Intermittent Hypoxia and Caffeine in Infants Born Preterm (ICAF)

Boston University, Beth Israel Medical Center, University of Massachusetts, Worcester, American SIDS Institute, Walter Reed National Military Medical Center, Dartmouth-Hitchcock Medical Center, Children's Hospital of Philadelphia, Johns Hopkins All Children's Hospital, Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD)

Intermittent Hypoxia and Caffeine in Infants Born Preterm (ICAF) Our proposal will address the critical question: is persisting intermittent hypoxia (IH) in preterm infants associated with biochemical, structural, or functional injury, and is this injury attenuated with extended caffeine treatment? The investigators will study the effects of caffeine on IH in 220 preterm infants born at ≤30 weeks + 6 days gestation. Infants who are currently being treated with routine caffeine, and who meet eligibility criteria, will be enrolled between 32 weeks + 0 days and 36 weeks + 6 days PMA. At enrollment, infants will be started on continuous pulse oximeter recording of O2 saturation and heart rate. If, based on standard clinical criteria, the last dose of routine caffeine is given on or before the day the infant is 36 weeks + 5 days PMA, then on the day following their last dose of routine caffeine treatment, infants will be randomized (110/group) to extended caffeine treatment or placebo. Randomized infants should begin receiving study drug (i.e. 5 mg/kg/of caffeine base, or equal volume of placebo) on the day of randomization, but no later than the third calendar day following the last dose of routine caffeine. Prior to 36 weeks + 0 days PMA, study drug will be given once daily (i.e. 5mg/kg/day) and beginning at 36 weeks + 0 days PMA, study drug will be given twice daily (i.e. 10 mg/kg/day). The last dose of study drug will be given at 42 weeks + 6 days PMA. Pulse oximeter recordings will continue 1 additional week after discontinuing study drug. Two caffeine levels will be obtained, the 1st at one week after beginning study drug, and the 2nd at a target date of 40 weeks + 0 days PMA, but no later than the last day of study drug, whether in hospital or at home. Inflammatory biomarkers will be measured at study enrollment and again at 38 weeks + 0 days PMA, or within 2 calendar days prior to hospital discharge, whichever comes first. Quantitative MRI/MRS should be obtained between study enrollment and 3 calendar days after starting study drug and again at a target date of 43 weeks + 0 days, but no later than 46 weeks + 6 days PMA.

