A new publication from the trial funded by Prolacta looking at possible benefits of their donor-human-milk derived preparations. This time it is the other comparison from the registered trial. That is; mothers who did not plan to breast feed were approached to randomize their infants, birth weights 500 to 1250 g.

(Cristofalo EA, Schanler RJ, Blanco CL, Sullivan S, Trawoeger R, Kiechl-Kohlendorfer U, Dudell G, Rechtman DJ, Lee ML, Lucas A et al: Randomized trial of exclusive human milk versus preterm formula diets in extremely premature infants. The Journal of pediatrics 2013)

The 2 groups received either processed, standardized donor human milk, with human milk based fortifier; or a preterm formula (which one is not stated). It is not clear what calorie density was given at what stage during the feeding regimes, although both seem to have started at 20 kcal/oz, and increased later on to 24 kcal/oz.
The primary outcome variable, as in the other part of this study, was the duration of parenteral nutrition, compared using a Kaplan Meier analysis. The sample sizes were tiny. Only 26 babies per group were planned, which was based on a very hopeful 50% reduction in the duration of PN which they thought would average 35 days in the formula group (SD 22 days).

This study was masked, which, as mentioned in the previous post, is rather important when many of the outcomes are based on subjective factors.

The enteral nutrition regime was supposed to start with 1 to 4 days npo, followed by up to 5 days of trophic feeding, followed by progression at 10 to 20 mL/kg/d up to 150 mL/kg/d.

Before getting to the clinical outcomes, it is interesting to note that the actual timing of introduction of the first feed was actually 6.5 days on average for the cow’s milk group and 4.0 days for the human. So most of the babies were started on feeds after the upper limit of the planned duration of being npo; it then took 25 days for the human milk fed babies to reach full feeds, and 29 days for the cow’s milk. I think we can do better than that. I think most very preterm babies can start feeds on day 1, and there is no good justification for not immediately starting to increase feeds, aiming for 30 mL/kg/d.  The data show that having trophic feeds is better than no feeds, but there is no data to show that trophic feeds are better than immediately increasing feed volume. There is no data that different rates of feed advancement affect NEC.

In this study the parenteral nutrition duration was longer with cow’s milk, (p=0.4). There was a lot of NEC in the bovine group, 5 cases, 21%, and 1 case in the human milk group, which was not statistically significant. 4 of the NEC cases in the cow’s milk group went for surgery, but not the human milk baby (p=0.036).

That is an awful lot of NEC, and an awfully high proportion needing surgery. I don’t think I have ever seen that in a publication, and I also used to work in a center where there was very little breast milk when I arrived, and we didn’t have anywhere near that incidence (even though it was too high).

I think we do need to be a little skeptical about this result, it is, again a secondary outcome, in a very small trial which was seriously underpowered, with a very small number of total events, but it is suggestive that this approach may be a big advance, if it can be confirmed in further, larger trials.

One thing also worth noting is that the incidence of late onset sepsis was not affected in either of the  Prolacta trials. In this new publication a whopping 79% of the bovine and 55% of the human milk babies got at least one late onset infection (difference not significant). In the previous study it was between 19 and 28% (highest in one of the human milk fortifier groups). Clearly human milk is not the answer to infection prevention, we need other approaches.