- Volume 44, Issue 3, 1996

Staphylococcus aureus is an important cause of bone and joint infections. In recent years, significant changes in the incidence of septic arthritis and osteomyelitis have occurred. Haematogenous osteomyelitis is now less common during childhood, but secondary spread of infection to bone or joint from a contiguous site in adults is increasing in incidence. Infection introduced at the time of surgery or arising by the haematogenous route is a significant complication of prosthetic joint implantation, and the effect of bone cement on local immune function may be important in this setting. Although S. epidermidis is a more common cause of prosthetic joint infection, S. aureus is more difficult to treat. S. aureus produces a number of extracellular and cell-associated factors, but it is unclear what role these have as virulence factors in vivo. Furthermore, it is difficult in animal models to simulate transient bacteraemia followed by non-fulminating septic arthritis or osteomyelitis, as occurs in the patient. Surface factors which may be important in pathogenesis include the cell wall (activates complement and stimulates cytokine release), capsular polysaccharide (promotes adhesion to host cell surfaces), collagen receptors and fibronectin-binding protein. Staphylococcal toxic shock syndrome toxin (TSST-1) and the enterotoxins are super-antigens and have the potential to suppress plasma cell differentiation and antibody responsiveness. TSST-1-positive isolates have been shown to cause more severe joint infection in one animal model, but most other studies to date have focused on in-vitro rather than in-vivo effects. There is little evidence supporting a role for coagulase, lipase and the haemolysins in staphylococcal bone and joint infections. Despite the clinical importance of these infections, surprisingly little is known about pathogenesis at the cellular level. Future research should focus on the role of the host immune system in limiting spread of infection, and the expression of virulence factors in animal or other models incorporating isogenic mutant strains.

Verocytotoxin-producing Escherichia coli O157 (O157 VTEC) has become well recognised as an important enteric pathogen. The number of organisms present in environmental and clinical samples may be low and efforts have been made to increase the sensitivity of O157 VTEC detection. Immunomagnetic separation (IMS) has been shown to improve O157 VTEC detection in bovine faeces and food samples. A milk-borne outbreak of O157 VTEC infection allowed us to compare the isolation rates from human faeces by IMS, direct faecal culture on sorbitol-MacConkey agar and a PCR test for verotoxin gene carriage. Of 142 faecal samples examined, 20 were positive on both direct culture and IMS and a further 13 on IMS alone. Therefore, IMS increased the detection rate of individual cases of O157 VTEC infection and also compared well with PCR. We recommend IMS for use in routine diagnostic laboratories where a more sensitive method than direct faecal culture is required for O157 VTEC isolation.

The results of pulsed-field gel electrophoresis (PFGE) of chromosomal DNA of the same 12 methicillin-resistant S. aureus (MRSA) strains of diverse geographical origin, performed in three different laboratories were compared; one laboratory used field-inversion gel electrophoresis (FIGE), one used contour clamped homogeneous electrophoresis (CHEF) and one used both (all manufactured by BioRad Laboratories Inc., Hercules, CA, USA). No single method produced the maximum number of chromosomal fragments from all isolates. In only four instances were the same number of fragments identified by any two techniques. Although there were similar trends in strain identification the results showed many discrepancies even with a three-band difference rule to discriminate between strains. Plasmids in seven of the isolates produced a fragment, but this did not affect discrimination of the study isolates. There is a great need to standardise methodology and produce a standard set of strains to assist in this process.

In Omori Hospital, Toho University School of Medicine, relatively low-virulence blood isolates, including coagulase-negative staphylococci (CNS), enterococci and non-fermentative gram-negative rods other than Pseudomonas aeruginosa comprised c. 60% of total blood isolates. A retrospective study was conducted to assess their clinical significance by reviewing a total of 91 hospital charts. The physicians’ assessments of these positive blood cultures as recorded in the charts were classified into four categories—sepsis, possible sepsis, contamination and no comment. The episodes classified as sepsis accounted for 5.0-19.6%. These episodes were also evaluated by a graded clinical significance score based on multiple factors, including number of positive cultures and clinical signs. The scores for the 91 episodes covered a wide range from 1 to 9, indicating that both contaminants and causative organisms may have been involved. The episodes judged as sepsis or possible sepsis tended to have higher scores. The scores for the episodes associated with enterococci were also higher than those involving CNS or non-fermentative gram-negative rods. The scores for episodes associated with intravenous hyperalimentation catheters were higher than those not associated with the catheters.

