Revisão Acesso aberto Revisado por pares

Listeriosis During Pregnancy

2002; Wolters Kluwer; Volume: 81; Issue: 4 Linguagem: Inglês

10.1097/00005792-200207000-00002

ISSN

1536-5964

Autores

Eleftherios Mylonakis, Maria Paliou, Elizabeth Hohmann, Stephen B. Calderwood, Edward J. Wing,

Tópico(s)

Salmonella and Campylobacter epidemiology

Resumo

Introduction Listeriosis has a unique predilection for pregnant women, with an estimated 17-fold increase in incidence (118,126). Maternal listeriosis is a diagnostic challenge, and intrauterine infection can lead to severe complications such as amnionitis, preterm labor, spontaneous abortion, stillbirth, or infection of the neonate. Over the past 20 years, epidemics of listeriosis have lead to an increased number of cases and higher awareness (90). We examined the clinical characteristics of perinatal listeriosis and to evaluate possible risk factors, clinical and laboratory findings, response to therapy, and outcome of perinatal listeriosis in mothers and neonates. Methods Based on a computerized study of discharge diagnoses, review of the records of the Infection Control Department, and questioning of infectious disease specialists, we identified cases of pregnant women with perinatal listeriosis who were treated at 4 hospitals (Brigham and Women’s Hospital, Boston, MA; Massachusetts General Hospital, Boston, MA; Mount Auburn Hospital, Cambridge, MA; and Women and Infants Hospital, Providence, RI) between January 1990 and January 2000. All cases were in pregnant women who had Listeria monocytogenes isolated from cultures of normally sterile body sites or vaginal swabs. In addition, we reviewed the English-language literature published between January 1980 and July 2000 to identify similar cases of listeriosis during pregnancy. For this review we conducted a computer-based search of MEDLINE (National Library of Medicine, Bethesda, MD), using the key words listeriosis and pregnancy. We complemented this computer-based search with manual search of Index Medicus (National Library of Medicine, Bethesda, MD) and by cross-referencing published articles. We excluded duplicate cases and reports without adequate clinical data. In the evaluation of antimicrobial therapy we did not include antibiotics used for 2 or fewer days in our calculations. Results Case study Maternal infection: We identified 11 pregnant women with listeriosis (mean age, 30.4 yr; range, 20–39 yr), 3 in the state of Rhode Island and 8 in Massachusetts (Table 1). Three women had no previous pregnancies and 2 women had a history of spontaneous abortions (1 woman with 2 and 1 woman with 3 previous abortions). One woman was on immunosuppressive drugs for heart transplantation secondary to cardiomyopathy, and 1 had developed gestational diabetes during the first trimester of pregnancy. None had human immunodeficiency virus (HIV) or liver disease. Two women had a history of consuming foods that have been linked to Listeria infection (cole slaw and hot dogs) a few days before developing symptoms of listeriosis. Cultures from these foods were not available.TABLE 1: Features of 11 episodes of perinatal listeriosis, treated at 4 hospitals in Massachusetts and Rhode IslandThe clinical and laboratory findings included leukocytosis in 8 of 10 patients tested, abnormal liver function tests in 3, and tenderness in the right lower quadrant in 1. There were 6 preterm deliveries (4 by vaginal delivery and 2 by cesarean section), and 1 patient had a cesarean section due to fetal demise. Five of 6 patients had premature contractions not responding to treatment. When the symptoms of listeriosis developed, 9 of 11 women were in the third trimester of pregnancy while 2 of 11 were in the 17th and 18th week of gestation, respectively. These 2 pregnancies ended in, respectively, a spontaneous abortion and early death of a preterm delivered baby. The time from initial presentation of Listeria infection to diagnosis varied. In 8 of 11 cases, the diagnosis was made postpartum, while in 3 patients, symptoms of Listeria infection appeared at least 1 week before delivery and were diagnosed prepartum. All women were treated with intravenous antibiotics. Initial treatment included ampicillin with gentamicin (6 patients), ampicillin or amoxicillin monotherapy (2 patients), tobramycin with clindamycin, ceftriaxone with erythromycin, and vancomycin with gentamicin (1 patient each). The duration of intravenous antibiotics ranged from 3 to 12 days. Three women were discharged on additional oral antibiotics: 2 on erythromycin and 1 on amoxicillin. No follow-up is available. Neonatal infection: As noted above, 2 of 11 pregnancies resulted in fetal death: a spontaneous abortion and birth of a fetus that died immediately after birth. Findings in the latter case were consistent with “granulomatosis infantisepticum.” Three neonates born to infected mothers were not infected by Listeria spp. These infants were born at term, had no symptoms or clinical findings