Ultrasound Obstet Gynecol 2015; 46: 414–418 Published online 31 August 2015 in Wiley Online Library (wileyonlinelibrary.com). DOI: 10.1002/uog.14889
Early prediction and aspirin for prevention of pre-eclampsia (EPAPP) study: a randomized controlled trial A. O. ODIBO*, K. R. GOETZINGER†, L. ODIBO* and M. G. TUULI† *Division of Maternal Fetal Medicine, Department of Obstetrics and Gynecology, University of South Florida, Morsani College of Medicine, Tampa, FL, USA; †Division of Maternal Fetal Medicine, Department of Obstetrics and Gynecology, Washington University School of Medicine, St Louis, MO, USA
K E Y W O R D S: acetylsalicylic acid; ASA; aspirin; high-risk; pre-eclampsia
ABSTRACT
INTRODUCTION
Objective To estimate the effect of early administration of aspirin on the prevention of pre-eclampsia in high-risk women.
Pre-eclampsia complicates up to 8% of pregnancies and is a major contributor to maternal mortality and morbidity. The only effective treatment is delivery, which leads to significant neonatal morbidity and mortality if carried out preterm, especially when the disease occurs early in pregnancy. Several recent studies have shown improved predictive ability for pre-eclampsia of models that combine risk factors, biophysical measures and maternal serum analytes1,2 . However, the utility of these prediction tests in themselves is limited by the absence of studies demonstrating effective interventions to reduce the risk of pre-eclampsia in women identified in this manner to be at high risk for pre-eclampsia. Low-dose aspirin has the potential to prevent preeclampsia through the inhibition of thromboxanemediated vasoconstriction and immune-modulation. In a recent systematic review, when studies were stratified based on gestational age at which aspirin was initiated, a risk reduction of over 50% was noted when aspirin was initiated prior to 16 weeks while no benefit was seen when it was initiated after 16 weeks3 . We conducted a randomized controlled trial to estimate the efficacy of low-dose aspirin for preventing pre-eclampsia in women identified as high risk. We hypothesized that the risk of pre-eclampsia in women identified by a first-trimester multiparameter predictive model to be at high risk would be significantly reduced by initiating low-dose aspirin early in pregnancy.
Methods This was planned as a randomized, double-blind, placebo-controlled trial of aspirin for women with risk factors for pre-eclampsia. Participants were randomized to start either aspirin (81 mg/day) or placebo at 11 + 0 to 13 + 6 weeks of gestation. The primary outcome was pre-eclampsia and secondary outcomes included gestational hypertension and small-for-gestational age at birth. Results The trial suffered from slow recruitment, leading to a protocol change to broaden the inclusion criteria (from a minimum score of multiple risk factors to at least one risk factor for pre-eclampsia). The trial was then terminated prematurely due to continuing slow recruitment and a lack of equipoise given a change in national guidelines to administer aspirin to high-risk women. From the 53 women who were randomized, 30 were included in the final analysis. There was no evidence that the primary outcome of pre-eclampsia was prevented by low-dose aspirin (relative risk (RR) 0.88, 95% CI 0.21–3.66). Gestational hypertension was seen in two women, both in the aspirin group. There was no evidence that the occurrence of small-for-gestational age was reduced by aspirin (RR 0.88, 95% CI 0.06–12.72). Conclusions Although this study was underpowered to show effectiveness of aspirin compared to placebo due to the premature termination and difficulties encountered, it highlights practical issues to inform future studies. Copyright © 2015 ISUOG. Published by John Wiley & Sons Ltd.
METHODS This was a randomized control trial to estimate the efficacy of low-dose aspirin in preventing pre-eclampsia in women
Correspondence to: Dr A. Odibo, Division of Maternal Fetal Medicine, Department of Obstetrics and Gynecology, University of South Florida, Morsani College of Medicine, STC suite 6000, 2 Tampa General Circle, Tampa, FL 33606, USA (e-mail: [email protected]) Accepted: 21 April 2015
Copyright © 2015 ISUOG. Published by John Wiley & Sons Ltd.
ORIGINAL PAPER
Early prediction and aspirin for prevention of pre-eclampsia Table 1 Risk factor scoring system for first-trimester prediction of pre-eclampsia2 Risk factor Chronic hypertension History of prior pre-eclampsia Diabetes mellitus Obesity (BMI > 30) Bilateral uterine artery notches Low PAPP-A (< 0.52 MoM)
Score 4 3 2 2 1 1
BMI, body mass index; MoM, multiples of the median; PAPP-A, pregnancy-associated plasma protein-A.
identified in the first trimester to be at high risk. The study was approved by Washington University School of Medicine Institutional Review Board, and registered with clinicaltrials.gov (NCT01547390). The inclusion criteria were singleton pregnancy undergoing ultrasound examination at 11 + 0 to 13 + 6 weeks and deemed to be at high risk for pre-eclampsia by the criteria listed in Table 12 . We excluded pregnancies with multiple gestation, fetal aneuploidy, major fetal structural anomaly and bleeding disorder, and women with allergy to aspirin or already on aspirin or heparin. All eligible women who presented for first-trimester ultrasound examination were approached for consent to participate in the study. Consenting women had the following assessments between 9 + 0 and 13 + 6 weeks: history for risk factors, uterine artery Doppler, and measurement of pregnancy-associated plasma protein-A (PAPP-A). A total score of at least 6 based on a risk factor scoring system in first trimester (Table 1) was initially needed to be eligible for randomization. However, due to slow recruitment 2 years into the study (with only 23 women enrolled) and following the recommendation from the data safety monitoring committee, the Institutional Review Board was asked for permission to revise the inclusion criteria to the presence of any of the risk factors (as listed in Table 1). Women meeting the inclusion criteria were randomized to receive daily 81 mg aspirin or a placebo pill from 11 + 0 to 13 + 6 weeks until 37 weeks or delivery, whichever occurred first. Randomization was 1:1 using a computerized randomization program by the pharmacist who was otherwise uninvolved in the study. The study team and the patient were blinded to the intervention the patient was receiving. The placebo/aspirin was dispensed by pharmacy at randomization and then again at the 18–24-week ultrasound visit and at the 28–32-week ultrasound visit. The women were encouraged to take the study medication with their regular prenatal vitamins. No specific instructions were given as to the timing of ingestion of the medication. Compliance was assessed by having the women complete a pill diary, which they were asked to return along with any unused pills, which were counted and returned to pharmacy. Demographic, previous medical and obstetric history and current obstetric information were collected from patient questionnaires and chart abstraction.
