Does Pentaerytrithyltetranitrate reduce fetal growth restriction in pregnancies complicated by uterine mal-perfusion? Study protocol of the PETN-study: a randomized controlled multicenter-trial

The aim of this study is to confirm the safety and effectiveness of PETN for secondary prophylaxis of FGR in a prospective, multicenter, randomized, double-blind, placebo controlled, parallel group, interventional trial.

Ethical approval has been obtained from the ethical committee of the University Hospital of Jena as leading committee (Institutinal review board (IRB) Friedrich Schiller University, Jena) (5085–02/17). Each of the 14 German study centers obtained ethical approval of their local committees (IRB Universitäts-Kinderklinik, Kiel, IRB Hannover Medical School, Hannover, IRB Medical School, Rheinischen Friedrich-Wilhelms-University, Bonn, IRB State Office for Health and Social Affairs (LAGeSo) Berlin (two centres), IRB of the state of Saxony-Anhalt, Halle (two centres), IRB Medical School, Leipzig, IRB Technical University Dresden, IRB Medical School, Eberhard-Karls-University, Tübingen, IRB University of Ulm, IRB Ludwig Maximilian University, Munich, IRB national Medical association Bavaria, Munich), which than were referred to the Jena committee before approval was issued. Written informed consent must be obtained before the performance of any protocol related procedures that are not part of usual subject care.

Eligible patients for the proposed trial are pregnant women recognized to be at risk of adverse pregnancy outcome associated with FGR at mid gestation. Since poor placental function and mal-perfusion of the placenta is recognized to cause FGR [3] impaired placental flow measurement is strongly associated with the development of FGR. A mean pulsatility index >95th percentile according to Cnossen et al. predicts FGR with a sensitivity of 18% with a specificity of 95% in a general population and a sensitivity of 58% for any FGR in a high risk population with a specificity of 75% [4]. The risk to develop FGR is in general higher in women with a medical history of chronic hypertension, diabetes mellitus type 2, systemic lupus erythematodes or antiphopholipid antibody syndrome and in those with prior history of adverse pregnancy outcome associated with FGR (late pregnancy loss, stillbirth, abruption placentae, preterm delivery, preeclampsia or HELLP-syndrome) [13]. Accordingly, at least one third of the pregnant women diagnosed with impaired perfusion of the uterine arteries at mid gestation will develop FGR and associated adverse pregnancy outcomes. In the presence of additional risk factors this likelihood will rise considerably.

Consequently, the Inclusion criteria are impaired uterine artery Doppler ultrasound, defined by a mean Pulsatility Index (PI) greater than 1.6 (> 95th centile of German reference population, PIA Fetal Database, Version 4.00 and higher, Viewpoint, Germany; [14]), singleton pregnancy at 19⁰ to 22⁶ weeks of gestation who are over 18 years old and provide informed consent.

To ensure the inclusion of solely placental reasons for FGR and for ethical reasons, exclusion criteria are multiple pregnancies, fetal chromosomal or suspected major structural defects, circumstances predicting adverse pregnancy outcome other than due to FGR, like premature rupture of membrane or cervical insufficiency at the time of enrolment, participation in any other clinical trial and maternal disease defined as contraindication for intake of PETN.

As primary efficacy endpoint development of FGR (< birth weight below the 10th centile) or perinatal death as a composite outcome was proposed in accordance to the pilot study [12]. Key secondary endpoints are severe morbidity and mortality as composite of severe FGR, intrauterine or neonatal death or placental abruption, birth weight below the 3rd, 5th and 10th centile according to Voigt [15], development of FGR requiring delivery before 30 and before 34 weeks` gestation, spontaneous preterm delivery before 34 weeks and before 37 weeks gestation, rate of neonatal intensive care unit admission and neonatal complications like the need of ventilation, occurrence of intraventricular hemorrhage II – IV ° or necrotized enterocolitis.

The gestational age was calculated from a certain last menstrual period date and confirmed by first trimester measurement of the crown-rump-length. Weight percentiles are derived from gestational age and weight according to the German standards by Voigt [15]. Falling below the 10th weight percentile in combination with impaired uterine perfusion (inclusion criteria mean PI of > 1.6) is defined as FGR [16].

Development of severe FGR is defined as birth weight below the 5th centile. Preterm delivery is determined from gestational age at birth.

Patients will be randomly assigned to treatment groups in blocks of varying size. Randomization will be stratified by center and by an indicator for low or high risk of FGR. Patients will be classified as high risk in case of pre-existing hypertension, pre-existing diabetes or obstetric history of FGR, late pregnancy loss, stillbirth, abruption placentae, preterm delivery, preeclampsia or HELLP-syndrome. All others are at low risk. PETN and placebo will be administered in a double-blind fashion. Primary analysis will be performed in the intent-to-treat population.

Sample size calculation is based on results generated from our pilot trial [12]. Since additional risk factors increase the risk of FGR a stratified randomisation and analysis is planned for two strata. A 2:1 ratio was observed for low and high risk patients in the trial. Proportion of FGR or perinatal death was 0.42 in the low-risk and 0.69 in the high-risk stratum of the placebo group. We discovered an odds ratio of 0.41 (PETN vs. placebo) in the pilot trial. More conservatively, we want to be able to find a 50% risk reduction (i.e. an odds ratio of 0.5) in the here proposed study, which would still be of profound clinical relevance. Accordingly, we will need a sample size of 290 patients to detect this treatment effect by a Mantel-Haenszel test with a power of 80% at a two-sided alpha-level of 0.05 (nQuery 7.0). Assuming a dropout rate of about 10% a total of 324 patients should be randomized.

