Determination of the etiology of seroreversals in HIV testing by antibody fingerprinting.

Fifteen sets of seroreversal serum samples (i.e., an initial specimen confirmed as anti-human immunodeficiency virus, HIV, positive and a 2-4-week follow-up specimen that is anti-HIV negative) were identified from 711,684 military members tested by the U.S. Army HIV screening program from January 1990 to May 1991. We utilized the Miragen Antibody Fingerprinting Assay to determine what proportion of these seroreversals was caused by different or discordant patient samples as opposed to laboratory testing errors. Forty-two percent of these seroreversals demonstrated substantially discordant immunoblot strips representing serum samples from different patients. We recommend utilization of this antibody fingerprinting assay to help determine the etiology of HIV seroreversal specimens.

Acidification modified p24 antigen capture assay in HIV seropositives.

The p24 antigen is only detectable in a minority of asymptomatic HIV infected individuals, which is in part felt to be secondary to immune complex formation with p24 antibody. We sought to determine if the recovery of p24 antigen could be enhanced by an acidification procedure in an attempt to dissociate p24 antibody-antigen complexes. We tested 291 HIV antibody positive serum samples in duplicate, comparing the standard Coulter p24 Antigen Assay with the same assay which was modified by pretreating samples for 60 min with 1.5 M glycine (pH 1.85) and then neutralizing the sample with 1.5 M Tris (pH 9.0). Using the assay cutoff (approximately 7.8 pg/ml), the standard assayed sera and the acid pretreated sera were positive in 65/291 (22.3%) and 167/291 (57.4%) of the samples, respectively. This difference between procedures was statistically significant (p < 0.0001) by McNemar chi 2. There was also a statistically significant difference between Walter Reed stages for p24 antigen quantification by ANOVA (p < 0.0001). We have demonstrated that the detection of p24 antigen may be significantly enhanced by acid pretreatment of sera which dissociates immune complexed antigen. We have also shown significant differences exist between Walter Reed stages for quantitative p24 antigen levels. Enhanced detection of p24 antigen may be important prognostically and may allow for the monitoring of antiretroviral therapy.

Comparison of standard and quantitative blood cultures in the evaluation of children with suspected central venous line sepsis.

We reviewed our experience with paired quantitative and standard blood cultures in the evaluation of children with suspected central-line sepsis with the hypothesis that by employing both systems we would increase our yield of pathogenic isolates. A total of 913 paired cultures were reviewed, representing 267 pathogenic isolates and 58 individual episodes of sepsis. The isolates were analyzed for recovery rates for each system and by combining both systems. The Isolator system proved to be equal to the BACTEC system for the recovery of all groups of pathogenic isolates. The combined use of both the quantitative and the standard culture systems demonstrated a statistically significant advantage (p less than 0.001) for the recovery of pathogens as compared with either system alone. The use of either system alone would have missed 15% of the total pathogenic isolates. Quantitative colony counts were helpful in identifying the line as the source of infection in 35 to 58 episodes of sepsis and were often beneficial in the clinical management central venous line infection. We recommend the use of the Isolator 1.5 ml combined with a conventional broth-bottle system in selected pediatric patients to enhance the recovery of pathogenic organisms.

Incidental paranasal sinus abnormalities on CT of children: clinical correlation.

The paranasal sinuses were prospectively evaluated by CT, clinical history, and physical examination in infants and children having cranial CT for indications unrelated to upper respiratory inflammation (URI). One hundred and one CT scans were studied, and sinus abnormalities were detected in 18% of patients older than 1 year and without signs or symptoms of URI. When signs and/or symptoms of recent URI were present, the incidence of abnormalities was 31%. Maxillary antral were not identifiable or were opacified in 72% of all infants under 1 year old. Because of the high incidence of sinus abnormalities on CT in children with and without evidence of recent URI, abnormalities should not be ascribed to sinusitis without close clinical correlation.

Failure of intrapartum antibiotics to prevent culture-proved neonatal group B streptococcal sepsis.

