Do we need to reconsider our antibiotic regimen for preterm premature rupture of membranes? Pathogens and perinatal outcome in PPROM

Abstract

Background

The antibiotic regimen currently recommended in most guidelines in case of preterm premature rupture of membranes (PPROM) with an aminopenicillin and a macrolide is mainly based on studies performed more than 25 years ago. In view of the changing pathogens in neonatal and perinatal infectious morbidity and their increasing aminopenicillin-resistance in recent decades, the question arises as to whether these data are still applicable to our current situation and whether an adjustment of our antibiotic standard regimen with amoxicillin and azithromycin is necessary.

Methods

We retrospectively analyzed the pathogens and the perinatal outcome, particularly the perinatal and neonatal infectious morbidity, in 293 pregnancies with PPROM < 37 + 0 weeks and delivery in our department from 01/2020 to 12/2023.

Results

In PPROM < 34+0 weeks, we found neonatal infection (defined by clinical infection criteria) in 14.5%, neonatal early-onset sepsis (defined by clinical infection criteria plus invasive pathogen detection) in 3.6% and one case of maternal sepsis. The causative pathogens of the 6 neonatal early-onset sepsis cases were in 3 cases (50%) E.coli (2/3 amoxicillin-resistant) and in one case each Enterococcus faecalis, Streptococcus mitis and Haemophilus influenzae. The most common pathogens in the vaginal swabs were Ureaplasma species (25%). The most common pathogens in the placental swabs were, in general, coagulase-negative Staphylococci (21.2%), followed by E.coli (10.2%) and Ureaplasma species (8.8%) and, in case of histological confirmed chorioamnionitis, Ureaplasma species (21.4%), followed by E. coli (17.9%) and coagulase-negative Staphylococci (14.3%).

In PPROM ≥ 34+0 weeks, we found neonatal infection in 2.6% and there was no case of neonatal or maternal sepsis.

Conclusions

E.coli and coagulase-negative Staphylococci, both predominantly resistant to aminopenicillins, as well as Ureaplasma species seem to be the currently most relevant pathogens in PPROM. Therefore, azithromycin should be administered, but amoxicillin no longer appears to be the optimal regimen. Further research comparing different antibiotic regimens with better coverage of the currently relevant pathogens such as co-amoxiclav (amoxicillin + clavulanic acid) or cephalosporins is needed.

Background

Preterm premature rupture of membranes (PPROM) is one the most prevalent complications in pregnancy, occurring in up to 3% of all pregnant women and causing approximately one-third of preterm births [1]. PPROM is highly associated with maternal and neonatal infectious morbidity, such as chorioamnionitis, endometritis and septicemia, occurring in approximately one-third of the PPROM-cases and leading to potentially serious sequalae, particularly in preterm neonates (increased risk of neurodevelopmental impairment, perinatal death, respiratory distress syndrome, intraventricular hemorrhage, necrotizing enterocolitis and sepsis) [2–6]. The administration of antibiotics in case of PPROM is standard practice. In a systematic review of 22 randomized controlled trials involving more than 6800 women, antibiotic therapy was associated with significant reduction in intra-amniotic infection, neonatal infection, infants born within 48 h, neonatal oxygen therapy and use of surfactant [7]. However, the evidence is insufficient to determine the optimal antibiotic regimen [8]. Mainly based on the Mercer protocol [2] and the ORACLE I trial [3], guidelines mostly recommend the administration of a combination of an aminopenicillin and a macrolide and the avoidance of co-amoxiclav (amoxicillin + clavulanic acid) due to its possible association with necrotizing enterocolitis (NEC) in preterm neonates. According to the AWMF-guideline [9], the guideline of the German-speaking countries, the antibiotic regimen in our department, a tertiary perinatal center, consists of oral administration of amoxicillin for a total of 7 days and oral single-dose azithromycin. Therapy is adjusted based on pathogen detection in the vaginal swab.

To determine the optimal prophylactic antibiotic regimen for PPROM, it is necessary to identify the current local spectrum of pathogens that commonly cause perinatal and neonatal infectious morbidity and to assess their susceptibility to the standard antibiotic regimen. In recent years, both in the German-speaking countries as well as at a global scale, we see a change in the pathogen spectrum with an increase in Gram-negative bacteria, in particular E. coli, as the causative agent of neonatal early-onset sepsis (EOS) and a general increase in antibiotic resistance of relevant pathogens to aminopenicillins. In a nationwide prospective cohort study with inclusion of all neonates admitted to a tertiary care neonatal intensive care unit in Switzerland between 2011 and 2015, the most common causative pathogens of neonatal EOS were overall (term and preterm neonates) group B Streptococci (38%) and E. coli (23%). In preterm neonates, E. coli was the most common pathogen (37%), group B Streptococci were responsible for 27% of the EOS-cases [10]. According to the data from the German Neonatal Network from 2009 to 2017, the most common causative pathogens of neonatal EOS in preterm infants < 32 weeks of gestation and with a birth weight < 1500 g were E. coli (34.6%) followed by coagulase-negative Staphylococci (CoNS) (23.9%), group B Streptococci (16.1%) and Enterococci (11.9%) [11]. In Central Switzerland the current resistance rates of E. coli and CoNS to aminopenicillins are 49% and 86%, respectively [12].

The aim of this study was to determine whether our current standard antibiotic regimen with amoxicillin and azithromycin is appropriate to cover the current spectrum of relevant pathogens in case of PPROM.

Methods

We retrospectively analyzed the spectrum of pathogens and the perinatal outcome, particularly the perinatal and neonatal infectious morbidity, in all pregnancies complicated by PPROM and delivery in our department between January 1, 2020, and December 31, 2023. All patients who signed the hospital`s general consent were included in this analysis.

