Hypothermia at NICU admission in very preterm infants

August 2016  Volume 175

Blunted response: smoke, illicit substances, and how babies breathe — Clyde J. Wright, MD

Is it better to overtreat or undertreat? — Janet H. Silverstein, MD

Copyright ª 2016 by Elsevier Inc.

D

espite our best efforts, rates of death due to sudden infant death syndrome (SIDS) remain unacceptably high. A litany of risk factors predisposing babies to SIDS has been identified, including maternal smoking and substance misuse. However, the mechanisms linking maternal smoking and substance misuse to SIDS are unknown. In this volume of The Journal, Ali et al demonstrate that during the time when the risk of dying from SIDS peaks (6-12 weeks of age), term infants born to mothers who smoke or misuse substances have a blunted ventilatory response to inhaled carbon dioxide when compared with controls. Specifically, control infants demonstrate an expected and reliable increase in mintue ventilation when challenged with either 2% or 4% carbon dioxide. In contrast, the change in minute ventilation in similarly exposed infants born to mothers who smoke or misuse substances was significantly less. It is attractive to hypothesize that infants exposed to smoke and other illicit substances—previously shown to carry a significant SIDS burden—are predisposed to complications arising from situations that require an intact venilatory response. These data provide important insights into the potential mechanisms linking prenatal exposures and SIDS. How will we use this information? What is behind the impaired response – and is it permanent? Can we modify this impaired response, especially during the period of highest risk of dying of SIDS? Should more babies be studied to confirm this association, or should we begin to cousel families based on this observation? While we work to answer these questions, it seems appropriate to emphasize that the families taking care of these babies must take great care when creating the sleep environment for these incredibly high-risk infants. Article page 224<

Y

ogi Berra, an exceptionally gifted baseball player, was a font of wisdom. One of his most famous quotes, “When you come to a fork in the road, take it,” applies to the conundrum tackled by Paone et al in this volume of The Journal. Approximately 43% of infants and 10% of older children with congenital hypothyroidism have elevated levels of both thyrotropin (TSH) and thyroxine (T4). When both T4 and triiodothyronine (T3) are elevated, pediatricians are in the untenable position of having to determine if they should overtreat in an attempt to normalize the TSH at the possible expense of increasing the T4 above normal, when studies have demonstrated impaired school performance, lower IQ, and a higher risk of attention deficit in children who had high T4 concentrations in infancy (J Pediatr 2000;136:292-7; Pediatrics 2003;112:923-30). Or, should they attempt to maintain normal T4 values while TSH concentrations remain mildly elevated? Although the evidence that low T4 levels during the time of rapid brain growth affects intellectual functioning, the evidence that elevated TSH values during the first 3 years of life is less robust, though one study reported poorer school performance in children who had TSH elevations in infancy (Acta Paediatr 2001;90:1249-56). Due to the lack of data, we do not know which fork to take in this particular road. The cause of the abnormal lack of suppression of TSH with adequate levothyroxine administration is unknown. Affected infants might have abnormally low thyroid gland secretion of T3, abnormally low conversion of T4 to T3, or abnormal central feedback with lack of normal TSH suppression (ie, central resistance to thyroid hormone). (Continues on next page) 1

Paone et al hypothesized that the lack of suppression of TSH by appropriate doses of levothyroxine (LT4) could be overcome with the addition of liothyroninie (LT3), as shown in other studies (Horm Res Paediatr 2010;73:108-14; J Pediatr Endocrinol Metab 2011;24:347-50). In order to assess this hypothesis, the authors performed a retrospective analysis of clinical data of 12 patients with CH with both high T4 and TSH values. Six of them had been treated with LT4 alone followed by LT4 + LT3 and 6 only received LT4. The addition of LT3 resulted in the normalization of TSH (mean 4.3 mIU/L, down from 10 mIU/L pre-T3 treatment , vs 8.5 mIU/L in the T4 monotherapy group). The T4 values were likewise normalized, with a mean decrease of 23%  9% from pretreatment values. T3 concentrations remained normal for age. Because this was a retrospective study, the groups were neither matched for the age of treatment initiation nor given a neurodevelopmental assessment. However, this is a promising treatment for infants with persistent elevation of TSH and T4 despite appropriate treatment with levothyroxine. Long-term randomized controlled trials are needed to assess the effect of this regimen on thyroid hormone values. But, the real question that must be answered is not whether the laboratory values can be normalized but whether this regimen can improve neurocognitive functioning in these children, including not only intelligence but also school performance. And, hopefully, this study will help us to determine which fork in this particular road we should take. Article page 167<

