Abstract
BACKGROUND
In high-resource settings, biomarkers of angiogenic balance, such as the soluble fms-like tyrosine kinase-1 (sFlt1)/placental growth factor (PlGF) ratio, have been studied extensively to aid in evaluation of patients with suspected preeclampsia (PE), and have been incorporated into the 2021 International Society for the Study of Hypertension in Pregnancy definition of PE. The utility in under-resourced settings has not been as well characterized.
OBJECTIVE
This analysis sought to identify the role of the sFlt1/PlGF ratio in the evaluation of patients with or without hypertension who are suspected of having PE without other diagnostic information.
STUDY DESIGN
This is a secondary analysis of a prior prospective study of patients who were presented with suspected PE at ≥20+0 weeks’ gestation at a single academic tertiary care center. Patients were recruited in the parent study from July 2009 to June 2012. In the original study, clinicians were masked to biomarker results, and patients were followed by chart review. In this analysis, the performance of the sFlt1/PlGF ratio (≤38, >38, or >85) was assessed alone in identifying both hypertensive and non-hypertensive patients at risk of evolving into PE with severe features (PE-SF; American College of Obstetricians and Gynecologists’ definition) within two weeks of the triage visit (PE-SF2). Hypertension was defined as a blood pressure (BP)≥140/90 mmHg.
RESULTS
There were 1043 patients included in the analysis; of whom, 579 (55.5%) and 464 (44.5%) presented with or without hypertension, respectively. In triage, 332 (75.4%) of hypertensive patients presented due to BP concerns, and the remainder were evaluated due to other features (new-onset headache, proteinuria, or edema). On triage evaluation, 66.8% of all patients had a normal sFlt1/PlGF ratio ≤38, and 17.0% had an elevated ratio >85. Among hypertensive patients, a sFlt1/PlGF ratio ≤38 was a good rule-out test for PE-SF2 (negative likelihood ratio [LR-] of 0.15), and a ratio >85 was a good rule-in test (positive likelihood ratio [LR+] of 5.75). Among normotensive patients, sFlt1/PlGF was useful as a rule-in test for ratio >38 (LR+ 5.13) and >85 (LR+ 12.80). Stratified by gestational age, sFlt1/PlGF continued to be a good rule in and good rule out test at <35 weeks among those with hypertension but did not have good test performance ≥35 weeks. sFlt1/PlGF had a good test performance as a rule in test for >85 regardless of gestational age. In triage, 4.3% (30/693) of patients with sFlt1/PlGF ratio <38 had concurrent laboratory evidence of PE, compared with 15.9% (28/176) patients with ratio >85.
CONCLUSION
These findings support the potential for the use of sFlt1/PlGF and BP measurement alone in resource-limited settings where other laboratory tests or clinical expertise are unavailable for risk stratification. Performance of the biomarker varied by the presence of hypertension and gestational age.
Key words: angiogenic biomarkers, low resource settings, preeclampsia
AJOG Global Reports at a Glance.
Why was this study conducted?
-
•
This study was conducted to simulate the predictive value of sFlt1/PlGF with and without hypertension in a low resource setting.
What are the key findings?
-
•
sFlt1/PlGF is a good rule in test for patients regardless of hypertension, and a good rule out test for those with hypertension.
What does this add to what is known?
-
•
This study identifies potential utility of angiogenic markers in a low resource setting.
Introduction
Hypertensive disorders of pregnancy (HDPs) affect at least 10% of pregnancies worldwide, with reports of 46,000 maternal and 500,000 fetal/neonatal deaths from preeclampsia annually.1,2 The World Health Organization (WHO) estimates that 16% of maternal deaths in low- and middle-income countries (LMICs) are a result of HDPs, with eclampsia being the most common direct cause of death.3,4 In high-income countries (HICs), progression from preeclampsia to eclampsia complicates fewer than 1% of cases; however, in LMICs, progression is at least double that, at approximately 2.3%.4
Prediction of adverse maternal outcomes in preeclampsia can be achieved with integration of maternal laboratory testing and clinical assessment.5,6 In under-resourced settings where hematologic and biochemical laboratory tests are not routinely available, maternal signs and symptoms can be used, but with reduced predictive performance.7
Among patients with suspected or confirmed preeclampsia, angiogenic biomarkers can improve the identification of patients at highest risk of adverse outcomes.8,9 The two biomarkers most widely studied for their predictive performance are soluble fms-like tyrosine kinase-1 (sFlt1) and placental growth factor (PlGF).10,11 Increased sFlt1 levels, decreased PIGF levels, and an increased sFlt1/PIGF ratio are seen in preeclampsia and are thought to contribute to its characteristic endothelial dysfunction.10, 11, 12 Based on demonstration that, among patients with suspected preeclampsia, an increased sFlt1/PlGF ratio improves identification of patients most likely to progress to preeclampsia within one week, as well prediction of related adverse outcomes, the 2021 International Society for the Study of Hypertension in Pregnancy (ISSHP) guidelines incorporated angiogenic imbalance into their definition of preeclampsia.12, 13, 14 A recent large multisite study in the United States found high positive and negative predictive values for the sFlt1/PlGF ratio to predict progression to preeclampsia with severe features (PE-SF; American College of Obstetrics and Gynecology definition) within 2 weeks among patients admitted with HDPs.15,16
Prior studies of angiogenic biomarkers have been performed primarily in HICs. However, angiogenic biomarkers have potential in LMICs where there is limited availability of trained health professionals and limited ability to perform and interpret a full laboratory evaluation. The early identification of LMIC-resident patients at greatest risk for adverse outcomes from HDPs may reduce maternal and perinatal mortality, by facilitating earlier referral to better-resourced hospitals and optimizing the allocation of limited resources.17 Potentially, assessing angiogenic imbalance could identify those patients who would benefit cost-effectively from full laboratory workup to identify complications of preeclampsia (e.g., hemolysis, elevated liver enzymes and low platelet (HELLP) syndrome, disseminated intravascular coagulation).
