Eur Radiol (2014) 24:1487–1496 DOI 10.1007/s00330-014-3190-1
GASTROINTESTINAL
Perforation rate in CT colonography: a systematic review of the literature and meta-analysis Davide Bellini & Marco Rengo & Carlo Nicola De Cecco & Franco Iafrate & Cesare Hassan & Andrea Laghi
Received: 17 January 2014 / Revised: 5 April 2014 / Accepted: 15 April 2014 / Published online: 10 May 2014 # European Society of Radiology 2014
Abstract Purpose The primary aim was to assess the perforation rate of CTC; the secondary aim was to identify potential clinical/ technical predictors of this complication. Methods Methods for analysis were based on PRISMA (preferred reporting items for systematic reviews and metaanalyses). From the selected studies, the rate of CTC perforation and patient/technical characteristics potentially associated with this event were extracted. Forest plots showing individual and pooled estimates of the perforation rate were obtained for all analyses. I2 was used to evaluate heterogeneity between studies. Results Eleven articles out of the 187 initially identified were selected for the analysis (103,399 patients). There were 29,048 (28 %) asymptomatic individuals and 30,773 (30 %) symptomatic patients; this characteristic was not reported in the remaining subjects (42 %). Colon distension was obtained manually in 69,222 (67 %) and using an automated carbon dioxide insufflator in 26,479 (26 %) patients; in the remaining 7 % of patients, this information was missing. Twenty-eight colonic perforations were reported, with the CTC perforation rate estimated to be 0.04 % (95 % CI. 0.00-0.10), 19-fold higher in symptomatic than in screening subjects (OR: 19.2, D. Bellini (*) : M. Rengo : C. N. De Cecco : A. Laghi Department of Radiological Sciences, Oncology and Pathology, “Sapienza” University of Rome, ICOT Hospital, Via Franco Faggiana 34, 04100 Latina, Italy e-mail: [email protected] F. Iafrate Department of Radiological Sciences, Oncology and Pathology, “Sapienza” University of Rome, Policlinico Umberto I Hospital, Rome, Italy C. Hassan Gastroenterology and Digestive Endoscopy Unit, Nuovo Regina Margherita Hospital, Rome, Italy
CI 3.3-108 and P=0.001). The surgical rate was 0.008 %. No CTC-related deaths were reported. Conclusions The perforation rate in CTC is very low, particularly considering asymptomatic individuals. Key Points • This is the first meta-analysis on this topic, based on 100,000 patients. • The CTC-related colorectal perforation rate is 0.04 %, 0.02 % in asymptomatic subjects. • The CTC-induced surgery rate is 0.008 % (1:12,500). • The perforation rate in CTC is low, particularly in averagerisk, asymptomatic individuals. Keywords Computed tomographic colonography . Intestinal perforation . Mass screening . Insufflation . General surgery
Abbreviations and acronyms CTC Computed tomographic colonography CRC Colorectal cancer CC Conventional colonoscopy ES Effect size OR Odds ratio PR Perforation rate DCBE Double-contrast barium enema
Introduction Computed tomographic colonography (CTC) is a minimally invasive and accurate imaging technique, able to detect already developed colorectal cancer (CRC) and adenomatous polyps in both symptomatic and asymptomatic patients [1–5]. According to recent guideline recommendations, CTC,
1488
performed every 5 years, is now recognised as an acceptable alternative screening option by both radiological and gastrointestinal societies [6–8]. CTC has been shown to be highly accurate [9] for the detection of CRC and clinically relevant polyps [10], and is also comparable to conventional colonoscopy (CC). Moreover, CTC is generally preferred by patients, especially because of its lighter bowel preparation, faster noninvasive implementation and very uncommon related adverse events [11–15]. Despite being generally regarded as safer than CC [16], CTC has been shown to be associated with potentially serious adverse events, mainly represented by large bowel perforation [17, 18]. Until now, few research articles have aimed to estimate the rate of bowel perforation and few other additional case reports have been published [18–27]. Cumulative data from multicentre series documented a perforation rate for CTC in a range from 0.009 % [20] to 0.05 % [21]. However, when considering the relative rarity of the event, any individual study was underpowered to provide a reliable estimate of the real rate of CTC perforation. The primary aim of this systematic review and meta-analysis was to assess the perforation rate of CTC. The secondary aim was to identify potential clinical/technical predictors of this complication.
