Keywords
gastrostomy - postoperative complications - endoscopy
Introduction
The standard management of pediatric patients requiring long-term enteral nutritional
support is a gastrostomy feeding tube, and percutaneous endoscopic gastrostomy (PEG)
remains the technique of choice. Although different procedures for PEG placement have
been described, the pull technique (P-PEG) and the T-fasteners technique (T-PEG) are
the most frequent ones used in the pediatric population.[1] The comparison of T-PEG and P-PEG is clinically significant because, despite their
widespread use, there is limited reported experience with the use of T-PEG in pediatric
patients.[1]
The P-PEG technique, also known as the Ponsky technique, is the most commonly used,
involving the anterograde insertion of a gastrostomy tube. Its main disadvantage is
the solid internal bolster, which requires general anesthesia or sedation for its
removal (either endoscopically or by traction) to perform the first tube replacement.[2]
[3]
[4] In contrast, the T-PEG technique allows for the direct insertion of a balloon gastrostomy
tube. This is facilitated by T-fasteners anchoring the stomach to the abdominal wall,
enabling its retrograde insertion through a Seldinger technique. The main advantage
of T-PEG is the absence of the need for sedation or general anesthesia to replace
the balloon gastrostomy tube, as it can be easily done by deflating it.[2]
[4]
[5]
[6]
[7]
Despite the growing popularity of the T-PEG technique, there is limited research available
on its safety in pediatric patients, with only a few studies published.[2]
[4]
[6]
[8]
[9] Additionally, there is a shortage of literature comparing the incidence of complications
between P-PEG and T-PEG in children, with only two studies published.[2]
[4] It is worth mentioning that complications related to PEG placement are common,[10] underscoring the critical importance of analyzing PEG, especially regarding pediatric
patient outcomes.
This study aimed to analyze the complication rates in P-PEG and T-PEG placement procedures
in pediatric patients, focusing on those complications that required anesthetic interventions
for management.
Methods
We conducted a retrospective, observational cohort analysis of 146 pediatric patients
who underwent PEG placement procedures between 2010 and 2021. The inclusion criteria
comprised all pediatric patients under 16 who were recommended to undergo PEG placement.
Each patient was evaluated by a multidisciplinary team of pediatric gastroenterologists
and pediatric surgeons to assess the indication for PEG placement. The exclusion criteria
involved neonates under 3 months old due to our consistent recommendation for the
placement of a P-PEG within this age group. The limited abdominal space in neonatal
patients may not allow for the placement of the three anchors required for T-PEG,
and the use of a long dilator could potentially injure the posterior wall of the stomach.
These anatomical space limitations regarding T-PEG in neonatal patients were previously
documented.[4] Our center initially used P-PEG as the primary method for PEG placement (2010–2016).
However, since 2016, we have shifted our preference to T-PEG as the primary technique.
All PEG placement procedures were performed using an endoscopic, percutaneous approach.
Laparoscopy was employed in selected cases, such as those requiring concomitant antireflux
surgery or those with severe hepatomegaly and scoliosis, to ensure greater intraoperative
safety. The gastrostomy placement technique and surgical steps remained identical,
whether or not laparoscopy was employed simultaneously. The stomach was insufflated,
and the abdominal wall was transilluminated. Translighting and digital pressure showed
the best site for gastrostomy placement. Intraoperative intravenous antibiotics (Cefoxitin)
were administered and maintained for 24 hours, according to the guidelines specified
in our center's protocol for surgical antibiotic prophylaxis in pediatric patients.[11] Progressive enteral tolerance was initiated 24 hours after the surgical procedure.
After achieving adequate enteral intake, all patients were discharged and subsequently
received follow-up at the outpatient clinic, where a multidisciplinary team consisting
of a specialist nurse, a pediatric surgeon, and a gastroenterologist conducted regular
assessments. Complications were recorded either during outpatient clinic visits or
when patients presented to the emergency department. The postoperative evolution was
recorded over 1.5 years in all patients.
Patients were divided into two groups depending on the technique used (P-PEG vs. T-PEG).
