Indian Journal of Surgery

https://doi.org/10.1007/s12262-018-1732-5

ORIGINAL ARTICLE

Completely Linear Stapled Versus Handsewn Cervical Esophagogastric

Anastomosis After Esophagectomy
Tarun Kumar 1 & Ravi Krishanappa 2 & Esha Pai 3 & Raxith Sringeri 2 & T. B. Singh 4 & Jyoti Swain 1 &
Sindhuri Kondapavuluri 2 & Manoj Pandey 1

Received: 17 April 2017 / Accepted: 22 January 2018

# Association of Surgeons of India 2018

Abstract
Very limited data is present which compares completely linear stapled to handsewn cervical esophagogastric anastomosis.
Primary objective was to determine whether linearly stapled (LS) anastomosis has lower clinically apparent leaks, when com-
pared to handsewn anastomosis (HS). Secondary objectives were morbidity, mortality, overall leak and stricture rates, and
presence of a symptomatic cervical stricture. This is a comparative study of 77 patients who underwent LS (n = 29) and HS
(n = 48) cervical anastomosis. Anastomotic leak was found to be 19.4% (15/77). In the HS group, 27.08% (13/48) and in the LS
group, 6.89% (2/29), respectively, leaked (p = 0.03), relative risk (RR)—3.93 (95% CI 1.21–15.25). 32.5% (23/77) patients
remained admitted for more than 14 days. 52.1% (25/48) patients in the HS group were discharged within 14 days of surgery;
whereas; 93.1% (27/29) were discharged in LS group (p = 0.001), RR—6.95 (95% CI 2.13–25.94). Overall, 90-day mortality
was 7.8% (6/77). In the HS group, 8.3% (4/48) patients died while in the LS group, 6.8% (2/29) patients died (p = 0.82), RR—
1.21(95% CI 0.27–5.53). In the HS group, 6.25% (3/48) patients were diagnosed with stricture compared to 6.8% (2/29) patients
in the LS group (p = 0.9), RR—0.91 (95% CI 0.19–4.44). Overall stricture rate was 6.4% (5/77). Cervical anastomosis done with
linear staplers has less leak rates compared to handsewn anastomosis.

Keywords Linear stapled esophagogastric anastomosis . Completely stapled cervical anastomosis . Mechanical cervical
esophagogastric anastomosis . Stapled esophagogastric anastomosis

Introduction whether in the neck or thorax, is one of the major causes of

postoperative morbidity [4].
Surgery along with neoadjuvant therapy remains the mainstay There is a paucity of published world literature comparing
of treatment for localized middle- and lower-third esophageal completely linear-stapled and handsewn cervical anastomoses,
cancers [1, 2]. One of the major factors which affects treat- with only two retrospective studies that have focused on this
ment outcomes remains anastomotic leaks, which have been aspect. End-to-end anastomosis versus end-to-side versus side-
reported to be between 15 and 25% [3]. Anastomotic leak, to-side, handsewn versus stapled versus hybrid*, anastomosis
on anterior wall versus posterior wall of stomach, and their
various combinations segregate the data leaving us with limited
* Tarun Kumar numbers in each subset for analysis. Therefore, deriving an
[email protected] unequivocal and meaningful conclusion as to the optimal tech-
nique that minimizes leak rates and morbidity remains elusive.
1
Department of Surgical Oncology, Institute of Medical Sciences, It was this lack of robust evidence that inspired this retrospec-
Banaras Hindu University, Varanasi 221005, India tive analysis of our work, as a basis to know if the same war-
2
Department of Surgical Oncology, JSS Medical College, Mysore rants observation in a prospective controlled setting.
Bangalore Road, Bannimantap, Mysuru 570015, India With the above background, we conducted this retrospec-
3
Department Of Surgical Oncology, Tata Memorial Hospital, E tive analysis of a prospectively maintained database from
Borges Road, Parel, Mumbai 400012, India August 2012 to July 2016 with the aim to compare end-to-
4
Department of Biostatistics, Institute of Medical Sciences, Banaras side linearly stapled (LS) anastomosis versus end-to-side
Hindu University, Varanasi 221005, India handsewn (HS) anastomosis in the neck.
 Indian J Surg

