AUTHORS
Nancy S. Green I, MD, Vincent Mboizi, MMED, MS2, Grace Nambatya, RN, MPH3, Dennis Kalibbala, MSc4, Michael Kawooya, MD5, Samson Kamya Lubowa, MD6, Frank Minja, MD7, Seyedmehdi Payabvash, MD8, Hediyeh Baradaran, MD, MS9, Deogratias Munube, MD10, Paul Bangirana, PhD11, Lawrence S. Honig, MD, PhD12, Richard Idro, MD, PhD13
AFFILIATIONS
1 Columbia University Irving Medical Center, New York, New York, United States
2 Global Health Uganda, Kampala, Uganda
3 Global Health Uganda, Kampala, Uganda
4 Global Health Uganda, Kampala, Uganda
5 Ernest Cook Ultrasound Research and Education Institute, Kampala, Uganda
6 Makerere University College of Health Sciences, Kampala, Uganda
7 Emory University School of Medicine, Atlanta, Georgia, United States
8 Columbia University Irving Medical Center, New York, New York, United States, Columbia University Irving Medical Center, New York, New York, United States
9 Columbia University Irving Medical Center, New York, New York, United States, Columbia University Irving Medical Center, New York, New York, United States
10 Mulago National Referral Hospital, Kampala, Uganda
11 Makerere University College of Health Sciences, Kampala, Uganda
12 Columbia University Irving Medical Center, New York, New York, United States
13 Makerere University College of Health Sciences, Kampala, Uganda
ABSTRACT
Introduction: Progressive cerebrovascular injury in children with sickle cell anemia (SCA) is common in Uganda and other countries in sub-Saharan Africa (SSA). Clinically silent cerebral infarcts (SCI) are the most frequent childhood MRI-detected findings, yet MRI is not generally available in SSA. In an earlier cross-sectional study, we reported >60% SCI prevalence and higher plasma levels of CNS injury-associated neuron-specific neurofilament light chain (NfL) in those with SCI. Here we report a secondary analysis of our 2.5-year Ugandan hydroxyurea (HU) open label, single arm trial with escalation to maximum dose by standard CBC indices. We aimed to assess whether NfL or other blood-based brain biomarkers of brain injury were associated with existing or progressive SCI and potential effects of HU.
Methods: The trial was approved by the Uganda and U.S. Institutional Review Boards for child entry at ages 3-9 years, laboratory-documented SCA, Hb >6.0g/dL and prior total HU exposure of <6 months. In all, 264 children enrolled from Mulago Hospital Sickle Cell Clinic in Kampala. Entry required a well visit without recent illness or transfusion, with normal neurologic history and stroke-focused neurologic exam (PedNIHSS) at each of 3 timepoints: entry, 1.5 (midpoint) and 2.5 years (completion). Transcranial doppler (TCD) velocity (TAMV) was assessed at the same timepoints. A randomly selected MRI sub-sample (N=89) was aged 5-9 to avoid sedation. Brain MRI at enrollment and completion used a 1.5 T scanner with 1mm views. Scans were reviewed by 2 independent neuroradiologists, with a 3rd neuroradiology reviewer for adjudications, each blinded to clinical data and other reads. Infarcts were defined as >3mm lesions in any 1 dimension, with larger infarcts >5mm. Plasma samples were obtained at entry, midpoint and completion, plus from non-SCD sibling controls of similar ages. Samples were analyzed using Simoa Human Neurology 4-Plex A assay (N4PA) kit (Quanterix Corp) for central nervous system markers of injury: NfL, Glial fibrillary acidic protein (GFAP), Tau protein and Ubiquitin carboxy-terminal hydrolase L1 (UCH-L1). Biomarker levels were log-transformed for analyses when not normally distributed.
Results: We analyzed all 84 participants who were scanned at both trial entry and 2.5 year completion: 46 (54.8%) females, mean age 6.7 years. In all, 44 (52.4%) had >1 infarct at entry; 12 (14.3%) had progressed on repeat scan, defined by new or enlarged infarct(s). 1. NfL: 1a) At entry MRI, mean NfL levels were 27% higher (p=.036) in participants with larger infarcts versus no or small infarcts; 1b) The group with infarct progression had strongly trended higher NfL than levels for no or small infarcts at entry and midpoint, and was 34% higher (p=.026) at HU trial completion; 2. GFAP, a marker of astrocyte injury or inflammation, 2a) was 9% higher (p=.021) in the subset with larger infarcts; 2b) GFAP was 10% higher (p=0.44) in the progression group at trial completion; 3. At study entry, TCD velocity did not correlate with NfL or GFAP levels, but became highly correlated with NfL (p<.001) and GFAP (p=.004) at trial completion; 4. Levels of tau protein and UCH-L1, both markers of brain injury (e.g. trauma), did not correlate with MRI findings in this trial or our prior cross-sectional sample; 5. Neither sex nor age correlated with biomarker levels; 6. Non-SCD controls had lower levels of NfL and GFAP.
Conclusions: 1. Elevated NfL plasma levels in those with larger infarcts confirmed and extended our prior findings in an independent pediatric Ugandan SCA sample; 2. Persistently elevated NfL levels were correlated with on-going cerebrovascular injury, suggesting that higher levels may reflect more recent cerebral vascular injury; 3. Timing of GFAP elevations suggest a delayed elevation associated with infarcts at trial entry and progression, perhaps reflecting post-infarct inflammation rather than acute astrocyte injury; 4. Elevated TCD velocity despite prolonged HU therapy was associated with higher levels of both biomarkers. In sum, our results provide novel evidence that NfL, and perhaps GFAP, may be useful for screening children for SCIs and for those at higher risk for progressive infarcts and stroke despite HU therapy. These brain biomarkers, especially NfL, along with TCD for screening for large vessel disease, may provide more accessible cerebrovascular screening than MRI imaging for children with SCA in SSA.
DISCLOSURES
Nancy S. Green I, MD: No relevant conflicts of interest to declare.