Predicting adverse long-term neurocognitive outcomes after pediatric intensive care unit admission.

BACKGROUND AND OBJECTIVE: Critically ill children may suffer from impaired neurocognitive functions years after ICU (intensive care unit) discharge. To assess neurocognitive functions, these children are subjected to a fixed sequence of tests. Undergoing all tests is, however, arduous for former pediatric ICU patients, resulting in interrupted evaluations where several neurocognitive deficiencies remain undetected. As a solution, we propose using machine learning to predict the optimal order of tests for each child, reducing the number of tests required to identify the most severe neurocognitive deficiencies. METHODS: We have compared the current clinical approach against several machine learning methods, mainly multi-target regression and label ranking methods. We have also proposed a new method that builds several multi-target predictive models and combines the outputs into a ranking that prioritizes the worse neurocognitive outcomes. We used data available at discharge, from children who participated in the PEPaNIC-RCT trial (ClinicalTrials.gov-NCT01536275), as well as data from a 2-year follow-up study. The institutional review boards at each participating site have also approved this follow-up study (ML8052; NL49708.078; Pro00038098). RESULTS: Our proposed method managed to outperform other machine learning methods and also the current clinical practice. Precisely, our method reaches approximately 80% precision when considering top-4 outcomes, in comparison to 65% and 78% obtained by the current clinical practice and the state-of-the-art method in label ranking, respectively. CONCLUSIONS: Our experiments demonstrated that machine learning can be competitive or even superior to the current testing order employed in clinical practice, suggesting that our model can be used to severely reduce the number of tests necessary for each child. Moreover, the results indicate that possible long-term adverse outcomes are already predictable as early as at ICU discharge. Thus, our work can be seen as the first step to allow more personalized follow-up after ICU discharge leading to preventive care rather than curative.

Abnormal DNA methylation within HPA-axis genes years after paediatric critical illness.

BACKGROUND: Critically ill children suffer from impaired physical/neurocognitive development 2 years later. Glucocorticoid treatment alters DNA methylation within the hypothalamus-pituitary-adrenal (HPA) axis which may impair normal brain development, cognition and behaviour. We tested the hypothesis that paediatric-intensive-care-unit (PICU) patients, sex- and age-dependently, show long-term abnormal DNA methylation within the HPA-axis layers, possibly aggravated by glucocorticoid treatment in the PICU, which may contribute to the long-term developmental impairments. RESULTS: In a pre-planned secondary analysis of the multicentre PEPaNIC-RCT and its 2-year follow-up, we identified differentially methylated positions and differentially methylated regions within HPA-axis genes in buccal mucosa DNA from 818 former PICU patients 2 years after PICU admission (n = 608 no glucocorticoid treatment; n = 210 glucocorticoid treatment) versus 392 healthy children and assessed interaction with sex and age, role of glucocorticoid treatment in the PICU and associations with long-term developmental impairments. Adjusting for technical variation and baseline risk factors and correcting for multiple testing (false discovery rate < 0.05), former PICU patients showed abnormal DNA methylation of 26 CpG sites (within CRHR1, POMC, MC2R, NR3C1, FKBP5, HSD11B1, SRD5A1, AKR1D1, DUSP1, TSC22D3 and TNF) and three DNA regions (within AVP, TSC22D3 and TNF) that were mostly hypomethylated. These abnormalities were sex-independent and only partially age-dependent. Abnormal methylation of three CpG sites within FKBP5 and one CpG site within SRD5A1 and AKR1D1 was partly attributable to glucocorticoid treatment during PICU stay. Finally, abnormal methylation within FKBP5 and AKR1D1 was most robustly associated with long-term impaired development. CONCLUSIONS: Two years after critical illness in children, abnormal methylation within HPA-axis genes was present, predominantly within FKBP5 and AKR1D1, partly attributable to glucocorticoid treatment in the PICU, and explaining part of the long-term developmental impairments. These data call for caution regarding liberal glucocorticoid use in the PICU.

Tight Blood-Glucose Control without Early Parenteral Nutrition in the ICU.

BACKGROUND: Randomized, controlled trials have shown both benefit and harm from tight blood-glucose control in patients in the intensive care unit (ICU). Variation in the use of early parenteral nutrition and in insulin-induced severe hypoglycemia might explain this inconsistency. METHODS: We randomly assigned patients, on ICU admission, to liberal glucose control (insulin initiated only when the blood-glucose level was >215 mg per deciliter [>11.9 mmol per liter]) or to tight glucose control (blood-glucose level targeted with the use of the LOGIC-Insulin algorithm at 80 to 110 mg per deciliter [4.4 to 6.1 mmol per liter]); parenteral nutrition was withheld in both groups for 1 week. Protocol adherence was determined according to glucose metrics. The primary outcome was the length of time that ICU care was needed, calculated on the basis of time to discharge alive from the ICU, with death accounted for as a competing risk; 90-day mortality was the safety outcome. RESULTS: Of 9230 patients who underwent randomization, 4622 were assigned to liberal glucose control and 4608 to tight glucose control. The median morning blood-glucose level was 140 mg per deciliter (interquartile range, 122 to 161) with liberal glucose control and 107 mg per deciliter (interquartile range, 98 to 117) with tight glucose control. Severe hypoglycemia occurred in 31 patients (0.7%) in the liberal-control group and 47 patients (1.0%) in the tight-control group. The length of time that ICU care was needed was similar in the two groups (hazard ratio for earlier discharge alive with tight glucose control, 1.00; 95% confidence interval, 0.96 to 1.04; P = 0.94). Mortality at 90 days was also similar (10.1% with liberal glucose control and 10.5% with tight glucose control, P = 0.51). Analyses of eight prespecified secondary outcomes suggested that the incidence of new infections, the duration of respiratory and hemodynamic support, the time to discharge alive from the hospital, and mortality in the ICU and hospital were similar in the two groups, whereas severe acute kidney injury and cholestatic liver dysfunction appeared less prevalent with tight glucose control. CONCLUSIONS: In critically ill patients who were not receiving early parenteral nutrition, tight glucose control did not affect the length of time that ICU care was needed or mortality. (Funded by the Research Foundation-Flanders and others; TGC-Fast ClinicalTrials.gov number, NCT03665207.).

