top of page
Horiz_red_podcast.png

#341 - 📑 Journal Club - The Complete Episode from August 17th 2025

Updated: Sep 4

ree

Hello friends 👋

This week on Journal Club, we review five important studies with direct implications for your work in the NICU. First, we examine the newly released third-generation Fenton growth charts—how do they differ from previous versions, and what do they reveal about how we may have underestimated early growth trajectories in preterm infants? Next, we evaluate the association between retinopathy of prematurity and structural brain abnormalities on term-equivalent MRI—can ROP severity serve as a marker for broader neurologic vulnerability? We also look at a randomized trial assessing the safety and effectiveness of using 100% oxygen during deferred cord clamping in extremely preterm infants. Does this strategy safely reduce early hypoxemia without increasing the risk of hyperoxia? Then, we explore early neurodevelopmental outcomes following autologous cord blood stem cell infusions in preterm infants—what signals are emerging, and how close are we to bedside applications? Finally, we unpack two diaphragm-focused investigations: one challenging the long-held practice of using rib counts on chest radiographs to estimate lung volumes, and another showing how even short-term sedation can induce measurable diaphragmatic dysfunction. These studies raise important questions—how should we adapt our practice in light of this data?



----


The articles covered on today’s episode of the podcast can be found here 👇


Fenton TR, Elmrayed S, Alshaikh BN.Paediatr Perinat Epidemiol. 2025 Jun 19. doi: 10.1111/ppe.70035. Online ahead of print.PMID: 40534585 Review.


Fenton TR, Alshaikh B, Kusuda S, Helenius K, Modi N, Norman M, Lui K, Lehtonen L, Battin M, Klinger G, Vento M, Lastrucci V, Gagliardi L, Adams M, Marba STM, Isayama T, Hakansson S, Bassler D, Shah PS; International Network for Evaluation of Outcomes (iNeo) of Neonates Investigators.Arch Dis Child Fetal Neonatal Ed. 2025 Jun 19;110(4):401-408. doi: 10.1136/archdischild-2024-327845.PMID: 39762001


Roy S, Peterson L, Kline-Fath B, Parikh NA; Cincinnati Infant Neurodevelopment Early Prediction Study (CINEPS) Investigators.J Pediatr. 2025 Jun 27;286:114711. doi: 10.1016/j.jpeds.2025.114711. Online ahead of print.PMID: 40582695 Free article.


Katheria AC, Ines F, Lee HC, Sollinger C, Vali P, Morales A, Sanjay S, Dorner R, Koo J, Gollin Y, Das A, Poeltler D, Steinhorn R, Finer N, Lakshminrusimha S.JAMA Pediatr. 2025 Jul 21:e252128. doi: 10.1001/jamapediatrics.2025.2128. Online ahead of print.PMID: 40690234


Zhou L, Razak A, McDonald CA, Yawno T, McHugh DT, Whiteley G, Connelly K, Sackett V, Miller SL, Jenkin G, Novak I, Hunt RW, Malhotra A.JAMA Netw Open. 2025 Jul 1;8(7):e2521158. doi: 10.1001/jamanetworkopen.2025.21158.PMID: 40608334 Free PMC article. Clinical Trial.


Faix RG, Laptook AR, Shankaran S, Eggleston B, Chowdhury D, Heyne RJ, Das A, Pedroza C, Tyson JE, Wusthoff C, Bonifacio SL, Sånchez PJ, Yoder BA, Laughon MM, Vasil DM, Van Meurs KP, Crawford MM, Higgins RD, Poindexter BB, Colaizy TT, Hamrick SEG, Chalak LF, Ohls RK, Hartley-McAndrew ME, Dysart K, D'Angio CT, Guillet R, Kicklighter SD, Carlo WA, Sokol GM, DeMauro SB, Hibbs AM, Cotten CM, Merhar SL, Bapat RV, Harmon HM, Sewell E, Winter S, Natarajan G, Mosquera R, Hintz SR, Maitre NL, Benninger KL, Peralta-Carcelen M, Hines AC, Duncan AF, Wilson-Costello DE, Trembath A, Malcolm WF, Walsh MC; Eunice Kennedy Shriver National Institute of Child Health and Human Development Neonatal Research Network.JAMA Pediatr. 2025 Apr 1;179(4):396-406. doi: 10.1001/jamapediatrics.2024.6613.PMID: 39992674 Free PMC article. Clinical Trial.


Spahic H, Zoubovsky SP, Dietz RM.Acta Paediatr. 2025 Jul;114(7):1742-1743. doi: 10.1111/apa.70098. Epub 2025 Apr 18.PMID: 40251839 No abstract available.


Dahm SI, Sett A, Gunn EF, Ramanauskas F, Hall R, Stewart D, Koeppenkastrop S, McKenna K, Gardiner RE, Rao P, Tingay DG.JAMA Pediatr. 2025 Jul 21:e252108. doi: 10.1001/jamapediatrics.2025.2108. Online ahead of print.PMID: 40690243 Free PMC article.


Hoshino Y, Arai J, Hirono K, Maruo K, Miura-Fuchino R, Yukitake Y, Kajikawa D, Kamakura T, Hinata A, Okada Y, Sato Y.Pediatr Pulmonol. 2025 May;60(5):e71126. doi: 10.1002/ppul.71126.PMID: 40365938


----


Watch this week's Journal Club on YouTube 👇



----


The transcript of today's episode can be found below 👇


Ben Courchia: Hello everybody, welcome back to the Incubator Podcast. We're back today for an episode of Journal Club. Daphna, how's your summer going?


