SMFM Abstracts 437
www.AJOG.org
ANGIOGENIC FACTORS AS BIOMARKERS FOR PREECLAMPSIA: THE FINAL STORY? JACOB LARKIN1, SINDHU SRINIVAS2, MARY SAMMEL3, ELLA OFORI4, MICHAL ELOVITZ1, 1University of Pennsylvania, Philadelphia, Pennsylvania, 2Society for Maternal-Fetal Medicine, Philadelphia, Pennsylvania, 3University of Pennsylvania, Biostatistics, Philadelphia, Pennsylvania, 4University of Pennsylvania, Pennsylvania OBJECTIVE: Increased sFlt1 and Endoglin (ENG), and decreased Placental Growth Factor (PGF), have been associated with preeclampsia (PEC). A recent publication showed high sensitivity (sens)/specificity (spec) using sFlt1 & ENG to predict PEC. These studies sought to confirm this association and develop a predictive model for PEC using angiogenic factors. STUDY DESIGN: Cases (108) are prospectively identified with PEC or gestational HTN. Controls (160) were admitted at term. Clinical data and serum were collected. ELISAs measured levels of sFlt1, ENG and PGF. Multivariate analyses were performed comparing cases to controls and severe PEC to controls controlling for confounders (race, GA at delivery, age, history of PEC, CHTN). ROC curves were developed using best models. RESULTS: Protein levels for all factors were different in cases vs. controls (TABLE). Analytes were different between controls and severe PEC. CHTN significantly modifies the association with PGF (interaction P⫽0.03); in the absence of CHTN, the lowest quartile of PGF increases the odds of being a case by 5.5 (P⬍.001). In the presence of CHTN, decreased PGF is not associated with PEC. The best model had 80% spec and 53% sens. CONCLUSION: This study confirms that angiogenic factors are associated with PEC, but that maternal disease modulates the association. Unlike recent reports (AJOG July, 2007), this larger study suggests that these factors may have limited diagnostic utility in predicting PEC for all populations.
439
Systolic BP (mmHg) Diastolic BP (mmHg) Mean BP (mmHg) Heart rate (beats/minute) Cardiac output (L) TPR (mmHg 䡠 s/min)
Angiogenic factors Group Controls (n ⫽ 160) Cases (n ⫽ 108) Severe Cases (n ⫽ 64) P (control vs. PEC, control vs. severe PEC)
sFlt mean ⫹/⫺ SD (median) 6.8 ⫹/⫺ 6.0 (5.5) 12.1 ⫹/⫺ 13.6 (7.0) 14.7 ⫹/⫺ 15.9 (9.9) ⬍0.001, ⬍0.001
Eng mean ⫹/⫺ SD (median)
PGF mean ⫹/⫺ SD (median)
14.0 ⫹/⫺ 11.8 (10.8) 24.2 ⫹/⫺ 21.7 (18.2) 28.3 ⫹/⫺ 23.5 (20.1) ⬍0.001, ⬍0.001
204 ⫹/⫺ 263.9 (130.5) 104 ⫹/⫺ 195.8 (48.8) 90.3 ⫹/⫺ 145.1 (48.6) 0.002, 0.003
438
FIRST TRIMESTER AMBULATORY BLOOD PRESSURE MONITORING WITH CALCULATION OF THE HYPERBARIC INDEX FOR PREDICTION OF PREECLAMPSIA OR PREGNANCY INDUCED HYPERTENSION KARLIJN VOLLEBREGT1, ILJA GUELEN2, JANNEKE GISOLF3, KEES BOER1, HANS WOLF1, 1Academic Medical Center, Obstetrics, Amsterdam, Netherlands, 2BMEYE BV, Amsterdam, Netherlands, 3Academic Medical Center, physiology, Amsterdam, Netherlands OBJECTIVE: Elevated blood pressure is the most powerful risk indicator for pregnancy induced hypertension or preeclampsia. This study evaluates if 48-hour ambulatory blood pressure measurement in women with normal auscultatory blood pressure can be used for prediction of preeclampsia (PE) or pregnancy induced hypertension (PIH). STUDY DESIGN: 48 hour ambulatory blood pressure measurements with Spacelab 90207 were performed at 8-10 weeks gestational age in 190 healthy women (151 nullipara and 39 multipara with previous PE/PIH or fetal growth restriction). All women had normal auscultatory blood pressure (⬍ 140/90 mmHg). Women with antihypertensive medication and pregnancies with congenital malformations were excluded. In the first 90 healthy nulliparous women, who delivered after an uncomplicated pregnancy of an infant with normal birth weight, 24 hour-timeadjusted 90% tolerance limits were calculated according to Hermida et al (Biomed. Instrum.Technol. 1996;30:267). In the remaining 100 women the time-adjusted blood pressure excess over the tolerance limit was calculated (Hyperbaric Index (HBI) in mmHg / hour) for systolic, diastolic and mean blood pressure. The largest value of these parameters (HBImax) was used for classification. RESULTS: Fifteen women developed PE and 13 PIH. HBImax was significantly higher in women with PE (median 28; range 0-178) or PIH (median 3, range 0-95) than in the remaining women (median 0, range 0-333). HBImax ⬎ 15 mmHg had a sensitivity / specificity of 73% / 87% for PE and 50% / 89% for PE or PIH. CONCLUSION: Elevated HBI in the first trimester was associated with later development of PE or PIH. Hermida et al (Hypertension 1998;31:83) published a sensitivity of 93% and specificity of 100 % for prediction of PE or PIH by HBImax in the first trimester with a cut off at 15 mmHg/hour. We could not confirm this high sensitivity and specificity. 0002-9378/$ - see front matter doi:10.1016/j.ajog.2007.10.457