Primary objective of the study was to determine wheth- taken from each patient before any surgical or clinical proce-
er linearly stapled (LS) cervical anastomosis have lower dure. Waiver of consent was obtained from the ethical com-
rates of clinically apparent anastomotic leaks following mittee of respective institutions. Literature was searched using
esophagogastric anastomosis after esophagectomy for can- the string (esophagogastric[All Fields] OR
cer, when compared to handsewn anastomosis (HS). esophagogastric[All Fields]) AND anastomosis[All Fields]
Secondary objectives of the study were to study the mor- AND ((Bneck^[MeSH Terms] OR Bneck^[All Fields]) OR
bidity, mortality, overall leak and stricture rates, and presence (Bneck^[MeSH Terms] OR Bneck^[All Fields] OR
of a symptomatic cervical stricture. Bcervical^[All Fields])) AND (handsewn[All Fields] OR
*Posterior layer stapled, anterior layer handsewn. mechanical[All Fields] OR stapled[All Fields] OR
stapler[All Fields]) NOT (partial[All Fields] OR partially[All
Fields] OR (Bchimera^[MeSH Terms] OR Bchimera^[All
Materials and Methods Fields] OR Bhybrid^[All Fields])) AND ((Clinical
Trial[ptyp] OR Comparative Study[ptyp] OR Review[ptyp]
This manuscript has been written in accordance with the OR systematic[sb]) AND English[lang] AND cancer[sb]).
Strengthening the Reporting of Observational studies in Total of 13 publications were found; all the other articles used
Epidemiology (STROBE) statement: guidelines for reporting as references were either hand searched or were cross
observational studies [5], with the details as follows. references.
We conducted an observational study retrospectively ana- Diagnostic workup and administration of neoadjuvant ther-
lyzing a prospectively maintained database from August 2012 apy was as per the in-keeping with international standards and
to July 2016 that included 77 patients at two different centers: evidence-based guidelines. We specifically compared two
Banaras Hindu University, Varanasi, India, and JSS Hospital, techniques of anastomosis which have been described below
Mysore, India, to compare leak rates of linearly stapled (LS) (Fig. 2).
versus handsewn (HS) cervical anastomosis, both end-to-side,
in esophagectomy done for cancer (Fig. 1). a) LS anastomosis was done using two 55-mm-long linear
The study protocol conforms to the ethical guidelines of the cutters of 3.8-mm (blue) limb length in an end-to-side
BWorld Medical Association Declaration of Helsinki-Ethical fashion on the anterior wall of stomach.
Principles for Medical Research Involving Human Subjects^ b) HS anastomosis was done using single layer, simple
adopted by the 18th WMA General Assembly, Helsinki, interrupted stiches, using polyglactin 3-0, on the posterior
Finland, June 1964, and amended in Fortaleza, Brazil, 2013 wall of stomach in an end to side fashion. Anastomoses
[6]. The data was collected prospectively during routine clin- were done by consultant surgeons with prior experience
ical practice and accordingly, signed informed consent was in esophageal surgery.

