Original Research ajog.

org

OBSTETRICS
Prospective, randomized, double-blind,
placebo-controlled evaluation of the Pharmacokinetics,
Safety and Efficacy of Recombinant Antithrombin
Versus Placebo in Preterm Preeclampsia
Michael J. Paidas, MD; Allan T. N. Tita, MD, PhD; George A. Macones, MD, MSCE; George A. Saade, MD;
Richard A. Ehrenkranz, MD1; Elizabeth W. Triche, PhD; James B. Streisand, MD; Garrett K. Lam, MD; Everett F. Magann, MD;
David F. Lewis, MD; Mitchell P. Dombrowski, MD; Erika F. Werner, MD; David W. Branch, MD; Mounira A. Habli, MD;
Chad A. Grotegut, MD, MBA, MHSc; Robert M. Silver, MD; Sherri A. Longo, MD; Erol Amon, MD, JD; Kirsten L. Cleary, MD;
Helen Y. How, MD; Sarah R. Novotny, MD; William A. Grobman, MD, MBA; Valerie E. Whiteman, MD1;
Deborah A. Wing, MD, MBA; Christina M. Scifres, MD; Baha M. Sibai, MD

BACKGROUND: Despite expectant management, preeclampsia remote RESULTS: There was no difference in median gestational age at
from term usually results in preterm delivery. Antithrombin, which displays enrollment (27.3 weeks’ gestation for the recombinant human anti-
antiinflammatory and anticoagulant properties, may have a therapeutic role in thrombin group [range, 23.1e30.0] and 27.6 weeks’ gestation for the
treating preterm preeclampsia, a disorder characterized by endothelial placebo group [range, 23.0e30.0]; P¼.67). There were no differences in
dysfunction, inflammation, and activation of the coagulation system. median increase in days gained (5.0 in the recombinant human anti-
OBJECTIVE: This randomized, placebo-controlled clinical trial aimed to thrombin group [range, 0e75] and 6.0 for the placebo group [range,
evaluate whether intravenous recombinant human antithrombin could 0e85]; P¼.95). There were no differences between groups in composite
prolong gestation and therefore improve maternal and fetal outcomes. neonatal morbidity scores or in maternal complications. No safety issues
STUDY DESIGN: We performed a double-blind, placebo-controlled related to recombinant human antithrombin were noted in this study,
trial at 23 hospitals. Women were eligible if they had a singleton preg- despite the achievement of supraphysiological antithrombin
nancy, early-onset or superimposed preeclampsia at 23 0/7 to 30 0/7 concentrations.
weeks’ gestation, and planned expectant management. In addition to CONCLUSION: The administration of recombinant human anti-
standard therapy, patients were randomized to receive either recombinant thrombin in preterm preeclampsia neither prolonged pregnancy nor
human antithrombin 250 mg loading dose followed by a continuous improved neonatal or maternal outcomes.
infusion of 2000 mg per 24 hours or an identical saline infusion until
delivery. The primary outcome was days gained from randomization until Key words: anticoagulation, antiinflammatory, human, hyperten-
delivery. The secondary outcome was composite neonatal morbidity score. sion, phase III clinical trial, pregnancy, premature birth,
A total of 120 women were randomized. prematurity

P reterm preeclampsia (PPE) at 23 to

30 weeks’ gestation, although rela-
tively rare in relation to the overall
experience a high rate of perinatal death
and severe short- and long-term
morbidities.1e4 Expectant management
anticoagulant effects of antithrombin
(AT) are hypothesized to have a thera-
peutic role in PE.10e14 Nonclinical
incidence of preeclampsia (PE), affects is a recommended practice in selected studies support the hypothesis and pro-
approximately 6000 to 8000 pregnancies patients with PPE and consists of careful vide evidence of the efficacy of AT in
per year in the United States and results medical surveillance of the mother and preeclamptic animal models.15,16 The
in the delivery of high-risk and signifi- the fetus with the goal of extending evidence of benefit has been observed in
cantly premature neonates who pregnancy to decrease the prematurity preliminary trials demonstrating that
complications.5,6 plasma-derived AT prolonged gesta-
The current understanding of the PPE tional age (GA) and increased neonatal
Cite this article as: Paidas MJ, Tita ATN, Macones GA, pathophysiology points to a number of weight at birth.17e21
et al. Prospective, randomized, double-blind, placebo-
derangements instigated by the placenta The prospective randomized, double-
controlled evaluation of the Pharmacokinetics, Safety and
Efficacy of Recombinant Antithrombin Versus Placebo in that affect the maternal endothelium and blind, placebo-controlled evaluation of
Preterm Preeclampsia. Am J Obstet Gynecol the kidney, including impaired vascular the Pharmacokinetics, Safety and Effi-
2020;223:739.e1-13. endothelial growth factor signaling, cacy of Recombinant Antithrombin
0002-9378
increased sensitivity to angiotensin II, Versus Placebo in Preterm Preeclampsia
ª 2020 The Author(s). Published by Elsevier Inc. This is an endothelial dysfunction, and inflamma- (PRESERVE-1) trial was designed to test
open access article under the CC BY-NC-ND license (http:// tion and activation of both the coagula- the hypothesis that patients with PPE
creativecommons.org/licenses/by-nc-nd/4.0/).
https://doi.org/10.1016/j.ajog.2020.08.004
tion system and platelets.7e9 The between 23 and 30 weeks’ gestation
pleiotropic antiinflammatory and randomized to receive recombinant

