The rate of marrow recovery and extent of donor engraftment following transplantation of ex vivo-expanded bone marrow cells are independently influenced by the cytokines used for expansion.

Successful stem cell transplantation depends on cell dose, and this is particularly true for placental/cord blood transplantation in which it has been clearly shown that both the success of engraftment as well as the speed of white cell and platelet recovery are dependent on the nucleated cell dose in the graft. Thus, if stem cell numbers could be increased, the speed as well as the likelihood of engraftment might be improved. We studied the effect of two different cytokine combinations--kit ligand (KL), interleukin-3 (IL-3), and Flt-3 ligand supplemented with thrombopoietin and IL-11 (combination 1) or granulocyte/macrophage colony-stimulating factor (GM-CSF) and G-CSF (combination 2)--for their ability to affect speed and extent of engraftment using limited numbers (5 x 10(4)) of murine bone marrow (BM) light-density (LD) cells or their progeny expanded ex vivo in the presence one or the other cytokine combination for 6 days. With combination 1, we found that speed of platelet recovery was enhanced, but at the expense of white blood cell (WBC) recovery and percent donor engraftment. Furthermore, the cytokine combination that best maintained donor engraftment, combination 2, did so at the expense of platelet recovery. In no case was percent donor engraftment improved over 5 x 10(4) unmanipulated LD BM cells. These results are consistent with the interpretation that immediate recovery of blood cells of different lineages and longterm donor engraftment are separate functions that can be influenced by the choice of cytokines used during the ex vivo expansion process.

Rapid simultaneous diagnosis of infections with respiratory syncytial viruses A and B, influenza viruses A and B, and human parainfluenza virus types 1, 2, and 3 by multiplex quantitative reverse transcription-polymerase chain reaction-enzyme hybridization assay (Hexaplex).

A multiplex quantitative reverse transcription-polymerase chain reaction-enzyme hybridization assay (Hexaplex; Prodesse, Milwaukee) was developed and used to rapidly detect and quantitate RNA of respiratory syncytial viruses A and B, influenza viruses A and B, and human parainfluenza virus types 1, 2, and 3 in nasal wash specimens in a single test. Primers and probes originated from highly conserved regions of each viral genome. Six and a half primer pairs were mixed for the simultaneous detection and quantitation of RNA from seven different respiratory viruses. We tested 109 clinical samples with this assay. Twenty-nine virus culture-positive samples were all positive by Hexaplex. Samples from 40 symptomatic patients were negative by virus culture, but eight of these were positive by Hexaplex. Forty samples from asymptomatic children were negative by both virus culture and Hexaplex. No cross-reactions were noted among 17 different respiratory viruses with use of this assay. Hexaplex was 100% sensitive (95% confidence interval [CI], 0.88-1.0) and 98% specific (95% CI, 0.97-0.99). All eight "false-positive" Hexaplex results (in comparison with negative viral culture results) were for symptomatic patients with low numbers of virus RNA copies. This finding suggests that Hexaplex may be more sensitive than virus culture. Our data demonstrate that Hexaplex is a rapid, sensitive, and specific quantitative test for the diagnosis of infections with these seven common respiratory viruses.

Antigenic structure, function, and evolution of the hemagglutinin-neuraminidase protein of human parainfluenza virus type 1.

Twenty-two monoclonal antibodies directed to the hemagglutinin-neuraminidase protein of human parainfluenza virus type 1 (HPIV-1) were used in competition assays to create an antigenic map of neutralization sites. Eighty-seven clinical strains isolated over 35 years from multiple geographic regions were reacted in ELISA, hemagglutinin-inhibition, and microneutralization assays with these monoclonal antibodies. Together these assays revealed 21 epitopes on five nonoverlapping antigenic sites (I, III-VI) with a sixth (II) bridging site connecting sites I, III, and IV. Only 7 (33%) of these epitopes were conserved among all isolates. Previously described HPIV-1 genotypes were associated with the presence or absence of specific antigenic sites and evidence of probable immune selection within genotypes. Two sites were present on all isolates tested (III, V), and one (VI, genotype A) has not been found for 15 years. Forty hemagglutinin-neuraminidase nucleotide sequences were analyzed in terms of homology, structure, and evolution. These data may be useful in future epidemiologic, therapeutic, or vaccine-related work.

Two distinct human parainfluenza virus type 1 genotypes detected during the 1991 Milwaukee epidemic.

