Higher Red-Cell Transfusion Limit for Preemies Offers No Survival, Developmental Benefit

By Gene Emery

December 31, 2020

(Reuters Health) - Aggressive use of red-cell transfusion for extremely-low-birth-weight newborns with anemia does not improve survival and fails to stem neurodevelopmental impairment by age 2, doctors behind a new study reported Wednesday.

The TOP trial of 1,824 infants with a birth weight of 1,000 grams or lower used an algorithm to assess when a transfusion was advised based on infant age and respiratory support. For example, at three weeks or older, babies on respiratory support were given a transfusion if their hemoglobin dropped below 11.0 g/dL if they were in the high threshold group or if it fell below 8.5 for infants in the low threshold group. With no respiratory support, the thresholds were 10.0 and 7.0 respectively.

The odds of death, cognitive delay, cerebral palsy, hearing loss or vision loss at the 22- to 26-month mark were 50.1% in the high-threshold group and 49.8% in the low-threshold group (P=0.93). The rates were corrected for prematurity.

Death claimed 16.2% of high-threshold babies and 15.0% of low-threshold infants. Neurodevelopmental impairment was seen in 39.6% and 40.3%, respectively.

By the time they were discharged from the hospital, 28.5% of the infants in the high-threshold group had survived without severe complications versus 30.9% in the low-threshold group.

Based on the results, "I would argue that less is better," chief author Dr. Haresh Kirpalani, professor emeritus of neonatology at the Children's Hospital of Philadelphia in Pennsylvania, told Reuters Health in a telephone interview.

"There's a couple of caveats to add. If a baby was going for surgery, the algorithm was put on hold until the physician felt the baby recovered or didn't need any further intervention. But otherwise, you can allow the infant to go down in hemoglobin values," he said.

The study was done at 41 neonatal intensive care units with randomization from 2012 to 2017. Followup was through February. The National Institutes of Health funded the test.

Red cell concentration can be a problem in these babies because their bodies may not be up to speed for generating the cells, and they may desperately need the blood that is often drawn to guide their treatment.

Debate has raged for years over the safest level to tolerate and Dr. Kirpalani said he undertook the study because "I got fed up on ward rounds arguing whether a baby should or should not get blood."

A 2005 study had also suggested little benefit for a more-aggressive transfusion practice but there was a hint in the data that using a higher threshold might prevent some mild-to-moderate cognitive delay.

The TOP researchers in this much-larger trial didn't see it.

They did see more transfusions in the high-threshold group and that the difference in the transfusion standard produced a mean difference in hemoglobin levels of 1.9 grams per deciliter throughout the treatment period, with no improvement in death or disability.

"There was no evidence that the effects of the transfusion strategy on the primary outcome differed according to center, birthweight group, or sex," the researchers said.

There was no significant difference in the odds of serious adverse events.

There's still work to be done, said Dr. Kirpalani. "The most obvious follow-up that's needed is to see if intellectual outcomes are no different by the age of 5 years."

It is not known how many neonatal intensive care units currently use the lower or higher standard, or how many of them will adjust their rules based on the TOP study.

"However, I think neonatologists have been primed by previous studies and much debate on this issue to be looking for an answer. I think this trial is sufficiently large to be convincing," said Dr. Kirpalani.

It might also prompt blood banks to question the need for a transfusion in these children, especially during times of shortage, he said.

SOURCE: https://bit.ly/3pnLjH3 The New England Journal of Medicine, online December 30, 2020.

Comments

3090D553-9492-4563-8681-AD288FA52ACE

processing....