Intrauterine Transmission of Coccidioidomycosis

Neonatal coccidioidomycosis, although rare, has been reported.1-9 Aspiration of infectious vaginal secretions during birth appears to be the major mode of transmission.8, 10 Transplacental infection has been thought not to occur because extensive coccidioidal placentitis is found without transmission of disease to the fetus.7, 11, 12 We report a case of neonatal coccidioidomycosis in an infant delivered by cesarean section. There was no labor and fetal membranes were intact at birth, indicating intrauterine acquisition of Coccidioides. Case report.Mother. The infant's mother was a 20-year-old gravida 1, para 0, Hispanic woman. She had a positive vaginal culture swab for group B Streptococcus at 28 weeks of gestation and received 10 days of ampicillin treatment. After stopping the antibiotic she developed a cough, fever and shortness of breath. Parenteral antibiotics were then begun for a presumed community-acquired pneumonia. Because she did not improve an extensive workup was done. This was negative, except for positive serum titers for Legionella and varicella. Titers for Coccidioides immitis were sent but pending. The mother's status deteriorated. She was intubated and, at 31 weeks and 4 days of gestation, transferred to the Medical Center at the University of California at San Francisco. At University of California at San Francisco she was reevaluated for infection and given betamethasone to induce fetal lung maturation. Three days after arrival fetal heart rate decelerations, in conjunction with maternal oxygen desaturation, led to an urgent cesarean section. At delivery the mother was not in labor and fetal membranes were intact. The anterior placenta was removed before delivery of the fetus, who was in breech position. The placenta appeared abnormal and subsequent pathologic examination revealed multiple foci of acute inflammation with numerous Coccidioides spherules. On the day of delivery the mother's qualitative immunodiffusion test for Coccidioides was reported positive, and amphotericin B therapy was started. The next day her serum antibody by complement fixation was reported positive with a titer of 1/64, and a previous endotracheal culture was positive for C. immitis. She also developed oral herpes simplex lesions. During the ensuing weeks, despite intensive medical treatment, her respiratory and overall status declined and she died. On autopsy disseminated coccidioidomycosis was found, with involvement of her lungs, diaphragm, kidneys, adrenal, liver and lymph nodes. Infant. At cesarean delivery the female infant had an estimated weight of 1800 g and a physical examination compatible with a 32-week gestation. She was depressed at birth, made no spontaneous respiratory efforts and was immediately intubated and ventilated. She gradually responded to resuscitation and a dose of naloxone (Apgar scores of 2, 4 and 5 at 1, 5 and 10 min). Her initial chest radiograph showed a diffuse, reticular granular, pulmonary pattern, compatible with mild respiratory distress syndrome. To rule out group B streptococcal or other bacterial infection, a blood culture was done. Ampicillin and gentamicin were given. The chest radiograph cleared rapidly and she was extubated in a day. The blood culture remained negative and antibiotics were discontinued after 72 h. During the next 2 weeks she appeared well. A Coccidioides antibody titer, drawn on the third day of life, was reported positive by complement fixation with a titer of 1/16. This was thought to reflect transplacentally acquired maternal antibody. However, a follow-up chest radiograph, an abdominal sonogram and liver enzymes were normal. Between Days 15 and 16 of life, she became febrile and lethargic and had episodes of apnea and bradycardia. Cerebrospinal fluid (CSF) had normal protein and glucose and one white blood cell per high power field; on smear the cells were mainly lymphocytes and monohistiocytes. The chest radiograph showed possible consolidation and, given her history, she was treated with ampicillin, gentamicin, acyclovir and amphotericin B. By the next day the chest roentgenogram showed a diffuse interstitial pneumonia. Eosinophilia was noted in the peripheral blood. Cultures and tests done on the infant's blood, CSF, nasal and conjunctival swabs were negative for bacteria, Chlamydia and viruses (herpes, cytomegalovirus, respiratory syncytial virus and other respiratory organisms). A deep pharyngeal aspirate was positive by culture for C. immitis, which was identified by morphology and DNA probe. Her CSF fungal cultures remained negative. She was treated with intravenous amphotericin B (1 mg/kg/day). Antibiotics and acyclovir were stopped. Her fever disappeared, activity improved and the chest radiograph normalized within 1 week. Five days after