Two independent investigators performed the literature search, using PubMed, EMBASE, and Ovid database. The search terms were “laparoscopy”, “surgery”, “children”, and ”Inflammatory Bowel Disease”, or “Ulcerative colitis”, “Crohn Disease”, and “Indeterminate Colitis”. Inclusion criteria were: (1) paper fully written in English language; and (2) patients younger than 18 years of age. All prospective, observational, and retrospective studies were included. Case reports were excluded.

Data were extracted from articles that fulfilled inclusion and exclusion criteria and entered into tables. These data included first Author, country of origin, years of data collection, series size, PIBD types, and incidence of complications.

The medical records of all patients affected by PIBD (CD, UC, IBD-U) from January 2006 to December 2014 who underwent minimally invasive surgery (laparoscopic or laparoscopic-assisted) at Giannina Gaslini Children Hospital were reviewed. We recorded demographic data, type of procedure performed (single or staged), operating time, morbidity and length of hospital stay from our centralized operating room database.

Descriptive statistics were reported as absolute frequencies and percentages for qualitative data, mean ± SD or median and range (based on variability) were used to describe quantitative variables. Differences in the frequencies of each variable were evaluated by the χ² test, or by Fisher’s exact test, when appropriate. All the statistical tests were two sided and a P value lower than 0.05 was considered as statistically significant.

On the basis of the available literature data, we could address minimally invasive surgery for PIBDs focusing on diagnostic laparoscopy before 2000 and on laparoscopic treatment, afterwards. Furthermore, therapeutic laparoscopy can be divided basing on the adopted surgical procedure[8-26].

The first report regarding diagnostic laparoscopy dates back to 1975 when Leape et al[27] reported the first series of children and infants undergoing diagnostic laparoscopy for various issues, including CD. Diagnostic laparoscopy was used to determine what surgical treatment was eventually required with a conventional laparotomy. Later on, in 1996, Miller and colleagues reported the use of minimally invasive surgery to detect the presence of abnormal mesenteric fat (“creeping fat”) in patients with suspected CD. The Authors described 6 children who underwent diagnostic laparoscopy. Three of them were suspected of having CD and underwent resection confirming the diagnosis[28]. Similarly, in 1998, Schier et al[29] reported a series of 11 children who underwent diagnostic laparoscopy without major complication. The Authors confirmed the usefulness of direct images in the early stages of CD for a better implementation of adequate medical treatments.

Laparoscopy has been adopted in adults since early 90s’ either for the treatment of UC or CD with good results[30,31]. In children, therapeutic laparoscopy for IBDs has been introduced since early 2000s’. The first report of laparoscopic treatment of CD dates back to 2002, when Rothemberg reported his preliminary experience with his first 15 segmental bowel resections for CD[10].During the same year, Georgeson reported his preliminary experience with 18 patients with UC who underwent Laparoscopic assisted total colectomy with pouch reconstruction[9]. Both authors reported their preliminary experiences on relatively small pediatric series demonstrating the feasibility of minimally invasive surgery for the treatment of IBDs in children, either CD or UC. Similarly, Proctor in 2002 and Dutta in 2003 reinforced these considerations in two separate reports[8,12]. In particular, Proctor and colleagues reported a comparative retrospective analysis of the results of open vs laparoscopic subtotal colectomy for UC, treated in their institution between 1999 and 2001. The authors concluded that laparoscopy requires longer surgery but better cosmetic results with shorter return to normal activities and bowel function, being the incidence of major complications unaffected by the chosen approach. Furthermore, the authors underlined how length of surgery may be improved during the learning curve, as previously reported in adult literature[8].

Up to present years, a number of publications reported extensive use of laparoscopy for the treatment of PIBD and larger series have been reported. So far, the largest series published in international literature is that by Diamond and colleagues who reported a series of 136 patients who underwent 154 laparoscopic procedures for PIBD in the period between 1999 and 2007. The authors reported improved cosmetic results, a reduced hospitalisation but an incidence of major complications comparable to that of open surgery, particularly intestinal obstruction[18] (Table 1).

Small bowel and/or ileocolic resection: Either laparoscopically-assisted or total intracorporeal ileo-colic resection and segmental ileal resection have been reported. Those procedures have been adopted in patients with CD and consist of laparoscopy with 3 to 4 ports. The whole bowel is inspected and the site of resection is determined by preoperative imaging matched to intraoperative evidences. The resection-anastomosis can be accomplished via a mini-laparotomy (either extending umbilical port incision or by means of a small modified Pfannenstiel incision) through which the bowel is exteriorized, resected and anatomised[13,18]. Alternatively, total intracorporeal resection and anastomosis can be performed as described by Rothemberg and Dutta in their previous reports[10,12].

