|Home | About | Journals | Submit | Contact Us | Français|
Streptococcus pneumoniae is a rarely recognized cause of neonatal sepsis. We present a recent case of S. pneumoniae bacteremia acquired on the first day of life in a neonate born at 30 weeks of gestation to a mother without prenatal care who had prolonged rupture of the membranes and received intravenous ampicillin prior to delivery. The isolate was resistant to penicillin, with a MIC of the drug of 4 μg/ml. The child responded to a 7-day course of intravenous vancomycin. S. pneumoniae was recovered from the vagina of the mother on a swab culture collected prior to delivery, and isolates from mother and child were confirmed to be identical on the basis of pulsed-field gel electrophoresis. Although neonatal sepsis due to the peripartum transmission of S. pneumoniae is rare, this case highlights the concern that increasing efforts to prevent group B streptococcus neonatal disease may lead to an increase in neonatal infections due to resistant organisms.
The mother was a 20-year-old employee of a fast-food restaurant who presented in premature labor with rupture of her membranes that occurred during sexual intercourse on the day of admission. This was her fourth pregnancy; she had undergone normal vaginal delivery of a healthy, full-term male infant 2 years earlier and had had two voluntary abortions. She presented at her 30th week of pregnancy after having had only one prenatal visit approximately 1 month prior to her presentation. Her only medications were prenatal vitamins and metronidazole that she had been taking for an episode of bacterial vaginosis diagnosed 1 to 2 weeks earlier. She denied history of fever, cough, headache, neck stiffness, dysuria, abdominal pain, or diarrhea.
On examination, she was alert and awake, with painful uterine contractions every 3 to 4 min. Her vital signs were normal and she was afebrile. The results of heart and lung examinations were normal, and her abdomen was gravid and nontender. She had pooling of clear amniotic fluid in her vaginal vault, and her cervix was dilated between 8 and 9 cm. Fetal heart sounds were present at a rate of approximately 160 beats per min, with variable decelerations during contractions. Routine laboratory testing that was performed at the time of admission included a normal white blood cell count as well as a negative rapid plasma reagin test, hepatitis B surface antigen test, human immunodeficiency virus serology, cervical swabbing for Neisseria gonorrhoeae and Chlamydia trachomatis for a nucleic acid amplification test, urine toxicology screening, and vaginal culture testing for group B streptococcus. Except for the white blood cell count, none of these test results were available until at least 1 day following admission of the patient.
The mother was admitted with anticipated vaginal delivery, and at 24 h after the rupture of her membranes, she was given 2 g of intravenous ampicillin with plans to continue this at a dose of 1 g every 4 h for the prevention of group B streptococcus early onset disease. Approximately 2 h after her initial dose of ampicillin (26 h after membrane rupture), she delivered a live female infant. Although the mother later required surgical dilatation and curettage for removal of a retained placenta, antibiotics were discontinued after her delivery and she was discharged on her fourth hospital day. Blood cultures (collected from the mother only after testing of the infant's culture returned a positive result) were negative.
The child weighed 1,410 g and had Apgar scores of 1 at 1 min, 3 at 5 min, and 7 at 10 min; she had poor tone and no spontaneous respirations, so she was intubated in the delivery room and taken to the neonatal intensive care unit. Although there were no findings specific for sepsis and her white blood cell count was normal (5,100), the infant's prematurity and history of intrapartum antibiotic exposure led to the initiation of intravenous ampicillin and gentamicin administration on the day of birth (after a blood culture was obtained). After 2 days, the child was switched to vancomycin when blood culture results indicated the presence of penicillin-resistant S. pneumoniae. The child's chest radiograph was normal, and with receipt of a surfactant for suspected respiratory distress syndrome, she was successfully weaned from the ventilator by her third day of life. Vancomycin was continued to complete a 7-day course; a repeat blood culture gave negative results. The child's course was further complicated by hyperbilirubinemia and metabolic abnormalities requiring 12 days of parenteral nutrition followed by nasogastric feeds. She was discharged home to the care of her mother on the 35th day of hospitalization.
Cell cultures from a vaginal swab collected from the mother were both grown on Trypticase soy agar containing 5% sheep blood agar and LIM broth (BBL-Becton Dickinson Bioscience, Cockeysville, Md.) and incubated in 5% CO2 at 35°C for 24 h. Unless otherwise specified, group B streptococcus was the primary organism targeted for complete workup of vaginal cultures in this laboratory. Blood cultures were collected from the newborn (BacTec Peds Plus) and the mother (BacTec Plus aerobic; Becton Dickinson Bioscience) and incubated in the BacTec 9240 instrument for continuously monitoring blood culture (Becton Dickinson Bioscience). The newborn's blood culture signaled positive in less than 24 h; the mother's blood culture was negative after 5 days of incubation. An aliquot of the positive blood culture was Gram stained, and subcultured to sheep blood agar. The Gram staining revealed gram-positive cocci in pairs and short chains. On subculture, colonies suspected of consisting of S. pneumoniae organisms were Gram stained and confirmed as S. pneumoniae by the bile solubility test. However, because the child's blood culture was flagged as positive prior to the mother's vaginal culture having been completed and group B streptococci were not recovered from the LIM broth, the blood agar plate containing the vaginal specimen was reviewed specifically for the presence of S. pneumoniae and was noted to contain a confluent growth of colonies suspected of harboring S. pneumoniae.