Our department is a tertiary perinatal center and the Center hospital of Central Switzerland with more than 2000 births (270–280 preterm births) per year and around 770 hospitalized neonates (270–290 preterm neonates) in the neonatology department per year. In case of diagnosis of PPROM, cervical-vaginal swabs for pathogen detection including general bacteria, fungi, Chlamydia, Mycoplasma and Ureplasma species (= Ureaplasma urealyticum and Ureaplasma parvum), a rectal-vaginal swab for group B Streptococci and a urine probe are taken, and blood sampling (infectious parameter: white blood cell count and c-reactive protein) is performed. Additionally, the fetus is monitored with ultrasound and, in case of PPROM after fetal viability (≥ 23 + 0–24 + 0 weeks of gestation), with cardiotocography. In case of PPROM after fetal viability, the woman is hospitalized, an antibiotic therapy with single-dose azithromycin 1 g orally and amoxicillin 1 g orally every 8 h for 48 h followed by 500 mg orally every 8 h for 5 days is administered and, in case of PPROM < 34 + 0 weeks of gestation, lung maturation (+/- tocolysis or 48 h) is performed. Depending on pathogen detection in the swabs, antibiotic therapy is adjusted. During hospitalization, the women and the fetus are closely monitored with regular blood samplings and cardiotocography. In case of uncomplicated clinical course, the patient is subsequently considered for outpatient management with planned check-ups two times per week. In case of PPROM before 23 + 0–24 + 0 weeks of gestation, an outpatient observational management is used until fetal viability is achieved. The baby will be delivered at any time in case of suspected chorioamnionitis, other complications or, after completion of lung maturation or gestational age ≥ 34 + 0 weeks of gestation, in case of spontaneous onset of labor. If there is no spontaneous onset of labor, induction of labor or, if indicated, cesarean section is performed at the latest at 37 + 0 weeks of gestation.

The primary outcome of our study was defined as the occurrence of neonatal infection, neonatal early-onset sepsis (EOS) or maternal sepsis with its pathogens and their antibiotic susceptibility. Neonatal infection was defined by clinical infection criteria without invasive pathogen detection (two or more of the following criteria: fever > 38 °C, hypothermia < 36.5 °C, temperature instability, bradycardia < 80/min, tachycardia > 200/min, poor capillary refill > 2 s, increased frequency of apnea > 20 s, metabolic acidosis with BE < −10 mmol/l, hyperglycemia > 140 mg/dl, respiratory distress, lethargy, irritability, hypotonia, seizures, poor feeding, oxygen requirement, elevated white cell count, high immature-to-total neutrophil ratio, elevated C-reactive protein). Sepsis was defined by clinical signs of infection plus invasive pathogen detection in the blood cultures, in case of neonatal EOS within the first 72 h of life. Secondary outcomes included the pathogen spectrum in vaginal and placental swabs and the latency period between PPROM and birth. In addition, the following general perinatal outcome parameters were evaluated: delivery mode, induction of labor, chorioamnionitis, completion of lung maturation prior to birth, pathological CTG, meconium-stained amniotic fluid, placental abruption, prolapse of the umbilical cord, blood loss, birth weight percentile/SGA-fetus, gestational age at birth/preterm birth, arterial pH, 5-minute APGAR, admission to neonatal intermediate care and/or intensive care unit (NICU), intubation, CPAP-therapy, neonatal late-onset sepsis (LOS), postnatal antibiotic therapy in the neonate, maternal and neonatal death, maternal and neonatal inpatient stay. Chorioamnionitis was suspected based on clinical criteria (maternal fever ≥ 38 °C plus one of the following: maternal leukocytosis (white blood cell count > 15,000 cells/mm³), fetal tachycardia (> 160 beats per minute) and/or purulent vaginal discharge) or was histologically confirmed. Neonatal LOS was defined as sepsis with occurrence after 72 h of life.

Regarding our microbiology techniques: After collection (in case of the placental swabs under most sterile conditions as possible), the vaginal and placental swabs were cultured on different culture media and the evaluation of the pathogens were carried out semiquantitative. To test for Chlamydia we used polymerase chain reaction (PCR) method. Regarding the Ureaplasma species and Mycoplasma hominis, we were able to differentiate between Ureaplasma species und Mycoplasma hominis with the “Mycoplasma Duo Kit” (company Bio-Rad), but we could not differentiate between Ureaplasma urealyticum and Ureplasma parvum. In general, all vaginal and placental swabs and all blood cultures were tested with different culture media for both aerobes and anaerobes. To assess the antibiotic susceptibility of the pathogens we used microdilution methods (BD Phoenix™ M50 system and Sensititre Vizion Digital MIC Viewing system), Kirby-Bauer disk diffusion tests and MIC assays. If pathogens were detected in blood cultures, antibiotic susceptibility testing was performed automatically; if pathogens were detected in vaginal and placental swabs, this testing was only performed upon special request.

The database was built in REDCap and statistical analysis was conducted using SPSS software (IBM SPSS Statistics version 29.0.2.0). Continuous variables are presented as mean and standard deviation or as median and range. Categorical variables are presented as frequencies and percentages.

The study was approved by the local ethics committee (Project-ID 2023 − 01319, date of approval July 14, 2023) and was performed in accordance with the Declaration of Helsinki.

Results

Of the 417 pregnant patients with PPROM and delivery in our department between January 1, 2020, and December 31, 2023, we included a total of 293 patients. 124 patients with unsigned general consent were excluded.

Of the 293 included pregnancies, in 156 (53.2%) PPROM occurred < 34 + 0 weeks of gestation and in 137 (46.8%) between 34 + 0 and 36 + 6 weeks of gestation. Due to the different management with administration of corticosteroids for lung maturation (+/- tocolysis or 48 h) in case of PPROM < 34 + 0 but not in case of PPROM ≥ 34 + 0 weeks, the outcome of both groups was assessed separately.

The baseline characteristics of all included patients are summarized in Table 1.

Table 1.

Demographic and clinical baseline of included patients

	PPROM < 34+0 Mean (SD) or N (%)	PPROM > 34+0 Mean (SD) or N (%)
Maternal age (years)	32.8 (+/- 4.7)	32.8 (+/- 4.1)
BMI
Normal	125 (80.1%)	79 57.7%)
Underweight	0	1 (0.7%)
Overweight	14 (9%)	20 14.6%)
Obese	17 (10.9%)	37 (27%)
Grade 1	10 (58.8%)	26 0.3%)
Grade 2	4 (23.5%)	6 (16.2%)
Grade 3	3 (17.6%)	5 (13.5%)
Parity
0	95 (60.9%)	90 65.7%)
1	41 (26.3%)	33 (24.1%)
> 1	20 (12.8%)	14 (10.2%)
History of PPROM	7 of N = 61 (11.5%)	2 of N = 47 (4.3%)
History of preterm birth	15 of N = 61 (24.6%)	8 of N = 47 (17%)
Conception method
Spontaneous	146 (93.6%)	1304.9%)
Stimulation	2 (1.3%)	2 (1.5%)
IVF/ICSI	8 (5.1%)	5 (3.6%)
Egg donation	0	0
Singleton/Multiples
Singleton	137 (87.8%)	1186.1%)
Twins	18 (11.5%)	19 13.9%)
Triplets	1 (0.7%)	0
Smoking in pregnancy	3 (1.9%)	3 (2.2%)
Pre-existing diabetes	0	0
Gestational diabetes	18 (11.5%)	9 (6.6%)
Diet	9 (50%)	2 (22.2%)
Insulin	9 (50%)	7 (77.8%)
Gestational hypertension	1 (0.6%)	2 (1.5%)
Preeclampsia	1 (0.6%)	0