Copyright © 2015 ISUOG. Published by John Wiley & Sons Ltd.
415
The primary outcome was pre-eclampsia diagnosed per ACOG criteria at the onset of the study4 . Details of secondary outcomes were also collected. These included small-for-gestational-age neonate (birth weight < 10th percentile for gestational age on the Alexander growth standard5 ); early pre-eclampsia (delivery < 34 weeks); severe pre-eclampsia (blood pressure > 160/110 or symptoms including persistent headaches, visual disturbance, or evidence of abnormal renal failure, abnormal liver enzymes or thrombocytopenia); gestational hypertension (elevated blood pressure occurring after 20 weeks’ gestation with no evidence of pre-eclampsia); preterm birth (delivery < 37 weeks); stillbirth; antepartum hemorrhage (hemorrhage after 20 weeks’ gestation); neonatal death; neonatal intensive care unit admission; and miscarriage.
Statistical analysis Statistical analyses for the randomized controlled trial were originally planned on the ‘intention-to-treat’ principle but, based on the high withdrawal rate, the analysis was performed on an ‘as-treated’ principle. The primary outcome and other categorical variables were compared across groups using the chi-square or Fisher’s exact test as appropriate. We calculated the 95% confidence intervals around relative risk (RR) of the primary outcome in the intervention and placebo groups. We calculated the required sample size for the trial based on estimates using our scoring system. Using a receiver–operating characteristics (ROC) curve we determined the overall accuracy of our risk factor scoring system to be 78% (95% CI, 70–86%) and the optimal cut-off for predicting pre-eclampsia to be a total score of 62 . Based on the anticipated baseline rate of 36% in the placebo group, a total of 186 women would be needed for 80% power to detect a 50% reduction in the risk of pre-eclampsia (two-tailed α of 0.048; incorporating one interim analysis and using the O’Brien–Fleming stopping rule6 ). To account for an estimated 15% dropout rate, we initially estimated that 220 screen-positive women would need to be randomized. Following the revised inclusion criteria, a revised sample size of 684 women needed to be randomized to detect the same 50% reduction of pre-eclampsia with 80% power and 15% anticipated dropout rate; this was based on the revised accuracy of the model of 65%. However, the trial was terminated prematurely due to slow recruitment and lack of equipoise given a change to national guidelines to administer aspirin to high-risk women. All statistical analyses were performed using STATA software version 13.1 (STATA Corp., College Station, TX, USA).
RESULTS Between April 2012 and March 2014, a total of 1470 women were screened. Four hundred and forty-one were screened with the original algorithm and 1029 with the
Ultrasound Obstet Gynecol 2015; 46: 414–418.
416
modified one. Of these, 138 were screen positive (41 when using the original algorithm and 97 with the modified one) and 53 were eligible and consented to participate in the study and were randomized. Common reasons for patients declining to participate or being excluded from the study included the following: already being on aspirin, allergy to aspirin, already on anticoagulants for other indications such as a thrombophilia, concerns about safety of aspirin, and wanting to be guaranteed aspirin rather than a placebo. Some women with only one risk factor such as morbid obesity had the perception that their risk for pre-eclampsia was not high and therefore declined to participate. Ten women in the aspirin group and 13 in the placebo group dropped out of the study after randomization. Seven withdrawals in the aspirin group were by patient request, one pregnancy was terminated and two were lost to follow-up (Figure 1). The baseline characteristics of the 30 women who completed the study are shown in Table 2. Baseline characteristics were similar in the two groups except for a higher proportion of African Americans in the placebo group. The primary outcome of the study, pre-eclampsia, was seen in six of the 30 women (20%) with complete follow-up, three in the aspirin group and three in the placebo group (RR 0.88, 95% CI 0.21–3.66). All cases with pre-eclampsia occurred and were delivered before 34 weeks’ gestation. Gestational hypertension was seen in two women, both in the aspirin group. Small-for-gestational age was seen in two of the randomized women (6.7%), one in the aspirin and one in the placebo group (RR 0.88, 95% CI 0.06–12.72). With the exception of the three women reporting rash/hives, no other adverse effect was reported in either of the groups. The mean risk score was significantly higher (9.2 ± 1.2) in the six women who developed early pre-eclampsia vs 4.5 ± 0.4 in the other 24 women enrolled in the study with complete follow-up (P = 0.0001).
DISCUSSION This randomized trial, which was terminated prematurely, was affected by several practical issues that raise questions regarding the feasibility of conducting future studies with similar objectives. The findings, which should be interpreted with caution given that the study was underpowered to test our original hypothesis, are, however, consistent with the majority of trials using low-dose aspirin for preventing pre-eclampsia prior to 16 weeks7 – 11 . The inconclusive results reported by most trials aiming to use aspirin to prevent pre-eclampsia are mostly due to the sample size of these studies, as has been demonstrated by subsequent meta-analyses12 – 16 . Despite the conclusions of these meta-analyses as to the modest benefit for preventing all forms of pre-eclampsia, international guidelines have been published recommending the use of low-dose aspirin17 – 19 . The main limitation of our study is the small sample size, similar to other low-dose aspirin trials in early pregnancy. Our initial scoring system that was validated
Copyright © 2015 ISUOG. Published by John Wiley & Sons Ltd.
