Background: Preeclampsia is a multisystem disorder defined by new onset of hypertension with proteinuria after 20 weeks gestation. In part due to dysregulation of pro-angiogenic factors (e.g., placental growth factor [PlGF]) and anti-angiogenic factors (e.g., soluble fms-like tyrosine kinase 1 [sFlt-1]), preeclampsia results in decreased placental perfusion. An increased sFlt-1:PlGF ratio is associated with increased risk of preeclampsia. In this study, we evaluated sFlt-1:PlGF cutoffs and evaluated the clinical performance of sFlt-1:PlGF for predicting preeclampsia. Methods: sFlt-1:PlGF results from 130 pregnant females with clinical suspicion of preeclampsia were used to evaluate the diagnostic accuracy of different sFlt-1:PlGF cutoffs and to compare the clinical performance of sFlt-1:PlGF to traditional preeclampsia markers (proteinuria and hypertension). Serum sFlt-1 and PlGF were measured using Elecsys immunoassays (Roche Diagnostics) and preeclampsia diagnosis was verified by expert chart review. Results: A sFlt-1:PlGF cutoff of >38 yielded the greatest diagnostic accuracy of 90.8% (95% CI, 85.8%–95.7%). Using a cutoff of >38, sFlt-1:PlGF exhibited a greater diagnostic accuracy than traditionally used parameters such as new or worsening proteinuria or hypertension (71.9% and 68.6%, respectively). sFlt-1:PlGF >38 exhibited a negative predictive value (NPV) of 96.4% for rule-out of preeclampsia within 7 days, and a positive predictive value (PPV) of 84.8% for predicting preeclampsia within 28 days. Conclusions: Our study shows the superior clinical performance of sFlt-1:PlGF over hypertension and proteinuria alone to predict preeclampsia at a high-risk obstetrical unit.

BACKGROUND: Our objective was to investigate the predictive and diagnostic accuracy of the angiogenic proteins sFlt-1 (soluble fms–like tyrosine kinase-1) and PlGF (placental growth factor) for preterm preeclampsia and explore the relationship between renal function and these proteins. METHODS: We completed a blinded, prospective, longitudinal, observational study of patients with chronic kidney disease followed at a tertiary center (2018–2023). Serum samples were obtained at 3 time points along gestation (planned sampling): 12–16, 18–22, and 28–32 weeks. In addition, samples were obtained whenever preeclampsia was suspected (indicated sampling). sFlt-1 and PlGF levels remained concealed until the study ended. The primary outcome was preterm preeclampsia. The planned and indicated samples were used to estimate the predictive and diagnostic accuracy of the angiogenic proteins, respectively. RESULTS: Of the 97 participants, 21 (21.6%) experienced preterm preeclampsia. In asymptomatic patients with chronic kidney disease, the angiogenic proteins were predictive of preterm preeclampsia only when sampled in the third trimester, in which case the sFlt-1/PlGF ratio (false positive rate of 37% for a detection rate of 80%) was more predictive than either sFlt-1 or PlGF in isolation. In patients with suspected preeclampsia, the diagnostic accuracy of the sFlt-1/PlGF ratio (false positive rate of 26% for a detection rate of 80%) was higher than that of sFlt-1 and PlGF in isolation. Diminished renal function was associated with increased levels of PlGF. CONCLUSIONS: sFlt-1 and PlGF can effectively predict and improve the diagnostic accuracy for preterm preeclampsia among patients with chronic kidney disease. The optimal sFlt-1/PlGF ratio cutoff to rule out preeclampsia may need to be lower in patients with impaired renal function.