presenting with pulmonary symptoms, she developed discrete, indurated areas of erythema on her right wrist and left heel. These were possibly former iv sites, but the question of cutaneous Coccidioides or erythema nodosum was raised. A wrist biopsy showed inflammation but no organisms; cultures were negative. There was no radiographic evidence of osteomyelitis at any site. Liver function remained normal. However, after 2 weeks of amphotericin B therapy, a deep pharyngeal culture again grew C. immitis. When she became ill her serum was positive by immunodiffusion for complement-fixing IgG antibodies to C. immitis but her CSF was negative. After 2 weeks of amphotericin B therapy quantitative titers of 1/2 were found by immunodiffusion and complement fixation. Later samples also had the same concentration of complement-fixing IgG antibodies. Coccidioidal precipitin IgM was not identified. After 3 weeks of intravenous amphotericin B treatment, the infant's therapy was changed to oral fluconazole (6 mg/kg/day). She was clinically asymptomatic and her serum fluconazole concentration was verified to be in the therapeutic range (15 μg/ml). She was discharged home at 45 days of age. Serology at 3 months of age showed unchanged antibody titers. A deep pharyngeal culture at 4 months of age showed no fungal growth. At last follow-up the infant was 6 months old. She was growing and doing well and was still on fluconazole therapy. Discussion. Prior cases of coccidioidomycosis have been reported in infants <1 month old.1-9 In the majority of cases there has been associated maternal infection.2, 3, 5, 6, 8 Pregnancy is a known risk factor for dissemination of coccidioidomycosis, and in endemic areas disseminated disease leads to maternal mortality rates of 20 to 60%.5, 13 Disseminated disease can involve the placenta, but in a number of documented cases of placentitis, fetal infection did not occur.7, 11, 12 This has led to the concept that Coccidioides does not cross the placenta.7, 8, 11, 12 In the prior neonatal cases of coccidioidomycosis, the reported delivery mode was vaginal.1-3, 6-8Coccidioides is known to infect the female genital tract and has been cultured from the cervix postnatally.8, 10 This has suggested that neonatal infection is acquired at birth, by contact with or aspiration of infectious material originating in the mother's vagina.7, 8, 11 The possibility of in utero infection has been considered but the circumstances of birth for the prior cases has left this unanswered.3, 5, 6 The concept that neonatal infection is acquired at birth is reinforced by the pattern of disease. Neonates have presented with respiratory symptoms, pneumonitis, decreased activity, poor feeding and fever at 5 to 21 days of age.1-9 Most have developed meningitis or disseminated disease, with involvement of the liver, spleen, other visceral organs and lymph nodes.1-4, 6-9 Prior cases of documented neonatal Coccidioides infection have been rapidly fatal.1-8 Three additional infants, diagnosed at 2 to 6 months of age, have survived.14-16 Neonatal infection was presumed in these latter patients based on historical information. Our patient presented with typical symptomatology and C. immitis infection was proved on culture. However, given her birth by cesarean section and lack of labor or ruptured membranes, she could not have acquired the organism from contact with maternal vaginal secretions. She was not discharged from the nursery before becoming ill, and San Francisco is not a usual Coccidioides environment. The normal incubation period is 10 to 16 days (range, 1 week to 1 month).17 Our infant's birth history and illness at 15 days indicate that she was infected in utero, likely close to birth. This was probably from infected amniotic fluid that was swallowed (a normal fetal occurrence) or inhaled (during gasping with the heart rate decelerations). Also of interest were our patient's eosinophilia and the erythema nodosum-like skin lesions, both of which can occur with primary coccidioidomycosis.13, 18 The initial drop in complement-fixing antibody titer was likely a result of loss of maternal antibody. The lack of IgM antibodies was not surprising, because these are only detected transiently, in 75% of primary infections.18 In most of the prior neonatal cases of coccidioidomycosis, the underlying pathogenic process was not recognized until late in the disease or at autopsy.1-9 When disease has been identified in older infants, amphotericin B has been beneficial.9, 15, 16 In neonates with compatible symptoms, the intrauterine transmission of Coccidioides should be considered and appropriate diagnostic and therapeutic measures begun. Valerie Charlton, M.D., M.P.H. Kay Ramsdell, M.S.N. Sally Sehring, M.D. Department of Pediatrics; University of California at San Francisco; San Francisco, CA