In 1959, Kasai and Suzuki described 12 cases of biliary atresia (BA) treated at Sendai in Japan over the previous 8 years. The first six were treated by his chief, Prof. Katsoura and the remainder by Kasai, still very much a junior member of the team and only 37 years old.1 Only 3 actually had any form of long-term survival, one ultimately dying of cholangiocarcinoma at the age of 63 years.2 So, what was so radical? Well, these cases were described using the pathological epithet “uncorrectable” and, almost certainly, was a sentence of death from end-stage liver failure a few months later. Four years following publication of this landmark, in Denver CO, and completely unaware of each other, a team led by a young surgeon, also co-incidentally 37 years old, Thomas Starzl, transplanted a liver into a child born with BA. They showed that the surgery was technically possible though the patient died on-table of uncontrolled bleeding. Nonetheless, from such early roots emerged a surgical strategy which has been the mainstay of treatment for BA for about 30 years.

There are many aetiological variants of BA, united by an obstructive biliary tract and a propensity to cholestatic liver injury, fibrosis and ultimately cirrhosis.3 The former is irreversible, the later quixotic at presentation but progressive and in only one direction.

The aim of this chapter is two-fold; to summarise our attempts at prognosis at the time of presentation and to present the latest evidence-based adjuvant strategies to maximise outcomes of the Kasai portoenterostomy (KPE).

There has long been an association between early diagnosis of BA and improved prognosis, not least due to earlier intervention in the form of KPE.4 However, the mainstay diagnostic investigations remain invasive and fail to accurately predict which infants will clear their jaundice and will not require a liver transplantation (LT) post-KPE. We describe below, and summarise in Table 1, ongoing efforts to accurately diagnose BA and provide insight in the clinical outcomes post-KPE.

APRi has the virtue of being an off-the-shelf laboratory value, and is readily calculated. The prognostic significance of biochemical values in BA both before and after KPE has been the subject of a number of studies. In our own series from the 1990s, presented at an American monothematic conference in 2002, we identified the components (low serum aspartate aminotransferase (AST) and high platelet count) as independent markers of prognosis.5 With worsening liver disease relative thrombocytopaenia occurs as a result of not only increased sequestration of platelets by an enlarging spleen, but also perhaps by decreased thrombopoietin production by the liver.6 Soon after, the AST-to-platelet ratio index (APRi) was introduced as a surrogate of liver fibrosis, initially with adult chronic hepatitis C patients.7 It has since been validated in children.8,9 Pre-KPE APRi has been shown to be indicative of the state of liver at presentation.10 Indeed, older infants with BA pre-KPE tend to have a higher APRi, reflecting a more progressed liver disease.10 In a recent series we presented 473 infants who underwent KPE and subsequently followed up for up to 10 years.11 Their pre-KPE APRi was stratified into quartiles and clinical outcomes analysed. Statistically significant differences of 5-year native liver survival (NLS) and clearance of jaundice (COJ) were identified between these APRi quartiles, with the highest quartile having the worst 5 year NLS and COJ rates at 34 % and 49 % respectively, and the lowest quartile the best, at 54 % and 67 % respectively (Fig. 1).

Though often described as a surrogate of liver fibrosis there have been only inconsistent correlations between APRi and histological liver fibrosis scores (Metavir or Ishak), probably owing to the limited nature of these histological grading systems.8,12 There has, however, been a correlation between APRi and degree of portal α-SMA and CD34+ expression, markers of myofibroblast activity and neovascularisation,13 which is evidence of ongoing fibrosis.

Matrix Metalloproteinases (MMP), produced by macrophages and connective tissue cells, are a group of enzymes responsible for remodelling the extracellular matrix (ECM). MMP-7 has been found to be expressed as a result of inflammation and activation of the β-catenin pathway.14 As a result, it degrades several building blocks of the ECM such as type IV collagen, laminin and elastin.14 MMP-7 has been implicated in the pathogenesis of BA, and its serum levels investigated not only for diagnosing BA, but also for predicting the clinical outcome post-KPE.