Although some Authors suggested to resort to the “safer” extracorporeal anastomoses (laparoscopic-assisted approach) due to the inflamed and fragile bowel to be anastomized with staplers[13,14], both alternatives have proved to be safe and effective in experienced hands and are now used worldwide in CD.

Strictureplasty: The Heineke-Mikulicz strictureplasty can be performed with a minimally invasive approach, either totally intracorporeal or laparoscopic-assisted. Although mostly limited to upper gastrointestinal tract or to multisite CD involvement for intestinal preservation, in paediatric settings this procedure can be accomplished with good results. Of note, Romeo and colleagues demonstrated that the recurrence rate of strictureplasty overlaps that of resection-anastomosis, thus making this technique a valid option for selected CD patients at risk for short bowel syndrome following resection[32].

Total or subtotal colectomy: Either in elective or emergency setting, total or subtotal colectomy can be carried out with results that overlap and/or overcome those of conventional open surgery[18]. This procedure has been mostly used for the treatment of UC or IBD-U (occasionally for the treatment of pancolonic CD) and consists of a 4 to 5 ports laparoscopy. The left colon is approached first and divided close to peritoneal reflection with straight or angled stapling devices. Mesentery is divided using available laparoscopic sealing devices (Ligasure^® in our experience) and colectomy is carried on in anticlockwise direction. Particular care must be taken when dividing midcolonic vessels in order to avoid lesions of first jejunal loop at Treitz ligament. Similarly, another key-point of colectomy is the right hepatic flexure when surgeons must spare and protect the duodenum and hepatic hilum. Once colonic isolation and resection reaches the caecum, the whole colon can be extracted through the right iliac fossa port and the same wound can be used to fashion the ileostomy. Alternatively, the colon can be extracted by everting the rectum through the anus and stapling the rectum from outside. This alternative approach can turn useful in the elective setting in case of small children, particularly those with IBD-U.

Straight or J-pouch reconstruction? Although a debate regarding the indication to fashion a pouch during reconstruction after colectomy in PIBD exists, most surgeons resort to this technique. A straight ileo-anal or ileo-rectal anastomosis is used only by a limited number of Authors. In 2006, Tilney and colleagues performed a meta-analysis demonstrating that, though basing on a very few good-quality studies, pouch procedures should be preferred in order to achieve better survival and functional outcome[33]. On the basis of these considerations, although pouch implies frequent endoscopic follow up and a relatively high incidence of pouchitis, (reported in up to 50% of patients[34-36]), pouch procedures represent at the moment the gold standard for reconstruction after colectomy in children with PIBD.

J-pouch ileo-rectal or ileo-anal anastomosis? This is one of the most controversial topics in the surgical treatment of PIBD and represents a key element in the radical treatment of UC/IBD-U and of pancolonic CD. Some surgeons opt for a mucosectomy, endorectal pull-through and J-pouch ileo-anal anastomosis[9]. Others do prefer subtotal proctocolectomy and J-pouch ileo-rectal anastomosis 2-3 cm above the dentate line. The most relevant issues leading surgeons’ choice are: (1) relatively high risk of leaving small islands of rectal diseased mucosa in the rectal cuff, in case of endorectal dissection; and (2) relatively high risk of dysplasia and cancer in the 2-3 residual centimetres of rectum left in place in case of J-pouch ileo-rectal anastomosis[9,37]. Regardless of the choice for reconstruction, both options are amenable of laparoscopic approach. The mesentery of the terminal ileum is widened and lengthened and a J-pouch is created through the site of the ileostomy. J-pouch length varies between short pouches[38] and longer ones[18] according to surgeons. The pouch is then returned into the abdomen and the rectal pouch is dissected with a sealing device (i.e., Ligasure^®) down to the elevator ani and stapled. A circular stapling device is then inserted into the anus and the ileal-pouch is anastomized 2-3 cm above the dentate line. Alternatively, the endorectal dissection can be accomplished starting 1-2 mm above the dentate line in order to perform a classic endorectal pull-through, as described for Hirschsprung’s disease[39].

