Our new ESPGHAN position paper and the reasoning behind the recommendations.
I’m very excited to share what is almost certainly the most impactful piece of work I’ve done so far: a position paper on the screening and diagnosis of MASLD in children, from the ESPGHAN Steatotic Liver Disease Special Interest Group. We’ve been actively working on it for more than two years, with at least another year of build-up before that.
(If MASLD is a new term to you, I’ve written before about the name change from NAFLD to MASLD.)
The paper makes a large number of recommendations. Here I want to highlight just a few.
[But first, I should say that these are my personal reflections on this paper, and do not necessarily reflect those of my other co-authors or ESPGHAN.]
Who to screen. On balance, we recommend screening for paediatric MASLD, but only in selected children. Almost every international paediatric obesity and liver society already recommends screening, so we’re not unusual. Where we differ is that we recommend screening only in children affected by obesity, not those who are affected by being overweight alone. The rationale is that more severe metabolic syndrome (particularly insulin resistance) correlates with more severe liver disease, and typically tracks with the degree of BMI elevation.
Stepping back from more lenient screening has large public-health implications. Services are already full of referrals for paediatric MASLD, and we need to focus on the cases that carry meaningful diagnostic information i.e. only give a child a label if it’s actually going to change what happens to them. I’m not convinced that labelling a child with a BMI z-score of just over 1.0 and an ALT of 31 as having MASLD changes anything we’d practically do for them: they’re highly unlikely to be offered any pharmacotherapy for their liver (there is none currently licensed in children), and the lifestyle advice for their BMI would be the same regardless.
Which ALT cut-off. We’ve also used a different ALT cut-off for screening than the AASLD guidance. We use a flat cut-off of 30 for everyone (boys and girls and at any age within the paediatric window). This is, above all, to make the guideline simple, practical and usable. I’m meant to be an expert in this condition, and I still forget the exact age- and sex-specific cut-offs derived from the NHANES data (a US study that has very accurately defined normal ALT levels in children). If you look at that data, the upper limit of normal ALT moves around within roughly 22–35 depending on whether you’re a boy or a girl, and on other factors such as ethnicity.
The degree of ALT elevation also correlates with histological severity in paediatric (but not adult) MASLD, replicated in multiple studies. So if we’re using a slightly higher cut-off than some other guidance, the children we’d miss on screening alone are those with the lowest ALT values, who have a lower probability of histologically advanced disease. I do appreciate there’s good data showing that some children have completely normal ALT and significant fibrosis. But this cut-off is for screening only; it is not the only route by which a child can be investigated.
Note, if a child has an ALT over 80, don’t delay investigation or referral, in case it reflects more significant liver dysfunction. If the ALT is between 30 and 80, it’s probably worth repeating after one to two months in case it was a transient blip (assuming a suitable clinical assessment as well).
Diagnosis vs screening. We’ve set a different BMI threshold for diagnosis than for screening. Screening requires obesity; for diagnosis, a child who is at least overweight can be diagnosed with MASLD. This differs from some other guidelines: the original Delphi consensus that introduced the NAFLD-to-MASLD name change required only any one of several metabolic criteria. A subsequent updated international consensus held that, for the vast majority of paediatric cases, the child must be at least overweight, because not being overweight is the single strongest predictor of an alternative diagnosis rather than MASLD. Lean MASLD (a BMI z-score below 1.0) can still be diagnosed, but only by a specialist after extensive investigation, which is why it sits outside the standard pathway we set out here.
Monitoring. Previous ESPGHAN guidance generally held that the degree of change in transaminases and GGT wasn’t helpful. Data accumulated over the last few years from multiple cohorts (and in particular paired-biopsy studies from the US) now show that improvement in ALT and GGT does correlate with histological improvement, as does a reduction in ALT and AST. This is in complete contradiction to adult MASLD. It’s unclear at what age or pubertal stage this stops being true in children — at some point they transition from paediatric to adult guidance — and that’s an important open question, alongside many others about MASLD in young adults.
Non-invasive tests. In adults there are well-defined cut-offs, derived from thousands of histologically validated patients, for the ELF test and for transient elastography (most often FibroScan). In children the data is much more limited. We’ve concluded that transient elastography is probably useful at the extremes: a very normal reading (under 5.5 kPa) or a high one (over 10 kPa) probably do predict a normal or an abnormally stiff liver respectively. In between, it’s difficult, and we’ll need more data to refine cut-offs for moderate and significant fibrosis. This contrasts with the AASLD guidance, which doesn’t recommend transient elastography at all. My view is that it’s already such a routine part of our practice — not just in MASLD but across paediatric hepatology — that we can reasonably offer some cut-offs, while acknowledging that the majority of patients will fall in the range where we don’t yet have the data to interpret the degree of stiffness meaningfully. We also don’t yet know whether treatment-related changes in elastography correlate with histological improvement, or whether elastography correlates with clinical endpoints in children. I would imagine it does to some extent, by extrapolation from adults, where it correlates with both histology and clinical outcomes. So there’s some assumptions involved, but we think there’s utility at the extreme ends of transient elastography readings.
Liver biopsy. Finally, the most important role of liver biopsy remains diagnostic uncertainty. Previous ESPGHAN guidance worded it such that abnormal liver enzymes persisting for six to twelve months were themselves an indication for biopsy: that’s no longer the case. That said, a child with very persistently abnormal enzymes (e.g. ALT persistetly >250 IU/L) may well, in my view, fall into the “diagnostic uncertainty” category, and after reasonable non-invasive testing I would probably recommend a biopsy to be sure we’re not missing another diagnosis. We’ve also recommended considering biopsy around the time of transition to adult services, given the uncertainty about non-invasive testing in this group, and because a clear diagnosis can have real implications for follow-up and for eligibility for future therapeutic options in adulthood.
It has been a genuine pleasure and privilege to work on this position paper, and I hope it proves useful to clinicians, families and patients across Europe and beyond. I’d welcome any feedback on how it works in practice.
