Treatment of classic mid-trimester preterm premature rupture of membranes (PPROM) with oligo/ anhydramnion between 22 – 26 weeks’ gestation by means of continuous amnioinfusion: a randomized multicentric prospective controlled TRIAL

Background: The classic mid-trimester preterm premature rupture of membranes (PPROM), is dened as rupture of fetal membranes prior to 28 weeks’ gestation (WG) with oligo/ anhydramnion, complicates approximately 0.4-0.7% of all pregnancies and is associated with very high neonatal mortality and morbidity. Antibiotics have limited success to prevent bacterial growth, chorioamnionitis and fetal inammation. The repetitive amnioinfusion doesn’t work because of immediately uid lost after the intervention. The continuous amnioinfusion through the transabdomianal port system or catheter in patients with classic PPROM shows promise by ush out of bacteria and inammatory components from the amniotic cavity, replacing amniotic uid and thus prolonging PPROM-to-delivery interval. Aim: This multicenter trial tests the effect of continuous amnioinfusion on the neonatal survival without the typical major morbidities, like severe bronchopulmonary dysplasia, intraventricular hemorrhage, cystic periventricular leukomalacia and necrotizing enterocolitis one year after the delivery. Methods/Design: randomized multicenter trial; two-arm parallel design. Control group: PPROM patients between 22/0 (20/0) -26/0 WG treated with antibiotics and corticosteroids in accordance to guidelines of German Society of Obstetrics and Gynecology (standard PPROM therapy). In the interventional group the standard PPROM therapy will be complemented by “Amnion Flush Method” with the amnioinfusion of articial amniotic uid (up to 100 ml/h, 2400 ml/day). Subjects: 68 patients with classic PPROM between 22/0 WG. TRIAL-registration:


Introduction
Mid-trimester PPROM complicates approximately 0.4-0.7% of all pregnancies and is associated with very high neonatal mortality and severe morbidity [1][2][3][4]. The causes of the mid-trimester PPROM are multifactorial [4]. The mechanism of PPROM is discussed as altered membrane morphology including marked swelling and disruption of the collagen network, which can be triggered by bacterial products and pro-in ammatory cytokines and involves the activation of matrix metalloproteinases. [4][5][6][7].
The propagation of bacteria is an important contributing factor in adverse neonatal and maternal outcomes after PPROM [1,4,8]. Gomez et al reported that antibiotics failed to eliminate the amniotic infection in 83% of PPROM cases [9]. The "classic PPROM" with oligo-/anhydramnion is associated with short latency period and worse neonatal outcome compared to similar gestational aged neonates [4].
In Germany, the number of pregnant women with mid-trimester classic PPROM has been calculated to be about 5-7 thousand per year. Generally, about 40% of very preterm infants who survive the initial NICU die during next 5 years of life and the long-term morbidity of the survivors remain high. Furthermore, more than 40% of surviving neonates following PPROM prior to 25 weeks of gestation develop bronchopulmonary dysplasia (BPD) later on [4,10].

Background And Rationale
The fetal in ammatory response syndrome (FIRS), a systemic in ammation determined by elevated fetal plasma IL-6 concentrations and histological signs of funisitis in the umbilical cord, is an additional independent risk factor for the occurrence of severe neonatal morbidity [11].
The lavage effect of continuous amnioinfusion could prevent the fetus and amniotic cavity from bacterial colonization, reduce the in ammatory response and thus protect the neonate from major complications, such as pulmonary hypoplasia, sepsis, cerebral palsy and joint deformities [4,[12][13][14][15].
Serial amnioinfusions with standard electrolyte solutions have seen little or no effects onto neonatal outcome probably because of immediately uid lost after the intervention and possible negative effect of longtime saline solution onto the fetus [4,14,[16][17][18][19].
Tchirikov et al. published the prolongation of the PPROM-to-delivery interval to 49 days in average of patients with classic PPROM using the intensive lavage of the amniotic cavity with Amnion Flush solution (pump rate 100 ml/h) throughout a transabdominal catheter with an additionally designed anker system [4,12,20,21].
However, in recently published retrospective results from Japan the authors did not improve the neonatal outcome using low rate of continuous amnioinfusion with Ringer's lactate solution (40 ml/h) [14,15]. In the current German medical guideline to prevention of the preterm birth the amnioinfusion for the treatment of the PPROM is mentioned, but prospective randomized studies must be performed to clarify the positive effect of the continuous amnioinfusion onto the fetal survival [22].