Acute morbidities can contribute to adverse neurodevelopment outcomes in preterm infants born at ≤30 wks gestation, but neural damage occurring after resolution of acute morbidities may be more subtle and related to cycles of inflammation and repair in the developing brain. One possible contributor to these more subtle injuries is intermittent hypoxia (IH), defined as repetitive cycles of hypoxia and re-oxygenation, which occur commonly in convalescent premature infants. Caffeine treatment can improve both motor and cognitive neurodevelopmental outcome in premature infants, especially at higher doses, but mechanisms are unclear.The Caffeine for Apnea of Prematurity (CAP) Trial in infants born preterm at <1250 g reported 1) shorter duration of positive pressure ventilation and reduced rate of bronchopulmonary dysplasia (BPD) in infants treated with caffeine during the early postnatal wks prior to 34-35 wks postmenstrual age (PMA), 2) improved motor function and reduced rates of developmental coordination disorder at 5 years, and 3) diffusion changes by MRI consistent with improved white matter microstructural development. Although potential mechanisms for this caffeine effect were not studied in these reports, a recent study of very preterm infants in postnatal weeks 1-10 showed for the first time a direct link between IH and motor, cognitive and language impairment at 18 months corrected age (adjusted risk gradient p<0.005).Notably, the greatest risk gradient was at postnatal ages 9-10 wk, consistent with a contributory role of later IH present after stopping routine caffeine treatment. These data emphasize the potential importance of recurrent episodes of IH, as convalescing infants approach term-equivalent age, on later cognitive, language and motor impairments. Studies of IH during the early postnatal wks of life in very preterm infants may be due to other mechanisms, including ineffective ventilation or other acute morbidities. However, H in spontaneously breathing convalescing infants is due to ventilatory immaturity with associated respiratory pauses or brief apneas, and has a characteristic pattern of brief desaturation from a normoxic baseline followed by reoxygenation and return to normoxia. This study will assess IH only during spontaneous breathing in infants after resolution of acute morbidities or need for supplemental O2, and approaching term-equivalent age, a time when IH has been shown by other studies to be the consequence of immature breathing regulation. IH during spontaneous breathing related to ventilatory immaturity requires continuous high resolution pulse oximetry recordings for detection, and consists of brief, repetitive cycles of O2 desaturation from a normoxic baseline, followed by return to baseline saturations. These repetitive cycles of reoxygenation following each IH episode are pro-inflammatory and cause oxidative stress, free radical production, and release of pro-inflammatory cytokines. Studies show increased levels of inflammatory biomarkers in animal models of IH-associated obstructive sleep apnea (OSA) and in human subjects with OSA. Although inflammatory biomarkers may be elevated in the first 2-3 postnatal weeks in very preterm infants who develop BPD and neurodevelopmental sequelae, it is unknown if later IH during spontaneous breathing in convalescing preterm infants is associated with inflammation or other biochemical, structural or metabolic acute injury or adverse consequences. Clinically unrecognized IH events are still common after discontinuing routine caffeine treatment, typically at 34-35 weeks PMA. Except for 1 study, however, the potential adverse consequences of IH have not been investigated in human infants. In obstructive sleep apnea, however, even modest amounts of chronic IH have been associated with significant neurocognitive morbidity. Evidence from animal models also shows that IH has significant and long lasting effects on multiple physiological control mechanisms and neurological outcomes. It's hypothesized that persistent IH in spontaneously breathing preterm infants after stopping routine caffeine treatment is associated with acute adverse consequences. The relationship between IH, adenosine, caffeine and brain development is complex and not fully understood. At clinically effective doses, caffeine exerts effects in the brain by blocking adenosine (Ado) A1 and A2A receptors, resulting in respiratory stimulation and increased alertness, vigilance and arousal.44-60 Ado A1 receptor activation contributes to hypoxia-induced reduction in cerebral myelination and ventriculomegaly. Caffeine treatment attenuates the effects of hypoxia, presumably through blockade of Ado A1 receptors. It is thus reasonable to hypothesize that similar mechanisms may be active in the human preterm infant. Caffeine may thus be neuroprotective through two major mechanisms: 1) reducing incidence and severity of IH due to its respiratory stimulatory effects, and 2) reducing pre- and immature oligodendrocyte injury. Brain development progresses through a highly programmed series of events. Myelination in the cerebral hemispheres begins to accelerate at ~30-32 wks and continues to term and beyond, and disturbances in these late gestation developmental processes often result in failure of normal brain growth, abnormal cortical organization, impaired myelination, and connectivity, commonly observed in surviving preterm infants. Persisting IH thus has even greater potential for later neurodevelopmental disability than the IH associated with obstructive sleep apnea. Since IH can be attenuated with extended caffeine, persisting IH may thus be a modifiable cause of a previously unrecognized additional risk for disabilities associated with preterm birth. The period from 33-35 to 42 weeks PMA is a critical time for brain development, and is also a time when significant IH during spontaneous breathing is present, but the adverse effects of this IH are unknown. As the 1st step in understanding acute injury from IH, the investigators address a fundamental and critically important question with high potential public health benefit: does continued caffeine treatment after receiving the last dose of routine caffeine at 32 weeks + 0 days PMA and 36 weeks + 5 days PMA reduce extent of IH and attenuate indicators of acute injury at 43-44 wks PMA? The investigators will assess injury in 4 domains: biochemical (inflammation), structural (MRI), functional and metabolic (MRS). Our proposed study thus has the potential to have major impacts on clinical practice: 1) how clinicians assess and interpret IH, and 2) duration of pharmacological treatment with caffeine. This will be the 1st study in human infants to assess the effects of continuing caffeine treatment in attenuating acute injury indicators associated with IH.