Abnormal DNA methylation within genes of the steroidogenesis pathway two years after paediatric critical illness and association with stunted growth in height further in time.

BACKGROUND: Former critically ill children show an epigenetic age deceleration 2 years after paediatric intensive care unit (PICU) admission as compared with normally developing healthy children, with stunted growth in height 2 years further in time as physical correlate. This was particularly pronounced in children who were 6 years or older at the time of critical illness. As this age roughly corresponds to the onset of adrenarche and further pubertal development, a relation with altered activation of endocrine pathways is plausible. We hypothesised that children who have been admitted to the PICU, sex- and age-dependently show long-term abnormal DNA methylation within genes involved in steroid hormone synthesis or steroid sulphation/desulphation, possibly aggravated by in-PICU glucocorticoid treatment, which may contribute to stunted growth in height further in time after critical illness. RESULTS: In this preplanned secondary analysis of the multicentre PEPaNIC-RCT and its follow-up, we compared the methylation status of genes involved in the biosynthesis of steroid hormones (aldosterone, cortisol and sex hormones) and steroid sulphation/desulphation in buccal mucosa DNA (Infinium HumanMethylation EPIC BeadChip) from former PICU patients at 2-year follow-up (n = 818) and healthy children with comparable sex and age (n = 392). Adjusting for technical variation and baseline risk factors and corrected for multiple testing (false discovery rate < 0.05), former PICU patients showed abnormal DNA methylation of 23 CpG sites (within CYP11A1, POR, CYB5A, HSD17B1, HSD17B2, HSD17B3, HSD17B6, HSD17B10, HSD17B12, CYP19A1, CYP21A2, and CYP11B2) and 4 DNA regions (within HSD17B2, HSD17B8, and HSD17B10) that were mostly hypomethylated. These abnormalities were partially sex- (1 CpG site) or age-dependent (7 CpG sites) and affected by glucocorticoid treatment (3 CpG sites). Finally, multivariable linear models identified robust associations of abnormal methylation of steroidogenic genes with shorter height further in time, at 4-year follow-up. CONCLUSIONS: Children who have been critically ill show abnormal methylation within steroidogenic genes 2 years after PICU admission, which explained part of the stunted growth in height at 4-year follow-up. The abnormalities in DNA methylation may point to a long-term disturbance in the balance between active sex steroids and mineralocorticoids/glucocorticoids after paediatric critical illness, which requires further investigation.

Development and validation of the creatinine clearance predictor machine learning models in critically ill adults.

BACKGROUND: In critically ill patients, measured creatinine clearance (CrCl) is the most reliable method to evaluate glomerular filtration rate in routine clinical practice and may vary subsequently on a day-to-day basis. We developed and externally validated models to predict CrCl one day ahead and compared them with a reference reflecting current clinical practice. METHODS: A gradient boosting method (GBM) machine-learning algorithm was used to develop the models on data from 2825 patients from the EPaNIC multicenter randomized controlled trial database. We externally validated the models on 9576 patients from the University Hospitals Leuven, included in the M@tric database. Three models were developed: a "Core" model based on demographic, admission diagnosis, and daily laboratory results; a "Core + BGA" model adding blood gas analysis results; and a "Core + BGA + Monitoring" model also including high-resolution monitoring data. Model performance was evaluated against the actual CrCl by mean absolute error (MAE) and root-mean-square error (RMSE). RESULTS: All three developed models showed smaller prediction errors than the reference. Assuming the same CrCl of the day of prediction showed 20.6 (95% CI 20.3-20.9) ml/min MAE and 40.1 (95% CI 37.9-42.3) ml/min RMSE in the external validation cohort, while the developed model having the smallest RMSE (the Core + BGA + Monitoring model) had 18.1 (95% CI 17.9-18.3) ml/min MAE and 28.9 (95% CI 28-29.7) ml/min RMSE. CONCLUSIONS: Prediction models based on routinely collected clinical data in the ICU were able to accurately predict next-day CrCl. These models could be useful for hydrophilic drug dosage adjustment or stratification of patients at risk. TRIAL REGISTRATION: Not applicable.

The association of hypoglycemia with outcome of critically ill children in relation to nutritional and blood glucose control strategies.

