Abstract
Background/Aim: Gastric cancer is the fifth most frequently diagnosed cancer and the second most common cause of cancer-related death. The only potentially curative treatment is surgical resection, which is associated with potentially severe complications, such as anastomotic leakage. The aim of this systematic review was to evaluate the relationship between anastomotic leakage and overall and disease-free survival after surgery for gastric cancer. Materials and Methods: A systematic literature search was performed and 7 articles published between 2010 and 2019 were included, including a total of 7,167 patients. Results: Among the included studies the frequency of anastomotic leakage ranged from 6 to 41%. Patients affected by anastomotic leakage had an overall survival ranging between 4.1 and 97.6 months, whereas patients who did not experience anastomotic leakage had an overall survival between 23 and 109.5 months. Conclusion: Closer follow-up or even more aggressive oncological therapy may be considered for patients affected by anastomotic leakage after surgery for gastric cancer.
Gastric cancer is the fifth most frequently diagnosed cancer and the second most common cause of cancer-related death (1-2). In the world, annually, more than one million diagnoses of gastric cancer are made and the incidence of gastric cancer is highly dependent on region and culture (2). The cumulative risk of developing gastric cancer from birth to age 74 is higher in males, amounting for 1.87% against a cumulative risk of 0.79% in females worldwide.
The management of patients with gastric cancer is interdisciplinary, including surgical oncology, medical oncology, gastroenterology, radiation oncology, radiology, and pathology. Nowadays, surgery remains the primary treatment option for patients with resectable tumors (2, 3).
For patients with cT2 or even more advanced gastric tumors, perioperative chemotherapy and preoperative chemoradiation (category 2B) should always be considered, even if surgical resection with negative margins (R0) remains the most important step of the treatment (3).
However, gastric cancer surgery may be associated with several complications. The overall complication rate after radical gastrectomy for gastric cancer is 18.7%. The most common complication is anastomotic leakage (AL), which affects up to 3.4% of patients (3, 4).
The long-term consequences of AL following gastric cancer surgery are still debating. The aim of our work was to analyse the impact of AL on disease-free-survival (DFS) and overall survival (OS) in patients who underwent gastrectomy for cancer, through a careful analysis of the results currently available in the literature.
Patients and Methods
Search strategies. We conducted a systematic search in PubMed, Embase, Cochrane Library, CILEA Archive, BMJ Clinical Evidence and Up ToDate databases with the following search terms: “cancer” AND “gastric” (or “stomach”) AND “anastomotic leakage” (or fistula or dehiscence) AND “overall survival “ AND “disease-free survival ” AND “recurrence” AND “long-term-results”.
We also scanned all the reference lists to identify other relevant studies. Unpublished reports were not considered eligible.
Compliance with ethical standards. All procedures performed in the studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.
Study selection. Inclusion criteria were: 1) articles written in English and published between January 1998 and June 2019. 2) Open or minimally invasive surgery with curative intent (with histological confirmation of R0 status). 3) Cases of gastric adenocarcinoma and/or gastroesophageal adenocarcinoma type III according to Siewert classification (5). 4) Studies mentioning overall survival and/or disease-free survival.
Exclusion criteria were: 1) treatment for disease recurrence; 2) case report; 3) letter to the Editor; 4) Review.
Data extraction and quality assessment. Two authors (GM, MC) independently screened the articles by title, abstract and keywords, and then selected and analysed the relevant articles. Any disagreement was resolved by discussion with the senior author (PA).
PRISMA statement guidelines for conducting and reporting systematic reviews were followed (6).
The research protocol was registered at the International Prospective Register of Systematic Reviews with the following registration number: CRD42019120781 (6). The Newcastle-Ottawa Scale (NOS) was used for quality assessment (7). Cochrane Collaboration guidelines were followed (8). Three broad perspectives were used to judge a study evaluating the selection of the study groups, comparability, ascertainment of either the exposure or outcome of interest for case-control or cohort studies, respectively (7, 8). Results of quality assessment scale (7) are reported in Table I. The study was designed and reported according to the STROCSS criteria (9).
Results
The literature search yielded 135,194 items; 135,001 were not considered suitable for abstract or integral reading, including duplicates removal. Then, 193 titles and abstracts were reviewed (Figure 1). A total of 180 papers were excluded because they did not match our inclusion criteria. Finally, only 13 articles were selected for full-text review and of these, six more were excluded for the following reasons: Five because they did not report OS or DFS rates in patients with occurrence of anastomotic leakage and one because it did not report AL rates. There was no disagreement regarding eligibility of full-text articles.
