Factors influencing HIV-1 banding patterns in miniaturized western blot testing of dried blood spot specimens.

In the HIV Seroprevalence Survey among Childbearing Women (SCBW), antibodies to human immunodeficiency virus type 1 are detected using enzyme immunoassays (EIA) and Western blot (WB) methods modified to accommodate samples of blood dried on special collection paper. Dried blood spot (DBS) eluates positive by EIA are tested by one of two WB methods, the miniblot technique using equipment from Immunetics Corporation and the PBS Integra assay (pageblot) from Genetic Systems. In this report we compared the performance of the two WB methods. The identity and position of the viral proteins on the WB were identified using monoclonal antibodies and monospecific antisera. The blots differed substantially in their composition and concentration of viral glycoproteins. Performance of the WB assays with DBS elution buffers from different EIA kits was equivalent except for samples eluted in the Abbott buffer, which reduced detection of antibodies to the p31, p51, p55, and p66 viral proteins. Case classification of DBS, positive sera, dilution curve samples, and seroconversion panels was equivalent by both tests in the presence of all elution buffers. Proficiency evaluation panels sent to SCBW participating laboratories over a 3-year period were used to note the differences between the two WB methods in detection of antibodies to the viral glycoproteins.

HIV-1 serologic test results for one million newborn dried-blood specimens: assay performance and implications for screening.

In a population-based national survey conducted in 1988-90, more than one million neonatal dried-blood specimens were tested for maternal antibody to human immunodeficiency virus type 1 (HIV-1). Enzyme immunoassays (EIA) and Western blot tests were performed in 20 state laboratories following standardized procedures. The observed predictive value of a repeatedly reactive EIA results closely coincided with that expected on the basis of manufacturer's estimates of test sensitivity and specificity for dried-blood specimens. Of the 2,845 EIA-reactive specimens tested by Western blot, 1,323 (47%) were positive, 1,270 (45%) were negative, and 252 (9%) were indeterminate. False-positive EIA and indeterminate Western blot results occurred at rates independent of seroprevalence. These data help characterize the results to be expected from screening of similar low-seroprevalence populations and constitute a base line for the detection of systematic testing errors.

Prevalence of HIV infection in childbearing women in the United States. Surveillance using newborn blood samples.

A national, population-based survey was initiated in 1988 to measure the prevalence of human immunodeficiency virus (HIV) infection in women giving birth to infants in the United States. Following standardized procedures, residual dried-blood specimens collected on filter paper for newborn metabolic screening were tested anonymously in state public health laboratories for maternal antibody to HIV. As of September 1990, annual survey data were available from 38 states and the District of Columbia. The highest HIV seroprevalence rates were observed in New York (5.8 per 1000), the District of Columbia (5.5 per 1000), New Jersey (4.9 per 1000), and Florida (4.5 per 1000). Nationwide, an estimated 1.5 per 1000 women giving birth to infants in 1989 were infected with HIV. Assuming a perinatal transmission rate of 30%, we estimate that approximately 1800 newborns acquired HIV infection during one 12-month period. Preventing transmission of HIV infection to women and infants is an urgent public health priority.

KIE: The authors present the initial results from a national population-based survey initiated in 1988 to measure the prevalence of HIV infection in women giving birth in the United States. Residual dried-blood specimens collected for newborn metabolic screening were tested anonymously in state public health laboratories for maternal antibody to HIV. Basing their estimates on survey data from 38 states and the District of Columbia, the authors estimate the incidence of HIV infection in infants. They urge making the prevention of transmission of HIV infection to women and infants an urgent public health priority.

Epidemiologic, clinical, and laboratory findings of human ehrlichiosis in the United States, 1988.

In 1988, the Centers for Disease Control and the Oklahoma State Department of Health identified 40 patients who had a fourfold or greater change in antibody titer in response to Ehrlichia canis. The median age of these patients was 42 years, 83% were male, 76% became ill between May and July, and 92% reported recent exposures to ticks. Patients resided in or were exposed to ticks in 14 states, including five where ehrlichiosis had not been reported before 1988. Thirty-four patients (85%) were hospitalized, and many had serious complications, including acute respiratory failure (seven patients), encephalopathy (six patients), and acute renal failure (four patients). Pulmonary infiltrates were demonstrated in 14 patients, cerebrospinal fluid pleocytosis was seen in 10 patients, and elevated levels of serum creatinine were demonstrated in eight patients. Two patients, both of whom had preexisting medical problems, died. Nonhospitalized patients received tetracycline therapy earlier in the course of their illness than hospitalized patients. There was no significant difference in the interval from initiation of antibiotic therapy to the first day of defervescence between patients treated with tetracyclines and those treated with chloramphenicol.