Data collected included age, weight, sex, and associated comorbidities (classified
as neurological, cardiac, respiratory, and others). Complications were categorized
based on the Clavien–Dindo classification[12] and divided into three primary groups: overall complications, complications associated
with the first gastrostomy tube replacement, and those requiring urgent endoscopic,
laparoscopic, or laparotomy interventions.
For the statistical analysis, qualitative variables were analyzed using the chi-square
and, when necessary, the Fisher's test. For quantitative variables, nonparametric
tests were employed, as they did not meet the criteria for normal distribution.
The study was approved by the institutional ethics committee and complied with the
good clinical practice guidelines.
Pull-Technique Gastrostomy
P-PEG was performed by placing a single percutaneous puncture under endoscopic control,
according to the method described by Ponsky[2] ([Fig. 1]). The procedure involved the utilization of a MIC Percutaneous Endoscopic Gastrostomy KIT (AVANOS®; GA). A percutaneous guide was inserted into the stomach and subsequently
removed through the mouth using an endoscope. Next, a 14-Fr gastrostomy tube was attached
to the guide and pulled through the stomach in an anterograde fashion. The length
of the intracorporeal portion of the gastrostomy tube was then adjusted based on the
abdominal wall thickness of each patient.
Fig. 1 Steps to perform the P-PEG. (A) A guide is inserted percutaneously into the stomach and removed through the esophagus
assisted by endoscopy. (B) The gastrostomy tube is tied to the guide and the guide is pulled, allowing the
gastrostomy tube placement in an antegrade way. (C) Endoscopic P-PEG final appearance. (D) Extracorporeal P-PEG final appearance. P-PEG, pull-technique-percutaneous endoscopic
gastrostomy.
T-Fasteners Gastrostomy Technique
T-PEG was performed by placing three percutaneous T-fasteners under endoscopic control
([Fig. 2]). The procedure involved the utilization of an Introducer Kit for MIC-KEY Gastrostomy Feeding Tube (AVANOS®; GA). The fasteners were released from the needle tip using a stylet and
secured in place. A guide and an 18-Fr dilator were then inserted between the T-fasteners
using a modified Seldinger technique.[5] Subsequently, a 14-Fr balloon gastrostomy tube was retrogradely placed. The length
of the gastrostomy tube varied (12, 15, or 17 mm) depending on the abdominal wall
thickness of each patient. T-fasteners typically dislodge spontaneously during the
first postoperative month.
Fig. 2 Steps to perform the T-PEG. (A) The percutaneous T-fasteners are placed. (B) The guide and the dilator are inserted, allowing the gastrostomy tube placement
in a retrograde way. (C) Endoscopic T-PEG final appearance. (D) Extracorporeal T-PEG final appearance. T-PEG, T-fasteners percutaneous endoscopic
gastrostomy technique.
Gastrostomy Replacement
In both the P-PEG and T-PEG groups, the first gastrostomy replacement was conducted
6 months after the placement, and a MIC-KEY gastrostomy feeding tube was inserted
in all cases. During the replacement, the tube size varied according to the patient's
needs, with either a tube of the same size (14 Fr) or a larger one (16 Fr). In the
P-PEG group, the replacement was performed endoscopically or by traction under sedation;
in the T-PEG group, the balloon gastrostomy was replaced without sedation in the outpatient
clinic.
Results
One hundred and forty-six patients underwent PEG placement, 70 (48.0%) P-PEG and 76
(52.0%) T-PEG. The follow-up period for all patients in both groups was 1.5 years.
The mean weight at the time of surgery was 14.1 kg (standard deviation [SD] = 9.6)
and the mean age was 5.1 years (SD = 5.3).
The main indication for PEG placement was a diagnosed swallowing disorder. The most
frequent cause of swallowing disorder was neurological disease (62.3%), followed by
cardiopathy (8.9%) and severe respiratory disease (8.9%). The remaining 19.8% showed
a swallowing disorder related to various syndromic and genetic conditions.
No significant differences between both groups in terms of age, weight, sex, and associated
comorbidities were found (p > 0.05; [Table 1]).