Objectives
Principal research question: Does the intervention (continuous amnioinfusion -" ush out" method) reduce adverse neonatal outcome of patients with PPROM with oligo/anhydramnion and prolong the pregnancy.

The full infrastructure of Coordination Center for Clinical Trials (KKS) of Martin-Luther-University Halle
Wittenberg will be involved in ICH-GCP compliant data management and handling, randomization, safety management and monitoring of the study.

Study setting
The study will be performed in 10 tertiary German perinatal centers including six leading German Universities (Fig. 1).

Eligibility criteria
Inclusion criteria: single pregnancy (from 22/0 to 26/0 weeks' gestation), with classic PPROM with oligo/ anhydramnion between 20/0 to 26/0 weeks' gestation. From the beginning of 22/0 weeks' gestation the PPROM duration must be lower as of ve days till randomization. The patients receive RDS prophylaxis before randomization. The vertical deepest amniotic uid depot must be <2 cm, and the amniotic uid Index < 3. The PPROM will be con rmed by positive testing of the placental alpha microglobulin-1 and clearly amniotic uid lost (Fig 2, Flow diagram).
The authorized physicians of perinatal centers taking part in the trial will identify the potential trial participants with classic PPROM. Eligible women with classic mid-trimester PPROM will be asked for their written informed consent at the time of admission to the hospital -after having received oral and written information. Patients, who are not able to understand the informed consent, will not been included. Additionally for ve years all patients undergoing a fetal surgery therapy in hospital had led a motion for a compassionate use at ethical review committee. Recruitment for " ush out" method via a port-system is ethically sound for PPROM patients with oligo-/anhydramnion based on available research evidence.
In case of complications all equipment for emergency interventions will be easily accessible in the perinatal centers. Additional "safety net" for nalization of the study and induction of delivery has been established to avoid any complications.

Control group
The PPROM patients of the control group receive the standard therapy related to the German Guidelines including antibiotic therapy (Amoxicillin/Clarithromycin therapy or 7 days Amoxicillin and once Azithromycin 1 g per os) or other antibiotics according to recommendations of the physicians from the microbiology department based on the culture results for at least 7 days [22]. Daily blood analysis (Hb, WBC, Th, CRP, IL-6), cervical smear every 5 days. The RDS prophylaxis must be performed before randomization.

Amnion Flush interventional group
The PPROM patients of the Amnion Flush Group receive the standard therapy related to the German Guidelines like the control group [22]. Additionally, under local anaesthesia and ultrasound guiding the perinatal catheter with an anchor system (0.65 mm Diameter, CE 0481, PakuMed GmbH, Essen, Germany) will be inserted into the amniotic cavity for the continuous amnioinfusion with 100 ml/h of Amnion Flush Solution, (Serumwerk Bernburg AG, Germany) until delivery (latest on beginning of 34/0 weeks of gestation) [4,12,13] (Fig. 3).

Detailed description of the perinatal catheter implantation
The catheter implantation will be performed in the operation room in aseptic conditions. Before the catheter implantation the amnioinfusion with 300 ml of Amnion Flush Solution with be performed using 22 gauge needle under ultrasound control. The gynaecologist must have an experience of at least of 100 amniocenteses because of di cult conditions for the puncture related to anhydramnion.
The operation table will be kept for maximum Trendelenburg patients' position avoiding completely lost of infused Amnion ush Solution because of PPROM. If patients will complain about uterine contractions, the single short of Partusisten 25µg will be slowly infused i.v.
The appropriate location for the catheter implantation (preferred area is the uterine fundus without placenta) will be identi ed using ultrasound. The local anaesthesia with Xylocaine 1% 10 ml will be injected [20]. The 18 gauge needle will be inserted into the amniotic cavity. The anchor catheter will be inserted throughout the needle. The 18 G needle will be removed. The catheter will be xed up using the sterile transparent foil. The Amnion Flush Solution will be infused using standard medical pumps with the rate of 100 ml/h (+/-20 ml). Daily ultrasound control must be performed to validate the intraamniotic catheter position and the deepest deport of the amniotic uid. The deepest pool of amniotic fluid should be stabilized (target 4 cm). The catheter should be replaced if indicated (e.g. delivery or catheter dislocation or plugging) but latest every 30 days to avoid contamination and/or in ammation.
Secondary outcomes: Gestational age at delivery, duration of PPROM-to-delivery interval, perinatal mortality, incidence of histological a rmed chorioamnionitis, incidence of FIRS and/or maternal sepsis, cord-artery pH, APGAR scores, neonatal outcome at calculated age 36 weeks of gestation: duration of O 2supplementation, retinopathy of prematurity. Participants with surviving babies will send a prepaid validated outcome questionnaire at 12 months after the birth of their baby. The independent University Center of Clinical Studies of Martin-Luther-University Halle-Wittenberg will receive the copies of pediatric examinations performing monitoring and nal statistical evaluation. Neurodevelopmental assessments as well as the check-up of the surviving children will be performed in tertian medical centers by trained pediatric specialists.

