Abstract
Background/Aim: Colon interposition counts among the most common techniques for reconstruction after esophagectomy. Availability of data on metachronous mucosal pathologies is weak. The aim of this review was to identify all reports on the development of metachronous adenoma and adenocarcinoma in colon interposition after esophagectomy in adulthood. Materials and Methods: A comprehensive search was conducted in MEDLINE/PubMed, Science Direct, Cochrane Library, Bayerische Staatsbibliothek München. All studies reporting on patients who received colon interposition as substitute after esophagectomy in adulthood for benign and malignant reasons were included. Results: Five retrospective studies were included, reporting on 1016 patients. Therein, no interval lesion was identified. One further study, which formally must be excluded for a misfit to inclusion criteria reports on three interval carcinomas within 365 patients. Because these lesions were the only ones found within a cohort analysis, results were supplementary reported in this review. Additionally, 31 case reports including 32 patients with benign (n=7) or malignant (n=25) findings were analyzed. Median age was 63.5 years (interval carcinoma) and 69 years (benign lesion). Benign and malignant lesions were diagnosed after a median of 8.5 years. Conclusion: Due to the rareness of respective cohort studies, the frequency of metachronous lesions cannot be calculated accurately. The estimated rate of interval carcinoma is 0-0.22%. Life-long endoscopic surveillance of patients with colon interposition is recommended.
Esophagectomy with replacement by colon interposition was first described by Kelling and Vuillet in 1911 (1, 2). Since then, the colon developed into the most commonly used esophageal substitute if the stomach is not available for reconstruction (3). The majority of patients undergoing esophagectomy with colon interposition for malignant disease are older than 55 years and thereby at risk for the development of colonic polyps. Further, patients with adenocarcinoma of the esophagus showed a significantly higher risk for the occurrence of colonic polyps (4). Patients, who receive the interposition procedure for benign pathologies are relevantly younger in common, therefore their lifetime risk for the development of mucosal changes in the substitute is relatively higher. To date, no systematic review focuses on the rate of mucosal changes such as adenoma and adenocarcinoma in colon interposition in adulthood.
This review was performed to identify available evidence on the development of metachronous adenoma and adenocarcinoma in colon interposition after esophagectomy for benign (e.g. Boerhaave-Syndrom, trauma, corrosive injury) and malignant reasons (esophageal cancer) in adulthood.
Materials and Methods
Protocol and registration. The systematic review is based on a review protocol, designed in accordance to the PRISMA-P-Statement (5-7).
The protocol is registered with the PROSPERO International Prospective Register of Systematic Reviews as CRD42017082144.
Eligibility criteria. All studies (randomized, non-randomized, case series, case reports), reporting on patients who received colon interposition after esophagectomy for benign (e.g. Boerhaave-Syndrom, trauma, caustic injury) and malignant reasons (adenocarcinoma or squamous cell carcinoma of the esophagus) were eligible for data analysis. The search was not limited to a certain time frame. Studies had to be fully published, available as full text and written in English or German language.
Exclusion criteria. Exclusion from the analysis: colon interposition in children, abstract/full-text was not available, not in English or German language, review articles.
Information sources and search strategy. A bibliographic search was conducted in Medline/PubMed, Science Direct, Cochrane database and Bayerische Staatsbibliothek from 12-1-2017 to 4-1-2018. In two cases, authors were contacted for full text transfer. In 13 cases of older full-text articles, not available in one of the abovementioned databases, full-text articles were directly requested from editorial offices of the publishing journals. Combinations of the following terms were used: “polyp”, “interposition”, “colon” (“polyps”[MeSH Terms] OR “polyps”[All Fields] OR “polyp”[All Fields]) AND “interposition”[All Fields] AND (“colon”[MeSH Terms] OR “colon”[All Fields]). The PubMed similar articles function was used to extend potential hits. References of all eligible reports were screened for further information.
Study selection. Selection process. Three authors (SM, GF, PR) independently assessed titles and/or abstracts of all identified reports and excluded those considered irrelevant. Afterwards, full-text papers were screened on their accordance to inclusion criteria. One author (SM) extracted data from the included studies and a second (GF) checked extracted data. Disagreements were discussed and solved.
