The systematic searches returned a total of 731 citations. Following deduplication, 441 citations were identified. Of the 441 citations, 73 full manuscripts were identified as potentially eligible with a total of 17 RCTs meeting our eligibility criteria (n = 3,157 patients). Eleven studies were included in our meta-analysis as shown in our PRISMA flow diagram (Figure 1).
Studies were published between 1996 and 2019, with sample sizes ranging from 32 to 516 participants. The mean age of participants was similar across studies ranging from 50.8 to 67.5 years old. Follow-up periods were highly variable ranging from 3 months to 3 years. Most studies were performed in China (5 studies, 19%), India (4 studies, 24%), Japan (3 studies, 18%), or other (5 studies [Iran, Hong Kong, Thailand, Netherlands, France], 29%). The number of male participants was higher than female participants in all studies except one. The incidence of AL ranged from 1.4 to 17%. The patient characteristics of the included studies are provided in Table 1.
Table 1. Characteristics of the Included Studies and Participants Studies
First author (year)
|
Country
|
Total participants (N)
|
Total participants in intervention groups (n)
|
Prevalence AL (%)
|
Male-to-female ratio evaluated
(T, I, C)
|
Age, years (y)
Mean ± SD
|
Neoadjuvant therapy
|
Follow-up, months (mo, wk, or y)
Mean ± SD
|
Bhat 20066
|
India
|
194
|
97
|
8.8
|
T: 3:1.8
|
T: 52.5
|
Excluded patients with previous neoadjuvant Tx
|
F/U every 3-mo for 3 y post-op every 4- to 6-mo post-op
|
Dai 20115
|
China
|
253
|
127
|
3.1
|
T: 4:01
|
T: 63.5
|
Excluded patients with previous neoadjuvant Tx
|
*22 mo (3-52 mo)
|
Daryaei 200819
|
Iran
|
40
|
18
|
15
|
NR
|
T: 58.4 ± 10.3
|
NP
|
NP
|
Gupta 200120
|
India
|
100
|
48
|
12
|
I: 0.8:1
C: 0.7:1
|
I: 51.3 ± 13.0
C: 50.8 ± 13.2
|
Rad ± Chemo: 22/100 and Chemo alone: 56/100
|
3 mo or more
|
Hayashi 201921
|
Japan
|
71
|
34
|
5.6
|
T: 6:8
|
T: 63.04
|
NP
|
NP
|
Law 199722
|
Hong Kong
|
122
|
61
|
3.3
|
I: 7.7:1
C: 6.6:1
|
I: 64 ± 1.2
C: 63 ± 1.0
|
NP
|
20 (SD 2.2) mo hand-sewn and 19 (2.2) mo stapled group (p= NS)
|
Liu 20147
|
China
|
378
|
188
|
4.2
|
T: 3:01
|
T: 64
|
Excluded patients with previous neoadjuvant Tx
|
NP
|
Liu 201523
|
China
|
432
|
219
|
5
|
I: 3:1
C: 3:1
|
I: 62 ± 8
C: 61 ± 9
|
Rad + chemo: 64/478
|
17.8 (3.2) mo hand-sewn and 18.3 (3.4) mo stapled
|
Luechakiettisak 200824
|
Thailand
|
104
|
52
|
4.8
|
I: 4.8:1
C: 5.6:1
|
I: 63.6
C: 62.0
|
NP
|
NP
|
Mistry 201225
|
India
|
253
|
127
|
3.1
|
T: 2.1:1
|
I: 53.4
C: 56.7
|
Rad ± Chemo: 2/150 and Chemo alone: 72/150
|
NP
|
Nederlof 201126
|
Netherlands
|
128
|
64
|
31
|
I: 2:1
C: 7:1
|
I: 60
C: 63
|
Rad + Chemo: 27/64 and Chemo alone: 17/64
|
3- or 6-wk outpatient visit. 3 mo first y post-op. Every 4 mo second y post-op.