Daphna Yasova Barbeau: Summer’s gone, sadly. But we had a very nice summer catching up on some of these papers that we didn’t get to earlier. We had quite a backlog, I think. Yeah, we’ll need to do an extra journal club. It looks like our colleagues are quite prolific, even over the summer.


Ben Courchia: That's right. I hope that the audience has enjoyed the series of episodes we’ve released so far over the summer, which have a very educational flavor. We’ve really tried to deliver content that is diverse and relevant both for a new crop of budding neonatologists joining the field and for everyone else as well. I enjoy these episodes tremendously, and I'm not even in fellowship, so I think that’s a good sign.

Also, just a quick mention: we had two great episodes recently in collaboration with the NEC Society. If you’re interested in learning more about NEC and fostering collaboration for research and advocacy, definitely check out the 2025 NEC Symposium. It will take place September 7–10, and you can learn more at necsociety.org/NEC-Symposium. Don’t forget to use the promo code incubator for 10% off registration.

We also have an EBNEO segment today, which I’m very excited about, part of our ongoing collaboration with the Evidence-Based Neonatology team. And one last announcement: check out our website and also DelphiConference.org, where the updated agenda for the January 26–28, 2025 conference in Florida is now live. It’s going to be very exciting, and tickets are already available.

Okay, so the first place I want to start is with the new Fenton growth curves that have been published. We use the Fenton growth curves every day in the NICU, so it’s always a big deal when an update to such a core tool comes out. These new third-generation Fenton growth charts represent a change in how we evaluate growth in preterm infants. For many years, the Fenton curve has helped us approximate what a baby’s growth would look like if they had continued developing in utero. The goal has always been to nourish preterm infants in a way that mirrors normal fetal growth, but that requires data that truly reflects healthy growth.

Previous Fenton charts—especially the second-generation edition—relied heavily on population-level data that included many infants with abnormal fetal growth (growth restriction, preeclampsia, maternal diabetes, and other complications). This skewed the lower percentiles downward, especially between 22 and 28 weeks. That’s the gap these new papers set out to address.

The first paper was published in Pediatric and Perinatal Epidemiology and provides the scientific and statistical foundation for the updated curves, describing the meta-analysis and modeling techniques used. The second paper, in Archives of Disease in Childhood: Fetal and Neonatal Edition, presents the final charts for clinical use, along with new tools and recommendations.

Together, these offer both the background and the practical implementation. The team conducted a systematic review and meta-analysis under PRISMA (Preferred Reporting Items for Systematic reviews and Meta-Analyses) guidelines, pooling data from Medline, Embase, gray literature, U.S. Vital Statistics, the iNEO Consortium, and studies from 2013 (the time of the last chart update) through November 2024.

To ensure the new charts reflected healthy fetal growth, strict inclusion criteria were used: only population-based cohorts of preterm infants born 22–42 weeks, without abnormal fetal growth. They excluded congenital anomalies, hypertensive disorders, gestational diabetes, smoking, and provider-initiated preterm births. They also only included studies with at least 500 infants born before 30 weeks. In total, seven studies spanning 15 countries and 4.8 million births were included, with 174,000 infants born before 30 weeks. Data on weight, head circumference, and length were extracted, stratified by sex, and harmonized with WHO standards using advanced modeling.

A clinically important change in the final charts: a 1.5% postnatal weight reduction was applied for infants 23–32 weeks, reflecting early fluid loss. This makes plotting SGA infants and aligning expectations more realistic.

Results:

  1. More consistent growth across percentiles. The new curves are smoother and more realistic, particularly at the lower percentiles. For example, the old 10th percentile flattened between 22–28 weeks and then steepened at 30-36 weeks, while the new one follows a more consistent trajectory. At 28 weeks, the new 10th percentile matches the 27th–29th percentile of the old chart, suggesting the old curves may have underestimated growth. In the journal article, they plotted the new Fenton chart over the old ones, and it's very interesting to see that. I'm going to point this out to you. can see for the lower percentiles, you can see that they have the older (second generation) curve that are dotted, and the newer (third generation) that have lines. You can see that it is much smoother.

  2. Growth velocity trends. Growth velocity declines steadily from 22 to 50 weeks. Weight gain falls from 21 g/kg/day (22–27 weeks) to under 6 g/kg/day by 49 weeks. Head circumference growth drops from 1.1 cm/week to 0.3–0.4 cm/week later one, and length from 1.5 cm/week to 0.7–0.8 cm/week by 49 weeks. And that was a consistency that was observed in the 3rd, 50th, and 97th percentile for both sexes, confirming that the chart represent probably a more proportional predictable growth across gestational ages.

  3. Alignment with fetal ultrasound estimates. The 50th percentile on the new charts actually aligns closely with fetal ultrasound standards (Hadlock, WHO, Nicolaides), reinforcing the goal of mirroring fetal growth.

So in conclusion, the third-generation Fenton curves provide a more accurate representation of preterm growth by excluding infants with abnormal fetal growth. They show improved growth velocity patterns and align better with fetal ultrasound estimates. PediTools has already updated their website with the new charts, so if you use paper charts in the unit, it’s time to print new ones!


Daphna Yasova Barbeau: I miss the paper growth charts! You could really see it in real time. And it’s such a valuable tool for families when you’re rounding and have the growth chart right there.