Data items. Data were extracted using the PICO-System (8, 9). The PICO tool focuses on the Population, Intervention, Comparison and Outcomes of scientific publications. It is used to identify components of clinical evidence within the literature. Additionally, the following items were extracted: first author, year of publication, study type, patient age, patient gender, underlying disease, intervention, main outcome measures, evidence of polyps and/or adenocarcinoma, interval between colon interposition and diagnosis of polyps/adenocarcinoma, evidence of colonic polyps/adenocarcinoma within preoperative colonoscopy, metastatic disease in case of carcinoma, therapy of polyps/adenocarcinoma prior to interposition, therapy of polyps/adenocarcinoma in interposition.
Risk of bias in individual studies. Risk of bias of non-randomized studies was assessed using the Newcastle-Ottawa Scale (NOS). NOS was designed as a quality assessment tool to evaluate both non-randomized cohort and case control studies. The quality of assessed studies was evaluated, using a star rating system. The overall maximum score was nine stars. The following domains were included: selection of study groups (maximum score: four stars), comparability of study groups (maximum score: two stars), ascertainment of exposures (for case control studies) or outcome of interest (for cohort studies) (maximum score: 3 stars) (10) (Table I).
Synthesis of results. A narrative synthesis of the results was performed since the number of cohort studies was too low for meta-analysis. Further, inclusion criteria as well as follow-up workflow and follow-up examinations were inhomogeneous.
Results
Study selection. Overall, 157 studies were screened for eligibility. Thereof, 72 records remained after exclusion of duplicates. After screening and full-text assessment, 37 records remained (case reports: n=31; non-randomized cohort studies: n=6). No randomized trials were identified according to the inclusion criteria. Six studies were excluded within the screening process and 29 more after full-text assessment (Figure 1, Table II).
Characteristics of cohort studies. Detailed study characteristics are depicted in Tables III and IV.
Study 1. In 1982, Akiyama and coworkers published data from a questionnaire-based national survey (9). Therein, 223 members of the Japanese society for Esophageal Diseases were asked to report on interval carcinoma within the substitute after esophagectomy. In 9.5% (n=617), colonic interposition was applied for intestinal reconstruction. In these cases, no carcinoma occurred. The authors did not indicate the time frame in which the abovementioned operations were performed. Moreover, it is not clearly described whether all patients were followed up consequently or if only symptomatic patients received further diagnostic work up (11).
Study 2. Eleftheriadis reported on a series of 13 patients who received colonic interposition after esophagectomy for benign disease from 1967 to 1983. Eight of them were available for endoscopic follow up, applied within an average time of 7.2 years after initial surgery. While endoscopic examination did not render relevant evidence of colitis, histologic samples of all but one patient showed microscopic changes: inflammation (n=6) and fibrosis (n=1). No cases of malignancy or pre-malignant adenoma were found (12).
Studies 3+4. Two studies by Isolauri et al. focused on both an isolated cohort of patients who received colon interposition for benign disease (Timeframe: 1964-1985) and a mixed cohort of patients with underlying malignant and benign pathologies (Timeframe: 1964-1985). According to the period of patient inclusion, the benign cohorts in both studies seem to be the same. Therefore, a certain overlap must be assumed within the results. Retrospectively, it is not possible to explain slight differences between both cohorts. All patients available for follow up underwent endoscopic examination of the interposition. Therein, the mucosa was evaluated macroscopically. Moreover, bacterial and fungal specimens were obtained. Overall, no graft-malignancies were found after a median follow up of 57 months, respectively 67 months in both studies. However, within the 1991 analysis, one carcinoma was found, classified as “recurrent carcinoma” and not as malignancy, originating from the interposition. Since no histopathologic analysis was available, this carcinoma could not be clearly assigned to the organ it originated from (13, 14).
PRISMA flow sheet.
