We report a case of a low-birth-weight infant with an infection of the urinary tract with Trichomonas vaginalis, who later developed cystic chronic lung disease suggestive of Wilson-Mikity syndrome. Although she had mild respiratory distress syndrome at birth, the extent of the chronic lung disease was out of proportion to the initial illness. We speculate that maternal infection with this organism may have resulted in an inflammatory response that led to its development.
A 956-g, appropriate-for-gestational-age, African–American female was delivered by Cesarean section following 27 5/7 weeks of gestation in breech presentation after a period of advanced cervical dilatation and uterine contractions. Her mother was a 20-year-old gravida 5, para 2022 woman whose prenatal laboratory data were significant for vaginal colonization with Streptococcus agalactiae, treatment for Chlamydia trachomatis, and a history of cocaine and marijuana usage confirmed by urine toxicology. The amniotic membranes were ruptured for 23 days prior to delivery. On admission to the hospital, she was treated with betamethasone and ampicillin/sulbactam prior to delivery. A nonreassuring biophysical profile of 4/10 (0 for fluid; 0 for nonstress test; +2 for gross movements; +2 for fine movements; and 0 for breathing) and vaginal bleeding with suspected abruption resulted in delivery of the infant by Cesarean section. The Apgar scores were 1, 5, and 9 at 1, 5, and 10 minutes of life, respectively.
After delivery, the infant was managed with mechanical ventilation with pressure support and volume guarantee for respiratory distress syndrome. She received exogenous surfactant therapy. On the third day of life, she was extubated, at which time she was placed on nasal continuous positive airway pressure. She was weaned to supplemental oxygen delivered by nasal cannula on the following day. The lung fields showed resolution of the acute lung disease (Figure 1). Apnea of prematurity was treated with methylxanthines. Echocardiography failed to reveal a patent ductus arteriosus. Parenteral nutrition, antibiotics, and fluid therapy were provided. Ampicillin and gentamicin were discontinued after 72 hours because the sepsis screen, consisting of a C-reactive protein, complete blood cell count with differential, and blood culture were not supportive of infection. A sample of sputum obtained on the second day of life through the endotracheal tube was negative for C. trachomatis, Mycoplasma hominis, and Ureaplasma urealyticum. The placental pathology report, however, documented an immature placenta consistent with 27 weeks of gestation with acute amnionitis and funisitis.
On the infant's 15th day of life, a urinalysis was obtained to determine the etiology of leukocytes detected by a routine urine dipstick. A urinalysis collected in a bag was within normal limits except for the presence of numerable red blood cells and 8 to 12 white blood cells per high power field. A repeat sample obtained by urethral catheterization revealed the presence of urine that was yellow in color and hazy in appearance with a specific gravity of 1.020 and pH of 6. It was negative for glucose, bile, ketones, nitrates, and reducing substances, but it contained moderate blood, protein (100 mg/dl), moderate leukocyte esterase, 20 to 25 red blood cells per high power field, 8 to 10 white blood cells per high power field, occasional bacteria, moderate calcium oxalate crystals, and trichomonads (unspun specimen). The bacterial culture was negative and the infant was otherwise asymptomatic. Metronidazole therapy was initiated after discussion with a pediatric infectious disease specialist. Urinalysis was obtained on the infant's 19th day of life, which showed absence of the Trichomonas.
The initial blood urea nitrogen and creatinine values were 24 and 1 mg/dl, respectively, on the second day of life, and 6 and 0.6 mg/dl on the 18th day of life. On the infant's 18th day of life, hyponatremia developed. Sodium chloride therapy was initiated. Urine electrolytes revealed sodium 26 mEq/l, potassium 29 mEq/l, chloride 28 mEq/l, and the urine creatinine value was less than 10 mg/dl. The fractional excretion of sodium was 1.1 and the ratio of urine creatinine to plasma creatinine was 35, the latter value consistent with a prerenal state. The fractional excretion of sodium calculation was equivocal because the urine creatinine was reported as less than 10 mg/dl and, therefore, could not be accurately calculated. Ultrasonography had been performed on the infant's fifth day of life due to the presence of hematuria and excluded the presence of renal vein thrombosis or other abnormalities. Urinalysis at that time revealed a yellow but hazy appearance, negative glucose, negative bile, negative ketones, specific gravity 1.015, small blood, pH 5.5, negative protein, negative nitrate, negative leukocyte esterase, negative Clinitest test, occasional red blood cells per high power field, two to four white blood cells per high power field, and slight mucus, but no trichomonads.
On the infant's 21st day of life, she was transferred to a tertiary care pediatric hospital near her home in order to facilitate parental visitation. Renal ultrasonography and a voiding cystourethrogram were within normal limits. Although she had minimal lung disease and was tolerating enteral feedings at the time of transfer, she developed cystic chronic lung disease suspicious for the Wilson-Mikity Syndrome (Figure 2) and a persistent oxygen requirement delivered by nasal cannula.
DENOUEMENT AND DISCUSSION
Although infection of the vagina with Trichomonas vaginalis is common among pregnant women, it is not commonly cultured from neonates, nor is it a highly suspected pathogen in the intensive care nursery. However, infants with vaginal discharge have been reported to be infected with T. vaginalis,1,2,3,4 and the organism has been cultured from tracheal aspirates in infants with respiratory diseases such as pneumonia.5,6 In one series from Poland, it accounted for 17.2% of cases in infants less than 3 weeks of age who developed vaginal discharge.7 Despite its detection in urine samples, the “pyuria” most likely has its origin in the vagina. It is believed that the effect of maternal estrogens on the vaginal epithelium may predispose newborn female infants to infection.1 Successful treatment has been reported with metronidazole.8,9
We speculate that the development of a “bubbly” radiographic appearance of this patient's lungs, together with persistent oxygen requirement after several weeks of life following a relatively benign acute phase of her illness, is reminiscent of the condition described by Wilson and Mikity10 in 1960, and later discussed by Hodgman et al.11 This late-developing syndrome, characterized by cyanosis, retractions, and tachypnea in small premature infants, was reported to have its onset after several weeks of life in infants without respiratory distress at birth. The “bubbly” radiographic appearance of the lungs may initially contain cysts and streaky infiltrates that later resolve and develop basilar hyperinflation and apical atelectasis. Respiratory mechanics are also altered,12 though not determined in this patient. Although this infant did manifest respiratory distress syndrome, the degree of lung disease as the infant aged appeared out of proportion to the acute phase of her illness. Although T. vaginalis was not isolated from this infant's airway, it is plausible that it contributed to her chronic lung disease by inflammation or other unexplained phenomenon. As maternal infection with this organism is common and is associated with premature rupture of membranes and premature birth in some populations,13 it may be worthwhile to prospectively study its association with chronic lung disease in premature infants with and without respiratory distress syndrome.