Introduction
Endoscopic mucosal resection (EMR) and submucosal dissection (ESD) are widely used
as therapeutic methods for resection of non-polypoid lesions of the gastrointestinal
tract, including early cancer. Both techniques require a submucosal injection to separate
the lesion from the muscularis propria and thus reduce complications such as thermal
injury and perforation [1 ]. However, available solutions for submucosal injection are not very long lasting,
require multiple injections in order to adequately perform a complete ESD or a large
EMR, and none have consistently been shown to be more effective or safer than others
[2 ]. Aside from duration, the different compositions and viscosities of these solutions
could potentially damage the submucosa and muscularis propria resulting in impaired
healing or delayed complications [3 ]
[4 ]
[5 ]
[6 ]
[7 ].
Recently, a new treatment for gastroesophageal reflux that combines hyaluronic acid,
chondroitin sulfate, and poloxamer 407 (Ziverel, Norgine, UK) was approved in Spain.
Poloxamers are reverse-phase polymers that undergo a temperature-dependent liquid-to-gel
transformation and can be used for medical applications. Poloxamers have also been
demonstrated to provide a more durable submucosal cushion in ex vivo and in vivo porcine
models [8 ]. However, there is no information about potential tissue damage by these polymers.
The objective of this study was to assess the efficacy and safety of this new substance
as a submucosal injection solution.
Methods
Evaluation of the efficacy
In the ex vivo experiment, a total of 48 gastric submucosal cushions were created
in fresh ex vivo porcine stomachs using gelafundin (Gelaspan, Braun, Germany) (n = 16)
or Ziverel diluted with saline solution at two different concentrations: 50 % (n = 16)
and 80 % (n = 16). Each submucosal injection was performed by injecting 2 mL of solution
using a 21-gauge intramuscular needle and the duration of the cushions was measured
[9 ]. The time for the cushion to disappear was recorded using a stopwatch. Timing began
immediately after withdrawal of the needle and ended when the bleb had flattened completely
[10 ]. All stomachs were placed on thermal pads at a temperature of 37°C [8 ]. The investigator was blinded to the type of solution injected.
Evaluation of immediate tissue damage
In the in vivo model, we created 10 submucosal cushions in the upper third of a single
pig’s stomach by injecting 2 mL of 80 % Ziverel at separate sites using a standard
gastroscope and a 23-gauge catheter injection needle. After a visually adequate submucosal
elevation had been achieved, endoclipping was carried out near the injection sites
for further identification of the locations. The pig was euthanized 30 minutes after
the last injection, and a necropsy was performed [9 ]. The retrieved stomach was fixed with formalin and was cut at the different injection
sites, taking 3-mm-thick slices of each cushion for histology. Paraffin tissue blocks
were obtained from each injection site. Afterwards, 2-µm-thick histologic sections
from each paraffin block were stained with H&E and examined. The study was approved
by the Ethics Animal Committee of Barcelona University.
Statistical analysis
Statistical comparisons between the three groups were made using the Kruskal-Wallis
non-parametric test followed by a one-way analysis of variance (ANOVA) for independent
data and a Tukey test for post-ANOVA pairwise comparisons. A P value < 0.05 was considered statistically significant. Calculations were performed
with SPSS software (SPSS 21.0 for Windows, SPSS, Inc., Chicago, IL, United States).
Results
In the ex vivo study, submucosal cushions created using Ziverel with the two different
concentrations lasted longer than those created using gelafundin ([Table 1 ], [Fig. 1 ]). We did not observe any leakage of fluid from the injection puncture with any solution.
Table 1
Duration of the submucosal cushions.
Gelafundin
50 % Ziverel
80 % Ziverel
ANOVA
Time, mean ± SD (range), min
3.9 ± 1.5 (2 – 7)
13.1 ± 6.6 (4 – 25)
23.1 ± 15.6 (7 – 68)
0.000[1 ]
1 Post-ANOVA test:P = 0.000;P = 0.018;P = 0.03.
Fig. 1 Chronological disappearance of submucosal cushions. Both formulations of Ziverel
maintained the mucosal elevation longer than gelafundin (P = 0.000).
In the in vivo study, the whole procedure lasted 30 minutes from the first injection
to the last one. Each submucosal cushion had a median duration of 49.5 minutes (range
35 – 65 minutes) before the animal was euthanized. All of the cushions apart from
the first one were macroscopically visible at the time of necropsy ([Fig. 2 ]). Macroscopic examination of the specimens showed a gross submucosal bleb in them
all apart from the first one ([Fig. 3 ]). The most significant histological change was edema and vascular ectasia in the
submucosa which progressively decreased with time after injection ([Fig. 4 ]). Mucosal changes consisted of mild edema and vascular ectasia. No muscle damage
or necrosis was identified in any section.
