Failure of bacterial screening to detect Staphylococcus aureus: the English experience of donor follow-up.

BACKGROUND AND OBJECTIVES: Between February 2011 and December 2016, over 1·6 million platelet units, 36% pooled platelets, underwent bacterial screening prior to issue. Contamination rates for apheresis and pooled platelets were 0·02% and 0·07%, respectively. Staphylococcus aureus accounted for 21 contaminations, including four pooled platelets, one confirmed transfusion-transmitted infection (TTI) and three 'near-miss' incidents detected on visual inspection which were negative on screening. We describe follow-up investigations of 16 donors for skin carriage of S. aureus and molecular characterisation of donor and pack isolates. MATERIALS AND METHODS: Units were screened by the BacT/ALERT 3D detection system. Contributing donors were interviewed and consent requested for skin and nasal swabbing. S. aureus isolates were referred for spa gene type and DNA macrorestriction profile to determine identity between carriage strains and packs. RESULTS: Donors of 10 apheresis and two pooled packs screen positive for S. aureus were confirmed as the source of contamination; seven had a history of skin conditions, predominantly eczema; 11 were nasal carriers. The 'near-miss' incidents were associated with apheresis donors, two donors harboured strains indistinguishable from the pack strain. The TTI was due to a screen-negative pooled unit, and a nasal isolate of one donor was indistinguishable from that in the unit. CONCLUSION: Staphylococcus aureus contamination is rare but potentially harmful in platelet units. Donor isolates showed almost universal correspondence in molecular type with pack isolates, thus confirming the source of contamination. The importance of visual inspection of packs prior to transfusion is underlined by the 'near-miss' incidents.

Microbiological risk assessment for personal care products.

Regulatory decisions regarding microbiological safety of cosmetics and personal care products are primarily hazard-based, where the presence of a potential pathogen determines decision-making. This contrasts with the Food industry where it is a commonplace to use a risk-based approach for ensuring microbiological safety. A risk-based approach allows consideration of the degree of exposure to assess unacceptable health risks. As there can be a number of advantages in using a risk-based approach to safety, this study explores the Codex Alimentarius (Codex) four-step Microbiological Risk Assessment (MRA) framework frequently used in the Food industry and examines how it can be applied to the safety assessment of personal care products. The hazard identification and hazard characterization steps (one and two) of the Codex MRA framework consider the main microorganisms of concern. These are addressed by reviewing the current industry guidelines for objectionable organisms and analysing reports of contaminated products notified by government agencies over a recent 5-year period, together with examples of reported outbreaks. Data related to estimation of exposure (step three) are discussed, and examples of possible calculations and references are included. The fourth step, performed by the risk assessor (risk characterization), is specific to each assessment and brings together the information from the first three steps to assess the risk. Although there are very few documented uses of the MRA approach for personal care products, this study illustrates that it is a practicable and sound approach for producing products that are safe by design. It can be helpful in the context of designing products and processes going to market and with setting of microbiological specifications. Additionally, it can be applied reactively to facilitate decision-making when contaminated products are released on to the marketplace. Currently, the knowledge available may only allow a qualitative or semi-quantitative rather than fully quantitative risk assessment, but an added benefit is that the disciplined structuring of available knowledge enables clear identification of gaps to target resources and if appropriate, instigate data generation.

Bacillus cereus in personal care products: risk to consumers.

