Percutaneous Endoscopic Gastrostomy in Children : Experience from Single Center of a Developing Country

Address for correspondence: Dr. Mohit Kehar Registrar, Division of Paediatric Gastroenterology and Hepatology, Institute of Child Health, Sir Ganga Ram Hospital, Delhi. E-mail: mohitkehar86@yahoo.co.in Tel No: +91 9810266452 1Dr. Mohit Kehar, MBBS, DNB Paediatrics, Fellowship in Paediatric Gastroenterology and Hepatology, Registrar, 2Dr. Satyam Upadhyay, MBBS, MD, Fellow division of Paediatric Gastroenterology and Hepatology, 3Dr. Nishant Wadhwa, MBBS, DCH, DNB, Consultant and Chief. All from the Division of Paediatric Gastroenterology and Hepatology, Institute of Child Health, Sir Ganga Ram Hospital, Delhi. Abstract


Introduction
N utri on is of paramount importance for adequate growth and development of a child.Enteral nutri on is the ideal mode of nutrient delivery for children 1,2 The various routes of providing enteral nutri on to a paediatric pa ent are by nasogastric tube inser on, nasojejeunal tube inser on, gastrostomy which can be placed surgically or endoscopically.Nasogastric tube feeding is most o en used for short-term periods, however, there are several limita ons for its long-term use including nasal discomfort, blockage or displacement of the tube, irrita on or penetra on of the larynx and recurrent pulmonary aspira ons 3 .
Gastrostomy has various advantages over nasogastric tube as there is less chance of blocking and tube displacement and these tubes donot require frequent revisions.Gastrostomy tubes can be placed surgically or endoscopically.The superiority of percutaneously placed gastrostomies compared to former surgical gastrostomy procedures (i.e.Witzel, Stamm, Janeway techniqueues) has been shown clearly in many clinical studies 4,5 .For these obvious reasons, percutaneous endoscopic gastrotomy (PEG) is now the preferred op on for providing medium and longterm nutri onal support in pa ents with impairment of feeding abili es leading to under nutri on and its complica ons.
PEG was fi rst introduced by Gauderer in 1980, the fi rst PEG inser on was performed in University hospital of Cleveland, USA, on a four and-half monthold baby 6 .In addi on to improving the nutri on and growth of subjects, the use of the PEG can signifi cantly reduce feeding me and ease drug administra on.There has been a consistent improvement in the social func oning, mental, general health percep on, and quality of life of caregivers in prospec ve cohort studies 7 .It has been found in various studies that the nutri onal status of unwell children is a common cause of anxiety for parents and feeding mes can be stressful 8 .
The impact of PEG feeding is posi ve with many parents repor ng a high level of sa sfac on and wishing the interven on to have taken place earlier 9,10 .Nutri onal support with the use of the PEG has been demonstrated in children with neurodisability 10,11 , cys c fi brosis 12,13 , neonatal pulmonary disease 14 , congenital heart disease (CHD) 15,16 , Crohn disease 17 , oncological condi ons 18 , metabolic disease, gene c chromosomal, and degenera ve diseases 19 .PEG is widely being used for paediatric pa ents in developed countries for various indica ons.
The aim of our study were to review cases with percutaneous endoscopic gastrostomy (PEG) procedure and to review pa ent characteris cs, indica ons, and complica ons and outcome of PEG tube inser on in children at our center.