BACKGROUND: Withholding parenteral nutrition (PN) until one week after PICU admission facilitated recovery from critical illness and protected against emotional and behavioral problems 4 years later. However, the intervention increased the risk of hypoglycemia, which may have counteracted part of the benefit. Previously, hypoglycemia occurring under tight glucose control in critically ill children receiving early PN did not associate with long-term harm. We investigated whether hypoglycemia in PICU differentially associates with outcome in the context of withholding early PN, and whether any potential association with outcome may depend on the applied glucose control protocol. METHODS: In this secondary analysis of the multicenter PEPaNIC RCT, we studied whether hypoglycemia in PICU associated with mortality (N = 1440) and 4-years neurodevelopmental outcome (N = 674) through univariable comparison and multivariable regression analyses adjusting for potential confounders. In patients with available blood samples (N = 556), multivariable models were additionally adjusted for baseline serum NSE and S100B concentrations as biomarkers of neuronal, respectively, astrocytic damage. To study whether an association of hypoglycemia with outcome may be affected by the nutritional strategy or center-specific glucose control protocol, we further adjusted the models for the interaction between hypoglycemia and the randomized nutritional strategy, respectively, treatment center. In sensitivity analyses, we studied whether any association with outcome was different in patients with iatrogenic or spontaneous/recurrent hypoglycemia. RESULTS: Hypoglycemia univariably associated with higher mortality in PICU, at 90 days and 4 years after randomization, but not when adjusted for risk factors. After 4 years, critically ill children with hypoglycemia scored significantly worse for certain parent/caregiver-reported executive functions (working memory, planning and organization, metacognition) than patients without hypoglycemia, also when adjusted for risk factors including baseline NSE and S100B. Further adjustment for the interaction of hypoglycemia with the randomized intervention or treatment center revealed a potential interaction, whereby tight glucose control and withholding early PN may be protective. Impaired executive functions were most pronounced in patients with spontaneous or recurrent hypoglycemia. CONCLUSION: Critically ill children exposed to hypoglycemia in PICU were at higher risk of impaired executive functions after 4 years, especially in cases of spontaneous/recurrent hypoglycemia.

Long-term impact of paediatric critical illness on the difference between epigenetic and chronological age in relation to physical growth.

BACKGROUND: Altered DNA-methylation affects biological ageing in adults and developmental processes in children. DNA-methylation is altered by environmental factors, trauma and illnesses. We hypothesised that paediatric critical illness, and the nutritional management in the paediatric intensive care unit (PICU), affects DNA-methylation changes that underly the developmental processes of childhood ageing. RESULTS: We studied the impact of critical illness, and of the early use of parenteral nutrition (early-PN) versus late-PN, on "epigenetic age-deviation" in buccal mucosa of 818 former PICU-patients (406 early-PN, 412 late-PN) who participated in the 2-year follow-up of the multicentre PEPaNIC-RCT (ClinicalTrials.gov-NCT01536275), as compared with 392 matched healthy children, and assessed whether this relates to their impaired growth. The epigenetic age-deviation (difference between PedBE clock-estimated epigenetic age and chronological age) was calculated. Using bootstrapped multivariable linear regression models, we assessed the impact hereon of critical illness, and of early-PN versus late-PN. As compared with healthy children, epigenetic age of patients assessed 2 years after PICU-admission deviated negatively from chronological age (p < 0.05 in 51% of bootstrapped replicates), similarly in early-PN and late-PN groups. Next, we identified vulnerable subgroups for epigenetic age-deviation using interaction analysis. We revealed that DNA-methylation age-deceleration in former PICU-patients was dependent on age at time of illness (p < 0.05 for 83% of bootstrapped replicates), with vulnerability starting from 6 years onwards. Finally, we assessed whether vulnerability to epigenetic age-deviation could be related to impaired growth from PICU-admission to follow-up at 2 and 4 years. Multivariable repeated measures ANOVA showed that former PICU-patients, as compared with healthy children, grew less in height (p = 0.0002) and transiently gained weight (p = 0.0003) over the 4-year time course. Growth in height was more stunted in former PICU-patients aged ≥ 6-years at time of critical illness (p = 0.002) than in the younger patients. CONCLUSIONS: As compared with healthy children, former PICU-patients, in particular those aged ≥ 6-years at time of illness, revealed epigenetic age-deceleration, with a physical correlate revealing stunted growth in height. Whether this vulnerability around the age of 6 years for epigenetic age-deceleration and stunted growth years later relates to altered endocrine pathways activated at the time of adrenarche requires further investigation.

Development and validation of clinical prediction models for acute kidney injury recovery at hospital discharge in critically ill adults.

PURPOSE: Acute kidney injury (AKI) recovery prediction remains challenging. The purpose of the present study is to develop and validate prediction models for AKI recovery at hospital discharge in critically ill patients with ICU-acquired AKI stage 3 (AKI-3). METHODS: Models were developed and validated in a development cohort (n = 229) and a matched validation cohort (n = 244) from the multicenter EPaNIC database to create prediction models with the least absolute shrinkage and selection operator (Lasso) machine-learning algorithm. We evaluated the discrimination and calibration of the models and compared their performance with plasma neutrophil gelatinase-associated lipocalin (NGAL) measured on first AKI-3 day (NGAL_AKI3) and reference model that only based on age. RESULTS: Complete recovery and complete or partial recovery occurred in 33.20% and 51.23% of the validation cohort patients respectively. The prediction model for complete recovery based on age, need for renal replacement therapy (RRT), diagnostic group (cardiac/surgical/trauma/others), and sepsis on admission had an area under the receiver operating characteristics curve (AUROC) of 0.53. The prediction model for complete or partial recovery based on age, need for RRT, platelet count, urea, and white blood cell count had an AUROC of 0.61. NGAL_AKI3 showed AUROCs of 0.55 and 0.53 respectively. In cardiac patients, the models had higher AUROCs of 0.60 and 0.71 than NGAL_AKI3's AUROCs of 0.52 and 0.54. The developed models demonstrated a better performance over the reference models (only based on age) for cardiac surgery patients, but not for patients with sepsis and for a general ICU population. CONCLUSION: Models to predict AKI recovery upon hospital discharge in critically ill patients with AKI-3 showed poor performance in the general ICU population, similar to the biomarker NGAL. In cardiac surgery patients, discrimination was acceptable, and better than NGAL. These findings demonstrate the difficulty of predicting non-reversible AKI early.

