To the Editor,

Handa et al. have added a genuinely useful dataset to a field that still runs largely on case reports. ^{Reference Handa, Krishnaswamy and Faisal1} Their series of 11 infants confirms what smaller reports have suggested piecemeal: that sirolimus produces prompt clinical stabilisation and substantial tumour shrinkage and it does something more valuable still: it puts a number on regrowth after stopping the drug. A 27% recurrence rate, with all three children responding again on reinstitution, is exactly the kind of finding that should shape how the next set of infants is managed and we would like to build on it.

Our observation concerns the discontinuation threshold itself. The authors stopped sirolimus once tumour size fell by more than 75%, alongside resolution of arrhythmia, obstruction and ventricular function. This is a reasonable pragmatic rule, but the paper’s own table hints that it may not fit every patient equally. Cases with confirmed TSC1/TSC2 mutations and those without a demonstrable mutation appear scattered fairly evenly among both the complete responders and the ones who later regrew, which raises the question of whether genotype, rather than percentage regression alone, should eventually inform how aggressively a clinician chases complete disappearance before withdrawing therapy. A dataset of this size cannot settle that question, but it is large enough to make the question worth asking prospectively.

The regrowth pattern the authors describe several months after cessation sits comfortably alongside what has now been reported on the prenatal side of this same biology. In a recent letter we discussed a case of transplacental everolimus for fetal cardiac rhabdomyoma in which regrowth occurred at 1, 3, and 5 months after birth, a pattern consistent with the roughly one-third postdiscontinuation regrowth rate reported across the prenatal mTOR inhibitor literature. ^{Reference Jindal and Kamal2,Reference Mustafa, Javinani and Morning3} Read together, the postnatal and prenatal experiences suggest that regrowth after mTOR inhibition is not a quirk of dosing or adherence but a predictable feature of the underlying tumour biology once the drug is withdrawn, whichever side of birth treatment began on.

The near-total absence of adverse effects in this cohort, with only mild hypertriglyceridaemia in two infants, is reassuring but also somewhat atypical against a literature in which hypertriglyceridaemia has been reported in up to 60% of infants, and outcomes in other neonatal series have varied more widely. ^{Reference Handa, Krishnaswamy and Faisal1,Reference Ng, McGuinness and Prendiville4} This may simply reflect careful trough-level titration in a busy tertiary unit, but neonates have markedly immature CYP3A4 activity, and even small differences in monitoring intervals could plausibly account for divergent toxicity rates between series. A pooled analysis of trough levels against adverse events across published cohorts would help clinicians titrate with more confidence than any single-centre series can offer.

None of this diminishes the value of the paper. If anything, the honesty with which the authors flag an uncertain optimal duration, rather than overselling a fixed protocol, is what makes this series worth building on. A prospective multicentre registry, capturing genotype, trough levels, discontinuation criteria and regrowth timing in a standardised way, seems the natural next step and one this group is well placed to lead. ^{Reference Muschel, Höck and Griesmaier5}