A total of seven articles fulfilled the selection criteria and were included in this systematic review (10-16). This pool of articles consisted of seven retrospective reports. Characteristics of the included studies are presented in Table I. Table II shows the outcomes.
A total of 7,167 adult patients who underwent surgery for gastric and GEJ cancer (type III according to Siewert Classification) were included.
Tsujimoto et al. (10) have evaluated the influence of postoperative infection on long-term survival after potentially curative gastrectomy in 1,332 patients. Two groups of patients were considered: 141 patients composed the first group, who had postoperative complications (10.6%). The second group, without postoperative complications, included the other 1,191 patients. The OS of patients with stage II gastric cancer was 65.6% (no postoperative complications) versus 49.7% if they experienced postoperative complications (p=0.04). OS was 41.2% for patients with stage III gastric cancer but if they experienced postoperative complications it dropped to 20.6%, (p=0.002). Multivariate analysis demonstrated that AL was significantly associated with cancer-specific survival (10).
Sierzega et al. (11) included 699 patients of whom 41 were affected by AL (5.8%). The group with AL had 5-years DFS of 11 months. The group without complications had a 5-year DFS of 19 months (p=0.021). The OS was 4.1 months and 23 months, respectively (p=0.001).
Han Mo Yoo et al. (12) included 478 patients. AL was found in 35 patients (6.7%). In the group of patients affected by leakage, the overall mean survival was significantly lower: 30.5 versus 96.2 months (p<0.001). AL was an independent factor predictive for overall survival (HR=3.58, 95%CI=2.29-5.59).
Kim et al. (13) included a total of 3,827 patients. They reported an AL rate of 1.88%. They did not report the OS, and they showed a median DFS of 97.6 months in patients with AL and a median DFS of 109.5 months in patients, who did not experience AL (p=0.076).
Samples et al. (14), collected data of 102 patients. Six ALs occurred (5.8%). The 5-year DFS was not reported.
Andreou et al. (15) collected data from 471 patients. 41 patients were affected by AL (8.7%), with a 5-year overall survival rate of 35% versus not AL of 58% (p=0.005). OS in patients with AL was 39% versus 61% in those without AL (p=0.005).
Barchi et al. (16) enrolled 258 patients. They had 15 patients with AL (5.8%). In this manuscript the 5-year DFS was not reported. The OS was 50 months in patients with AL versus 100 months in patients without AL.
Discussion
The aim of our work was to analyse the impact of AL on DFS and OS after gastric cancer surgery. The relationship between AL and long-term results after surgery for gastric cancer is still the subject of debate. Our group has recently published a review regarding the relationship between AL and oesophageal cancer, showing how, with a proper literature analysis, is it possible to highlight the relationship between AL and survival (17). A similar investigation was performed on colorectal surgery, concluding that AL after proctectomy for cancer is associated with worse long-term DFS and overall survival (18).
Beecher et al. (19) have demonstrated in their review how surgical trauma, inflammatory response, intraoperative tumor manipulation, growth factors released during the procedure and tumor growth have a significant impact on OS and recurrence rates. Surgery can indeed bring to suppression cell-mediated immunity and diffusion of malignant cells into the bloodstream, therefore, increasing the metastatic potential. In our research, five articles showed that AL had a statistically significant impact on the OS and DFS (10-12, 15, 16), and two articles written by Kim et al. and Samples et al. did not find any statistically significant relationship (13, 14).
The studies showing that AL had a statistically significant impact on OS and DFS included 3,238 patients, whereas the study that did not found any correlation between AL and long-term results (13, 14) included a total of 3,929 patients. In absolute numbers, articles showing no correlation between AL and worse long-term outcomes have a higher number of enrolled patients. However, the studies showing a significant correlation between AL and survival included more patients with AL: 273 patients had AL (8.4%, range=5.8-10.6%) in this group versus 78 patients (1.9%, range=1.88-5.8%) in the articles not showing any significant relationship. The smaller number of patients with AL may limit the statistical power of studies not showing any correlation.
Indeed, a global analysis of the data shows that there is a negative trend in long-term outcomes (worse DFS and worse OS) in patients with AL.