Materials and Methods
The study was a prospec ve study carried out in one of the ter ary care hospitals, Sir Ganga Ram Hospital, Division of Paediatric Gastroenterology and Hepatology, New Delhi, India for a period of two years from August 2010 to August 2012.
All pa ents who underwent PEG tube placement and have had at least one year of follow up were enrolled for the study a er informed consent from parents/guardians of the pa ents.As exclusion criteria all pa ents who had less than one year of follow up a er PEG placement were excluded from the study.Pa ents with acute systemic illness, with deranged coagula on profi le or/ and thrombocytopenia, were also excluded for PEG placement.
The study was approved from ethical commi ee as per hospital research commi ee protocol.Pa ent demographics, principal diagnosis and indica on for PEG placement, length of hospital stay a er PEG and post procedure complica ons were recorded for all the pa ents from the medical charts.The pa ents were followed up and weight gain and height gain was periodically checked a er 3 months, 6 months and 12 months of PEG placement.
PEG placements were performed under moderate seda on using Midazolam and Ketamine in the endoscopy suite by a team consis ng of a pediatric gastroenterologist and accompanying team.Standard ''pull technique" was used for all PEG placements.As per protocol all pa ents underwent a pre procedure work up consis ng of a complete hemogram and coagula on profi le.
All pa ents received a preopera ve single dose of an bio c a third genera on cephalosporin, immediately before PEG placement.Kimberley clark / Freka PEG tubes were used of size 14-24 F depending on the age and weight of the pa ent.In pa ents aged 0-1 yr 14 Fn tube was used, in 1-6 yrs 20Fn tube and in older than 6 yrs 24 Fn tube was used to provide enteral nutri on support.A er the PEG tube placement, feeds were ini ated a er 4-6 hrs of procedure and parents were explained the feeding process and handling of the tube.
All pa ents were started on measure to reduce gastroesophageal refl ux like head end eleva on and proton pump inhibitors for at least three months.A er this they were con nued on posi onal measures and PPI were stopped and the pa ents were followed up for symptoms and signs of refl ux.
Permanent PEG removals were performed when children no longer required the PEG for feeding support as decided following a detailed mul disciplinary assessment which includes a detailed dietary history assessing the op mum calories, macronutrients and drugs the pa ent is having orally along with extend of the recovery of primary illness for which PEG tube was placed.PEG removal was performed endoscopically under seda on.Whenever long term support was required PEG tube was replaced with a Balloon replacement tube (BRT) or Mickey bu on a er 3-5 months of primary tube inser on.Tube was replaced because of wear and tear in the tube; it was either blocked or disfi gured so it was replaced with a replacement tube.
The pa ents were divided in four groups according to the age of pa ent ie 0-1 yr, 1.1-5 yr, 5.1-10 year and 10.1-16 years and the follow up weight and height was compared in all the four groups.Data analysis between groups has been compared using ANOVA and within groups across follow-ups has been done using paired t-test.SPSS 15.0.

Results
Forty six PEG tube placements were performed during the study period.Amongst the pa ent, 30 were male and 16 were females.The median age of pa ents at me of PEG inser on was 6.75 years (range is 5 weeks -16 years) and mean weight was 10.3 kg (range 2.4 kg -40 kg).The commonest indica on for PEG inser on were Cerebral palsy (CP) (47.8 %) and failure to thrive with feeding diffi culty and recurrent aspira ons (13%) followed by congenital heart disease requiring adequate weight gain (8.6%).(Table 1) Three pa ents had head injury with diff use neuronal involvement (6.5%), two pa ents were suff ering from tubercular meningi s with hydrocephalus (4.3%).There were two pa ents each suff ering from polytrauma and intraventricular bleeding (4.3%).There were two pa ents of medulloblastoma (4.3%), and one pa ent each of Re s syndrome, Duchenne muscular dystrophy, and Gullein Barre syndrome.(2.1% each) The youngest pa ent was a fi ve week old male child who was suff ering from retrognathia with cle palate with feeding diffi culty with recurrent aspira ons whose weight at the me of PEG tube placement was 2.4 kg.Our series includes ten infants, the oldest being nine month old at the me of PEG inser on.
The me for PEG placement was SD 14.39+_4.14range 10-20 minute).The dura on of hospital stay following PEG inser on was one day with a range of one to four days.No pa ent remained in hospital beyond this me having had a PEG as the sole procedure Erythema at the PEG inser on site was the most common complica on post procedure (10/46, 21 %).In all instances this was treated with oral an bio cs for 5 days with no progression.Two pa ents had refl ux of feeds from the PEG tube which improved a er giving laxa ves as both these pa ent were suff ering from chronic cons pa on.No procedure-related mortality was no ced.
Thirty six children underwent PEG removal during the study period.These are pa ents in whom either PEG tube was changed with a BRT or in whom it was not required anymore so it was removed permanently.In twenty six pa ents, the PEG was subs tuted with BRT/mickey bu on, whereas in ten children the PEG was removed because it was no longer required for feeding support.The average dura on between the PEG inser ons to BRT conversion was 4.7 months.
Over the study period ten pa ents had the PEG permanently removed because it was no longer required for feeding support (average dura on a er PEG inser on 1.3 years).Out of these ten pa ents, two were suff ering from tubercular meningi s, two with head injury, two with intracranial bleeding, two pa ent of polytrauma, one pa ent each of medulloblastoma and Gullein Barre syndrome.No pa ent had PEG removal before one year of tube placement.
All PEG tube removals were done endoscopically under seda on with no complica ons.Trac on technique for tube removal was not used in any of the pa ent as the tubes used in our unit were not trac on removable.In follow up out of forty six pa ents in whom PEG tube was placed, ten pa ents lost to our follow up and rest thirty six children were under follow up for atleast one year during which there weight and height were measured.The ten pa ents who were lost to follow up included four pa ents with cerebral palsy, two pa ents with feeding diffi culty with failure to thrive, one pa ent suff ering from head injury, one pa ent with patent ductus arteriosus with severe failure to thrive, one pa ent of medulloblastoma and one pa ent of Re syndrome.
Three pa ents had an increase in amount of refl ux following PEG tube placement, which manifested as recurrent apneic a ack in one pa ent who was admi ed in neonatal intensive care unit suff ering from hypoxic ischemic encepathalopathy sequel with cle lip and palate.The other two pa ents had repeated vomi ng and aspira on pneumonia.For these pa ents a feeding tube was guided endoscopically through the PEG tube feeding channel into the jejunum under fl uoroscopy (Jejunal tube percutaneous endoscopic gastrotomy /JET -PEG).Post placement of JET-PEG, there was consistent weight gain and improvement in refl ux symptoms in these pa ents.There were thirty six pa ents who were under follow up for one year.A er three month of PEG tube placement average weight gain no ced was 1.3 kg and height gain of 1.6 cm.A er six months the average weight gain was 2.8 kg with height gain of 2.5 cm, a er twelve average weight gain was 4.2 kg and height gain of 4.13 cm (Fig 1).
As the growth velocity is diff erent in diff erent age group of pa ents, four age group were formed and the average weight and height of all these 4 groups were compared.
In all the four groups there was consistent height and weight gain at 3, 6 and 12 months with p <0.05 in all groups when there average weight and height was compared with their presenta on parameters (Table 2  and 3).
In age group 0-1 yr the percentage improvement in height and weight at 3, 6 and 12 months was best amongst the en re four groups owing to high growth velocity in this age group of pa ents.(Figure 1