During the pilot study the reported compliance was markedly high. Of the 111 patients included in the study only 5 discontinued intervention (5%), of whom 2 in the verum group stopped intervention due to potential side effects (dizziness and headache). This high compliance was likely to be related to the high burden derived from the worry about the unborn.

The number of patients calculated for allocation to the PETN trial accounts for a more conservative dropout rate of 10%, due to the particularly good compliance of pregnant women, especially those to be at risk for neonatal complications. For the primary outcome, robustness of the results regarding missing values will also be analyzed.

Patients eligible are recruited from those pregnant women presenting during prenatal care for routine or risk guided midtrimester ultrasound screening for fetal structural anomalies including the assessment of utero-placental perfusion. 5% are expected to demonstrate impaired utero-placental flow with a mean PI exceeding the 95th centile indicating a risk for IUGR. In our pilot trial an estimate of 50% of the eligible patients were enrolled per year. Accordingly, to recruit 324 patients within 18 months we need to identify 650 women with impaired utero-placental perfusion at mid gestation. Based on the fact, that of all women presenting for mid gestation ultrasound 5% will meet the inclusion criteria (mean PI >95th centile (> 1.6)) a cohort of about 13,000 women will be needed to recruit from. To enlarge the cohort, study centers are asked to cooperate with affiliated outpatient facilities, run by gynecologist holding a DEGUM II or III qualification. Patients recruited in the affiliated facilities will then be sent to the participating study center for verification of eligibility and study inclusion. (Fig. 1).

The study started recruiting in August 2017. The coordinating hospital and trial coordinator is the University Hospital of Jena, PD Dr. Tanja Groten. Participating hospitals and local trial coordinators are University Hospital Schleswig Holstein, PD Dr. Ulrich Pecks; Medical University Hanover, Prof. Dr. Constantin von Kaisenberg; University Hospital Bonn, Mrs. Mateja Condic; Charité Berlin, PD Dr. med. Stefan Verlohren; Vivantes Network of Health GmbH, Clinic Neukoelln, PD Dr. med. Dietmar Schlembach; University Hospital and Polyclinic of Obstetrics and Prenatal Medicine, Dr. med. Gregor Seliger; Obstetric Clinic St. Elisabeth and St. Barbara Hospital Halle, Dr. Sven Seeger; University of Leipzig Medical Center, Division of Obstetrics, Dr. Anne Tauscher; University Hospital Dresden Carl Gustav Carus, Dr. Matej Komar; University Women’s Hospital Tuebingen, Prof. Dr. Karl Oliver Kagan; Ulm University, Clinic of Gynecology and Obstetrics, PD Dr. Ulrike Friebe-Hoffmann; University Hospital of Munich, Dr. med. Christoph Hübener and Municipal Hospital Munich, Dr. Laura de Vries. As of January 2019 206 patients have been enrolled in the study. Enrollment has been expanded to Dezember 2019.

Following randomisation, oral intake of PETN or placebo two times daily will be continued until 36⁶ weeks. Maternal and fetal status and condition will be assessed every 4 weeks during the treatment period and at 38⁶ to 40⁰ weeks. (Table 1) Outcome data will be collected after delivery for neonatal (weight, height, APGAR 1′/5′/10′, umbilical artery pH, gestational age at birth, admission to the neonatal intensive care unit, respiratory distress syndrome grade III or IV, occurrence of intraventricular haemorrhage grade III or IV or parenchymal bleeding or periventricular leukomalacia, necrotized enterocolitis, neonatal death) and maternal pregnancy associated morbidity. Additionally, data on child development will be documented on growth, health status and intellectual development at the age of 12 months, by collecting the routine check-up data from the paediatricians. (Table 1).

Monitoring of selected safety data (maternal serum and clinical parameters, fetal growth control and control of utero-placental and fetal perfusion) is scheduled every 4 weeks. Monitoring of Adverse Events (AEs) and Serious Adverse Events (SAEs) is established. A data safety monitoring board is constituted to supervise safety events and outcome data of the neonates born during the ongoing study. The data safety monitoring board has met on a regular basis; three times so far, did not find any peculiarities and had no recommendations on protocol changes.

Baseline characteristics of patients will be compared by descriptive statistics to assess the success of randomization. The primary endpoint (binary variable) will be presented by group-specific incidences. The confirmatory test of the treatment effect stratified for risk-groups will be performed by the Mantel-Haenszel test. Mantel-Haenszel estimates of relative risks and their 95% confidence intervals will be calculated. Primarily, data will be analyzed according to the intent-to-treat (ITT) principle. The ITT-population will include all randomized patients with valid primary outcome data who have received at least one dose of study medication (full analysis data set). The drop-out rate is expected to be 10%. Missing values of the primary endpoint are multiply imputed to confirm the robustness of the main results (sensitivity analysis). All secondary endpoints (binary variables) will be analysed by Mantel-Haenszel methods as described above. In a sensitivity analysis generalized mixed linear models with treatment and risk indicator as fixed, and centre as random factor, will be fitted to adjust for variability between study centres. Explorative analyses of endpoints will be additionally performed for subgroups according to time of delivery and maternal age. AEs/SAEs will be listed. Incidence of events and descriptive statistics (quartiles, mean, standard deviation) of laboratory safety data will be displayed in treatment groups. No adjustment for endpoint multiplicity is planned. All tests will be done at a comparison-wise significance level of 0.05.