NOVEMBER 2020 American Journal of Obstetrics & Gynecology 739.e1

Original Research OBSTETRICS ajog.org

of 2000 mg rhAT per 24 hours for the

AJOG at a Glance duration of treatment. The total daily
Why was this study conducted? dose was 2250 mg on the first day of
This randomized, placebo-controlled clinical trial aimed to evaluate whether treatment and 2000 mg on subsequent
intravenous recombinant human antithrombin (rhAT) could prolong gestation days. Subjects randomized to placebo
and therefore improve maternal and fetal outcomes. received an identical administration of
normal saline. Continuous infusion
Key findings dosing was planned to continue until a
Compared with placebo, the administration of rhAT in preterm preeclampsia decision to deliver or the woman reached
neither prolonged pregnancy nor improved neonatal or maternal outcomes. a GA of 34 0/7 weeks.
What does this add to what is known? Follow-up visits
Antithrombin use did not lead to pregnancy prolongation or improvement in Posttreatment assessments of the mother
neonatal or maternal outcomes despite previous research suggesting that this were performed at hospital discharge
intervention could be beneficial in the management of early-onset preeclampsia. and approximately 4 to 6 weeks after
delivery. Information on the neonates
was collected when they reached a
human antithrombin (rhAT) in addition Throughout the trial, the randomization postmenstrual age of 36 weeks. If the
to standard of care would demonstrate a scheme was generated by Medidata neonate reached 36 weeks’ post-
longer interval from randomization to (New York City, NY). Only unblinded menstrual age at <28 days after delivery,
delivery (latency) than patients with PPE Medidata staff, selected contract research the final neonatal follow-up was done at
receiving placebo and standard of care. organization, sponsor team, and phar- the 4- to 6-week postdelivery visit.
macy staff at each site had access to the
Methods randomization scheme. Women and in- Outcome measures
Trial design and participants vestigators were blinded to treatment The primary outcome was gestational
We randomly assigned women aged at assignment. Unblinded pharmacy staff days gained, which was defined as the
least 16 years with PPE and who were at prepared the study drug. The definitions GA at delivery minus the GA at
23 0/7 to 30 0/7 weeks’ gestation with a of PE and superimposed PE are pre- randomization. The secondary
singleton fetus to receive either rhAT or sented in the Box. outcome was a 5-point composite
placebo. The primary efficacy endpoint score based on specific fetal or
was the difference in GA from the time of Selection of recombinant human neonatal adverse outcomes associated
randomization into the study to delivery antithrombin dosing for the trial with prematurity (Supplemental
between rhAT-treated and placebo- Mean baseline levels in previous pro- Table 1). Neonatal outcomes of spe-
treated women. Only subjects with PPE spective or interventional studies of AT cific interest were assessed from birth
who were admitted to the hospital and in PE were just below normal or low to the 36 weeks’ postmenstrual age
had a period of evaluation and stabili- normal (60%e82%).18e22 The effect of and the 36 weeks’ postmenstrual age
zation, ideally <48 hours from hospi- several rhAT treatment scenarios on the visit or to the 4 to 6 weeks’ post-
talization, were considered for this study AT activity in pregnant women with PPE delivery visit (if both the 36 weeks’
(Figure). This trial targeted early PPE, was simulated using population phar- postmenstrual age and the 36 weeks’
diagnosed between 23 and 30 weeks’ macokinetic modeling. Predicted cu- postmenstrual age visit occurred <28
gestation, given the significant perinatal mulative AT activities were compared days after delivery). Maternal out-
morbidity and mortality associated with with those reported for the treatment comes of specific interest are pre-
prematurity. We chose weekly cutoffs with plasma-derived AT in PPE.18,21 The sented in Supplemental Table 2.
because we believed these would be the dose of rhAT was targeted to the same All investigators and their corre-
most clinically relevant time periods. We concentrations from plasma AT after sponding participating centers were
sought and received institutional review bolus administration that yielded an ef- skilled in the evaluation and manage-
board approval from 26 participating ficacy signal in the Maki trial.18 These ment of PPE and agreed to follow the
centers. Exclusion criteria are presented target concentrations were determined accepted Society for Maternal-Fetal
in Supplemental Box. through a pharmacokinetic model of Medicine and American College of Ob-
plasma AT. stetricians and Gynecologists (ACOG)
Randomization and study-group The trial drug (either rhATor placebo) published guidelines for delivery. Several
assignments was administered through a dedicated of the centers were also part of the
Eligible women were randomly assigned intravenous line. Subjects randomized to Maternal-Fetal Medicine Units Network.
in a 1:1 ratio to either rhAT or placebo. rhAT received an initial loading dose of Similarly, the neonatal intensive care
Randomization was blocked by study site 250 mg rhAT for 15 minutes, immedi- units (NICUs) of the participating cen-
using variable block sizes of 2 and 4. ately followed by a continuous infusion ters were part of the national NICU

739.e2 American Journal of Obstetrics & Gynecology NOVEMBER 2020

ajog.org OBSTETRICS Original Research

FIGURE
Eligibility, randomization, and assessment

1,321 Women were evaluated for eligibility

1,190 Were excluded

1,120 Did not meet eligibility criteria
438 Incorrect gesta onal age
213 No preeclampsia or superimposed
preeclampsia
126 Required Delivery
119 Were not stable
52 Not a candidate for expectant
management
45 Mul fetal pregnancy
36 Renal disease
19 Treated with heparin
16 Fetal anomaly
14 Alcohol or drug use
6 Seizure disorder
2 An phospholipid an body
syndrome
41 Other
70 Declined to par cipate

120 Underwent randomiza on

62 Were assigned to expectant management

58 Were assigned to expectant
& treatment with recombinant human
management & normal saline as placebo
an thrombin

0 lost to 3 lost to
follow-up follow-up

62 were treated in the ITT popula on 58 were treated in the ITT popula on
60 in the modified ITT popula on 54 in the modified ITT popula on
51 in the per protocol popula on 40 in the per protocol popula on

Patients were randomized at a 1:1 ratio with recombinant antithrombin as a 250 mg loading dose followed by continuous infusion at 2000 mg per 24
hours, or an identical saline infusion. Efficacy endpoints and outcomes were based on the ITT population: all women randomized according to their
treatment assignment. Modified ITT population refers to all women in the ITT population who received study treatment. Per-protocol population refers to
all women in the ITT population with no major protocol deviations.
AT, antithrombin; ITT, intent to treat.
Paidas et al. Recombinant AT vs placebo in preterm preeclampsia. Am J Obstet Gynecol 2020.

networks, which followed commonly normally distributed. An analysis of on logistic regression models in terms of
accepted disease definitions and practice covariance was used to test for a differ- treatment, type of PE (PE vs super-
guidelines. ence in treatment effects between rhAT imposed PE), and GA at randomization.
and placebo, in terms of treatment, type The sample size estimation was based on
Adverse events of PE (PE vs superimposed PE), and GA the results of the study by Maki et al.18
Safety assessments were performed at study randomization (measured in We estimated that 34 women per study
throughout the treatment and follow-up days as a continuous variable). Contin- group would provide a power of 90% to
periods. uous variables were summarized using detect a relative increase of 33% in the
descriptive statistics. Categorical data duration of the primary outcome, that is,
Statistical analysis were summarized by frequencies and days gained in GA, from 7.4 days in the
The Shapiro-Wilk test was used to percentages. For dichotomous out- placebo group to 9.7 days in the rhAT
determine whether the data were comes, odds ratios were estimated based group. We increased the sample size to

NOVEMBER 2020 American Journal of Obstetrics & Gynecology 739.e3

Original Research OBSTETRICS ajog.org

BOX
Definition of preeclampsia and superimposed preeclampsia

Preeclampsia was defined as:

 Gestational hypertension defined as a recorded SBP of 140 mm Hg or DBP of 90 mm Hg on 2 occasions at least 4 h apart (since the
commencement of medical intervention in any facility)
Or
 Severe gestational hypertension defined as SBP of 160 mm Hg or DBP of 110 mm Hg, confirmed with a second assessment within
a short interval (min)
And new onset of any of the following:
 Proteinuria defined as 0.3 g protein per 24 h in a 12e24 h urine collection or PCR of 0.3 mg/mg (on a random sample or any
collection period)
 Platelet count of <100,000/mL
 Serum creatinine concentrations of >1.1 mg/dL in the absence of other renal diseases
 Elevated liver transaminases to twice the upper limit of normal
 Cerebral or visual symptoms
Superimposed preeclampsia was defined as:
 The start of antihypertensive medication, increasing the dose of a currently administered antihypertensive medication or adding a
second antihypertensive medication after 20 wk of pregnancy for SBP of 160 mm Hg or DBP of 105 mm Hg in a subject who had a
history of controlled hypertension before 20 wk of pregnancy
And new onset of any of the following:
 Proteinuria defined as 0.3 g protein per 24 h in a 12e24 h urine collection or PCR ratio of 0.3 mg/mg (on a random sample or any
collection period)
 Platelet count of <100,000/mL
 Serum creatinine concentrations of >1.1 mg/dL in the absence of other renal diseases
 Elevated liver transaminases to twice the upper limit of normal
 Cerebral or visual symptoms
Paidas et al. Recombinant AT vs placebo in preterm preeclampsia. Am J Obstet Gynecol 2020.