Materials and methods Methods for analysis and inclusion criteria were based on PRISMA (preferred reporting items for systematic reviews and meta-analyses) recommendations for systematic reviews and meta-analyses [28]. Literature search The MEDLINE, Cochrane Library, Sumsearch2 and Web of Science databases (from inception to August 2013) were searched independently by two observers for studies that reported perforation rates in CTC. The following keywords were used for the search: virtual colon* (thus including colonoscopy, colonography), computed tomographic colon* (thus including colonoscopy, colonography) and CT colon* (thus including colonoscopy, colonography), combined with perforation, intestinal perforation and bowel perforation. Our selection criteria were broad in order to include as many studies as possible. The search was performed without any language restriction, with the presence of the search terms in the title or abstract of the article, but it was limited to human subjects. Comments, letters, review articles, case reports and unpublished data were excluded by the search. Additional exclusion criteria included studies with fewer than 500 patients. Potentially eligible papers were initially screened by two reviewers (A.L., an abdominal radiologist with more than 15 years of experience with CTC, and D.B., a radiology
Eur Radiol (2014) 24:1487–1496
resident with 4 years of CTC experience) on the basis of title, abstract and keywords; full articles were retrieved for further assessment if given information suggested that the study might include perforations in CTC and the references of the collected articles were crosschecked for further relevant studies. Evaluation of eligibility and selection of articles was performed independently by the two radiologists; different opinions in study selection were resolved by consensus. Data extraction Relevant data of selected studies were independently extracted by the radiologists using a data extraction form. Differences in data collection were resolved by consensus with a third reviewer (M.R., an abdominal radiologist with more than 10 years of experience in CTC) referring back to the original article. From each primary study, the following sample characteristics were extracted: (1) year of publication; (2) country where the study was performed; (3) whether it was a single or a multicentre study; (4) if multicentre, number of centres involved; (5) number of patients; (6) number of patients who underwent CTC for screening; (7) number of patients who underwent CTC because symptomatic; (8) technique for colon distension, whether manual with room air or automatic with carbon dioxide; (9) number of perforated patients; (10) temporal distribution of perforations; (11) modality for data collection (self-reported or questionnaire); (12) centre where the experience was recorded. Additionally, for each one of the perforated patients, (1) the clinical indication for CTC, (2) age, (3) whether CC was performed and how long before CTC, (4) whether colon diseases were present (colon cancer, acute diverticular disease, active inflammatory bowel disease, colonic hernia, bowel obstruction), (5) institutional experience, (6) site of perforation, (7) distribution of gas, (8) symptoms associated with perforation, if any, and (9) type of treatment (conservative versus surgery). Multiple attempts were made to contact authors if data presentation was incomplete or if it was necessary to resolve an apparent conflict or inconsistency in the article. To assess the methodological quality of the included primary studies and to detect potential bias, we used items from the Quality Assessment of Diagnostic Accuracy Studies, or QUADAS, tool that were relevant for our analysis [29]. Moreover, we recorded whether patients had undergone CC before CTC. Summary measures The primary end point of this systematic review and metaanalysis was to assess the perforation rate in CTC. Secondary end points were: (1) assessment of CTC perforation risk in symptomatic patients versus asymptomatic screening individuals and (2) the possible association between the modality of
Eur Radiol (2014) 24:1487–1496
colon distension (manual air insufflation versus automatic insufflation of carbon dioxide) and risk of CTC perforation. Statistical methods For each analysis, the effect size (ES), reported as the Z value, and percentage of heterogeneity between studies, computing I2 values, were calculated. Between-study heterogeneity was analysed using the following equation: I2 =[(Q−df)/Q]× 100%, where Q was the chi-squared statistic and df was the degree of freedom. Values of I2 equal to 25 %, 50 % and 75 % were assumed to represent low, moderate and high heterogeneity respectively. This describes the percentage of the variability in effect estimates resulting from heterogeneity rather than sampling error (chance). First, meta-analyses were performed by computing the event rate in one group (patients who underwent CTC, patients who underwent CTC for screening and patients who underwent CTC because symptomatic) using a random effects model. The perforation rate and 95 % confidence intervals (CIs) were calculated. Second, meta-analyses were performed by computing the odds ratio Fig. 1 Flowchart of metaanalysis (PRISMA flow diagram)