Table 1
Demographics and surgical indications by groups
|
P-PEG (N = 70)
|
T-PEG (N = 76)
|
Significant difference (p-value)
|
|
Age (years)
|
4.8[a] (SD = 5.1)
|
5.4[a] (SD = 5.5)
|
0.87
|
|
Weight (kg)
|
12.6[a] (SD = 8.6)
|
15.6[a] (SD = 10.3)
|
0.27
|
|
Sex
|
|
Female
|
55.7%
|
47.4%
|
0.31
|
|
Male
|
44.3%
|
52.6%
|
0.31
|
|
Indication of PEG
|
|
Neurological disease
|
55.7%
|
68.4%
|
0.11
|
|
Cardiopathy
|
11.4%
|
6.7%
|
0.30
|
|
Respiratory disease
|
11.4%
|
6.7%
|
0.30
|
|
Other
|
21.4%
|
18.4%
|
0.65
|
Abbreviations: PEG, percutaneous endoscopic gastrostomy; P-PEG, “pull-technique” gastrostomy;
SD, standard deviation; T-PEG, T-fasteners gastrostomy technique.
Percentages are based on the number of study participants in respective groups.
a Values expressed in mean and SD.
All gastrostomies were assisted by endoscopy and laparoscopy was also performed in
31.5%. Of these, 25.3% underwent antireflux surgery during the same intervention (8.2%
in the P-PEG group and 17.1% in the T-PEG group).
Overall Complications
The total number of complications in our series was 33 (22,6%). Seventeen (24.2%)
complications were found in the P-PEG group and 16 (21.0%) in the T-PEG group (p > 0.05). All complications were documented within the initial 6 months following
gastrostomy placement.
The specific types of complications recorded in each group are shown in [Table 2]. The most frequent complication registered in the P-PEG group was buried bumper
with 8/70 cases (11.4%), whereas in the T-PEG group was granuloma related to the T-fastener
with 7/76 cases (9.2%; [Fig. 3A]). Other complications registered included infection, bleeding, device dislodgement,
early button extrusion, pneumoperitoneum, prolapse, and abdominal wall necrosis ([Fig. 3B]).
Table 2
Overall complications related to each type of percutaneous endoscopic gastrostomy
placement according to Clavien–Dindo
|
Overall complications
|
P-PEG (N = 70)
|
T-PEG (N = 76)
|
|
Type I
|
3 (4.3%)
|
6 (7.9%)
|
|
Granuloma
|
–
|
3 (3.9%)
|
|
Infection
|
1 (1.4%)
|
3 (3.9%)
|
|
Bleeding
|
2 (2.9%)
|
–
|
|
Type II
|
–
|
1 (1.3%)
|
|
Bleeding
|
–
|
1 (1.3%)
|
|
Type III
|
13 (18.6%)
|
9 (11.8%)
|
|
Buried bumper
|
8 (11.4%)
|
–
|
|
Granuloma
|
–
|
4 (5.3%)
|
|
Device dislodgement
|
2 (2.9%)
|
1 (1.3%)
|
|
Early button extrusion
|
–
|
2 (2.6%)
|
|
Pneumoperitoneum
|
2 (2.9%)
|
–
|
|
Abdominal wall necrosis
|
–
|
1 (1.3%)
|
|
Bleeding
|
–
|
1 (1.3%)
|
|
Prolapse
|
1 (1.4%)
|
–
|
|
Type IV
|
1 (1.4%)
|
–
|
|
Device dislodgement
|
1 (1.4%)
|
–
|
|
Total
|
17 (24.2%)
|
16 (21.0%)
|
Abbreviations: PEG, percutaneous endoscopic gastrostomy; P-PEG, “pull-technique” gastrostomy;
T-PEG, T-fasteners gastrostomy technique.
Percentages are based on the number of study participants in respective groups.
Fig. 3 Images of postoperative complications. (A) Granuloma related to the T-fastener in the T-PEG group. (B) Abdominal wall necrosis in the T-PEG group. T-PEG, T-fasteners percutaneous endoscopic
gastrostomy technique.
When comparing complications using the Clavien–Dindo classification system, it is
clinically significant to note that we observed a higher incidence of severe complications
in the P-PEG group, as indicated in [Table 2]. Type I and II complications were more frequent in the T-PEG group (9.2 vs. 4.3%),
whereas type III and IV complications were more common in the P-PEG group (20.0 vs.