60 women per study group to assess either rhAT (62 women) or placebo (58 Primary and secondary outcomes
safety, and we also employed an adaptive women). Demographic and baseline For the primary outcome, there was no
study design to better address the sec- clinical characteristics were similar be- significant difference in the median in-
ondary outcome, that is, the 5-point tween the 2 treatment groups (Table 1). crease in GA between treatment groups
composite score based on specific fetal (5.0 days [range, 0e75] in the rhAT
or neonatal outcomes, given that Delivery information group and 6.0 days [range, 0e85] in the
contemporary robust data exist for pre- The indications for delivery, mode of placebo group; P¼.95). For the sec-
maturity, but not specifically for PPE.23 delivery, and maternal complications of ondary outcome, there was no signifi-
An interim analysis was performed interest are presented in Table 2. There cant difference between treatment
when primary efficacy endpoint data were no significant differences between groups in the mean composite fetal or
were available for the first 60 enrolled the 2 study groups. Neonatal assessments neonatal outcome score. The mean
subjects (50%). immediately after delivery, including composite fetal or neonatal outcome
Apgar scores and body measurements, score was 0.7 (SD, 1.0) in the rhAT
Results were comparable in the rhAT and placebo group and 0.6 (SD, 0.9) in the placebo
Characteristics of the participants groups. The mean birthweight was 1056.9 group (Tables 3 and 4). There was no
From July 2014 to May 2016, 1321 g (standard deviation [SD], 508.16) in the difference in the median increase in
women were screened and a total of 120 rhAT group and 1177.4 g (SD, 572.93) in gestational days gained between the
women were randomized to receive the placebo group. rhAT and placebo groups when the

739.e4 American Journal of Obstetrics & Gynecology NOVEMBER 2020

ajog.org OBSTETRICS Original Research

Compliance and adverse events

TABLE 1
Most women completed the study drug
Maternal demographics and baseline clinical characteristics
treatment (82% overall), with a slightly
Parameter rhAT (n¼61) Placebo (n¼58) higher proportion of study drug
Age at screening (y) completion in the rhAT group (90%)
than the placebo group (72%). Similar
Mean (SD) 29.0 (6.1) 29.3 (6.7)
proportions of neonate subjects in each
Median 29.0 30.0 study group completed the study (93%
Minimum to maximum 18e50 18e44 overall). There were no differences in
Mean (SD) 27.18 (1.96) 27.34 (2.06)
maternal adverse events between
treatment groups. The most frequently
Median 27.29 27.57 observed maternal treatment-emergent
Minimum to maximum 23.1e30.0 23.0e30.0 adverse events (TEAEs) in both treat-
GA at randomization, n (%) ment groups were worsened hyperten-
sion (rhAT, 53.3%; placebo, 46.3%)
23 wk (23 0/7e23 6/7) 4 (6.5) 3 (5.2)
and headache (rhAT, 40.0%; placebo,
24 wk (24 0/7e24 6/7) 5 (8.1) 6 (10.3) 35.2%). No fetus experienced a TEAE
25 wk (25 0/7e25 6/7) 9 (14.5) 6 (10.3) that the investigator considered to be
26 wk (26 0/7e26 6/7) 8 (12.9) 7 (12.1) related to the study drug. There were
no fetal deaths. Among neonates, at
27 wk (27 0/7e27 6/7) 11 (11.7) 12 (20.7)
least 1 serious adverse event was
28 wk (28 0/7e28 6/7) 11 (11.7) 5 (8.6) experienced by 53.3% of neonates in
29 wk (29 0/7e29 6/7) 13 (21.0) 16 (27.6) the rhAT group and 44.4% of neonates
30 wk (30 0/7e30 6/7) 1 (1.6) 3 (5.2) in the placebo group. A total of 5 ne-
onates died, of whom 3 were in the
Race, n (%)
rhAT group and 2 were in the placebo
Asian 2 (3.2) 1 (1.7) group. No neonate experienced a
African American 27 (43.5) 27 (46.6) serious adverse event or death that the
Native American 1 (1.6) 0 (0.0) investigator considered to be related to
the study drug.
White 29 (46.8) 30 (51.7)
RhAT administration was not asso-
Other 3 (4.8) 0 (0.0) ciated with an increased risk of
Ethnicity, n (%) bleeding. The estimated blood loss
Hispanic 6 (9.7) 10 (17.2) after delivery was 500 to <1000 cc for
most subjects in each group (rhAT, 43
Not Hispanic 56 (90.30) 48 (82.8)
subjects [69.4%]; placebo, 34 subjects
Type of preeclampsia, n (%) [59.6%]), and only 2 subjects (3.2%)
Preeclampsia 39 (62.9) 32 (55.2) (both in the rhAT group) had an
Superimposed preeclampsia 23 (37.1) 26 (44.8) estimated blood loss of 1500 cc.
2 Mean changes from the baseline for
BMI (kg/m )
prothrombin time and activated partial
Mean 36.92 (10.20) 35.86 (7.84) thromboplastin time were comparable
Median 34.87 35.61 between the rhAT and placebo groups.
Minimum to maximum 21.3e67.8 19.7e58.3 No statistically significant differences
between the treatment groups in mean
BMI, body mass index; GA, gestational age; rhAT, recombinant human antithrombin; SD, standard deviation.
Paidas et al. Recombinant AT vs placebo in preterm preeclampsia. Am J Obstet Gynecol 2020. changes from the baseline in coagula-
tion parameters were observed. Over-
all, there was a low rate of transfusion
(3 [4.8%] and 0 [0%]) in the rhAT and
groups were stratified at randomization all women in the ITT population who placebo groups, respectively, which
based on the diagnosis of PE vs super- received any study treatment, or per- was well within the accepted previous
imposed PE; presence or absence of protocol population, defined as pa- published experience.
proteinuria; or presence or absence of tients who completed the study without
severe-range blood pressures (BPs). any major protocol deviations, did not Antithrombin activity
Subgroup analyses of modified intent- indicate significant differences between At baseline, the 2 treatment groups had a
to-treat population (ITT), defined as treatment groups. similar mean AT activity (%) (rhAT,