Assessment of aberrant DNA methylation two years after paediatric critical illness: a pre-planned secondary analysis of the international PEPaNIC trial.

Critically ill children requiring intensive care suffer from impaired physical/neurocognitive development 2 y later, partially preventable by omitting early use of parenteral nutrition (early-PN) in the paediatric intensive-care-unit (PICU). Altered methylation of DNA from peripheral blood during PICU-stay provided a molecular basis hereof. Whether DNA-methylation of former PICU patients, assessed 2 y after critical illness, is different from that of healthy children remained unknown. In a pre-planned secondary analysis of the PEPaNIC-RCT (clinicaltrials.gov-NCT01536275) 2-year follow-up, we assessed buccal-mucosal DNA-methylation (Infinium-HumanMethylation-EPIC-BeadChip) of former PICU-patients (N = 406 early-PN; N = 414 late-PN) and matched healthy children (N = 392). CpG-sites differentially methylated between groups were identified with multivariable linear regression and differentially methylated DNA-regions via clustering of differentially methylated CpG-sites using kernel-estimates. Analyses were adjusted for technical variation and baseline risk factors, and corrected for multiple testing (false-discovery-rate <0.05). Differentially methylated genes were functionally annotated (KEGG-pathway database), and allocated to three classes depending on involvement in physical/neurocognitive development, critical illness and intensive medical care, or pre-PICU-admission disorders. As compared with matched healthy children, former PICU-patients showed significantly different DNA-methylation at 4047 CpG-sites (2186 genes) and 494 DNA-regions (468 genes), with most CpG-sites being hypomethylated (90.3%) and with an average absolute 2% effect-size, irrespective of timing of PN initiation. Of the differentially methylated KEGG-pathways, 41.2% were related to physical/neurocognitive development, 32.8% to critical illness and intensive medical care and 26.0% to pre-PICU-admission disorders. Two years after critical illness in children, buccal-mucosal DNA showed abnormal methylation of CpG-sites and DNA-regions located in pathways known to be important for physical/neurocognitive development.

Impact of tight blood glucose control within normal fasting ranges with insulin titration prescribed by the Leuven algorithm in adult critically ill patients: the TGC-fast randomized controlled trial.

BACKGROUND: It remains controversial whether critical illness-related hyperglycemia should be treated or not, since randomized controlled trials (RCTs) have shown context-dependent outcome effects. Whereas pioneer RCTs found improved outcome by normalizing blood glucose in patients receiving early parenteral nutrition (PN), a multicenter RCT revealed increased mortality in patients not receiving early PN. Although withholding early PN has become the feeding standard, the multicenter RCT showing harm by tight glucose control in this context has been criticized for its potentially unreliable glucose control protocol. We hypothesize that tight glucose control is effective and safe using a validated protocol in adult critically ill patients not receiving early PN. METHODS: The TGC-fast study is an investigator-initiated, multicenter RCT. Patients unable to eat, with need for arterial and central venous line and without therapy restriction, are randomized upon ICU admission to tight (80-110 mg/dl) or liberal glucose control (only initiating insulin when hyperglycemia >215 mg/dl, and then targeting 180-215 mg/dl). Glucose measurements are performed on arterial blood by a blood gas analyzer, and if needed, insulin is only administered continuously through a central venous line. If the arterial line is no longer needed, glucose is measured on capillary blood. In the intervention group, tight control is guided by the validated LOGIC-Insulin software. In the control arm, a software alert is used to maximize protocol compliance. The intervention is continued until ICU discharge, until the patient is able to eat or no longer in need of a central venous line, whatever comes first. The study is powered to detect, with at least 80% power and a 5% alpha error rate, a 1-day difference in ICU dependency (primary endpoint), and a 1.5% increase in hospital mortality (safety endpoint), for which 9230 patients need to be included. Secondary endpoints include acute and long-term morbidity and mortality, and healthcare costs. Biological samples are collected to study potential mechanisms of organ protection. DISCUSSION: The ideal glucose target for critically ill patients remains debated. The trial will inform physicians on the optimal glucose control strategy in adult critically ill patients not receiving early PN. TRIAL REGISTRATION: ClinicalTrials.gov NCT03665207. Registered on 11 September 2018.

Impact of duration of critical illness and level of systemic glucocorticoid availability on tissue-specific glucocorticoid receptor expression and actions: A prospective, observational, cross-sectional human and two translational mouse studies.