This study tested whether the use of the sFlt1/PlGF ratio and blood pressure (BP) measurement alone could provide sufficient information to best direct patient care in patients with suspected preeclampsia, according to whether they were normotensive or hypertensive in triage.
Materials and Methods
This is a secondary analysis of a previously published prospective cohort study of patients with a singleton gestation who presented to obstetric triage for evaluation of suspected preeclampsia at ≥20+0 weeks’ gestation, at a single academic tertiary care center. In brief, patients were recruited in the parent study from July 2009 to June 2012 at a medical center in Boston, Massachusetts (research ethics approval 2009P-000084),18,19 underwent standard preeclampsia evaluation and a blood draw for sFlt1 and PlGF levels, and were managed according to institutional standards of care. Clinicians were masked to sFlt1/PlGF results, and the electronic health record was reviewed for maternal and neonatal care and maternal and perinatal outcomes.
The primary outcome was the sFlt1/PlGF ratio test performance (utilizing cutoffs of ≤38, >38, or >85) in identifying patients at low or high risk of developing PE-SF within two weeks of the triage visit (PE-SF2), stratified by hypertension status on enrollment. Hypertension was defined as a single BP recording of a systolic BP (sBP)≥140 mmHg or diastolic BP (dBP)≥90 mmHg in obstetric triage, regardless of the HDP type or HDP diagnosis. The normotensive group included patients without elevated BP, including those with either chronic or gestational hypertension whose BP was controlled by antihypertensive medication.
The secondary outcomes, stratified by hypertension status in triage, were the sFlt1/PlGF ratio (≤38, >38, or >85) test performance in identifying patients at low or high risk of adverse maternal or perinatal outcomes. Adverse maternal outcomes included at least one of the following: severe hypertension (defined as sBP≥160 mmHg or dBP≥110 mmHg), HELLP syndrome, disseminated intravascular coagulation (DIC), placental abruption, intracranial hemorrhage, pulmonary edema, eclampsia, acute renal failure, or death. Adverse perinatal outcomes included at least one of the following: fetal death, neonatal death, or birthweight less than the 10th percentile for gestational age and sex.20
Statistical Analysis
Baseline characteristics, the nature of presentation to obstetric triage, the assessment results, and pregnancy outcomes were evaluated descriptively, according to hypertension status in obstetric triage. Results were presented as frequency counts and proportions for categorical variables, and median (interquartile range [IQR]) for continuous variables. Skewed distributions of continuous data were observed visually and confirmed using the Shapiro-Wilk test. Differences between those with and without hypertension were evaluated with a Mann Whitney U test, chi-square, or Fisher's Exact test, as appropriate.
To evaluate the diagnostic test properties of the sFlt1/PlGF ratio, sensitivity, specificity, likelihood ratios (negative [LR-] and positive [LR+]), and area under the receiver operating curve were calculated for PE-SF2. Cutoffs for sFlt1, PlGF, and clinical markers were based on previously established cutoffs.11,16,18,19 A LR- <0.20 was considered to demonstrate good diagnostic test performance in ruling out PE-SF2, whereas a LR+ >5.0 was considered to demonstrate good performance in ruling in the possibility of PE-SF2; these cut-offs correspond to a 30% difference in the probability of the outcome.21,22 Areas under the receiver-operator curve (AUC), and associated 95% confidence intervals (CI), were calculated using logistic regression.
Sensitivity analyses were undertaken for the primary and secondary outcomes, stratified by hypertension in triage for test performance of PlGF alone (using the published cut-off of >100pg/mL as normal),23, 24, 25 and dichotomizing the cohorts into patients who presented from 20+0 to 34+6 weeks gestation and those who presented ≥35+0 weeks.
Results
Overall, 1043 patients were included in the analysis, of whom 464 patients presented with normotension and 579 with hypertension. On triage evaluation, 697 (66.8%) patients had a normal sFlt1/PlGF ratio ≤38, and 177 (17.0%) had an elevated ratio >85. Table 1 shows that there were many differences in baseline patient characteristics, nature of presentation and evaluation in triage, and pregnancy outcomes according to the presence or absence of hypertension in triage.
Table 1.
Characteristics of Study Subjects Stratified by Hypertension in Triage
| Entire Cohort N = 1043 | No Hypertension N = 464 | Hypertension N = 579 | P-Value | |
|---|---|---|---|---|
| Demographics | ||||
| Maternal age, years | 32 (28, 35) | 32 (28, 35) | 32 (29, 36) | .01 |
| Body mass index, kg/m2 | 32.4 (28.7, 37.1) | 32.4 (28.5, 37.4) | 32.3 (28.9, 36.9) | .81 |
| Nulliparous | 601 (57.6%) | 242 (52.2%) | 359 (62.0%) | .001 |
| Current smoker | 46 (4.4%) | 23 (5.0%) | 23 (4.0%) | .74 |