Fig.1 Observational study

Indian J Surg

Fig.2 Two techniques of anastomosis

Postoperatively, patients were followed up clinically Results

and any suspicion of leak was confirmed radiologically
by CT scan. Morbidity and mortality: anastomotic leaks From August 2012 to July 2016, 85 patients were explored—
were classified as per components of definition of anas- 3 patients were inoperable due to metastases and local
tomotic leaks as described by Bruce et al. [7] (Table 1). unresectability while in 5 patients, an intrathoracic anastomo-
Morbidity was considered only when the patient sis was performed during Ivor-Lewis/lower thoraco-
remained admitted for more than 14 days or was abdominal approach esophagectomy, leaving behind 77
readmitted. Minor morbidities which did not prolong esophagectomised patients with cervical anastomoses, either
hospital stay for more than 14 days were not consid- handsewn or linearly stapled, both in end-to-side fashion.
ered. Mortality included all-cause death within 90 days Descriptive data has been provided below (Table 2).
of surgery. During long-term follow-up, patients with Anastomotic leak was found to be 19.4% (15/77). In the
dysphagia who required therapeutic dilatation were con- HS group, 13 out of 48 (27.08%), and in the LS group, 2 out of
sidered to have a stricture. Asymptomatic patients were 29 (6.89%), respectively, leaked (p = 0.03, CI 0.51–36.24%).
not investigated for stricture. Eleven and 4 out of 15 leaks were grade clinical minor and
The data has been presented in mean ± standard de- major respectively according to components of definition of
viation for quantitative variables normally distributed anastomotic leaks. Median duration of hospital stay after sur-
and student’s t test has been used to find out significant gery was overall 12 days (10–15 days)—14 days (12–16 days)
difference between the mean values. Quantitative vari- in the HS group and 10 days (9–11.75 days) in the LS group.
ables not normally distributed have been defined as me- Overall morbidity rates were 32.5%, i.e., 25 out of 77 patients
dian ± interquartile range (IQR) and Chi square test has were admitted for more than 14 days. Twnety-five out of 48
been used as a test of significance. Qualitative variables (52.1%) patients in the HS group were discharged within
have been presented in numbers and percentage and Z 14 days of surgery; whereas, 27 out of 29 (93.1%) were
test/Fisher’s exact probability tests have been used to discharged in LS group (p = 0.001, CI 16.6–28.39%).
calculate significant difference between the proportions Overall, 90-day mortality was 7.8% (6/77). In the HS group,
as per suitability. p < 0.05 was considered significant. 4 out of 48(8.3%) patients died secondary to anastomotic leak
All p values were derived from two-tailed tests. 95% (3/4) and pulmonary complications (1/4), respectively, while
confidence intervals (CI) of the difference between the in the LS group, 2 out of 29 (6.8%) patients died secondary to
proportions and the relative risk (RR) with 95% confi- anastomotic leak (1/2) and pulmonary complications (1/2),
dence intervals to compare the two techniques for vari- respectively. The difference in 90-day mortality was not found
ous outcome variables have also been presented. For to be significant between the HS and LS cohorts (p = 0.82, CI
calculation of relative risk, admission days being a con- 16.77–15.21%). All four patients with grade III leak died. In
tinuous variable have been categorized into two groups, the HS group, 3 out of 48 (6.25%) patients were diagnosed
less than12 days and more than 12 days. Twelve is the with stricture at 3 months compared to 2 out of 29 (6.8%)
median number of admission days for total number of patients in the LS group (p = 0.9, CI 12.6–18.57%). Overall
cases. stricture rate at 3 months was 6.4% (5/77) (Table 3).
This research did not receive any specific grant from The RR of leak rates was 3.93 times higher in HS group as
funding agencies in public, commercial, or not—for—profit compared to LS with 95% CI 1.21–15.25. The median days of
sectors. admission were 2.02 times higher with HS group compared to
 Indian J Surg

Table 1 Classification of leaks

Grade Definition Treatment

Radiologic Detected only on routine imaging, no clinical signs. No change in management

Clinical minor Presence of luminal contents through the drain or wound No change in management or intervention but may
site causing local inflammation, e.g., fever (temperature have prolonged hospital stay or delay in resuming
> 38 °C) or leukocytosis (white cell count > 10,000/l). oral intake
Leak may also be detected on imaging studies.
Clinical major As clinical minor. Severe disruption of anastomosis. Leak Change in management and intervention required
may also be detected on imaging studies.

LS group, 95% CI 1.36–29.7. In the HS group, risk of mor- partially stapled anastomosis. They also concluded increased
bidity was found 6.95 times higher than LS group having 95% odds for leak with handsewn anastomosis compared to fully
CI 2.13–25.94 (Table 3). stapled anastomosis i.e. 64.3% vs. 13.2% (p = 0.001). They
noted a non-significant difference in stricture rates between
HS and LS anastomoses (35.1 vs. 21.5%, p = 0.92), with the
Discussion fully stapled cohort having a lower stricture rate.
Though our stricture rates do not parallel our leak rates, all
A review of literature found only two retrospective studies [8, our patients who developed strictures had clinical evidence of
9] which compared completely linear stapled anastomosis to anastomotic leak in the perioperative period as seen in the two
handsewn anastomosis in the neck, while all other studies in- studies by Singh et al. [9] and Price et al. [8].
cluded the hybrid technique or use of circular stapler in the Increased morbidity in HS group is clearly seen to reflect
Bstapled anastomosis^ cohort [10–15] or included intrathoracic increased leak rates. All patients with anastomotic leak had
site of anastomosis [3, 16], precluding comparison to our study. hospital stay more than 14 days. Five out of 13 leaks in HS
Singh et al. [9] studied 93 patients out of which 43 underwent group had intrathoracic collection explaining the morbidity.
handsewn anastomosis, 34 underwent stapled anastomosis, and All leaks were managed by drainage either in the neck or by
remaining 16 had partially stapled hybrid anastomosis. Similar CT-guided drain placement.
to our observation, they showed significantly reduced leak and An interesting observation is the significant reduction in
stricture rates in the stapled group, i.e., 23 vs. 3% (p < 0.05) and operative time in the stapled cohort as noted in the study by
58 vs.18% (p < 0.05) respectively. However, one difference in Saluja et al. [11], where they used the partially stapled or
their technique was that they did not form a stomach tube and hybrid technique where only one layer of the anastomosis is
pulled up a non-tubularised stomach in the neck. stapled. Though we performed a fully stapled anastomosis in
In the review by Price et al. [8], 164 patients out of 432 our study, our data on duration of surgery was incomplete. A
esophagectomies underwent cervical anastomoses - 83 LS and faster stapled anastomosis is fathomable; assuming the swift-
14 HS anastomosis while the remaining 67 had hybrid or ness of the anastomosis may be another factor that improved