of listeriosis, and results of blood (all 3) and cerebrospinal fluid evaluations (performed in 2 of 3 cases) were negative. Two of these infants received intravenous antibiotic therapy with ampicillin and gentamicin until culture results were finalized. The remaining 6 neonates were infected by Listeria. All (6/6) were delivered preterm and developed symptoms of listeriosis at or immediately after birth. The placenta was examined pathologically in 8 cases and revealed acute chorioamnionitis in all (including the 2 infants who did not develop Listeria infection). Clinically infected neonates had blood cultures and evaluation of cerebrospinal fluid performed as part of their laboratory evaluation. Based on cultures and clinical presentation, 4 neonates were diagnosed with Listeria bacteremia/sepsis, 1 with bacteremia/ sepsis and pneumonia, and 1 with bacteremia/sepsis and meningitis. All infected neonates were treated with intravenous ampicillin and gentamicin for 14–20 days. Short-term follow-up was available for 6 of 7 infected neonates and in all (6/6), the infection resolved successfully with antibiotic treatment. No long-term follow-up is available. Literature review We identified 222 patients with perinatal listeriosis reported in the English-language literature between January 1980 and January 2000 for whom adequate clinical data were available (7,9,10,13,14–17,22–26,29,30,32,33,35,37–39,40,41,46–50,52,54,58–66,68,71,73,74,78,79,81,84,86–89,91,93–95,97–103,105–110,113,115,117,119–123,125,127,128,130–132,136,137). Maternal infection: Mean age of infection was 27.5 years (range, 17–38 yr). As in our case series, most patients were otherwise healthy. Possible predisposing conditions (in addition to pregnancy) included use of corticosteroids (3 cases), diabetes mellitus (2 cases), systemic lupus erythematosus (2 cases), and HIV infection (2 cases). Data on previous spontaneous abortions were available in 96 cases, and 18 of 96 (18.8%) women had 1 or more previous miscarriages (range, 1–4; mean, 1.6). The most common symptoms and signs on presentation are summarized in Table 2. Fever, the most common symptom, ranged between 38.2 °C and 41.2 °C (mean, 38.9 °C), while in 61 of 191 patients (32%), the clinical presentation was described as a flu-like syndrome. The average duration of symptoms before diagnosis was 6.6 days (range, 1–21 d). In 42 cases with a peripheral white blood count reported on admission or at diagnosis, the mean was 16,300 cells/mm3 (range, 3,900 to 33,800 cells/mm3). The most common means of diagnosis included culture of blood, cervix, or placenta (Table 3).TABLE 2: Presenting symptoms and signs of pregnant women with listeriosis*TABLE 3: Microbiologic diagnosis of Listeria infection in pregnant women*Data on therapy were available in 75 cases. The most common agent used was ampicillin or amoxicillin alone (31 patients) or with an aminoglycoside (19 patients). Other antibiotics included penicillin combined with an aminoglycoside (6 patients), penicillin alone (4 patients), erythromycin with an aminoglycoside (4 patients), trimethoprim-sulfamethoxazole (3 patients), erythromycin with an aminoglycoside (4 patients), or erythromycin alone (1 patient). Three patients received a cephalosporin. The mean duration of therapy was 13.2 days (range, 5 d to 3 mo). Details on the outcome of the pregnancy were available in 178 cases and in 36 of 178 (20.2%) of those cases, pregnancy resulted in spontaneous abortion or stillbirth. Among the remaining 142 cases, 97 neonates were infected (68.3%) and 45 (31.7%), were not infected. Neonatal infection: Among the 97 neonates infected with listeriosis, details on clinical presentation were available for 94. The most common signs and symptoms among neonates with early listeriosis included respiratory distress, fever, and/or neurologic abnormalities (Table 4). Microbiologic diagnosis was established based on culture of blood, surface site (such as the skin, conjunctiva, or ear), cerebrospinal fluid, or oropharyngeal secretions (Table 5). Of note is that 23 neonates (24.5%) had meningitis as part of the infection. Based on microbiologic and clinical data, infected neonates were diagnosed clinically with pneumonia (n = 19; 20.2%), bacteremia/sepsis (n = 24; 25.5%), pneumonia with bacteremia/sepsis (n = 28; 29.8%), meningitis and bacteremia/sepsis (n = 9; 9.6%), meningitis alone (n = 5; 5.3%), pneumonia, meningitis, and bacteremia/ sepsis (n = 5; 5.3%), and meningitis with pneumonia (n = 4; 4.3%).TABLE 4: Presenting symptoms and signs of neonates with acute listeriosis*TABLE 5: Microbiologic diagnosis of Listeria infection in neonates*Data on antibiotic therapy for neonates were available in 64 cases. Most affected neonates received ampicillin/amoxicillin alone (21 cases) or combined with an aminoglycoside (30 cases). Other antibiotic regimens included penicillin with an aminoglycoside (4 cases) and ampicillin with a cephalosporin (2). The duration of antibiotics was reported in 21 