Odibo et al. in our population2 appeared too strict, resulting in very few high-risk women being assigned a large enough score to be randomized into the trial. The model was designed to be highly accurate, balancing the trade-off between sensitivity and specificity. Our model had a sensitivity/specificity that depended on the number of risk factors present and their weighted score. As our analysis was not based on sensitivity for a fixed false-positive rate, but on the accuracy of the overall model, it is difficult to compare the effectiveness of the model to other published series. Following the change in the inclusion criteria, the recommendations of the United States Preventive Services Task Force that women at risk for pre-eclampsia should receive low-dose aspirin early in pregnancy was published19 . With that recommendation, the investigation team met and determined that we had no equipoise to continue the trial. Given the small sample size, the primary goal of randomization, which is to eliminate potential biases, may not have been achieved. The United States Preventive Services Task Force recommendation was based on a systematic review including most of the data discussed above. It brings up the controversy of whether clinical decisions should be based on the results of individual trials or on meta-analysis20 . In addition, recent reviews have suggested using higher doses of aspirin and administering these at night time. It is not certain if taking these steps would have impacted the outcome of our trial21 – 23 . We noted a significant dropout rate in the trial. Of the 53 women who were randomized, 23 (43.3%) dropped out for a number of reasons including the development of rash or hives or itching, ‘forgetting to take pills’, being on ‘too many pills’ and termination of pregnancy. The high dropout rate may be a reflection of our study population, which includes a high proportion of inner city low income or indigent patients with social issues that can reduce attendance for prenatal care. To improve study retention, our study coordinators called enrolled patients on a weekly basis to encourage continuation of study medication and remind them of prenatal visit appointments. These efforts had little impact on the overall success of the study. While dropouts were balanced between the two groups and would not be expected to affect the results, given the overall small sample size it resulted in our performing an ‘as-treated’ rather than an ‘intention-to-treat’ analysis as is the usual case with trials. The high withdrawal rate also raises a question of compliance among women prescribed aspirin for prevention of pre-eclampsia. Interestingly, all cases of pre-eclampsia in our study occurred before 34 weeks. The 20% incidence is much higher than would be expected for an unselected population. We were unable to estimate the overall performance of our screening model. This information would have been helpful if we had been able to collect it. However, because we are ethically not allowed to follow those who screened negative (and did not consent to participate) or withdrew from the study, it is impossible
Ultrasound Obstet Gynecol 2015; 46: 414–418.
Early prediction and aspirin for prevention of pre-eclampsia
417
Multiparameter first-trimester pre-eclampsia screening (n = 1470)
Screen negative (n = 1332) Screen positive (n = 138)
79 screen positive declined randomization, 16 found ineligible*
Randomized (n = 53)
Dropouts (n = 10) (7 on patient request, 1 termination, 2 LTF)
Treatment group (n = 26)
Placebo group (n = 27)
Treatment group with complete follow-up data (n = 16)
Placebo group with complete follow-up data (n = 14)
Dropouts (n = 13) (10 on patient request, 3 rash/hives)
Primary outcome: pre-eclampsia (n = 6)
Figure 1 Flowchart of study procedures and enrollment. *Ten women met both criteria. LTF, lost to follow-up. Table 2 Demographic data and inclusion criteria for the Early Prediction and Aspirin for Prevention of Pre-eclampsia (EPAPP) study
Maternal age (years) Body mass index (kg/m2 ) Nulliparous Race/ethnicity Black White Other Risk score* GA at randomization (weeks) Mean UtA-PI Chronic hypertension Pregestational diabetes
Placebo (n = 14)
Aspirin (n = 16)
P
31.6 ± 6.1 36.6 ± 6.9 3 (21.4)
30.0 ± 5.0 37.4 ± 8.9 5 (31.2)
0.45 0.78 0.54 0.013
11 (78.6) 3 (21.4) 0 (0) 5.9 ± 3.2 12.1 ± 1.0 1.68 ± 0.4 10 (71.4) 2 (14.3)
4 (25) 1 (6.2) 11 (68.8) 5.0 ± 2.9 11.7 ± 1.3 1.57 ± 0.8 6 (37.5) 4 (25.0)
0.40 0.35 0.69 0.06 0.46
Data are given as mean ± SD or n (%). *Risk score as defined by Goetzinger et al.2 . GA, gestational age; UtA-PI, uterine artery pulsatility index.
to calculate the overall performance of the model. The mean screening score of the women who developed early pre-eclampsia was significantly higher than that of the other women enrolled in the study. It is possible that these women’s risk was so high that low-dose aspirin could not
Copyright © 2015 ISUOG. Published by John Wiley & Sons Ltd.
prevent them from developing pre-eclampsia. The above limitations should inform investigators contemplating future studies for pre-eclampsia prevention. Based on current guidelines, which may be premature, we wonder if future trials focused on the primary question of
Ultrasound Obstet Gynecol 2015; 46: 414–418.
418
pre-eclampsia prevention with aspirin are ethical or practical given the problems we encountered with this trial. The failure to detect a difference in our study should be interpreted with caution due to the difficulties encountered with the trial and the study being underpowered to show a difference. Future trials should focus on the optimal dose of aspirin or combination of antiplatelet therapy to potentially optimize the prevention of pre-eclampsia.