Discussion
Gastrostomy tubes were placed exclusively by laparotomy un l 1980, when Gauderer showed that the PEG technique was more cost-eff ec ve and safer than surgical gastrostomy 6 .PEG technique is be er because it avoids the morbidity associated with laparotomy, causes less incisional pain, has a quicker recovery period, and can be performed more rapidly, with the average PEG taking less than 15 minutes 20 .
Neurodisabilty was the main indica on for PEG inser on in our experience.Cerebral palsy was the single most important indica on for PEG inser on.Craig et al have reported PEG experience in a North London cohort where the predominant indica on for inser on of PEG was CP followed by gene c syndromes, metabolic syndromes, and progressive degenera ve disorder 21 .Feeding diffi culty was the main indica on for PEG inser on in a South African series 22 , whereas neuromuscular and metabolic causes 23 and faltering growth 24 were the most important indica on in other studies.
There was consistent weight and height gain in all age group pa ents a er PEG inser on with best percentage was seen in pa ents aged less than one year owing to their normal high growth velocity.
In our study, no major complica on was seen and there were no procedure related mortality.Only minor complica on were seen which was erythema of the skin, similar complica ons were also reported by a group from Liverpool, United Kingdom 25 .
In our series following 46 new PEG tube placements three pa ent (6.5%) had increase in amount of refl ux manifes ng as apnea, vomi ng and aspira on pneumonia.In all these cases a JET PEG was done following which there was improvement in symptoms and growth velocity.There have been confl ic ng studies on the risk of gastroesophageal refl ux a er PEG tube placement however, the role of PEG as a cause of new onset gastoesophageal refl ux in children remains controversial 26,27 .We have not studied this parameter systemically, but as our unit protocol, all pa ents following PEG tube placement are placed on measures to prevent refl ux.
The dura on for PEG tube feeding depends upon the indica on for which it was placed; pa ents suff ering from cerebral palsy require long term support for op mal nutri on.Pa ents suff ering from GBS, polytrauma, TBM require short term support owing to recovery of their neurological insult.

Conclusion
There is an increasing demand for PEGs in management of chronic pediatric ailments who have nutri onal challenge and when feeding problems become a hindrance in growth and development.There has been a consistent weight gain in the pa ents on PEG tube, as calorie dense feeds can be given easily which would otherwise not be possible with a nasogastric tubes for prolonged periods.Our complica on rates compare favorably with those reported in other series with minor complica ons like local skin infec ons easily amenable to treatment.
Fig 1: Percentage weight improvement in diff erent age groups at 3, 6 and 12 months a er PEG inser on Fig 2: Percentage height improvement in diff erent age groups at 3, 6 and 12 months a er PEG inser on

Table 2 :
Mean weight 3, 6 and 12 months with p-values of diff erent age groups