Serum Protein Abundance in Children with Sickle Cell Disease at Baseline, during Acute Pain Crisis, and on Hydroxyurea - Compared to Children with Other Pediatric Diseases

Introduction: Sickle cell disease (SCD) is a multisystem disease, with substantial variation in presentation and clinical course. Many biomarkers have been described both during steady state and with complications, but comparison between SCD and other pediatric disease are lacking. In this prospective study we used a multiplex dual antibody sandwich immunoassay and compared the abundance of 103 serum proteins in 130 children with SCD, 51 morbidly obese children (BMI percentile > 99%), 29 children with inflammatory bowel disease at initial diagnosis, 186 children with pediatric solid tumors and 151 healthy controls. Serum samples of children with SCD were either taken at baseline (59 samples), on the day of admission for pain crisis (27 samples), or while adherent to hydroxyurea for at least 9 months (44 samples). For each set of 5 disease samples, 3 healthy control samples were selected that were matched by age, gender and race. Type of SCD, laboratory values and disease severity based on history of SCD complication were recorded.

Methods: The relative concentration of 103 different serum proteins was measured with a customized quantibody array (Raybiotech, Inc.). Disease and control samples were processed in a random block design stratified by disease, sex, and age to minimize the effect of sample preparation and technician processing. The slides were developed with Alexa Flour 555- conjugated streptavidin and the signal was quantified using a laser scanner (GenePix® 4000A, Molecular Devices Corporation). Measurements from the different PMT voltages were combined into a single composite scan, which was then background corrected and normalized by setting the mean of each array block to the global mean. A linear mixed effects model was constructed for each protein that included gender, race, and disease subgroups as fixed effects factors, with the sample (nested within subject) treated as a random factor. This model was fitted to each protein by restricted maximum likelihood using the lmer function in the R package lme4. Contrasts were constructed using the multcomp package to generate p-values for comparisons of interest. Within each comparison, the false discovery rate was controlled at 5% using the method of Benjamini and Hochberg.

Results: Sixty-one serum proteins were significantly different between children with SCD and healthy subjects, 57 serum proteins between children with SCD and solid tumors, 38 serum proteins between children with SCD and obesity, and 14 serum proteins between children with SCD and inflammatory bowel disease. One-third of the serum proteins (19) were unique for SCD, as they were not significantly different from healthy children in any other of the pediatric diseases examined. Compared to serum of healthy children, 44 proteins were significantly different in children with severe SCD, 46 proteins in children with SCD suffering from an acute pain crisis, 45 proteins in children with SCD on hydroxyurea, and 42 proteins in children with clinically mild sickle cell disease (predominantly Hemoglobin SC disease). Surprisingly only one protein (CCL2) was significantly different between children with SCD at baseline and during an acute pain crisis. Hydroxyurea therapy only caused seven proteins to become significantly different from their pretreatment baseline; and only eight proteins were statistically different at a well baseline visit between children classified as having clinically mild versus severe disease. When compared to children with severe sickle cell disease, only 2 proteins (HGF and BDNF) were lower in both children with clinically mild SCD and in those taking hydroxyurea.

Conclusion: Distinct SCD-specific abundance changes in serum proteins were identified compared to healthy controls and other pediatric disease. Perhaps most interestingly, the protein abundance profiles of the 103 studied proteins in pediatric SCD patients are almost identical between their baseline state and during an acute pain crisis; and minimal protein changes occurred in children with sickle cell disease who were stably taking hydroxyurea with an otherwise good laboratory response (increased MCV and Hgb F). Differences in serum protein abundance profiles could help distinguish mild from severe SCD and should be explored in larger studies.