Participant timeline
The study site visits for monitoring the clinical trial will be carried out by chief and assistant monitors. During the clinical phase, the chief monitor will visit the medical centers every 6 months (see table 1), thereby being available to staff, supervising the assistant monitors, and in order to address and solve any problems arising. They will review the progress of recruitment and ensure participants' rights and safety, adherence to protocol regulatory compliance. During query process by the KKS, which implies another month of study site visits (beyond the end of recruitment), they will verify unclear data and obtain corrections. Chief monitor will also make sure that plenty of CRF are available to staff, and questionnaires are given to mothers for lling them in prior to discharge from hospital.
The intervention is expected to prolong the gestation by around two weeks resulting in a increase of the one year survival rate without major morbidities after 27 gestational weeks of 67% (hazard=0.0334).
Based on Chen et al. (2016) the one year survival rate without major morbidities after 25 weeks gestation is around 36% (hazard=0.0851) [23]. To be more conservative sample size calculation was based on a hazard=0.035 for the interventional group. Using a recruiting period of 21 months (four women per month) and an individual follow-up of one year 34 women are needed in each group for a two-sided log-rank test with 5% signi cance level, a power of 80% and a loss to follow up rate of 10% per year (hazard=0.0088). Calculations were performed by PASS software.

Assignment of interventions: allocation, concealment mechanism and implementation
The randomization will be performed using the randomizer of the Center of the clinical studies of the Martin-Luther University Halle-Wittenberg, Germany (computer-generated random numbers). Before the randomization the inclusion and exclusion criteria of the study will be printed by investigators into the randomizing table online. Matching criteria will be automatically analyzed and the group of the study and the number will be generated online.

Data collection and management
The trial is registered with full description at the German Clinical Trials Register (Deutsches Register Klinischer Studien, DRKS) and at Clinical Trials Register (ClinicalTrial.gov). The study's protocol is published at the start of the trial.
The Coordination Center for Clinical Trials, University of Halle (Saale), is responsible for data management and archiving. The data will be managed and archived according to existing standards and regulations, in a data format allowing long-term preservation and future reuse.
The data sheet will be completed when the baby will discharge home or after death. Any missing data will be reconciled by the chief investigator and trial administrator by contact with the PIs and examination of the hospital case notes. The data pack will be returned to the trial co-ordination centers after the baby will discharge home or after death.
Post-trial follow-up is at 6 and 12 months postpartum when questionnaires are being sent to mothers' home addresses for the assessment of subjective variables as well as variables pertaining to neonatal medical care for the health-economic evaluation.
CRF will include data for recruitment and con rmation of eligibility, informed consent, baseline data, process, data during treatment, primary and secondary outcomes, other outcomes, safety variables, prognostic factors, various questionnaires, as well as data of the pediatric evaluations.
The KKS will receive the copies of pediatric examinations. Neurodevelopmental assessments as well as the check-up of the surviving children will be performed in medical centers by trained pediatric specialists.