Ben Courchia: Exactly. We’re lucky to have a dedicated neonatal dietitian—Rebecca—who sits down with parents to go over the growth curves and explain nutritional adjustments. Parents love it, because it mirrors what they do at the pediatrician’s office. It’s not always easy for us to find time to do this ourselves, so having Rebecca is a luxury.


Daphna Yasova Barbeau: Yeah, if you’re lucky enough to have a Rebecca, then good for you. That’s the take-home point here: you can get the new graphs, or you can get a Rebecca!

Okay, well thank you for sharing that. I’m going to do an article you actually highlighted for me. It’s in the Journal of Pediatrics and it’s entitled: Retinopathy of Prematurity and Risk of Structural Brain Abnormalities on MRI at Term Among Infants Born at ≀32 Weeks of Gestation. And now you know everything the paper is about.


Ben Courchia: The association is not something you would expect. It’s quite intriguing and makes you want to dive into the paper.


Daphna Yasova Barbeau: Exactly. Lead author Shalini Roy, senior author Nahal Parikh, from the Cincinnati Infant Neurodevelopmental Early Prediction Study (CINEPS) Investigative Group. There’s been a lot of discussion, including animal studies, about whether retinopathy of prematurity (ROP) is associated with neurodevelopmental outcomes. Is one causing the other? Is it just abnormal maturation? Many hypotheses exist. So, the goal of this study was to determine if ROP severity is associated with brain abnormalities on MRI at term-equivalent age in infants born preterm.

This was a prospective cohort study of 395 infants born 23+0 through 32+6 weeks, between September 2016 and November 2019, across five NICUs. Exclusions: chromosomal or CNS congenital anomalies, cyanotic heart disease, or if still hospitalized/ventilated on >50% oxygen at 44 weeks PMA. All infants underwent routine ROP screening, classified as no ROP, any ROP (stage 1–5), or severe ROP (stage 3–5 or type 1 requiring treatment).

All infants then underwent MRI at 39–44 weeks PMA. Brain abnormalities were scored using the Kidokoro Global Brain Abnormality Score (GBAS). Subcomponents include cortical gray matter, cerebral white matter, deep gray matter, and cerebellum (each subcomponent was rated on a scale of 0 = normal, 4 = severely abnormal). A total score of 0–3 is considered normal, 4–7 is mildly abnormal, 8–11 is moderately abnormal, and ≄12 is severely abnormal.

They also controlled for many confounders: chorioamnionitis, hypertensive disorders, smoking, gestational age, neonatal comorbidities (grade 3 or 4 IVH, PVHI, cystic PVL, SGA, NEC, SIP, sepsis).


Ben Courchia: Yeah, it’s obviously a challenge because ROP is a marker of disease severity in general. To tease both things apart is not easy.


Daphna Yasova Barbeau: No, it’s not. But I think they did a good job identifying those characteristics and trying to account for them.

They found that 66% had no ROP, 34% had any ROP, and 4.8% had severe ROP. Of the severe ROP cases: 3 had retinal ablation, 6 received anti-VEGF, 1 timolol eye drops. No scleral buckle or vitrectomy needed. Infants with any ROP were more likely to have chorioamnionitis (22% versus 10%), lower GA, early white matter injury (18% versus 5%), sepsis (22% versus 5%), and NEC/SIP (10% versus 3%).

Any ROP was associated with a 1.5-point increase in global brain abnormality score. 1.5-point increase is meaningful—it moves a baby halfway toward the next severity category! Severe ROP was associated with a 2.3-point increase in GBAS. In babies <30 weeks GA, severe ROP remained significant, but any ROP did not. Sex and outborn status did not meaningfully affect any relationships.

Any ROP and severe ROP exhibited a significant relationship with higher deep gray matter and cerebellar scores: any ROP increased the scores by +0.7 and +0.5 respectively; severe ROP increased by +0.9 and +0.9 respectively. Interestingly, neither any ROP nor severe ROP was associated with total white matter or cortical gray matter scores. That was interesting, since those are the areas we often think of when we see preterm brain injury (e.g., IVH, PVL).

So, overall, their message was that there were associated factors (chorio, lower GA, sepsis, NEC/SIP) with ROP. ROP is associated with increased brain abnormalities, and severe ROP even more so. But the question remains—is it cause, consequence, or just a marker of underlying disease severity? It’s the chicken-or-egg problem. But either way, it’s a strong signal that these babies are at risk for more global issues.


Ben Courchia: Yes. At the very least, it highlights that ROP severity may help us identify babies at risk for adverse neurodevelopmental outcomes and that they may benefit from closer follow-up.

The next article I’d like to review looks at delayed cord clamping with high oxygen in extremely preterm infants. I think it’s an important and evolving question in neonatal resuscitation: what is the optimal concentration of inspired oxygen that we should be using during deferred cord clamping, or DCC as it is often known, in extremely preterm infants? Neonatal resuscitation guidelines currently recommend initiating resuscitation of extremely low gestational age neonates, often known as ELGANs, with about 21-30% oxygen to minimize the risk of oxygen toxicity. The problem that has arisen in recent years is that we have increasingly adopted the practice of doing delayed cord clamping. There is a delay in administering any form of oxygen to these babies during DCC. Data from NICUs in California, where resuscitation typically begins with 30% oxygen, shows that infants with low oxygen saturation are at increased risk of death or severe IVH, and for the combined outcomes of early death and severe IVH. There’s a meta-analysis of individual patient data of infants less than 32 weeks, which we reviewed on the podcast, that found lower mortality rates in those resuscitated initially with high oxygen saturation.