cases. Most neonates (12 of 21) received antimicrobial therapy for 14 days, and 4 of 21 received treatment for 3 weeks. Among the remaining 5 cases, 3 received treatment for 12 or fewer days and 2 for more than 3 weeks. The duration of follow-up varied among reports. Fifty-nine of 94 infected neonates (62.8%) recovered completely, 23 (24.5%) died, and 12 (12.7%) had neurologic sequelae or other long-term complications. Meningitis alone or in combination with bacteremia/sepsis or pneumonia was associated with the worst prognosis (Figure 1).Fig. 1: Outcome based on clinical and microbiologic diagnosis among 100 neonates born to mothers with perinatal listeriosis and who developed early-onset listeriosis (6 cases from the present case series and 94 cases from the literature).Discussion We reviewed 11 cases of maternal listeriosis from 4 hospitals in Massachusetts and Rhode Island and compared these data with the cases reported in the English-language literature between 1980 and 2000. Epidemiology Human listeriosis has been associated with a number of predisposing conditions such as age older than 50 years, diabetes mellitus, immune suppression, liver failure, HIV infection, and splenectomy (90). Pregnant women and neonates are also more susceptible to listeriosis, probably due to depressed cell-mediated immunity during pregnancy (55). Animal studies also support a causative role for depressed cell-mediated immunity increasing susceptibility to listerial infection during pregnancy, and T cells are the major component required for acquired protective immunity against Listeria (44,116). The mode of transmission of L. monocytogenes to the fetus is either transplacental via the maternal blood stream, or ascending from a colonized vaginal canal (118). Given the ability of L. monocytogenes to spread from cell to cell, it is easy to speculate how transplacental infection might occur, something uncommon in most other causes of maternal bacteremia (12,96,118). Cell-mediated immunity is also impaired during the neonatal period with documented decreased natural killer cell activity and defective macrophage-T cell interactions (55,118). The incidence of this infection among pregnant women has been estimated at 12 per 100,000, compared to 0.7 per 100,000 in the general population (118,126). A number of sporadic cases reported in the literature over the last few years and the probable increase of at least some forms of listeriosis has been attributed to higher awareness among health care providers and changes in diet or food processing (5,76,90,124). Among the 11 cases in our series, 1 woman was receiving immunosuppressive drugs for heart transplantation and 1 had developed gestational diabetes during the first trimester of pregnancy. Other cases of listeriosis among pregnant patients with immune suppression have been reported (25,33). However, as seen in our case series and the review of the literature, perinatal listeriosis mainly affects women without additional predisposing factors. Alimentary transmission is considered the most common route of acquisition of Listeria infection. Most cases of human listeriosis, including those during pregnancy, have been sporadic, with occasional common-source outbreaks. Such outbreaks have been associated with various foods such as coleslaw, Mexican-style soft cheese, milk, and pâté (1,4,8,18,19,27,28,31,34,36,43,56,72,75,77,92,111,114,116,129,133). Two women in our case series had a history of consuming foods linked to listeriosis a few days before developing symptoms of Listeria infection. Fecal carriage of L. monocytogenes occurs in 1%– 15% of the population (70); the incidence of women carrying L. monocytogenes in the vagina is lower (21). Pregnancy does not affect the fecal, cervicovaginal or oropharyngeal carriage rate of L. monocytogenes (69), but carriage of Listeria has been reported as a possible additional predisposing factor for perinatal listeriosis. In Israel, the carriage of L. monocytogenes in the genital tract was associated with a history of recurrent abortion (104), but this was not confirmed in Britain (80) or the United States (3). In another study (82) from the United States, done from 1979 to 1989, Listeria was not isolated from the cervix or the endometrium of any of the 86 patients with 2 or more fetal losses. The data on neonatal infection reported herein focus on early neonatal disease. In general, neonatal listeriosis is classified as early or late infection, depending on timing of the onset of symptoms. Early infection occurs in the first 5–7 days following delivery and is almost always associated with maternal illness. Late infection develops more than 5–7 days after delivery, often presents as a meningeal illness, and some cases have been associated with nosocomial transmission (58). Infants with late-onset listeriosis are generally full term, healthy at birth, and delivered to mothers who have had uncomplicated pregnancies. The source of L. monocytogenes in this group is