Bacillus cereus is ubiquitous in nature and thus occurs naturally in a wide range of raw materials and foodstuffs. B. cereus spores are resistant to desiccation and heat and able to survive dry storage and cooking. Vegetative cells produce several toxins which on ingestion in sufficient numbers can cause vomiting and/or diarrhoea depending on the toxins produced. Gastrointestinal disease is commonly associated with reheated or inadequately cooked foods. In addition to being a rare cause of several acute infections (e.g. pneumonia and septicaemia), B. cereus can also cause localized infection of post-surgical or trauma wounds and is a rare but significant pathogen of the eye where it may result in severe endophthalmitis often leading to loss of vision. Key risk factors in such cases are trauma to the eye and retained contaminated intraocular foreign bodies. In addition, rare cases of B. cereus-associated keratitis (inflammation of the cornea) have been linked to contact lens use. Bacillus cereus is therefore a microbial contaminant that could adversely affect product safety of cosmetic and facial toiletries and pose a threat to the user if other key risk factors are also present. The infective dose in the human eye is unknown, but as few as 100 cfu has been reported to initiate infection in a susceptible animal model. However, we are not aware of any reports in the literature of B. cereus infections in any body site linked with use of personal care products. Low levels of B. cereus spores may on occasion be present in near-eye cosmetics, and these products have been used by consumers for many years. In addition, exposure to B. cereus is more likely to occur through other routes (e.g. dustborne contamination) due to its ubiquity and resistance properties of spores. The organism has been recovered from the eyes of healthy individuals. Therefore, although there may be a perceived hazard, the risk of severe eye infections as a consequence of exposure through contaminated near-eye cosmetics is judged to be vanishingly small. It is unlikely that more stringent microbiological standards for near-eye cosmetics will have any impact on the risk of severe eye infections caused by B. cereus, as these are not linked to use of personal care products.

Activity of four antimicrobial cocktails for tissue allograft decontamination against bacteria and Candida spp. of known susceptibility at different temperatures.

Several antimicrobial cocktail solutions of differing composition and concentrations are widely used to decontaminate viable banked tissue allografts at different temperatures and times of exposure. We compared the efficiency of four cocktails comprising nine antimicrobials to kill suspensions of a panel of 27 strains of 13 bacterial species, and 3 Candida spp. at 4, 22 and 37 °C for 24 h. All but one bacterial strains were susceptible to one or more of the agents tested individually at concentrations at least fourfold below the recommended susceptibility breakpoint minimum inhibitory concentrations for drug/species combinations. Candida lusitaniae was resistant to nystatin and amphotericin. The concentrations of several of the cocktail constituents were often greatly in excess (50-1,000-fold) of that required to inhibit the growth of susceptible strains. All cocktails were ineffective against a pan-resistant strain of Enterococcus faecium and one of the four cocktails failed to kill two strains of methicillin resistant Staphylococcus aureus. Each cocktail was most efficient at 37 °C, less so at 22 °C, and poorly active at 4 °C. We conclude that the practice of decontamination of tissues with antimicrobials at low temperatures is not supported by in vitro susceptibility tests.

Variable number tandem repeat loci providing discrimination within widespread genotypes of Acinetobacter baumannii.

Some genotypes of Acinetobacter baumannii, defined by pulsed-field gel electrophoresis (PFGE), have been found in many hospitals. Our aim was to find variable number tandem repeat (VNTR) loci capable of providing discrimination among isolates with highly similar or identical PFGE profiles, to gain insights into the epidemiology. Thirteen loci identified in A. baumannii ATCC 17978 were tested using a panel of isolates that included multiple representatives of genotypes belonging to the three European clonal lineages. Two loci, with repeat units of 9 and 6 bp respectively were selected. Repeat numbers varied between 3 and 29, and 9 and 26 respectively at the two loci. The repeat numbers of representatives of each genotype often differed between hospitals, providing a means of tracking patient transfers and possible transmissions between patients. The results suggest that this analysis accurately reflects the known epidemiological information, and provides a valuable tool for cross-infection studies.

Transmission of Pseudomonas aeruginosa in children with cystic fibrosis attending summer camps in The Netherlands.

BACKGROUND: Cross-infection of Pseudomonas aeruginosa has been reported to occur at holiday camps for children with Cystic Fibrosis (CF) with varying frequency. The study aimed to establish the degree of transmission resulting in subsequent infection of P. aeruginosa among CF children (n=80) attending holiday camps in The Netherlands. METHODS: The study was performed in the summer of 2001 in four camps organised simultaneously at different locations. Sputum was collected on day 1 of the holiday, and three and six months later. Different morphotypes of P. aeruginosa from sputum were genotyped by AFLP analysis. Criteria were defined for the degree of evidence of transmission. RESULTS: There were 18 cases possible, 2 cases of probable transmission and 1 case of highly probable transmission. Two predominant types of P. aeruginosa were found (types 18 and 23). Type 18 was already prevalent on day 1 mostly in younger children and was involved in eleven cases of transmission; type 23 was involved in six cases of transmission among older children. CONCLUSIONS: There was a considerable risk of transmission of P. aeruginosa during holiday camps for CF children in The Netherlands. Two genotypes of P. aeruginosa appeared to be easily transmissible, one of which seemed common in the Dutch CF population.