Monitoring
The coordinating investigator is responsible for implementing and maintaining quality assurance and quality control systems with written SOPs to ensure that the trial is conducted and data generated, documented and reported in compliance with the protocol, ICH-GCP and the applicable regulatory requirements. Tasks assigned to KKS Halle (see chapter 9, trial supporting facilities) will be ful lled according to the written SOPs from KKS Halle and German-wide KKS Network. Data from source documents will be transcribed onto the case report forms (CRF) by the clinical research staff of each trial site. All source documents are kept in the trial site and no information in source documents about the identity of the patients will be disclosed. Pre-trial visits and on-site monitoring will be conducted by trained CRAs from KKS Halle. 3-5 pre-trial visits will be performed to allow selection of 3 sites for participation. Based on the pre -trial visit reports the PI and the steering committee will decide on participation of the respective trial site. Because of previous on-site selection visits, for initiation of the study centers conference calls (each site separately) will be considered. Initiation will be supported by .ppt presentations and study materials (investigator's site le, protocol, informed consent, CRFs etc.) earlier supplied to the sites. On-site monitoring will be performed on a risk-based approach and is planned by means of regular visits from the beginning to the end of the trial to check the completeness of patient records, the consistency of entries on the CRFs and the adherence to the protocol and to Good Clinical Practice. Initially the informed consent forms and the adherence to the inclusion and exclusion criteria will be checked. Thereafter, a formal check of captured information for completeness and plausibility will be done followed by a check of the correct transfer of data from the source data. Full (100%) source data veri cation for the presence of informed consent and adherence to the inclusion/exclusion criteria will be performed. Data used for all primary and safety variables will also be 100% source data veri ed. The speci c extent of the monitoring and the source data veri cation will be speci ed in the monitoring manual. Each trial site will be monitored once after the rst trial participant has been enrolled and treated.
The early monitoring shall ensure prompt intervention by CRA or study center in case of any problems at the trial site such as major inconsistencies in trial conduction and CRF completion. Another 7 (average, individual number depending on recruiting rate and quality of CRF completion) monitoring visits at each trial site will be conducted during the enrolment period. Every trial site will undergo a close-out visit by the CRA after the last participant at that site has nished the follow up. The monitor's access to the trial documents and medical records is ensured by the investigator's agreement and the cooperation agreement between the sponsor and KKS Halle.

Statistical analysis
The collected data will be analyzed by specialists of statistic of the Institute of Medical Epidemiology, Biostatistics and Informatics, Martin-Luther-University Halle-Wittenberg, Germany Adverse event reporting harms The Data Safety Monitoring Board (DSMB) will follow the progress of the clinical trial, evaluate the safety and primary e cacy parameters and will propose to the sponsor whether to continue, modify or stop a trial and provide the funding agency with information and advice. The DMSB will set up telephone conferences twice a year to evaluate current safety data. Therefore, the sponsor will provide up-to-date safety data including a summary of AEs, line listing of SAEs and adverse device effects (ADEs). In case of a high number of severe unexpected events in between DMSB meetings or a case of a Suspected Unexpected Serious Adverse Reaction (SUSAR) or other medically important conditions the DMSB will be informed by the sponsor immediately and may give advice for further procedures if required. An interim analysis is not planned.
Stopping rules a) for patients: placental abruption, indication for termination of pregnancy (e.g HELLP syndrome, pathological CTG, fetal bradycardia, eclampsia, umbilical cord prolapse), signs of chorioamnionitis (maternal fever > 37.8°C and two or more of the following criteria: uterine tenderness, malodorous vaginal discharge, maternal leukocytosis > 15000 cells/mm 3 (without corticosteroids), maternal tachycardia > 100 beats/min, and fetal heart rate > 160 bpm), CRP > 20 mg/dl, increased concentrations of interleukine-6 (> 7 pg/ml), positive bacteriologic examination of the amniotic uid, pos. Gram-stained sediment or IL-6 concentration > 2600 pg/mL or WBC >30 mm 3 . b) for participating centers: inability to recruit the patients, complications during the catheter implantation, like placental abruption, damage of the fetus or umbilical cord, fetal bleeding or death (audit after rst complication, center ex in case 2 or more complications/year).
for the whole trial: high range of complications in interventional group compared to the control group (we are going to audit the study every 3 months discussing this point), unexpected adverse reactions to the medical products without correction possibilities, clear statistical bene t of " ush-out" treatment.

Frequency and plans for auditing trial conduct
The medical centers of clinical trial will undergo an audit twice a year by a study manager, KKS, Martin-Luther-University Halle-Wittenberg. To ensure quality of data and study performance the sponsor may conduct site visits by an independent auditor. An audit will only be performed after noti cation and arrangement with the investigator. An audit certi cate will be issued as quality proof and has to be led in the trial master le and as a copy in the investigator site le.