Its diagnostic value is beginning to be appreciated but is still controversial, not the least because of variably quoted cut-off levels, ranging from 1.9 ng/ml15 to 69 ng/ml16. The sensitivity and specificity of this test has been reported as high as 96 %17 and 93 %18 respectively. MMP-7 appears to have a fundamental role in the pathogenesis of the secondary liver fibrosis in BA, and measuring its serum concentration may offer a window into the degree of liver fibrosis, and its influence on native liver survival. Pre-KPE MMP-7 levels tend to correlate with the degree of liver fibrosis,19 and thus by the ongoing remodeling activity. It is therefore a reasonable hypothesis that higher pre-KPE MMP-7 levels predict worse outcomes post-KPE. In our own experience, a correlation between serum MMP-7 levels and liver fibrosis was noted at KPE, though statistical significance between MMP-7 levels and clinical outcomes, such as COJ and NLS, was not reached.16 Interestingly, Jiang et al. identified a subset of 40 out of 329 BA infants with relatively low MMP-7 (≤25 ng/ml).20 This group also had a significantly lower GGT than their high MMP-7 (>25 ng/ml) counterparts. As expected, their liver fibrosis score assessed at histology was significantly better than the group with high MMP-7 levels. However, their 3 and 6-month COJ rates were significantly worse (29.7 % vs. 46.9 %; P = 0.05, and 32.1 % vs. 54.5 %, P = 0.03 respectively). The same research group had previously identified a similar correlation with low GGT levels pre-KPE.21 It is not clear why infants with better liver fibrosis scores have worse outcomes post-KPE, though perhaps the unusually low MMP-7 levels give us some insight on the reduced remodeling capacity of the livers of these infants.

There is progressive liver fibrosis pre-KPE with the passage of time. Wu et al. have suggested to correct for the age of the infant at the point of MMP-7 measurement.22 Comparing 50 infants with BA with 120 non-BA infants, the age-adjusted MMP-7 levels were significantly higher (P < 0.001) in BA. On ROC analysis a cut-off value of 0.1 ng/ml/day-of-life returned a sensitivity, specificity, PPV and NPV of 88 %, 82 %, 66 % and 94 % respectively. This implies the relative specificity of MMP-7 for BA over other causes of cholestasis and liver disease in infants.

Serum concentration of MMP-7 may have prognostic value post-KPE. Chi et al. followed up 191 patients with BA for longer than 2 years with regular measurements of serum MMP-7.23 Four distinct patterns of MMP-7 concentration were identified and matched with clinical outcome. A decline of MMP-7 levels post-KPE had more favorable outcome (2-year NLS of 79 % - 100 %) than fluctuating, static or rising MMP-7 levels, even in the absence of jaundice (2-year NLS 6 % - 31 %).

Secretin receptors (SCTR) are usually expressed by mature cholangiocytes, rather than hepatocytes.24 These act as transmembrane receptors which bind to secretin. The subsequent intracellular mechanism ensures regulation of the secretion of bicarbonate, protecting bile duct epithelium from bile acids.25 A collaborative study between Helsinki and King's College Hospital investigated the link between SCTR expression and clinical outcome in 50 infants at KPE and 22 infants during follow-up post-KPE.26 At KPE, the degree of SCTR expression significantly correlated with the Metavir fibrosis score, as well as with the degree of α-SMA expression. Higher SCTR expression was also noted in infants undergoing KPE later, unsurprisingly, indicating more progressed ductular reaction. Interestingly, initial SCTR also provided some insight on prognosis post-KPE. When SCTR expression levels at KPE were divided into tertiles, the highest tertile was found to have a significantly lower NLS. Significantly lower SCTR expression was also found in patients who cleared their jaundiced post-KPE (Fig. 2).

SCTR expression seems to be corroborated to ductular reaction with more extensive liver disease leading to higher degrees of SCTR expression. Nonetheless SCTR expression is an immunohistochemical test, and therefore invasive and remains relatively specialised requiring further studies to confirm its utility.