Table 2 Descriptive data

Total HS (% of n) LS (% of n) 95% CI of difference in proportions p

Number of patients, n 77 48 29
M: F 54:33 34:14 20:9 19.45–25% 0.8
Mean age in years 58.49 ± 8.1 58.08 ± 7.6 58.9 ± 9.0 4.7–3.09 0.7
Level of growth Middle-third 46(59.7%) 27(56.25%) 19(65.5%) 15.01–31.09% 0.4
Lower-third 31(40.3%) 21(43.75%) 10(34.4%) 15.01–31.09%
Histology SCC 61(79.2%) 37(77.1%) 24(82.8%) 16.12–23.91% 0.5
Adeno 16(20.8%) 11(22.9%) 05(17.2%) 16.12–23.91%
Neoadjuvant chemotherapy Given 38(49.4%) 26(54.1%) 12(41.3%) 11.72–35.16% 0.27
Not given 39(51.6%) 22(45.8%) 17(58.6%) 11.72–35.16%
Surgery Open TTE 19(24.6%) 16(33.3%) 03(10.3%) 0.93–39.98% < 0.01
MIS 40(51.9%) 16(33.3%) 24(82.7%) 24.91–65.91%
THE 18(23.3%) 16(33.3%) 02(6.89%) 5.07–42.65%

Adeno adenocarcinoma, F female, M male, SCC squamous cell carcinoma, MIS mininally invasive surgery, THE transthoracic hiatal esophagectomy,
TTE transthoracic esophagectomy
Indian J Surg

Table 3 Outcome data and results

Outcome variables Total (77) HS (48) LS (29) 95% CI of difference in proportions p value Relative Risk 95% CI

Leak rates 15(19.4%) 13(27.08%) 02(6.89%) 0.51–36.24% 0.03 3.93 1.21–15.25

Median days of admission 12(09–16) 14(12–16) 10(9–11.75) 16.6–28.39% < 0.001 2.02 1.36–2.97
Morbidity 25(32.5%) 23(47.9%) 02(6.8%) 18.5–56.81% < 0.01 6.95 2.13–25.94
Mortality 06(7.8%) 04(8.3%) 02(6.89%) 16.77–15.21% 0.82 1.21 0.27–5.53
Stricture formation 05(6.4%) 03(6.25%) 02(6.89%) 12.6–18.57% 0.90 0.91 0.19–4.44

