Page 7 - JCTR-9-5

Songtanin et al. | Journal of Clinical and Translational Research 2023; 9(5): 308-316 309
alter the balance that controls immune tolerance from gut antigens cancer vs. rectal cancer). The proximal colon consists of cecum,
and gut inflammation, a process considered to be a mechanism for ascending colon, hepatic flexure, and transverse colon whereas
the development of colorectal cancer [7]. Appendectomy may be the distal colon consists of splenic flexure, descending colon, and
a potential risk factor in developing colorectal cancer; however, sigmoid colon.
differences in study design, such as the definition of the interval The consistency of the findings of all meta-analyses was
between appendectomy and the initial diagnosis of colorectal further examined using leave-one-out sensitivity analyses. The
cancer, follow-up period, and location of colorectal cancer, need likelihood of publication bias was explored using funnel plots,
consideration. Therefore, we performed a systematic review and and the effect sizes of missing (i.e., unpublished/unreported)
meta-analysis to evaluate the association between appendectomy studies were imputed through the trim-and-fill method [11,12].
and the development of colorectal cancer. Funnel plot asymmetry was confirmed using Egger’s tests [13].
Heterogeneity of effect sizes was quantified by calculating the
2. Methods
Higgins’ I statistic [14,15]. Meta-regression analyses were not
2
We conducted this meta-analysis according to Cochrane’s attempted due to the small number of studies included in each
manual of diagnostic test accuracy, and the manuscript was meta-analysis [15,16].
prepared according to the preferred reporting items for systematic 3. Results
reviews and meta-analysis of diagnostic test accuracy (PRISMA-
DTA) guidelines [8,9]. Figure 1 provides a graphical representation of the study
A literature search of MEDLINE and EMBASE databases was screening and selection process. A total of 919 articles were found
conducted from their inception through September 2022. Search using the above search criteria. After removal of duplicates, 541
terms included (a) appendectomy, (b) colorectal cancer, colon articles were evaluated. Of these, 23 articles were potentially
cancer, and rectal cancer. Only studies evaluating adult population relevant to the study goals. Thirteen studies were removed
were included in the study. Case series, case reports, and non- because the study was a non-English publication (n = 3), used the
English publications, and studies with appendiceal carcinoma and same study cohort (n = 1), was a review article (n = 1) or a letter
mucinous appendiceal neoplasm were excluded from the study. to the editor (n = 1), or had missing data (n = 7). Therefore, 10
The titles and abstracts were reviewed by two independent authors studies were included in this meta-analysis [4,5,17-24]. Table 1
(B.S. and N.T.). Discrepancies were resolved through discussion reports the characteristics of each study. Of these studies, one was
between the two independent authors and the senior author abstract, and nine were full articles. A total of 39,711 colorectal
(K.N). Two independent authors compiled data from each study, cancers and a total of appendectomies 384,278 were analyzed.
including study characteristics, study population characteristics, Two studies did not classify patients based on the specific location
and study results (B.S. and N.T.). Study characteristics included but reported results just as colorectal cancer [5,22]; eight studies
author, year of publication, start and end dates for data collection, classified patients based on a specific location of colorectal
country, and type of study design. Study population characteristics cancer [4,17-21,23,24]. Table S1 demonstrates an association
included number of patients, age, gender, number of patients who
underwent appendectomy, number of control patients (those
without appendectomy), number of colorectal cancers, location
of colon cancers (classified as proximal colon which includes
cecum, ascending colon, hepatic flexure, and transverse colon
and distal colon which includes splenic flexure, descending colon,
and sigmoid colon) and rectal cancer if data were available),
and number of patients without colorectal cancers. The quality
of each study was independently evaluated by each investigator
using Newcastle-Ottawa quality assessment scale [10]. Any
discrepancies were resolved through discussion between the two
independent authors with the senior author (K.N.).
2.1. Statistical analysis

A DerSimonian-Liard random-effects meta-analysis was
performed using the “meta” package (version 5.0-1) in R
statistical software (version 4.2.2) to examine the pooled odds
ratio representing the associations between colorectal cancer
and the history of appendectomy. Additional random-effects
meta-analyses were performed to pool the odds ratios of studies
examining the associations between appendectomy based on the
anatomical site of colorectal cancer (i.e., proximal vs. distal colon Figure 1. A visual representation of the search strategy.
DOI: http://dx.doi.org/10.18053/jctres.09.202305.23-00090