Background
Bronchopulmonary dysplasia (BPD) is a pediatric chronic lung disease. Pulmonary hypertension (PH) can develop in BPD, greatly increasing morbidity and mortality. PH is the result of decreased pulmonary vascular lumen diameter from vasoconstriction and/or vascular remodeling. Vasoconstriction and vascular remodeling result from activation of abnormal signal transduction cascades including the mitogen‐activated protein kinases (MAPK). MAPK deactivation is regulated by a family of phosphatases called the dual specificity phosphatases (DUSPs). In the present study, we hypothesized that (1) single nucleotide polymorphisms (SNPs) within the DUSP gene family are differentially expressed in a cohort of BPD patients with and without PH and that (2) DUSP SNPs, when combined with clinical data, could be useful as a biomarker for PH development in BPD patients.
Methods
BPD patients (n=188) who were born at <35 weeks gestation and enrolled at Nationwide Children's Hospital were studied. PH was defined by established echocardiographic criteria. DNA from patient blood samples was assayed for 31 SNPs in DUSP genes by Agena Massarray. DUSP SNPs were selected from NCBI database for their putative functionality. Clinical characteristics and minor allele frequencies (MAF) were compared between BPD‐PH (cases) and BPD alone (control) groups using t‐test or chi‐square as appropriate with p‐value <0.05 as significant. Biomarker models were analyzed by calculating area under the curve (AUC) for clinical and SNP data separately and together.
Results
In a cohort of 188 patients with BPD, 61 (32%) showed echocardiographic evidence of PH. Patients with BPD‐PH had a lower birthweight (806g ± 343g vs. 972g ± 414g, p=0.004), gestational age (262 ± 24 vs. 270 ± 25, p=0.041), and were more often small for gestational age (20% vs. 7%, p=0.010) than BPD alone patients. Patients with BPD‐PH received less surfactant (48% vs. 63%, p=0.044), more post‐natal pre‐admit steroids (33% vs. 12%, p=0.001), and more mechanical ventilation (92% vs. 73%, p=0.003) than BPD alone patients. Of the 31 DUSP SNPs evaluated, DUSP‐1 SNP rs322351 was less common in cases than controls (MAF; 0.32 vs. 0.44, p=0.021), and DUSP‐5 SNPs rs1042606 (MAF; 0.44 vs. 0.34, p=0.042) and rs3793892 (MAF; 0.42 vs. 0.32, p=0.049) were more common in cases than in controls. The best fit biomarker model combines clinical (birthweight<700g + post‐natal, preadmission steroids + mechanical ventilation + no surfactant) and SNP data (DUSP1 rs322351 + DUSP5 rs1042606) with AUC 0.76.
Conclusions
In the present study, we identified several potential clinical and genetic biomarkers for BPD‐PH patients including birthweight, gestational age, small for gestational age, surfactant, post‐natal pre‐admit steroids, mechanical ventilation, DUSP‐1 SNP rs322351, and DUSP‐5 SNPs rs1042606 and rs3793892. Combining clinical and DUSP‐1 SNP data yields the most robust biomarker for PH in BPD.
Support or Funding Information
This research was funded by the National Heart, Lung, and Blood Institute (Gran...