IL-8 (aka neutrophil chemotactic factor) is a pro-inflammatory cytokine released by resident macrophages, which attracts neutrophils to the site of injury, potentiates phagocytosis and initiates angiogenesis. Serum IL-8 levels are significantly increased at the time of KPE compared with normal controls, both reach a peak one year after KPE, and correlate with liver fibrosis.27,28

IL-8 combined with other biomarkers such as MMP-7 and GGT have shown an excellent accuracy in diagnosing BA,29 but despite these promising findings IL-8 serum levels only have a weak association with native liver survival.27

Osteopontin is a proinflammatory cytokine synthesized in multiple cell types including the bile duct epithelium. It is involved in macrophage recruitment during inflammation, plays a role in hepatic regeneration and is part of a generic fibrotic response. Previous studies have demonstrated that it is over-expressed in proliferating interlobular biliary epithelium and bile duct structures.30 In a recent study, serum osteopontin levels were found to be increased in BA compared to patients with other causes of cholestasis, nonetheless there is weak association with clinical outcomes such as COJ and NLS.16

Fibroblast Growth Factor 19 (FGF 19) is abnormally elevated in the gallbladder epithelium of patients with cholestatic conditions and is 3.7-fold higher in BA compared with cholestatic-age matched controls. While an association with liver fibrosis has not been found, increased serum FGF 19 positively correlates with bile ductal proliferation. Also, high serum FGF-19 levels independently predict native liver survival regardless of age at KPE and failure to resolve cholestasis in a retrospective cohort of patients collected in two centres26,31 (Fig. 2). Larger prospective studies are still needed to validate these results.

Inflammation indexes based on routine blood results have been developed to serve as cheap and easy-to-calculate markers of inflammation and immune response and have been used as predictors in hepatocellular carcinoma and cholangiocarcinoma.32 The systemic Immune-Inflammation Index (SII), defined as the peripheral platelet counts X neutrophil count-to-lymphocyte ratio, was calculated in a retrospective cohort of 168 patients with BA and correlated with 24-month native liver survival, COJ and cholangitis. Patients with a preoperative SII cut-off point of < 140 had a 5 times higher chance of reaching 24-month native liver survival and better COJ, but no difference in cholangitis rates. The SII also proved to have a higher predictive value over malnutrition, GGT, liver fibrosis grading and lymphocyte count.33

Study of the small molecules, or metabolites, smaller than 1.5 KDa, usually go unnoticed in plasma samples during routine laboratory tests. There has been recent success in identifying the “metabolomic” phenotype for several pathologies such as colorectal cancer34 and thyroid cancer.35 Looking for a metabolomic description of BA, a team from Zhengzhou China analysed the plasma samples of 90 patients with BA, and compared their findings with 48 infants with cholestasis and 47 healthy controls.36 327 metabolites were identified. Further analysis on the difference in the relative concentration of these molecules in each group led to the proposition of a diagnostic model for BA containing three metabolites: lyso-phosphocholine, phosphocholine, and biliverdin-IX-α. A ROC analysis on their proposed diagnostic formula combining the relative concentration of these 3 metabolites returned a sensitivity of 92 % and a specificity of 98 %. Although impressive results of a novel diagnostic tool, these are not routinely measured values, which implies that widespread adoption and validation of this research may have to wait.

Mac-2 binding protein glycan isomer (M2BPGi) has been found in studies to be secreted by hepatic stellate cells,37 inducing Mac-2 expression by Kupffer cells, which in turn increases α-SMA expression and myofibroblast differentiation,37 driving the process of liver fibrosis. Serum concentration of M2BPGi has been investigated as a marker of liver fibrosis in children with BA.38

Yamada et al. histologically assessed the liver of 64 patients with BA who were undergoing a liver transplant, and correlated the serum concentration of M2BPGi with the degree of liver fibrosis on histology39 At a cut-off index of 3.53, grade F4 (Metavir) fibrosis score could be predicted with a sensitivity of 94 % and specificity of 92 %.39 Investigating its prognostic potential, Ueno et al. retrospectively reviewed the serum M2BPGi concentration of 11 patients with BA from the point they were diagnosed with liver cirrhosis until they received a transplant, or the end of the 2-year study period if sooner. They found that a two-fold increase in M2BPGi, as occurred in 2 of their patients, was likely predictive of liver cirrhosis decompensation, and need for liver transplantation.40 Although interesting, much bigger numbers are required before widespread clinical adoption of M2BPGi testing.

