 |
|
Case Reports Indian Pediatrics 2003; 40:261-264 |
|
|
Drug Resistant Neonatal Candida Tropicalis Septicemia. Did it cause Diaphragmatic Hernia? |
|
|
Sourabh Dutta From the Division of Neonatology, Department of Pediatrics, Postgraduate Institute of Medical Education and Research, Chandigarh 160 012, India. Correspondence to: Dr. Sourabh Dutta, Assistant Professor, Deparment of Pediatrics, PGIMER, Chandigarh 160 012, India. E-mail: [email protected] Manuscript received: August 12, 2002; Initial
review completed: August 26, 2002; A full-term, 3 kg baby girl developed early onset Candida tropicalis septicemia. The fungus was resistant to amphotericin B, fluconazole and itraconazole. She developed an acquired diaphrag-matic hernia during the course of the infection. The possible association of the hernia with the fungal sepsis is discussed. She improved on treatment with 4-flucytosine and after being operated for the hernia.
Neonatal systemic fungal infections may be either nosocomial or congenital(1). Noso-comial fungemias usually manifest beyond the first week of life, whereas congenital infections present early, in the setting of maternal vulvo-vaginitis. Drug resistance has been reported, but multi-drug resistance is rare. We report an unusual case of neonatal Candida tropicalis septicemia that had early onset with no apparent predisposition, primary multi-drug resistance and acquired diaphragmatic hernia. Case Report M, a term, 3 kg female, was delivered vaginally to a 23-year-old, immunocompetent mother. The pregnancy and delivery was supervised at another hospital. She had an episode of urinary infection in the seventh month. There was no history of diabetes or candida infection. Membranes ruptured in the second stage of labor. At 3 hours the baby started grunting and became cyanosed. She was started on intravenous fluids and referred to our level III unit. Here she was ventilated and antibiotics were started after drawing cultures. The chest X-ray showed a diffuse haze bilaterally. An echocardiogram showed a normal heart. By day 4, the chest X-ray showed right lower lobe consolidation. The gram stain and culture of tracheal secretions did not yield any organism. A chest ultrasound done the same day showed no pleural effusion or extra-thoracic structures. The diaphragm was intact and normal in position. On day 5 a urine smear for fungal hyphae was reported +++. Intravenous fluconazole was added. An ultrasound abdomen showed no evidence of focal fungal infection. The cerebrospinal fluid did not grow any micro-organisms. Indirect ophthalmoscopy showed no fungal endophthalmitis. The blood culture that had been sent on day 1 of life was reported on day 7 to be growing a yeast species, which was later identified to be Candida tropicalis. As the baby continued to worsen, fluconazole was emperically changed to amphotericin B. The fungal blood culture sent prior to this change grew the same fungus. Since she remained sick, and did not improve on amphotericin, itraconazole was added. A blood culture sent at this time again grew Candida tropicalis, and methicillin resistant staphylococcus aureus for which teicoplanin was added. At the end of two weeks the fungi were all reported to be Candida tropicalis resistant to amphotericin B, itraconazole, fluconazole and ketocona-zole. Investigations for immunodeficiency drew a blank. Since the fungus was sensitive to 5-flucytosine, 100 mg/kg/day oral 5-flucytosine was added to amphotericin B in place of itraconazole. The fungal blood culture became sterile 5 days after starting 5-flucytosine. Urine for fungus was also reported negative on day 20. A cumulative dose of 24 mg/kg of amphotericin B and 2 weeks of oral-5-flucytosine were adminis-tered. Renal and liver functions and hematological parameters were normal. However, she had increasing respiratory distress, and serial chest X-rays showed that the liver had moved up into the right hemithorax. Repeat chest ultrasound on day 22 showed a right-sided diaphragmatic hernia, with the liver herniated into the thorax. She was successfully operated upon the next day. A 5 × 3 cm diaphragmatic defect was repaired. In the meantime, her mother underwent gynecological and dermatological evalua-tions, which did not yield evidence of vaginal, cutaneous or systemic candidiasis. Her vaginal swab culture did not grow any pathogens. There had been no outbreak of fungal infections in the referring hospital before or after this patient. Similarly, in our own unit there had been no patient with fungemia in the preceding months. Our environmental surveillance cultures had also not grown fungi for several months preceding this episode. Surface cultures were taken from all the personnel, but we were not able to isolate any fungus. Fresh environmental cultures drew a blank. She was extubated to CPAP on day 34 of life, oxygen was omitted on day 45 and she was discharged on day 51. At 12 