Diagnosis of human ehrlichiosis with the indirect fluorescent antibody test: kinetics and specificity.

Human ehrlichiosis, an acute febrile illness caused by Ehrlichia canis or a closely related rickettsial organism, was first identified in 1986. From 1986 through 1988, sera from 85 patients demonstrated a fourfold rise or fall in antibody titer to E. canis. Seven (22%) of 32 patients initially tested during the first week after onset of illness. 17 (68%) of 25 tested during the second week, and all 18 tested during the third week had titers that exceeded the minimum positive titer of greater than or equal to 80. Of the 85 confirmed ehrlichiosis patients, 31 (36.5%) also had indirect fluorescent antibody titers considered diagnostic of infection with Rickettsia rickettsii, Rickettsia typhi, or Coxiella burnetti, but in most these diagnoses were not supported by epidemiologic, clinical, or serologic evidence. These results emphasize that patients suspected of having a tick-borne infection should be tested for antibodies to E. canis as well as for those to other rickettsiae.

Human ehrlichiosis: prospective active surveillance in febrile hospitalized patients.

Between 1 April 1987 and 30 September 1988 prospective active surveillance of human ehrlichiosis was conducted among febrile patients hospitalized in southeast Georgia. If disease of the immune system, bacterial infection, pneumonia, or surgical illness was documented by the end of the first full day of hospitalization, the patient was excluded. Of 75 patients enrolled in the study, 8 (10.7%) had a fourfold rise or fall in titer to Ehrlichia canis (case-patients). The rate was 5.3 cases/100,000 population. Case-patients were more likely than noncase patients to have become ill in May or June (P = .008) and to report a recent tick bite (P less than .001). At hospital admission, case-patients had lower white blood cell counts (median, 4,450/mm3; P = .001), platelet counts (median, 133,000/mm3 P less than .001), and higher aspartate aminotransferase (median, 68 IU/ml; P = .004) and alanine aminotransferase levels (median, 62 IU/ml; P = .038).

Addition of monoclonal antibodies specific for Rickettsia akari to the rickettsial diagnostic panel.

Monoclonal antibodies were produced from mice infected with Rickettsia akari (the etiologic agent of rickettsialpox) and evaluated for specificity in indirect fluorescent-antibody tests with 23 different rickettsial antigens. Of the nine antibodies that were evaluated, two were specific for R. akari and four reacted with R. akari and all other spotted fever group rickettsiae. The remaining three antibodies reacted with some, but not all, members of the spotted fever group. None of the antibodies reacted with typhus, scrub typhus, trench fever, or Q fever rickettsiae. Adding these antibodies to the list of available diagnostic reagents will facilitate identification of rickettsial diseases, particularly those caused by members of the spotted fever group, where the clinical presentations are similar and the etiologic agents are closely related antigenically.

Prevalence and distribution of spotted fever and typhus infections in Sierra Leone and Ivory Coast.

A serosurvey for evidence of rickettsial infections was conducted in the rural populations of several tropical rain forest areas in Sierra Leone and Ivory Coast. Seropositivity rates were surprisingly high in both countries, with more than 7% of the individuals in some districts having antibodies to spotted fever-group rickettsiae. No significant difference was found in the overall prevalence of diagnostic antibody titers to spotted fever-group rickettsiae in Sierra Leone (5.3%) and Ivory Coast (6.2%). However, there was a significant difference (p less than 0.001) in the prevalence of diagnostic antibody titers to typhus rickettsiae in the two countries. There were no marked geographic differences within either country in overall prevalence of rickettsial infections, but there were possible area differences in specific seropositivity rates to typhus- and spotted fever-group rickettsiae in Sierra Leone. In both countries, age and sex differences were important in determining seropositivity, but there was no indication of an age-sex interaction. In Sierra Leone, 59 of the 80 positive sera (73.8%) were from persons age 15 or above (p less than 0.001), and 50 of the 80 (62.5%) were from males (p = 0.05). In Ivory Coast, 33 of the 37 positive sera (89.2%) were from the greater than or equal to 15-age group, and 28 of the 37 (75.7%) were from males (p less than 0.001 for both age and sex). The identification of specific areas endemic for these rickettsial diseases should facilitate the diagnosis and treatment of patients with rickettsial illnesses in West Africa.