 Data are reported as mean and SD or number and percentages where appropriate

PPROM < 34+0: N = 156 women, PPROM > 34+0: N = 137 women

The general course of pregnancy with fetal and neonatal survival and primary outcome is presented in Fig. 1. In PPROM < 24 + 0 weeks, there was a 72% survival rate in live-born neonates in whom neonatal intensive care was instituted. The neonates, who died within the first month of life, were born between 23 + 2 and 24 + 2 weeks of gestation. In PPROM ≥ 24 + 0–33 + 6 weeks, the neonate survived in 99.3% of the cases and there was one case of neonatal death. In this case, the neonate was diagnosed with esophageal atresia and a genetic disorder was strongly suspected. In the cohort with PPROM ≥ 34 + 0 weeks of gestation, 99.4% of the neonates survived and there was also one case of neonatal death. In this case, the neonate suffered from a severe diaphragmatic hernia (see Fig. 1).

Fig. 1.

Flow chart with primary outcome

Regarding the primary outcome, the occurrence of neonatal infection, neonatal EOS or maternal sepsis, there were 24 cases (14.5%) of neonatal infection, six cases (3.6%) of neonatal EOS and one case (0.7%) of maternal sepsis in the cohort with PPROM < 34 + 0 weeks of gestation and live birth. In case of neonatal infection the two most common pathogens detected in the placental swab were E. coli and CoNS in 20.8% of cases each (see Table 2). The causative pathogen in the one case of maternal sepsis and in three cases (50%) of EOS was E. coli (in 2/3 resistant to aminopenicillins, in 1/3 resistant to co-amoxiclav and in all cases susceptible to cephalosporins). The causative pathogens in the other three cases of EOS were Enterococcus faecalis (susceptible to aminopenicillins), Streptococcus mitis (susceptible to aminopenicillins) and Haemophilus influenzae (susceptible to aminopenicillins), respectively. One of the neonates with EOS caused by E. coli died. This neonate was born at 23 + 4 weeks of gestation. The other neonates with EOS survived. In the cohort with PPROM ≥ 34 + 0 weeks of gestation, there were four cases (2.6%) of neonatal infection, but no case of neonatal EOS or maternal sepsis (see Fig. 1; Table 2).

Table 2.

Primary outcome: occurrence of neonatal infection, neonatal early-onset sepsis or maternal sepsis

	PPROM < 34 + 0 and live birth N (%)	PPROM ≥ 34 + 0 N (%)
Neonatal infection	24 (14.5%) Pathogens in the placental swabs*: - No pathogen detected: 6 (25%) - Coagulase-negative Staphylococci: 5 (20.8%) - E. coli: 5 (20.8%) - Streptococcus mitis: 3 (12.5%) - Streptococcus oralis: 2 (8.3%) - Enterococcus faecalis: 2 (8.3%) - Gardnerella vaginalis: 2 (8.3%) - Ureaplasma species: 1 (4.2%) - Haemophilus parainfluenzae: 1 (4.2%) - Kytococcus schroeteri: 1 (4.2%) - Klebsiella oxytoca: 1 (4.2%) - Klebsiella pneumoniae: 1 (4.2%) - Corynebacterium amycolatum: 1 (4.2%) - Gleimia europaea: 1 (4.2%) - Bacteroides fragilis: 1 (4.2%)	4 (2.6%) Pathogens in the placental swabs**: - No pathogen detected: 1 (25%) - Coagulase-negative Staphylococci: 1 (25%) - Ureaplasma species: 1 (25%) - Streptococcus anginosus: 1 (25%) - Bacillus circulans: 1 (25%) - Capnocytophaga sputigena: 1 (25%)
Neonatal early-onset sepsis (EOS)	6 (3.6%) Case 1: pathogen: E. coli (resistant to aminopenicillins and co-amoxiclav, susceptible to cephalosporins); PPROM at 23 + 2, birth at 23 + 4; sex: female; birth weight: 570 g; lung maturation not completed; delivery mode: c-section (indication: chorioamnionitis), neonatal death at 1 st day of life; vaginal swab: no pathogen, placental swab: E. coli; chorioamnionitis (clinically and histologically) Case 2: pathogen: E. coli (resistant to aminopenicillins, susceptible to co-amoxiclav and cephalosporins); PPROM at 33 + 2, birth at 34 + 3; sex: male; birth weight: 2500 g; lung maturation completed; delivery mode: vaginal spontaneous; vaginal swab: Ureaplasma species, Candida albicans, placental swab: E. coli; chorioamnionitis (clinically, no histological examination of the placenta performed) Case 3: pathogen: E. coli (susceptible to aminopenicillins); PPROM at 26 + 4, birth at 29 + 1; sex: female; birth weight: 1500 g; lung maturation completed; delivery mode: c-section (indication: chorioamnionitis); vaginal swab: no pathogen, placental swab: E. coli; chorioamnionitis (clinically and histologically); additional maternal E. coli sepsis Case 4: pathogen: Enterococcus faecalis (susceptible to aminopenicillins); PPROM at 26 + 5, birth at 27 + 0; sex: female; birth weight: 1000 g; lung maturation completed; delivery mode: c-section (indication: history of c-section with longitudinal uterotomy); vaginal swab: no pathogen, placental swab: Enterococcus faecalis; chorioamnionitis (clinically, no histological examination of the placenta performed) Case 5: pathogen: Streptococcus mitis (susceptible to aminopenicillins); PPROM at 30 + 3, birth at 33 + 2; sex: male; birth weight: 2380 g; lung maturation completed; delivery mode: c-section (indication: history of c-section); vaginal swab: no pathogen, placental swab: Streptococcus mitis; chorioamnionitis (clinically, no histological examination of the placenta performed) Case 6: pathogen: Haemophilus influenzae (susceptible to aminopenicillins); PPROM at 28 + 6, birth at 30 + 0; sex: male; birth weight: 1370 g; lung maturation completed; delivery mode: c-section (indication: chorioamnionitis); vaginal swab: Ureaplasma species, placental swab: Haemophilus influenzae; chorioamnionitis (clinically and histologically)	0
Maternal sepsis	1 (0.7%) Pathogen: E. coli (susceptible to aminopenicillins), neonate with EOS (= case 3)	0