REFERENCES 1. Poon LC, Kametas NA, Maiz N, Akolekar R, Nicolaides KH. First-trimester prediction of hypertensive disorders in pregnancy. Hypertension 2009; 53: 812–818. 2. Goetzinger KR, Tuuli MG, Cahill AG, Macones GA, Odibo AO. Development and validation of a risk factor scoring system for first-trimester prediction of preeclampsia. Am J Perinatol 2014; 31: 1049–1056. 3. Bujold E, Roberge S, Lacasse Y, Bureau M, Audibert F, Marcoux S, Forest JC, Gigu`ere Y. Prevention of preeclampsia and intrauterine growth restriction with aspirin started in early pregnancy: a meta-analysis. Obstet Gynecol 2010; 116: 402–414. 4. ACOG Practice Bulletin. Diagnosis and management of preeclampsia and eclampsia. Number 33, January 2002. Obstet Gynecol 2002; 99: 159–167. 5. Alexander GR, Himes JH, Kaufman RB, Mor J, Kogan M. A United States national reference for fetal growth. Obstet Gynecol 1996; 87: 163–168. 6. Pocock SJ. When to stop a clinical trial. BMJ 1992; 305: 235–240. 7. Ebrashy A, Ibrahim M, Marzook A, Yousef D. Usefulness of aspirin therapy in high-risk pregnant women with abnormal uterine artery Doppler ultrasound at 14–16 weeks pregnancy: randomized controlled clinical trial. Croat Med J 2005; 46: 826–831. 8. Bakhti A, Vaiman D. Prevention of gravidic endothelial hypertension by aspirin treatment administered from the 8th week of gestation. Hypertens Res 2011; 34: 1116–1120. 9. Villa P, Kajantie E, Raikkonen K, Pesonen AK, Hamalainen E, Vainio M, Taipale P, Laivuori H; PREDO Study group. Aspirin in the prevention of pre-eclampsia in high-risk women: a randomised placebo-controlled PREDO Trial and a meta-analysis of randomised trials. BJOG 2013; 120: 64–74. 10. Vainio M, Kujansuu E, Iso-Mustajarvi M, Maenpaa J. Low dose acetylsalicylic acid in prevention of pregnancy-induced hypertension and intrauterine growth retardation in women with bilateral uterine artery notches. BJOG 2002; 109: 161–167.
Copyright © 2015 ISUOG. Published by John Wiley & Sons Ltd.
Odibo et al. 11. August P, Helseth G, Edersheim T, Hutson J, Druzin M. Sustained release, low-dose aspirin ameliorates but does not prevent preeclampsia (PE) in a high risk population. Proceedings of the 9th International Congress, International Society for the Study of Hypertension, 1994; 72. 12. Imperiale TF, Petrulis AS. A meta-analysis of low-dose aspirin for the prevention of pregnancy-induced hypertensive disease. JAMA 1991; 266: 260–264. 13. Askie LM, Duley L, Henderson-Smart DJ, Stewart LA. Antiplatelet agents for prevention of pre-eclampsia: a metaanalysis of individual patient data. Lancet 2007; 369: 1791–1798. 14. Bujold E, Morency AM, Roberge S, Lacasse Y, Forest JC, Gigu`ere Y. Acetylsalicylic acid for the prevention of preeclampsia and intra-uterine growth restriction in women with abnormal uterine artery Doppler: a systematic review and meta-analysis. J Obstet Gynaecol Can 2009; 31: 818–826. 15. Roberge S, Villa P, Nicolaides K, Gigu`ere Y, Vainio M, Bakthi A, Ebrashy A, Bujold E. Early administration of low-dose aspirin for the prevention of preterm and term preeclampsia: a systematic review and meta-analysis. Fetal Diagn Ther 2012; 31: 141–146. 16. Roberge S, Nicolaides KH, Demers S, Villa P, Bujold E. Prevention of perinatal death and adverse perinatal outcome using low-dose aspirin: a meta-analysis. Ultrasound Obstet Gynecol 2013; 41: 491–499. 17. NICE clinical guideline No. 107. Hypertension in pregnancy. Management of hypertensive disorders in pregnancy. Issued 2010, updated January 2011. http://www.nice.org. uk/nicemedia/live/13098/50418/50418.pdf [Accessed 20 March 2013]. 18. WHO Recommendations for Prevention and Treatment of Preeclampsia and Eclampsia. World Health Organization, 2011. http://whqlibdoc. who.int/publications/2011/9789241548335_eng.pdf [Accessed 20 March 2013]. 19. Henderson JT, Whitlock EP, O’Conner E, Senger CA, Thompson JH, Rowland MG. Low-Dose Aspirin for the Prevention of Morbidity and Mortality From Preeclampsia: A Systematic Evidence Review for the U.S. Preventive Services Task Force. Rockville (MD, USA): Agency for Healthcare Research and Quality (US); 2014. 20. Scifres CM, Iams JD, Klebanoff M, Macones GA. Meta-analysis vs large clinical trials: which should guide our management? Am J Obstet Gynecol 2009; 200: 484.e1–5. 21. Caron N, Rivard GE, Michon N, Morin F, Pilon D, Moutquin JM, Rey E. Low-dose ASA response using the PFA-100 in women with high-risk pregnancy. J Obstet Gynaecol Can 2009; 31: 1022–1027. 22. Hermida RC, Ayala DE, Fernandez JR, Mojon A, Alonso I, Silva I, Ucieda R, Codesido J, Iglesias M. Administration time-dependent effects of aspirin in women at differing risk for preeclampsia. Hypertension 1999; 34: 1016–1023. 23. Park F, Russo K, Williams P, Pelosi M, Puddephatt R, Walter M, Leung C, Saaid R, Rawashdeh H, Ogle R, Hyett J. Prediction and prevention of early-onset pre-eclampsia: impact of aspirin after first-trimester screening. Ultrasound Obstet Gynecol 2015; 46: 419–423.