not known but is possibly the mother’s alimentary tract or the environment, as the organism is rarely isolated from the mother’s genital tract (122). In our study, we used 5 days as the cutoff between early and late listeriosis. However, there is some overlap in the definition of the 2 groups in the literature (29). Clinical manifestations As seen in our case series and review, a common presentation of illness in the mother is that of a flu-like syndrome (70). In our study, there were no cases of severe systemic listeriosis in the pregnant women, although in 36% of cases, blood cultures were positive. However, there are rare case reports of Listeria meningoencephalitis and severe maternal infection during pregnancy (85,121). Early-onset neonatal listeriosis most often presents as pneumonia and/or sepsis, but, as seen in our review, neonates also can present with meningitis. Clinical manifestations of early neonatal listeriosis are variable; however, most babies are premature and most cases are clinically apparent at the time of delivery, with respiratory distress (cyanosis, apnea, etc.), meconium-staining, fever, rash, jaundice, or lethargy. As in Patient 9 in our report, there may be a characteristic rash that facilitates early diagnosis of listeriosis. These lesions may be seen on the skin, usually on the trunk or extremities, and are maculopapular or papulovesicular in nature. Microabscesses may be seen on the fetal surface of the placenta as well. A discrete roseolar or pustular eruption may also be seen on the skin or posterior pharynx. Slightly elevated pale patches measuring 1 to 2 mm in diameter with a bright erythematous base have been described (11). Granulomatosis infantisepticum occurs less frequently, and the classic syndrome includes granulomatous pneumonia and granulomas in skin, liver, and other locations (50,113). Diagnosis As there are no pathognomonic clinical features associated with perinatal listeriosis, the diagnosis depends on bacteriologic findings. L. monocytogenes resembles a diphtheroid morphologically (short, Gram-positive bacillus) and therefore can be mistaken for a contaminant if cultures are not examined carefully. A high index of suspicion is a prerequisite for the diagnosis of listeriosis, which can occur at any stage in pregnancy. When suspected, perinatal listeriosis can be diagnosed easily by cultures of stool, urine, cervical swab, amniotic fluid, or maternal blood. In both our case series and the literature, microbiologic diagnosis is most commonly made by culturing of blood (about 35% of cases). Blood cultures should be considered in the evaluation of pregnant women with a flu-like syndrome and should be obtained before antibiotic therapy is started (112). After delivery, the placenta should be examined whenever congenital infection is suspected. In our case series, examination of the placenta revealed chorioamnionitis in all cases with early neonatal listeriosis (and also in 2 neonates that did not develop listeriosis). In addition to Gram stain and culture of surface sites, frozen sections of the placenta may enable a rapid diagnosis (15). In cases of early-onset disease, L. monocytogenes can be isolated from the blood, superficial sites (skin, conjunctiva, external ear, etc.), cerebrospinal fluid, oropharyngeal secretions, gastric aspirate, amniotic fluid, or placenta. Biopsy of skin lesions may also reveal a leukocytic infiltrate with bacteria (11). In 1 series, 12 of 23 neonatal patients exhibited meconium staining of the amniotic fluid despite the often premature delivery (22). In early neonatal listeriosis, amniotic fluid tends to be “old and dark,” while sometimes it is simply discolored or brownish and does not appear to be meconium stained (83). The presence of meconium macroscopically and Gram-positive rods with tumbling motility microscopically in the amniotic fluid should alert the clinician (22,74,98). Treatment All women from our case series and almost all women in the literature review recovered. However, fatalities among pregnant women with listeriosis have been reported (33,136). Antibiotic therapy is recommended for treatment of the mother and may also prevent neonatal illness (60,75,112), even early in pregnancy (79). Maternal listeriosis at 26 weeks gestation was successfully treated with ampicillin, with delivery of a normal infant at term (35). Fuchs and coworkers (40) described a pregnant woman with Listeria bacteremia during the 13th week of gestation. She received ampicillin and gentamicin for 3 weeks and delivered a healthy boy in the 36th week of pregnancy. Maternal listeriosis at 19 weeks of gestation was successfully treated with resultant birth of an unaffected infant at term (53). It is noteworthy that case reports suggest even untreated maternal listeriosis may, in rare instances, be associated with the birth of an unaffected infant (60,75,112). High-dose ampicillin is the treatment of choice for Listeria infection, and the synergistic and bactericidal