Randomized, placebo-controlled, double-blinded clinical trial A site-stratified randomly-permuted blocked randomization design will be used to ensure balance at each site. Randomization will also be stratified into 2 birth gestational age categories of <28 wks and 28 wks + 0 days to 30 wks + 6 days gestation to ensure balance. The randomization scheme will be developed by the DCC using specialized software, and clinical sites will be provided appropriate enrollment logs and randomization procedures from the research pharmacy.

All clinical team personnel and all research staff (except research pharmacist) will remain blinded to study group until study completion.

Infants in the extended caffeine treatment arm will, beginning the next day after stopping routine caffeine treatment, receive 5 mg/kg/day of caffeine base and increase to 5 mg/kg/twice-a-day (BID) of caffeine base beginning at 36 weeks + 0 days PMA and continuing the BID doses through 42 weeks + 6 days PMA.

Infants in the placebo arm will, beginning the next day after stopping routine caffeine treatment, receive the equivalent (to study drug) volume of placebo daily and increase to the equivalent (to study drug) volume placebo BID through 42 weeks + 6 days PMA.

Infants will be started on oral caffeine base at 5 mg/kg/day. At 36 weeks + 0 days PMA drug dose will be increased to 5 mg/kg BID (total daily dose 10 mg/kg). Dose will be weight-adjusted weekly until NICU (neonatal intensive care unit) discharge. After discharge, all new doses will be calculated from the last weight recorded prior to discharge. The research pharmacy will prepare a bulk oral solution with active drug (caffeine base). While in the hospital, a daily 24-hour supply will be prepared and dispensed. For home administration of study drug, the research pharmacy at each clinical site will prepare and dispense a sufficient quantity of caffeine base solution for outpatient treatment up to 42 weeks + 6 days.

SyrSpend SF Unflavored will be used as the placebo for the control group infants. The volume of the placebo will match the volume of the study drug.

Compare the extent of IH exposure, from randomization through 42 weeks + 6 days PMA (within each gestational week and overall), in infants randomized to extended caffeine treatment to infants assigned to receive placebo.

Extent of IH as measured by seconds below 90% saturation per 24 hours of recorded oximetry data within each week PMA

Compare changes in a panel of inflammation-related cytokines and chemokines, from enrollment to the target age of 38 weeks + 0 days PMA, in infants randomized to extended caffeine treatment to infants assigned to receive placebo.

Plasma concentration of each inflammatory biomarkers between baseline and 38 weeks + 0 days PMA in caffeine-treated compared to placebo-treated infants measured by blood sample. Bulk analyses of inflammatory biomarker plasma concentrations will be performed at Children's National Medical Center using a commercially available 40-plex V-Plex ELISA multi-spot assay (MesoScale Diagnostics, Rockville, MD) to measure inflammation-related proteins from different functional categories of cytokines and chemokines, including growth factors, and adhesion molecules, IFN-alpha, IL-6, Gro/CXCL1, IL-1β, IL-4-6, IL-6 receptor, IL-8, IL-10, IL-13, ICAM-1, myeloperoxidase, CRP, MCP, IGFBP-1, MIP-1a, RANTES, and TNFa. Analytes that show strong trends or significance with this assay may be further analyzed with individual ELISA assays, to confirm the original result.

Inflammatory biomarkers will be measured at study enrollment and again at 38 weeks + 0 days PMA, or within 2 calendar days prior to hospital discharge, whichever comes first.

Compare changes in quantitative MRI structural, microstructural from enrollment to 43-46 weeks PMA, in infants randomized to extended caffeine treatment to infants assigned to receive placebo.

• MRI changes in microstructural measures between baseline and end of study (between 43 weeks + 0 days and 46 weeks + 6 days PMA). MRI acquisition protocols will be standardized and field-tested across all sites in Y1 Q1-3, and MRI calibration studies will be performed to ensure that the MRI scanner properties and parameter settings during the acquisition phase are correct. Conventional MRI studies will be reviewed in a standardized fashion by a pediatric neuroradiologist at each site blinded to study randomization. All MRI data sets will be processed at the Advanced Pediatric Brain Imaging Research Laboratory (DBRL) at Children's National Medical Center. All quantitative MRI outcome measures are continuous and will be performed by a single investigator masked to randomization. Abnormalities of brain development, maturation, the presence of focal destructive ischemic or hemorrhagic lesions, will be documented.