Ultrasound Obstet Gynecol 2015; 46: 414–418.
Early prediction and aspirin for prevention of pre-eclampsia (EPAPP) study: a randomized controlled trial A. O. ODIBO*, K. R. GOETZINGER†, L. ODIBO* and M. G. TUULI† *Division of Maternal Fetal Medicine, Department of Obstetrics and Gynecology, University of South Florida, Morsani College of Medicine, Tampa, FL, USA; †Division of Maternal Fetal Medicine, Department of Obstetrics and Gynecology, Washington University School of Medicine, St Louis, MO, USA
K E Y W O R D S: acetylsalicylic acid; ASA; aspirin; high-risk; pre-eclampsia
ABSTRACT
INTRODUCTION
Objective To estimate the effect of early administration of aspirin on the prevention of pre-eclampsia in high-risk women.
Pre-eclampsia complicates up to 8% of pregnancies and is a major contributor to maternal mortality and morbidity. The only effective treatment is delivery, which leads to significant neonatal morbidity and mortality if carried out preterm, especially when the disease occurs early in pregnancy. Several recent studies have shown improved predictive ability for pre-eclampsia of models that combine risk factors, biophysical measures and maternal serum analytes1,2 . However, the utility of these prediction tests in themselves is limited by the absence of studies demonstrating effective interventions to reduce the risk of pre-eclampsia in women identified in this manner to be at high risk for pre-eclampsia. Low-dose aspirin has the potential to prevent preeclampsia through the inhibition of thromboxanemediated vasoconstriction and immune-modulation. In a recent systematic review, when studies were stratified based on gestational age at which aspirin was initiated, a risk reduction of over 50% was noted when aspirin was initiated prior to 16 weeks while no benefit was seen when it was initiated after 16 weeks3 . We conducted a randomized controlled trial to estimate the efficacy of low-dose aspirin for preventing pre-eclampsia in women identified as high risk. We hypothesized that the risk of pre-eclampsia in women identified by a first-trimester multiparameter predictive model to be at high risk would be significantly reduced by initiating low-dose aspirin early in pregnancy.
Methods This was planned as a randomized, double-blind, placebo-controlled trial of aspirin for women with risk factors for pre-eclampsia. Participants were randomized to start either aspirin (81 mg/day) or placebo at 11 + 0 to 13 + 6 weeks of gestation. The primary outcome was pre-eclampsia and secondary outcomes included gestational hypertension and small-for-gestational age at birth. Results The trial suffered from slow recruitment, leading to a protocol change to broaden the inclusion criteria (from a minimum score of multiple risk factors to at least one risk factor for pre-eclampsia). The trial was then terminated prematurely due to continuing slow recruitment and a lack of equipoise given a change in national guidelines to administer aspirin to high-risk women. From the 53 women who were randomized, 30 were included in the final analysis. There was no evidence that the primary outcome of pre-eclampsia was prevented by low-dose aspirin (relative risk (RR) 0.88, 95% CI 0.21–3.66). Gestational hypertension was seen in two women, both in the aspirin group. There was no evidence that the occurrence of small-for-gestational age was reduced by aspirin (RR 0.88, 95% CI 0.06–12.72). Conclusions Although this study was underpowered to show effectiveness of aspirin compared to placebo due to the premature termination and difficulties encountered, it highlights practical issues to inform future studies. Copyright © 2015 ISUOG. Published by John Wiley & Sons Ltd.
METHODS This was a randomized control trial to estimate the efficacy of low-dose aspirin in preventing pre-eclampsia in women
Correspondence to: Dr A. Odibo, Division of Maternal Fetal Medicine, Department of Obstetrics and Gynecology, University of South Florida, Morsani College of Medicine, STC suite 6000, 2 Tampa General Circle, Tampa, FL 33606, USA (e-mail: [email protected]) Accepted: 21 April 2015
Copyright © 2015 ISUOG. Published by John Wiley & Sons Ltd.
ORIGINAL PAPER
Early prediction and aspirin for prevention of pre-eclampsia Table 1 Risk factor scoring system for first-trimester prediction of pre-eclampsia2 Risk factor Chronic hypertension History of prior pre-eclampsia Diabetes mellitus Obesity (BMI > 30) Bilateral uterine artery notches Low PAPP-A (< 0.52 MoM)
Score 4 3 2 2 1 1
BMI, body mass index; MoM, multiples of the median; PAPP-A, pregnancy-associated plasma protein-A.
identified in the first trimester to be at high risk. The study was approved by Washington University School of Medicine Institutional Review Board, and registered with clinicaltrials.gov (NCT01547390). The inclusion criteria were singleton pregnancy undergoing ultrasound examination at 11 + 0 to 13 + 6 weeks and deemed to be at high risk for pre-eclampsia by the criteria listed in Table 12 . We excluded pregnancies with multiple gestation, fetal aneuploidy, major fetal structural anomaly and bleeding disorder, and women with allergy to aspirin or already on aspirin or heparin. All eligible women who presented for first-trimester ultrasound examination were approached for consent to participate in the study. Consenting women had the following assessments between 9 + 0 and 13 + 6 weeks: history for risk factors, uterine artery Doppler, and measurement of pregnancy-associated plasma protein-A (PAPP-A). A total score of at least 6 based on a risk factor scoring system in first trimester (Table 1) was initially needed to be eligible for randomization. However, due to slow recruitment 2 years into the study (with only 23 women enrolled) and following the recommendation from the data safety monitoring committee, the Institutional Review Board was asked for permission to revise the inclusion criteria to the presence of any of the risk factors (as listed in Table 1). Women meeting the inclusion criteria were randomized to receive daily 81 mg aspirin or a placebo pill from 11 + 0 to 13 + 6 weeks until 37 weeks or delivery, whichever occurred first. Randomization was 1:1 using a computerized randomization program by the pharmacist who was otherwise uninvolved in the study. The study team and the patient were blinded to the intervention the patient was receiving. The placebo/aspirin was dispensed by pharmacy at randomization and then again at the 18–24-week ultrasound visit and at the 28–32-week ultrasound visit. The women were encouraged to take the study medication with their regular prenatal vitamins. No specific instructions were given as to the timing of ingestion of the medication. Compliance was assessed by having the women complete a pill diary, which they were asked to return along with any unused pills, which were counted and returned to pharmacy. Demographic, previous medical and obstetric history and current obstetric information were collected from patient questionnaires and chart abstraction.