months follow-up, her physical growth was normal, developmentally she was near normal and there were no episodes of recurrent respira-tory infections, wheezing or mucocutaneous candidiasis. Discussion Our patient was a full term, normal weight baby with no immunodeficiency disease. Her blood culture drawn within a few hours of birth grew a multi-drug resistant Candida tropicalis species. The source of the fungus remained a mystery despite a thorough search. Vertically acquired candidemia from the mother could not be established because the mother was completely normal and her investigations for candida were negative. It is difficult to accept that it was nosocomial, for the following reasons: blood culture drawn on day 1 grew the fungus, there was no other patient with a similar infection and environmental cultures and surface cultures from the staff were non-contributory. We were left with a theoretical possibility that the mother had a silent fungal vaginal colonization, which spontaneously cleared. This may have followed the antibiotic course for her urinary infection. However even this cannot explain why a community-acquired fungus should be multi-drug resistant. The second unusual feature was that the fungal species was resistant to 4 anti-fungals, belonging to 2 different classes. There is no reported case of multi-drug resistant Candida tropicalis infection in neonates so far. Among neonatal fungemias due to other fungal species, there are reports of resistance to azoles and rare reports of resistance to amphotericin B. A case of sepsis due to azole-resistant Candida albicans in a high-risk premature infant was reported in 1999. In another report of 3 vertically acquired Candida albicans infections in neonates, one type isolated from a neonate and two types isolated from the mother, were resistant to fluconazole(2). Rown et al.(3) reported high MICs (>8 µg/mL) of fluconazole for seven non-Candida albicans isolates. Yoss et al.(4) reported 2 cases of neonatal fungemias resis-tant to amphotericin B with Trichosporon beigelli. Our patient’s fungal blood cultures became sterile only after 5-flucytosine was added to the treatment regime. Serum concentrations after oral administration of 5-fluorocytosine (dose 25 to 100 mg/kg/day) are known to be detectable in all infants, but there is inter-individual variability in half-life, volume of distribution, and clearance(5). We do not have facilities to monitor flucytosine levels in the blood. In retrospect we feel that the dose of 10 mg/kg/day for 2 weeks was adequate because our patient improved without relapse. The course of the index patient was complicated by an unanticipated problem-a right-sided diaphragmatic hernia. The chest X-rays on the first few days of life and an ultrasound done on day 5 showed no evidence of a diaphragmatic hernia. This suggested that it was an acquired diaphragmatic hernia (ADH) rather than congenital (CDH). The infection with which ADH has been associated most frequently in literature is group B streptococcus (GBS) infection(6-7). In the reported cases of ADH secondary to GBS infection, the hernia is invariably right sided. After a GBS infection, an abnormal shadow in the right lower zone is seen on the chest X-ray and an elevation of the bowel and liver into the right hemi-thorax gradually appears. Postoperative result is generally excellent. The pathogenesis of ADH follow-ing GBS sepsis is not well explained. The clinical characteristics of the ADH in our case followed exactly the same course as the cases following GBS infection. There is no case of ADH following a fungal infection reported in literature so far. Since we were able to isolate only Candida tropicalis in the initial days of life, we propose that the ADH was related to the fungal infection. It is instructive to note that there are reports of 2 cases of non-GBS infections associated with ADH, the clinical picture being similar to the one described with GBS(8,9). A late presentation of unrelated CDH was also considered, but the point against this possibility is that late presenters are usually radiologically evident from birth. So far, there is only one confirmed case of possible late radiological presentation of CDH. This was a left-sided CDH in a 9-month-old boy. His chest X-ray after birth was entirely normal and there was no other explanation for the dia-phragmatic hernia that manifested later(10). This alternative but rather remote possibility, that a right sided CDH in our patient was radiologically silent in the initial weeks of life, cannot be excluded with certainty. We conclude that our newborn patient had a primary multi-drug resistant Candida tropicallis infection, the first case of its kind. The infection responded to a combination of 5-flucytosine and amphotericin B. There are reasons to believe that the patient developed an acquired diaphragmatic hernia, possibly related to the fungal sepsis.
|
