Inadequate management of pregnancy-associated listeriosis: lessons from four case reports
Listeria monocytogenes infection during pregnancy can lead to dramatic fetal or neonatal outcomes. No clinical trial has evaluated treatment options, and retrospective studies of cases are therefore important to define optimal regimens. We report four cases of materno–neonatal listeriosis illustrating inadequate antimicrobial therapy management and discuss recommended treatment options.
Listeria monocytogenes is a gram-positive bacillus responsible for listeriosis, a rare yet potentially severe food-borne infection  . Pregnant women are at particularly high risk for invasive listeriosis, with mild or even asymptomatic maternal presentation contrasting with the potentially dramatic fetal or neonatal outcomes, including abortion, stillbirth, preterm birth and severe neonatal infection . Fetal or neonatal loss occurs in 20% of cases .
Prevention in pregnant women relies on adherence to strict diet recommendations such as reheating of leftovers until steaming and avoidance of unpasteurized dairy products, cold delicatessen items, meat spreads and pâtés [http://www.cdc.gov/Listeria/prevention.html 1999].
Early diagnosis remains challenging: 30% of infected mothers are asymptomatic and blood cultures are positive in only 36% of symptomatic mothers . Treatment recommendations are based on in vitro and in vivo experimental studies, and heterogeneous historical cohorts . Amoxicillin is the keystone of listeriosis treatment and should be swiftly prescribed when L. monocytogenes is isolated during pregnancy. In addition, some experts recommend empiric amoxicillin therapy in the presence of high-level clinical suspicion, without consensus on treatment dosage and duration . In the absence of clinical trials addressing this issue, retrospective studies of cases are instrumental in clarifying the situations requiring antibiotics, optimal regimen, dosage and duration . We report four cases of pregnancy-associated listeriosis with inadequate antibiotic therapy to help in defining optimal treatment options.
Patients and Methods
Surveillance of human listeriosis in France is based on mandatory reporting of cases and voluntary submission of L. monocytogenes isolates to the French National Reference Centre (NRC) for Listeria. We defined a case of pregnancy-associated listeriosis as L. monocytogenes isolation from a pregnant woman, her fetus, or from a newborn <28 days old. All 146 pregnancy-associated cases reported to the NRC, including those included in the prospective national study conducted by the NRC (MONALISA, Clinical trials NCT01520597) from January 2009 to September 2012, were reviewed: those who reported fever or who had received antimicrobial therapy before listeriosis were identified and their medical records were analysed. Four cases with medical/obstetric conditions compatible with listeriosis were retrieved and presented here. Isolates were identified as L. monocytogenes at the NRC. No other pathogen was identified.
A 37-year-old G8P4 (gestations 8, parity 4) obese pregnant woman from Cameroon presented with genital hemorrhage at 33 weeks of gestation. Medical history included three spontaneous abortions and myomectomy by laparoscopy. She was afebrile and blood pressure was normal. Endocervical bleeding was evidenced, without evidence for placenta praevia or placental abruption. Vaginal swab and urine samples evidenced no pathogen. C-reactive protein (CRP) was 13 mg/L. No blood culture was performed. Bleeding spontaneously waned and she was discharged with amoxicillin 3 g/day for 5 days according to local protocols for genital haemorrhages. At 37 weeks of gestation she gave birth by cesarean section (uterus scar) to a eutrophic male baby. Infant clinical examination and placenta macroscopic aspect were normal. Meconium culture yielded numerous colonies of L. monocytogenes. Infant blood culture was sterile and his CRP <1 mg/L. He was treated with intravenous amoxicillin 100 mg/kg/day for 7 days and gentamicin for 2 days. His clinical examination and CRP level remained normal, stool cultures remained negative for L. monocytogenes and he was discharged home.
A 35-year-old G1P0 pregnant woman without medical history presented with fever (38.5°C) at 29 weeks of gestation. Cervical and fetal assessments were unremarkable. No focal infection was found. Leucocytosis was 20 × 109/L, CRP was 50 mg/L, vaginal swab found no pathogen. Blood cultures grew L. monocytogenes, for which oral amoxicillin 3 g/day was prescribed for 5 days, followed by cefpodoxime 200 mg twice a day for a further 5 days. She became afebrile and was discharged home. She consulted at 31 weeks of gestation for fever and uterine contractions, and gave birth prematurely to a female eutrophic baby whose gastric aspiration, placenta and blood cultures were positive for L. monocytogenes. Infant lumbar puncture was haemorrhagic. Ultrasound brain imaging displayed ventriculitis consistent with meningitis. Maternal vaginal swab and urine sample tested positive for L. monocytogenes. The infant received intravenous amoxicillin 200 mg/kg/day for 21 days and tobramycin for 5 days, with favourable outcome.
A 24-year-old G1P0 pregnant woman presented with fever (38°C), uterine contractions and vomiting at 34 weeks of gestation. Medical history included smoking cessation and nephrolithiasis. At examination no focal infection, vaginal discharge or rupture of membranes was found. Cervical and fetal assessments were unremarkable. Vaginal swab and urine sample cultures were sterile. No blood culture was performed. Renal and urinary tract ultrasonography was normal. Leucocytosis was 8.2 × 109/L, CRP was 25 mg/L and liver function tests were normal. She was discharged with cefixime 200 mg twice daily for 7 days. At 36 weeks of gestation she consulted for fever (38°C) and uterine contractions. Leucocytosis was 20 × 109/L and CRP was 36 mg/L. She prematurely gave birth by vaginal delivery to a female febrile (38°C) eutrophic baby who presented respiratory failure. Amniotic fluid was meconium-stained. Gastric aspirate, ear swabs, blood culture and cerebrospinal fluid all yielded L. monocytogenes. The baby was successfully treated with amoxicillin 200 mg/kg/day for 21 days.