Use of sequence-based typing and multiplex PCR to identify clonal lineages of outbreak strains of Acinetobacter baumannii.

Representatives (n = 31) of outbreak strains of Acinetobacter baumannii from five countries fell into three clear groups, designated Groups 1-3, based on their ompA (outer-membrane protein A), csuE (part of a pilus assembly system required for biofilm formation) and bla(OXA-51-like) (the intrinsic carbapenemase gene in A. baumannii) gene sequences. With the exception of the closely related alleles within the Group 1 clonal complex, alleles at each locus were highly distinct from each other, with a minimum of 14 nucleotide differences between any two alleles. Isolates within a group shared the same combination of alleles at the three loci, providing compelling evidence that the outbreak strains investigated belonged to three clonal lineages. These corresponded to the previously identified European clones I-III. Sequence differences among the alleles were used to design multiplex PCRs to rapidly assign isolates belonging to particular genotypes to sequence groups. In the UK, genotypes belonging to the Group 1 clonal complex have been particularly successful, accounting for the vast majority of isolates referred from hospitals experiencing problems with Acinetobacter.

Characterisation of Exiguobacterium aurantiacum isolates from blood cultures of six patients.

Exiguobacterium spp. are alkaliphilic, halotolerant, non-spore-forming Gram-positive bacilli, hitherto uncharacterized from human infections. Six isolates of Exiguobacterium aurantiacum were obtained from patients with bacteraemia, three of whom had myeloma. All isolates formed orange-yellow pigmented colonies on blood agar, were catalase- and DNase-positive, and grew on nutrient agar at pH 10 and in the presence of NaCl 6% w/v. The six isolates were susceptible to all antimicrobial agents tested and were uniform in their fatty acid and mass spectrum profiles.

Development of extended-spectrum activity in TEM beta-lactamases in hyper-mutable, mutS Escherichia coli.

TEM-1 and TEM(pUC19)beta-lactamases can gain activity against ceftazidime and other expanded-spectrum cephalosporins via point mutation. The frequency of emergent resistance to ceftazidime at 4 x MIC was elevated >or= 250-fold in hyper-mutable, MutS-deficient Escherichia coli harbouring these beta-lactamase genes on high- or low-copy plasmids. Moreover, although ceftazidime-resistant mutants, or those with reduced susceptibility, were selected in both the wild-type and mutS hosts, many more mutants in the mutS host showed ceftazidimase-type extended-spectrum beta-lactamase (ESBL) activity. This correlated with a G-A point mutation at position 484 in the bla(TEM-1) and bla(TEM-pUC19) genes, conferring the Arg164His amino-acid substitution found in the TEM-29 ESBL. Non-ESBL mutants lacked changes in bla(TEM).

Persistence of Pseudomonas aeruginosa strains in respiratory infection in AIDS patients.

OBJECTIVES: To establish the clinical pattern of Pseudomonas aeruginosa respiratory infections in HIV-seropositive patients and to determine whether repeated isolation of the organism represents reinfection or recurrence and to assess whether common source, nosocomial infection occurred. DESIGN AND METHODS: Evaluation of the clinical pattern of P. aeruginosa respiratory infections by case note review and epidemiological characterization of P. aeruginosa by serotype determination and Xbal DNA macrorestriction analysis. Serum sensitivity testing of strains was performed to further define phenotypic characteristics of the isolated organisms. RESULTS: Seventy-three per cent (29 out of 40) of individuals had P. aeruginosa isolated on two or more occasions in the setting of clinical respiratory infection. Overall, 85% had evidence of P. aeruginosa to within 2 months of study completion or death. Epidemiological characterization revealed persistence of unique single strains in 93% of individuals where multiple isolates were available for testing, whereas only two patients harboured a common strain. The serotype distribution of strains was similar to that reported from non-HIV-positive patients. CONCLUSIONS: Once established, eradication of P. aeruginosa from the respiratory tract of HIV-seropositive individuals with advanced immunosuppression is problematic and a chronic infective state appears common. There was no evidence of nosocomial transmission. Serotype loss and development of sensitivity to normal human serum were both observed and were highly correlated. This represents truncation of O-antigenic lipopolysaccharide on the cell surface of P. aeruginosa and may reflect progression to phenotypes commonly associated with chronic infection in other clinical settings such as cystic fibrosis.