11.8%). We emphasize that although these differences were notable, they did not reach
statistical significance (p = 0.24 and p = 0.17, respectively). Only one life-threatening complication (type IV) was observed
in our series, diagnosed in the P-PEG group: a tube dislodgement during the first
tube replacement led to peritonitis, septic shock, and urgent laparotomy. There were
no reported deaths related to PEG placement.
Among the complications, those intraoperative were minimal, with only one in each
group. In the P-PEG group, a Clavien–Dindo type III complication was documented during
gastrostomy placement: a pneumoperitoneum, necessitating laparoscopy to confirm the
absence of complications. In the T-PEG group, a Clavien–Dindo Type II complication
was recorded: an episode of intraoperative self-limited bleeding, which required a
blood transfusion.
Complications Related to First Gastrostomy Tube Replacement
As previously described, the primary difference between the two groups is the sedation
requirement for the first replacement, with no need for anesthesia in the T-PEG group.
Our study confirms that the P-PEG group had a significantly higher incidence of sedation
requirement for the first replacement compared with the T-PEG group (97 vs. 2.6% [p < 0.05]). Two cases in the T-PEG group necessitated sedation for tube replacement
due to the complication of early button extrusion, occurring at 1 and 4 months, respectively.
In both cases, sedation was necessary to facilitate the dilation of the stoma, enabling
gastrostomy replacement.
We explicitly analyzed complications related to the first tube replacement and found
a higher incidence in the P-PEG group (7.2 vs. 1.4%; p = 0.092). Although the difference was not statistically significant, it is worth
noting that in the T-PEG group, the only dislodgement was resolved endoscopically,
while in the P-PEG group, two out of five complications resulted in peritonitis and
required emergent laparotomy, as shown in [Table 3].
Table 3
Complications related to the first tube replacement conducted 6 months after its placement
in both groups
|
Complication related to
the first tube replacement
|
P-PEG (N = 70)
|
T-PEG (N = 76)
|
Treatment
|
|
Dislodgment with peritonitis
|
2
|
–
|
Laparotomy
|
|
Dislodgment without peritonitis
|
1
|
1
|
Endoscopy
|
|
Pneumoperitoneum
|
1
|
–
|
Endoscopy
|
|
Undiagnosed buried bumper[a]
|
1
|
–
|
Surgical gastrostomy site revision
|
|
Total
|
5 (7.2%)
|
1 (1.4%)
|
–
|
Abbreviations: PEG, percutaneous endoscopic gastrostomy; P-PEG, “pull-technique” gastrostomy;
T-PEG, T-fasteners gastrostomy technique.
Percentages are based on the number of study participants in respective groups.
a In this case, while removing the PEG, its intragastric part became lodged in the
abdominal wall as a result of an undetected buried bumper. To extract the fragment,
it was necessary to perform a surgical revision of the affected area.
Complications Treated with Urgent Surgical/Endoscopic Interventions
Patients in the P-PEG group needed more urgent interventions (either by endoscopy,
laparoscopy, or laparotomy) compared with the T-PEG (18.6 vs. 6.6% [p < 0.05]). The most frequent cause of an urgent intervention was a buried bumper,
found in eight patients in the P-PEG group and none in the T-PEG group (11.4 vs. 0%
[p < 0.05]).
Management of Complications Treated with Urgent Interventions
In the following section, we provide a detailed account of the management of complications
that necessitated urgent interventions in both groups.
In the P-PEG group, the majority of buried bumper cases (six out of eight) underwent
endoscopic repositioning. The two remaining cases of buried bumper necessitated a
surgical revision to extract the intragastric portion of the PEG and replace the gastrostomy.
Additional complications requiring urgent interventions in the P-PEG group included
one case of pneumoperitoneum immediately after PEG placement leading to exploratory
laparoscopy with no findings, two cases of gastrostomy dislodgement leading to peritonitis
requiring laparotomy to clean the abdominal cavity, and one case of dislodgement without
peritonitis requiring endoscopic replacement. It is noteworthy that all three previously
mentioned cases of gastrostomy dislodgement occurred during the first tube replacement.