Ethics approval and consent to participate
The study will be conducted in compliance with the Declaration of Helsinki and all amendments.
The TRIAL has got a permission of the Ethical Committee of the Halle University Medical Center (2020-185, January 25, 2021).
The trial was registered with ClinicalTrials.gov ID: NCT04696003 and German Clinical Trials Register: DRKS00024503.
Written informed consent of the patients as well as the parents is required for enrolment in the study and can be withdrawn at any time without reasons or disadvantages. Eligible women with classic midtrimester PPROM will be asked for their written informed consent at the time of admission to the hospital -after having received oral and written information. Patients, who are not able to understand the informed consent, will not been included. Recruitment for " ush out" method via a port-system is ethically sound for PPROM patients with oligo-/anhydramnion based on available research evidence.
In case of complications all equipment for emergency interventions will be easily accessible in the perinatal centers. Only hospitals which possess a Level 1 perinatal center in accordance to German law (tertian perinatal center) will receive permission to participate on this trial.
Any modi cations to the protocol will be sent to the ethics committees before they are implemented within the study and communication changes that impact the patients would require signing of a revised consent form.

Discussion
Mid-trimester preterm premature rupture of membranes (PPROM) together with oligohydramnios before 26 weeks affects about 0.4%-0.7% of all pregnancies. A high number of unreported cases should be expected in cases of PPROM priory to 22 weeks of gestation. Mid-trimester PPROM with anhydramnion is associated with a very high neonatal mortality rate as well as an increased risk of long-and short-term severe neonatal morbidities, physical and developmental disabilities, including chronic respiratory disease, neurodevelopmental or behavioral effects (impairment of visual/hearing/executive functioning, global developmental delay and psychiatric/behavioral sequelae) and cardiovascular diseases [3,4,24]. The early birth due to PPROM with anhydramnion during the canalicular stage of the lung development between 16 th and 26 th weeks' gestation leads to pulmonary hypoplasia [4,25]. Prolonged anhydramnion after PPROM is associated with a four-fold increased risk of composite adverse outcomes, including death, BPD, severe neurological disorders, severe retinopathy, when compared to an age-adjusted control group [8,26].
The causes of the mid-trimester PPROM are multifactorial including local in ltration by bacteria with reaction of pro-in ammatory cytokines, pathologic anatomical remodeling of the amniotic membranes (contribution of MMPs), invasive procedures and fetoscopic surgery, genetic and iatrogenic factors, smoking, vaginal bleeding etc. [4,7,[27][28][29][30][31][32][33]. In ammatory mediators likely play a causative role in both disruption of fetal membrane integrity and activation of uterine contraction.
The diagnosis of PPROM is classically established by identi cation watery leakage from the cervical canal, positive swab assay for placental alpha macroglobulin-1-and indigo carmine-tests [4,34,35]. The management of the PPROM requires balancing the potential neonatal bene ts from prolongation of the pregnancy and reduction of the adverse effects of newborn immaturity with the risk of intra-amniotic infection and its consequences for the mother and infant [2,4,22,36]. Close monitoring for signs of chorioamnionitis (e.g. body temperature, CTG, CRP, leucocytes, IL-6, procalcitonine and amniotic uid examinations) is necessary to minimize the risk of neonatal and maternal complications.
Despite the fact that broad spectrum antibiotics are routinely used in the therapy of PPROM based on the guidelines' recommendations, it must be considered that the maternally applied drugs only hardly reach the place of bacterial colonization [4,19,22]. The amniotic membranes and the umbilical cord do not have an effective capillary net which would supply the surfaces with the antibiotics from the maternal circulation [37]. The placenta is a selective barrier for foreign substances and the bioavailability among antibiotics is very low for most antibiotic agents [4,[38][39][40][41][42][43]. On the other hand, the healthy fetus without any infection after PPROM probably does not need any longtime trans-placental antibiotic treatment leading to possible change of fetal programming [4,12,13,44].