The Genie HIV-1/2 kit (Sanofi Diagnostics Pasteur, Montreal, Quebec), a synthetic-peptide solid-phase enzyme immunoassay, was evaluated as a confirmatory assay for HIV-1 antibodies in comparison with Western blot (BioRad, Hercules, CA, USA) on 50 stored HIV-1 antibody-positive sera and the 137 sera yielding repeated positive results in the conventional EIA screen out of 13405 fresh patient sera from Saskatchewan in 1993. The stored HIV-1-positive sera were uniformly positive in the Genie test. Of the 137 EIA screen-positive sera, 33 were uniformly positive and 64 were uniformly negative in Genie and Western blot; 36 were Genie-negative and indeterminate by Western blot; and four were Genie indeterminate, of which one was negative and three were indeterminate by Western blot. All HIV-1 Western blot-indeterminate and Genie-indeterminate sera were negative in radio-immunoprecipitation assay (RIPA) and Western blot for HIV-1 and HIV-2 antibodies performed by a reference laboratory. Genie gave an accurate definitive result for 97% of EIA positive sera compared with 71% for Western blot. There was excellent correlation between Genie, Western blot and RIPA results. However, the Genie assay was faster, less costly and yielded fewer indeterminate results than Western blot in confirmatory testing for HIV-1 antibodies.

Multi-resistant Klebsiella pneumoniae have recently occured in several nosocomial outbreaks of septicaemia. An animal model resembling the pathophysiology of these infections in man would be very useful. A new model of endogenous septicaemia caused by K. pneumoniae and Escherichia coli strains in mice has been established. The mortality rate of conventional ddY mice given cyclophosphamide (CY) or fluorouracil (5-FU), each 200 mg/kg intraperitoneally, every other day was 70 and 100%, respectively. Pseudomonas aeruginosa septicaemia was observed in all dead mice treated with CY, whereas Enterobacteriaceae, including K. pneumoniae, were isolated from 90% of mice given 5-FU. Specific-pathogen-free mice, decontaminated with ampicillin and ceftazidime, were given multi-resistant K. pneumoniae CF504, CF514 or E. coli CF604, or CF614 carrying CAZ-1/TEM-5 plasmid by oral inoculation. Subsequent dosing with 5-FU induced lethal septicaemia caused by the inoculated strains in most of these mice, whereas CY did not regularly induce septicaemia. This model with 5-FU is considered to resemble closely the situation observed in man and to be beneficial for investigating pathophysiology and therapeutic strategies.

IgA and IgG antibodies to cytoplasmic and secreted antigens of serotypes 4 and 8 of Mycobacterium avium and the percentage of agalactosyl immunoglobulin (%Gal[0]) were measured by ELISA in groups of blood donors, HIV seronegative persons, HIV seropositive persons with CD4+ cell counts >300/mm3 and AIDS patients co-infected with M. avium. No differences were found between the control groups, but HIV seropositive persons were distinguished by their increased %Gal[0] (p < 0.001) and increased IgA titre (p < 0.05) to secreted antigens of both serotypes of M. avium. Patients going on to develop aviumosis differed from other HIV-positive individuals, having more IgA to secreted antigens of serotype 8 (p < 0.03) and more IgG to secreted antigens of both serotypes (p < 0.001). After developing M. avium infection there was a further increase in %Gal[0] (p < 0.0001), IgA titres fell to both types of antigen from serotype 4 (p < 0.01) and sonicate antigen of serotype 8 (p < 0.001) and IgG fell to the secreted antigens of serotype 4 (p < 0.03). One interpretation of these observations is that antibody profiles to M. avium might be used to identify healthy persons at special risk of developing HIV seropositivity, and to identify persons with early AIDS who are likely to develop aviumosis.