NOVEMBER 2020 American Journal of Obstetrics & Gynecology 739.e5

Original Research OBSTETRICS ajog.org

TABLE 2
Delivery information (ITT population)
Parameter rhAT (n¼62) Placebo (n¼58)
GA at time of delivery (wk)
Mean (SD) 28.77 (2.66) 29.64 (3.24)
Median 28.79 29.64
Minimum to maximum 24.0e34.9 23.1e40.0
Indications for delivery (maternal, fetal, or both), n (%)
Maternal 43 (69.4) 40 (69.0)
Fetal 5 (8.1) 9 (15.5)
Both 14 (22.6) 9 (15.5)
Indications for delivery, n (%)
>34 wk gestation 3 (4.8) 6 (10.3)
Refractory hypertension despite maximal medical intervention 36 (58.1) 28 (48.3)
HELLP 1 (1.6) 0 (0.0)
Thrombocytopenia 5 (8.1) 1 (1.7)
Elevated liver enzymes 10 (16.7) 4 (7.1)
Elevated LDH 5 (8.3) 1 (1.8)
Elevated total bilirubin 0 (0.0) 0 (0.0)
Oliguria with evidence of acute renal failure 0 (0.0) 0 (0.0)
Persistent visual symptoms 0 (0.0) 2 (3.4)
Placental abruption 0 (0.0) 0 (0.0)
Oligohydramnios 2 (3.2) 2 (3.4)
IUGR below fifth percentile 1 (1.6) 0 (0.0)
Pulmonary edema 2 (3.2) 3 (5.2)
Reverse end-diastolic flow on umbilical Doppler ultrasound 2 (3.2) 1 (1.7)
Biophysical score of 4/10 on 2 occasions 0 (0.0) 1 (1.7)
Nonreassuring fetal heart rate tracing 15 (24.2) 12 (20.7)
Intractable headache unrelieved with analgesia 8 (12.9) 10 (17.2)
Intractable right upper quadrant abdominal pain or vomiting 5 (8.1) 1 (1.7)
Other 17 (27.4) 14 (24.1)
Maternal complications of interest
Eclamptic seizure 0 (0.0) 1 (1.7)
HELLP 1 (1.6) 0 (0.0)
Oliguria with evidence of acute renal failure 0 (0.0) 0 (0.0)
Pulmonary edema 2 (3.2) 3 (5.2)
Stroke 0 (0.0) 0 (0.0)
Mode of delivery, n (%)
Vaginal delivery 5 (8.1) 11 (19.0)
Cesarean delivery 57 (91.9) 47 (81.0)
Blood transfusion required within 24 h after delivery, n (%) 3 (4.8) 0 (0.0)
GA, gestational age; HELLP, hemolysis, elevated liver enzymes, and low platelet; ITT, intent to treat; IUGR, intrauterine growth restriction; LDH, lactate dehydrogenase; rhAT, recombinant human
antithrombin; SD, standard deviation.
Paidas et al. Recombinant AT vs placebo in preterm preeclampsia. Am J Obstet Gynecol 2020.

739.e6 American Journal of Obstetrics & Gynecology NOVEMBER 2020

ajog.org OBSTETRICS Original Research

data suggest a more severe obstetrical

TABLE 3
history among subjects from the rhAT
Secondary outcome: composite endpoint of fetal and neonatal outcomes
group. In contrast, the percentage of
(ITT)
subjects who had previously experienced
Adjusted PE was lower in the rhAT group than the
Parameter difference, placebo group (28% and 40%, respec-
population Placebo Difference 95% CI tively). Although the diagnostic criteria
statistic rhAT (n¼60) (n¼56) (rhATplacebo) (rhATplacebo) P value
used for subjects with superimposed PE
Composite endpoint of fetal and neonatal outcomes were comparable between groups,
Mean (SD) 0.7 (1.0) 0.6 (0.9) 0.1 0.0 (0.3 to 0.4) .7824 women in the rhAT group who were
Median 0.0 0.0 — — — diagnosed as having PE may have been in
a more severe condition at the time of
Minimum to 0e4 0e4 — — — diagnosis than subjects in the placebo
maximum
group. In the superimposed PE sub-
Fetal and neonatal events include BPD, IVH (grade 3), cystic PVL, ROP (stage 3), late sepsis, and NEC (Bell’s stage 2).
Scores: 0, no events, no mortality; 1, 1 event, no mortality; 2, 2 events, no mortality; 3, 3 or more events, no mortality; 4, death. group, nearly twice as many patients in
The secondary efficacy endpoint is a composite endpoint of specific fetal and neonatal outcomes consisting of BPD, IVH, PVL, the rhAT group were diagnosed as hav-
ROP, late sepsis, NEC, and mortality (fetal and neonatal). Multiple imputation methods were used for the imputation of any
missing data for the secondary efficacy endpoint. The regression model included treatment, type of PE (PE vs superimposed PE), ing the criteria of severe gestational hy-
and GA at randomization (measured in days as a continuous variable). For subjects for whom both 36-week PMA and the 36- pertension than the placebo group (17 vs
week PMA visit occurred in <28 days after delivery, the secondary outcome endpoint includes events occurring during the 4e6
week postdelivery visit. 9, respectively). In addition, the mean
BPD, bronchopulmonary dysplasia; CI, confidence interval; GA, gestational age; ITT, intent to treat; IVH, intraventricular systolic and diastolic BP used for diag-
hemorrhage; NEC, necrotizing enterocolitis; PE, preeclampsia; PMA, postmenstrual age; PVL, periventricular leukomalacia; nosis was higher in the rhAT group
rhAT, recombinant human antithrombin; ROP, retinopathy of prematurity; SD, standard deviation.
Paidas et al. Recombinant AT vs placebo in preterm preeclampsia. Am J Obstet Gynecol 2020.
(rhAT, 165.2/97.5 mm Hg; placebo,
159.1/95.7 mm Hg). Ten patients in the
rhAT group developed thrombocyto-
penia or hemolysis, elevated liver en-
94.70 [SD, 14.60]; placebo, 93.05 [SD, In our study, the rhAT and placebo zymes, and low platelet (HELLP)
14.29]). Maximum ATactivity (%) in the groups were generally well balanced at syndrome compared with 2 patients in
rhAT group was 291.7 (observed at day baseline. The treatment groups had a the placebo group. Overall, these few
8). A total of 15 patients had baseline AT similarly high mean body mass index differences between the rhAT and pla-
levels below 80% (8 in the rhAT group (BMI) at screening (36e37 kg/m2). We cebo are unlikely to account for the lack
and 7 in the placebo group), and their do acknowledge that there were some of efficacy of rhAT in prolonging gesta-
results were consistent with the overall potentially important differences in the tion in this trial. We were unable to
results of the study. The mean ATactivity respective treatment group profiles at detect any benefit of rhAT administra-
in umbilical cord blood was 28.25 (SD, baseline. For example, there were dif- tion in any PE subgroup.
8.63) in the rhAT group and 31.44 (SD, ferences between the groups in obstet- The efficacy results in our trial are not
14.46) in placebo group subjects, indi- rical history. The percentage of first-time consistent with earlier results from the
cating that rhAT did not cross the pregnancies was higher in the rhAT Maki study in Japan18 that showed a
placenta. group (37%) than the placebo group significant and clinically relevant pro-
(22%), with an even greater difference in longation of pregnancy of approximately
Comment the number of subjects who had no 1 week with ATreplacement (16.8 days vs
The key findings of our study are as previous live births (rhAT, 60%; placebo, 10.2 days for the placebo group). The
follows: (1) rhAT administration to 28%), corresponding with higher rates Maki study and the current study pop-
women with singleton gestation diag- of previous neonatal morbidity, miscar- ulation differed in background charac-
nosed as having PPE between 23 and 30 riage, and abortion in the rhAT group. teristics, most notably in GA difference,
weeks’ gestation did not extend their The mean number of previous live births racial composition, and BMI, which
pregnancy more than the expectant in the placebo group was double than were substantially higher in our study. In
management alone; (2) the median in- that of the rhAT group (rhAT, 0.7; pla- the Maki study, the mean GA in the AT-
crease in GA was 5.0 days in the rhAT cebo, 1.4). Of the subjects who did have treated and placebo groups at randomi-
group and 6.0 days in the placebo group; previous deliveries, the mean minimum zation was 31.8 (SD, 3.2) and 31.7 (SD,
(3) rhATadministration did not improve GA for previous deliveries was 4 weeks 2.7) weeks, respectively, which was
individual or composite fetal or neonatal less in the rhAT group in comparison nearly 1 month later than that in our
outcomes; and (4) there were no safety with the placebo group (31 weeks vs 35 study population. In the Maki study,18
issues related to rhAT administration in weeks). The intrauterine growth re- patients diagnosed as having PE be-
the setting of PPE, despite supra- striction at baseline was more frequent tween 24 to 35 weeks’ gestation were
physiological AT levels and prolonged in the rhAT group than the placebo included if the patients experienced a
administration. group (27% vs 19%, respectively). These systolic BP of 160 mm Hg and/or