A 35-year-old G8P3 pregnant woman from Cameroon presented with fever (38°C), vomiting and diarrhoea at 29 weeks of gestation. She was human immunodeficiency virus (HIV) -infected (CD4+ T-cell blood count 577/mm3, HIV viral load < 20 copies/mL) and her antiretroviral regimen included tenofovir, emtricitabine and lopinavir/ritonavir. Fetal assessment was normal. Vaginal swab evidenced no pathogen; stool, urine and blood samples cultured negative. Leucocytosis was 5.2 × 109/L, CRP was 47 mg/L. She was discharged without antimicrobial therapy and her symptoms resolved. At 38 weeks of gestation, she was admitted to the emergency room for fever and uterine contractions. She gave birth by cesarean section to a male eutrophic baby whose clinical examination was normal. Maternal blood cultures grew L. monocytogenes and urine cultures were negative. The mother received intravenous amoxicillin 8 g/day for 14 days and gentamicin for 2 days, with good outcome. Newborn blood, ear swab, placenta and gastric aspiration samples cultured negative. The baby received amoxicillin 100 mg/kg/day for 6 days and gentamicin for 3 days. He remained asymptomatic and HIV-negative.
These observations present an array of maternal clinical situations that ended in overt neonatal listeriosis (cases 2–4) or neonatal L. monocytogenes colonization (case 1). Together, they illustrate that maternal listeriosis optimal treatment requires the recognition of maternal clinical settings evocative of listeriosis and the timely prescription of an active drug, amoxicillin, at the right dose and time and for adequate duration (above 3 g/day, for more than 5 days).
Cases 1, 3 and 4 were documented postnatally, without direct assessment of maternal infection. Case 3 in particular, although febrile, had no blood cultures collected. Maternal clinical presentation of cases 3 and 4 was evocative of listeriosis, although the 9-week delay between gastrointestinal symptoms and premature birth in case 4 was higher than the median incubation time reported in pregnancy . In case 1, the link between isolated genital haemorrhage and maternal listeriosis remains unclear; however, the exceptional observation of L. monocytogenes in the first meconium without neonatal systemic infection suggests, although it does not demonstrate, an incomplete effect of maternal amoxicillin.
When to treat?
Early diagnosis and therapy are the cornerstone of listeriosis management. Available recommendations favour empirical treatment in high-risk situations , such as fever during labour, preterm labour or uterine irritability, especially during an ongoing listeriosis outbreak . Other local practices, such as in France—where public health authorities dealt in the mid-1990s with repeated listeriosis outbreaks and numerous sporadic cases—consider amoxicillin in each case of unexplained fever during pregnancy to prevent overt listeriosis [7, 8]. The relevance of this practice is debated. Pros include the delayed and low frequency of positive maternal blood cultures in maternal listeriosis , the potentially severe outcome of untreated listeriosis, and the efficacy of prompt therapy . Cons include the unevaluated cost-effectiveness of this strategy given the low incidence of pregnancy-associated listeriosis against the high incidence of undocumented fever during pregnancy—in 2011, 31 cases of pregnancy-associated listeriosis were reported to the French NRC for a total of c.800 000 live births during the same period—the risk of antibiotic adverse effects (allergy), the ineffectiveness of amoxicillin on other common and potentially severe bacterial infections (Enterobacteriaceae), and the selection of resistant enteric bacteria possibly transmitted to the neonate. These reports highlight situations of L. monocytogenes suspicion that would have or did benefit from adequate diagnosis strategy based on blood cultures and swift empiric treatment: unexplained isolated fever (case 2), fever with gastrointestinal symptoms and uterine contractions (cases 3 and 4). Listeriosis, although rare, should be kept in mind along with other febrile diseases requiring specific therapies: gastroenteritis, influenza or flu-like infections, pyelonephritis, chorioamnionitis. In case of fever, blood cultures should be performed.
How to treat?
First, when empiric therapy is decided, active antimicrobials should be chosen, unlike cases 2 and 3 who received cephalosporin to which L. monocytogenes is naturally resistant . Recommended dosage is mostly based on L. monocytogenes meningitis treatment recommendations  and ranges from 3 to 12 g/day for 7–28 days, or even until delivery [4, 10, 11]. French recommendations rely on amoxicillin 100 mg/kg/day for 2 weeks, associated with gentamicin 5 mg/kg/day for 3 to 5 days . Cases 1 and 2 suggest that high-dose (>3 g/day) and prolonged (>5 days) therapy are needed for cure. These mothers received amoxicillin 3 g/day for 5 days, which turned out to be ineffective for case 2, and partly ineffective for case 1, where L. monocytogenes was cultured from meconium, without overt infection. Gentamicin is synergic with amoxicillin and could have been added in case 2 .
In conclusion, these cases illustrate that L. monocytogenes should be kept in mind as a cause of unexplained fever during pregnancy; they provide data that underline the benefit of early adequate empiric therapy in well-chosen situations, and the need for high-dose prolonged amoxicillin therapy. Data from larger cohorts are needed to clarify optimal L. monocytogenes therapy in pregnancy.
We thank Mathieu Tourdjman (InVS) for critical reading and helpful discussions, and Viviane Chenal-Francisque (NRC) and Véronique Goulet (InVS) for their respective contributions to the national surveillance of listeriosis. This study received financial support from Institut Pasteur, Inserm and Institut de Veille Sanitaire.
The authors have no conflicts of interest to declare.