Background: The ratio of the antiangiogenic factor, soluble fms-like tyrosine kinase 1 (sFlt-1), to the proangiogenic factor, placental growth factor (PlGF), is associated with increased risk of preeclampsia. Here, we describe an analytical evaluation of the Elecsys sFlt-1 and PlGF assays at the first North American site in which they were clinically implemented. Methods: The analytical evaluation included short- and long-term imprecision, method comparison, accuracy, linearity, sample stability, limit of quantification verification, and measurement uncertainty. Quality indicators were also evaluated, including turnaround time and repeat test frequency. Results: Short-term (13-day) and long-term (12-month) imprecision for sFlt-1 and PlGF were 0.99 and bias 38, >85, and >110) exhibited negative percent agreement and positive percent agreement of 100%, respectively (n = 40). During the first 12 months, 257 orders were placed, repeat test frequency was 17.5%, mean time between repeat orders was 23 days, and 72.0% of results were reported within 2 h from sample receipt when quality control was run continuously. Conclusions: We describe analytical performance parameters and quality indicators of the Elecsys sFlt-1 and PlGF assays, which was the first North American clinical laboratory site to implement these assays in support of the institution’s high-risk obstetrical unit.

Background Recent studies have reported elevated urinary vitamin D binding protein (uVDBP) concentrations in patients with diabetic kidney disease, although the utility of uVDBP to predict deterioration of kidney function over time has not been examined. Our objective was to assess the association of uVDBP with longitudinal changes in kidney function. Methods Adults at-risk for type 2 diabetes from the Prospective Metabolism and Islet Cell Evaluation (PROMISE) study had 3 assessments over 6 years (n = 727). Urinary albumin-to-creatinine ratio (ACR) and estimated glomerular filtration rate (eGFR) were used as measures of kidney function. Measurements of uVDBP were performed with ELISA and normalized to urine creatinine (uVDBP:cr). Generalized estimating equations (GEE) evaluated longitudinal associations of uVDBP and uVDBP:cr with measures of kidney function, adjusting for covariates. Results Renal uVDBP loss increased with ACR severity at baseline. Subjects with normoalbuminuria, microalbuminuria, and macroalbuminuria had median log uVDBP:cr concentrations of 1.62 μg/mmol, 2.63 μg/mmol, and 2.48 μg/mmol, respectively, and ACR positively correlated with uVDBP concentrations (r = 0.37, p < 0.001). There was no significant association between uVDBP and eGFR at baseline. Adjusted longitudinal GEE models indicated that each SD increase in both baseline and longitudinal uVDBP:cr was significantly associated with higher ACR over 6 years (β=30.67 and β=32.91, respectively). Conversely, neither baseline nor longitudinal uVDBP:cr measures showed a significant association with changes in eGFR over time. Significance These results suggest that loss of uVDBP:cr over time may be a useful marker for predicting renal tubular damage in subjects at risk for diabetes.

Background: The advancement of next-generation sequencing (NGS) technologies provides opportunities for large-scale Pharmacogenetic (PGx) studies and pre-emptive PGx testing to cover a wide range of genotypes present in diverse populations. However, NGS-based PGx testing is limited by the lack of comprehensive computational tools to support genetic data analysis and clinical decisions. Methods: Bioinformatics utilities specialized for human genomics and the latest cloud-based technologies were used to develop a bioinformatics pipeline for analyzing the genomic sequence data and reporting PGx genotypes. A database was created and integrated in the pipeline for filtering the actionable PGx variants and clinical interpretations. Strict quality verification procedures were conducted on variant calls with the whole genome sequencing (WGS) dataset of the 1000 Genomes Project (G1K). The accuracy of PGx allele identification was validated using the WGS dataset of the Pharmacogenetics Reference Materials from the Centers for Disease Control and Prevention (CDC). Results: The newly created bioinformatics pipeline, Pgxtools, can analyze genomic sequence data, identify actionable variants in 13 PGx relevant genes, and generate reports annotated with specific interpretations and recommendations based on clinical practice guidelines. Verified with two independent methods, we have found that Pgxtools consistently identifies variants more accurately than the results in the G1K dataset on GRCh37 and GRCh38. Conclusions: Pgxtools provides an integrated workflow for large-scale genomic data analysis and PGx clinical decision support. Implemented with cloud-native technologies, it is highly portable in a wide variety of environments from a single laptop to High-Performance Computing (HPC) clusters and cloud platforms for different production scales and requirements.