| Race | .33 | |||
| White/Caucasian | 692 (66.4%) | 317 (68.3%) | 375 (64.9%) | |
| Black/African American | 166 (15.9%) | 73 (15.7%) | 93 (16.1%) | |
| Asian | 73 (7.0%) | 24 (5.2%) | 49 (8.5%) | |
| Other | 43 (4.1%) | 20 (4.3%) | 23 (4.0%) | |
| Unspecified | 68 (6.5%) | 30 (6.5%) | 38 (6.6%) | |
| Hispanic ethnicity | 104 (10.0%) | 49 (10.6%) | 55 (9.5%) | .80 |
| Past medical history | ||||
| Chronic hypertension | 256 (24.5%) | 82 (17.7%) | 174 (30.1%) | <.0001 |
| Diabetes | 88 (8.4%) | 41 (8.8%) | 47 (8.1%) | .68 |
| Nature of suspected PE(Presenting characteristics) | ||||
| GA at presentation, weeks | 35.6 (32.6, 37.1) | 35.6 (32.0, 37.4) | 35.6 (32.7, 37.1) | .79 |
| BP in triage | ||||
| Systolic BP, mmHg | 138 (128, 149) | 127 (120, 134) | 148 (141, 156) | <.0001 |
| Diastolic BP, mmHg | 87 (79, 95) | 79 (72, 84) | 94 (90, 99) | <.0001 |
| Severe hypertension (≥ 160/110) | 110 (10.6%) | 0 (0%) | 110 (19.0%) | <.0001 |
| Presenting symptoms | ||||
| BP concerns | 670 (65.2%) | 240 (52.4%) | 430 (75.4%) | <.0001 |
| Proteinuria | 145 (14.3%) | 72 (15.9%) | 73 (13.0%) | .20 |
| Headaches | 301 (29.4%) | 142 (31.1%) | 159 (28.0%) | .27 |
| Visual changes | 103 (10.1%) | 51 (11.2%) | 52 (9.2%) | .29 |
| Persistent epigastric/RUQ Pain | 74 (7.3%) | 34 (7.5%) | 40 (7.1%) | .79 |
| Edema | 141 (13.9%) | 63 (13.9%) | 78 (13.8%) | .96 |
| Routine laboratory values in triage | ||||
| Platelet count, K/μL | 237 (196, 281) | 241 (199, 285) | 233 (192, 276) | .09 |
| Low platelet count, <100 K/μL | 11 (1.1%) | 4 (0.9%) | 7 (1.2%) | .76 |
| Creatinine, mg/dL | 0.6 (0.5, 0.7) | 0.6 (0.5, 0.6) | 0.6 (0.5, 0.7) | <.0001 |
| Elevated creatinine, >1.1 mg/dL | 21 (2.0%) | 7 (1.5%) | 14 (2.4%) | .30 |
| AST, U/L | 22 (17, 31) | 21 (17, 26) | 24 (18, 34) | .02 |
| ALT, U/L | 16 (12, 23) | 15 (12, 21) | 17 (12, 24) | .04 |
| Elevated AST or ALT, >80 U/L | 33 (3.2%) | 12 (2.6%) | 21 (3.6%) | .34 |
| Hematocrit, % | 34.5 (32.4, 36.5) | 34.2 (32.1, 36.2) | 34.9 (32.6, 36.8) | .003 |
| Uric acid, mg/dL | 4.6 (3.9, 5.5) | 4.3 (3.7, 5.1) | 4.9 (4.1, 5.9) | <.0001 |
| Other triage characteristics | ||||
| ACOG severe PE definitiona | 105 (10.1%) | 22 (4.7%) | 83 (14.3%) | <.0001 |
| Angiogenic biomarkers in triage | ||||
| sFlt1 | 3458 (2054, 6451) | 2613 (1669, 4606) | 4514 (2591, 8090) | <.0001 |
| PlGF | 203 (105, 433) | 301 (152, 607) | 160 (85, 313) | <.0001 |
| sFlt1/PlGF | 16.6 (4.3, 57.5) | 9.4 (3.1, 25.7) | 31.1 (8.6, 81.7) | <.0001 |
| ≤ 38 | 693 (66.8%) | 379 (82.0%) | 314 (54.5%) | <.0001 |
| 39–85 | 169 (16.3%) | 46 (10.0%) | 123 (21.4%) | |
| > 85 | 176 (17.0%) | 37 (8.0%) | 139 (24.1%) | |
| Antenatal management following triage | ||||
| Admitted following triage | 457 (43.8%) | 133 (28.7%) | 324 (6.0%) | <.0001 |
| Delivery and pregnancy outcomes | ||||
| Final hypertensive diagnosis | <.0001 | |||
| Chronic hypertension | 171 (16.4%) | 66 (14.2%) | 105 (18.1%) | |
| Gestational hypertension | 258 (24.7%) | 97 (20.9%) | 161 (27.8%) | |
| PE or superimposed PE | 364 (34.9%) | 87 (18.8%) | 277 (47.8%) | |
| Severe features | 182 (50.0%) | 35 (40.2%) | 147 (53.1%) | .04 |
| Normal | 233 (22.3%) | 201 (43.3%) | 32 (5.5%) | |
| Other (Gestational Proteinuria) | 17 (1.6%) | 13 (2.8%) | 4 (0.7%) | |
| PE-SF within two weeks of triage | 145 (13.9%) | 23 (5.0%) | 122 (21.1%) | <.0001 |
| Outcomes | ||||
| GA at delivery, weeks | 37.9 (36.3, 39.1) | 38.6 (37.1, 39.4) | 37.3 (35.7, 38.6) | <.0001 |
| Preterm delivery <37 weeks | 142 (13.6%) | 30 (6.5%) | 112 (19.3%) | <.0001 |
| Preterm delivery <34 weeks | 99 (9.5%) | 21 (4.5%) | 78 (13.5%) | <.0001 |
| Delivered within two weeks | 610 (59.2%) | 214 (46.7%) | 396 (69.1%) | <.0001 |
| Indicated delivery | 425 (72.4%) | 102 (50.8%) | 323 (83.7%) | <.0001 |
| Caesarean delivery | 562 (54.4%) | 238 (51.9%) | 324 (56.4%) | .15 |
| Birthweight, grams | 3010 (2355, 3505) | 3280 (2730, 3625) | 2815 (2185, 3360) | <.0001 |
| Macrosomia | 39 (6.6%) | 18 (8.8%) | 21 (5.4%) | .12 |
| Maternal adverse outcomeb | 181 (17.4%) | 23 (5.0%) | 158 (27.3%) | <.0001 |
| Severe hypertension | 146 (14.0%) | 13 (2.8%) | 133 (23.0%) | <.0001 |
| HELLP | 42 (4.0%) | 7 (1.5%) | 35 (6.0%) | .0002 |
| DIC | 2 (0.2%) | 1 (0.2%) | 1 (0.2%) | .88 |
| Placental abruption | 14 (1.3%) | 3 (0.7%) | 11 (1.9%) | .08 |
| Pulmonary edema | 5 (0.5%) | 0 (0%) | 5 (0.9%) | .07 |
| Cerebral hemorrhage | 1 (0.1%) | 0 (0%) | 1 (0.2%) | .37 |
| Death | 1 (0.1%) | 1 (0.2%) | 0 (0%) | .44 |
| Eclampsia | 1 (0.1%) | 0 (0%) | 1 (0.2%) | .37 |
| Acute renal failure | 5 (0.5%) | 2 (0.4%) | 3 (0.5%) | .84 |
| Fetal adverse outcomeb | 38 (3.6%) | 12 (2.6%) | 26 (4.5%) | .10 |
| Fetal death | 4 (0.4%) | 0 (0%) | 4 (0.7%) | .13 |
| Neonatal death | 5 (0.5%) | 2 (0.4%) | 3 (0.5%) | .84 |
| Birthweight<10th percentile | 31 (5.3%) | 10 (4.9%) | 21 (5.5%) | .78 |
Defined as platelet count<100, AST or ALT>80, or creatinine>1.1 either (a) in the presence of chronic hypertension, or (b) severe hypertension (BP>160/110) without chronic hypertension.