Do we end life well? — Paul G. Fisher, MD

Iron deficiency anemia: like mother, like child — William S. Ferguson, MD

2

M

any general pediatricians and subspecialists believe that their clinical acumen caring for a child while ill extends equally well to the end of life. But do we end life well? Do we meet the needs of families and their dying child? In this volume of The Journal, L€ ovgren et al report their results from a cross-sectional survey of Swedish parents whose 40 children died from spinal muscular atrophy types I or II. Although end of life from this neurodegenerative disease may not exemplify all other pediatric deaths, and some practices in this study could be specific to Swedish culture, the findings regarding how we deliver care are only slightly reassuring and perhaps more heart wrenching. Among parents who talked with a physician about how they wanted their child to die, all but 2 of 26 had their wishes fulfilled. All those parents who wanted their child to pass away in the hospital had their request met, but 6 of 16 who wanted their child to die at home did not have their wish fulfilled. Siblings were almost always not engaged in the dying process. More than one-quarter of parents reported that health care staff did or said something stressful at the end of life. These findings study should make us want to do better. As the investigators note, communication with the parents is paramount. We should not assume but rather explore parents’ wishes, values, and spirituality. Cultural differences need to be appreciated. Our everyday clinical practices need to be altered and sometimes radically changed at the end of life. The actions of health care staff can be particularly lasting memories. As the authors note, especially supportive efforts from providers after death were dressing the child and making the room beautiful, collecting mementos, showing feelings openly, giving small gifts, and opening the window to release the soul. Article page 201<

I

ron deficiency remains a leading cause of anemia worldwide. The prevalence among infants and toddlers in the US remains close to 10% despite widespread adoption of iron-fortified formulas and the recommendation to delay the introduction of cow’s milk until 1 year of age; in developing countries the frequency is often much higher. In addition to anemia, there is evidence that iron deficiency causes long-term neurocognitive deficits that can persist despite iron replacement therapy. Although the most recent US Preventive Services Task Force guidelines recommend iron supplements for high-risk infants (low birth weight or premature), there is a surprising paucity of

Volume 175

high-quality data regarding the value of routine screening or iron supplementation for low-risk infants. In this volume of The Journal, Zhang et al report on a retrospective review of mother-infant pairs in northern China who had enrolled in a large prospective study of iron and vitamin supplementation during pregnancy. Although they did not directly measure iron status in their subjects, this population has very low frequencies of hemoglobinopathies and infections that might cause anemia; thus, it is logical to conclude that nutritional factors (eg, iron deficiency) were a major cause for anemia when it was observed. The investigators found that infants born to women with mild anemia at the end of the second trimester (hemoglobin levels 100-109 g/L, approximately 6% of the total population) had almost double the risk of overt anemia at 5-7 and 11-13 months of age, compared with infants born to women with hemoglobin levels $ 110 g/L. These results suggest that maternal anemia during late pregnancy could reasonably be used as an additional risk factor to identify infants who are at increased risk of nutritional anemia and who might benefit from targeted efforts at detection and prevention, especially among resource-limited populations for whom the relative ease and low cost of hemoglobin/hematocrit measurements would be an important consideration. Article page 106<

Nonsteroidal anti-inflammatory drugs: risks outweigh benefits — Sarah S. Long, MD