Ki-67 is a protein expressed by dividing cells in varying amounts during the cell cycle, peaking during mitosis. It has been popularised as a marker of tumour proliferation and has long been validated in breast cancer.41 Yoshii et al. investigated the expression of Ki-67 in 35 liver specimens of infants with BA, including 17 who underwent LT after KPE, and 5 primary LT.42 At KPE, the expression of Ki-67 in infants who subsequently required LT was significantly higher than those who still had their native liver at follow-up. Interestingly, at the point of liver transplantation post-KPE the expression of Ki-67 was significantly lower than at KPE. When compared against the Inuyame fibrosis score (from F0, healthy, to F4, severe fibrosis), samples with F0 and F4 scores had the lowest Ki-67 expression, while F1 had the highest Ki-67 expression, with increasing fibrosis scores thereafter following an inverse correlation to the expression of Ki-67. Given the small number of samples included in the study it is difficult to differentiate between Ki-67 expression and fibrosis in their ability to predict prognosis post-KPE. The patients with the most severe fibrosis (F4) mostly received a primary LT (5/6), and the patients most likely to not require a LT had lower fibrosis scores (F0-F3). Indeed, our unpublished data of 11 samples showed no difference in prognosis post-KPE and Ki-67 expression at KPE. More data is needed to accurately describe the relationship between Ki-67 expression at KPE and prognosis.

This is a controversial section as every unit believes that they have already adopted the best operative strategy and post-operative adjuvant therapy for their set of patients. We set out areas worthy of closer inspection –can we do anything to get just a little better. The strategy may be thought of as an example of the theory of “marginal gains” –small improvements in many areas leading to significant gains overall. Table 2 summarises novel therapies discussed in detail below.

Centralisation is a relatively recent concept in surgery. It is a response to the effects of infrequent operations for rare diseases and allows economies of scale in terms of multi-specialty input where there are on-going complex needs. It may be constrained by geography obviously, or by a political desire to maintain as many complex operations locally to maintain standing and the financial income that accompanies them.

Evidence of patchy outcomes of KPE in the UK were available from two national outcome surveys in the 1980s and 1990s.43,44 The national government of the day only decided to act following the latter and restricted who could do the KPE and where it could be done, mandating three English centres as centres of expertise (i.e. King's College Hospital in London, Birmingham and Leeds) from 1999. The other key requirement was co-location with liver transplantation facilities.

Early studies justified the policy with a clear improvement in national results45 and there has been a recently published 20-year review of the experience of 867 infants.46 This showed a significant improvement in COJ post-KPE over that period (54 % to 69 %) and reduction of the median age at KPE (53d to 48d) over time. Native liver survival and overall survival remained static (e.g. 10-year overall and NLS was 90 % and 46 % respectively).

Other northern European countries adopted the same pattern (Fig. 3). Following centralisation of care in 2005 to Helsinki, Finland showed a significant improvement in both COJ and 2-year NLS (27 % to 75 % and 38 % to 75 % respectively).47 Combining data from all Nordic countries, including Denmark, Finland, Iceland, Norway and Sweden, there are now comparable 5-year NLS and COJ rates at 55 % and 65 % respectively.48 In countries where centralization of care is not feasible due to political, monetary or logistical hindrances, limited evidence suggests that the formation of a dedicated BA team within a centre can sufficiently concentrate and accumulate experience to improve clinical outcomes.49

Reduction of the age of the infant at the time of potentially corrective surgery should be regarded as fundamental to achieving optimum results in any unit, anywhere. However, much of this is pre-presentation and out of surgeons’ hands. The role of screening in BA (by stool colour charts, early estimation of conjugated bilirubin levels etc.) has been considered in other articles in this issue and will not be considered further here. Similarly, abbreviation of the medical diagnostic pathway and exclusion of medical or metabolic causes of a conjugated jaundice are vital but outside the remit of this article.

One of the key outcomes of a centralised programme of management in England and Wales46 was a progressive reduction in age at KPE due principally to the elimination of the role of the “middle-man” i.e. the reduced responsibility of the local non-specialist paediatrician to be expected to achieve a diagnosis of BA. Fig. 4 illustrates the relationship of outcome with age. A multi-institutional study from the American Western Pediatric Surgical Research collaboration50 suggested that each additional day of age toward KPE was associated with a 2 % decrease in likelihood of transplant-free survival.

Biliary atresia pathogenesis remains an area of active research in which inflammation has an important role. Unsurprisingly, corticosteroids are used in the postoperative period due to their anti-inflammatory properties and their potential to promote bile flow.51, 52, 53, 54, 55

There is considerable variation in the type of steroid, length of course and dosing used, but most use initial IV methylprednisolone and continuation with oral prednisolone. Rectal budesonide54 has been used by one German centre, with the suggestion that it has a higher affinity to the steroid receptor and as there is a predominant first-pass liver metabolism a higher concentration can be achieved with less systemic effects.