Categories are non-exclusive, where subjects can experience multiple adverse outcomes.
Abbreviations: ACOG (American College of Obstetricians and Gynecologists), ALT (alanine aminotransferase), AST (aspartate aminotransferase), BMI (body mass index), BP (blood pressure), DIC (disseminated intravascular coagulation), GA (gestational age), HELLP (Hemolysis, Elevated Liver Enzymes Low Platelet) syndrome, PCR (protein:creatinine ratio), PE (preeclampsia), PE-SF (preeclampsia with severe features) PlGF (placental growth factor), RUQ (right upper quadrant) abdominal, sFlt1 (soluble fms-like tyrosine kinase-1).
In this cohort, patients presented at a median age of 32 years (IQR 28, 35) and had a median body mass index (BMI)>30 kg/m2; 4.4% were smokers and 15.9% were Black/African American. 24.5% of patients had a history of chronic hypertension and 8.4% had a history of pre-gestational diabetes. Patients with hypertension were slightly older, more often nulliparous, and more often had a history of chronic hypertension, compared with normotensive patients.
At presentation to obstetric triage with suspected preeclampsia, the median gestational age was 35.6 weeks (IQR 32.6, 37.1), median sBP 138 mmHg (IQR 128, 149), median dBP 87 mmHg (IQR 79, 95). Generally, patients were assessed in triage due to concerns about BP, headache, and visual symptoms. As expected, hypertensive patients had higher BP levels than normotensive patients, with more frequent severe hypertension, and more frequent presentations with concerns about BP. However, just over half of the normotensive patients presented because of BP concerns, and there were no differences with regards to other presenting symptoms. Patients with hypertension had higher sFlt1-, lower PlGF-levels, and a higher sFlt1/PlGF ratio, although 54.5% of values were normal (≤38) in this cohort.
Following assessment in obstetric triage, hypertensive patients were more frequently admitted to hospital, and their pregnancy outcomes were more likely to be unfavorable. This included more HDPs (preeclampsia and PE-SF2, specifically), earlier gestational ages at giving birth (with a higher frequency of preterm birth and birth within two weeks), and more adverse maternal outcomes, particularly severe hypertension and HELLP syndrome.
Among hypertensive patients, an sFlt1/PlGF ratio ≤38 was a good rule-out test for the development of PE-SF2, (as highlighted in green, reflecting a LR- <0.20), and a ratio >85 was a good rule-in test for PE-SF2 (as highlighted in red, reflecting a LR+ >5.00). This was confirmed in the good discriminative ability of the models in hypertensive patients, both for an sFlt-1/PlGF ratio >38 (AUC 0.78, 95% CI: 0.75, 0.82) and an sFlt1/PlGF ratio >85 (AUC 0.78, 95% CI: 0.74, 0.83) (Table 2). In normotensive patients, an sFlt1/PlGF ratio >38 and >85 were good rule-in tests for PE-SF2. Similar results were seen for patients at 20+0 - 34+6 weeks, irrespective of whether or not hypertension was present. At or after 35+0 weeks, only for normotensive patients at a cut-off of >85, was the ratio a good rule-in test for PE-SF2.
Table 2.
Test Performance of sFlt1/PlGF for PE-SF Within Two Weeks of Triage, Stratified by Blood Pressure in Triage
 |
Green shading reflects use to reassure (i.e., as a good “rule-out” test), whereas red shading reflects use to confirm concern (i.e., as a “rule-in” test). Values in yellow may be considered to confirm concern based off a positive likelihood ratio cutoff of 3.0. Indented values are reported among those individuals meeting the definition of the larger category heading. Hypertension included patients with BP≥140/90 mmHg in triage, whereas patients with no hypertension had BP < 140/90.
Abbreviations: PE (preeclampsia), SENS (sensitivity), SPEC (specificity), LR- (negative likelihood ratio), LR+ (positive likelihood ratio), AUC (area under the receiver operating characteristic curve), BP (blood pressure), PlGF (placental growth factor), sFlt1 (soluble fms-like tyrosine kinase-1), PE-SF (preeclampsia with severe features).
In terms of prediction of maternal adverse outcomes, for those with hypertension, an sFlt-1/PlGF ratio of >85 did not perform well as a rule-in test at any gestational age. In contrast, for normotensive patients, an sFlt-1/PlGF ratio >85 was a good rule-in test for prediction of maternal adverse outcomes in patients at both 20+0 - 34+6 weeks (LR+ 5.38) and ≥ 35+0 weeks (LR+ 5.00; Supplementary Table 1).
Table 3 shows that PlGF alone was not a good rule-in (at <100pg/mL) or rule-out (at ≥100pg/mL) test for PE-SF2, in hypertensive or normotensive patients, at any gestational age, with all LR- results ≥0.20 and all LR+ results ≤5.0. Similar findings were seen for the use of PlGF alone for the prediction of maternal adverse outcomes (Supplementary Table 1). In contrast, PlGF≤100pg/mL was a good rule-in test for prediction of adverse perinatal outcomes in normotensive patients (+LR 6.70; Supplementary Table 2).