Study 5. In 1996, Pompeo et al. published data on 100 consecutive patients having received esophagocoloplasty (13). Patients were divided into three groups: group A: congenital disease (n=22), group B: acquired benign lesions (n=36), group C: malignant disease (n=42). Out of the 42 patients belonging to group C only 5 survived more than 5 years. These patients were excluded from the analysis by the authors. Moreover, all patients with follow-up examinations of less than one year and two cases of late deaths were excluded (n=7). Thus, data from the 51 remaining patients were reviewed. A subdivision into pediatric (n=25) and adult (n=26) patients was made. The follow up rate was 100%. Due to inclusion criteria of the presented study, only information on adult patients was included into the analysis. Within the endoscopic follow up, no cases of interval carcinoma in the interposition were found. Since a differentiation between group A and B was not given in regard to benign interval lesions, those were not included in the review (15).
Newcastle-Ottawa Scale for assessing the quality of nonrandomized studies.
Excluded studies with reasons for exclusion.
Jeyasingham et al. (14) reported on two cases (0.6%) of adenocarcinoma at the cologastric junction and one case of adenocarcinoma of the interposed colon (0.3%) within a cohort of 365 patients who survived hospital stay after esophagectomy with colon interposition for benign reasons. The cohort consisted of pediatric and adult patients and was therefore formally excluded from the analysis, because allocation of respective cases to the adult or pediatric cohort was not completely comprehensible. Because the described interval carcinomas are the only ones reported in the context of a cohort study and not as single case reports or case series, we decided to report on the analysis even though it did not accurately fit our inclusion criteria (16).
PICOS question for cohort studies.
Overall, five retrospective cohort studies by four authors were included in the analysis, reporting on 1016 patients. Therein, no case of an interval carcinoma or premalignant adenoma was diagnosed in the interposed colon. Adding results of Jeyasingham to the analysis, interval carcinoma would have been detected in 0.22% of analyzed patients.
Risk of bias within studies. Per protocol, the risk of bias within retrospective studies was assessed, using the Newcastle-Ottawa scale (10). No study was rated with the highest possible score of 9 stars (Table I).
Characteristics of case reports. Overall, 31 case reports including 32 patients who developed benign or malignant interval pathologies within their colon interposition were identified (17-47). Baseline characteristics and findings of the specific lesions are depicted in Table V. Therein, reports on benign (n=7) and malignant (n=25) findings are presented separately. Median age was 63.5 years (range=18-80 years) in patients with interval carcinoma and 69 years (range=60-73 years) in cases of benign lesions of the interposition. Carcinomas as well as benign lesions of the interposition were diagnosed after a median of 8.5 years (7 months to 55 years in cases of malignant findings, 1 to 16 years in cases of benign findings). Information on endoscopic findings of the colon before being used as interposition were only available in one report, showing an adenomatous polyp. Underlying disease leading to esophagectomy was malignant in 42.9% of cases with benign interval lesion and in 52% of reports on interval carcinoma.
Discussion
Esophageal replacement using the colon is one of the most common techniques of reconstruction after esophagectomy for both benign and malignant diseases. In general, the colon is proven to be a long lasting and well fit substitute, showing excellent postoperative function as well as alimentary satisfaction and quality of life (3, 48). However, apart from functional issues, mucosal pathologies of the interposition have been described but are yet to be systematically assessed. Therefore, we performed the presented systematic review.
Characteristics of cohort studies.