The extent of genetic and antigenic variation found in a population of human parainfluenza virus type 1 (HPIV-1) during a single local epidemic was investigated. Fifteen HPIV-1 strains isolated from children in 1991 were analyzed. Nucleotide sequence variation in the hemagglutinin-neuraminidase protein (HN) gene demonstrated two distinct genotypes (genotypes C and D). Unique patterns were identified involving 62 nucleotide and 10 amino acid positions. These patterns represented 40% of all mutations within the HN gene. The remaining mutations were randomly distributed, and 74% involved only one (55%) or two isolates. Genotypes were statistically different from each other at both the nucleotide (P = 0.001) and amino acid (P = 0.001) levels and demonstrated unique potential N-linked glycosylation patterns. Thirty-eight monoclonal antibodies (MAbs) made to four different viral proteins (22 HN, 2 fusion [F], 1 phosphoprotein, and 13 nucleoprotein) (originating from two different genotypes [genotypes A and D]) were compared for their ability to bind to the clinical isolates in enzyme-linked immunosorbent assays (ELISAs) and hemagglutinin-inhibition (HI) assays. Twenty-one MAbs bound well to all clinical isolates in ELISAs and HI assays. The remaining 17 MAbs showed variation in all four structural proteins. Three HN MAbs demonstrated genotype C- and D-specific antigenic and neutralization differences. Evolutionary analysis using parsimony methods confirmed the differences between the two genotypes. No differences in either clinical presentation or disease severity between the two genotypes were found. Geographically localized HPIV-1 epidemics can be caused by at least two distinct genotypes with minor but specific antigenic changes. The clinical and immunologic roles of HPIV-1 genotypes have not been determined.

Epidemiology and cost of infection with human parainfluenza virus types 1 and 2 in young children.

To determine the morbidity, costs, and epidemiological features of lower respiratory tract infections (LRIs) due to human parainfluenza virus types 1 and 2 (HPIV-1 and HPIV-2), we evaluated 1,213 children < 6 years of age who were seen for LRIs in the emergency room of the Children's Hospital of Wisconsin and/or were admitted to the hospital for LRIs during the fall quarter of 1991. The age, sex, race, and respiratory syndrome were recorded for each child; 158 patients (13%) had respiratory samples cultured for viruses and were followed clinically for the duration of their illness. Caucasian children had croup diagnosed more often than did African-American children (relative risk [RR] = 3.12; 95% confidence interval [CI], 2.43-4.00; P < .001), while African-American children more often had pneumonia (RR = 1.85; 95% CI, 1.36-2.5; P < .001). Forty-five of 70 viruses recovered were HPIV-1 (17 cases) or HPIV-2 (28 cases). Together these two viruses were recovered from 49% of children presenting with croup, 10% of those presenting with bronchiolitis, and 12% of those presenting with pneumonia. Gender- and race-associated differences were documented in the group of children infected with HPIV-2: specifically, this group included more girls than boys (RR = 1.99; 95% CI, 1.02-3.88; P < .04) and more Caucasian than African-American children (RR = 2.64; 95% CI, 1.05-6.63; P = .027). These data extrapolate nationally to approximately 250,000 emergency-room visits and approximately 70,000 hospitalizations due to HPIV-1 and HPIV-2, with a cost of $50 million for the former and $140 million for the latter.

Recovery of human parainfluenza virus types one and two.

The ability to recover human parainfluenza virus types 1 and 2 (HPIV-1, 2) from infected individuals has been highly variable. During the autumn of 1991, 158 nasal wash specimens collected from children with lower respiratory symptoms were split and cultured independently at two laboratories using different tissue culture techniques. Immunofluorescent antibody (IFA) and hemadsorption (HAd) assays were compared for their speed and efficiency in viral detection. 45 isolates [HPIV-1 (17) and HPIV-2 (28)] were recovered by one laboratory and only one (HPIV-2) by the other. IFA was the most sensitive assay detecting 87% of HPIV-1 and 70% of HPIV-2 by the fourth day of culture. HAd assay detected 87% of HPIV-1 isolates by the time they were positive by IFA, but only 35% of the HPIV-2 isolates. Significant methodologic differences between laboratories were then compared simultaneously for effect on virus recovery from culture positive frozen clinical specimens. Recovery of 100% of the isolates was achieved. Factors that contributed to differences in recovery of HPIV-1 and 2 were: (1) primary African green monkey (AGMK) cells were inferior to cynomolgus monkey kidney or LLC-MK2 cells, (2) addition of trypsin to culture medium for AGMK and LLC-MK2 cells enhanced recovery, (3) use of IFA was essential for rapid detection of HPIV-2, and (4) use of microtiter plate culture without specimen dilution enhanced virus recovery. A survey of clinical virology laboratories demonstrated considerable variability in the use of these techniques for routine respiratory virus culture.

Genetic variation and evolution of human parainfluenza virus type 1 hemagglutinin neuraminidase: analysis of 12 clinical isolates.

The extent of genetic variation and evolution in a population of human parainfluenza virus type 1 was investigated. The hemagglutinin neuraminidase genes of 13 isolates collected over a 26-year period were sequenced and compared. All isolates except the 1957 type strain were from a single geographic location and demonstrated significant consistent genetic change from the type strain (47/7 [nucleotide/amino acid] substitutions). Antigenic subgroup A isolates demonstrated minor intragroup differences (9/1 substitutions). However, 18/7 unique substitutions separated subgroup A from B regardless of geographic location or year of isolation. Multiple strains of both subgroups appeared and reappeared over decades with only minor variation. There may be significant genetic differences between clinical isolates based on geographic location, and progressive mutational change may occur. Previously defined antigenic and now genetic subgroups were stable and at least regional in distribution over the period studied. The biologic implications and extent of this variation need further evaluation.