Table 3.
Test Performance of PlGF for PE-SF Within Two Weeks of Triage, stratified by Blood Pressure in Triage
| SENS | SPEC | LR - | LR + | AUC (95% CI) | |
|---|---|---|---|---|---|
|
All
patients (N=145/1043, 13.90%) | |||||
| Hypertension | 0.84 | 0.49 | 0.33 | 1.65 | 0.67 (0.63, 0.70) |
| PlGF≤100 | 0.68 | 0.77 | 2.96 | 0.42 | 0.72 (0.68, 0.77) |
| No hypertension | 0.16 | 0.51 | 1.65 | 0.33 | 0.67 (0.63, 0.70) |
| PlGF≤100 | 0.50 | 0.90 | 5.00 | 0.56 | 0.70 (0.59, 0.81) |
|
20–34
weeks (N=85/459, 18.52%) | |||||
| Hypertension | 0.82 | 0.52 | 0.35 | 1.71 | 0.67 (0.62, 0.72) |
| PlGF≤100 | 0.77 | 0.73 | 2.85 | 0.32 | 0.75 (0.69, 0.81) |
| No hypertension | 0.18 | 0.48 | 1.71 | 0.35 | 0.67 (0.62, 0.72) |
| PlGF≤100 | 0.57 | 0.90 | 5.70 | 0.48 | 0.74 (0.60, 0.87) |
|
≥35
Weeks (N=60/584, 10.27%) | |||||
| Hypertension | 0.87 | 0.47 | 0.28 | 1.64 | 0.67 (0.62, 0.72) |
| PlGF≤100 | 0.56 | 0.79 | 2.67 | 0.56 | 0.68 (0.60, 0.75) |
| No hypertension | 0.13 | 0.53 | 1.64 | 0.28 | 0.67 (0.62, 0.72) |
| PlGF≤100 | 0.38 | 0.90 | 3.80 | 0.69 | 0.64 (0.46, 0.82) |
Green shading reflects use to reassure (i.e., as a good “rule-out” test), whereas red shading reflects use to confirm concern (i.e., as a “rule-in” test). Indented values are reported among those individuals meeting the definition of the larger category heading.
Abbreviations: PE (preeclampsia), SENS (sensitivity), SPEC (specificity), LR- (negative likelihood ratio), LR+ (positive likelihood ratio), AUC (area under the receiver operating characteristic curve), BP (blood pressure), PlGF (placental growth factor), sFlt1 (soluble fms-like tyrosine kinase-1), PE-SF (preeclampsia with severe features).
Hypertensive patients with an sFlt1/PlGF ratio >85 were more likely to have elevated liver function tests (LFTs) in triage than those with a ratio ≤38 (12.23% vs 0.64%, P<.0001) and more likely to have thrombocytopenia (2.88% vs 0.64%, respectively, P=.06). Similarly, amongst normotensive patients, those with an sFlt1/PlGF ratio >85 were more likely to have elevated LFTs than those with an sFlt1/PlGF ratio ≤38 (8.11% vs 2.11%, P=.10) and more likely to have thrombocytopenia (5.41% vs 0.53%, respectively, P=.045), Supplementary Table 3.
The test performance of biomarkers to predict abnormal laboratory results in preeclampsia was then investigated. Overall, 21 (2.01%) patients had abnormal creatinine, 11 (1.05%) had abnormal platelet counts, and 33 (3.16%) had abnormal LFTs in triage. In terms of prediction of abnormal laboratory results within two weeks, for those with hypertension, a ratio of >85 performed well as a modest rule-in test at any gestational age for transaminitis (LR+ 3.19) and thrombocytopenia (LR+ 3.74). A cut-off of 38 had a good rule out for thrombocytopenia among all patients (LR- <0.2). In contrast, for normotensive patients, an sFlt1/PlGF ratio >85 was a good rule-in test for prediction of both transaminitis and thrombocytopenia (LR+ 5.71 and 8.14, respectively; Table 4). In addition, among those who had normal laboratory values in triage and with an sFlt1/PlGF ratio <38 only 0.90% developed any abnormal laboratory results within 2 weeks compared with 9.93% among those with an sFlt1/PlGF ratio >85 (P<.0001).
Table 4.
Test Performance of sFlt1/PlGF for Predicting Abnormal Labs Within Two Weeks, Stratified by Blood Pressure in Triage
 |
Green shading reflects use to reassure (i.e., as a good “rule-out” test), whereas red shading reflects use to confirm concern (i.e., as a “rule-in” test). Indented values are reported among those individuals meeting the definition of the larger category heading.
Abnormal laboratory values are defined as platelet count<100, AST or ALT>80, or creatinine>1.1.
Abbreviations: PE (preeclampsia), SENS (sensitivity), SPEC (specificity), LR- (negative likelihood ratio), LR+ (positive likelihood ratio), AUC (area under the receiver operating characteristic curve), PlGF (placental growth factor), sFlt1 (soluble fms-like tyrosine kinase-1).
Discussion
Principal Findings
Among both normotensive and hypertensive patients with suspected preeclampsia, our findings suggest that the sFlt1/PlGF ratio is a useful test for risk-stratification. Among hypertensive patients, an sFlt1/PlGF ratio <38 performed well as a rule-out test for subsequent development of PE-SF2, especially for those presenting at <35+0 weeks’ gestation, and an sFlt1/PlGF ratio >85 was a good rule-in test. Among normotensive patients, an sFlt1/PlGF ratio >38 was a good rule-in test (and >85 was excellent) for subsequent development of PE-SF2, especially for those presenting at <35+0 weeks’ gestation; test performance was poor as a rule-out test. A PlGF threshold of 100pg/mL did not perform well as either a rule-in or rule-out test. Finally, although very few (<2%) of patients with suspected preeclampsia had concurrent abnormal laboratory results, or went on to develop them, at <35+0 weeks’ gestation, an sFlt1/PlGF ratio <38 was reassuring regardless of BP, and a ratio >85 was concerning.