Following the selection of representative colonies, identification of S. pneumoniae was confirmed by Gram staining and a positive bile solubility result. Antimicrobial susceptibility testing was performed on both isolates, and the respective drug MICs for each were determined with Microscan MICrostrep panels (DadeBehring, Sacramento, Calif.). Both isolates were resistant to penicillin (MIC, 4.0 μg/ml), intermediately susceptible to cefotaxime and ceftriaxone (MIC, 1.0 μg/ml), and susceptible to vancomycin (MIC, 0.25 μg/ml) per National Committee for Clinical Laboratory Standards criteria (14).
The isolates were both serotyped and found to be of serotype 19F. Repetitive sequence PCR (rep-PCR) was used to produce molecular fingerprints of the respective isolates as previously described (15). By this technique, the two isolates possessed identical rep-PCR patterns (Fig. (Fig.1).1). In a follow-up to this case, a total of 100 sequential vaginal cultures performed for group B streptococcus screening in our laboratory were examined specifically for colony morphologies suggestive of the presence of S. pneumoniae; none were found.
The prevention of group B streptococcus neonatal sepsis through widespread implementation of maternal screening guidelines (5) has been heralded as a major public health prevention success (7, 17, 18). However, bacterial pathogens other than Streptococcus agalactiae can also spread from mother to child at the time of birth and result in neonatal pneumonia, meningitis, sepsis, and even death (13, 19). Streptococcus pneumoniae is an uncommon cause of serious, invasive disease during the neonatal period (1, 4, 9, 10, 12, 16). However, much like group B streptococcus, it may cause early onset infection and has been suspected to spread from mothers who are colonized in the vagina at the time of delivery (4, 16). Although Geelen and colleagues reported the largest series of seven cases (6), there exist other series consisting of two (12) or three cases (4) in addition to multiple single-case reports (1, 9, 10, 16). Reviews of the medical literature up to 1990 identified between 23 and 32 reported cases of neonatal sepsis due to S. pneumoniae (11, 20); however, a more recent review of all neonatal infections up to and including 1997 identified a total of 171 cases (8).
We report a recent case of neonatal sepsis due to a penicillin-resistant strain of S. pneumoniae and provide molecular evidence to suggest that the organism spread from mother to child around the time of delivery. Although other investigators have found evidence to suggest that S. pneumoniae is acquired from the mother during childbirth in a manner similar to that by which group B streptococcus is acquired (6, 10), this is the first case in which molecular methods were employed to confirm clonal spread of a drug-resistant strain from mother to child. Moreover, although there is at least one previous report of neonatal disease due to drug-resistant S. pneumoniae (16), our case is unique in that it may indicate the failure of intravenous antibiotics administered during labor to prevent neonatal sepsis due to penicillin-resistant S. pneumoniae.
We do not know the full extent of the mother's risk factors for colonization with drug-resistant S. pneumoniae. Although it is conceivable that the metronidazole she was taking for bacterial vaginosis led to vaginal overgrowth of aerobic flora (2), use of this antibiotic has not been identified as a risk factor for drug resistance in S. pneumoniae. It is unknown whether she received any other antibiotics during her pregnancy. Although she did not personally work at a day care or nursery, it is unknown whether her 2-year-old son was exposed to such settings and could have exposed her to drug-resistant S. pneumoniae.
The use of intravenous antibiotics during labor is recommended for women who either are found upon routine vaginal screening at 35 to 37 weeks of gestation to be colonized with group B streptococcus or, in cases in which screening results are unavailable, are known to have risk factors for neonatal disease (5). Risk factors include delivery at less than 37 weeks of gestation, prolonged rupture of amniotic membranes (≥18 h), and intrapartum fever (≥38°C). This case highlights the concern that increasing efforts to prevent group B streptococcus early onset disease may lead to an increase in neonatal infections due to resistant organisms.
Although previous reports suggest that S. pneumoniae is responsible for only 1 to 9% of all cases of neonatal sepsis (8, 11), there is some concern that this disease is increasing in incidence (8, 12). S. pneumoniae has in the past been reported only rarely as part of the normal vaginal flora; however, changes in sexual practice during pregnancy (i.e., increased orogenital sex) and improved isolation techniques for S. pneumoniae have been suggested as potential causes for an increase in neonatal disease (8). It also appears plausible that increased use of antimicrobials for the prevention of group B streptococcal disease can select for more-resistant pathogens such as S. pneumoniae, as has been suggested for gram-negative pathogens (13, 19).
It is interesting that if the incidence of neonatal sepsis due to S. pneumoniae does increase to the point where specific prevention measures become justified, maternal immunization with pneumococcal vaccine could prove useful. Although neonates born to mothers with low levels of antibodies against group B streptococcus have an increased risk of disease, the use of vaccination in the prevention of neonatal disease has been thus far hampered by the lack of an effective vaccine (3). If an effective group B streptococcus vaccine were available, its use could reduce the need for intrapartum antibiotics. In contrast, there are available effective pneumococcal vaccines and, given the fact that the maternal and neonate isolates in this case were of a serotype included in the vaccine currently available in the United States (i.e., 19F), it is conceivable that maternal immunization with the 23-valent polysaccharide vaccine (or, potentially, with the protein conjugate vaccine) would have protected the child. At present, however, there are no data to support routine pneumococcal immunization of pregnant women.
One additional issue this case raises is whether clinical microbiology laboratories should report on the presence of S. pneumoniae recovered incidentally on vaginal cultures used to screen for group B streptococcus. Based upon results of our prospective survey of 100 vaginal screening cultures, the recovery of S. pneumoniae from these cultures appears to be infrequent. Therefore, despite the overall rarity of neonatal disease, the fact that S. pneumoniae is an increasingly recognized neonatal pathogen that is infrequently encountered on screening cultures leads us to believe that laboratories should at least consider reporting its presence in a vaginal screening culture performed for group B streptococcus.