NOVEMBER 2020 American Journal of Obstetrics & Gynecology 739.e7

Original Research OBSTETRICS ajog.org

diastolic BP of 110 mm Hg and/or which was similar to our study. In our the overall rate of neonatal morbidity
proteinuria of 2 g/L of protein in a 24- study, we enrolled singleton patients was quite low, and the mean composite
hour urine collection and a Gestosis In- with PE diagnosed between 23 to 30 score was <1.
dex of  6 on 2 occasions at least 6 hours weeks’ gestation who were being expec-
apart despite bed rest. Patients with tantly managed. We included patients Limitations
chronic hypertension, renal disease, with chronic hypertension, and AT levels Although 23 sites participated in the
diabetes mellitus, systemic lupus ery- were not a criterion for entry into the trial, only 7 (30%) enrolled more than 5
thematosus, multiple pregnancy, and study. We excluded patients with multi- patients, which attests to the challenge of
other severe medical conditions were ple gestation. an interventional trial in a pregnant
excluded. Patients with AT deficiency We found that baseline AT activity population with a relatively rare disease.
were also excluded. Compared with our levels were higher than expected, Initial enrollment into the trial was slow,
study, Maki et al18 included patients at a consistent with the findings of the so the upper limit of GA inclusion
later GA and required higher BPs. D’Angelo study and at odds with the criteria was extended from 28 to 30
Other supportive evidence for the use Maki study and the initial study weeks in an effort to accelerate study
of AT replacement in PPE came from the describing AT activity in PE.25 D’Angelo enrollment. However, this change
study of Paternoster et al21 who found et al24 found normal mean (SD) AT ac- contributed to a very low neonatal
that higher doses of AT replacement in tivity levels at baseline in most ran- composite score. It would have been
patients with severe PE between 24 and domized patients with PE (88%16% difficult to show a neonatal benefit even
33 weeks’ gestation were associated with and 85%85% for the patients receiving if rhATextended pregnancy significantly.
greater prolongation of pregnancy. In AT and placebo, respectively). Peak AT The diagnosis of PE was based on the
the Paternoster study,21 subjects diag- activity levels in our study were consis- definition of PE in place at the time,
nosed as having PE between 24 to 33 tent with the Maki study, which found a which allows for some patients to be
weeks were included in the study on the peak plasma AT activity of 200% diagnosed as having PE based on sub-
basis of the following criteria: diastolic immediately after the AT infusion, and jective criteria (eg, severe headache)
arterial pressure of >90 mm Hg and the D’Angelo study (244%). Hence, along with BP criteria. Another poten-
systolic arterial pressure of >140 mm Hg insufficient dosing is unlikely to account tially challenging diagnostic criterion
on more than 3 occasions, proteinuria for our negative efficacy results and the relates to the BP criteria for super-
(in a 24-hour urine sample) of >0.3 g/L, study by D’Angelo. Moreover, our imposed PE. The guidelines suggested
and AT of 75%. Patients with chronic continuous daily infusion of rhAT would that “a sudden increase in blood pressure
hypertension, renal disease, diabetes also have mitigated against AT elimina- that was previously well controlled”
mellitus, systemic lupus erythematosus, tion losses previously reported in pa- satisfies the PE criteria,6 which is
multiple pregnancies, and other severe tients with PE.19 imprecise. We insisted on an increase in
medical conditions were excluded. Another consideration that might BP requiring either a start of an antihy-
Compared with our study, Paternoster explain our negative findings is the lack pertensive agent, increasing the dose of a
et al21 included patients at a later GA and of homogeneity of our patient popula- current agent, or adding a second agent,
required decreased AT activity for tion. Just before the study started, the which increased the likelihood of
enrollment. ACOG released new diagnostic criteria correctly identifying patients with
Toward the end of our study, D’An- for PE.6 We incorporated the new superimposed PE.
gelo et al24 reported on a multicenter criteria into our current protocol. Pro-
randomized trial with negative results, teinuria was no longer required for the Strengths
using high-dose AT replacement in PPE. diagnosis of PE. The median prolonga- The PRESERVE-1 trial represents the
These investigators had planned to enroll tion of pregnancy in the subgroup of largest, most comprehensive study of
240 patients from 13 centers but only patients without proteinuria at baseline PPE ever completed in the United States.
enrolled 38 patients; the sponsor termi- was 24 days in contrast to 5 days in the The trial was completed in <2 years,
nated the trial because of poor enroll- group that had proteinuria. which is remarkable considering the
ment. In the D’Angelo study,24 patients Another possibility for the lack of ef- rarity of the patient population and the
diagnosed as having PE at <30 weeks’ fect in our study may be related to the operational complexity of the trial
gestation were included. The criteria for timing of study drug administration in design. The study was adequately pow-
the diagnosis of PE were diastolic BP relation to the clinical status of the pa- ered to address the primary outcome. An
repeatedly at >90 mm Hg plus a daily tient and when the patient was admitted adaptive study design enabled a sample
proteinuria of 0.3 g. The exclusion to the hospital. It is possible that the size reestimation. Only principal in-
criteria were conditions requiring im- delay in the timing of study drug vestigators and centers comfortable with
mediate delivery. Compared with our administration affected the ability to expectant management participated in
study, D’Angelo et al24 included patients detect a treatment difference. Owing to the trial. The study inclusion and
at a similar GA and did not require enrollment challenges, the GA was exclusion criteria were based on the
decreased AT activity for enrollment, extended to 30 weeks. Consequently, current guidelines at the time.5,6 The