This body of evidence prompted ILCOR, the International Liaison Committee on Resuscitation of Neonatal Life Support Task Force, to suggest that it is reasonable to begin resuscitation of infants born preterm before 32 weeks of gestation with a relatively higher oxygen concentration. This matters because during DCC, the use of 100% oxygen may actually cause pulmonary vasodilation without inducing systemic hyperoxia, as the pulmonary venous return mixes with the umbilical venous blood. This could enhance pulmonary blood flow and oxygenation.

While we know that DCC reduces mortality, the lack of effective breathing before cord clamping remains a risk factor for poor outcomes. This is the window that has opened up with the acceptance of DCC as a quality measure. Physiologic evidence suggests that to overcome hypoxia in this population, we may need both positive pressure and high oxygen.

This study aimed to evaluate whether using 100% oxygen during DCC in infants born between 22 and 28 weeks would reduce the incidence of hypoxemia by five minutes of life, defined as having a saturation of less than 80%, compared to the use of just 30% oxygen. It also looked at the safety and feasibility of delivering oxygen during DCC, especially during C-sections.

The study author was Dr. Anup Katheria, who has been on the podcast before. It was a double-blind randomized clinical trial involving preterm infants born between 22 weeks and 0 days and 28 weeks and 6 days at birth. The study was conducted across hospital centers in California from November 2021 to October 2024. Participants were randomized prior to delivery into one of two groups. One group received 90 seconds of DCC with 100% oxygen, considered the high oxygen group, while the other received 30% oxygen, known as the low oxygen group. Both groups received respiratory support via masks, CPAP, or positive pressure ventilation during DCC, using a Neopuff.

They had some blinding by covering certain settings on the Neopuff. There were several exclusion criteria: pregnancies with membrane rupture before 20 weeks, monochorionic twins with twin-to-twin syndrome, placenta accreta, known congenital anomalies, etc. They used a “Life Start” bed, basically a table where resuscitation could be performed at the bedside, with a chemical warming mattress and sterile drapes. Everything, including CPAP oxygen tubing in a sterile sleeve, was set up before the OB began.

Infants remained attached to the umbilical cord for a total of 90 seconds, receiving stimulation as neonatal resuscitation recommendations state, followed by CPAP for approximately 60 seconds. No changes were made to the oxygen concentration or pressure settings during the intervention. After the cord was clamped and cut, the infant was taken to a separate warmer where resuscitation continued. The initial setting on that warmer was always 30%, with titration up or down at the clinician’s discretion.

The primary outcome of the study was the incidence of hypoxemia at five minutes of life, defined as sats below 80%. Exploratory outcomes included heart rate, oxygen saturation in the first 10 minutes, need for intubation in the delivery room, surfactant administration, chest compressions, epinephrine use, hyper- or hypoxemia direction, cerebral oxygenation, hemodynamic support, and several short-term morbidities like IVH, NEC, PDA, and bronchopulmonary dysplasia.

A total of 140 infants were included in the study: 68 in the high oxygen group and 72 in the low oxygen group. The average gestational age was 26 weeks. Because this was a prospective randomized study, baseline characteristics were well-matched.

For the primary outcome, 39% of infants in the low oxygen group achieved saturations of 80% or higher by five minutes, compared to 69% in the high oxygen group—a 30% difference. This yielded an adjusted odds ratio of 3.74 with a 95% confidence interval of 1.8 to 7.8, and a p-value of less than 0.001.

Secondary and exploratory outcomes: the time to reach sats of 80% was shorter in the high oxygen group, four minutes versus six minutes. Mean oxygen saturation in the first five minutes was 69% in the high oxygen group versus 67% in the low oxygen group. Heart rate in the first five minutes was similar. The proportion of infants breathing before cord clamping was comparable, and rates of CPAP and PPV before and after cord clamping were similar.

Delivery room interventions: intubation rates were similar, 38% and 39%. Surfactant administration was also similar, 24% and 28%. A few infants required chest compressions and epinephrine: two in the high oxygen group received chest compressions and one received epinephrine. None in the low oxygen group received these interventions.

As for NICU outcomes, severe IVH occurred in three infants in the high oxygen group and six in the low oxygen group. Death before 40 weeks was similar. Grade 2–3 BPD was lower in the high oxygen group, 22% versus 39%, but not statistically significant enough to make a recommendation. PDA requiring pharmacological treatment was 31% in the high oxygen group and 44% in the low oxygen group. Rates of NEC, late-onset sepsis, and ROP requiring surgery were similar. There was no significant difference in cerebral tissue oxygenation or systemic hyperoxia during the first 10 minutes or the first 24 hours of life. Importantly, despite transient use of 100% oxygen, there was no signal of increased systemic oxygen exposure, hyperoxia, or associated complications.

The authors concluded that using 100% oxygen during DCC significantly reduced early hypoxemia without increasing systemic hyperoxia or adverse outcomes. These findings support the potential revision of resuscitation guidelines to incorporate higher oxygen concentration during DCC for extremely preterm infants, and further larger multi-center trials are warranted to validate these results and assess short- and long-term outcomes.

 

So, it's happening. We're going to get into that sterile field. It's exciting.


Daphna Yasova Barbeau: The other night, in the middle of the night, I attended a C-section and I was like, oh I'm so glad I don't have to gown up
but, but if it helps the babies, then I’ll do it!