Copyright © 2015 ISUOG. Published by John Wiley & Sons Ltd.
415
The primary outcome was pre-eclampsia diagnosed per ACOG criteria at the onset of the study4 . Details of secondary outcomes were also collected. These included small-for-gestational-age neonate (birth weight < 10th percentile for gestational age on the Alexander growth standard5 ); early pre-eclampsia (delivery < 34 weeks); severe pre-eclampsia (blood pressure > 160/110 or symptoms including persistent headaches, visual disturbance, or evidence of abnormal renal failure, abnormal liver enzymes or thrombocytopenia); gestational hypertension (elevated blood pressure occurring after 20 weeks’ gestation with no evidence of pre-eclampsia); preterm birth (delivery < 37 weeks); stillbirth; antepartum hemorrhage (hemorrhage after 20 weeks’ gestation); neonatal death; neonatal intensive care unit admission; and miscarriage.
Statistical analysis Statistical analyses for the randomized controlled trial were originally planned on the ‘intention-to-treat’ principle but, based on the high withdrawal rate, the analysis was performed on an ‘as-treated’ principle. The primary outcome and other categorical variables were compared across groups using the chi-square or Fisher’s exact test as appropriate. We calculated the 95% confidence intervals around relative risk (RR) of the primary outcome in the intervention and placebo groups. We calculated the required sample size for the trial based on estimates using our scoring system. Using a receiver–operating characteristics (ROC) curve we determined the overall accuracy of our risk factor scoring system to be 78% (95% CI, 70–86%) and the optimal cut-off for predicting pre-eclampsia to be a total score of 62 . Based on the anticipated baseline rate of 36% in the placebo group, a total of 186 women would be needed for 80% power to detect a 50% reduction in the risk of pre-eclampsia (two-tailed α of 0.048; incorporating one interim analysis and using the O’Brien–Fleming stopping rule6 ). To account for an estimated 15% dropout rate, we initially estimated that 220 screen-positive women would need to be randomized. Following the revised inclusion criteria, a revised sample size of 684 women needed to be randomized to detect the same 50% reduction of pre-eclampsia with 80% power and 15% anticipated dropout rate; this was based on the revised accuracy of the model of 65%. However, the trial was terminated prematurely due to slow recruitment and lack of equipoise given a change to national guidelines to administer aspirin to high-risk women. All statistical analyses were performed using STATA software version 13.1 (STATA Corp., College Station, TX, USA).
RESULTS Between April 2012 and March 2014, a total of 1470 women were screened. Four hundred and forty-one were screened with the original algorithm and 1029 with the
Ultrasound Obstet Gynecol 2015; 46: 414–418.
416
modified one. Of these, 138 were screen positive (41 when using the original algorithm and 97 with the modified one) and 53 were eligible and consented to participate in the study and were randomized. Common reasons for patients declining to participate or being excluded from the study included the following: already being on aspirin, allergy to aspirin, already on anticoagulants for other indications such as a thrombophilia, concerns about safety of aspirin, and wanting to be guaranteed aspirin rather than a placebo. Some women with only one risk factor such as morbid obesity had the perception that their risk for pre-eclampsia was not high and therefore declined to participate. Ten women in the aspirin group and 13 in the placebo group dropped out of the study after randomization. Seven withdrawals in the aspirin group were by patient request, one pregnancy was terminated and two were lost to follow-up (Figure 1). The baseline characteristics of the 30 women who completed the study are shown in Table 2. Baseline characteristics were similar in the two groups except for a higher proportion of African Americans in the placebo group. The primary outcome of the study, pre-eclampsia, was seen in six of the 30 women (20%) with complete follow-up, three in the aspirin group and three in the placebo group (RR 0.88, 95% CI 0.21–3.66). All cases with pre-eclampsia occurred and were delivered before 34 weeks’ gestation. Gestational hypertension was seen in two women, both in the aspirin group. Small-for-gestational age was seen in two of the randomized women (6.7%), one in the aspirin and one in the placebo group (RR 0.88, 95% CI 0.06–12.72). With the exception of the three women reporting rash/hives, no other adverse effect was reported in either of the groups. The mean risk score was significantly higher (9.2 ± 1.2) in the six women who developed early pre-eclampsia vs 4.5 ± 0.4 in the other 24 women enrolled in the study with complete follow-up (P = 0.0001).
DISCUSSION This randomized trial, which was terminated prematurely, was affected by several practical issues that raise questions regarding the feasibility of conducting future studies with similar objectives. The findings, which should be interpreted with caution given that the study was underpowered to test our original hypothesis, are, however, consistent with the majority of trials using low-dose aspirin for preventing pre-eclampsia prior to 16 weeks7 – 11 . The inconclusive results reported by most trials aiming to use aspirin to prevent pre-eclampsia are mostly due to the sample size of these studies, as has been demonstrated by subsequent meta-analyses12 – 16 . Despite the conclusions of these meta-analyses as to the modest benefit for preventing all forms of pre-eclampsia, international guidelines have been published recommending the use of low-dose aspirin17 – 19 . The main limitation of our study is the small sample size, similar to other low-dose aspirin trials in early pregnancy. Our initial scoring system that was validated
Copyright © 2015 ISUOG. Published by John Wiley & Sons Ltd.