The anatomical conditions result in the inability to eradicate the colonization of bacteria in the amniotic cavity which means that infectious complications remain a major problem after PPROM. De Santis et al. found out that the patients with PPROM did not appear to demonstrate any bene t from this repetitive amniotic uid (AF) replacement (250 mL per intervention) as measured by post procedure because uid loss occurred within 6 hours of instillation [19]. Locatelli et al. found that the serial amnioinfusions could improve the neonatal outcome primarily by prolonging latency. The patients with PPROM and oligohydramnios had a signi cantly shorter interval to delivery, lower neonatal survival of 20%, 62% of pulmonary hypoplasia and 60% rate of neurological handicap [16].
Roberts et al. concluded that serial transabdominal amnioinfusions showed no signi cant difference in the outcome of maternal and perinatal outcome. The perinatal mortality was 19/28 vs 19/28 [17,18]. The positive fetal survival effect of serial amnioinfusions was unfortunately counterbalanced by an increased risk of neonatal death: 14 neonates died in AI's group vs. 9 in control group. It is possible that the use of saline solution for AI was inappropriate due to the large deviations from normal human amniotic uid [45].
Thus, accumulation of sodium and chloride would disturb the sodium potassium pump located in plasma membrane of human cells and could be in uencing organs' performance, e.g. cardiac, lung and brain. These adverse effects of the instillation uid could explain the high mortality rate after AI in this study [4,17,18].
Tchirikov 2018 has summarized literature between 2000 and 2017 on PPROM between 18 and 28 weeks [4]. The positive effect of continuous pump-regulated amnioinfusion of 100 ml/h, 2400 ml/d ("Amnion Flush Method") through the subcutaneously implanted perinatal catheter system with an anker with Amnion Flush Solution in "classic PPROM" less than 28/0 weeks' gestation showing promise since 2008.
However, two groups from Japan in recently published retrospective studies (2020) were not able improve the neonatal outcome using the low rate amnioinfusion with Ringer's lactate solution (40 ml/h) [14,15]. The authors mentioned the high frequency of catheter dislocation (60%) and lack of pump infusion in many cases [14,15]. The missed positive effect of the prolongation of the PPROM-delivery interval by two weeks onto healthy neonatal survival in these studies could be explained by "Salzgurken-Effekt" of offlabel used saline solution for a long time with a very high deviation of many parameters compared to human amniotic uid (Table 1) [4].
Fetal skin in second trimester is still very thin and permeable. It is a matter of common knowledge that the fetus swallows and inspirates/expirates amniotic uid. Gilbert and Brace published that the fetus swallows 200-250 ml/kg/day amniotic uid [46] . Continuous amnioinfusion with normal saline solution signi cantly increased plasma Na + and Clconcentrations in fetal sheep [47] . The amniotic uid is a very complex hypoosmotic solution with an alkaline pH, low concentration of the elements Cl -, K + and Na + , presence of trace elements, surfactants and many other substances [48] . The change of physiological fetal surroundings for a long period using amnioinfusion with simple electrolyte solutions could destroy in all probabilities the fetal programming. Hypernatremia nduced osmotic demyelization disorders, central pontine and extrapontine myelinolysis are well known [49] . Chhabra et al. described the extrapontine myelinolysis induced by hypernatremia [50] . The combination of very immature brain-blood barrier of the fetus with the permeable skin and swallowing of relative large amount of electrolyte solution deteriorates the uctuations of the NaCl concentration of the fetal brain. This could explain the high neonatal mortality rate after serial amnioinfusion in the Roberts et al study and very moderate success of continuous amnioinfusion using Ringer's lactate solution (inappropriate to human amniotic uid) described in the retrospective Japanese studies [14,15,17,18] .
Summarized, the standard treatment options of the classic mid-trimester PPROM with oligoanhydramnion including RDS prophylaxis and broad spectrum antibiotics hardly improve the neonatal outcome. The extension of the standard therapy by continuous amnioinfusion -"Amnion-Flush" method could improve the neonatal survival without major morbidities. In 2020, the German Federal Ministry of Education and Research funded this prospective multicentre randomized study investigating the effect of continuous amnioinfusion in patients with classic PPROM with oligo/ anhydramnion between 22/0 (20/0) -26/0 WG with Amnion Flush Solution (100 ml/h, 2400 ml/d) throughout transcutaneously placed perinatal catheter with an anchor system compared with the standard PPROM therapy (ClinicalTrials.gov: NTC04696003). Figure 1 German tertian perinatal centers taking part in the TRIAL ClinicalTrials.gov: NTC04696003