According to data extracted from the included cohort studies, the risk of interval lesions was found to be extremely low. With inclusion criteria applied strictly, a rate of 0% can be calculated from a cohort of 1016 patients. Beyond this, 7 case reports present data on benign interval lesions and 25 report on interval carcinoma in the esophageal substitute. In this context, a certain risk of mucosal changes must be estimated. Characteristics given by the analyzed case reports are all limited to the index case. Information on the total cohort of esophagectomies with consecutive interposition of the colon from the respective institution is lacking. Since the overall number of operations is not known, calculation of the occurrence risk of metachronous lesions is not possible. According to the presented data, the development of mucosal changes takes several years, showing a median of 8.5 years after index operation. A potential rearrangement of the mucosal microbiota following the positional change of the colon, with respective alteration of mucosal resistance may be assumed. Yet, no experimental studies report on this thesis. However, the interposed colon is exposed to an altered dietary composition as well as to potential gastric and biliary reflux in the distal part of the substitute. These aspects potentially facilitate the additional development of mucosal changes (14, 49). Improved diagnostic techniques as well as the improvement of perioperative (neoadjuvant and adjuvant) therapies lead to earlier detection and a significantly higher rate of long-term survival in patients suffering from esophageal cancer (37). Moreover, patients operated for benign disease have no relevant limitations of their life expectancy in case of survival of the acute postoperative phase and adequate anastomotic healing. Consecutively, a theoretical risk of mucosal interval pathologies exists in both groups. Moreover, evidence was found, that patients, suffering from esophageal cancer are at higher risk for colonic lesions such as adenoma and carcinoma (4, 50). In this context, the need for a structured endoscopic follow up must be highlighted. According to current international guidelines, follow-up colonoscopy for colorectal cancer screening is recommended 10 years after the first screening colonoscopy without pathologic findings (51). The risk of interval pathologies after colonic interposition, derived from included cohort studies, is extremely low. However, following the authors' personal opinion, follow up endoscopy should be applied within a closer interval, such as five years, even if preoperative colonoscopy was without pathologic findings and no risk factors were identified, as suggested by Imperiale and coworkers (52). Moreover, it has been previously shown before, that patients with adenocarcinoma of the esophagus have a significantly higher risk for the occurrence of colonic polyps (4). Because interval lesions are typical late complications and the risk increases over a longer time after index surgery, a life-long endoscopic follow-up must be postulated. In addition, the individual risk can be assessed by anamnestic factors such as colitis, history of colonic polyps, positive family history for colonic carcinoma (43). Positive information should lead to a closer follow-up interval. In addition, it is of significant importance to perform preoperative colonoscopy before using the colon as esophageal substitute.
Case reports.
Limitations
The presented systematic review is biased by the lack of randomized and structured prospective observational trials. None of the included cohort studies was rated with the full score of nine stars according to the Newcastle-Ottawa scale. This study is thereby at risk of being biased by the relatively low quality of these studies (e.g. lack of controls, short period of structured follow up, symptom-based follow-up). Data, extracted from case reports are all limited to the index case and give no information on the whole cohort. In this context, no risk calculation was possible.
Authors' Recommendations
The risk for pre-malignant and malignant interval pathologies within the colon interposition is extremely low. Therefore, the authors of the present review recommend:
Preoperative assessment of the risk of development of colonic neoplasia;
Standard preoperative colonoscopy;
Lifelong endoscopic assessment of the interposed colon within an interval of five years;
Adjustment of follow-up endoscopy to the individual's risk, preoperative endoscopic findings within the colon and potential interval pathologies.
Conclusion
In conclusion, esophageal substitute by parts of the colon is associated with a low but important risk of mucosal interval pathologies such as adenoma and carcinoma. This may be caused by the positional change of the colon as well as by individual risk factors. Therefore, preoperative endoscopic assessment of the colon must be strictly claimed, as well as a structured and life-long endoscopic follow up of the interposition. To date, no randomized trials or structured prospective observational studies are available on this topic. Overall, the available evidence is relevantly low, wherefore further investigations are of significant importance.
Footnotes
Authors' Contributions
MS and PR designed the review, wrote the manuscript and analyzed results of literature research, AA revised the review and screened full text articles, ST did literature research and selected studies as well as data items, CM did literature research and selected studies as well as data items, FG did literature research and selected studies as well as data items, AH did native revision of the manuscript and designed the review together with MS and PR, JP revised the design and the manuscript, FA reviewed full text articles and selected data items. All Authors finally revised and approved the manuscript and agreed to be accountable for all aspects of the review.
This article is freely accessible online.
Conflicts of Interest
The Authors have no conflicts of interest to declare regarding this study.
- Received October 18, 2019.
- Revision received November 5, 2019.
- Accepted November 6, 2019.
- Copyright© 2019, International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved
References
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.