Fig. 2 The retrieved stomach is stretched flat on a cork board with pins. It was possible
to identify the 10 endoclips that indicate the sites of the injections.
Fig. 3 Macroscopic examination of the specimens showing a gross submucosal bleb in all specimens
apart from the first one (top left; number B16-16915-1).
Fig. 4 Microscopic findings at the level of the injections show that the thickness of the
submucosa decreases with time after injection. a normal mucosa without injection: 0.5 mm; b 65 minutes after injection: 1.3 mm; c 50 minutes after injection: 5.27 mm. d 35 minutes after injection: 10.29 mm.
Discussion
The most effective and simplest way to prevent complications during EMR and ESD is
to create a sufficiently thick submucosal layer by endoscopic injection of fluid into
the submucosa. In two comparative studies with several solutions performed in porcine
models, hyaluronic acid lasted longer [10 ]
[11 ]. In our study, the combination of hyaluronic acid with chondroitin sulfate and poloxamer
407 lasted longer than the colloid solution and the duration of elevation was similar
to the results obtained in other ex vivo studies using hyaluronic acid alone [10 ]. Remarkably, in the in vivo stomachs, the duration of elevation was twice as long
and the cushion was still visible up to 60 minutes after the injection. To our knowledge,
these results demonstrate the longest durability reported so far compared with other
solutions.
Gelafundin (succinylated gelatin) was used as the control injection solution because
it is used to perform large EMRs in our unit due to its durability, widespread availability
in our country, and low price.
Damage to the resected specimen can limit precise histological diagnosis of a targeted
lesion, and hyaluronic acid has been shown to be the least harmful [12 ]. Furthermore, tissue damage to the muscle layer may result in delayed bleeding or
perforation. The addition of poloxamer to hyaluronic acid in the present study did
not lead to any damage to the mucosa, muscularis mucosa, or the muscularis propria,
confirming its safety.
Ziverel has the beneficial effect of reducing the permeability of injured mucosa which
is enhanced by the fact that poloxamer 407 facilitates product adhesion to the esophageal
mucosa prolonging its action [13 ]. Based on these properties, we hypothesize that Ziverel could also facilitate the
healing process in the scar left after EMR or ESD and expect an early proliferation
of collagen and elastic fibers as described with other solutions [14 ]
[15 ]. Similar to a previous study in which Ziverel was applied over the mucosa, we do
not expect any organ-specific toxicity. In addition, the cost of Ziverel is significantly
lower than hyaluronic acid (0.085 and 10 euro per mL, respectively) and its use could
decrease the final cost of the procedure (for instance, considering that for a gastric
ESD, at least 50 mL of submucosal solution would be needed, the total cost would be
4.25 euro with Ziverel compared with 500 euro with hyaluronic acid).
The main inconvenience of injecting poloxamer is its high viscosity and this makes
injection fairly difficult without a balloon dilator gun [8 ]. Although we were able to inject Ziverel through a 23G sclerotherapy needle at both
concentrations, the injection was subjectively easier with the more dilute concentration
but, despite this, the duration of the submucosal cushion was still longer than with
the control solution. In the ex vivo study, we used a 21G needle in order to facilitate
the injection.
In this study, we did not perform any endoscopic resection and we do not know how
use of this solution could affect the outcomes and speed of EMR and ESD. We also do
not know how this solution might affect the settings of the electrosurgical units
although, based on initial experience with poloxamer [8 ], we do not expect any differences. Moreover, we do not know the impact of Ziverel
in the prevention of bleeding and perforation, which are the main clinical complications
of endoscopic resection.
This study has several strengths. One is that the endoscopist who produced the submucosal
cushions and timed the blebs was blinded to which solution was being tested. However,
there were noticeable differences in the viscosities of the different solutions during
injection and this made it easy to guess which one was being used. The second strength
is that the study was performed in resected and living stomachs in order to assess
the influence of blood flow and absorption from the tissue confirming that in living
animals, the duration of mucosal elevation with 80 % Ziverel was even longer than
in explanted stomachs.
With regard to limitations, we did not measure the change in height of cushions as
reported in other studies [11 ]
[16 ]. This could be a more objective measure of cushion duration. On the other hand,
we only analyzed the immediate potential tissue damage and this does not exclude long-term
tissue damage, especially after performing an endoscopic resection. However, long-term
(1 week) tissue damage has only been associated with the use of hypertonic solutions
with high osmolality.
In conclusion, the present study demonstrates the potential use of a combination of
hyaluronic acid, chondroitin sulfate, and poloxamer 407 as a solution for submucosal
injection, not only because it lasted longer than other solutions but also because
it did not induce acute tissue damage. Further studies are needed to assess its efficacy
and safety when performing in vivo endoscopic resections.