739.e8 American Journal of Obstetrics & Gynecology NOVEMBER 2020

ajog.org OBSTETRICS Original Research

TABLE 4
Fetal or neonatal outcomes of specific interest
rhAT (n¼62), Placebo (n¼58), Difference (%) Odds ratio (95% CI)
Parameter n (%) n (%) (rhATplacebo) (rhATplacebo) P value
Fetal death (including stillbirth) 0 (0.0) 0 (0.0) 0.0 N/A —
Neonatal death 3 (4.8) 2 (3.4) 1.4 1.541 (0.197e12.033) .6799
SGA of <10% 12 (19.4) 9 (15.5) 3.9 1.384 (0.514e3.726) .5201
RDS 59 (95.2) 51 (87.9) 7.3 2.724 (0.665e11.154) .1635
BPD 22 (35.5) 20 (34.5) 1.0 0.989 (0.456e2.142) .9774
ROP stage 3 0 (0.0) 0 (0.0) 0.0 N/A —
NEC (Bell’s grade 2) 1 (1.6) 2 (3.4) 1.8 0.493 (0.043e5.642) .5692
IVH 3 3 (4.8) 1 (1.7) 3.1 2.606 (0.257e26.460) .4179
Cystic PVL 2 (3.2) 1 (1.7) 1.5 1.250 (0.101e15.475) .8618
Early sepsis (positive blood culture not 0 (0.0) 0 (0.0) 0.0 N/A —
regarded as contaminant within 72 h of
delivery)
Late sepsis (positive blood culture not 2 (3.2) 4 (6.9) 3.7 0.540 (0.060e3.232) .5001
regarded as contaminant >72 h of delivery)
Meningitis 0 (0.0) 1 (1.7) 1.7 N/A .9441
Avoidance of neonatal morbidity (BPD, IVH 6 (58.1) 33 (56.9) 1.2 1.114 (0.521e2.382) .7813
grade 3, cystic PVL, ROP stage 3, late
sepsis, and NEC Bell’s grade 2) and of
fetal or neonatal mortality
For subjects who did not complete the study, data up to their discontinuation were used. The odds ratio and 95% CI and the P value are for the treatment effect from a logistic regression model in terms
of treatment, type of preeclampsia (preeclampsia vs superimposed preeclampsia), and GA at randomization.
BPD, bronchopulmonary dysplasia; CI, confidence interval; IVH, intraventricular hemorrhage; N/A, not available; NEC, necrotizing enterocolitis; PVL, periventricular leukomalacia; RDS, respiratory
distress syndrome; rhAT, recombinant human antithrombin; ROP, retinopathy of prematurity; SGA, small for gestational age.
Paidas et al. Recombinant AT vs placebo in preterm preeclampsia. Am J Obstet Gynecol 2020.

decision to deliver was made uniquely by with singleton gestation, without preex- preeclampsia at less than 27 weeks’ gesta-
a maternal-fetal medicine specialist. The isting chronic hypertension or other tion: maternal and perinatal outcomes ac-
cording to gestational age by weeks at onset
assumptions underlying the pharmaco- medical conditions, diagnosed as having of expectant management. Am J Obstet
kinetic modeling were proven to be ac- PE between 23 and 28 weeks with AT Gynecol 2008;199:247.e1–6.
curate. This trial establishes the safety of activity levels of <80% would constitute 3. Bombrys AE, Barton JR, Habli M,
rhAT in PPE even with supra- the ideal patient population in a ran- Sibai BM. Expectant management of severe
physiological levels of rhAT. domized clinical trial to determine preeclampsia at 27(0/7) to 33(6/7) weeks’
gestation: maternal and perinatal outcomes
whether rhAT antenatal administration according to gestational age by weeks at
Conclusion will prolong gestation. n onset of expectant management. Am J Per-
The administration of rhAT to women inatol 2009;26:441–6.
with singleton pregnancy diagnosed as Acknowledgments 4. Stevens W, Shih T, Incerti D, et al. Short-term
having early-onset PE between 23 and 30 We thank the patients and team members costs of preeclampsia to the United States
Kathleen Murphy and Laura Massey. health care system. Am J Obstet Gynecol
weeks’ gestation was not associated with 2017;217:237–48.e16.
pregnancy prolongation and improve- 5. Publications Committee, Society for
ment in maternal or neonatal outcomes. References Maternal-Fetal Medicine, Sibai BM. Evaluation
rhAT is safe for pregnant women with 1. Haddad B, Deis S, Goffinet F, Paniel BJ, and management of severe preeclampsia before
PPE and their fetus or neonates. Cabrol D, Sibai BM. Maternal and perinatal 34 weeks’ gestation. Am J Obstet Gynecol
outcomes during expectant management of 239 2011;205:191–8.
Given the conflicting results regarding
severe preeclamptic women between 24 and 33 6. American College of Obstetricians and Gy-
AT supplementation in PPE, additional necologists, Task Force on Hypertension in
weeks’ gestation. Am J Obstet Gynecol
studies are indicated to determine its 2004;190:1590–5; discussion 1595e7. Pregnancy. Hypertension in pregnancy. Report
potential benefit in this vulnerable pa- 2. Bombrys AE, Barton JR, Nowacki EA, et al. of the American College of Obstetricians and
tient population. Nulliparous women Expectant management of severe Gynecologists’ Task Force on Hypertension in