BACKGROUND: Reduced glucocorticoid-receptor (GR) expression in blood suggested that critically ill patients become glucocorticoid-resistant necessitating stress-doses of glucocorticoids. We hypothesised that critical illness evokes a tissue-specific, time-dependent expression of regulators of GR-action which adaptively guides glucocorticoid action to sites of need. METHODS: We performed a prospective, observational, cross-sectional human study and two translational mouse studies. In freshly-isolated neutrophils and monocytes and in skeletal muscle and subcutaneous adipose tissue of 137 critically ill patients and 20 healthy controls and in skeletal muscle and adipose tissue as well as in vital tissues (heart, lung, diaphragm, liver, kidney) of 88 septic and 26 healthy mice, we quantified gene expression of cortisone-reductase 11ß-HSD1, glucocorticoid-receptor-isoforms GRα and GRß, GRα-sensitivity-regulating-co-chaperone FKBP51, and GR-action-marker GILZ. Expression profiles were compared in relation to illness-duration and systemic-glucocorticoid-availability. FINDINGS: In patients' neutrophils, GRα and GILZ were substantially suppressed (p≤0·05) throughout intensive care unit (ICU)-stay, while in monocytes low/normal GRα coincided with increased GILZ (p≤0·05). FKBP51 was increased transiently (neutrophils) or always (monocytes,p≤0·05). In patients' muscle, 11ß-HSD1 and GRα were low-normal (p≤0·05) and substantially suppressed in adipose tissue (p≤0·05); FKBP51 and GILZ were increased in skeletal muscle (p≤0·05) but normal in adipose tissue. GRß was undetectable. Increasing systemic glucocorticoid availability in patients independently associated with further suppressed muscle 11ß-HSD1 and GRα, further increased FKBP51 and unaltered GILZ (p≤0·05). In septic mouse heart and lung, 11ß-HSD1, FKBP51 and GILZ were always high (p≤0·01). In heart, GRα was suppressed (p≤0·05), while normal or high in lung (all p≤0·05). In diaphragm, 11ß-HSD1 was high/normal, GRα low/normal and FKBP51 and GILZ high (p≤0·01). In kidney, 11ß-HSD1 transiently increased but decreased thereafter, GRα was normal and FKBP51 and GILZ high (p≤0·01). In liver, 11ß-HSD1 was suppressed (p≤0·01), GRα normal and FKBP51 high (p≤0·01) whereas GILZ was transiently decreased but elevated thereafter (p≤0·05). Only in lung and diaphragm, treatment with hydrocortisone further increased GILZ. INTERPRETATION: Tissue-specific, time-independent adaptations to critical illness guided GR-action predominantly to vital tissues such as lung, while (partially) protecting against collateral harm in other cells and tissues, such as neutrophils. These findings argue against maladaptive generalised glucocorticoid-resistance necessitating glucocorticoid-treatment. FUNDING: Research-Foundation-Flanders, Methusalem-Program-Flemish-Government, European-Research-Council, European-Respiratory-Society.

Impact of critical illness and withholding of early parenteral nutrition in the pediatric intensive care unit on long-term physical performance of children: a 4-year follow-up of the PEPaNIC randomized controlled trial.

BACKGROUND: Many critically ill children face long-term developmental impairments. The PEPaNIC trial attributed part of the problems at the level of neurocognitive and emotional/behavioral development to early use of parenteral nutrition (early-PN) in the PICU, as compared with withholding it for 1 week (late-PN). Insight in long-term daily life physical functional capacity after critical illness is limited. Also, whether timing of initiating PN affects long-term physical function of these children remained unknown. METHODS: This preplanned follow-up study of the multicenter PEPaNIC randomized controlled trial subjected 521 former critically ill children (253 early-PN, 268 late-PN) to quantitative physical function tests 4 years after PICU admission in Leuven or Rotterdam, in comparison with 346 age- and sex-matched healthy children. Tests included handgrip strength measurement, timed up-and-go test, 6-min walk test, and evaluation of everyday overall physical activity with an accelerometer. We compared these functional measures for the former critically ill and healthy children and for former critically ill children randomized to late-PN versus early-PN, with multivariable linear or logistic regression analyses adjusting for risk factors. RESULTS: As compared with healthy children, former critically ill children showed less handgrip strength (p < 0.0001), completed the timed up-and-go test more slowly (p < 0.0001), walked a shorter distance in 6 min (p < 0.0001) during which they experienced a larger drop in peripheral oxygen saturation (p ≤ 0.026), showed a lower energy expenditure (p ≤ 0.024), performed more light and less moderate physical activity (p ≤ 0.047), and walked fewer steps per day (p = 0.0074). Late-PN as compared with early-PN did not significantly affect these outcomes. CONCLUSIONS: Four years after PICU admission, former critically ill children showed worse physical performance as compared with healthy children, without impact of timing of supplemental PN in the PICU. This study provides further support for de-implementing the early use of PN in the PICU. Trial registration ClinicalTrials.gov, NCT01536275 ; registered on February 22, 2012.

Physical, Emotional/Behavioral, and Neurocognitive Developmental Outcomes From 2 to 4 Years After PICU Admission: A Secondary Analysis of the Early Versus Late Parenteral Nutrition Randomized Controlled Trial Cohort.

