A New Publication

Janvier A, Farlow B, Barrington KJ. Cardiac surgery for children with trisomies 13 and 18: Where are we now? Seminars in Perinatology. 2016. Annie and Barb and I have just published this review of the current situation regarding cardiac surgery for infants with the aneuploidies in the title. (For those who don’t know, Barb Farlow is a mother of a child, who has sadly died, who had trisomy 13.) In addition to discussing the literature we make some recommendations about how to help families making difficult decisions. The literature review is limited by the common practice of lumping together all kinds of cardiac surgery, from ASD repairs to Fontan procedures and Norwoods. Also few papers describe whether there was withholding or withdrawal of life-sustaining interventions at any time.

Nevertheless, it is clear that prolonged survival after cardiac surgery is possible.

Our hope is that we will move away from the universal denial of cardiac surgery, which still occurs in some places. On the other hand it is not appropriate, for any patient, to ignore their other problems pretending they don’t exist, infants with aneuploidies often have other difficulties, which may be serious, but which are very variable. Only by making a decision with the parents which is medically appropriate, likely to improve their life, and consistent with the families values, will we be able to truly act in their best interests.

We offer some guidance which will hopefully be helpful, this is an abbreviated version of our suggestions.

(1) Words are important, avoid these words: “vegetable,” “futile,” “lethal,” “incompatible with life,” “waste of time/resources/emotions,” “Tee 13 or 18,” “it, that,” “product of conception,” “hopeless,” “nothing we can do.”

(2) Avoid unfounded and biased predictions: do not tell parents that this child will have a negative impact on their family, their other children, or their lives.

(3) Recognize that each child and family is unique

(4) Provide balanced counseling: which includes recent medical information, as well as the experience of families who live(d) with children with these conditions.

(5) Personalize the information: all infants are different, even if they carry the same label.

(6) One step at the time: decisions about potential complex cardiac interventions need not be discussed and decided upon prenatally, but be forthright about care options for stable children.

(7) Provide parents with different possible outcomes for their fetus/child (including stillbirth): emphasize the spectrum and uniqueness of each child and the lack of control you and they have on many of these outcomes.

(8) Empower parents: recognize that this diagnosis is difficult for parents, that they are doing their best.

(9) Offer reasonable hope:  hope that their child might defy the odds and hope that her life, however long, will have an enriching and positive influence on all who love her.

(10) Support parents: ask them how you can help them.

We hope that families facing these decisions will find caregivers who are sensitive to their needs, willing to consider the options, and supportive.

Posted in Neonatal Research | 2 Comments

The death knell for Xenon?

Azzopardi D, et al. Moderate hypothermia within 6 h of birth plus inhaled xenon versus moderate hypothermia alone after birth asphyxia (TOBY-Xe): a proof-of-concept, open-label, randomised controlled trial. The Lancet Neurology. 2016;15(2):145-53.  Babies who undergo therapeutic hypothermia for perinatal encephalopathy are still at high risk of significant long term impairments. Other therapies to add to hypothermia are being sought and tested, one of them being Xenon. Inhaled xenon gas has neuroprotective effects in many models; but it is expensive and difficult to use, in order to make it affordable for use over several hours you need to recirculate the exhaled xenon, so you need a special ventilator, which has been developed for this trial. In this “pilot’ RCT, eligible babies were typical of infants who are cooled, and had to have started on hypothermia within 6 hours after birth. As is usual, most of the 92 enrolled babies (2/3) were born in peripheral hospitals and transported in, many babies were cooled quite quickly, 93% before 4 hours of age. Xenon or standard care was started after randomization which was after arrival in the study centre, so the assigned treatment didn’t start until an average of 10 hours of age.  Xenon (or no xenon) was then continued for exactly 24 hours of age.

The primary outcomes of the study were MR findings; using spectroscopy they calculated the ration of lactate to N-acetyl aspartate in the thalamus, and using diffusion tensor imaging they calculated the fractional inosotropy of the posterior limb of the internal capsule. Scans were performed after the end of cooling at about 6 days of age. Because of deaths and a small number of scans not done in survivors, they ended up with around 75 babies with data for each of the two primary outcomes, data from the MRI were analyzed by a masked individual (images of the lone ranger… radiologist) . Basically the study showed no effect of Xenon.

Which is a bummer.