1489
(OR) using a random effects model. Odds ratios and 95 % confidence intervals (CIs) for PR in CTC were calculated between groups (screening versus symptomatic population and carbon dioxide versus manually distended population). Statistical significance was assigned at P
GASTROINTESTINAL
Perforation rate in CT colonography: a systematic review of the literature and meta-analysis Davide Bellini & Marco Rengo & Carlo Nicola De Cecco & Franco Iafrate & Cesare Hassan & Andrea Laghi
Received: 17 January 2014 / Revised: 5 April 2014 / Accepted: 15 April 2014 / Published online: 10 May 2014 # European Society of Radiology 2014
Abstract Purpose The primary aim was to assess the perforation rate of CTC; the secondary aim was to identify potential clinical/ technical predictors of this complication. Methods Methods for analysis were based on PRISMA (preferred reporting items for systematic reviews and metaanalyses). From the selected studies, the rate of CTC perforation and patient/technical characteristics potentially associated with this event were extracted. Forest plots showing individual and pooled estimates of the perforation rate were obtained for all analyses. I2 was used to evaluate heterogeneity between studies. Results Eleven articles out of the 187 initially identified were selected for the analysis (103,399 patients). There were 29,048 (28 %) asymptomatic individuals and 30,773 (30 %) symptomatic patients; this characteristic was not reported in the remaining subjects (42 %). Colon distension was obtained manually in 69,222 (67 %) and using an automated carbon dioxide insufflator in 26,479 (26 %) patients; in the remaining 7 % of patients, this information was missing. Twenty-eight colonic perforations were reported, with the CTC perforation rate estimated to be 0.04 % (95 % CI. 0.00-0.10), 19-fold higher in symptomatic than in screening subjects (OR: 19.2, D. Bellini (*) : M. Rengo : C. N. De Cecco : A. Laghi Department of Radiological Sciences, Oncology and Pathology, “Sapienza” University of Rome, ICOT Hospital, Via Franco Faggiana 34, 04100 Latina, Italy e-mail: [email protected] F. Iafrate Department of Radiological Sciences, Oncology and Pathology, “Sapienza” University of Rome, Policlinico Umberto I Hospital, Rome, Italy C. Hassan Gastroenterology and Digestive Endoscopy Unit, Nuovo Regina Margherita Hospital, Rome, Italy
CI 3.3-108 and P=0.001). The surgical rate was 0.008 %. No CTC-related deaths were reported. Conclusions The perforation rate in CTC is very low, particularly considering asymptomatic individuals. Key Points • This is the first meta-analysis on this topic, based on 100,000 patients. • The CTC-related colorectal perforation rate is 0.04 %, 0.02 % in asymptomatic subjects. • The CTC-induced surgery rate is 0.008 % (1:12,500). • The perforation rate in CTC is low, particularly in averagerisk, asymptomatic individuals. Keywords Computed tomographic colonography . Intestinal perforation . Mass screening . Insufflation . General surgery
Abbreviations and acronyms CTC Computed tomographic colonography CRC Colorectal cancer CC Conventional colonoscopy ES Effect size OR Odds ratio PR Perforation rate DCBE Double-contrast barium enema
Introduction Computed tomographic colonography (CTC) is a minimally invasive and accurate imaging technique, able to detect already developed colorectal cancer (CRC) and adenomatous polyps in both symptomatic and asymptomatic patients [1–5]. According to recent guideline recommendations, CTC,
1488
performed every 5 years, is now recognised as an acceptable alternative screening option by both radiological and gastrointestinal societies [6–8]. CTC has been shown to be highly accurate [9] for the detection of CRC and clinically relevant polyps [10], and is also comparable to conventional colonoscopy (CC). Moreover, CTC is generally preferred by patients, especially because of its lighter bowel preparation, faster noninvasive implementation and very uncommon related adverse events [11–15]. Despite being generally regarded as safer than CC [16], CTC has been shown to be associated with potentially serious adverse events, mainly represented by large bowel perforation [17, 18]. Until now, few research articles have aimed to estimate the rate of bowel perforation and few other additional case reports have been published [18–27]. Cumulative data from multicentre series documented a perforation rate for CTC in a range from 0.009 % [20] to 0.05 % [21]. However, when considering the relative rarity of the event, any individual study was underpowered to provide a reliable estimate of the real rate of CTC perforation. The primary aim of this systematic review and meta-analysis was to assess the perforation rate of CTC. The secondary aim was to identify potential clinical/technical predictors of this complication.