|
Okuyama 200727
|
Japan
|
32
|
14
|
12
|
I: 13:1
C: 16:2
|
I: 63.6
C: 64.3
|
Excluded patients with previous neoadjuvant Tx
|
5 y
|
Saluja 201228
|
India
|
174
|
87
|
17
|
I: 2.3:1
C: 1.6:1
|
I: 51.4 ± 12
C:50.9 ± 14
|
Rad + Chemo: 107/174
|
NP
|
Zhang 201029
|
China
|
516
|
272
|
1.4
|
I: 1.4:1
C: 1.4:1
|
I: 59 ± 1.2
C: 60 ± 1.3
|
Excluded patients with previous neoadjuvant Tx
|
12 mo
|
Zheng 20134
|
China
|
164
|
82
|
8.5
|
I: 1.6:1
C: 1.4:1
|
I: 67.5 ± 11.2
C: 65.7 ± 9.4
|
None of the patients received chemotherapy or radiotherapy pre-op
|
3 y
|
Tabira 200430
|
Japan
|
44
|
22
|
14
|
I: 6.3:1
C: 10:1
|
I: 64 ± 8
C: 60 ± 8
|
NP
|
NP
|
Valverde 199631
|
France
|
152
|
78
|
16
|
I: 9.6:1
C: 10.1:1
|
I: 59 ± 9
C: 59 ± 10
|
NP
|
NP
|
Abbreviations: C, control; Chemo, chemotherapy; I, intervention; mo, months; NP, not provided; pre-op, pre-operatively; post-op, post-operatively; Rad, radiation; SD, standard deviation; Tx, therapy; wk, weeks; Y, years
*Median (range)
Seven studies (41%) investigated stapled (vs. hand sewn) anastomosis, three studies (18%) investigated omentoplasty (vs. hand-sewn or stapled) anastomosis, three studies (18%) investigated early removal (postop day 1 or 2 days) or no nasogastric tube (vs. conventional 7 to 10 days to nasogastric tube removal), two studies (12%) that investigated subtotal gastric resection (vs. slender gastric tube) reconstruction, one study (6%) investigated valvuloplasty (vs. stapled) anastomosis, and one study (6%) that compared end-to-end (vs. end-to-side) anastomosis. Sixteen studies (94%) used contrast to diagnose AL and six studies (35%) used additional endoscopy and/or chest tube or drain output. Seven studies (41%) administered medical management, three studies (18%) administered surgical management, and two studies (12%) administered endoscopic management for the treatment of AL. The length of stay in hospital postoperatively varied from 10.7 to 29.4 days. The study intervention characteristics are outlined in Table 2.
Table 2. Intervention Characteristics of the Included Studies
First author (year)
|
Intervention and control groups
|
Surgical approach to intervention
|
Length of Stay,
days
|
Diagnostic modality for anastomosis
|
Medical management
|
Endoscopic management
|
Surgical management
|
Bhat 20066
|
Omentoplasty (I) vs hand-sewn anastomosis alone (C)
|
Cervical: 102
Thoracic: 92
|
NP
|
Water-soluble contrast
|
Abx, bronchodilators, chest physiotherapy
|
Re-insertion NG tube
|
Re-exploration, refashioning anastomosis
|
Dai 20115
|
Omentoplasty (I) vs stapled anastomosis alone (C)
|
Cervical: 75
Thoracic: 180
|
I: 20.4 (11.5)**
C: 23.1 (15.2)**
|
Contrast
|
NP
|
NP
|
NP
|
Daryaei 200819
|
NG tube (I) vs Metoclopramide (C)
|
Cervical: 20
Thoracic: 20
|
I: 10.9 (3.5)**
C: 13.9 (8.2)**
|
Gastrografin contrast
|
Metoclopramide (C)
|
NP
|
NP
|
Gupta 2001V20
|
Subtotal (I) vs slender anastomosis (C)
|
Cervical only
|
I: 10.7 (3.6)**
C: 11.9 (5.6)**
|
Water-soluble contrast
|
Not reported
|
NP
|
NP
|
Hayashi 2019V21
|
No or early NG tube removal (I) vs prolonged NG tube (C)
|
Thoracic only
|
I: 25.7 (12.76)**
C: 29.4 (18.06)**
|
Contrast agent
|
All patients received PPI, ICU admission post-op
|
Re-insertion of NG tube
|
Tracheostomy, mini-tracheostomy