OBJECTIVES: PICU patients face long-term developmental impairments, partially attributable to early parenteral nutrition (PN) versus late-PN. We investigated how this legacy and harm by early-PN evolve over time. DESIGN: Preplanned secondary analysis of the multicenter PEPaNIC-RCT (ClinicalTrials.gov, NCT01536275) that enrolled 1,440 critically ill children from 2012 to 2015 and its 2- (2014-2018) and 4-year (2016-2019) cross-sectional follow-up studies. SETTING: PICUs of Leuven (Belgium), Rotterdam (The Netherlands), and Edmonton (Canada). PATIENTS: Patients and demographically matched healthy control children that underwent longitudinal assessment for physical/emotional/behavioral/neurocognitive functions at both follow-up time points. INTERVENTIONS: In the PEPaNIC-RCT, patients were randomly allocated to early-PN versus late-PN. MEASUREMENTS AND MAIN RESULTS: This within-individual longitudinal study investigated changes in physical/emotional/behavioral/neurocognitive functions from 2 to 4 years after PICU admission for 614 patients (297 early-PN and 317 late-PN, tested at mean ± sd age 5.4 ± 4.2 and 7.3 ± 4.3 yr) and for 357 demographically matched healthy children tested at age 5.6 ± 4.3 and 7.5 ± 4.3 years. We determined within-group time-courses, interaction between time and group, and independent impact of critical illness and early-PN on these time-courses. Most deficits in patients versus healthy children remained prominent over the 2 years ( p ≤ 0.01). Deficits further aggravated for height, body mass index, the executive function metacognition, intelligence, motor coordination (alternating/synchronous tapping), and memory learning-index, whereas verbal memory deficits became smaller (working/immediate/delayed memory) ( p ≤ 0.05). Adjustment for risk factors confirmed most findings and revealed that patients "grew-into-deficit" for additional executive functions (flexibility/emotional control/total executive functioning) and "grew-out-of-deficit" for additional memory functions (recognition/pictures) ( p ≤ 0.05). Time-courses were largely unaffected by early-PN versus late-PN, except for weight loss and limited catch-up for visual-motor integration and alertness in early-PN patients ( p ≤ 0.05). CONCLUSIONS: From 2- to 4-year post-PICU admission, developmental impairments remained prominent. Within that time-window, impaired growth in height, executive functioning and intelligence aggravated, and impaired memory and harm by early-PN only partially recovered. Impact on development into adulthood requires further investigation.

DNA methylation alterations in muscle of critically ill patients.

BACKGROUND: Intensive care unit (ICU)-acquired weakness can persist beyond ICU stay and has been associated with long-term functional impairment of ICU survivors. Recently, DNA methylation alterations were found in the blood of ICU patients, partially explaining long-term developmental impairment of critically ill children. As illness-induced aberrant DNA methylation theoretically could also be involved in long-term weakness, we investigated whether the DNA methylation signature in muscle of adult critically ill patients differs from that in muscle of healthy controls. METHODS: Genome-wide methylation was determined (Infinium® HumanMethylationEPIC BeadChips) in DNA extracted from skeletal muscle biopsies that had been collected on Day 8 ± 1 in ICU from 172 EPaNIC-trial patients [66% male sex, median age 62.7 years, median body mass index (BMI) 25.9 kg/m2 ] and 20 matched healthy controls (70% male sex, median age 58.0 years, median BMI 24.4 kg/m2 ). Methylation status of individual cytosine-phosphate-guanine (CpG) sites of patients and controls was compared with F-tests, using the Benjamini-Hochberg false discovery rate to correct for multiple comparisons. Differential methylation of DNA regions was assessed with bump hunting, with 1000 permutations assessing uncertainty, expressed as family-wise error rate. Gene expression was investigated for 10 representative affected genes. RESULTS: In DNA from ICU patients, 565 CpG sites, associated with 400 unique genes, were differentially methylated as compared with controls (average difference 3.2 ± 0.1% ranging up to 16.9%, P < 0.00005). Many of the associated genes appeared highly relevant for muscle structure and function/weakness, including genes involved in myogenesis, muscle regeneration, nerve/muscle membrane excitability, muscle denervation/re-innervation, axon guidance/myelination/degeneration/regeneration, synapse function, ion channelling with especially calcium signalling, metabolism (glucose, protein, and fat), insulin signalling, neuroendocrine hormone regulation, mitochondrial function, autophagy, apoptosis, oxidative stress, Wnt signalling, transcription regulation, muscle fat infiltration during regeneration, and fibrosis. In patients as compared with controls, we also identified two hypomethylated regions, spanning 18 and 3 CpG sites in the promoters of the HIC1 and NADK2 genes, respectively (average differences 5.8 ± 0.01% and 12.1 ± 0.04%, family-wise error rate <0.05). HIC1 and NADK2 play important roles in muscle regeneration and postsynaptic acetylcholine receptors and in mitochondrial processes, respectively. Nine of 10 investigated genes containing DNA methylation alterations were differentially expressed in patients as compared with controls (P ≤ 0.03). CONCLUSIONS: Critically ill patients present with a different DNA methylation signature in skeletal muscle as compared with healthy controls, which in theory could provide a biological basis for long-term persistence of weakness in ICU survivors. TRIAL REGISTRATION: ClinicalTrials.gov: NCT00512122, registered on 31 July 2007.

Phasing out DEHP from plastic indwelling medical devices used for intensive care: Does it reduce the long-term attention deficit of critically ill children?

BACKGROUND: Children who have been critically ill face long-term developmental impairments. Iatrogenic exposure to di(2-ethylhexyl)phthalate (DEHP), a plasticizer leaching from plastic indwelling medical devices used in the pediatric intensive care unit (PICU), has been associated with the pronounced attention deficit observed in children 4 years after critical illness. As concerns about DEHP toxicity increased, governmental authorities urged the phase out of DEHP in indwelling medical devices and replacement with alternative plasticizers. We hypothesized that exposure to DEHP decreased over the years, attenuating the pronounced long-term attention deficit of these vulnerable children. METHODS: We compared plasma concentrations of 3 oxidative DEHP metabolites (5cx-MEPP, 5OH-MEHP, 5oxo-MEHP) on the last PICU day in 216 patients who participated in the Tight Glucose Control study (2004-2007) and 334 patients who participated in the PEPaNIC study (2012-2015) and survived PICU stay. Corresponding minimal exposures to these metabolites (plasma concentration multiplied with number of days in PICU) were also evaluated. In patients with 4-year follow-up data, we compared measures of attention (standardized reaction times and consistency). Comparisons were performed with univariable analyses and multivariable linear regression analyses adjusted for baseline risk factors. RESULTS: In the PEPaNIC patients, last PICU day plasma concentrations of 5cx-MEPP, 5OH-MEHP, 5oxo-MEHP and their sum, and corresponding minimal exposures, were reduced to 17-69% of those in the Tight Glucose Control study (p < 0.0001). Differences remained significant after multivariable adjustment (p ≤ 0.001). PEPaNIC patients did not show better attention than patients in the Tight Glucose Control study, also not after multivariable adjustment for risk factors. CONCLUSION: Exposure of critically ill children to DEHP in the PICU decreased over the years, but the lower exposure did not translate into improved attention 4 years later. Whether the residual exposure may still be toxic or whether the plasticizers replacing DEHP may not be safe for neurodevelopment needs further investigation.