Why didn’t it work? I think first off we have to be careful in saying it didn’t work, there was no effect on the primary outcomes, but the primary outcomes are surrogates. Surrogates should always be mistrusted, even when they are called “biomarkers”. Is the surrogate an accurate enough predictor of good or adverse clinically important outcomes? I think that is questionable here, mostly because I don’t know the data well enough to answer the question, but is it possible that a clinically significant benefit of xenon will be shown if (hopefully when) these babies are followed up? My guess is that such an outcome is quite unlikely, but possible. In fact I think this study is a good opportunity to prove the value of the MR surrogates. If the authors are right (and usually Dennis Azzopardi, Dave Edwards and the many associated luminaries who wrote this article are indeed right) then using similar surrogates in future trials will help to screen for effective adjunctive therapies in cooled babies, and more quickly than waiting 2 years or so for follow up.

Maybe starting xenon at 10 hours of age is just too late? As the authors point out, they performed a trial in a real world environment, it would be possible, if you had the ventilator always ready and available, to start Xenon the moment a baby enters the referral NICU, but that would still lead to significant delays of evaluation and transport. Maybe 24 hours is too short? It was based on the best previously available literature, and again technically feasible, before doing another study with longer Xenon administration I think we would need some very good rationale.

In the end, this real-life application of xenon in cooled babies didn’t show any sign of being effective. We should look elsewhere I guess, something that could be given very quickly when a baby is cooled, such as melatonin, or erythropoietin look like they are the most worthy of further investigation. A review of the literature from 3 to 4 years ago concluded that, and I haven’t seen much to change the situation since then. Robertson NJ, et al. Which Neuroprotective Agents are Ready for Bench to Bedside Translation in the Newborn Infant? The Journal of pediatrics. 2012;160(4):544-52.e4.

Posted in Neonatal Research | Tagged , | 4 Comments

Surfactant without intubation, where are we?

A recent trial from Germany tried to answer the question of whether giving surfactant through a thin catheter while the baby was on CPAP would reduce “death or BPD” compared to intubation for surfactant. Kribs A, et al. Nonintubated Surfactant Application vs Conventional Therapy in Extremely Preterm Infants: A Randomized Clinical Trial. JAMA Pediatr. 2015;169(8):723-30. Infants of less than 28 weeks gestation, who were spontaneously breathing and needing more than 30% oxygen at between 10 and 120 minutes of age were enrolled, both groups received the same dose of Curosurf.

The main result of the trial is that to answer important questions like this in the early 21st century, you need big trials. This one was underpowered to find all except very large effects, with just over 100 babies per group. So although the primary outcome was improved in the LISA (less-invasive surfactant application) group, 67% surviving without BPD compared to 59% in the controls, this may have been due to chance, (p=0.2). If this was a real effect of the treatment, it would certainly be clinically worthwhile. The trial was high quality, just too small.

This brings me to another general point, the sample size was calculated on the estimate of a control rate of survival without BPD of 47% and an improvement to 69% with LISA. Apparently this was based on “data from an earlier feasibility study” but the reference they give was not to an earlier feasibility study, so I tried to find it, and the previous studies of LISA that I can find do not produce the figures they present. In any case using small pilot studies as a way of estimating a treatment effect is inherently unreliable, the smaller a study, the wider the confidence intervals around any result; sample sizes should be based on a realistic and clinically important estimate of an effect. William Tarnow-Mordi’s article in a recent Seminars in Fetal and Neonatal Medicine is a good place to read about some of these issues. For a trial of LISA if the expected control group outcome is around 50%, and increase in survival without BPD to 60% would be worthwhile, especially if there are no adverse consequences. To have an 80% chance of detecting such an effect you need around 400 per group, to have a 90% chance of doing so you need around 500 per group.

To return to the new trial, the secondary outcomes looked better in the LISA group, with less severe intraventricular hemorrhage, fewer intubations and fewer pneumothoraces. The main secondary outcome was pre-specified as survival without “major morbidity” which was BPD, severe IVH, cystic PVL, surgically treated retinopathy, or surgical NEC or perforation. That outcome was increased in the LISA group.

I would say the answer to the question in the title, is “not there yet” so far there seems to be no downside, none of the studies have shown an increase in any adverse outcome, but there is no clear evidence of benefit for BPD.

I am still concerned about performing a laryngoscopy for the procedure without premedication; laryngoscopy is very unpleasant, and is responsible for most of the adverse physiologic response seen during intubation. Apparently premedication was not routine in either group in this trial, which I think is very inappropriate. Once a baby has vascular access, there is no good reason for not premedicating prior to endotracheal intubation;  finding an analgesia/sedation routine to use during LISA (or MIST if you prefer that acronym) is, I think, a priority.

Posted in Neonatal Research | Tagged , , , | 1 Comment

When to deliver after Preterm Ruptured Membranes

I have often wondered why my obstetrical colleagues would often induce labour once a woman with ruptured membranes reached 34 weeks. I wasn’t aware of any data to support doing this, or, on the other hand, any good data to say that you shouldn’t.