Results in the Context of What is Known
Many studies have shown that angiogenic biomarkers, specifically sFlt-1, PlGF, and their ratio, can be used to predict development of preeclampsia and adverse maternal and perinatal outcomes. Previous studies have shown that a low ratio has a high negative predictive value for development of preeclampsia within one week of presentation and prediction of adverse outcomes within two weeks.26, 27, 28 In addition, the ratio has additive value in patients with confirmed preeclampsia for the prediction of adverse outcomes.9 Most published studies have been conducted in HICs where a multitude of tests and monitoring would be available; however, ready access to testing and monitoring may not be the case in LMICs.7,18,19,29
Our findings demonstrate utility of these angiogenic makers for prediction of preeclampsia with severe features regardless of hypertension status in triage. In addition, our findings show that these biomarkers can be used for risk assessment amongst those with suspected preeclampsia. Although not directly, our study supports findings published in a prior study that showed utility of biomarkers for risk stratification and determine who needs further surveillance or transfer to higher level of care.30
Clinical and Research Implications
With increasing knowledge and availability of interventions, preeclampsia-related maternal and perinatal mortality have decreased significantly in HICs; however, the same cannot be said for LMICs.31 According to the WHO, the disease burden in LMICs is seven times higher than in HICs.32 Protocols are needed for early detection and prognostication to improve evidence-based resource allocation in LMICs.
The current results demonstrate that using a BP cuff and the sFlt1/PlGF ratio, without recourse to further testing, can identify those patients who will develop PE-SF2; this will enable the identification of those patients who need to be transferred to higher levels of maternal and neonatal care. In addition, the results demonstrate that abnormal laboratory results (thrombocytopenia, transaminitis) are uncommon among patients with a low sFlt1/PlGF ratio. Consideration can be made for using the sFlt1/PlGF ratio alongside BP measurement as opposed to a full laboratory panel to optimize the use of limited resources. For example, sFlt1/PlGF testing might be performed as the first test and a full battery of preeclampsia laboratory tests limited to patients with a sFlt1/PlGF ratio >85. In addition, a sFlt1/PlGF ratio <38 on initial evaluation would cost-effectively eliminate repeated clinical laboratory testing among patients with normal initial evaluation laboratory results.
Studies in Europe, South America, and the United States have looked at the cost-effectiveness of adding the sFlt1/PlGF ratio with a cut-off of 38 to standard diagnostic criteria. The results were notable for a significant reduction in hospitalizations, resulting in substantial cost-saving.28,33, 34, 35 Implementing the sFlt1/PlGF ratio with a rule-in at >85 may be lifesaving in LMIC settings, regardless of the presence of hypertension. Although not directly evaluated by this study, the test could be used in high-risk pregnancies as well, such as multiple gestation. As a rule-out test among those with hypertension, this ratio may best allocate resources to those who truly are at higher risk of adverse outcomes. Additionally, the diagnosis of hypertensive disorders currently requires clinical expertise, but discrete cutoff points for the sFlt1/PlGF ratio and hypertension would allow those with limited training to assess and triage patients. There have been prior studies on the cost-effectiveness of the use of biomarkers; however, none have been done in the LMIC setting.35,36 Further studies are needed to evaluate the cost effectiveness of implementation in the LMIC setting, given the need for running the tests and setting up the test platform.
Strengths and Limitations
A major strength of this study is the large sample size. Additionally, the data from the parent study were collected in a prospective manner by trained research staff using standardized forms; thus, there were minimal missing data. Limitations of this study include that the absolute number of adverse events was small. The parent data were collected in the United States in a tertiary care center. Real world data from an LMIC is needed for validation of these results. The tests performed in this study were on a laboratory grade system - a point of care system may be more applicable to the LMIC setting and test performance needs to be evaluated in that setting. Another limitation of the study is that time to blood draw from the time of corticosteroid administration was not collected in the original study; however, we do not anticipate that many, if any, patients would have received corticosteroids in triage when the blood was drawn. Furthermore, the specific diagnosis in triage is limited by the data collected during the original study, as the diagnosis assigned by the clinicians was not collected as part of the study.
Conclusions
Our findings support the use of the sFlt1/PlGF ratio for both normotensive and hypertensive patients with suspected preeclampsia, in resource-limited settings where other laboratory tests or clinical expertise may not be readily available. Test performance as a rule out and rule in test varied according to gestational age and cutoff used. The ratio can be used for risk stratification and resource allocation to the most at-risk patients.
CRediT authorship contribution statement
Easha Patel: Writing – review & editing, Writing – original draft, Visualization, Validation, Methodology. Sunitha Suresh: Writing – review & editing, Writing – original draft, Visualization, Validation, Methodology, Conceptualization. Ariel Mueller: Writing – review & editing, Writing – original draft, Visualization, Validation, Methodology, Formal analysis, Conceptualization. Courtney Bisson: Writing – review & editing, Writing – original draft, Visualization, Validation, Methodology, Conceptualization. Katherine Zhu: Writing – review & editing, Writing – original draft. Stefan Verlohren: Writing – review & editing, Writing – original draft, Visualization, Validation, Methodology, Conceptualization. Peter Von Dadelszen: Writing – review & editing, Writing – original draft, Visualization, Validation, Methodology, Conceptualization. Laura Magee: Writing – review & editing, Writing – original draft, Visualization, Validation, Methodology, Conceptualization. Sarosh Rana: Writing – review & editing, Writing – original draft, Visualization, Validation, Supervision, Resources, Methodology, Investigation, Formal analysis, Data curation, Conceptualization.
Declaration of competing interest
S. Rana reports serving as a consultant for Roche Diagnostics, Thermo Fisher, Beckman Coulter, Siemens and has received research funding from Roche Diagnostics and Siemens for work related to angiogenic biomarkers unrelated to the submitted work. A. Mueller reports serving as a statistical consultant to Roche Diagnostics for a study related to the use of angiogenic factors in preeclampsia. S.V. received speaker fees and participated in advisory boards from Roche Diagnostics, ThermoFisher, Siemens and Beckman Coulter outside the submitted work. All other authors report no conflict of interest. No commercial organization was involved in the conception or delivery of this work.