Inhibition of Pseudomonas aeruginosa from cystic fibrosis by selective media.

Pseudomonas Isolation Agar (selective agent, Irgasan, 25 mg/1) and Pseudomonas Selective Agar (selective agents, cetrimide 200mg/1 and nalidixic acid 15 mg/1) inhibited some strains of P aeruginosa from cystic fibrosis sputum but did not inhibit isolates from other sources. Of 200 cystic fibrosis isolates, 22 were inhibited by 16 mg/1 Irgasan, 45 by 8 mg/1 nalidixic acid, and 15 by 128 mg/1 cetrimide. We recommend that cystic fibrosis sputum should be cultured on selective and non-selective media to maximise the isolation of P aeruginosa.

Antibody response of fibrocystic patients to homologous 0-typable and 0-defective isolates of Pseudomonas aeruginosa.

Antibody titres to Pseudomonas aeruginosa of sera from 60 adult fibrocystic patients were determined in an enzyme linked immunosorbent assay (ELISA) with whole cells of homologous isolates which had been classified according to 0-antigen state by their reactivity with 0-typing antisera. Patients who were continuously colonised with Ps aeruginosa gave the highest titres: range 1500-64000 (mean 11000) and 500-48000 (mean 9000) with homologous 0-typable and 0-defective isolates, respectively. Lower titres to both varieties of isolates were obtained with recently colonised patients, and non-colonised patients gave titres with reference laboratory strains marginally above those of healthy controls. Serum titres of patients with sequential isolates were strain dependent and did not correlate with the 0-antigen state of the strain. Sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDS-PAGE) and immunoblot analysis of these sera and strains showed antibody binding primarily to high molecular weight 0-repeating units of lipopolysaccharide. It is concluded that the 0-antigen of the strain of Ps aeruginosa used in the ELISA test does not influence the titre obtained with fibrocystic sera, and it is recommended that serum titres should be assessed with a panel of homologous isolates from patients.

Latex agglutination test for identification of Pseudomonas pseudomallei.

A latex agglutination test was developed and evaluated for the rapid presumptive identification of Pseudomonas pseudomallei, the causative organism of melioidosis. The test was 100% sensitive for 52 isolates of Ps pseudomallei and 100% specific when tested with other medically important Pseudomonas species and Enterobacteriaceae. A subsequent field trial, with clinical specimens from patients with suspected melioidosis, confirmed the sensitivity and specificity of the test.

Evaluation of three oligonucleotide primer sets in PCR for the identification of Burkholderia cepacia and their differentiation from Burkholderia gladioli.

AIMS: To evaluate three oligonucleotide primer pairs--two specific for 16S and 23S rRNA sequences of Burkholderia cepacia, and the third specific for internal transcribed spacer region of 16S-23S sequences of B gladioli--for the identification and differentiation of reference and clinical strains of these and other species. METHODS: The three primers sets were applied in polymerase chain reaction (PCR) to a collection of 177 clinical isolates submitted for identification from diagnostic laboratories as presumed B cepacia. RESULTS: At an annealing temperature of 63 degrees C, all eight B cepacia and four B gladioli reference strains reacted with their specific primers. B vandii was the only other species that was positive with both B cepacia primers but five Burkholderia or Ralstonia species reacted with one of these primers. Seventy eight isolates were typical of B cepacia in biochemical tests and 75 of these reacted with specific primers; three, however, were positive with the B gladioli primers. Fifteen asaccharolytic isolates were confirmed as B cepacia by PCR but other non-fermenting Gram negative species were negative with each of the primers. CONCLUSIONS: PCR using 16S rRNA sequences is recommended for identification of B cepacia that give atypical results in biochemical tests.