It turns out that I was well-informed, there just wasn’t any good data, until now. Morris JM, et al. Immediate delivery compared with expectant management after preterm pre-labour rupture of the membranes close to term (PPROMT trial): a randomised controlled trial. Lancet. 2015;387(10017):444-52. In this study over 1800 women with singleton pregnancies and ruptured membranes without labour, who were between 34 and 37 weeks gestation, were randomized to either immediate delivery (induction or cesarean) or expectant management, in which case the woman and her obstetrician waited for spontaneous labour or another indication to deliver. Women whose membranes ruptured before 34 weeks became eligible when they hit that mark.

This remarkable study took 10 years to recruit their subjects. It was run out of Sydney, and funded by the NHMRC of Australia and enrolled mothers from 11 different countries.

What did they find? Well, neonatal sepsis occurred in 23 (2%) of the babies in the immediate delivery group, 29 (3%) of the expectant group, a difference which could easily be due to chance. There were 3 neonatal deaths in each group. On the other hand, expectantly managed pregnancies ended up with a significantly higher gestational age and birthweight, not surprisingly, and as a result less NICU admission, less respiratory distress, less assisted ventilation,  and fewer days in hospital, all of which were highly statistically significant. For the mothers there were some downsides, there was a slight increase in  antepartum or intrapartum haemorrhage from 3% to 5% and they had one day more of hospital stay with expectant management, but they had many fewer cesarean deliveries. 19% in the expectant group compared to 26%.

This is very high quality evidence that we should not be doing what ACOG currently states, which is to deliver immediately because of the risk of neonatal sepsis. If things are going well, and there is no sign of infection, pre-labour preterm rupture of membranes can be followed closely, with delivery for other obstetric indications.

Posted in Neonatal Research | Tagged , , | 2 Comments

How frequent is Acute Kidney Injury in the NICU? Not as frequent as some publications would have you believe.

One recent publication states that 30% of all their babies under 30 weeks gestation had Acute Kidney Injury (AKI from now on). Another from a couple of years ago reports 40% of all their babies under 1500g had AKI.

Both of these publications based their definition of AKI on an increase in serum creatinine concentrations, an increase of over 0.3 mg/dl (that is about 26 micromol/L in modern units). Both of them included very preterm babies in the first couple of days of life. Both of them ignored the usual increase in creatinine that occurs during the first 2 to 3 days of the very immature infant.

The increase in serum creatinine has been described for several years, (Miall LS, et al. Plasma creatinine rises dramatically in the first 48 hours of life in preterm infants. Pediatrics. 1999;104(6):e76.) the mechanisms have been investigated (they include creatinine re-absorption from the tubules) and the usual profile of creatinine clearly described. This is what normal creatinine values in very preterm babies look like, according to (Thayyil S, et al. A gestation- and postnatal age-based reference chart for assessing renal function in extremely premature infants. J Perinatol. 2008;28(3):226-9).

Creatinine 1st week

As you can see, on average plasma creatinine increases by about 40 micromol/L in the first 48 hours of life, which means, according to the definitions used in several recent articles, on average all very preterm babies have AKI!

The day to day changes were plotted like this (in the same publication).

creatinine changes

So you can’t use a definition based on creatinine alone to diagnose AKI in very immature infants. According to one recent publication, 100% of babies at 22 and 23 weeks gestation had AKI, and 80% of 24 week infants. Most of those babies have good urine output, follow the usual postnatal weight loss trajectory, do not become hyperkalemic, in short they have no sign of renal dysfunction.

By day 2 to 3 tubular creatinine reabsorption is ending, and the creatinine starts to fall, after that you might be able to diagnose AKI from changes in the creatinine concentration alone, but it would be great to have some definition of AKI in the preterm that actually reflected renal injury and which was correlated with later renal problems, without that, it is hard to know what to do about a creatinine increase. Certainly creatinine increases in the first 2 days of life are almost universal under 28 weeks and do not mean Kidney Injury. Some years ago it was suggested that an increase in creatinine over 0.5 mg/dl, or 44 micromol/L could be used in the first 3 days, these graphs show that would be a more appropriate threshold, at least from a statistical point of view, if such a threshold is a good discriminator between those babies with and without true AKI needs to be investigated.


Posted in Neonatal Research | 7 Comments

All results must be published, and all data available.

Support for the All trials campaign has been growing, but progress seems to be slow. There are still many trials that are not published, and the impacts on science based medicine are substantial.

Recently, two articles by former participants in trials have been published, a viewpoint that is very important for these campaigns, I believe.

The first is called “The Outcome of My Clinical Trial Is a Mystery” it is by someone who was part of a trial to evaluate an intracardiac device to close an ASD, the author says :

It’s more than a little unsettling to think that I’ll live for another 70 years or so with this device in my heart, but may never know what the research said about how it works.