NOVEMBER 2020 American Journal of Obstetrics & Gynecology 739.e9

Original Research OBSTETRICS ajog.org

Pregnancy. Obstet Gynecol 2013;122: antithrombin concentrate: results of a prospec- Memphis, TN (Dr Lam); University of Arkansas for Medical
1122–31. tive feasibility study. Semin Thromb Hemost Sciences, Little Rock, AR (Dr Magann); University of South
7. Maynard SE, Min JY, Merchan J, et al. 2003;29:645–52. Alabama Children’s and Women’s Hospital, Mobile, AL
Excess placental soluble fms-like tyrosine ki- 18. Maki M, Kobayashi T, Terao T, et al. Anti- (Dr Lewis); St. John Hospital and Medical Center, Detroit,
nase 1 (sFlt1) may contribute to endothelial thrombin therapy for severe preeclampsia: MI (Dr Dombrowski); Women and Infants Hospital of
dysfunction, hypertension, and proteinuria in results of a double-blind, randomized, placebo- Rhode Island, Providence, RI (Dr Werner); University of
preeclampsia. J Clin Invest 2003;111: controlled trial. BI51.017 Study Group. Thromb Utah Health, Salt Lake City, UT (Dr Branch); Cincinnati
649–58. Haemost 2000;84:583–90. Children’s Hospital, Cincinnati, OH (Dr Habli); Duke Uni-
8. Mol BWJ, Roberts CT, Thangaratinam S, 19. Weiner CP, Herrig JE, Pelzer GD, Heilskov J. versity Health System, NC (Dr Grotegut); The University of
Magee LA, de Groot CJM, Hofmeyr GJ. Pre- Elimination of antithrombin III concentrate in Utah, Salt Lake City, UT (Dr Silver); Ochsner Clinic
eclampsia. Lancet 2016;387:999–1011. healthy pregnant and preeclamptic women with Foundation, New Orleans, LA (Dr Longo); St. Louis Uni-
9. Robertson SA. Preventing preeclampsia by an acquired antithrombin III deficiency. Thromb versity School of Medicine, St. Louis, MO (Dr Amon);
silencing soluble Flt-1? N Engl J Med 2019;380: Res 1990;58:395–401. Columbia University Irving Medical Center, New York, NY
1080–2. 20. Paternoster DM, De Fusco D, Tambuscio B. (Dr Cleary); Norton Healthcare, Louisville, KY (Dr How);
10. Bauer KA, Rosenberg RD. Role of anti- Treatment with antithrombin III (AT III) concen- University of Mississippi Medical Center, Jackson, MS (Dr
thrombin III as a regulator of in vivo coagulation. trate in pre-eclampsia. Int J Gynaecol Obstet Novotny); Northwestern University School of Medicine,
Semin Hematol 1991;28:10–8. 2000;71:175–6. Chicago, IL (Dr Grobman); University of South Florida,
11. Harada N, Okajima K, Uchiba M, 21. Paternoster DM, Fantinato S, Manganelli F, Tampa, FL (Dr Whiteman); University of California Irvine,
Kushimoto S, Isobe H. Antithrombin reduces Milani M, Nicolini U, Girolami A. Efficacy of AT in Irvine, CA (Dr Wing); University of Oklahoma Health Sci-
ischemia/reperfusion-induced liver injury in rats pre-eclampsia: a case-control prospective trial. ences Center, Oklahoma City, OK (Dr Scifres); and Uni-
by activation of cyclooxygenase-1. Thromb Thromb Haemost 2004;91:283–9. versity of Texas Health Sciences Center at Houston,
Haemost 2004;92:550–8. 22. Weiner CP, Kwaan HC, Xu C, Paul M, Houston, TX (Dr Sibai).
1
12. Komura H, Uchiba M, Mizuochi Y, et al. Burmeister L, Hauck W. Antithrombin III activity Deceased.
Antithrombin inhibits lipopolysaccharide- in women with hypertension during pregnancy. Received Feb. 24, 2020; revised May 20, 2020;
induced tumor necrosis factor-alpha produc- Obstet Gynecol 1985;65:301–6. accepted Aug. 6, 2020.
tion by monocytes in vitro through inhibition of 23. Stoll BJ, Hansen NI, Bell EF, et al. Trends in This study was sponsored by rEVO Biologics and LFB
Egr-1 expression. J Thromb Haemost 2008;6: care practices, morbidity, and mortality of USA, Inc, which are subsidiaries of LFB SA, who
499–507. extremely preterm neonates, 1993e2012. contributed to the study design and data collection and
13. Ornaghi S, Barnhart KT, Frieling J, JAMA 2015;314:1039–51. analysis but played no role in the interpretation of the data
Streisand J, Paidas MJ. Clinical syndromes 24. D’Angelo A, Valsecchi L; ATIII-Early Pre- or the writing of the report. M.J.P. and B.M.S. report
associated with acquired antithrombin defi- eclampsia Study Group (ATIII-EPAS). High dose grants from LFB SA during the conduct of the study.
ciency via microvascular leakage and the related antithrombin supplementation in early pre- Outside of the submitted work, M.J.P. reports grants from
risk of thrombosis. Thromb Res 2014;133: eclampsia: a randomized, double blind, placebo- LFB USA, Inc; rEVO Biologics; BioIncept, LLC; Progenity;
972–84. controlled study. Thromb Res 2016;140:7–13. and GestVision. B.M.S. reports grants from LFB USA, Inc,
14. Ornaghi S, Paidas MJ. Upcoming drugs for 25. Weenink GH, Treffers PE, Vijn P, Smoren- during the conduct of the study and from Progenity
the treatment of preeclampsia in pregnant berg-Schoorl ME, Ten Cate JW. Antithrombin III outside of the submitted work. G.A.M. and G.K.L. report
women. Expert Rev Clin Pharmacol 2014;7: levels in preeclampsia correlate with maternal grants from rEVO Biologics during the conduct of the
599–603. and fetal morbidity. Am J Obstet Gynecol study. J.B.S., E.F.M., Kathleen Murphy, and Laura
15. Shinyama H, Yamanaga K, Akira T, et al. 1984;148:1092–7. Massey report funding from LFB USA, Inc, and rEVO Bi-
Antithrombin III prevents blood pressure 26. Fenton TR, Kim JH. A systematic review and ologics during the conduct of the study and outside of the
elevation and proteinuria induced by high salt meta-analysis to revise the Fenton growth submitted work. The other authors report no conflict of
intake in pregnant stroke-prone spontane- chart for preterm infants. BMC Pediatr 2013; interest.
ously hypertensive rats. Biol Pharm Bull 13:59. Author disclosure forms for M.J.P. and B.M.S. are
1996;19:819–23. attached. All other author forms are complete and
16. Shinyama H, Akira T, Uchida T, Hirahara K, Author and article information available for submission to the American Journal of Ob-
Watanabe M, Kagitani Y. Antithrombin III pre- From the Yale University School of Medicine, New Haven, stetrics and Gynecology.
vents renal dysfunction and hypertension CT (Drs Paidas and Ehrenkranz); University of Alabama at Clinical Trial Registration: ClinicalTrials.gov,
induced by enhanced intravascular coagulation Birmingham School of Medicine, Birmingham, AL NCT02059135.
in pregnant rats: pharmacological confirmation (Dr Tita); Washington University School of Medicine, St. The results of the Pharmacokinetics, Safety and Effi-
of the benefits of treatment with antithrombin III Louis, MO (Dr Macones); University of Texas Medical cacy of Recombinant Antithrombin Versus Placebo in
in preeclampsia. J Cardiovasc Pharmacol Branch, Galveston, TX (Dr Saade); Brown University Preterm Preeclampsia study have not been published or
1996;27:702–11. School of Medicine, Providence, RI (Dr Triche); rEVO Bi- presented at any scientific meetings.
17. Kobayashi T, Terao T, Ikenoue T, et al. ologics and LFB USA, Inc, Framingham, MA (Dr Strei- Corresponding author: Michael J. Paidas, MD.
Treatment of severe preeclampsia with sand); University of Tennessee College of Medicine, [email protected]

739.e10 American Journal of Obstetrics & Gynecology NOVEMBER 2020

ajog.org OBSTETRICS Original Research

SUPPLEMENTAL TABLE 1
A priori fetal or neonatal outcomes of specific interest
Perinatal complication Definition
Fetal death Includes stillbirth
Neonatal death Death from day 0 (not stillbirth) to the last study
assessment
All death Fetal and neonatal death
SGA Below the tenth percentile weight by gender from
Fenton et al26
RDS Clinical features of RDS and oxygen or respiratory
support for 6 h of the first 24 h of life
BPD Oxygen requirement at 36 wk postmenstrual age
ROP Severe (stage 3) retinopathy of prematurity
NEC Bell’s stage 2
IVH Cranial ultrasound, CT, or MRI evidence for IVH grade
3
Cystic PVL Cranial ultrasound, CT, or MRI evidence for cystic PVL
Early sepsis Early sepsis (positive blood culture not regarded as
contaminant), within 72 h of delivery
Late sepsis Late sepsis (positive blood culture not regarded as
contaminant) more than 72 h after delivery
Meningitis Positive culture of cerebrospinal fluid
BPD, bronchopulmonary dysplasia; CT, computed tomography; IVH, intraventricular hemorrhage; MRI, magnetic resonance imaging; NEC, necrotizing enterocolitis; PVL, periventricular leukomalacia;
RDS, respiratory distress syndrome; ROP, retinopathy of prematurity; SGA, small for gestational age.
Paidas et al. Recombinant AT vs placebo in preterm preeclampsia. Am J Obstet Gynecol 2020.