Ben Courchia: All right, we're going to take a quick break and then we will be back with Robert Dietz from the EBNeo community to review the paper that we've talked about already on the podcast, but has made the rounds about whole-body hypothermia for neonatal encephalopathy in babies born between 33 and 35 weeks.


--------


Okay, we are back today for a new EBNeo segment. We're always happy about our partnership with the EBNeo community that identifies an article or manuscript most likely to impact clinical care. Members of their team devote a significant amount of time reviewing that paper and writing a thorough commentary. We are joined today by Dr. Robert Dietz. Robert, welcome to the program.


Robert Dietz, MD, PhD: Hey, thanks, Ben. Glad to be here.


Ben Courchia, MD: Robert is a neonatologist from the Department of Pediatrics at the University of Colorado in the US. Robert, you're reviewing for us today a paper that has really made the rounds and created a lot of buzz. It's a paper published in JAMA Pediatrics called Whole Body Hypothermia for Neonatal Encephalopathy in Preterm Infants 33 to 35 Weeks Gestation: A Randomized Clinical Trial. We reviewed this paper on the podcast before. But for the people who can't remember the details of the trial so much, can you give us a quick overview of the data presented?


Robert Dietz, MD, PhD: Yeah, sure. These authors compared whether babies between 33 and 35 weeks who met criteria for therapeutic hypothermia—based on fairly standard HIE criteria—benefited from therapeutic hypothermia started within six hours of birth, compared to babies who were not cooled. The main outcome was whether hypothermia reduced the risk of death or disability at 18 to 22 months. This was a randomized control trial with about 88–90 patients. Demographics were pretty equal between groups.

The biggest buzz in the field came from the 35-weekers. They showed that in 33-weekers there was a signal for increased death with cooling, and this was also seen in 34-weekers. In 35-weekers, they had 48 patients: 28 randomized to hypothermia, 20 to normothermia. They found about 25% of cooled babies at 35 weeks died or had moderate-to-severe disability, compared to 20% with normothermia. For death alone, it was 18% in the hypothermia group versus 15% in the normothermia group.


Ben Courchia, MD: Right. The paper, as you said, was a randomized trial with 88 babies in the hypothermia group and 80 in the normothermia group. One of the interesting aspects was the Bayesian approach. They presented information not only as relative risk with confidence intervals, but also probabilities of benefit versus harm. For the primary outcome—combined outcome of death or moderate-to-severe disability—the probability of harm was about 75%. For mortality alone, the probability of increased death with hypothermia was about 87%, with only a 13% probability of benefit. Like you mentioned, it’s rare that one of the supplemental tables gets so much attention. But here it did. It broke down outcomes by gestational age—33, 34, and 35 weeks—looking separately at combined outcomes and mortality alone. The rates of death were especially higher in the 34-week group. The 33-week group had too few babies to draw strong conclusions, and the 35-week outcomes were similar across groups. So, while many thought 35-week babies might be safe to cool, this paper raises doubts. I’m curious—what were your main clinical practice takeaways in your commentary?


Robert Dietz, MD, PhD: Yeah, even beyond the "therapeutic creep,” it’s actually an AAP guideline to cool 35-week babies and above. That’s what makes this paper so important. In Colorado, we changed our inclusion criteria from 36 to 35 weeks back in 2016, based on ICE trial data that included down to 35 weeks. The AAP likely used that data in their recommendations. But when you look back, the ICE trial only had a handful of late preterm babies, so the field shifted based on very little evidence.

This trial suggests we may need to rethink. Late preterm babies may be at higher risk for harm due to factors like immature epidermal barrier, higher surface-area-to-weight ratio, and greater susceptibility to cold stress, which could lead to worse outcomes.

Another under-discussed point in this paper is overcooling. About 35–40% of babies randomized to hypothermia were cooled below 32°C. They were using older blanket-control systems rather than the servo-controlled systems most units use now.


Ben Courchia, MD: Exactly. We know overcooling is dangerous, which is why maintaining a tight temperature range is critical. That could very well have played a role here.


Robert Dietz, MD, PhD: Yes, and that’s true across all gestational ages, even full-term babies. I’m very interested in neuroscience (I have a PhD in neuroscience). And based on this data, many groups are rethinking their cutoff—whether to keep cooling at 35 weeks or move to 36 weeks and above. I think the AAP is actively reevaluating their guidelines as well.

That said, this trial still had small numbers—25 babies in one arm and 20 in the other for the 35-week group. Some had temperature management issues. In Colorado, we’ve decided to continue cooling 35-weekers, until the AAP updates recommendations or new data emerges. We’re also part of the CHNC, so we are looking at our larger consortium data—several thousand babies cooled over the last 15 years—to better understand outcomes.


Ben Courchia, MD: That’s really valuable. The CHNC work you’re doing is important because every detail in the cooling process matters—how quickly babies are cooled, whether within an hour or nearly six hours later. And as you said, with such small numbers in this trial, it’s hard to know if that’s enough to change practice.

What I appreciated in your commentary was your effort to connect the findings with possible physiologic mechanisms, like immature skin barriers, higher surface-area-to-weight ratio, and higher cold stress in late preterms. That kind of thinking could guide tailored protocols. Do you see the future of hypothermia including different protocols for late preterm versus full-term babies?


Robert Dietz, MD, PhD: I think so. There’s growing interest in late preterm babies, as well as in mild HIE. We’re moving beyond the original randomized trial populations, and thoughtful people are asking whether these groups should be managed differently. Targeted protocols could certainly be the way forward.


