Keywords
esophageal biomechanics - surgical suturing - esophageal surgery - swine
Introduction
Long-gap forms of esophageal atresia represent a separate entity within the esophageal
atresia disease spectrum and pose a variety of challenges in their management.[1] Foker described holding the ends of the esophagus under tension for several minutes
to enabling a primary repair of long-gap esophageal atresia by the resulting elongation.[2] However, as he acknowledges, this maneuver results in anastomotic tension.[3] Tension on the other hand has been blamed to be a reproducible risk factor for anastomotic
stricture[4]
[5] as well as anastomotic leakage[5]
[6]
[7] in esophageal anastomoses, among others.[8]
[9] We have recently been able to prove experimentally that several minutes of constant
traction results in more pronounced esophageal elongation than a similar steady increase
of traction forces.[10] In the same report, we also investigated the durability of different anastomotic
techniques and found that the simple continuous suture outperforms the simple interrupted
suture in terms of anastomotic resilience.[10] We conducted a secondary analysis of these data to investigate whether the more
durable simple continuous anastomosis would also allow for more esophageal elongation
than the simple interrupted suture anastomosis.
Materials and Methods
The data on the resistance to traction forces of several anastomotic techniques is
available elsewhere.[10] In this experiment, we measured the elongation of the specimens during application
of traction forces until failure of the anastomoses for simple continuous and simple
interrupted suture anastomoses. Our methodology has been described in detail before.[10]
[11] In brief, esophagi of swine (Sus scrofa domestica, weighing 100–120 kg at an age of 100–120 days) were obtained from a local abattoir
(Färber, Alzey, Germany), stored in a humid atmosphere, and processed to the end of
the experiment within 5 hours of slaughtering as described previously for the investigation
of biomechanical tissue properties.[11]
[12]
[13]
[14]
[15] Esophagi were freed from its surrounding tissue, mounted in a motorized horizontal
test stand (Sauter THM, Sauter, Balingen, Germany), and subjected to linear traction
until visible disruption of the circumferential muscular layer occurred. This was
objectified by traction forces that did not increase beyond this point. A video of
the endpoint has been published before to make this abstract description more accessible
for the reader by its visual depiction.[10] All anastomoses were constructed by the same surgeon (C.O.) at 13 cm distal to the
upper esophageal sphincter using USP 4–0 polydioxanone (Ethicon, Norderstedt, Germany)
sutures. Initial length of all esophagi was around 26 cm as depicted elsewhere.[10] Suture bite length was 8 mm to the cut surface, with a 5-mm lateral advancement
([Fig. 1]). This summed up to a relatively uniform number of 10 stitches per anastomosis.[10] All endpoints were verified by a second researcher (A.L.).
Fig. 1 Photographic documentation of the experimental procedures. (A) Simple interrupted suture anastomosis of 8 mm distance to cut surface and 5 mm lateral
distance between the stitches. (B) Simple continuous anastomosis of 8 mm distance to the cut surface and 5 mm lateral
distance between the stitches.
We used an exploratory investigation of five specimens per group to conduct an a priori
power analysis using G*Power[16] (version 3.9.1.2) to achieve statistical power of 80% for a two-tailed t-test for independent groups with an α-level of 5%. The measurements for the simple continuous anastomoses were randomly
drawn from the basis of 28 measurements in the original study[10] using R's random number generating algorithm to achieve a balanced experimental
design. We conducted all statistical analyses using R (version 3.4.3) with the stats4
package (version 3.4.3). A Gaussian distribution within our results was confirmed
with the Shapiro–Wilk test and homoscedasticity was tested by Bartlett's test. Comparisons
were conducted by using Student's t-test. We used the bias corrected, accelerated bootstrap procedure[17] to calculate the 95% confidence interval of the group mean with 10,000 iterations
as recommended elsewhere[18] using the groupwiseMean function from the rcompanion package (version 1.13.2).[19]
Discussion
Pigs were the traditional model in experimental pediatric esophageal surgery,[20] but have largely been replaced by rodent models.[21] Recently, swine experienced a revival as a favorite model for experimental esophageal
surgery,[22]
[23]
[24]
[25]
[26]
[27]
[28]
[29]
[30] because porcine closely mimics human anatomy.[20]
[21]
[22] Moreover, tissue equivalence of porcine and human esophagi[31] supports using swine as an experimental model, in particular because tissue equivalence
is a feature unique to the esophagus, but not to other organs, for example, the colon.[32]
There has been an ongoing debate on the right anastomotic technique in adult esophageal
surgery. The most common techniques have narrowed to stapled versus continuous hand-sewn
anastomoses,[33] but recently also included robotic-assisted anastomoses.[34] In pediatric esophageal surgery, robot-assisted esophageal surgery has only been
conducted experimentally in swine.[35]
[36] Concerning staplers, we have previously shown that their use is currently not advisable
due to size discrepancy between staplers and patient's anatomy.[37] Moreover, the vast majority of operations for esophageal atresia are still conducted
via the open approach,[38]
[39] which also suggests the comparison of hand-sewn techniques would have the widest
relevance for the practicing pediatric surgeon.
Interestingly, the continuous outperformed the simple interrupted suture anastomosis
for esophageal elongation, most probably due to a different dispersal of traction
forces within the tissue: Traction might propagate along the suture into the organ
and because the continuous suture has a higher contact surface compared with the simple
interrupted suture anastomosis, enabling larger amplitude of esophageal elongation.
This might explain why simple interrupted suture anastomoses did not achieve anastomotic
resiliencies similar to the native organ.[10]
[40]
Our experiment has a statistical power sufficient enough to ensure that we were not
investigating noise but a true effect based on the a priori power analysis. Given
the equivalence of reaction of porcine and human esophageal tissue to traction, our
results strongly encourage the transition into an experiment with live animals to
further investigate the phenomenon before transition into everyday clinical practice.
The implementation of Myers' decades old dictum “the best oesophagus is the patient's
own oesophagus”[41] might be achieved by esophageal elongation following several minutes of traction,[2] but it also requires securing the length gain by prevention of anastomotic leakages.
They are linked to anastomotic tension,[5]
[6]
[7] which could be modified by using continuous suturing that tolerates more anastomotic
tension[10] and thereby esophageal elongation before anastomotic failure.
Conclusion
The simple continuous esophageal anastomosis outperforms the simple interrupted suture
anastomosis in esophageal elongation and resistance to traction forces ex vivo. This
result should be corroborated in an in vivo experiment to explore its suitability
for clinical practice.