In the T-PEG group, the following complications necessitating urgent interventions
were observed: one case of dislodgement without peritonitis (during the first replacement)
requiring endoscopic replacement, two cases of early tube dislodgement requiring endoscopic
replacement, one case of bleeding following PEG placement managed with cold saline
solution by endoscopy, and one case of wall necrosis in a patient diagnosed with microcephalic
primordial dwarfism, requiring surgical revision ([Fig. 3B]).
Discussion
The main advantage of the T-PEG technique, also called one-step PEG, is that it directly
inserts a balloon gastrostomy tube, allowing the gastrostomy replacement without the
need for sedation or general anesthesia.[2]
[4]
[5]
[6]
[7] Despite the increasing popularity of the T-PEG technique, there have been only a
few studies evaluating its safety in pediatric patients, as shown in [Table 4]. To date, only two studies have compared the incidence of complications between
P-PEG and T-PEG in children.[2]
[4] However, none of these sources offer recommendations regarding the utilization of
T-PEG versus P-PEG based on complications development.
Table 4
Summary of published studies on T-fasteners gastrostomy technique placement in pediatric
patients
|
Authors
|
Year
|
Number of patients
|
Comparative analysis with P-PEG
|
Age
(years)
|
Weight
(kg)
|
Follow-up
|
Major
complications
related to T-PEG[a]
|
|
Moreno Montero et al[6]
|
2013
|
8
|
No
|
4.6[b] (0.7–4.6)
|
12.4[b]
(5.5–36)
|
–
|
None
|
|
Kvello et al[8]
|
2010–2014
|
87
|
No
|
1.9[b] (0.2–16.4)
|
10.4[b]
(5.4–33.0)
|
1 month
|
14 (16.0%)
|
|
Jacob et al[2]
|
2007–2010
|
73
|
Yes (n = 55)
|
4.5[c]
(0.1–17)
|
13.6[c]
(2.4–57)
|
3 years
|
–
|
|
Göthberg and Björnsson[4]
|
2005–2012
|
206
|
Yes (n = 168)
|
3.9[c]
(± 4.7)
|
13.5[c]
(± 9.9)
|
6 months
|
2 (1.0%)
|
|
Dahlseng et al[9]
|
2017–2022
|
82
|
No
|
2.0[b]
(0.3–18.1)
|
9.9[b]
(3.5–51.5)
|
3.5 months
|
7 (8.5%)
|
|
Our case series
|
2010–2021
|
76
|
Yes (n = 70)
|
5.1[b]
(0.2–21.0)
|
14.1[c]
(3.0–55.0)
|
1.5 years
|
9 (11.8%)
|
Abbreviations: PEG, percutaneous endoscopic gastrostomy; P-PEG, “pull-technique” gastrostomy;
T-PEG, T-fasteners gastrostomy technique.
a Major complications are defined as Clavien–Dindo types III and IV.
b Values expressed in median and (range).
c Values expressed in mean and (range).
The complication rates recorded in our study closely resemble those documented in
the existing literature.[2]
[4]
[6]
[8]
[9] Major complications in previous T-PEG studies ranged from 1 to 16%, while in our
series, it was 11.8% ([Table 4]).
In our series, after comparing P-PEG and T-PEG, we did not find significant differences
in the total number of complications (24.2 vs. 21.0%). Although we observed no significant
differences in overall complication rates between the two groups, we did observe a
higher frequency of severe complications in the P-PEG group, although these differences
did not reach statistical significance. It is important to emphasize the relevance
of severe adverse events classified as Clavien–Dindo III/IV complications from a patient
safety perspective, as they involve invasive interventions or life-threatening events.
Furthermore, we specifically evaluated the need for an urgent endoscopy, laparoscopy,
or laparotomy to solve the complication. Globally, patients in the P-PEG group needed
more urgent interventions than the T-PEG.