Ben Courchia, MD: Yeah, because their tolerance for acidosis varies so much by gestational age. A 23-week baby tolerates acidosis differently than a 39-week baby, and that should inform how we approach management.

This was a fascinating discussion, Robert. The commentary you co-wrote with Harisa Spahic and Sandra Zoubovsky is titled Is Therapeutic Hypothermia Beneficial to Infants Born Between 33 and 35 Weeks of Gestation? It’s available now on Acta Paediatrics. Thank you so much for your time and for sharing your insights.


Robert Dietz, MD, PhD: It’s been a pleasure. Thank you, Ben, for inviting me.


-----------------------------------


Ben Courchia: All right, we are back for the last half of Journal Club. Daphna, where are you taking us next?


Daphna Yasova Barbeau: Well, I'm taking us to another exciting area of opportunity. This was in JAMA Network Open Pediatrics called Early Neurodevelopment of Extremely Preterm Infants Administered Autologous Cord Blood Cell Therapy. This is a secondary analysis of a non-randomized clinical trial, the Cord Safe Study. Lead author was Lindsay Zhao and senior author Atul Malhotra. Lots of exciting new studies about stem cell therapy are coming from down under.

The Cord Safe Study was basically a pilot to look at safety outcomes for preterm infants receiving autologous (their own) cord blood stem cells. They then wanted to do a short-term follow-up looking at what they called “early neurodevelopment.” So again, it was a phase one trial. They wanted to look at feasibility and safety of this cord blood cell infusion, which we have discussed previously on the podcast. The babies enrolled were extremely preterm infants born before 28 weeks gestation and surviving at least two weeks of postnatal life. Parents obviously had to consent to participate. An important detail: they had to have adequate good-manufacturing-practice-grade autologous umbilical cord blood cells. The blood cells were collected, analyzed for quality, and only if they met standards were they used for infusion.


Ben Courchia: Yeah, that's a big concern. Just because you have stem cells doesn’t mean they’re good cells. They could be bad cells.


Daphna Yasova Barbeau: That's right. We described this before—maybe we can find the episode number—but the Cord Safe Study detailed all the criteria to ensure there were enough stem cells and that they were high enough quality.

Infants were excluded if they were outborn, if less than 7 mLs of cord blood were collected, if the cells weren’t GMP (Good Manufacturing Practices)-grade, if they had severe IVH reported by day 8 cranial ultrasound, or if they had culture-positive sepsis within 48 hours of the planned infusion.

If infants met all inclusion criteria between postnatal days 9–15, they were eligible for the autologous umbilical cord blood cell (UCBC) infusion. The cells were processed and prepared to a final volume of 10 mL/kg for infusion. They compared these babies to a contemporaneous cohort (all infants born <28 weeks during the study period who were not enrolled), using the same exclusion criteria.

The main outcomes were “early development,” not long-term follow-up. They used term-equivalent MRI brain scans and applied the Kidokoro brain injury score. Quick reminder that Kidokoro brain scores uses subscores of different brain areas, then is summed for a score: normal is 0–3, mild 4–7, moderate 8–11, severe ≄12.

They also did early neurodevelopmental assessments: general movements assessment (GMA), the Hammersmith Infant Neurological Exam (HINE), and hearing tests—all at post-menstrual age 52–54 weeks. Importantly, when MRI, GMA, and HINE are all abnormal, sensitivity for detecting cerebral palsy is 98%.

This was a small group, not powered to detect changes in neurodevelopment—it was really a pilot safety trial. Twenty-three infants received cord blood infusion, with a median gestational age of 26 weeks and median birth weight 748 grams. The contemporaneous cohort had 93 infants, median gestation also 26 weeks, median birth weight 769 grams. There was one death in the infusion group (from bacterial sepsis, not related to infusion) and two in the control group.

MRI brain injury scores: median of 2 for the cord blood group, 3 for controls. Adjusted analysis showed no statistically significant differences. Both groups fell within normal or mild categories. Brain structure measurements were also not significantly different.

HINE scores: median 57 for infusion group vs. 59 for controls—not significant. Absent fidgety movements: 4.5% in infusion group vs. 11.5% in controls (not statistically significant, though seemingly notable). No infants in the infusion group were assessed as high risk for CP, while 7% of the control group were. Again, not statistically significant, but notable.

In summary: this study wasn’t powered to detect neurodevelopmental differences, but follow-up is ongoing. Participants will undergo standardized assessments at 2–3 years using Bayley-IV. The next step is the Cord Cell Randomized Controlled Trial, randomizing infants to receive UCBC or not, looking for long-term neuroprotection.


Ben Courchia: I really appreciate you covering this paper. I think this is still a frontier of neonatology. The buzz about stem cells was probably too early 10 years ago—it was far from bedside reality. But now I think we are reaching the point where stem cell infusion will start happening at the bedside.

We strongly believe this is the next frontier. Dr. Malhotra in Australia has been at the cutting edge of a lot of this work. It took a lot of resources to bring him to Delphi this past year, and his talk on neonatal cell therapies—“Translating Promise Into Benefits”—was super interesting. He covered all the projects they’re working on. It’s striking to see how close they are to clinical applications.


Daphna Yasova Barbeau: Yes, and this is just one clinical application. Next steps will look at babies with IVH and with HIE. Lots of opportunities there.