There is now considerable evidence favouring the use of a high-dose (vs. low dose) steroid regimen in infants of <70 days at KPE51,52,55 (and not much with older infants56) with most studies showing an increased COJ of at least 15 %52,55 (Fig. 4). Even the North American START trial showed a difference of 15 % in this sub-group, though the difference did not achieve statistical significance.53 Nevertheless, it is noteworthy, that the first author of this trial has adopted a high-dose steroid protocol, albeit in selected infants, for his own institution showing dramatic improvements in their results.57

The largest (n = 198) randomised study is a recent one from Shanghai, China. Their results are in keeping with previous studies showing a statistically significant 23 % improvement in COJ at 6 months post-KPE and increased native liver survival.55

What is also clear is that the efficacy of steroids is improved with a younger age at KPE (Fig. 5). In one age cohort analysis from Kings College Hospital, 100 % CoJ was achieved for that cohort who were ≤30 days at KPE.56 Customization of post-operative corticosteroid regimes by direct observation of stool colour and further steroid pulses has become common in Japanese centres58 and as has been mentioned above in certain North American centres.57 This necessitates an inevitable increase in length of stay,58 but may be the way forward to maximise outcomes.

One of the reservations of using high dose steroids is the potential increase of adverse effects. In the START trial, comprehensive follow-up showed an increase in the number of “adverse effects” in the steroid arm, though after first 30 days the risk normalised thereafter.52

Ascending cholangitis is the most common postoperative complications post-KPE and recurrent events of cholangitis are associated with worse native liver survival.59,60 A retrospective observational study from Houston, TX showed that nearly half of all infections developed in the first three months post-KPE, with the most common being cholangitis (45 %), followed by viral respiratory illness. Older infants and those with lower platelet counts at KPE seemed to have a higher risk of developing any infection.61 The most significant factor in the ultimate failure of anicteric survivors is cholangitis, particularly if it is recurrent.60

With this in mind, a very low threshold for suspecting and aggressively treating cholangitis, especially in the first-year post-KPE, is advised. A recent expert consensus panel (predominantly European and North American) agreed on treating suspected cholangitis for at least 7–10 days and confirmed cholangitis for 14–21 days.59

It is common practice to use prophylactic antibiotics to prevent this potentially catastrophic scenario, but the available evidence of any effect is actually rather limited. A recent meta-analysis including studies with control groups showed that adding prophylactic antibiotics did not reduce the risk of cholangitis,62 though most studies included were retrospective and non-randomised. The aforementioned expert panel recommended using prophylaxis for 6–12 months after KPE.59

Viral infection has been long been suggested as a causative factor in BA. Out of the many viruses thought to be involved the current focus for most recent studies has been CMV. The incidence of CMV in patients with BA has been reported between 10 and 32 % in European centres,63,64 12 % in North America65 and up to 60 % in Chinese studies.66 It is still unclear if CMV is an initiating and causative factor or an aggravating factor to something else. There do appear to be clear clinical and laboratory differences between those deemed CMV +ve and CMV –ve. So, age at KPE is higher, they tend to have worse liver function tests and higher APRi levels.63, 64, 65, 66

There is histological evidence exists of increased inflammation and fibrosis in CMV +ve positive infants and specifically increased infiltration with CD8+ and Th1+ lymphocytes and CD56+ NK cells.67,68

CMV +ve infants appear to have worse outcomes post-KPE63,69; very bad at least in our original study with barely any keeping their native liver.64 Nonetheless, a retrospective review of outcomes on stored biobank samples derived from many North America centres (n = 220) did not identify differences in COJ. What they did find was a significantly higher risk of death regardless of age at KPE,65 an intriguing if unexplained finding originally reported in our own study.64

The effect of using antiviral therapy (AVT) as part of an adjuvant regimen is still evolving. A decrease of viremia without side effects was reported in one Swedish study,69 and in our own prospective database AVT significantly improved clearance of jaundice rates and native liver survival.70 Hopes for a randomised controlled trial, even a multicentre one, appear slim given that the CMV incidence is relatively low in Europe and North America. So, it looks if it is going to happen it will be Chinese-led.