C-reactive protein rise in response to macronutrient deficit early in critical illness: sign of inflammation or mediator of infection prevention and recovery.

PURPOSE: Withholding parenteral nutrition (PN) early in critical illness, late-PN, has shown to prevent infections despite a higher peak C-reactive protein (CRP). We investigated whether the accentuated CRP rise was caused by a systemic inflammatory effect mediated by cytokines or arose as a consequence of the different feeding regimens, and whether it related to improved outcome with late-PN. METHODS: This secondary analysis of the EPaNIC-RCT first investigated, with multivariable linear regression analyses, determinants of late-PN-induced CRP rise and its association with cytokine responses (IL-6, IL-10, TNF-α) in matched early-PN and late-PN patients requiring intensive care for ≥ 3 days. Secondly, with multivariable logistic regression and Cox proportional-hazard analyses, we investigated whether late-PN-induced CRP rises mediated infection prevention and enhanced recovery or reflected an adverse effect counteracting such benefits of late-PN. RESULTS: CRP peaked on day 3, higher with late-PN [216(152-274)mg/l] (n = 946) than with early-PN [181(122-239)mg/l] (n = 946) (p < 0.0001). Independent determinants of higher CRP rise were lower carbohydrate and protein intakes (p ≤ 0.04) with late-PN, besides higher blood glucose and serum insulin concentrations (p ≤ 0.01). Late-PN did not affect cytokines. Higher CRP rises were independently associated with more infections and lower likelihood of early ICU discharge (p ≤ 0.002), and the effect size of late-PN versus early-PN on these outcomes was increased rather than reduced after adjusting for CRP rise, not confirming a mediating role. CONCLUSIONS: The higher CRP rise with late-PN, explained by the early macronutrient deficits, did not relate to cytokine responses and thus did not reflect more systemic inflammation. Instead of mediating clinical benefit on infection or recovery, the accentuated CRP rise appeared an adverse effect reducing such late-PN benefits.

Persisting neuroendocrine abnormalities and their association with physical impairment 5 years after critical illness.

BACKGROUND: Critical illness is hallmarked by neuroendocrine alterations throughout ICU stay. We investigated whether the neuroendocrine axes recover after ICU discharge and whether any residual abnormalities associate with physical functional impairments assessed 5 years after critical illness. METHODS: In this preplanned secondary analysis of the EPaNIC randomized controlled trial, we compared serum concentrations of hormones and binding proteins of the thyroid axis, the somatotropic axis and the adrenal axis in 436 adult patients who participated in the prospective 5-year clinical follow-up and who provided a blood sample with those in 50 demographically matched controls. We investigated independent associations between any long-term hormonal abnormalities and physical functional impairments (handgrip strength, 6-min walk distance, and physical health-related quality-of-life) with use of multivariable linear regression analyses. RESULTS: At 5-year follow-up, patients and controls had comparable serum concentrations of thyroid-stimulating hormone, thyroxine (T4), triiodothyronine (T3) and thyroxine-binding globulin, whereas patients had higher reverse T3 (rT3, p = 0.0002) and lower T3/rT3 (p = 0.0012) than controls. Patients had comparable concentrations of growth hormone, insulin-like growth factor-I (IGF-I) and IGF-binding protein 1 (IGFBP1), but higher IGFBP3 (p = 0.030) than controls. Total and free cortisol, cortisol-binding globulin and albumin concentrations were comparable for patients and controls. A lower T3/rT3 was independently associated with lower handgrip strength and shorter 6-min walk distance (p ≤ 0.036), and a higher IGFBP3 was independently associated with higher handgrip strength (p = 0.031). CONCLUSIONS: Five years after ICU admission, most hormones and binding proteins of the thyroid, somatotropic and adrenal axes had recovered. The residual long-term abnormality within the thyroid axis was identified as risk factor for long-term physical impairment, whereas that within the somatotropic axis may be a compensatory protective response. Whether targeting of the residual abnormality in the thyroid axis may improve long-term physical outcome of the patients remains to be investigated. Trial registration ClinicalTrials.gov: NCT00512122, registered on July 31, 2007 ( https://www.clinicaltrials.gov/ct2/show/NCT00512122 ).

Aerobic exercise capacity in long-term survivors of critical illness: secondary analysis of the post-EPaNIC follow-up study.