 PPROM < 34+0 and live birth: N =146 women, 166 neonates, PPROM > 34+0: N = 137 women, 156 neonates

*In 1 out of 24 cases, no placental swab was available and in 5 out of 24 cases, more than one pathogen was detected

**In 1 out of 4 cases, more than one pathogen was detected

In the cohort with PPROM < 34 + 0 weeks of gestation, the mean gestational age at PPROM was 28 + 4 weeks (SD +/- 35 d) and at birth 31 + 0 weeks (SD +/- 28.8 d). In the cohort with PPROM ≥ 34 + 0 weeks of gestation, both were 35 + 4 weeks (SD +/- 6.1 d and SD +/- 5.8 d, respectively). The earlier the PPROM occurred, the longer the mother remained pregnant. The latency periods between PPROM and birth divided into different subgroups based on gestational age are presented in Fig. 2.

Fig. 2.

Secondary outcome: latency period between PPROM and birth

Vaginal swabs at the diagnosis of PPROM were available in 94.9% (148 out of 156) of the cases with PPROM < 34 + 0 weeks of gestation and in 38.7% (53 out of 137) of the cases with PPROM ≥ 34 + 0 weeks of gestation. The by far most commonly found pathogens in the vaginal swabs were Ureaplasma species (< 34 + 0: in 25%, ≥ 34 + 0: in 20.8%). E. coli was detected in three cases (2%) in the cohort with PPROM < 34 + 0 weeks and in one case (1.9%) in the cohort with PPROM ≥ 34 + 0 weeks. Swabs from the placenta were available in 87.8% (137 out of 156) of the cases with PPROM < 34 + 0 weeks and in 61.3% (84 out of 137) of the cases with PPROM ≥ 34 + 0 weeks. The most common pathogens were CoNS (< 34 + 0: 21.2%, ≥ 34 + 0: 17.9%), followed by E. coli (< 34 + 0: 10.2%, ≥ 34 + 0: 7.1%) and Ureaplasma species (< 34 + 0: 8.8%, ≥ 34 + 0: 7.1%). In two of the three cases with detection of E. coli in the vaginal swab, the women additionally received co-amoxiclav, since the requested antibiotic susceptibility testing showed aminopenicillin-resistance and susceptibility to co-amoxiclav or cephalosporins. In these two cases, E. coli could not be detected in the placental swab. In contrast, in the third case, in which antibiotic susceptibility testing was not performed and therapy was not adjusted, E. coli was detected in the placental swab and the neonate suffered from an infection. The requested antibiotic susceptibility testing of the E. coli in the placental swab showed a resistance to aminopenicillins and co-amoxiclav but susceptibility to cephalosporins. All pathogens detected in the vaginal and placental swabs are summarized in Table 3.

Table 3.

Secondary outcome: pathogens in the vaginal and placental swabs

	PPROM < 34 + 0 (including late miscarriage) N (%)	PPROM ≥ 34 + 0 N (%)
Vaginal swabs*	Ureaplasma species: 37 (25%) Group B Streptococci: 14 (9.5%) Candida albicans: 11 (7.4%) Gardnerella vaginalis: 9 (6.1%) Mycoplasma hominis: 5 (3.4%) E. coli: 3 (2%) Klebsiella pneumoniae: 1 (0.7%) Chlamydia: 1 (0.7%) Staphylococcus aureus: 1 (0.7%)	Ureaplasma species: 11 (20.8%) Group B Streptococci: 10 (18.9%) Candida albicans: 4 (7.5%) Gardnerella vaginalis: 2 (3.8%) Streptococcus pyogenes: 1 (1.9%) E. coli: 1 (1.9%)
Placental swabs**	In general: - Coagulase-negative Staphylococci: 29 (21.2%) - E. coli: 14 (10.2%) - Ureaplasma species: 12 (8.8%) - Enterococci species: 10 (7.3%) - Enterococcus faecalis: 9 - Enterococcus faecium: 1 - Streptococcus mitis: 10 (7.3%) - Streptococcus anginosus: 8 (5.8%) - Prevotella bivia: 8 (5.8%) - Klebsiella species: 5 (3.6%) - Klebsiella pneumoniae: 3 - Klebsiella variicola: 1 - Klebsiella oxytoca: 1 - Corynebacterium species: 5 (3.6%) - Lactobacillus species: 4 (2.9%) - Cutibacterium acnes: 4 (2.9%) - Finegoldia magna: 4 (2.9%) - Gardnerella vaginalis: 3 (2.2%) - Peptoniphilus species: 3 (2.2%) - Mixed anaerobic flora: 3 (2.2%) - Candida albicans: 2 (1.5%) - Mycoplasma hominis: 1 (0.7%) - Streptococcus oralis: 1 (0.7%) - Group B Streptococci: 1 (0.7%) - Haemophilus influenzae: 1 (0.7%) - Haemophilus parainfluenzae: 1 (0.7%) - Brevibacterium paucivorans: 1 (0.7%) - Bacteroides fragilis: 1 (0.7%)- A ggregatibacter aphrophilus: 1 (0.7%)- Schaalia odontolytica: 1 (0.7%)- Kytococcus schroeteri: 1 (0.7%)- Gleimia europaea: 1 (0.7%)- Bifidobacterium longum: 1 (0.7%)- Globicatella sanguines: 1 (0.7%)	In general: -Coagulase-negative Staphylococci: 15 (17.9%) - E. coli: 6 (7.1%) - Ureaplasma species: 6 (7.1%) - Enterococci species: 3 (3.6%) - Enterococcus faecalis: 2 - Enterococcus faecium: 1 - Lactobacillus species: 3 (3.6%) - Corynebacterium species: 2 (2.4%) - Streptococcus anginosus: 2 (2.4%) - Gardnerella vaginalis: 2 (2.4%) - Phocaeciola vulgatus: 2 (2.4%) - Streptococcus sanguinis: 1 (1.2%) - Bacteroides vulgaris: 1 (1.2%) - Bifidobacterium pseudocatenulatum: 1 (1.2%) - Cutibacterium acnes: 1 (1.2%) - Prevotella bivia: 1 (1.2%) - Finegoldia magna: 1 (1.2%) - Bacteroides uniformis: 1 (1.2%) - Bacillus circulans: 1 (1.2%) - Capnocytophaga sputigena: 1 (1.2%) In case of neonatal infection (4 cases): *** - No pathogen detected: 1 (25%) - Coagulase-negative Staphylococci: 1 (25%) - Ureaplasma species: 1 (25%) - Streptococcus anginosus: 1 (25%) - Bacillus circulans: 1 (25%) - Capnocytophaga sputigena: 1 (25%)
	In case of neonatal infection (24 cases): *** - No pathogen detected: 6 (25%) - Coagulase-negative Staphylococci: 5 (20.8%) - E. coli: 5 (20.8%) - Streptococcus mitis: 3 (12.5%) - Streptococcus oralis: 2 (8.3%) - Enterococcus faecalis: 2 (8.3%) - Gardnerella vaginalis: 2 (8.3%) - Ureaplasma species: 1 (4.2%) - Haemophilus parainfluenzae: 1 (4.2%) - Kytococcus schroeteri: 1 (4.2%) - Klebsiella oxytoca: 1 (4.2%) - Klebsiella pneumoniae: 1 (4.2%)- C orynebacterium amycolatum: 1 (4.2%)- Gleimia europaea: 1 (4.2%)- Bacteroides fragilis: 1 (4.2%)
	In case of neonatal early-onset sepsis (6 cases): - E. coli: 3 (50%) - Enterococcus faecalis: 1 (16.7%) - Streptococcus mitis: 1 (16.7%) - Haemophilus influenzae: 1 (16.7%) In case of histologically confirmed chorioamnionitis (28 cases): **** - No pathogen detected: 5 (17.9%) - Ureaplasma species: 6 (21.4%) - E. coli: 5 (17.9%) - Coagulase-negative Staphylococci: 4 (14.3%) - Prevotella bivia: 4 (14.3%) - Streptococcus anginosus: 3 (10.7%) - Streptoccous mitis: - Enterococcus faecalis: 2 (7.1%) - Peptoniphilus species: 2 (7.1%) - Klebsiella pneumoniae: 1 (3.6%) - Haemophilus influenzae: 1 (3.6%) - Gardnerella vaginalis: 1 (3.6%) - Bacteroides fragilis: 1 (3.6%)