Ben Courchia: Exactly. He mentioned the Cord Safe trial in that talk, as well as using umbilical cord blood cells for preterm lung injury. They’re involved in multiple trials. At Delphi this year, Dr. Bernard Thibault from Toronto will also update us on stem cell therapy. I’ll link Dr. Malhotra’s talk in the episode show notes—it’s only 20 minutes, very high quality, easy to stream. His wife, who’s an OB, also gave a great (and very different) talk.


Daphna Yasova Barbeau: Yes, they were wonderful—two very different but equally captivating talks. Opposites attract, but they were delightful.


Ben Courchia: They have very different personalities. He came across as this introverted bench scientist—eloquent but measured—while she was extroverted and full of performance energy. She’s an obstetrician, so that fit her style perfectly.


Daphna Yasova Barbeau: We were very lucky to have them both. Definitely a highlight of Delphi 2024.


Ben Courchia: My last paper for today is also coming from JAMA Pediatrics, and it's called Diaphragm Position on Chest Radiograph to Estimate Lung Volumes in Neonates. I was very interested in this paper. The first author is Sophia Dahm, and the introduction is very interesting because it talks about how many of us in neonatology have accepted the routine clinical practice of using the diaphragm position on chest radiograph to assess lung volume.

How many times have you looked at a chest X-ray and counted the ribs and thought: hypoinflation, hyperinflation? I do it all the time. It's funny because the introduction talks about this.


Daphna Yasova Barbeau: I counted them today. Multiple times. I almost felt a little offended with this intro! Like, okay, so show me what else to do!


Ben Courchia: This approach of counting ribs—the number of visible posterior ribs to estimate aeration—has been part of our toolkit for over a century and is embedded in numerous NICU guidelines. Typically, seven to nine ribs suggest normal aeration, fewer than seven is hypoinflation, and more than nine is hyperinflation. I do this all the time, by the way. I'm not being judgmental—I count ribs several times a day on a lot of babies.

However, despite how widespread this practice is, it has never actually been validated in infants. In fact, only one small study previously looked at this question using gas washout, and it found no meaningful association. Since methods like spirometry and plethysmography are too invasive or complex for bedside use, many have defaulted to the diaphragm position as a simple heuristic. But how reliable is it?

The authors of this study set out to answer that question by comparing diaphragm position to CT images looking at derived lung volumes—a more precise method of estimating aerated lung tissue. The primary objective was to describe the association between diaphragm position on chest radiograph and CT-measured total lung volume in infants. This was a retrospective cross-sectional study conducted at the Royal Children's Hospital in Melbourne, Australia. They included infants who had received a chest CT in the first 30 days after birth, between 2012 and 2022. Infants with congenital lung anomalies were excluded.

To estimate diaphragm position, they didn’t actually use chest X-rays. Instead, they used CT topograms, which they described as CRE (Chest Radiograph Equivalent). The CT topogram is the frontal scout image prior to the CT scan; they are slightly different from traditional CXR in terms of image resolution and technical factors, but they are considered reliable for assessing diaphragm and rib anatomy. This was clever, since it avoided exposing the baby to yet another X-ray.

They defined diaphragm position as the last posterior rib that completely intersected or contained the diaphragm. If the diaphragm was between ribs, they applied a rule based on how much of the intercostal space was filled with lung fields. Lung volumes were then calculated from the CT scans using semi-automated segmentation software and standardized to body weight. Lung volume was standardized to body weight and expressed in milliliters per kilogram. They also computed Hounsfield units for tissue density and assessed consolidation subjectively. The investigators were blinded to diaphragm measurements and CT results to avoid bias.

The primary outcome was the association between diaphragm position (from the 6th to 11th posterior rib) and total lung volume. They used Kendall's Tau to assess correlation, defining anything below 0.4 as weak.

Results: of 292 eligible CT scans, 218 infants were included. The median age was 11 days; most were term infants. The average weight at the time of the CT scan was about 3000 grams. 61% had a cardiac diagnosis as their primary diagnosis, which as we know, is not your routine chorio admission. 35% were intubated, and 61% had some degree of parenchymal lung opacification or consolidation.

Diaphragm position ranged from ribs 6 to 11, with 95% falling between 8 and 10. In over half of infants, both diaphragms were at the same rib level; in 35% the left was lower, and in 10% the right was lower.

The correlation between diaphragm position and lung volume was weak: Kendall's Tau was 0.23. Even excluding infants with consolidation, correlation only rose to 0.3. Left and right sides separately remained weak at 0.25 and 0.21. So we understand how to interpret these numbers, the Kendall's Tau test is a non-parametric statistic tool used to measure the strength of and direction of an association between two variables. So a perfect association would have been 1. Perfect disagreement is a minus 1. And 0 means no association. The study had defined that 0.4 was going to be their threshold. So basically all the numbers I've given you have fallen below that threshold of 0.4. So there is an association, but it's not very strong. It's actually quite weak.

They were also looked at several secondary analyses comparing the Hounsfield unit, which indicate the lung density and the correlation with diaphragm position was also very weak. The apex diaphragm vertical distance, another potential marker for lung expansion, had poor correlation with volume. There was no notable difference when the diaphragm position was measured from either the lower or the higher hemithorax. The range of total lung volume at any rib level was also broad (20–60 mL/kg), making it difficult to use the diaphragm position as a reliable estimator in any practical way.

So the results of this cross-sectional study suggest that despite long-standing clinical acceptance, the diaphragm position has measured through the number of posterior rib on a chest radiograph is not sufficiently accurate for the use in practice as a surrogate of long volume. They caution against the use to guide respiratory support decision until alternative non-invasive methods are available, like maybe electrical impedance tomography.