The second is called “A Broken Covenant with Patients” from a woman who was part of a preliminary evaluation of a drug for ovarian cancer, and who gives other examples of “broken covenants”.

As for sharing of data, the International Committee of Medical Journal Editors are proposing new rules, the meat of their proposal is here:

As a condition of consideration for publication of a clinical trial report in our member journals, the ICMJE proposes to require authors to share with others the deidentified individual-patient data (IPD) underlying the results presented in the article (including tables, figures, and appendices or supplementary material) no later than 6 months after publication. The data underlying the results are defined as the IPD required to reproduce the article’s findings, including necessary metadata.

I agree with this approach, and I plan to comply, there are details to be worked out, but one that I think is important is how will this be policed and enforced? Many rules of ICMJE are not well followed, including, for example, the requirement for registration of trials, and that publications should follow CONSORT guidelines. There is no point adding new rules if they are just going to be ignored, as happens already with data sharing, in those situation where it is already required.

As one example of this failure you could look at the COMPare website. Ben Goldacre has, with a few colleagues, started to review the randomized controlled trials published in 5 leading medical journals. His team compare the outcomes published in the articles describing the research results to those available in the trial registration documents, or in a published protocol, if published before the trial started. Of course the outcomes described should be those that were initially planned, and if for some reason that was not the case, then any difference should be explained in the trial report. Changing outcomes for a trial means that it cannot necessarily be trusted to answer the question that was initially asked. That is particularly true when the primary outcome is changed when the investigators look at the results. Clearly, if the initially planned primary outcome is not significantly different between groups, but the authors find something else which looks cool, and then claim that was the primary outcome, the risks of making type 1 errors are enormously increased.

Journals are not doing a very good job of comparing reported outcomes to the pre-specified outcomes, nor of insisting that CONSORT standards are followed, and any changes clearly described and justified. The COMPare group as of today have checked 67 trials, they have found only 9 of them reported all the outcomes that were planned. There were 301 outcomes that were planned to be reported by the trials that were not reported, and there were 357 outcomes that appeared for the first time in the published articles.

The editors of the Annals of Internal Medicine have answered in a very haughty and inconsistent way to the letters from COMPare that pointed out the problems in several publications. At one point pointing out their “long experience”, as if that somehow makes them less fallible.

I can only hope that the COMPare team are successful in making journals check publications more carefully for consistency with the CONSORT standards. There are already some signs that they are having an impact.


Posted in Neonatal Research | Leave a comment

Medium Term outcomes after neonatal hypoglycemia

A new publication from the CHYLD program that I blogged about previously:

Harris DL, et al. Outcome at 2 Years after Dextrose Gel Treatment for Neonatal Hypoglycemia: Follow-Up of a Randomized Trial. The Journal of pediatrics. 2015.

In this paper the authors describe the outcomes of infants who were in their Sugar Babies study, when they reached 2 years of age. The infants had Bayley scales of infant development version 3 performed, as well as tests of executive function and neurological exams.

To recap, the Sugar Babies study took infants at risk of hypoglycemia, and when the blood sugar was below 2.6 mmol/L (47 mg/dl) they were randomized to either get dextrose gel, or placebo. They could get treated more than once if necessary, and if the blood glucose fell to more severe levels the received standard therapy. 237 babies were in the original study, 78% of them returned for follow-up.

The babies of course, as mentioned, all had risk factors for hypoglycemia, so there were 40% whose mothers were diabetic, 1/3 were late preterms, 16% SGA, 8% LGA. The outcomes were not different between the two groups, but they were not very reassuring. 34% of the infants had a Bayley cognitive or language composite less than 85, (1 SD below the standardized mean value of 100, which we know is not the mean of the general New Zealand population) The mean language and cognitive composite scores of the subjects were between 94 and 96, Full term controls in a recent Australian study had means of 108 to 109.

So these babies, in both groups look like they did substantially worse on the Bayley test than you might expect. There were also a few babies with other problems, including 9 with seizures, and their executive function scores were abnormal also. Is this because the risk factors that they had select a group of babies who are higher risk? Is it because mild hypoglycemia, even if you treat with a rapidly effective intervention, such as the dextrose gel, has a direct adverse effect on the newborn brain? We know that infants who are small for gestational age, and late preterm infants are at risk for developmental delay, but there seems to be remarkably little data about the risks of infants of diabetic mothers.

It certainly looks like being a newborn infant in a group at risk for hypoglycemia, who then has a blood sugar under 2.6 is a risk factor for developmental delay and for executive function disturbances.

Posted in Neonatal Research | 1 Comment