PPROM < 34 + 0 including late miscarriage: N = 156 women, PPROM ≥ 34 + 0: N = 137 women

*PPROM < 34 + 0 including late miscarriage: In 148 out of 156 cases (94.9%), vaginal swab was available; PPROM ≥ 34 + 0: In 53 out of 137 cases (38.7%), vaginal swab was available

**PPROM < 34 + 0 including late miscarriage: In 137 out of 156 cases (87.8%), placental swab was available; PPROM ≥ 34 + 0: In 84 out of 137 cases (61.3%), placental swab was available

***PPROM < 34 + 0: In 1 out of 24 cases, no placental swab was available and in 5 out of 24 cases, more than one pathogen was detected; PPROM ≥ 34 + 0: In 1 out of 4 cases, more than one pathogen was detected

**** In 8 out of 28 cases, more than one pathogen was detected

Histological examination of the placenta was performed in 33.6% (49 out of 146) of the cases with PPROM < 34 + 0 weeks of gestation and live birth and in 70% (7 out of 10) in the subgroup with late miscarriage. Among the total of 28 cases of histologically confirmed chorioamnionitis in these two groups, the most common pathogens detected in the placental swabs were Ureaplasma species in six cases (21.4%), E. coli in five cases (17.9%) and either CoNS or the anaerobic bacteria Prevotella bivia in four cases (14.3%) each (see Table 3).

In addition to the six cases with neonatal EOS in PPROM < 34 + 0 weeks of gestation, there were six cases (3.6%) of neonatal late-onset sepsis (LOS) in this group. The causative pathogens were group B Streptococci (susceptible to aminopenicillins) in 50% of the cases and CoNS (all resistant to aminopenicillins and co-amoxiclav, cephalosporins not tested) with in one case additional E. coli (resistant to aminopenicillins and co-amoxiclav, susceptible to cephalosporins) in the other 50% of the cases. In case of PPROM ≥ 34 + 0 weeks, CoNS (resistant to aminopenicillins and co-amoxiclav, cephalosporins not tested) was detected in the one case of LOS as the causative pathogen. Overall, 78.3% of the neonates in case of PPROM < 34 + 0 weeks were treated with antibiotics postnatally due to suspected or confirmed infection/sepsis or with a prophylactic approach. In case of PPROM ≥ 34 + 0 weeks, 14.7% of the neonates received antibiotics postnatally. All general perinatal outcome parameters are summarized in Table 4.

Table 4.

Perinatal outcome

	PPROM < 34+0 and live birth Mean (SD) or N (%)	PPROM > 34+0 Mean (SD) or N (%)
Delivery mode
Vaginal spontaneous	61 (41.8%)	65 47.4%)
Vaginal assisted	5 (3.4%)	14 10.2%)
Cesarean section after trial of vaginal birth	22 (15.1%)	12 8.8%)
Planned cesarean section	58 (39.7%)	46 (33.6%)
Indication for assisted vaginal birth or cesarean section after trial of vaginal birth
Pathological CTG	12 (44.5%)	18 (69.2%)
Suspected chorioamnionitis (clinical criteria)	7 (25.9%)	1 (3.8%)
Vaginal bleeding/placental abruption	5 (18.5%)	2 (7.7%)
Failed induction of labor	2 (7.4%)	1 (3.8%)
Prolapse of umbilical cord	1 (3.7%)	1 (3.8%)
Arrest of labor	-	3 (11.5%)
Induction of labor in case of planned vaginal delivery	20 (22.7%)	44 (48.4%)
Lung maturation completed prior to birth	123 (84.2%)	-
Pathological CTG	21 (14.4%)	21 (15.3%)
Meconium-stained amniotic fluid	1 (0.7%)	1 (0.7%)
Placental abruption	11 (7.5%)	2 (1.5%)
Prolapse of the umbilical cord	1 (3.7%)	1 (3.8%)
Mean blood loss in ml (SD)	479 (+/- 234)	465 (+/- 242)
Suspected chorioamnionitis (clinical criteria)	28 (19.2%) In case of PPROM <24+0 and late miscarriage: 6 out of 10 cases	3 (2.2%)
Histologically confirmed chorioamnionitis	22 (15.1%)* In case of PPROM <24+0 and late miscarriage: 6 out of 10 cases (and out of 7 performed histological examinations of the placenta)	-
Mean prenatal inpatient stay in days (SD)	7.8 (+/- 10.9)	0.8 (+/- 1.2)
Mean postnatal inpatient stay in days (SD)	3.6 (+/- 1)	3.7 (+/- 0.9)
Fetal sex
Male	94 (56.6%)	83 (53.2%)
Female	72 (43.4%)	73 (46.8%)
Birth weight percentile (SD)**	51.4 (+/- 21.7)	44.4 (+/- 25.9)
SGA <3. percentile	0	1 (0.6%)
SGA <10. percentile	2 (1.2%)	9 (5.8%)
Preterm birth <28+0 weeks	28 (16.9%)	-
Preterm birth <32+0 weeks	66 (39.8%)	-
Preterm birth <34+0 weeks	126 (75.9%)	-
Preterm birth <37+0 weeks	165 (99.4%)	156 (100%)
Arterial pH <7.1	5 of N = 160 (3.1%)	2 of N = 148 (1.4%)
5 min APGAR <7	20 of N = 163 (12.3%)	6 (3.8%)
Intubation	53 (31.9%)	7 (4.5%)
CPAP therapy	124 (74.7%)	38 (24.4%)
Neonatal late-onset sepsis (LOS) (occurrence after 72 hours of life)	6 (3.6%) Pathogen: 3x group B Streptococci, 3x coagulase-negative Staphylococci (+ 1x additional E. coli)	1 (0.6%) Pathogen: coagulase-negative Staphylococci
Postnatal antibiotic therapy	130 (78.3%)	23 (14.7%)
Neonatal death	8 (4.8%)***	1 (0.6%)***
Mean neonatal inpatient stay at NICU**** in days (SD)	41.2 (+/- 29.1)	18.6 (+/- 17.4)

