There are 2 new trials addressing different aspects of the questions surrounding what to do about TPN for this group of babies.

My readers may remember my post about the PEPaNIC trial, which showed that there were disadvantages of early TPN, compared to TPN starting after 7 days of ICU among PICU patients. And the subgroup analysis which showed that the same results applied to the newborn infants in that trial, who were mostly babies who needed surgery. Babies receiving early TPN had more nosocomial sepsis, longer assisted ventilation and longer PICU stays.

The first of the 2 trials I wanted to discuss is a smallish single centre trial, with some similarities to PEPaNIC. (Moon K, et al. Early versus late parenteral nutrition in term and late preterm infants: a randomised controlled trial. BMJ Paediatr Open. 2024;8(1)) 60 babies of at least 35 weeks, admitted to the NICU, who were expected to not be fed for at least 3 to 5 days were randomized soon after admission to either start immediate TPN or to wait until day 6. The trial is clearly too small for clinical outcomes to be the primary, so the primary outcomes were serum phenylalanine concentrations (to see if the babies were tolerating the amino acids they were given) and F2-isoprostane (as an index of oxidative stress).

The early TPN babies received more protein and more total calories during the randomized period, but by day 8 there were no differences in intakes. TPN babies were more likely to have mild hyperglycaemia (>8.3 mmol/L), both groups had a lower body weight z-score at discharge than at admission, and the fall was greater in the late TPN group (-1.0 vs -0.6).

The publication states “all other clinical outcomes were similar between groups”, but I beg to differ! In the supplemental materials we find : 8/30 early TPN babies developed a hospital acquired infection, compared to 3/30 late TPN babies, 27% vs 10%, or more than double. In the published protocol we note that these are culture-positive blood or CSF infections. Duration of NICU stay was 5 days longer (21 vs 16) with early TPN, and total duration of hospitalisation was 4 days longer, 23 vs 19 d. Duration of respiratory support was 16 hours longer among babies who were on such support. There was also twice as many babies with hypoglycaemia in the early TPN group (4 of them vs 2)

Of course, the authors had to very careful to not make too much of those differences in such a small trial. They are, relatively speaking, large differences, but could have been due to chance. What is striking is that the differences are all consistent with what PEPaNIC showed, (apart from hypoglycaemia, which was more common in the late TPN group in PEPaNIC).

The primary outcomes did show a difference in phenylalanine levels, which were 30% higher in the early TPN group during the randomized period, but almost identical afterwards. Isoprostane levels were similar between groups throughout.

The practice of starting very early TPN was based on a physiologic rationale for the very preterm infant, with the desire to reduce catabolism among infants with very limited energy and protein stores at birth, and evidence that this physiologic goal was attained by immediate, or very early TPN. The practice has slowly spread to other babies in the NICU, and TPN is often started early among other infants, for whom there is no evidence of benefit. Among such infants, however, there appear to be negative impacts, and there is no clear evidence of benefit.

I think that term and near term babies, admitted to intensive care units, who cannot be full fed immediately, can wait several days before TPN is considered. Whether we wait 3 days or a week or even longer is uncertain, and will require larger studies, the current state of knowledge is that there is no need to hurry.