Quantitative MRI/MRS should be obtained between study enrollment and 3 calendar days after starting study drug and again at a target date of 43 weeks + 0 days, but no later than 46 weeks + 6 days PMA.

Examine the association between salivary caffeine concentrations and IH outcomes at the one week post randomization and 40 weeks + 0 days PMA assessments.

Determine whether caffeine effects on changes in inflammatory or MRI biomarkers from baseline to follow-up are mediated by caffeine-related reduced IH.

Changes in inflammatory biomarkers and MRI measures of regional tissue volume between baseline and end of study in relation to IH measures. Bulk analyses of inflammatory biomarker plasma concentrations will be performed at Children's National Medical Center using a commercially available 40-plex V-Plex ELISA multi-spot assay (MesoScale Diagnostics, Rockville, MD) to measure inflammation-related proteins from different functional categories of cytokines and chemokines, including growth factors, and adhesion molecules, IFN-alpha, IL-6, Gro/CXCL1, IL-1β, IL-4-6, IL-6 receptor, IL-8, IL-10, IL-13, ICAM-1, myeloperoxidase, CRP, MCP, IGFBP-1, MIP-1a, RANTES, and TNFa. Analytes that show strong trends or significance with this assay may be further analyzed with individual ELISA assays, to confirm the original result.

Inclusion Criteria: Male and female infants born preterm at ≤30 weeks + 6 days post menstrual age Current treatment with routine caffeine PMA 32 weeks + 0 days - 36 weeks + 6 days Anticipated last dose of routine caffeine will be by 36 weeks + 5 days At least 12 hours of breathing room air with no ventilatory support other than on room air nasal air flow therapy regardless of flow rate, or on room air and receiving nasal CPAP, and relapse not anticipated. Able to tolerate enteral medications It is feasible to administer the first dose of study drug no later than 36 weeks + 6 days PMA Exclusion Criteria: Intraventricular hemorrhage Grade III-IV or cystic periventricular leukomalacia Current or prior treatment for seizures Current or prior treatment for cardiac arrhythmias Known renal or hepatic dysfunction that in the opinion of the investigator would have a clinically relevant impact on caffeine metabolism Major malformation, inborn error of metabolism, chromosomal abnormality Presence of a condition for which survival to discharge unlikely Social, mental health, logistical or other issues that, in the opinion of the investigator, would impact the ability of the family to complete the study

The Executive Committee will be responsible for developing publication procedures and resolving authorship issues. This study will be conducted in accordance with the National Institutes of Health (NIH) Public Access Policy, ensuring that the public has access to the published results of NIH funded research. Scientists are required to submit final peer-reviewed journal manuscripts that arise from NIH funds to PubMed Central upon acceptance for publication. This study will comply with the NIH Data Sharing Policy and Policy on the Dissemination of NIH-Funded Clinical Trial Information and the Clinical Trials Registration and Results Information Submission rule. This trial will be registered at, and results from, submitted to ClinicalTrials.gov. Every attempt will be made to publish results in peer-reviewed journals. Data may be requested from other researchers starting 6 months after the completion of the primary endpoint by contacting the DCC(data collection center).

De-identified patient data will be made available to other investigators at their request providing written required IRB (institutional review board) approval in obtained.

Rhein LM, Dobson NR, Darnall RA, Corwin MJ, Heeren TC, Poets CF, McEntire BL, Hunt CE; Caffeine Pilot Study Group. Effects of caffeine on intermittent hypoxia in infants born prematurely: a randomized clinical trial. JAMA Pediatr. 2014 Mar;168(3):250-7. doi: 10.1001/jamapediatrics.2013.4371.

Dobson NR, Rhein LM, Darnall RA, Corwin MJ, Heeren TC, Eichenwald E, James LP, McEntire BL, Hunt CE; Caffeine Study Group. Caffeine decreases intermittent hypoxia in preterm infants nearing term-equivalent age. J Perinatol. 2017 Oct;37(10):1135-1140. doi: 10.1038/jp.2017.82. Epub 2017 Jul 27.