|
Law 199722
|
Stapler (I) vs hand sewn anastomosis (C)
|
Thoracic only
|
NP
|
Gastrografin contrast, endoscopy
|
NP
|
NP
|
NP
|
Liu 20147
|
Valvuloplasty (I) vs stapled anastomosis alone (C)
|
Cervical: 126
Thoracic: 259
|
I: 20.4 (11.5)**
C: 22.1 (15.2)**
|
Contrast, endoscopy
|
NP
|
NP
|
NP
|
Liu 201523
|
Stapler (I) vs hand sewn anastomosis (C)
|
Cervical: 113
Thoracic: 354
|
I: 20.1 (6.8)**
C: 18.9 (7.3)**
|
Barium swallow, endoscopy
|
Nutrition, chest tube drain
|
NP
|
NP
|
Luechakiettisak 200824
|
Stapler (I) vs hand sewn anastomosis (C)
|
Thoracic only
|
NP
|
Gastrografin contrast
|
NP
|
NP
|
NP
|
Mistry 201225
|
Short-term (I) vs prolonged NG tube (C)
|
Cervical: 33
Thoracic: 117
|
I: 12 (9 – 17)*
C: 12 (10 – 17)*
|
Contrast
|
NP
|
NP
|
NP
|
Nederlof 2011V26
|
End-to-end (I) vs side-to-end (C) anastomosis
|
Cervical: 88
Thoracic: 40
|
I: 15 (9 – 125)*
C: 22 (8 – 281)*
|
Contrast, endoscopy, neck wound saliva
|
NP
|
NP
|
Re-operation
|
Okuyama 200727
|
Stapler (I) vs hand sewn anastomosis (C)
|
Cervical: 18
Thoracic: 14
|
NP
|
Water-soluble contrast
|
Conservative
|
NP
|
NP
|
Saluja 201228
|
Stapler (I) vs hand sewn anastomosis (C)
|
Cervical only
|
I: 12.8 (8)**
C: 11.9 (6)**
|
Gastrografin contrast
|
Abx, opening neck wound
|
NP
|
NP
|
Zhang 201029
|
Stapler (I) vs hand sewn anastomosis (C)
|
Thoracic only
|
NP
|
Chest tube output, contrast barium, endoscopy
|
Nutrition, chest tube drain
|
NP
|
NP
|
Zheng 20134
|
Omentoplasty (I) vs hand-sewn anastomosis (C)
|
Thoracic only
|
I: 21 (5)**
C: 23 (6)**
|
Gastrografin contrast
|
NP
|
NP
|
NP
|
Tabira 200430
|
Subtotal (I) vs slender gastric tube (C)
|
Thoracic only
|
NP
|
NP
|
Conservative
|
NP
|
NP
|
Valverde 1996V31
|
Stapler (I) vs hand sewn anastomosis (C)
|
Cervical: 45
Thoracic: 107
|
NP
|
Swallow, methylene blue, interstitial fluid in drains
|
NP
|
NP
|
NP
|
Abbreviations: Abx, antibiotics; C, control; I, intervention; mo, months; NP, not provided; POD, post-operative day; TE, transesophageal; TH, transhiatal; TT, transthoracic; wk, weeks; Y, years
*Median (IQR)
**Mean (SD)
VExcluded from meta-analysis (Valverde 1996, Group results influenced by multiple additional interventions; Nederlof 2011, only study to report intervention type; Hayashi 2019, excluded from AL pooled results because of restriction to reporting grade 3+ AL only; Gupta 2001, only study to report intervention type)
Primary Outcome
Anastomotic leak
The pooled results for 11 meta-analyzed studies are summarized in Figure 2 and the descriptive results for single RCT interventions are summarized in Supplemental 2. Esophagectomy patients that received stapled esophagogastric anastomosis demonstrated a similar reduction in risk of AL (RR: 0.92; 95% CI: 0.45, 1.87; I2 40.1%) compared to hand-sewn (6 studies, n = 1,454 patients). Esophagectomy patients that received omentoplasty had a significant 78% lower risk of leakage (RR: 0.22; 95% CI: 0.10, 0.50; I2 0%) compared to hand-sewn or stapled anastomosis alone (3 studies, n = 611 patients). Esophagectomy patients with early removal or no nasogastric tube demonstrated a significant 62% reduction in risk of leakage (RR: 0.38; 95% CI: 0.02, 0.65; I2 0%) compared to prolonged nasogastric tube removal (2 studies, n = 293 patients).
