment; therefore, further double-blind RCTs should be conducted to con rm the positive effects of simethicone. Third, demographic and procedure data, such as race, diet, microbiota, and genes, may have been interesting to evaluate, but those data were not analyzed due to the limited condition. Fourth, although endoscopists were trained adequately, the effect of observer bias cannot be ignored.
In conclusion, we believe that simethicone might bene t to improve small ADR, especially in the proximal colon, for colonoscopy in Asian with low baseline ADR. The simethicone with low dose was not inferior to that with high dose with respect to the detection of colorectal benign tumours. Availability of data and materials

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.