PURPOSE: To evaluate aerobic exercise capacity in 5-year intensive care unit (ICU) survivors and to assess the association between severity of organ failure in ICU and exercise capacity up to 5-year follow-up. METHODS: Secondary analysis of the EPaNIC follow-up cohort (NCT00512122) including 433 patients screened with cardiopulmonary exercise testing (CPET) between 1 and 5 years following ICU admission. Exercise capacity in 5-year ICU survivors (N = 361) was referenced to a historic sedentary population and further compared to demographically matched controls (N = 49). In 5-year ICU survivors performing a maximal CPET (respiratory exchange ratio > 1.05, N = 313), abnormal exercise capacity was defined as peak oxygen consumption (VO2peak) < 85% of predicted peak oxygen consumption (%predVO2peak), based on the historic sedentary population. Exercise liming factors were identified. To study the association between severity of organ failure, quantified as the maximal Sequential Organ Failure Assessment score during ICU-stay (SOFA-max), and exercise capacity as assessed with VO2peak, a linear mixed model was built, adjusting for predefined confounders and including all follow-up CPET studies. RESULTS: Exercise capacity was abnormal in 118/313 (37.7%) 5-year survivors versus 1/48 (2.1%) controls with a maximal CPET, p < 0.001. Aerobic exercise capacity was lower in 5-year survivors than in controls (VO2peak: 24.0 ± 9.7 ml/min/kg versus 31.7 ± 8.4 ml/min/kg, p < 0.001; %predVO2peak: 94% ± 31% versus 123% ± 25%, p < 0.001). Muscular limitation frequently contributed to impaired exercise capacity at 5-year [71/118 (60.2%)]. SOFA-max independently associated with VO2peak throughout follow-up. CONCLUSIONS: Critical illness survivors often display abnormal aerobic exercise capacity, frequently involving muscular limitation. Severity of organ failure throughout the ICU stay independently associates with these impairments.

Five-year outcome of respiratory muscle weakness at intensive care unit discharge: secondary analysis of a prospective cohort study.

PURPOSE: To assess the association between respiratory muscle weakness (RMW) at intensive care unit (ICU) discharge and 5-year mortality and morbidity, independent from confounders including peripheral muscle strength. METHODS: Secondary analysis of the prospective 5-year follow-up of the EPaNIC cohort (ClinicalTrials.gov: NCT00512122), limited to 366 patients screened for respiratory and peripheral muscle strength in the ICU with maximal inspiratory pressure (MIP) after removal of the artificial airway, and the Medical Research Council sum score. RMW was defined as an absolute value of MIP <30 cmH2O. Associations between RMW at (or closest to) ICU discharge and all-cause 5-year mortality, and key measures of 5-year physical function, comprising respiratory muscle strength (MIP), hand-grip strength (HGF), 6 min walk distance (6MWD) and physical function of the SF-36 quality-of-life questionnaire (PF-SF-36), were assessed with Cox proportional hazards and linear regression models, adjusted for confounders including peripheral muscle strength. RESULTS: RMW was present in 136/366 (37.2%) patients at ICU discharge. RMW was not independently associated with 5-year mortality (HR with 95% CI 1.273 (0.751 to 1.943), p=0.352). Among 156five-year survivors, those with, as compared with those without RMW demonstrated worse physical function (MIP (absolute value, cmH2O): 62(42-77) vs 94(78-109), p<0.001; HGF (%pred): 67(44-87) vs 96(68-110), p<0.001; 6MWD (%pred): 87(74-102) vs 99 (80-111), p=0.009; PF-SF-36 (score): 55 (30-80) vs 80 (55-95), p<0.001). Associations between RMW and morbidity endpoints remained significant after adjustment for confounders (effect size with 95% CI: MIP: -23.858 (-32.097 to -15.027), p=0.001; HGF: -18.591 (-30.941 to -5.744), p=0.001; 6MWD (transformed): -1587.007 (-3073.763 to -179.253), p=0.034; PF-SF-36 (transformed): 1.176 (0.144-2.270), p=0.036). CONCLUSIONS: RMW at ICU discharge is independently associated with 5-year morbidity but not 5-year mortality.

Postoperative Cerebral Oxygen Saturation in Children After Congenital Cardiac Surgery and Long-Term Total Intelligence Quotient: A Prospective Observational Study.

OBJECTIVES: During the early postoperative period, children with congenital heart disease can suffer from inadequate cerebral perfusion, with possible long-term neurocognitive consequences. Cerebral tissue oxygen saturation can be monitored noninvasively with near-infrared spectroscopy. In this prospective study, we hypothesized that reduced cerebral tissue oxygen saturation and increased intensity and duration of desaturation (defined as cerebral tissue oxygen saturation < 65%) during the early postoperative period, independently increase the probability of reduced total intelligence quotient, 2 years after admission to a PICU. DESIGN: Single-center, prospective study, performed between 2012 and 2015. SETTING: The PICU of the University Hospitals Leuven, Belgium. PATIENTS: The study included pediatric patients after surgery for congenital heart disease admitted to the PICU. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: Postoperative cerebral perfusion was characterized with the mean cerebral tissue oxygen saturation and dose of desaturation of the first 12 and 24 hours of cerebral tissue oxygen saturation monitoring. The independent association of postoperative mean cerebral tissue oxygen saturation and dose of desaturation with total intelligence quotient at 2-year follow-up was evaluated with a Bayesian linear regression model adjusted for known confounders. According to a noninformative prior, reduced mean cerebral tissue oxygen saturation during the first 12 hours of monitoring results in a loss of intelligence quotient points at 2 years, with a 90% probability (posterior ß estimates [80% credible interval], 0.23 [0.04-0.41]). Similarly, increased dose of cerebral tissue oxygen saturation desaturation would result in a loss of intelligence quotient points at 2 years with a 90% probability (posterior ß estimates [80% credible interval], -0.009 [-0.016 to -0.001]). CONCLUSIONS: Increased dose of cerebral tissue oxygen saturation desaturation and reduced mean cerebral tissue oxygen saturation during the early postoperative period independently increase the probability of having a lower total intelligence quotient, 2 years after PICU admission.