Early-onset group B streptococci (GBS-EOS) sepsis may be prevented by intrapartum antibiotics administered for GBS maternal colonization, premature labor, or prolonged rupture of membranes. We sought to identify cases of neonatal GBS sepsis after apparent failure of intrapartum chemotherapy and to determine the factors associated with failure of intrapartum antibiotics in these cases. We identified 96 GBS blood culture-positive infants at five military medical centers from 1987 to 1990. Eighteen (18.7%) of these infants had mothers who had received intrapartum antibiotics; 16 of 18 cases were early-onset disease, 15 of which initially had symptoms at less than 1 hour of age. Two infants had late-onset disease develop at 3 weeks of age. At least one perinatal risk factor (prematurity, prolonged rupture of membranes > 12 hours, maternal fever) was present in each of the 16 cases. Indications for intrapartum antibiotics were suspected chorioamnionitis (13 cases), GBS colonization and prolonged rupture of membranes or prematurity (3), and GBS colonization alone (2). Maternal antibiotics included ampicillin (14 cases), cephadyl (1), vancomycin (1), clindamycin (1), and gentamicin alone (1). The median number of doses of ampicillin before delivery was 1 (range, 1 to 21), which was administered at a median of 4 hours (range, 1 to 84) before birth. The mean dose of ampicillin was 1.8 gm/dose (range, 1 to 2 gm/dose). Two of 16 (12.5%) infants with GBS-EOS died as a result of GBS sepsis. In our population of neonates with GBS-EOS, 18.4% (16 of 87) of the infants had positive blood cultures despite intrapartum antibiotics. Intrapartum antibiotics may fail to prevent GBS sepsis in a number of infants born to mothers colonized with GBS or to those with acute chorioamnionitis.

Persistent catheter-related bacteremia: clearance with antibiotics and urokinase.

A patient with terminal osteogenic sarcoma and catheter-related coagulase-negative staphylococcal bacteremia was treated with vancomycin and blood cultures were positive for 4 days documented with quantitative colony counts. Urokinase therapy was initiated and was associated with a transient bacteremia with markedly increased colony counts of coagulase-negative Staphylococcus prior to eradication of the catheter-related infection. We feel that the combination of urokinase and appropriate antibiotics may be an effective method to eradicate line-associated coagulase-negative staphylococcal infection in selected patients.

False-negative urine latex particle agglutination testing in neonates with group B streptococcal bacteremia. A function of improper test implementation?

Although blood cultures remain the most specific indicator of Group B streptococcus (GBS) sepsis, a potentially life-threatening infection in neonates, test results may not be available for 24 to 48 hours. Detection of GBS antigen in the urine by latex particle agglutination (LPA) may speed diagnosis. This study analyzed the sensitivity of the GBS urine LPA assay under clinical conditions. The urine of neonates with early-onset GBS bacteremia was analyzed for GBS antigen over a three-year period at six military medical centers. Overall, 53.5% (38/71) of infants with positive blood cultures had a positive urine LPA test. Only one medical center routinely followed manufacturer's recommendations to concentrate urine specimens before testing. These data suggest that the sensitivity for the urine LPA assay, when performed on unconcentrated urine, is lower than previously reported. Clinicians should insist that the laboratory maximize sensitivity by concentrating urine prior to GBS LPA testing.

Clinical and laboratory predictors of bacterial diarrhea in a tropical environment.

A total of 180 patients with diarrhea, 6 years and below, were seen and evaluated in our pediatric clinic during the months of July to December, 1986. A bacterial pathogen was isolated from the stools in 24/180 (13%). The best historical factors for predictive accuracies were abrupt onset, no vomiting before the onset of diarrhea, and greater than four stools per 24 hours. The presence of fecal leukocytes was the best screening tool for stool culture positivity. The patients identified with the above three historical factors and positive for fecal leukocytes had an 83% probability of having a positive bacterial stool culture compared to only a 5% probability if any one of the factors was absent. We were able to identify a subpopulation of American dependents in a tropical environment with diarrhea who had a high probability of having a bacterial stool pathogen based on historical factors and the fecal leukocyte test.

The use of intravenous gammaglobulin in dengue fever, a case report.

We managed a serologically documented case of dengue fever with thrombocytopenia with IV IgG. Acute and convalescent ELISA and HI titers indicated an acute dengue infection with a secondary response that clinically, and by laboratory parameters dramatically improved after IV IgG administration. The use of IV IgG in cases of thrombocytopenia associated with dengue has both theoretical advantages and disadvantages. IV IgG may have a role in the management of DHF/DSS because of the significant morbidity and mortality associated with DHF/DSS. A prospective randomized study should be developed to determine if IV IgG does indeed effect the course of thrombocytopenia in dengue infections.