S130
Before
8–10 weeks
After
112 (7.5) 65 (6) 84 (7) 71 (8) 5.8 (1.1) 0.86 (0.18)
114 (10) 62 (7) 83 (8) 76 (10) 7.3 (1.4)*& 0.72 (0.13)*
110 (13) 60 (11) 81 (12) 71 (8) 6.4 (1.1) 0.74 (0.18)*
*different from before pregnancy (p ⬍ 0.05) different from after pregnancy (p ⬍ 0.05) Test sensitivity and specificity for prediction of PE or PIH
&
CONCLUSION: Blood pressure and heart rate values were comparable before, at 8-10 weeks and after pregnancy. However, cardiac output and TPR did not return to values before pregnancy at 4-6 months after delivery. 0002-9378/$ - see front matter doi:10.1016/j.ajog.2007.10.458
440 0002-9378/$ - see front matter doi:10.1016/j.ajog.2007.10.456
LONGITUDINAL COMPARISON OF HEMODYNAMIC MEASUREMENTS BEFORE PREGNANCY, IN EARLY PREGNANCY AND AFTER DELIVERY. KARLIJN VOLLEBRECHT1, SASKIA RANG1, KEES BOER1, HANS WOLF1, 1Academic Medical Center, Obstetrics, Amsterdam, Netherlands OBJECTIVE: To evaluate if hemodynamic parameters before and 4-6 months after delivery are comparable. STUDY DESIGN: The study group consisted of 28 women with a first pregnancy, who participated in a longitudinal study for assessment of cardiovascular risk indicators for preeclampsia. All pregnancies were uneventful and resulted in the delivery of an infant with normal birth weight. Participants were examined before pregnancy, at 8-10 weeks gestational age and 3-6 months after delivery by finger arterial pressure waveform registration using Finometer (Finapres Medical Systems, Amsterdam, NL). Data collection was performed over a period of 10 minutes supine rest after a stable signal had been reached for at least 5 minutes. The pressure registration was corrected by the Return to Flow method. Beat to beat systolic, diastolic and mean blood pressure, cardiac output (CO) and total peripheral resistance (TPR) were averaged over the registration period. RESULTS: Measurement results are presented in the table.
BLOOD PRESSURE MEASUREMENT IN THE FIRST TRIMESTER FOR PREDICTION OF PREECLAMPSIA OR PREGNANCY INDUCED HYPERTENSION KARLIJN VOLLEBREGT1, ILJA GUELEN2, JANNEKE GISOLF3, KEES BOER1, HANS WOLF1, 1Academic Medical Center, Obstetrics, Amsterdam, Netherlands, 2BMEYE BV, Amsterdam, Netherlands, 3Academic Medical Center, physiology, Amsterdam, Netherlands OBJECTIVE: To assess if automated blood pressure measurement in the first trimester of pregnancy is more effective than auscultatory measurement for the prediction of preeclampsia or pregnancy induced hypertension. STUDY DESIGN: 190 healthy women (151 nulliparous, 39 multiparous) with normal blood pressure (⬍ 140/90 mmHg) were recruited for a prospective study with examination at 8-10 weeks, comprising 48-hour blood pressure monitoring (Spacelab 90207, Redmond, WA, USA), and three auscultatory measurements. Data of the first 90 healthy nulliparous women, who delivered after an uncomplicated pregnancy of an infant with normal birth weight, were used for calculation of 90% time adjusted tolerance limits of 48-hour measurement and for calculation of the 90th-percentile for mean 48-hour measurement, mean sleep-time measurement, and mean auscultatory blood pressure. The remaining women were classified according to their excess blood pressure values over the 90% limits. The highest blood pressure excess of systolic, diastolic or mean pressure was used for classification. For time adjusted 48-hour measurement an excess of 15 mmHg/hour was used as cut-off, for the other measurements the 90th percentile was used. RESULTS: Fifteen women developed PE and 13 PIH. Differences of blood pressure measurements between women with PE, with PE or PIH and the normal control group were statistically significant. Table I demonstrates sensitivity and specificity for the tests. CONCLUSION: Time adjusted measurement performs similar as mean sleeptime measurement, and slightly better than mean 48-hour measurement or auscultatory measurement. Test sensitivity and specificity for prediction of PE or PIH
Time adjusted 48-hour measurement Mean 48-hour measurement Sleep-time measurement Auscultatory measurement 0002-9378/$ - see front matter doi:10.1016/j.ajog.2007.10.459
American Journal of Obstetrics & Gynecology Supplement to DECEMBER 2007
PE
PE or PIH
sens. / spec. 0.73 / 0.87 0.67 / 0.84 0.73 / 0.75 0.60 / 0.87
sens. / spec. 0.50 / 0.89 0.54 / 0.88 0.54 / 0.76 0.39 / 0.88