Abstract
Objective To develop and validate mid-trimester preterm preeclampsia (PE) risk predictors combining clinical factors and serum protein analytes, and to compare their performance with those of widely used clinical and risk assessment algorithms endorsed by professional societies.
Methods This was a secondary analysis of data from two large, multicenter studies in pregnant individuals (PAPR, NCT01371019; TREETOP, NCT02787213), originally conducted to discover, verify, and validate a serum proteomic predictor of preterm birth risk. Serum protein abundances were determined by mass spectrometry. Classifier models combined one or two novel protein ratio(s) with a composite clinical variable, denoted as ClinRisk3, which included prior PE, pre-existing hypertension, or pregestational diabetes. Predictive performance was assessed for the full validation cohort and for a subset that had early gestational age (GA) dating via ultrasound. Classifier performance was compared directly to the U.S. Preventive Services Task Force (USPSTF) algorithm for identification of pregnancies that should receive low-dose aspirin (LDASA) for PE prevention.
Results Nine of nine prespecified classifier models were validated for risk of preterm PE with delivery <37 weeks’ gestation. Areas under the receiver operating characteristic curve ranged from 0.72-0.78 in the full validation cohort, compared to 0.68 for both ClinRisk3 alone and for the USPSTF algorithm. In the early GA dating subcohort, an exemplar predictor, ClinRisk3 + inhibin subunit beta C chain/sex hormone binding globulin (ClinRisk3+INHBC/SHBG) showed a markedly lower screen positive rate (11.1% vs 43.3%) and higher positive predictive value (13.0% vs 5.0%) and odds ratio (9.93 vs 5.24) than USPSTF. Its performance was similar in nulliparas and all parities.
Conclusion Nine preterm PE risk predictors were identified, validated in an independent cohort, and shown to be more predictive than the USPSTF-endorsed algorithm. Our results indicate that a single blood test performed in the first half of pregnancy can be used for personalized PE risk assessment, particularly for pregnancies with minimal or no identified clinical risk factors, including nulliparas. Results can be used to guide personalized pregnancy management, including but not restricted to LDASA for PE prophylaxis, and serve as a basis for developing new prevention strategies.
Funding Source Sera Prognostics, Inc.
Précis Combining novel serum protein biomarkers and selected clinical variables for preterm preeclampsia prediction outperforms widely used clinical risk assessment algorithms currently recommended by practice guidelines.
Competing Interest Statement
ACF, MBB, TCF, TJG, JJB, and PEK are employees of Sera Prognostics, Inc., and receive salary and stock options. ADP is a paid consultant to Sera Prognostics, Inc. GKL was an employee of Sera Prognostics, Inc., but not during the research period for these analyses, nor during manuscript construction and publication. MKH received an Investigator Initiated Grant from Sera Prognostics, Inc., to investigate preterm birth risk stratification and treatment. The other authors have no conflict of interest to declare.
Funding Statement
This study was funded by Sera Prognostics, Inc.
Author Declarations
I confirm all relevant ethical guidelines have been followed, and any necessary IRB and/or ethics committee approvals have been obtained.
Yes
The details of the IRB/oversight body that provided approval or exemption for the research described are given below:
The following is a listing of PAPR (NCT01371019) IRBs: IRB Intermountain Healthcare; IRB for Human Research, Medical University of South Carolina; IRB The University of North Carolina at Chapel Hill; IRB Maricopa Integrated Health System; IRB Baystate Medical Center; IRB Oregon Health & Science University; IRB University of Texas Medical Branch; IRB Christiana Care; WCG IRB (previously Western IRB, used by The Ohio State University, San Diego Perinatal Center, and Regional Obstetrical Consultants). The following is a listing of TREETOP (NCT02787213) IRBs: IRB for Human Research, Medical University of South Carolina; IRB The University of North Carolina at Chapel Hill; IRB Maricopa Integrated Health System; IRB Oregon Health & Science University; IRB University of Texas Medical Branch; IRB Boston Medical Center; IRB Ochsner Clinic Foundation; IRB University of California San Diego; Human Research Protections Program Northwestern University; IRB Indiana University; IRB Duke Medicine; IRB for Clinical Investigations, Greenville Health System; WCG IRB (previously Western IRB, used by Denver Health & Hospital Authority, University of Colorado-Denver, UC-Irvine, Thomas Jefferson University, Regional Obstetrical Consultants, and Baystate Medical Center).
I confirm that all necessary patient/participant consent has been obtained and the appropriate institutional forms have been archived, and that any patient/participant/sample identifiers included were not known to anyone (e.g., hospital staff, patients or participants themselves) outside the research group so cannot be used to identify individuals.
Yes
I understand that all clinical trials and any other prospective interventional studies must be registered with an ICMJE-approved registry, such as ClinicalTrials.gov. I confirm that any such study reported in the manuscript has been registered and the trial registration ID is provided (note: if posting a prospective study registered retrospectively, please provide a statement in the trial ID field explaining why the study was not registered in advance).
Yes
I have followed all appropriate research reporting guidelines and uploaded the relevant EQUATOR Network research reporting checklist(s) and other pertinent material as supplementary files, if applicable.
Yes
Footnotes
Financial support
Financial support for the study was provided by Sera Prognostics, Inc. The sponsor funded the study, generated biomarker data and discovered and blindly validated the predictors.
Added USPSTF algorithm predictive performance comparison.
Data Availability
All data produced in the present study are available upon reasonable request to the authors.