The pooled risk ratios (RR) for AL were sub-grouped according to the site of esophagogastric anastomosis among 2 studies (Table 3). The pooled risk ratio for esophagectomy patients grouped according to cervical esophagogastric anastomosis (2 studies, RR: 0.23; 95% CI: 0.069, 0.788; I2 0%) was similar to the pooled RR for thoracic esophagogastric anastomosis (2 studies, RR: 0.19; 95% CI: 0.034, 1.032; I2 0%). The pooled risk ratios (RR) for AL among omentoplasty patients were also sub-grouped according to whether comparison groups received a stapled or hand-sewn anastomosis (Supplemental 3). The risk ratio for patients in the stapled anastomosis study (1 study, n = 194 patients, RR: 0.214; 95% CI: 0.064, 0.722) was similar to the pooled risk of AL for patients in the hand-sewn anastomosis studies (2 studies, n = 417 patients, RR: 0.264; 95% CI: 0.089, 0.789). Due to a lack of reporting of AL according to neoadjuvant therapy type (radiation and/or chemotherapy), it was not possible to conduct a stratified analysis.
Table 3. Risk ratios for anastomotic leak for omentoplasty intervention (sub-group by cervical or thoracic approach)
Group
|
Study (Author, year)
|
Risk ratio
|
95% CI (lower, upper)
|
I2
|
Cervical*
|
Bhat 20066
|
0.22
|
0.080, 0.88*
|
-
|
Dai 20115
|
0.26
|
0.030, 2.08
|
-
|
Overall (n = 2 studies)
|
0.23
|
0.080, 0.88*
|
0
|
Thoracic*
|
Bhat 20066
|
0.19
|
0.020, 1.52
|
-
|
Dai 20115
|
0.18
|
0.010, 3.68
|
-
|
Overall (n = 2 studies)
|
0.19
|
0.030, 1.03
|
0
|
*omentoplasty vs. stapled or hand-sewn anastomosis
Secondary Outcomes
Anastomotic stricture
Esophagectomy patients that received stapled esophagogastric anastomosis had a 2-fold increased risk of stricture (RR: 2.11; 95% CI: 1.36, 3.26; I2 35.0%) compared to hand-sewn esophagogastric anastomosis (6 studies, n = 1,380 patients). Esophagectomy patients that received omentoplasty had an 8% lower and not significantly different risk of stricture (RR: 0.92; 95% CI: 0.33, 2.57; I2 65.1%) compared to no omentoplasty (3 studies, n = 613 patients). The pooled results are summarized in Figure 3.
Mortality rate
Esophagectomy patients that received stapled esophagogastric anastomosis had no statistically significant difference in risk of mortality (RR: 1.22; 95% CI: 0.75, 1.98; I2 0%) compared to hand-sewn esophagogastric anastomosis (6 studies, n =1,363 patients). Esophagectomy patients that received omentoplasty had a 20% lower risk of mortality (RR: 0.80; 95% CI: 0.32, 2.0; I2 0%) compared to no omentoplasty (3 studies, n = 736 patients). Esophagectomy patients with early removal or no nasogastric tube demonstrated no statistically significant difference in risk of mortality (RR: 0.90; 95% CI: 0.317, 2.55; I2 0%) compared to prolonged nasogastric tube removal (2 studies, n =190 patients). The pooled results are summarized in Figure 4.