Acknowledgments
Patient Consent
This is a secondary analysis of prior prospective study during which written informed consent was taken for collection of data.
Sources of Funding
This research received no grant funding from agencies in the public, commercial, or not-for-profit sectors.
Presentation
Preliminary data was presented as poster at the 23rd World Congress of the International Society for Study of Hypertension in Pregnancy from Aug 28-31, 2022, in Montpellier, Occitanie, France.
Condensation
sFlt1/PlGF and blood pressure measurement alone can be used in resource-limited settings to identify high risk patients.
Footnotes
Supplementary material associated with this article can be found in the online version at doi:10.1016/j.xagr.2024.100359.
Appendix. Supplementary materials
References
- 1.Shobo O.G., Okoro A, Okolo M, et al. Implementing a community-level intervention to control hypertensive disorders in pregnancy using village health workers: lessons learned. Implement Sci Commun. 2020;1:84. doi: 10.1186/s43058-020-00076-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Magee L.A., Nicolaides K.H., von Dadelszen P. Preeclampsia. N Engl J Med. 2022;386(19):1817–1832. doi: 10.1056/NEJMra2109523. [DOI] [PubMed] [Google Scholar]
- 3.Firoz T., Sanghvi H, Merialdi M, von Dadelszen P. Pre-eclampsia in low and middle income countries. Best Pract Res Clin Obstet Gynaecol. 2011;25(4):537–548. doi: 10.1016/j.bpobgyn.2011.04.002. [DOI] [PubMed] [Google Scholar]
- 4.Goldenberg R.L., McClure EM, Macguire ER, Kamath BD, Jobe AH. Lessons for low-income regions following the reduction in hypertension-related maternal mortality in high-income countries. Int J Gynaecol Obstet. 2011;113(2):91–95. doi: 10.1016/j.ijgo.2011.01.002. [DOI] [PubMed] [Google Scholar]
- 5.Verlohren S., Droge L.A. The diagnostic value of angiogenic and antiangiogenic factors in differential diagnosis of preeclampsia. Am J Obstet Gynecol. 2022;226(2S):S1048–S1058. doi: 10.1016/j.ajog.2020.09.046. [DOI] [PubMed] [Google Scholar]
- 6.Teela K.C., Ferguson RM, Donnay FA, Darmstadt GL. The PIERS trial: hope for averting deaths from pre-eclampsia. Lancet. 2011;377(9761):185–186. doi: 10.1016/S0140-6736(10)61463-8. [DOI] [PubMed] [Google Scholar]
- 7.Payne B.A., Hutcheon JA, Ansermino JM, et al. A risk prediction model for the assessment and triage of women with hypertensive disorders of pregnancy in low-resourced settings: the miniPIERS (Pre-eclampsia Integrated Estimate of RiSk) multi-country prospective cohort study. PLoS Med. 2014;11(1) doi: 10.1371/journal.pmed.1001589. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Cerdeira A.S., Agrawal S, Staff AC, Redman CW, Vatish M. Angiogenic factors: potential to change clinical practice in pre-eclampsia? BJOG. 2018;125(11):1389–1395. doi: 10.1111/1471-0528.15042. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Suresh S., Patel E, Mueller A, et al. The additive role of angiogenic markers for women with confirmed preeclampsia. Am J Obstet Gynecol. 2022;228(5) doi: 10.1016/j.ajog.2022.10.044. 573.e1–573.e11. [DOI] [PubMed] [Google Scholar]
- 10.Rana S., Burke S.D., Karumanchi S.A. Imbalances in circulating angiogenic factors in the pathophysiology of preeclampsia and related disorders. Am J Obstet Gynecol. 2022;226(2S):S1019–S1034. doi: 10.1016/j.ajog.2020.10.022. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Zeisler H., Llurba E, Chantraine F, et al. Predictive value of the sFlt-1:PlGF ratio in women with suspected preeclampsia. N Engl J Med. 2016;374(1):13–22. doi: 10.1056/NEJMoa1414838. [DOI] [PubMed] [Google Scholar]
- 12.Lim S., Li W, Kemper J, Nguyen A, Mol BW, Reddy M. Biomarkers and the prediction of adverse outcomes in preeclampsia: a systematic review and meta-analysis. Obstet Gynecol. 2021;137(1):72–81. doi: 10.1097/AOG.0000000000004149. [DOI] [PubMed] [Google Scholar]
- 13.Magee L.A., Brown MA, Hall DR, et al. The 2021 International Society for the Study of Hypertension in Pregnancy classification, diagnosis & management recommendations for international practice. Pregnancy Hypertens. 2022;27:148–169. doi: 10.1016/j.preghy.2021.09.008. [DOI] [PubMed] [Google Scholar]
- 14.Droge L.A., Perschel FH, Stutz N, et al. Prediction of preeclampsia-related adverse outcomes with the sFlt-1 (Soluble fms-Like Tyrosine Kinase 1)/PlGF (Placental Growth Factor)—ratio in the clinical routine: a real-world study. Hypertension. 2021;77(2):461–471. doi: 10.1161/HYPERTENSIONAHA.120.15146. [DOI] [PubMed] [Google Scholar]
- 15.Thadhani R L.E., Rana S, Costantine MM, et al. Circulating angiogenic factors stratify hypertensive disorders of pregnancy. NEJM Evid. 2022;1(12) doi: 10.1056/EVIDoa2200161. [DOI] [PubMed] [Google Scholar]
- 16.American College of Obstetricians and Gynecologists' Committee on Practice Bulletins–Obstetrics. Gestational hypertension and preeclampsia: ACOG practice bulletin, number 222. Obstet Gynecol. 2020;135(6):e237–e260. doi: 10.1097/AOG.0000000000003891. [DOI] [PubMed] [Google Scholar]
- 17.van der Meij E., Herklots T, Yussuf A, et al. Retrospective validation study of miniPIERS prediction model in Zanzibar. Int J Gynaecol Obstet. 2021;153(2):300–306. doi: 10.1002/ijgo.13493. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Rana S., Powe CE, Salahuddin S, et al. Angiogenic factors and the risk of adverse outcomes in women with suspected preeclampsia. Circulation. 2012;125(7):911–919. doi: 10.1161/CIRCULATIONAHA.111.054361. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Rana S., Salahuddin S, Mueller A, Berg AH, Thadhani RI, Karumanchi SA. Angiogenic biomarkers in triage and risk for preeclampsia with severe features. Pregnancy Hypertens. 2018;13:100–106. doi: 10.1016/j.preghy.2018.05.008. [DOI] [PubMed] [Google Scholar]
- 20.American College of Obstetricians and Gynecologists Committee on Practice Bulletins–Obstetrics and the Society for Maternal-Fetal Medicine Publications Committee. ACOG practice bulletin No. 204: fetal growth restriction. Obstet Gynecol. 2019;133(2):e97–e109. doi: 10.1097/AOG.0000000000003070. [DOI] [PubMed] [Google Scholar]
- 21.Mahutte, N.G, and Duleba AJ. evaluating diagnostic tests. 2021 Available at: https://www.uptodate.com/contents/evaluating-diagnostic-tests. Accessed 1 September 2023.