Detection and characterization of mouse monoclonal antibodies to epidemic typhus rickettsiae.

A solid-phase immunofluorometric assay was used to detect mouse monoclonal antibodies to epidemic typhus rickettsiae, Rickettsia prowazekii (the immunizing antigen), and to murine typhus rickettsiae, Rickettsia typhi, a related antigen. Of the 649 hybridoma cultures obtained, 628 contained antibodies either to R. prowazekii or to both R. prowazekii and R. typhi. A total of 72 cultures were cloned by limiting dilution and yielded 137 antibody-producing clones. Of these, 104 produced antibodies specific for R. prowazekii, 22 produced antibodies that reacted with R. prowazekii and R. typhi, and 11 produced antibodies that reacted with R. prowazekii, R. typhi, and R. canada. The immunoglobulin isotypes of the mouse monoclonal antibodies produced were identified by a related indirect immunofluorometric assay technique with fluorescein isothiocyanate-conjugated antisera specific for each isotype. Antibodies were also evaluated by indirect fluorescent antibody tests, and antibodies from selected clones were found to neutralize rickettsial toxic activity in mice.

Evidence of Rickettsia prowazekii infections in the United States.

From January 1976 through January 1979 serum specimens from 1,575 individuals were received at the Center for Disease Control and tested for antibodies to rickettsiae. Of these, sera from eight persons gave serological results indicative of recent infections with epidemic typhus rickettsiae (Rickettsia prowazekii). Five of the persons were from Georgia, and one each was from Tennessee, Pennsylvania and Massachusetts. The illnesses occurred during the winter, chiefly in persons living in a rural environment. The clinical picture was compatible with louse-borne epidemic typhus. There was no apparent contact with human body or head lice, and no cases occurred in patient contacts, indicating that infection was not associated with the classic man-louse-man cycle of epidemic typhus. Two of the eight patients had contact with flying squirrels suggesting that they became infected from this known extrahuman reservoir of R. prowazekii.

Legionnaires' disease: isolation of a bacterium and demonstration of its role in other respiratory disease.

To identify the etiologic agent of Legionnaire's disease, we examined patients' serum and tissue specimens in a search for toxins, bacteria, fungi, chlamydiae, rickettsiae and viruses. From the lungs of four of six patients we isolated a gram-negative, non-acid-fast bacillus in guinea pigs. The bacillus could be transferred to yolk sacs of embryonated eggs. Classification of this organism is incomplete. We used yolk-sac cultures of the bacillus as antigen to survey suspected serum specimens, employing antihuman-globulin fluorescent antibody. When compared to controls, specimens from 101 to 111 patients meeting clinical criteria of Legionnaires' disease showed diagnostic increases in antibody titers. Diagnostic increases were also found in 54 recent sporadic cases of severe pneumonia and, retrospectively, in stored serum from most patients in two other previously unsolved outbreaks of respiratory disease. We conclude that Legionnaires' disease is caused by a gram-negative bacterium that may be responsible for widespread infection.

Recent experience with the complement fixation test in the laboratory diagnosis of rickettsial diseases in the United States.

Sera from patients suspected of having rickettsial infections were tested in the complement fixation test with antigens prepared from the rickettsiae of Rocky Mountain spotted fever (SF), rickettsial pox (RP), murine typhus, epidemic typhus, and from Rickettsia canada (RC). Eight units of antigen were used in all cases and two units in man. Only those patients with antibody titers of 1:16 or higher were included in the study. Largely on the basis of comparative titers, the patients were divided into two groups: 102 with SF and 35 with infections by one of the members of the typhus group. The antibody titers were higher with SF antigen than RP antigen in 72% of the SF patients, and in only two SF patients was the RP titer higher, and then by only one tube (twofold dilution). There seemed little advantage in including the RP antigen in the battery of rickettsial antigens. Cross-reaction with at least one of the typhus antigens was observed in the sera from 64% of the SF patients. It was extensive enough to be confusing (within one tube) in 17% with eight units of antigen, but the differentiation was more distinct with two units of antigen. The cross-reaction with typhus antigens was as frequent in children with SF as it was in adults; thus, it is unlikely that these cross-reactions resulted from previous typhus vaccination. The serological differentiation between murine typhus and epidemic typhus was frequently difficult, but the epidemiological background was distinct. Five patients had higher titers to RC antigen, and four of these may possibly have had RC infections.