Data are reported as mean and SD or number and percentages where appropriate. PPROM < 34 + 0 and live birth: N = 146 women, 166 neonates, PPROM ≥ 34 + 0: N = 137 women, 156 neonates

*Only in 49 out of 146 cases histological examination of the placenta was performed, chorioamnionitis was found in 44.9% of the performed examinations and 15.1% of the overall cases, respectively

**To calculate the birth weight percentile Fenton growth charts 2013 were used

***Occurrence of 2 neonatal deaths in case of PPROM > 24 + 0 weeks, in these cases the neonate had a congenital anomaly (case 1: esophageal atresia, case 2: severe diaphragmatic hernia)

****NICU = neonatal intensive care unit and/or neonatal intermediate care unit

Discussion

Recently, E. coli, CoNS and Ureaplasma species seem to be the most relevant pathogens in pregnancies complicated by PPROM. Predominantly aminopenicillin-resistant E. coli was the most common causative agent of neonatal EOS and maternal sepsis and naturally mostly aminopenicillin-resistant CoNS followed by E. coli were the most frequently detected pathogens in the placental swabs in case of neonatal infection. Ureaplasma species were the most frequently detected pathogens in the vaginal swabs collected from the pregnant women at diagnosis of PPROM as well as in the placental swabs in case of chorioamnionitis (followed by E. coli and CoNS). Therefore, to cover Ureaplasma species, macrolides such as azithromycin should continue to be part of the prophylactic antibiotic regimen administered to pregnant women in case of PPROM, but aminopenicillins no longer appear to be the optimal choice.

Due to its proven positive effect on the perinatal outcome, prophylactic antibiotic therapy in case of PPROM has been standard practice worldwide for many years. However, the optimal antibiotic regimen remains unclear. In a network meta-analysis from 2020 with inclusion of 20 randomized controlled trials including more than 7100 women and 15 different antibiotic regimens, the authors concluded that several antibiotics appear to be more effective than no treatment or placebo in reducing the rate of chorioamnionitis, but none of the antibiotics is clearly and consistently superior compared to others [8]. The current recommendation of most guidelines, including the AWMF-guideline of the German-speaking countries [9] recommending a combination of an aminopenicillin and a macrolide and to avoid co-amoxiclav, is mainly based on two studies performed more than two decades ago: The first one is the Mercer trial, published in 1997 [2]. In this trial, Mercer et al. were able to demonstrate a significant reduction in perinatal morbidity and a significant prolongation of the latency period in pregnancies complicated by PPROM between 24 + 0 and 32 + 0 weeks of gestation with the administration of the “Mercer protocol” (intravenous ampicillin and erythromycin for 2 days followed by oral amoxicillin and erythromycin for 5 days) compared to placebo. The second one is the ORACLE I trial from 2001 [3], a randomized multicenter trial with inclusion of more than 4800 women with PPROM < 37 + 0 weeks of gestation, in which Kenyon et al. compared erythromycin, co-amoxiclav, a combination of both or placebo. They found the most favorable overall outcome in the group with erythromycin and, as one of the secondary outcomes, a significant increase in neonatal NEC in the groups with administration of co-amoxiclav compared to the groups with administration of erythromycin or placebo (1.8% of proven neonatal NEC in any group with co-amoxiclav versus 0.7% in any group without co-amoxiclav, p 0.0005) [3, 4]. Thus, the authors recommended to avoid co-amoxiclav in any case of preterm delivery such as PPROM, spontaneous preterm labor or cesarean section for preterm delivery, although co-amoxiclav was found to be more effective in prolonging pregnancy and reducing maternal infection compared to erythromycin or placebo in this trial [3]. In contrast, there are some subsequently performed studies, albeit of poorer scientific quality, which did not show this association [13–15]. Furthermore, some evidence suggests that an early treatment with co-amoxiclav and gentamicin in preterm neonates is negatively associated with NEC and may even be protective against it [16, 17]. However, the role of prenatal and early postnatal treatment with any type of antibiotics in the development of neonatal NEC is not yet fully understood and studies show inconsistent results [18, 19]. Today, we know that the microbiome in early life plays an important role for future health. Prenatal antibiotic treatment of the pregnant woman and early postnatal antibiotic treatment of the neonate can lead to increased antimicrobial resistance and perturbations of the developing microbiome, which is associated with the occurrence of chronic diseases in later life [20, 21]. This knowledge further emphasizes the importance of careful management of antibiotic use in the perinatal period. In contrast, however, in daily clinical practice the lack of predictive precision in current diagnostic tools, the fear to miss an evolving neonatal EOS and the importance of early antibiotics administration for an optimal outcome in EOS often leads to an overtreatment in early life. In a recently published study with analysis of antibiotic exposure and neonatal EOS in Europe, North America and Australia, the authors report that for one case of culture-proven sepsis, 58 newborns received antibiotics, and 273 antibiotic days were administered [22]. In our cohort, similar observations can be made: Almost 80% of all neonates in case of PPROM < 34 + 0 weeks received antibiotics postnatally, but only 14.5% were actually diagnosed with infection and only 3.6% with EOS. In the case of PPROM ≥ 34 + 0 weeks, almost 15% of the neonates were treated with antibiotics postnatally, but only in 2.6% of the cases infection was confirmed and there was no case of EOS. These findings suggest a huge potential to reduce postnatal antibiotic exposure in neonates born after PPROM. Furthermore, regarding the prenatal period in pregnancies complicated by PPROM, although there is no potential to directly reduce antibiotic exposure, there may be potential to optimize the prophylactic antibiotic regimen with better coverage of the relevant pathogens and consequently reduce perinatal and neonatal infectious morbidity and need for antibiotics in general.