August 2016

I

nvestigators from 15 medical centers in France performed a case-control study to evaluate a hypothesized association between use of nonsteroidal antiinflammatory drugs (NSAIDs) during viral illness and occurrence of pneumonia with empyema. Cases were consecutive children 3 months through 15 years of age hospitalized at participating centers with thoracentesis-proven bacterial pneumonia. Cases also had to have a physician-diagnosed acute viral illness during the 15 days prior to hospitalization. Controls were children without empyema (matched within 1 year to age of case), with physician-diagnosed acute viral illness in the case physician’s practice during the 15 days prior to case’s hospitalization. To minimize confounding for NSAID indication, cases were excluded if there was not at least 24 hours without fever after onset of an upper respiratory infection (URI), before onset of symptoms leading to hospitalization, or if hospitalization for empyema occurred within 72 hours of URI onset. Investigators went to great lengths to identify case and control subjects’ drug exposures and viral causation of URI, and cases’ causation of empyema. Molecular techniques were used to detect viruses, bacteria, and agents of atypical pneumonia. The study included 83 cases with matched controls. Respiratory viruses were detected in 47% of cases and 45% of controls, with almost identical specific virus detections across groups. Mean delay between onset of URI and empyema diagnosis was 10 days. In 79 (95%) cases, bacterial etiology was confirmed, which was Streptococcus pneumoniae (predominately serotypes 1, 3, and 19A) in 86%. Considering the intake of any drug within 72 hours of onset of acute viral illness, $1 day of NSAID exposure had $two-fold OR (95% CIs 1.37 5.61) for empyema. Le Bourgeois et al designed and performed a cleverly conceived study to address the question, attempting to minimize confounding factors that could plausibly explain a false association of bacterial infection with NSAID use. Readers still will want to dissect the report carefully. The finding of the same viruses in cases and controls eliminated confounding related to virulence of virus. Considering likely potentiation by NSAIDs of bacterial complication of viral illness along with clearly identifiable risks of drugassociated adverse events of severe acute renal injury and aseptic meningitis, we should stop the current stampede to NSAID use during acute viral illness in children. Article page 47< 3

A severity score in PGM-1 deficiency — Hans C. Andersson, MD, FACMG

Hypothermia at NICU admission in very preterm infants — Alan H. Jobe, MD, PhD

4

C

arbohydrate modifications on proteins have long been recognized as important to targeting and function. I-cell disease established the critical mannose signal for targeting lysosomal enzymes synthesized in the golgi apparatus to their functional site within the acidic lysosomal lumen. A rapidly growing field of congenital disorders of glycosylation are teaching us new lessons about other critical functions of carbohydrates bound to proteins. Disorders resulting from deficiencies of posttranslational carbohydrate modification or biosynthesis of these molecular markers result in a broad array of clinical symptoms of varying severity. In this volume of The Journal, Wong et al describe a clinical severity score in 27 patients deficient in phosphoglucomutase-1, a well-recognized enzyme that functions in glucose homeostasis and appears to play an important role in posttranslational modification of glycoproteins. Patients with phosphoglucomutase-1 deficiency in this cohort are categorized into severity groups and have specific clinical predictors of severity identified. Establishing a natural history in this rare condition allows reliable assessment of the emerging therapy with oral galactose. This severity score identifies standardized clinical and biochemical variables that clinicians can use in the care of patients with phosphoglucomutase-1 deficiency. Article page 130<

S

ince the reports by Silverman et al reporting that hypothermia was associated with increased mortality in preterm infants, neonatologists have tried to effectively regulate body temperature. The challenge that remains is that the smallest and most premature infants continue to have hypothermia on admission to the neonatal intensive care unit. In this volume of The Journal, Wilson et al report that hypothermia occurred in 53.4% of 5697 infants born at <32 weeks gestational age in a population-based study with samples from 11 European countries. More concerning is that 12.9% of the population had admission temperatures below 35.5 C. Thus, despite the recent focus on developing and testing new approaches to better maintain body temperature in this very high risk population, temperatures often were quite low soon after birth. In this large population, a low initial temperature was associated with increased death at 1 to 6 days and 7 to 28 days of age despite multiple adjustments for variables likely linked to low body temperature. Because later deaths with hypothermia were not associated with low Apgar scores or other complications of prematurity, the effect of admission temperature may not be explained by an association of low body temperature with a less favorable clinical status at birth. The causal link between hypothermia and mortality remains unexplained, but hypothermia following very preterm delivery is a preventable “disease” that requires constant attention to prevent. Article page 61<

Volume 175