Length of Stay
The weighted mean difference (WMD) in the postoperative length of stay in hospital was determined based on the mean (± SD) reported among the included studies. Esophagectomy patients that received stapled anastomosis did not have a significantly different mean length of stay in hospital with a stay of 1.1 days longer [95% CI: -0.01, 2.2; I2 0%] compared to hand-sewn (2 studies, n = 606 patients). Esophagectomy patients that received omentoplasty had a statistically significant 2.1 day shorter mean length of stay in hospital (WMD: -2.1; 95% CI: -3.6, -0.6; I2 0%) compared hand-sewn or stapled anastomosis alone (2 studies, n = 417 patients). Esophagectomy patients with early removal or no nasogastric tube demonstrated a non-significant 3.2 day shorter mean length of stay in hospital (WMD: -3.2; 95% CI: -6.5, 0.2; I2 0%) compared to prolonged nasogastric tube removal (2 studies, n = 111 patients). Mistry et al. 2012 was excluded from the pooled WMD estimate as the investigators reported a median (IQR). Mistry et al. 2012 reported that the early removal or no nasogastric tube and prolonged nasogastric tube removal groups both had a length of stay of 12 days with no statistically significant difference (P=0.18) [26].
Risk of Bias
Seven (64%) meta-analyzed studies did not report whether the allocation of participants was concealed. Nine (82%) meta-analyzed studies lacked any details surrounding blinding of outcome assessment was blinded. Ten (91%) meta-analyzed studies lacked reporting of outcome assessment blinding. The risk of bias results are summarized in Figure 5 (Individual study risk of bias summarized in Supplemental 4).
Grade
There was a high quality of evidence for AL in the omentoplasty intervention. The unclear risk of bias in omentoplasty studies was due to the lack of allocation concealment in one study decreased the quality of evidence by one level. The large magnitude of effect in the omentoplasty studies increased the quality of evidence by one level. There was a moderate quality of evidence for AL in the early removal or no nasogastric tube intervention. The high risk of bias due to both the lack of randomization and allocation concealment in all studies decreased the quality of evidence by two levels. The large magnitude of effect increased the quality of evidence by one level. There was a very low quality of evidence for AL in the stapled anastomosis intervention. The high risk of bias due to both the lack of randomization and allocation concealment in nearly all studies decreased the quality of evidence by two levels. The imprecision of the measure of effect due to the lack of statistical significance reduced the quality of evidence by one level. The moderate level of heterogeneity in the pooled estimate decreased the quality of evidence by one level. The evidence profile is summarized according to intervention type in Table 4 (GRADE summarized in Supplemental 5).
Table 4. Summary of Findings (11 meta-analyzed studies)
Intervention
|
No. Participants (studies)
|
Quality of Evidence
|
Measure of effect, RR (95% CI)
|
Omentoplasty vs conventional anastomosisℓ
|
611 (3 studies)
|
+ + + + (high quality)
-1: unclear risk of bias1
+1: large magnitude of effect
|
RR = 0.22 (78% risk reduction)
95% CI = 0.1, 0.5*
|
Early or no NG tube decompression vs standard
|
374 (2 studies)
|
+ + + - (moderate quality)
-2: high risk of bias2
+1: large magnitude of effect
|
RR = 0.38 (62% risk reduction)
95% CI = 0.02, 0.65*
|
Stapled vs hand-sewn anastomosis
|
1532 (6 studies)
|
- - - - (very low quality) -2: high risk of bias2
-1: imprecision in measure of effect3
-1: inconsistency across studies4
|
RR = 0.92 (8% risk reduction)
95% CI = 0.45, 1.87
|
RR: Risk ratio; CI: Confidence Interval
|
GRADE: working group grades of evidence
High quality (+ + + +): more research very unlikely to change the estimate of effect
Moderate quality (+ + + -): means further research is likely to have an important impact on our confidence in the estimate of effect and may alter the estimate
Low quality (+ + - -): means that the effect estimate is limited and may substantially differ from
Very low quality (+ - - - or - - - -): grade means that we have little confidence in the effect estimate
|
*statistically significant confidence interval
ℓstapled or hand-sewn anastomosis
1one study lacked allocation concealment
2lack of randomization and allocation concealment
3optimal information size not met (appendix 8) and the 95% CI for the effect estimate crosses the null (RR = 1.0)
4moderate heterogeneity (I2 = 40.1%)