In order to determine the optimal prophylactic antibiotic regimen for PPROM, it is necessary to identify the current local spectrum of relevant pathogens and to assess their resistance profiles. Our findings in this study with E. coli and CoNS as the most common pathogens in maternal and neonatal infection and sepsis correspond well with the global observation of the changing pathogen spectrum in perinatal and particularly neonatal infectious morbidity and the general increase in antimicrobial resistance in recent decades. While the incidence of EOS caused by group B Streptococci decreased, presumably mainly due to the implementation of universal screening and antibiotic prophylaxis, the incidence of E. coli as causative EOS pathogen and its resistance to aminopenicillins have increased and to date, E. coli is the most common pathogen in EOS in preterm neonates in Western high-income countries, while CoNS are the most common ones in LOS [10, 11, 23–31]. In contrast, in low- and middle-income countries and, partly, in Asian populations, Gram-positive bacteria such as Staphylococcus aureus and CoNS and, although to a lesser extent, the Gram-negative Klebsiella pneumoniae seem to play are more important role than E. coli in neonatal EOS [31–34]. Similarly, resistance rates to the commonly used antibiotics are high. This variance between different regions is plausible considering the differences in patients’ characteristics, normal flora and general policy of antibiotic use. Therefore, results of studies evaluating pathogens, antimicrobial resistance and the effect of different antibiotic therapies cannot automatically be generalized and transferred to all countries over the world. In Central Switzerland, the resistance rate of E. coli isolated from humans increased from 39% in 2004 to 49% in 2023 [12]. In our cohort in case of neonatal EOS with E. coli, two of three were resistant to aminopenicillins. This is slightly above the upper margin of the range of 35–65% resistance rates reported in other studies [24, 35–37]. Additionally, most CoNS, as the second common pathogen in perinatal and neonatal infectious morbidity in our cohort, produce penicillinase (an enzyme that inactivates beta-lactam antibiotics such as amoxicillin) and, therefore, are naturally resistant to this type of antibiotic therapy (86% resistance rate of CoNS in Central Switzerland in 2023) [12]. Thus, our standard antibiotic regimen with amoxicillin does not cover most of the relevant pathogens in case of PPROM in our cohort. Additionally, it is notable that in case of detection of E. coli in the vaginal swab and adjustment of the antibiotic therapy with additionally administered co-amoxiclav based on the special requested antibiotic susceptibility testing showing aminopenicillin-resistance and susceptibility to co-amoxiclav and cephalosporins, E. coli was not found in the placental swab. On the contrary, in the one case in which therapy was not adapted (no antibiotic susceptibility testing performed), E. coli was found in the placental swab showing a resistance to aminopenicillin and co-amoxiclav and susceptibility to cephalosporins and the neonate suffered from an infection. Apart from this one case, in all other cases with detection of E. coli as well as in all cases with detection of CoNS in the placental swab or in the blood culture in neonatal sepsis, both were not found in the vaginal swab. Although these findings must be viewed with caution due to the overall low number of cases, they indicate the following: Firstly, in case of E. coli, a prenatal antibiotic therapy adapted to the resistance profile of the pathogen is able to eliminate the pathogen. Secondly, it is necessary to administer a therapy that already covers the most common pathogens from the outset, since these cannot be detected in the vaginal swab at diagnosis of PPROM in most cases. Therefore, there is no possibility to subsequently adjust therapy.

The third relevant pathogen in our cohort were Ureaplasma species. They were by far the most common pathogens in the vaginal swab, detected in every fourth woman with PPROM < 34 + 0 and in every fifth woman with PPROM ≥ 34 + 0 weeks of gestation. Additionally, they were the most frequently detected pathogens in the placental swab in case of chorioamnionitis. The role of Ureaplasma species in human disease is controversial as it is considered to be part of the normal urogenital flora [38, 39]. In non-pregnant women routine testing or treatment in case of asymptomatic carriage is not recommended [40]. However, today there is considerable evidence linking the detection of Ureaplasma species in pregnant women with adverse outcomes such as spontaneous abortion, preterm birth, PPROM and chorioamnionitis [38, 39]. The detection of Ureaplasma species is not only to be understood in the sense of a co-pathogen that can be detected together with the causative agent, but the presence of Ureaplasma species themselves in the amniotic fluid or placental tissue can lead to synthesis and release of prostaglandins and metalloproteases by stimulating cytokine production and chemotaxis of neutrophils, which subsequently leads to cervical ripening, PROM and preterm labor [39, 41]. In addition, through either intrauterine infection or intrapartum transmission, neonates can acquire Ureaplasma infection, which is associated with several serious sequelae [39, 41, 42]. Thus, our findings underline the importance of administering macrolides such as azithromycin in case of PPROM. In recent years, increasing antimicrobial resistance to standard antibiotic therapies (including macrolides) has also been reported for Ureaplasma species [39, 43, 45]. In our cohort, Ureaplasma species could be detected in placental swabs despite an antibiotic therapy with azithromycin after diagnosis of PPROM. However, while Ureaplasma species were the most frequently detected pathogens in vaginal swabs (in 25% of cases of PPROM < 34 + 0 weeks and 20.8% of cases of PPROM ≥ 34 + 0 weeks), they were only detected in 8.8% and 7.1% of cases of PPROM < 34 + 0 weeks and ≥ 34 + 0 weeks, respectively, in placental swabs. This suggests at least some efficacy of the administered azithromycin. In general, four classes of antibiotics are recognized for the treatment of Ureaplasma infections, namely tetracyclines, fluoroquinolones, macrolides and chloramphenicol [43, 44]. Given the teratogenic risk of tetracyclines and the potential fetal toxicity of fluoroquinolones and chloramphenicol, therapeutic options in pregnant women are mainly restricted to macrolides, especially as first-line choice [43]. Regarding whether one macrolide is superior to another in case of PPROM, Seaman et al. found in their systematic review and meta-analysis that the administration of azithromycin was associated with a similar latency period but a lower rate of clinical chorioamnionitis than the administration of erythromycin [46]. The AWMF-guideline, the guideline of the German-speaking countries, recommends the administration of azithromycin in case of PPROM [9]. In view of the reported increasing resistance rates of Ureaplasma species to macrolides a recently performed systematic review and meta-analysis found that azithromycin has a positive therapeutic effect on Ureaplasma infection and that there is no correlation with the azithromycin dose, regardless of whether it is a single dose of 1 g or a multiple-dose regimen [47].