- 22.McGee S. Simplifying likelihood ratios. J Gen Intern Med. 2002;17(8):646–649. doi: 10.1046/j.1525-1497.2002.10750.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Duhig K.E., Myers JE, Gale C. Placental growth factor measurements in the assessment of women with suspected preeclampsia: a stratified analysis of the PARROT trial. Pregnancy Hypertens. 2021;23:41–47. doi: 10.1016/j.preghy.2020.10.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Hurrell A., Beardmore-Gray A, Duhig K, Webster L, Chappell LC, Shennan AH. Placental growth factor in suspected preterm pre-eclampsia: a review of the evidence and practicalities of implementation. BJOG. 2020;127(13):1590–1597. doi: 10.1111/1471-0528.16425. [DOI] [PubMed] [Google Scholar]
- 25.Duhig K.E., Myers J, Seed PT, et al. Placental growth factor testing to assess women with suspected pre-eclampsia: a multicentre, pragmatic, stepped-wedge cluster-randomised controlled trial. Lancet. 2019;393(10183):1807–1818. doi: 10.1016/S0140-6736(18)33212-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Verlohren S., Brennecke SP, Galindo A, et al. Clinical interpretation and implementation of the sFlt-1/PlGF ratio in the prediction, diagnosis and management of preeclampsia. Pregnancy Hypertens. 2022;27:42–50. doi: 10.1016/j.preghy.2021.12.003. [DOI] [PubMed] [Google Scholar]
- 27.Hund M., Allegranza D, Schoedl M, Dilba P, Verhagen-Kamerbeek W, Stepan H. Multicenter prospective clinical study to evaluate the prediction of short-term outcome in pregnant women with suspected preeclampsia (PROGNOSIS): study protocol. BMC Pregnancy Childbirth. 2014;14:324. doi: 10.1186/1471-2393-14-324. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Schlembach D., Hund M, Schroer A, Wolf C. Economic assessment of the use of the sFlt-1/PlGF ratio test to predict preeclampsia in Germany. BMC Health Serv Res. 2018;18(1):603. doi: 10.1186/s12913-018-3406-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.von Dadelszen P., Payne B, Li J, et al. Prediction of adverse maternal outcomes in pre-eclampsia: development and validation of the fullPIERS model. Lancet. 2011;377(9761):219–227. doi: 10.1016/S0140-6736(10)61351-7. [DOI] [PubMed] [Google Scholar]
- 30.Soundararajan R., Suresh SC, Mueller A, et al. Real life outpatient biomarker use in management of hypertensive pregnancies in third trimester in a low resource SeTting: ROBUST study. Pregnancy Hypertens. 2021;23:97–103. doi: 10.1016/j.preghy.2020.11.010. [DOI] [PubMed] [Google Scholar]
- 31.Ghulmiyyah L., Sibai B. Maternal mortality from preeclampsia/eclampsia. Semin Perinatol. 2012;36(1):56–59. doi: 10.1053/j.semperi.2011.09.011. [DOI] [PubMed] [Google Scholar]
- 32.Wagnew Meazaw M., Chojenta C, Taddele T, Loxton D, et al. Preeclampsia and eclampsia: its burden and distribution across facilities in Ethiopia. Pregnancy Hypertens. 2022;29:64–71. doi: 10.1016/j.preghy.2022.06.006. [DOI] [PubMed] [Google Scholar]
- 33.Chantraine F., Van Calsteren K, Devlieger R, et al. Enhancing the value of the sFlt-1/PlGF ratio for the prediction of preeclampsia: cost analysis from the Belgian healthcare payers' perspective. Pregnancy Hypertens. 2021;26:31–37. doi: 10.1016/j.preghy.2021.08.113. [DOI] [PubMed] [Google Scholar]
- 34.Ohkuchi A., Saito S, Yamamoto T, et al. Short-term prediction of preeclampsia using the sFlt-1/PlGF ratio: a subanalysis of pregnant Japanese women from the PROGNOSIS Asia study. Hypertens Res. 2021;44(7):813–821. doi: 10.1038/s41440-021-00629-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Khosla K., Espinoza J, Perlaza L, et al. Cost effectiveness of the sFlt1/PlGF ratio test as an adjunct to the current practice of evaluating suspected preeclampsia in the United States. Pregnancy Hypertens. 2021;26:121–126. doi: 10.1016/j.preghy.2021.10.009. [DOI] [PubMed] [Google Scholar]
- 36.Schlembach D., Hund M, Wolf C, Vatish M. Diagnostic utility of angiogenic biomarkers in pregnant women with suspected preeclampsia: a health economics review. Pregnancy Hypertens. 2019;17:28–35. doi: 10.1016/j.preghy.2019.03.002. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.