Materials and methods Methods for analysis and inclusion criteria were based on PRISMA (preferred reporting items for systematic reviews and meta-analyses) recommendations for systematic reviews and meta-analyses [28]. Literature search The MEDLINE, Cochrane Library, Sumsearch2 and Web of Science databases (from inception to August 2013) were searched independently by two observers for studies that reported perforation rates in CTC. The following keywords were used for the search: virtual colon* (thus including colonoscopy, colonography), computed tomographic colon* (thus including colonoscopy, colonography) and CT colon* (thus including colonoscopy, colonography), combined with perforation, intestinal perforation and bowel perforation. Our selection criteria were broad in order to include as many studies as possible. The search was performed without any language restriction, with the presence of the search terms in the title or abstract of the article, but it was limited to human subjects. Comments, letters, review articles, case reports and unpublished data were excluded by the search. Additional exclusion criteria included studies with fewer than 500 patients. Potentially eligible papers were initially screened by two reviewers (A.L., an abdominal radiologist with more than 15 years of experience with CTC, and D.B., a radiology
Eur Radiol (2014) 24:1487–1496
resident with 4 years of CTC experience) on the basis of title, abstract and keywords; full articles were retrieved for further assessment if given information suggested that the study might include perforations in CTC and the references of the collected articles were crosschecked for further relevant studies. Evaluation of eligibility and selection of articles was performed independently by the two radiologists; different opinions in study selection were resolved by consensus. Data extraction Relevant data of selected studies were independently extracted by the radiologists using a data extraction form. Differences in data collection were resolved by consensus with a third reviewer (M.R., an abdominal radiologist with more than 10 years of experience in CTC) referring back to the original article. From each primary study, the following sample characteristics were extracted: (1) year of publication; (2) country where the study was performed; (3) whether it was a single or a multicentre study; (4) if multicentre, number of centres involved; (5) number of patients; (6) number of patients who underwent CTC for screening; (7) number of patients who underwent CTC because symptomatic; (8) technique for colon distension, whether manual with room air or automatic with carbon dioxide; (9) number of perforated patients; (10) temporal distribution of perforations; (11) modality for data collection (self-reported or questionnaire); (12) centre where the experience was recorded. Additionally, for each one of the perforated patients, (1) the clinical indication for CTC, (2) age, (3) whether CC was performed and how long before CTC, (4) whether colon diseases were present (colon cancer, acute diverticular disease, active inflammatory bowel disease, colonic hernia, bowel obstruction), (5) institutional experience, (6) site of perforation, (7) distribution of gas, (8) symptoms associated with perforation, if any, and (9) type of treatment (conservative versus surgery). Multiple attempts were made to contact authors if data presentation was incomplete or if it was necessary to resolve an apparent conflict or inconsistency in the article. To assess the methodological quality of the included primary studies and to detect potential bias, we used items from the Quality Assessment of Diagnostic Accuracy Studies, or QUADAS, tool that were relevant for our analysis [29]. Moreover, we recorded whether patients had undergone CC before CTC. Summary measures The primary end point of this systematic review and metaanalysis was to assess the perforation rate in CTC. Secondary end points were: (1) assessment of CTC perforation risk in symptomatic patients versus asymptomatic screening individuals and (2) the possible association between the modality of
Eur Radiol (2014) 24:1487–1496
colon distension (manual air insufflation versus automatic insufflation of carbon dioxide) and risk of CTC perforation. Statistical methods For each analysis, the effect size (ES), reported as the Z value, and percentage of heterogeneity between studies, computing I2 values, were calculated. Between-study heterogeneity was analysed using the following equation: I2 =[(Q−df)/Q]× 100%, where Q was the chi-squared statistic and df was the degree of freedom. Values of I2 equal to 25 %, 50 % and 75 % were assumed to represent low, moderate and high heterogeneity respectively. This describes the percentage of the variability in effect estimates resulting from heterogeneity rather than sampling error (chance). First, meta-analyses were performed by computing the event rate in one group (patients who underwent CTC, patients who underwent CTC for screening and patients who underwent CTC because symptomatic) using a random effects model. The perforation rate and 95 % confidence intervals (CIs) were calculated. Second, meta-analyses were performed by computing the odds ratio Fig. 1 Flowchart of metaanalysis (PRISMA flow diagram)
1489
(OR) using a random effects model. Odds ratios and 95 % confidence intervals (CIs) for PR in CTC were calculated between groups (screening versus symptomatic population and carbon dioxide versus manually distended population). Statistical significance was assigned at P