leak rates by reducing the exposure to anesthetic agents and hospital stay probably explaining the similar results in our
intraoperative fluctuation of blood pressure. studies.
A stapled anastomosis also theoretically eliminates the Secondly, our incomplete data on factors such as blood
margin of human error in surgical technique towards the fag loss, hemoglobin levels, serum albumin, pulmonary function
end of long surgery. This may be a corollary from the obser- tests (forced expiratory volume at 1 s) and comorbidities,
vation that that transhiatal esophagectomy has a trend towards which are known to affect the leak rates in esophageal cancer
reduced leak rates compared to Mckeown’s three-stage proce- patients, precluded any possible regression analysis to know if
dure, where the latter is found to have a comparatively the technique of stapling truly affected outcomes [19].
prolonged duration of surgery. Another hypothesis for lower Tabatabai et al. [19] demonstrated serum albumin levels less
leaks rates historically in transhiatal esophagectomies may be than 3.5 g/dl as one of the significant factors leading to leak.
that all transhiatal esophagectomies were performed for pa- Other factors found significant were FEV1 less than 2 l, in-
tients with lower-third and GE junction tumors requiring creased blood loss during surgery, and pulmonary complica-
wider margins at the fundus resulting in a shorter tube forma- tions. Aminian et al. [20] demonstrated increased leak rates in
tion and hence lower chances of fundal ischemia. We noted a patients with diabetes mellitus and hypertension.
22.03% leak rate in the transthoracic group versus 11.1% in
the transhiatal group; however, this difference was not statis-
tically significant (p = 0.49, CI 15.73–26.85%).
Conclusion
In our series, all 77 anastomoses were performed by three
surgeons adequately trained in esophageal surgery but there is
In the light of above data, authors conclude that a total me-
literature to suggest that anastomosis performed by trainees
chanical cervical anastomosis done with linear staplers has
and residents under supervision of senior surgeons have sim-
less leak rates compared to handsewn anastomosis. While all
ilar outcomes [17]. Also, stapled anastomosis does help in
stricture patients do have clinical evidence of perioperative
standardization of technique across continents which help in
leak, vice versa does not hold true, i.e., all anastomotic leaks
obtaining better research-oriented results as it reduces inter-
may not develop strictures.
personal bias.
As this is a retrospective study, a randomized control trial
Manometric studies [18] on the two techniques of anasto-
with adequate sample size is proposed to derive practice
mosis showed that the mean diameter of anastomosis in the
changing-results.
handsewn group was 1.67 and 1.70 cm in patients with and
without dysphagia respectively. These dimensions were 3.000
and 3.014 cm for the stapled group with and without dyspha-
gia respectively. This observation does raise concern, if at all
diameter of the anastomosis above a critical level plays a role References
in the clinical presentation of dysphagia and warrants further
research. 1. D’Journo XB, Thomas PA (2014 May) Current management of
As with all the retrospective studies, this study also has its esophageal cancer. J Thorac Dis 6(Suppl 2):S253–S264
limitations of numbers and study design. Firstly, data is not 2. Sjoquist KM, Burmeister BH, Smithers BM, Zalcberg JR, Simes
RJ, Barbour A, Gebski V, Australasian Gastro-Intestinal Trials
homogenous between the two cohorts, i.e., in the HS group, Group (2011 Jul 1) Survival after neoadjuvant chemotherapy or
there is equal distribution of cases between the three types of chemoradiotherapy for resectable oesophageal carcinoma: an up-
surgeries performed in the group, i.e., 33% (16 of 48 patients) dated meta-analysis. Lancet Oncol 12(7):681–692. https://doi.org/
underwent transthoracic, minimally invasive, and transhiatal 10.1016/S1470-2045(11)70142-5
3. Deng X-F, Liu Q-X, Zhou D, Min J-X, Dai J-G (2015 Apr 21)
esophagectomy; whereas, in the LS group, majority of the Hand-sewn vs linearly stapled esophagogastric anastomosis for
surgeries performed were by minimally invasive techniques esophageal cancer: a meta-analysis. World J Gastroenterol WJG
which is inherently known to have lower morbidity and 21(15):4757–4764. https://doi.org/10.3748/wjg.v21.i15.4757
 Indian J Surg