The most frequent complication in the P-PEG group was buried bumper, whereas in the
T-PEG group was granuloma related to the T-fastener. A buried bumper occurs when the
external bumper of the PEG tube is tightly positioned against the abdominal wall,[13]
[14] and it typically requires endoscopic repositioning. On the other hand, the granuloma
formation around the T-fastener results from a local inflammatory reaction to a foreign
body.[9] Initially, we attempted surgical removal of the granulomas, but we found that this
approach was not always effective, as the T-fasteners often are difficult to locate.
Therefore, we now prefer to use topical silver nitrate, which can help to dissolve
the granulomas. Additionally, spontaneous ejection of the T-fasteners and resolution
of the granulomas are common over time.[4]
Our study did not identify any other T-fastener-related complications, such as bleeding
or subcutaneous migration.[7] However, the optimal timing for removing T-fasteners remains unclear. In our practice,
we have found that allowing the T-fasteners to fall out spontaneously, typically within
a few weeks of placement, is a safe and effective approach. In a retrospective study
involving 350 patients who underwent immediate removal of T-fasteners following gastrostomy,
it was found that the immediate removal of T-fasteners did not correlate with an increased
complication rate.[15] However, it is important to note that the previous study has a relatively short
follow-up period of 1 month. Although there is no consensus on the ideal timing for
removing the T-fasteners, several pediatric series routinely remove the sutures 3
to 4 weeks after the procedure.[4]
[8]
[9] We consider this timeframe to be a reasonable period for removing the T-fasteners.
Regarding first tube replacement, almost all P-PEG required sedation, whereas most
T-PEG were replaced as an outpatient procedure without sedation. Also, we found more
complications related to the first replacement in the P-PEG group. Although this difference
did not reach statistical significance, P-PEG complications during first tube replacement
were more severe and required more urgent laparotomies, while T-PEG complications
were solved endoscopically. This underscores the importance of conducting follow-up
until at least the first tube replacement, as complications can occur. It is important
to consider that most referenced articles do not include the first tube replacement
in their follow-up period, a factor that warrants consideration when interpreting
their findings.[4]
[6]
[8]
[9]
Some authors have reported favorable outcomes with laparoscopic gastrostomy using
transcutaneous U-Stitch fixation.[16] In our series, laparoscopy was employed in selected cases, to ensure greater intraoperative
safety. All PEG procedures were conducted using an endoscopic percutaneous approach,
regardless of whether laparoscopy was employed simultaneously. While laparoscopic
gastrostomy using U-Stitch fixation is described with good results, it is not our
center's preferred technique due to its perceived time-consuming aspect.
It is important to mention that initially, the authors utilized P-PEG for PEG placement
between 2010 and 2016. However, from 2016 onward, they shifted their preference to
T-PEG as the primary technique. Despite the new implementation of T-PEG in 2016, we
observed that T-PEG was associated with fewer complications requiring urgent interventions
compared with P-PEG. The fact that we encountered fewer complications with a new technique
suggests that the learning curve for T-PEG is notably brief. However, it is crucial
to highlight the importance of following the correct technique, ideally performed
by a skilled surgical team.
This study presents some limitations. First, as a retrospective study, it is subject
to recall bias, where participants may remember past events inaccurately, leading
to potential inaccuracies in the data analysis. Second, neonatal patients weighing
less than 3 kg were excluded from the study due to the potential risks associated
with using the T-fastener technique in this population. Specifically, the narrow abdominal
surface of these patients may not allow for the placement of three anchors, and the
use of a long dilator could potentially injure the posterior wall of the stomach.[4] Lastly, this study lacks analysis of overall patient outcomes beyond complications,
such as recovery times or the impact on patient quality of life, which could be interesting
for future investigations. However, despite these limitations, our study provides
valuable information on the complications of these two widely used procedures.
Conclusion
In our experience, P-PEG is associated with more severe complications, particularly
during the first button replacement. Compared with T-PEG, P-PEG is associated with
a higher incidence of sedation requirements and a greater need for urgent endoscopic
and surgical interventions to treat complications. Therefore, based on our findings,
we advocate for the use of T-PEG in pediatric patients. P-PEG may be preferred in
neonatal patients, while T-PEG could be favored in the remaining pediatric population.
Further research is warranted to investigate the long-term outcomes of both P-PEG
and T-PEG.