Odibo et al. in our population2 appeared too strict, resulting in very few high-risk women being assigned a large enough score to be randomized into the trial. The model was designed to be highly accurate, balancing the trade-off between sensitivity and specificity. Our model had a sensitivity/specificity that depended on the number of risk factors present and their weighted score. As our analysis was not based on sensitivity for a fixed false-positive rate, but on the accuracy of the overall model, it is difficult to compare the effectiveness of the model to other published series. Following the change in the inclusion criteria, the recommendations of the United States Preventive Services Task Force that women at risk for pre-eclampsia should receive low-dose aspirin early in pregnancy was published19 . With that recommendation, the investigation team met and determined that we had no equipoise to continue the trial. Given the small sample size, the primary goal of randomization, which is to eliminate potential biases, may not have been achieved. The United States Preventive Services Task Force recommendation was based on a systematic review including most of the data discussed above. It brings up the controversy of whether clinical decisions should be based on the results of individual trials or on meta-analysis20 . In addition, recent reviews have suggested using higher doses of aspirin and administering these at night time. It is not certain if taking these steps would have impacted the outcome of our trial21 – 23 . We noted a significant dropout rate in the trial. Of the 53 women who were randomized, 23 (43.3%) dropped out for a number of reasons including the development of rash or hives or itching, ‘forgetting to take pills’, being on ‘too many pills’ and termination of pregnancy. The high dropout rate may be a reflection of our study population, which includes a high proportion of inner city low income or indigent patients with social issues that can reduce attendance for prenatal care. To improve study retention, our study coordinators called enrolled patients on a weekly basis to encourage continuation of study medication and remind them of prenatal visit appointments. These efforts had little impact on the overall success of the study. While dropouts were balanced between the two groups and would not be expected to affect the results, given the overall small sample size it resulted in our performing an ‘as-treated’ rather than an ‘intention-to-treat’ analysis as is the usual case with trials. The high withdrawal rate also raises a question of compliance among women prescribed aspirin for prevention of pre-eclampsia. Interestingly, all cases of pre-eclampsia in our study occurred before 34 weeks. The 20% incidence is much higher than would be expected for an unselected population. We were unable to estimate the overall performance of our screening model. This information would have been helpful if we had been able to collect it. However, because we are ethically not allowed to follow those who screened negative (and did not consent to participate) or withdrew from the study, it is impossible
Ultrasound Obstet Gynecol 2015; 46: 414–418.
Early prediction and aspirin for prevention of pre-eclampsia
417
Multiparameter first-trimester pre-eclampsia screening (n = 1470)
Screen negative (n = 1332) Screen positive (n = 138)
79 screen positive declined randomization, 16 found ineligible*
Randomized (n = 53)
Dropouts (n = 10) (7 on patient request, 1 termination, 2 LTF)
Treatment group (n = 26)
Placebo group (n = 27)
Treatment group with complete follow-up data (n = 16)
Placebo group with complete follow-up data (n = 14)
Dropouts (n = 13) (10 on patient request, 3 rash/hives)
Primary outcome: pre-eclampsia (n = 6)
Figure 1 Flowchart of study procedures and enrollment. *Ten women met both criteria. LTF, lost to follow-up. Table 2 Demographic data and inclusion criteria for the Early Prediction and Aspirin for Prevention of Pre-eclampsia (EPAPP) study
Maternal age (years) Body mass index (kg/m2 ) Nulliparous Race/ethnicity Black White Other Risk score* GA at randomization (weeks) Mean UtA-PI Chronic hypertension Pregestational diabetes
Placebo (n = 14)
Aspirin (n = 16)
P
31.6 ± 6.1 36.6 ± 6.9 3 (21.4)
30.0 ± 5.0 37.4 ± 8.9 5 (31.2)
0.45 0.78 0.54 0.013
11 (78.6) 3 (21.4) 0 (0) 5.9 ± 3.2 12.1 ± 1.0 1.68 ± 0.4 10 (71.4) 2 (14.3)
4 (25) 1 (6.2) 11 (68.8) 5.0 ± 2.9 11.7 ± 1.3 1.57 ± 0.8 6 (37.5) 4 (25.0)
0.40 0.35 0.69 0.06 0.46
Data are given as mean ± SD or n (%). *Risk score as defined by Goetzinger et al.2 . GA, gestational age; UtA-PI, uterine artery pulsatility index.
to calculate the overall performance of the model. The mean screening score of the women who developed early pre-eclampsia was significantly higher than that of the other women enrolled in the study. It is possible that these women’s risk was so high that low-dose aspirin could not
Copyright © 2015 ISUOG. Published by John Wiley & Sons Ltd.
prevent them from developing pre-eclampsia. The above limitations should inform investigators contemplating future studies for pre-eclampsia prevention. Based on current guidelines, which may be premature, we wonder if future trials focused on the primary question of
Ultrasound Obstet Gynecol 2015; 46: 414–418.
418
pre-eclampsia prevention with aspirin are ethical or practical given the problems we encountered with this trial. The failure to detect a difference in our study should be interpreted with caution due to the difficulties encountered with the trial and the study being underpowered to show a difference. Future trials should focus on the optimal dose of aspirin or combination of antiplatelet therapy to potentially optimize the prevention of pre-eclampsia.