Tsujimoto et al. (10), did not report OS as months of survival, but in the multivariate analysis they demonstrated the negative impact of AL on OS (p=0.0015). Moreover, they analysed the correlation between AL and cancer-specific survival finding a statistically significant correlation (p=0.0358).
Sierzega et al. (11) demonstrated how AL has negative effects on OS and DFS. They showed that patients with AL had an OS of 4.1 months versus 23 months for patients who did not have AL (p=0.001). Regarding DFS, they have shown that patients with AL had worse overall survival: 11 months compared to 19 months in patients without AL (p=0.021). In this study, the factors predisposing to AL were also analysed; they demonstrated that poor performance status had the greatest impact as a predictive factor of AL.
Yoo et al. (12) demonstrated that AL has a negative impact on OS and DFS, in a multivariate analysis (p<0.001). They also identified AL as one of the causes of cancer-related death (32.2 vs. 99.5 months; p<0.001), and recognized poor performance status as the most important factor predisposing to AL.
Andreau et al. (15) showed that AL is a significantly predictor of worse OS (p=0.037) and DFS (p=0.004). In their analysis, AL turned out to be an unfavourable factor independently from tumor stage and biology, and in the multivariate analysis for factors associated with tumor recurrence, AL was an independent predictor of tumor recurrence (p=0.002).
Barchi et al. (16) demonstrated that OS is statistically worse in patients with AL than in non-AL patients. They reported 15 fistulas in 258 patients operated for gastric cancer, with decreased OS in patients with AL compared to non-AL (50 months and 100 months, respectively). Barchi et al. recognized the type of surgery as a major cause of AL (total vs. completion gastrectomy).
Kim et al. (13) analysed a cohort of 3827 patients. They had 72 fistulas, 1.88% of the total number of included patients. They showed that patients with leakage and without leakage did not differ significantly in terms of DFS (p=0.076). Their statistics also showed that the AL was not associated with cancer-related death.
Samples et al. (14), in their analysis, showed that there is no correlation between AL and worse OS.
During the selection of the articles, some were excluded because they did not strictly respect the inclusion criteria. They did not specifically investigate the relationship between AL and worse long-term survival, but they analysed the negative impact of AL on general complications, pathological stage, surgical technique, inflammation, and the long-term survival (20-27). The impact of AL on the prognosis of several cancers is still debated (28-31).
All the included studies were retrospective, carrying an intrinsic risk of selection bias, including heterogeneity linked to the changes in clinical practice, technology, surgical and chemoradiotherapy protocols. Indeed, changes in adjuvant chemotherapy practice may have included some selection bias, with patients treated in the most recent years showing better oncological outcomes due to more effective protocols. The definition of AL, reported in 5 out of 7 articles (11-16), appeared to be different among the studies, because it was based on different clinical parameters. In all the articles, AL, when recognized, was investigated radiologically. The different classification used for the gravity of AL evaluation could be a potential source of bias. Only 2 articles reported the Clavien-Dindo classification: Barchi et al. (16) reported 3 AL of grade I-II and 12 AL of grade III-V; Samples et al. (14) reported 6 ALs of grade III. Moreover, none of the articles stratified the risk based on tumor staging except for Barchi et al. (16).
Conclusion
The majority of reviewed studies showed an association between AL and DFS and OS. Closer follow up or an even more aggressive oncologic therapy may be considered for patients affected by postoperative anastomotic leakage.
Footnotes
Authors' Contributions
Paolo Aurello gave substantial contribution in the design of the work, drafting the work, final approval of the version to be published. Matteo Cinquepalmi, Niccolò Petrucciani and Giovanni Moschetta gave substantial contribution in the acquisition, analysis and interpretation of data for the work, drafting the work and final approval of the version to be published. Laura Antolino, Federica Felli, Diego Giulitti, Giuseppe Nigri, Francesco D'Angelo, Stefano Valabrega and Giovanni Ramacciato gave substantial contribution in critically revising the article for important intellectual content and final approval of the version to be published. All Authors agreed to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.
This article is freely accessible online.
Conflicts of Interest
None of the Authors had any conflicts of interest to declare.
- Received January 9, 2020.
- Revision received January 17, 2020.
- Accepted January 20, 2020.
- Copyright© 2020, International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved