Department of General and Digestive Surgery, Ramón y Cajal University Hospital, Madrid, Spain
Department of Clinical Biostatistics, Ramón y Cajal University Hospital, Madrid, Spain. IRYCIS. CIBERESP.
Department of Nursey, Alcalá University, Spain
Department of General and Digestive Surgery, Sanitas La Zarzuela Hospital, Madrid, Spain
Published Date: 04/08/2021.
*Corresponding author: Araceli Ballestero Pérez, Department of General and Digestive Surgery, Ramón y Cajal University Hospital, M-607 Km 9, Madrid CP 28034 Spain
Orcid: 0000-0002-4897-2154
Introduction: The aim of our study was to evaluate recurrence and long-term survival in patients with obstructive left colon cancer treated with colonic stent and elective surgery versus emergency surgery.
Methods: A retrospective cohort study, where all the patients with left-sided malignant colonic obstruction admitted in the Emergency Department between june 2006 and january 2014, were included. They were treated either with colonic stent as a brigde to elective surgery or with emergency surgery. The observation period was until september 2018. The main study variables were overall survival, local recurrence and distance recurrence. As secondary endpoints were the rate of anastomoses and stomas and the morbi-mortality associated.
Results: 53 patients were included in stent group and 40 in surgery group. Local recurrence was only presented in 2 patients of the surgery group. Distance recurrence was more frequent in the surgery group (34.2% versus 17.8%), although without significant differences, multiplying the risk of recurrence by 2.22 times (SHR 2.22 CI 95 % 0.92-5.30 p 0.07). Overall survival (OS) did not show significant differences. Significant differences were found in terms of the highest rate of anastomosis in the stent group (75% versus 52.5% p 0.02) and a lower proportion of stoma (27% versus 47.5% p 0.04). Postoperative morbi-mortality were greater in the stent group.
Discussion/Conclusion: The long-term results showed no significant differences between the two groups. However, there was a significant improvement in the short-term results in the stent group, in terms of the rate of anastomosis and stoma.
Keywords: Colonic Neoplasms; Elective Surgical Procedures; Intestinal Obstruction; Neoplasm Recurrence; Stents
Up to 30% of colorectal cancers may debut with an intestinal obstruction [1]. 70% of the cases are located in the left colon [2]. The classic treatment for this problem has been the performance of an emergency surgery (ES) with a 40-50% of morbidiy and a 15-20% of mortality [3]. The intratumoral stent was described by Dohmoto in 1991 as palliative treatment [4] and in 1994 Tejero published the first experience as a bridge to surgery (BTS) [5]. Since then, several studies have been published comparing the use of self-expandable metallic stent (SEMS) as BTS versus ES in obstructive left-sided colon cáncer (OLCC). It has been demonstrated good results improving the rate of anastomoses and decreasing the rate of stoma, without significant differences in morbidity and mortality [6]. In 2013, long-term studies appeared, showing an increase in tumor recurrence [2,7,8], with no difference in overall survival (OS) [2,7,9]. The aim of our study was to evaluate the treatment with SEMS as BTS versus ES in patients with OLCC in terms of long-term survival and recurrence results. And as secondary endpoints were the short-term results of each treatment as well as the associated morbidity and mortality.
Study design and partcipants
This is a retrospective cohort study approved by the Ethics Committee of the XXXXX). The patients went to the XXXX Emergency Service between June 2006 and January 2014, with an intestinal obstruction due to left colon cancer.
Inclusion criteria
Patients with an intestinal obstruction caused by a primary tumor of the left colon, sigma and at the rectum-sigmoid junction with potentially curative disease were included.
Exclusion criteria
Patients diagnosed with intestinal obstruction due to other processes or patients diagnosed with intestinal perforation were excluded. Also, we excluded patients treated with palliative intention.
Treatment description
Patients according to the treatment performed were included in the SEMS group or in the Surgery group.
SEMS Group: formed by those patients in whom a SEMS was placed as a BTS. Wallflex 25x90 mm or Wallflex 25x120 mm uncovered SEMS was used (Boston Scientific Co., Natick, MA). The correct position of the SEMS was verified by fluoroscopy and an evacuating solution was administered. When the intestinal obstruction was resolved, elective surgery was performed by laparotomy. Patients who did not present technical/ clinical success or those who presented any endoscopic complication, ES was performed.
Surgery Group: formed by those patients in whom intestinal obstruction was resolved by ES through laparotomy.
Follow up and data recording: The study observation period was from June 2006 to September 2018, with a minimum follow-up of 4 years.
Intention to treat analysis was performed. Quantitative variables were expressed as means ± standard deviation (SD) or median and interquartile range (IQR) depending on whether the variable followed a normal distribution. Qualitative variables were expressed as absolute and relative frequencies. The student’s t-test or the Mann-Whitney test were used to compare the quantitative variables between groups. Chi-square test or Fisher's test were used to compare qualitative variables. For the variables time to death and time to recurrence, the Kaplan-Meier method and a regression model of competitive events were used. To evaluate the possible confounding effect, a Cox multivariable regression model was proposed. To assess the causal effect of the stent and to correct it for confounding variables, a propensity score (PS) with those variables potentially associated with the main events (age, American Society of Anesthesiologists Classification (ASA Class), tumor location and white blood cell count at admission) was employed. A matching has been performed with a maximum caliper of 0.2 SD of the PS logit. The comparability assessment has been carried out by calculating the standardized differences for each variable included in PS score; the model has been considered adequate if the standardized differences are less than 10%. Once paired, the effect of the treatment on the paired samples was evaluated using a GEE model. The statistical package STATA 16.1 has been used for statistical analysis. A p value of less than 0.05 was considered as statistical significance.
Table 1: Demographic, diagnostic and treatment characteristics of the sample.
Table 2: Univariate analysis of postoperative mortality in SEMS Group.
Table 3: Oncological and evolutionary characteristics of the sample.
Table 4: Local Recurrence.
Between June 2006 and January 2014, 147 SEMS were placed endoscopically and 54 patients were operated. Finally, the SEMS Group consisted of 53 patients and the Surgery Group were 40 patients as shown in Figure 1.
The demographic variables are described in Table 1. In the univariable analysis, no statistically significant differences were observed between groups in terms of age, sex, and ASA Class. A higher rate of pneumatosis was observed in the Surgery Group (20% vs. 5% p 0.04). Regarding tumor location, the most frequent place was the sigma, showing significant differences between groups (p 0.00). In the SEMS Group, the tumor was more frequent located in the sigma and descending colon, while in the Surgery Group a 40% of the tumors were located in the splenic angle and 7.5% in the distal transverse colon.
Endoscopic variables
The endoscopic variables of the SEMS Group are showed in Table 1. Technical success was 96% (51/53) and clinical success was 87% (46/53). There were 4 endoscopic complications (7.5%). There were 3 intestinal perforations (5.7%) and 1 rectal bleeding (1.9%).
Surgical variables
Surgical procedure was performed in 92/93 patients (one patient in the SEMS Group rejected the surgical treatment). ES was performed in 100% of the Surgery Group and in 15% of the SEMS Group (8) p 0.00. Primary anastomosis was performed in 75% of the SEMS Group (39) versus 52.5% of the Surgery Group (21) p 0.02. A stoma was required in 27% (14) of the SEMS Group and 47.5% (19) of the Surgery Group p 0.04. During follow-up, surgical reconstruction could be achived in 2 patients in the SEMS Group (14.3%) and 7 in the Surgery Group (36.7%) p 0.24. Regarding surgical complications following the Clavien-Dindo classification, a higher rate of them, was observed in the SEMS Group (42% vs. 27.5% p 0.52) without significant differences. Analyzing the type of complications, the anastomosis leak was similar in both groups (5.8% SEMS Group vs. 5% Surgery Group). A secondary surgery was required in 8 patients, 6 in the SEMS Group (12%) and 2 in the Surgery Group (5%) without significant differences (p 0.46). Postoperative mortality was higher in the SEMS Group (17% vs. 10% p 0.38). A logistic regression model was used and it was observed that the SEMS Group multiplied the risk of dying during the postoperative period by 1.85 times (OR 1.85 95% CI 0.52-6.67 p 0.34). In the SEMS Group, a higher postoperative mortality was observed in patients with a higher mean age, in those with a lower hemoglobin value and a greater leukocytosis on admission at the emergency department and in those with a greater alteration in the preoperative INR value as it shown in Table 2. A higher postoperative mortality was also observed in the group of patients in whom primary anastomosis could not be performed and in those who presented some type of postoperative complication. The overall lenght of hospital stay was similar in the two groups, 13 days (IQR 11-19) for the SEMS Group versus 11 days (IQR 14.5-17.5) for the Surgery Group (p 0,95). The median follow-up time was 36.7 months (IQR 8-84.7) without significant differences between groups.
Oncological variables
10.8% (10) of the patients were diagnosed intraoperatively with metastatic disease without differences between both groups. The treatment is described in Table 3. Staging according to the TNM classification for colorectal cáncer 8th edition did not show statistically significant differences as it shown in Table 3. Lymphadenectomy did not show significant differences between groups, although the median number of nodes resected was greater in the SEMS Group (17 (RIC 12-23) than in 14 (RIC 9-18)). 51.6% of the sample received adjuvant treatment, without differences between groups.
Only 2 patients in the Surgery Group (5%) had local recurrence at mean of 17.6 months ± 2.7. The characteristics of these patients are shown in Table 4. Distant recurrence was lower in the SEMS Group (17.8% vs 34.2% p 0.09). When the competitive risks model was performed, it was observed that the surgery group multiplied by 2.22 times the risk of distant recurrence (SHR 2.22 95% CI 0.92-5.30 p 0.07). With the couples selected according to the PS, the risk of distant recurrence was 3.48 times greater in the Surgery Group, with significant differences (SHR 3.48 95% CI 1.01-12.04 p 0.04). The OS according to the Kaplan Meier curve as shown in Figure 2, was lower in the SEMS Group (p 0.36), being at 5 years 60.4% in the SEMS group versus 68.5% in the surgery group. The disease-free survival (DFS), however, was lower in the Surgery Group (p 0.07), being at 5 years 58.5% Surgery Group versus 72.9% SEMS Group as it shown in Figure 2. Death by any cause was evaluated by performing a Cox regression study and it was estimated that there was a 1.43-fold increased risk in the SEMS Group (HR 1.43 95% CI 0.67-3.03 p 0, 36). When this value was corrected for the variables that could affect this result (age, tumor location, ASA Class and white blood cell count preoperatively), the risk of death by any cause remained higher in the SEMS Group (HR 1.31 95% CI 0.58-3.03 p 0.51). No effect was observed with the PS-matched data. Mortality cancer specific, was higher in the SEMS Group in the first 4 years, becoming similar after the 5th year without statistical differences.
The aim of our study was to analyze the long-term results in terms of survival, local and distant recurrence in patients treated with SEMS as BTS versus ES. The OS in our study did not show significant differences between the 2 groups, as shown other studies [8-14], although in our study was lower in the SEMS Gruoup. In the randomized multicenter study published in 2017, Arezzo et al. [11] showed a comparable OS and DFS in both groups as also it is shown in a meta-analysis by Zhang et al. [15], where they included 8 clinical trials that reported that the SEMS did not affect the OS. In another meta-analysis published by Amelung et al in 2018 of 21 articles, it referred that the use of SEMS as a BTS could not have an influence on the long-term oncological results without differences in OS at 3 (69.7% versus 67.9% OR 0.85 95% CI 0.68-1.08) and at 5 years (63.5% versus 57.9% OR 1.04 95% CI 0.68-1.57) [16]. Other authors such as Sloothaak et al reported worse OS in the SEMS Group [7]. Like Kim et al, whom described a worse OS at 5 years in the SEMS Group [17]. Regarding DFS, in our study there were no statistically significant differences, being greater in the SEMS Group (72.9% versus 58.5%). In 3 trials in which the long-term results were studied, they showed a tendency to decrease DFS in the SEMS Group [7,10]. However, several articles suggested that there are no differences. Matsuda et al, in their meta-analysis where 432 patients in the SEMS Group versus 704 patients in the Surgery Group were compared, showed differences, although they were not significant. The 5-year DFS was 48.4% and 59%, respectively, with a relative risk of 1.06 (p 0.43) [18]. According to some authors, perforation caused by the stent could result in a spread of tumor cells and it could affect in the oncological results [17,19]. Maruthachalam et al measured CK20 mRNA levels in peripheral blood after stent insertion and they were higher in these patients [19]. Increased interstitial pressure in the tumor mass could cause embolization into the lymphatic vessels [20]. In our study, there were only 2 cases of local recurrence in the Surgery Group, so we cannot draw conclusions about it. In a systematic review and meta-analysis of 8 clinical trials, they showed that tumor recurrence was higher in the SEMS Group (40.5% versus 26.6%) [21]. Alcantara et al reported a high rate of recurrence in the SEMS Group 53.3% (8/15 patients) versus 15.4% (2/13 patients) [22]. However, we observed that in our study, distant recurrence was higher in the Surgery Group (34.2% versus 18% p 0.09) and it presented statistical significance when we applied the PS in the 24 selected couples. It is true that the patients in the SEMS Group had a higher rate of postoperative mortality and death by other causes than in Surgery Group (28.9% versus 13.2%), so there were less possibilities of recurrence.
Regarding secondary endpoints, a technical success of the endoscopic procedure was of 96% and a clinical success of 87% were observed, similar to the prospective multicenter study of 513 patients by Matsuzawa et al, where it showed a technical and clinical success up to 95% [23]. However, at the beginning, other series such as Pirlet et al prematurely closed their prospective randomized study due to the low technical success rate (47%) with 2 associated perforations [24]. A risk factor for these poor technical / clinical results is the placement of the stent in angled areas [25] as the splenic angle is considered. None of the patients in the study who were treated with SEMS at the splenic angle presented technical/clinical failure, and only one of them presented a microscopic intestinal perforation identified in the pathology result. There were 7.5% of endoscopic complications derived from the stent (4/53). Up to 20% the endoscopic complications have been described in the literature [26]. Intestinal perforation is considered the worst of endoscopic complications and it is described around 7.6% according to the European Society of Gastrointestinal Endoscopy (ESGE) [27] with a mortality of 20-30% [28].
In our study, the rate of clinical perforation was 5.7% (3/53), with an associated mortality of 33% (1/3). There were 10 patients, who showed microperforation in the pathology piece. In total 25.5% presented micro or macroscopic intestinal perforation, Pirlet et al described 27% of perforations in the anatomo-pathological results [24]. The median time interval between the endoscopic procedure and elective surgery was 4 days (RIC 3-5), being less than the recommended by the ESGE (5-10 days) [27]. It could have caused worse postoperative results. Eight of the patients in the SEMS Group (15.4%) required ES due to technical/clinical failure or to the presence of some complication. These results are similar than the ones described in a retrospective study of 253 patients, in which 12% (22/182) of the SEMS Group required ES [29]. Primary anastomosis could be performed in 75% of the SEMS Group versus 52.5% of the Surgery Group (p 0.02), as was demonstrated in the prospective randomized studies [8,10,22-24] and reviews with meta-analyzes [15,30] performed, where there is a greater number of primary anastomoses and significantly lower rate of stoma. Arezzo et al. published in their systematic review and meta-analysis of 8 clinical trials [21], a 70% proportion of primary anastomosis in the SEMS Group versus 54.1% in the Surgery Group. The incidence of stoma was significantly lower in the SEMS Group (27% versus 47.5%) as described by this same author in his 2017 meta-analysis, which included 497 patients from 8 prospective randomized studies, in which he reported a lower rate of stoma in the SEMS Group (33.9% versus 51.4% RR 0.67 p <0.001) [21]. The rate of permanent stoma was 23.1% (12/52) in the SEMS Group versus 30% in the Surgery Group (12/40) without significant differences (p 0.45), as it occurred in other study [31], in which it was higher in the Surgery Group but without significant differences. Postoperative short-term evolution has been discussed in the literature. When we analyzed our sample, we were surprised that the SEMS Group, like other authors [24,32-33], presented a higher rate of complications, although not significantly (42% versus 27.5%). Pirlet et al and Van Hooft et al closed their clinical trials prematurely due to an increase in the number of complications in the SEMS Group [24,32]. Min et al also published a higher proportion of complications in the SEMS Group that reached statistical significance (31.4% versus 8.3%) [33]. However, Alcantara et al discontinued their clinical trial due to an increase in the number of overall complications (53.8% versus 13.3%) and anastomotic leakage (30.7% versus 0%) in the Surgery Group [22]. In a systematic review and meta-analysis of 8 clinical trials, 51.2% of postoperative complications were described in the Surgery group versus 33.9% in the SEMS Group [21]. Arezzo et al reported that the rate of reoperation was higher in the SEMS Group (10.9% versus 8.7% RR 1.23 p 0.487) as shown in our study (12% versus 5%) without significant differences [21]. These same authors described a very similar postoperative mortality in both groups (9.6% SEMS Group versus 9.9% Surgery Group). However, postoperative mortality in our study was higher in the SEMS Group (17% versus 10% p 0,38).
In this study, the OLCC treatment with SEMS as a BTS versus ES has not shown statistically significant differences regarding in terms of OS, although was lower in SEMS Group. However, when we applied the PS, the distance recurrence was higher in the Surgery Group in a significant statistically way. A higher rate of primary anastomoses and a lower proportion of stoma have also been shown statistically significant in the SEMS group. Postoperative morbidity and mortality did not show statistically significant differences between groups.
Acknowledgements
The authors acknowledge the contribution of the different services of Ramon y Cajal University Hospital (Endoscopy, Radiology, Pathology Department and the Biostatistics Unit) that have allowed to carry out this study.
Statement of Ethics
The study was approved by the Ethics Committee from Ramón y Cajal Hospital in the 1st of April 2015.
Conflict of Interest Statement
There are no conflicts of interest
Funding Sources
We declare that we did not receive any funding to carry out our study.