4. Xu Q-R, Wang K-N, Wang W-P, Zhang K, Chen L-Q (2011 Jun) 13. Laterza E, de’ Manzoni G, Veraldi GF, Guglielmi A, Tedesco P,
Linear stapled esophagogastrostomy is more effective than hand- Cordiano C (1999 Nov) Manual compared with mechanical cervi-
sewn or circular stapler in prevention of anastomotic stricture: a cal oesophagogastric anastomosis: a randomised trial. Eur J Surg
comparative clinical study. J Gastrointest Surg Off J Soc Surg Acta Chir 165(11):1051–1054
Aliment Tract 15(6):915–921. https://doi.org/10.1007/s11605- 14. Hsu H-H, Chen J-S, Huang P-M, Lee J-M, Lee Y-C (2004 Jun)
011-1490-1 Comparison of manual and mechanical cervical esophagogastric
5. STROBE Statement: Home [Internet]. [cited 2016 Dec 25]. anastomosis after esophageal resection for squamous cell carcino-
Available from: http://strobe-statement.org/index.php?id=strobe- ma: a prospective randomized controlled trial. Eur J Cardio-Thorac
home Surg Off J Eur Assoc Cardio-Thorac Surg. 25(6):1097–1101.
6. WMA Declaration of Helsinki - Ethical Principles for Medical https://doi.org/10.1016/j.ejcts.2004.02.026
Research Involving Human Subjects [Internet]. 2013 [cited 2016 15. Wang Q, He X-R, Shi C-H, Tian J-H, Jiang L, He S-L, Yang KH
Dec 25]. Available from: http://www.wma.net/en/30publications/ (2015 Apr) Hand-sewn versus stapled esophagogastric anastomosis
10policies/b3/ in the neck: a systematic review and meta-analysis of randomized
7. Bruce J, Krukowski ZH, Al-Khairy G, Russell EM, Park KG (2001 controlled trials. Indian J Surg 77(2):133–140. https://doi.org/10.
Sep) Systematic review of the definition and measurement of anas- 1007/s12262-013-0984-3
tomotic leak after gastrointestinal surgery. Br J Surg 88(9):1157– 16. Walther B, Johansson J, Johnsson F, Von Holstein CS, Zilling T
1168. https://doi.org/10.1046/j.0007-1323.2001.01829.x (2003 Dec) Cervical or thoracic anastomosis after esophageal re-
8. Price TN, Nichols FC, Harmsen WS, Allen MS, Cassivi SD, Wigle section and gastric tube reconstruction: a prospective randomized
DA, Shen KR, Deschamps C (2013 Apr) A comprehensive review trial comparing sutured neck anastomosis with stapled intrathoracic
of anastomotic technique in 432 esophagectomies. Ann Thorac anastomosis. Ann Surg 238(6):803–812-814. https://doi.org/10.
Surg 95(4):1154–1160-1161. https://doi.org/10.1016/j.athoracsur. 1097/01.sla.0000098624.04100.b1
2012.11.045 17. Handagala SDM, Addae-Boateng E, Beggs D, Duffy JP, Martin-
9. Singh D, Maley RH, Santucci T, Macherey RS, Bartley S, Weyant Ucar AE (2012 Jan) Early outcomes of surgery for oesophageal
RJ, Landreneau RJ (2001 Feb) Experience and technique of stapled cancer in a thoracic regional unit. Can we maintain training without
mechanical cervical esophagogastric anastomosis. Ann Thorac compromising results? Eur J Cardio-Thorac Surg Off J Eur Assoc
Surg 71(2):419–424. https://doi.org/10.1016/S0003-4975(00) Cardio-Thorac Surg 41(1):31–35
02337-7 18. Deng B, Wang R-W, Jiang Y-G, Tan Q-Y, Zhao Y-P, Zhou J-H, Liao
10. Wang W-P, Gao Q, Wang K-N, Shi H, Chen L-Q (2013 May) A XL, Ma Z (2009 Jan) Functional and menometric study of side-to-
prospective randomized controlled trial of semi-mechanical versus side stapled anastomosis and traditional hand-sewn anastomosis in
hand-sewn or circular stapled esophagogastrostomy for prevention cervical esophagogastrostomy. Eur J Cardio-Thorac Surg Off J Eur
of anastomotic stricture. World J Surg 37(5):1043–1050. https://doi. Assoc Cardio-Thorac Surg. 35(1):8–12. https://doi.org/10.1016/j.
org/10.1007/s00268-013-1932-x ejcts.2008.09.008
11. Saluja SS, Ray S, Pal S, Sanyal S, Agrawal N, Dash NR, Sahni P, 19. Tabatabai A, Hashemi M, Mohajeri G, Ahmadinejad M,
Chattopadhyay TK (2012 Jul) Randomized trial comparing side-to- Khan IA, Haghdani S (2009) Incidence and risk factors pre-
side stapled and hand-sewn esophagogastric anastomosis in neck. J disposing anastomotic leak after transhiatal esophagectomy.
Gastrointest Surg Off J Soc Surg Aliment Tract. 16(7):1287–1295. Ann Thorac Med 4(4):197–200. https://doi.org/10.4103/1817-
https://doi.org/10.1007/s11605-012-1885-7 1737.56012
12. Valverde A, Hay JM, Fingerhut A, Elhadad A (1996 Sep) Manual 20. Aminian A, Panahi N, Mirsharifi R, Karimian F, Meysamie
versus mechanical esophagogastric anastomosis after resection for A, Khorgami Z et al (2011 Dec) Predictors and outcome of
carcinoma: a controlled trial. French Associations for Surgical cervical anastomotic leakage after esophageal cancer surgery.
Research. Surgery 120(3):476–483. https://doi.org/10.1016/ J Cancer Res Ther 7(4):448–453. https://doi.org/10.4103/
S0039-6060(96)80066-3 0973-1482.92016