REFERENCES 1. Poon LC, Kametas NA, Maiz N, Akolekar R, Nicolaides KH. First-trimester prediction of hypertensive disorders in pregnancy. Hypertension 2009; 53: 812–818. 2. Goetzinger KR, Tuuli MG, Cahill AG, Macones GA, Odibo AO. Development and validation of a risk factor scoring system for first-trimester prediction of preeclampsia. Am J Perinatol 2014; 31: 1049–1056. 3. Bujold E, Roberge S, Lacasse Y, Bureau M, Audibert F, Marcoux S, Forest JC, Gigu`ere Y. Prevention of preeclampsia and intrauterine growth restriction with aspirin started in early pregnancy: a meta-analysis. Obstet Gynecol 2010; 116: 402–414. 4. ACOG Practice Bulletin. Diagnosis and management of preeclampsia and eclampsia. Number 33, January 2002. Obstet Gynecol 2002; 99: 159–167. 5. Alexander GR, Himes JH, Kaufman RB, Mor J, Kogan M. A United States national reference for fetal growth. Obstet Gynecol 1996; 87: 163–168. 6. Pocock SJ. When to stop a clinical trial. BMJ 1992; 305: 235–240. 7. Ebrashy A, Ibrahim M, Marzook A, Yousef D. Usefulness of aspirin therapy in high-risk pregnant women with abnormal uterine artery Doppler ultrasound at 14–16 weeks pregnancy: randomized controlled clinical trial. Croat Med J 2005; 46: 826–831. 8. Bakhti A, Vaiman D. Prevention of gravidic endothelial hypertension by aspirin treatment administered from the 8th week of gestation. Hypertens Res 2011; 34: 1116–1120. 9. Villa P, Kajantie E, Raikkonen K, Pesonen AK, Hamalainen E, Vainio M, Taipale P, Laivuori H; PREDO Study group. Aspirin in the prevention of pre-eclampsia in high-risk women: a randomised placebo-controlled PREDO Trial and a meta-analysis of randomised trials. BJOG 2013; 120: 64–74. 10. Vainio M, Kujansuu E, Iso-Mustajarvi M, Maenpaa J. Low dose acetylsalicylic acid in prevention of pregnancy-induced hypertension and intrauterine growth retardation in women with bilateral uterine artery notches. BJOG 2002; 109: 161–167.
Copyright © 2015 ISUOG. Published by John Wiley & Sons Ltd.
Odibo et al. 11. August P, Helseth G, Edersheim T, Hutson J, Druzin M. Sustained release, low-dose aspirin ameliorates but does not prevent preeclampsia (PE) in a high risk population. Proceedings of the 9th International Congress, International Society for the Study of Hypertension, 1994; 72. 12. Imperiale TF, Petrulis AS. A meta-analysis of low-dose aspirin for the prevention of pregnancy-induced hypertensive disease. JAMA 1991; 266: 260–264. 13. Askie LM, Duley L, Henderson-Smart DJ, Stewart LA. Antiplatelet agents for prevention of pre-eclampsia: a metaanalysis of individual patient data. Lancet 2007; 369: 1791–1798. 14. Bujold E, Morency AM, Roberge S, Lacasse Y, Forest JC, Gigu`ere Y. Acetylsalicylic acid for the prevention of preeclampsia and intra-uterine growth restriction in women with abnormal uterine artery Doppler: a systematic review and meta-analysis. J Obstet Gynaecol Can 2009; 31: 818–826. 15. Roberge S, Villa P, Nicolaides K, Gigu`ere Y, Vainio M, Bakthi A, Ebrashy A, Bujold E. Early administration of low-dose aspirin for the prevention of preterm and term preeclampsia: a systematic review and meta-analysis. Fetal Diagn Ther 2012; 31: 141–146. 16. Roberge S, Nicolaides KH, Demers S, Villa P, Bujold E. Prevention of perinatal death and adverse perinatal outcome using low-dose aspirin: a meta-analysis. Ultrasound Obstet Gynecol 2013; 41: 491–499. 17. NICE clinical guideline No. 107. Hypertension in pregnancy. Management of hypertensive disorders in pregnancy. Issued 2010, updated January 2011. http://www.nice.org. uk/nicemedia/live/13098/50418/50418.pdf [Accessed 20 March 2013]. 18. WHO Recommendations for Prevention and Treatment of Preeclampsia and Eclampsia. World Health Organization, 2011. http://whqlibdoc. who.int/publications/2011/9789241548335_eng.pdf [Accessed 20 March 2013]. 19. Henderson JT, Whitlock EP, O’Conner E, Senger CA, Thompson JH, Rowland MG. Low-Dose Aspirin for the Prevention of Morbidity and Mortality From Preeclampsia: A Systematic Evidence Review for the U.S. Preventive Services Task Force. Rockville (MD, USA): Agency for Healthcare Research and Quality (US); 2014. 20. Scifres CM, Iams JD, Klebanoff M, Macones GA. Meta-analysis vs large clinical trials: which should guide our management? Am J Obstet Gynecol 2009; 200: 484.e1–5. 21. Caron N, Rivard GE, Michon N, Morin F, Pilon D, Moutquin JM, Rey E. Low-dose ASA response using the PFA-100 in women with high-risk pregnancy. J Obstet Gynaecol Can 2009; 31: 1022–1027. 22. Hermida RC, Ayala DE, Fernandez JR, Mojon A, Alonso I, Silva I, Ucieda R, Codesido J, Iglesias M. Administration time-dependent effects of aspirin in women at differing risk for preeclampsia. Hypertension 1999; 34: 1016–1023. 23. Park F, Russo K, Williams P, Pelosi M, Puddephatt R, Walter M, Leung C, Saaid R, Rawashdeh H, Ogle R, Hyett J. Prediction and prevention of early-onset pre-eclampsia: impact of aspirin after first-trimester screening. Ultrasound Obstet Gynecol 2015; 46: 419–423.
Ultrasound Obstet Gynecol 2015; 46: 414–418.