Single or staged procedures: Total colectomy and J-pouch reconstruction can be accomplished as a single stage (Total colectomy with J-pouch reconstruction and no protective ileostomy), two-stage (Total colectomy with ileostomy followed by J-pouch reconstruction without protective ileostomy) or three-stage procedure (Total colectomy with ileostomy followed by J-pouch reconstruction and protective ileostomy) depending on surgeons’ attitude and on patients’ general conditions. Protective ileostomy is chosen by most surgeons[8,9,18,38]. Nonetheless, as complications have been frequently related to the ileostomy (i.e., internal hernia, prolapse, adhesion, twisting)[18,19] most surgeons addressed this issue and questioned whether the routine use of protective ileostomy should be abandoned in favour of a strict selection of patients with the highest risk of anastomotic complications (fulminant colitis? very low body mass index? low albumin levels?). Anyway, a three-stage approach is recommended in all emergent situation, i.e., fulminant colitis, patients on high dose steroids, severe malnutrition and IBD-U.

Further technical improvements: Recently, SILS^TM devices to perform single incision laparoscopic surgery have been adopted to further contain the trauma of abdominal wall and to improve the outcome of the patients both in terms of reduced pain, shorter postoperative stay, earlier recovery of normal bowel functions and improved cosmetic appearance[25,38]. Finally, robotic surgery is on its way to be applied to J-pouch ileo-rectal or ileo-anal anastomosis, given the possibility to apply this innovative technologies to rectal pouch dissection in order to further minimize the risk of damaging perirectal structures (personal unreported experience).

Comparing open and laparoscopic surgery for PIBD it comes clear that the incidence of complications does not significantly differ[8,12,35]. Nonetheless, surgery for PIBD is somehow frustrating given the relatively high incidence of surgical problems that can occur in the postoperative period. In fact, complications requiring surgical intervention occurred with an average incidence of 29% (158 out of overall 541 pediatric patients in this review) (Table 1). This percentage is similar or even lower than that reported for open surgery. The most frequent issues are represented by intestinal obstruction, anastomotic leakage or stenosis, pouchitis and faecal incontinence[8,18,19,34,40-44].

In particular, the incidence of complications approaches 55% for UC, being that of intestinal obstruction of around 25% and that of pouchitis of nearly 50%[34]. Similarly, the incidence of complications following surgery for CD can be as high as 33% with both early and late complications reported in pediatric patients[40,41]. This should confirm a higher likelihood of complications for patients with UC.

Functional outcome is acceptable with a wide variability of outcomes in different literature reports. Stavlo et al[42] has reported normal continence in 100% of patients in 2003 and Wewer in less than 50% of patients in 2005[43]. This wide range of results is difficult to explain. Though, most surgeons report a relatively high incidence of soiling, urge incontinence and night-time faecal continence issues in the long term[18,19,42-44]. All these issues must be acknowledged to families approaching surgery for PIBD in order to achieve a better education and participation in the long term care of their relatives.

Since no effective and definitive therapeutic option have been identified for the treatment of perianal disease in children with CD but the need for reduced trauma and sphincters preservation[45], the issue of perianal abscesses and fistula remains difficult to deal with and extremely troublesome both for the patient and the care-giver. In 2013 the World Congress of Gastroenterology implemented shared guidelines and therapeutic options for the medical and surgical management of perianal fistulising CD. The most important aspect is to appropriately select patients for surgery. On this regards, it is widely accepted that surgery for fistulising CD should be used only following complete mucosal healing and no active disease[46]. Those patients with perianal fistulising CD and healthy rectal mucosa are amenable of various surgical options, namely fistulotomy, biological plugs, fibrin glue, advancement flaps, fistula resections (including the so-called “cone-like” resection), stem cells and gracilis muscle interposition[46]. Only some of them can be considered as truly minimally invasive, namely biological plugs, fibrin glue and stem cells based therapy. In particular, the promising fibrin glue treatment showed results similar to those obtained with other established surgical treatment[47]. Although stem cells based therapy proved to be similarly effective, its healing rate of roughly 80% showed a dramatic decrease to nearly 30% over time[46].

Recently, Meinero et al[45] described an innovative and minimally invasive technique for the treatment of anal fistula, named Video Assisted Anal Fistula Treatment. Though based on a small series of patients, Schwandner[48] showed the feasibility of this approach in fistulising CD and the possibility to treat transphincteric, suprashpincteric and rectovaginal fistulae with little to no complications and minimal discomfort. Although results are based on only 11 patients, over 80% success rate and absent continence deterioration are promising aspects for this innovative technique.