In recent years, some studies suggested that cephalosporins may be superior to aminopenicillins regarding the duration of the latency period between PPROM and birth and the general maternal and neonatal outcome. In a subgroup analysis of the EPIPAGE-2 cohort in France with 492 included women with PPROM at 24–31 weeks of gestation an antibiotic therapy with a third-generation cephalosporin was associated with improved survival without severe morbidity compared to antibiotic therapy with amoxicillin [48]. Likewise, in a small, randomized trial in Israel with 87 included women with PPROM at 24–37 + 0 weeks of gestation, Wolf et al. were able to show a benefit of therapy with cefuroxime + roxithromycin compared to ampicillin + roxithromycin in terms of prolonged latency period and a reduced number of Gram-negative neonatal EOS [49]. Shortly after, the same study group confirmed their data in a larger trial with 124 included women with PPROM at 24–37 + 0 weeks of gestation: The maternal infectious morbidity and the adverse neonatal outcome were significantly lower in the cefuroxime group and the proportion of primiparas with a latency period > 4 days was significantly higher in the cefuroxime group [50]. However, in the most current systematic review and network meta-analysis from 2023 [51] with inclusion of 23 randomized controlled trials with overall more than 7600 women, published from inception of the electronic databases until June 2021, the authors compared cephalosporins (+/- a macrolide) to 9 other antibiotic regimens regarding maternal clinical chorioamnionitis and neonatal sepsis as primary outcome but failed to validate superiority of cephalosporins relatively to a control or placebo group, presumably because there were only very few trials published within the last decade taking into account the currently relevant pathogens and their increased aminopenicillin-resistance.

Overall, we observed favorable perinatal outcomes in our cohort, especially low rates of neonatal death and infectious morbidity, similar or even lower compared to the rates reported in other studies [52–55]. Furthermore, the latency periods in our cohort with PPROM < 34 + 0 weeks of gestation, especially in the subgroup with PPROM before or near the limit of viability (< 24 + 0 weeks), were remarkably longer compared to these known from other studies [52, 55]. This can be explained by differences in treatment options and general management of prematurely born neonates in different countries, but also by the fact that advances in neonatal care, particularly in intensive care to those at the threshold of viability, have significantly enhanced survival rates over the last decades. Therefore, later studies such as ours naturally show better results than earlier ones.

Our study has several limitations, mainly its retrospective single-center design with a risk of bias and limitation of generalizability of the results. In addition, although we were able to include all pregnancies with neonatal sepsis (EOS and LOS) and maternal sepsis, a potential selection bias due to the exclusion of a significant number of women without signed general consent must be considered. Furthermore, in the group with PPROM ≥ 34 + 0 weeks of gestation, the findings regarding the pathogen spectrum in the vaginal and placental swabs and, in the entire cohort, the findings regarding the occurrence of chorioamnionitis and its pathogen spectrum, must be interpreted with caution due to the large number of missing swabs and histological examinations of the placenta. In addition, a certain degree of contamination of the placenta via the birth canal or contamination of the placental swabs during swab collection cannot be ruled out and must therefore be taken into account, especially when interpreting the results of the high detection rates of CoNS in the placental swabs. Nevertheless, our study includes a large cohort of pregnancies complicated by PPROM and, for the purpose of this study (to evaluate whether our current standard antibiotic regimen with amoxicillin and azithromycin is appropriate to cover the current spectrum of relevant pathogens in case of PPROM and therefore focus on the general pathogen spectrum and the perinatal and neonatal infectious morbidity), the study design is suitable. Additionally, our findings are in line with the global observations of changes in pathogen spectrum and antimicrobial resistance.

Conclusions

The results of our study with predominantly aminopenicillin-resistant E. coli and CoNS as well as Ureaplasma species as the currently most relevant pathogens in PPROM underline the need to reconsider our standard antibiotic regimen with a combination of an aminopenicillin and a macrolide as it is recommended in most guidelines today. To cover Ureaplasma species, macrolides such as azithromycin should continue to be administered, but aminopenicillins no longer appear to be the optimal choice. Currently, the optimal antibiotic regimen is simply unknown. Due to the changes in pathogen spectrum and antimicrobial resistance in recent decades, the results of the larger part of performed studies are outdated and the latest evidence is poor and, so far, insufficient to support a different regimen. Thus, further research is necessary and should focus on conducting randomized controlled trials with comparison of aminopenicillins to antibiotics with better coverage of the currently relevant pathogens such as co-amoxiclav or cephalosporins.

Abbreviations

CoNS

Coagulase-negative Staphylococci

CPAP

Continuous positive airway pressure

EOS

Early-onset sepsis

LOS

Late-onset sepsis

NEC

Necrotizing enterocolitis

NICU

Neonatal intensive care unit / neonatal intermediate care unit

PCR

Polymerase chain reaction

PPROM

Preterm premature rupture of membranes

Authors’ contributions

Livia Schuler: Writing – original draft, Investigation, Data curation. Martin Stocker: Writing – review & editing. Markus Hodel: Writing – review & editing. Christian Braun: Writing – review & editing. Corina Christmann: Writing – review & editing. Christine Brambs: Writing – review & editing, Supervision. Sara Ardabili: Writing – original draft, Project administration, Conceptualization, Methodology, Investigation, Data curation, Formal analysis, Supervision.

Funding

There are no funding sources.

Data availability

The datasets supporting the conclusions of this article are included within the article.

Declarations

Ethics approval and consent to participate

The study was approved by the local ethics committee “Ethikkommission Nordwest- und Zentralschweiz (EKNZ)” in Basel, Switzerland (Project-ID 2023 − 01319, date of approval July 14, 2023) and was performed in accordance with the Declaration of Helsinki. All participants signed our hospital`s general informed consent form, which allows the use of their health-related data and samples for research purposes.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.