In the discussion, I think they talk about something that I believe is quite interesting because it makes you think a little bit about the chest X-ray and it makes you think a little bit about what is the information that we're getting. They're really talking about the fact that when we're looking at the CXR, we're looking at like an AP film. You're looking at like the degree of change in area from top to bottom, but babies are lying flat and the expansion going from posterior to anterior is probably more important. I'll quote from the discussion: “Fundamentally a frontal chest radiograph is a two-dimensional coronal image of the lungs, leaving the expansion of the lungs in the anterior-posterior direction unable to be examined. The more compliant and flexible infant thorax allows greater displacement along the non-coronal plane compared with adults. Infants also perform more abdominal breathing and maintain a tonic diaphragm during respiration.” And so they're saying that this highlights the many factors that contribute to radiographic appearance of the lungs at any point in time, beyond the level of respiratory support.

It's a very interesting conversation. I love neonatal pulmonology. So I think this is all very interesting.


Daphna Yasova Barbeau: Okay, well, I have another diaphragm paper. My last paper today is in Pediatric Pulmonology, from Japan. They looked at the impact of sedation on ventilator-induced diaphragmatic dysfunction in extremely preterm infants. I thought this was interesting for two reasons. First, it’s well known in adults and older children that ventilation can lead to diaphragmatic dysfunction—a decrease in thickness or contractility that prolongs ventilator dependence. We suspect this in babies too: the longer you're on ventilation, the longer you stay on ventilation. Second, they looked at sedation use. In the U.S., we've moved away from using a lot of sedation in very small babies, but that’s not universal.

This was a prospective observational study at a single level III NICU (July 2020–September 2023). Infants <28 weeks were recruited within six hours of birth, but only during working hours (8 a.m.–6 p.m. - good for them!) Babies who had already received sedation prior to their first diaphragm ultrasound were excluded, as were those with chromosomal abnormalities, major malformations, congenital lung or heart disease, or anticipated extubation within 48 hours.

Their general respiratory strategy was non-invasive or CPAP support first. Their goals were pH 7.25-7.35, oxygen sats 88-94%, and CO2 of 40-60. Intubation was reserved for persistent apnea, high effort, or gas exchange failure. Once ventilated, they used tidal volumes of 4–6 mL/kg. Sedation wasn’t routine, but when used, fentanyl was the first choice—often to stabilize circulation or prevent IVH in <24-week infants during the first 72 hours. They also used sedation when desaturations were thought to be due to agitation or supplemental oxygen >40% was thought to be due to PPHN.

Each baby got 3 different ultrasounds by 3 different physicians. They performed diaphragm ultrasounds in the supine position with quiet breathing. They measured thickness of the right diaphragm at end inspiration (TDI) and end expiration (TDE), and calculated the diaphragm thickening fraction (DTF = (TDI – TDE)/TDE × 100%).

They obtained the first ultrasound as the baseline measurement performed after the initial respiratory steps (intubation, surfactant, etc.). The second diaphragm ultrasound was performed within a 24 (plus or minus four) hours, because they felt like some babies would already have been developing this diaphragmatic dysfunction.

The main outcome of the study were the differences in change in the TDE over time, the TDI over time, and in the diaphragm thickening fraction over time from day zero to day one. They had 52 babies admitted to the unit, 49 of them were intubated after birth, 12 were admitted outside of working hours, three had major abnormalities, and four were extubated within 40 hours. So they had 30 patients included in the study, 13 of whom received continuous fentanyl on day one, so that was the sedated group, and then they had 17 without sedation, the non-sedated group.

Baseline measurements showed no differences between sedated and non-sedated groups. But over 24 hours, sedated infants had a significant decrease in all three parameters (TDI, TDE, DTF). Non-sedated infants also had decreases from day zero to day one, but the sedated group’s decline was greater.

The sedated group was also younger (median 24.4 weeks vs. 26.3) and smaller (646 grams vs. 791 grams), so that’s a consideration. Still, sedation was associated with greater diaphragmatic dysfunction after just one day.

Take-home: sedation in preterm infants can worsen diaphragmatic dysfunction—an underrecognized complication. Even a single day of ventilation changes the diaphragm, so we should be cautious about "just one more night" on the ventilator.


Ben Courchia: Yeah, we don’t tend to appreciate the impact of our interventions at the organ or cellular level. I think this is such a good point. I've always been a bit skeptical about point-of-care lung ultrasound, but this application is interesting.


Daphna Yasova Barbeau: What, you don’t like lung ultrasound?


Ben Courchia: I don’t mind it. I just think it should provide new or superior information. Some of its uses, like for pneumothorax, I felt weren’t gaps we desperately needed to fill. Like we could just see it on CXR or transillumination. But using ultrasound to look at the diaphragm—that’s useful.


Daphna Yasova Barbeau: Yes, especially for predicting extubation success. Knowing more about diaphragmatic function or dysfunction is exciting.


Ben Courchia: And like you said, we’re working hard not to paralyze or oversedate our infants, and this supports that approach. It’ll be interesting to see if our colleagues in Japan adjust their practice based on this.


Daphna Yasova Barbeau: Yes. They seem very comfortable with diaphragmatic ultrasound, so hopefully we’ll see more studies.


Ben Courchia: Okay, Daphna, I think that does it for us. Thank you for listening and subscribing. We’ll see you next time on the Incubator Podcast.

Comments

bottom of page