effects of adding gentamicin may be desirable in treatment of newborns and patients with underlying disease (6). Penicillin can also be used, but microbiologic data and animal studies suggest that ampicillin may be superior (51,57,90). Previous studies have occasionally noted strains with resistance to penicillin and/or ampicillin, but those reports had limited data on the microbiologic methodology. Plasmidmediated resistance to chloramphenicol, erythromycin, and tetracycline in Listeria have been reported (101). Furthermore, strains of L. monocytogenes resistant to trimethoprim (19,20), gentamicin (134), streptomycin, kanamycin, sulfamethoxazole, or rifampin have also been described (19,20,90). Cephalosporins and clindamycin, which might otherwise be appropriate choices for Gram-positive pathogens typically associated with endomyometritis or septic abortion, are not effective antimicrobial agents for L. monocytogenes. In nonpregnant patients, trimethoprim-sulfamethoxazole is considered as a second-line agent for patients with listeriosis and severe allergy to penicillin who cannot undergo desensitization therapy (90). However, trimethoprim-sulfamethoxazole is Food and Drug Administration (FDA) category C for use in pregnancy (animal studies show toxicity, human studies inadequate, but benefit may exceed risk in some cases). The lack of a well-studied and clearly safe second-line agent for Listeria infection during pregnancy emphasizes the importance of a thorough evaluation of reported penicillin allergies. Allergy testing and/or desensitization should be considered (135). Failures of vancomycin, tetracycline, fluoroquinolones, and erythromycin have also been described (45,90), but such treatment can be used in selected patients. Outcome Approximately 1 of 5 pregnancies complicated by listeriosis results in spontaneous abortion or stillbirth. The prognosis is worse in early pregnancy. In our 11 cases, 2 pregnancies complicated by Listeria infection in the second trimester ended in a spontaneous abortion or in early death of a preterm infant. In our series and in the literature, approximately two-thirds of surviving infants born to mothers with listeriosis during pregnancy developed clinical neonatal listeriosis. However, even neonates who do not develop listeriosis may have a prolonged hospital stay, compared to neonates born to uninfected mothers (14). Prevention Recurrent miscarriage has been attributed to L. monocytogenes infection in some studies but not others. Giraud et al (42) detected L. monocytogenes in 8 of 388 (2.1%) miscarriages and in 5 liveborn and 5 stillborn infants out of 1,112 (0.9%) preterm deliveries. Listeriosis was diagnosed in 13 of the 422 (3.1%) second-trimester spontaneous abortions studied in another series (67). Pregnant women should avoid soft cheese and should reheat leftover foods. If vaginal cultures of a pregnant woman show growth of Listeria, some authors suggest attempting to eradicate the microorganism before delivery (29). Some reports recommend that vaginal swabs for culture of Listeria should be considered as part of the evaluation of women with recurrent miscarriages, mothers of infants dying early in life, and in some cases of pregnant women in regular contact with animal reservoirs of Listeria (2). Conclusions Perinatal listeriosis usually affects pregnant women without other predisposing factors, often leads to preterm delivery, and can cause fetal death, especially when the infection occurs during the second trimester of pregnancy. A common presentation is that of a flu-like illness, and listeriosis should be a diagnostic consideration in febrile pregnant women at any stage of pregnancy. Placental pathology may show acute chorioamnionitis and suggest the diagnosis. Approximately one-fifth of pregnancies affected by listeriosis result in spontaneous abortion or stillbirth. Among the remaining cases, approximately two-thirds of the neonates are infected. Infected neonates usually develop pneumonia or bacteremia/ sepsis, alone or in combination, while meningitis develops in 25% of cases. Summary Pregnant women have an increased incidence of listeriosis. We identified 11 patients with perinatal listeriosis who were treated at 4 hospitals in New England during a 10-year period. We also reviewed the English-language literature on perinatal listeriosis (January 1980–July 2000) and identified 222 previously reported cases. Most pregnant women with listeriosis had no additional predisposing factors. Fever was the most common symptom, and more than 30% of patients presented with a flu-like syndrome. In one-fifth of cases, pregnancy resulted in spontaneous abortion or stillbirth. Among the remaining 142 cases, 97 (68.3%) neonates were infected by Listeria. Infected neonates were most commonly diagnosed with pneumonia, bacteremia/sepsis, or meningitis. Meningitis alone, or in combination with bacteremia/sepsis or pneumonia, was associated with a worse prognosis.

Referência(s)
Altmetric
PlumX