The application of regenerative medicine in BA has gained traction in recent years.71 The liver's capacity to regenerate following injury or indeed surgical resection has long been appreciated.72 It has been shown that fully differentiated biliary epithelial cells can dedifferentiate and propagate to completely replace hepatocytes even after severe depletion.73 The opposite is also true,74 indicating cellular plasticity in a healthy hepatobiliary environment. Similar phenomena have been described in adult patients with chronic liver disease,75 indicating a retained capacity for the liver to regenerate even during disease.

More recently, a team from Pittsburgh PA has sought to further delineate this mechanism in paediatric [liver disease, including BA.76 LGR5 is a G-protein coupled receptor present in stem cells extensively studied in intestinal crypts.77 Although absent in a resting liver, LGR5+ Hepatic Progenitor Cells (HPC) are induced in BA, as well as in α1 anti-trypsin deficiency (A1ATD), and progressive familial intrahepatic cholestasis (PFIC).76 Indeed, the regenerative compartment induced by the ductal reaction in BA has been shown to be expanded in patients with more advanced disease undergoing liver transplantation.78 Efforts to shift the equilibrium from inflammation and hepatocytosis to regeneration are ongoing.

Already a mainstay of treatment in patients with congenital or acquired diseases of bone marrow suppression, granulocyte-colony stimulating factor (GCSF) has been shown in animal models to increase liver regeneration, as well as decelerate hepatic inflammation and fibrosis.79 A team from Hanoi, Vietnam have recently published results of a Phase 1 trial, examining the effects of GCSF in infants with BA.80 The trial established safety up to a dose of 10 μg/kg/d for 3 days post-KPE, with all patients mounting an appropriate increase of CD34+ cell count, indicative of bone-marrow derived haematopoietic stem cell mobilisation. Moreover, when the 5 study participants were compared to 10 controls who received KPE without GCSF, the group receiving GCSF had non-significant lower bilirubin levels at 1 month post-KPE (62 μmol/L vs 21 μmol/L).

Transplantation of autologous bone marrow mononuclear cells (BMMNC) has also been investigated for the treatment of chronic liver disease.81,82 Nguyen at al., again from Vietnam, recruited 19 patients post-KPE with established cirrhosis (aged 7 months to 15 years).83 Mononuclear cells were aspirated from bone marrow and infused through the common hepatic artery. This was repeated at 6 months. At 12 months, the degree of fibrosis and peri‑portal inflammation on liver biopsy and endoscopic appearance of varices was unchanged. However, there was a significant reduction of serum bilirubin and GGT (42 μmol/L to 6.8 μmol/L and 211 IU/L to 149 IU/L respectively). This is certainly a promising set of results, suggesting limitation of the progression of liver disease even in established liver cirrhosis.

There are minimally-invasive alternatives to the traditional open operation. Laparoscopic KPE was initially reported by a Brazilian team in 200284 with a number of case reports thereafter. However, poor results prompted an international moratorium in the 2000s and it has not really become re-established in North America or Europe. Nevertheless, its use over the past 10 years has been re-invigorated in large Chinese and Japanese centres.85, 86, 87, 88 There is no real suggestion that the surgery can obtain better outcomes that open surgery –how can it given it should be a replica of the open operation (biliary remnant resection and Roux-loop reconstruction).

One of the largest Chinese experiences is from Huai'an in Jianshu province that compared 45 open KPE (2012 to 2016) to 56 laparoscopic KPE (2015 to 2021). There was no difference in COJ group (74 % vs. 79 %) and a similar 2-year NLS (71 % vs. 67 %).85 Other groups failed to show a significant difference in long term outcomes.86,87 The most recent systematic review identified 23 comparative (non-randomised) studies and showed that there were no significant differences in main outcomes.89 Some centres, following a period of laparoscopic KPE reverted to the open approach because of worse results.90

A high case-load appears mandatory for satisfactory results in laparoscopic KPE, and therefore may well preclude its use in the West. Some authors have even put a figure of 40–50 patients91,92 as the learning curve before the clinical outcome of the operation becomes comparable to the established open technique.

Finally, robotic-assisted KPE, has now been re-reported, again from Wuhan, China93 with a series of 25 infants and more than acceptable outcomes almost 20 years after it was first reported from Stanford, CA.94 It may offer more intra-operative dexterity due to the robot's articulating wrists and 3D field of vision. Nevertheless, it remains out of reach for many centres and today's robotic technology is still not geared to that of an infant population.