NOVEMBER 2020 American Journal of Obstetrics & Gynecology 739.e11

Original Research OBSTETRICS ajog.org

SUPPLEMENTAL TABLE 2
A priori maternal outcomes of specific interest
Maternal complication Definition
Maternal death Documented death of any cause during the trial up to 28 d after delivery
Eclamptic seizure Presence of new-onset grand mal seizures in a woman with preeclampsia,
without evidence for other etiology (eg, bradycardia, tachycardia, hypotension,
expected or unexpected drug effect)
Myocardial infarction Detection of a rise and/or fall of cardiac biomarker values (cardiac troponin) with
at least 1 value above the 99th percentile upper reference limit and with at least
one of the following:
 Symptoms of ischemia
 New or presumed new significant ST-segment T-wave (ST-T) changes or new
left bundle branch block
 Development of pathologic Q waves in the ECG
 Imaging evidence of new loss of viable myocardium or new regional wall
motion abnormality
 Identification of an intracoronary thrombus by angiography or autopsy
CVA Acute onset of a focal neurologic deficit, with CT or MRI evidence for hemorrhagic
CVA or cerebral infarction (brain or retinal cell death caused by prolonged
ischemia)
Transient ischemic A brief episode (usually less than 1 or 2 h) of neurologic dysfunction caused by
attack focal brain or retinal ischemia without evidence of acute infarction (as
determined by CT or MRI)
Progressive renal  Serum creatinine of >1.1 mg/dL
insufficiency
Thrombocytopenia  Platelet count of <100,000/mL
(without HELLP)
HELLP A combination of:
 AST of 70 IU/L
 Platelet count of <100,000/mL
 Evidence of hemolysis on blood film plus either LDH of 600 IU/mL or total
bilirubin of 1.2 mg/dL
DIC Overt DIC is defined as a score of 5 derived from scoring the following
parameters:
 Platelet count of 100/mm3 (0), 50/mm3 but <100/mm3 (1), <50/mm3 (2)
 Elevated fibrin-related marker such as soluble fibrin monomer or fibrin
degradation products with no increase (0), moderate increase (2), strong
increase (3)
 Prolonged PT of 3 s (0), >3 but <6 s (1), and 6 s (2)
 Fibrinogen level of 1.0 g/L (0) and <1.0 g/L (3)
Pulmonary edema Clinical symptoms of pulmonary edema, confirmed by chest x-ray examination
Placental abruption Vaginal bleeding and uterine pain or tenderness accompanied by documented
fetal distress or uterine hypertonicity
Severe intra- and Estimated blood loss of 1500 mL within 24 h after delivery
postpartum hemorrhage
VTE including DVT and Clinically symptomatic VTE (eg, breathlessness, hemoptysis, leg pain, and
pulmonary embolism swelling) with confirmation by appropriate diagnostic tools (eg, venography,
Doppler ultrasound, ventilation or perfusion scan, MRI)
AST, aspartate aminotransferase; CT, computed tomography; CVA, cerebrovascular accident; DIC, disseminated intravascular coagulation; DVT, deep vein thrombosis; ECG, electrocardiogram;
HELLP, hemolysis, elevated liver enzymes, and low platelet; LDH, lactate dehydrogenase; MRI, magnetic resonance imaging; PT, prothrombin time; VTE, venous thromboembolism.
Paidas et al. Recombinant AT vs placebo in preterm preeclampsia. Am J Obstet Gynecol 2020.

739.e12 American Journal of Obstetrics & Gynecology NOVEMBER 2020

ajog.org OBSTETRICS Original Research

SUPPLEMENTAL BOX
Exclusion criteria

Criteria requiring immediate delivery of the fetus before randomization:

 Refractory hypertension despite maximal medical intervention of systolic blood pressure of 160 mm Hg or diastolic blood pressure of
110 mm Hg
 Thrombocytopenia (platelets <100,000/mL) with or without HELLP syndrome, defined as:
o AST of 70 units/L
o Platelet count of <100,000/mL
o Evidence of hemolysis on blood film plus either LDH of 600 IU/mL or total bilirubin of 1.2 mg/dL; oliguria (500 mL/24 h)
o Evidence of progressive renal insufficiency
 Serum creatinine concentration of >1.1 mg/dL
 Persistent visual disturbances
 Placental abruption
 Pulmonary edema
 Nonreassuring fetal heart rate tracing
 Intractable headache unrelieved with analgesia
 Intractable right upper quadrant abdominal pain or vomiting
 If umbilical Doppler ultrasound had been performed, the presence of an abnormal umbilical artery Doppler as defined by persistent,
absent, or reverse end-diastolic flow
 Biophysical score of 4/10 on 2 occasions
 Oligohydramnios (deepest vertical pocket of <22 cm on ultrasound)
 Other maternal or fetal conditions that would preclude expectant management
Other exclusion criteria:
 Known lethal or major fetal anomaly
 Recent (within 12 mo) history of maternal alcoholism or drug dependence
 Diagnosis of epilepsy
 Need for chronic therapy with NSAIDs including selective Cox-2 inhibitors, or unwilling to abstain from the use of NSAIDs during the
study treatment period (low-dose aspirin of 81 mg/d or less was allowed)
 Received within 72 h or had requirement for heparin
 Low-molecular-weight heparin such as enoxaparin or dalteparin
 Fondaparinux
 Antiplatelet agents such as clopidogrel, prasugrel, or high-dose aspirin (>81 mg/d)
 DTIs such as dabigatran
 Preexisting renal disease, documented prepregnancy or in pregnancy before 20 wk gestation (before the diagnosis of preeclampsia), or
24 h urine of 0.3 g/24 h, documented in pregnancy, before 20 wk gestation, or 2þ dipstick or 0.3 PCR, documented in pregnancy at
the last available test before 20 wk gestation
o In the case of conflicting results between dipstick, PCR, and timed urine collection to work up an episode of proteinuria, the timed
urine collection result would supersede other results.
 Multifetal pregnancy
 History of antiphospholipid antibody syndrome
 Known hypersensitivity to goat and goat milk proteins
 Participation in another interventional clinical trial of an investigational, unapproved therapy (drug, biologic, device) within 30 d of
consent
AST, aspartate aminotransferase; Cox, cyclooxygenase; DTI, direct thrombin inhibitor; HELLP, hemolysis, elevated liver enzymes, and low platelet; LDH, lactate dehydrogenase;
NSAID, nonsteroidal antiinflammatory drug; PCR, polymerase chain reaction. Paidas et al. Recombinant AT vs placebo in preterm preeclampsia. Am J Obstet Gynecol 2020.

NOVEMBER 2020 American Journal of Obstetrics & Gynecology 739.e13