Latex agglutination for rapid detection of Pseudomonas pseudomallei antigen in urine of patients with melioidosis.

A latex agglutination test for the detection of Pseudomonas pseudomallei antigen in urine was evaluated for the rapid diagnosis of melioidosis. With unconcentrated urine, antigen was detected in only 18% of patients with melioidosis overall. However, when urine was concentrated 100-fold, antigen was detected in 47% overall and in 67% of patients with septicaemia or disseminated infection, in whom a rapid diagnosis is most important. The specificity of the test was 100%. These results compared favourably with an enzyme immunoassay. This latex agglutination test is a simple, rapid and highly specific method of diagnosing melioidosis, and will be particularly useful in areas with limited laboratory facilities.

Serological relationships of the O antigens of Klebsiella pneumoniae O5, Escherichia coli O8 and a new O serotype of Serratia marcescens.

Klebsiella pneumoniae O5, Escherichia coli O8 and Serratia marcescens 3255 were shown to cross-react in both ELISA and immunoblotting. The cross-reaction appeared to be due to the O antigen of their lipopolysaccharide (LPS). In addition, there was evidence that the reactions of these strains with their homologous antisera were due, in part, to determinants other than O polysaccharide.

Auxotrophy of Pseudomonas aeruginosa in cystic fibrosis.

Seventy-four of 403 (18.4%) sputum isolates of Pseudomonas aeruginosa from 49 of 136 (36.0%) adults with cystic fibrosis (CF) were auxotrophic mutants. Two of 11 (18.2%) isolates of P. aeruginosa taken from patients with non-CF bronchiectasis were also auxotrophic. All 99 strains taken from non-bronchiectatic sources were prototrophic. Forty-six of 55 (83.6%) CF auxotrophs required one or more of 36 growth factors tested; the requirements for the remaining 9 isolates were not identified. Methionine was the sole factor required by 17 of 22 (77.3%) isolated which depended on a single factor. We conclude that auxotrophy is a feature of P. aeruginosa infection in cystic fibrosis.

Immunochemical characterization of two new O serotypes of Serratia marcescens (O27 and O28).

In an earlier study of the distribution of O-serotypes among clinical isolates of Serratia marcescens, two apparently new serotypes were identified, represented by strains S1254 and S3255. Studies using ELISA, immunoblotting and the Quellung reaction have shown that they qualify for inclusion in the O-antigenic typing scheme on three counts: (1) they possess chemically distinct O-antigenic repeating units, (2) the O-antigens are serologically distinguishable from all others, and (3) they are found in a significant proportion of clinical S. marcescens strains (13% and 6% respectively). S1254, the type strain for serotype O27, is an acapsular strain which expressed a glucorhamnan with a disaccharide repeating unit as its lipopolysaccharide side chain. It cross-reacts with serotype O4, the O antigen of which is an O-acetylated form of the O27 glucorhamnan, but this cross-reaction can be eliminated by reciprocal cross-absorption. S3255, the type strain for serotype O28, has a mannose homopolymer as its O-antigen and is the only S. marcescens serotype with a trimeric repeating-unit structure. However, it cross-reacts with the O5 serotype strain due to similarities in their acidic capsular polysaccharides. Cross-absorption and the production of serum to an acapsular variant of serotype strain O28 produced typing reagents which could differentiate serotypes O5 and O28.

Preparation of agglutinating antisera specific for the flagellar antigens of Pseudomonas aeruginosa.

H-specific agglutinating sera to Pseudomonas aeruginosa were prepared by immunisation with partially purified flagella. The results of agglutination and immobilisation tests with rabbit sera prepared against the flagella of six H-type strains showed that the sera had high titres and were H specific. Cross-absorption tests revealed that one strain (H-3) possessed a distinct antigen not present in any of the others. Two groups of strains (H-1, H-2 and H-5; H-4 and H-6) each possessed a distinct major antigen. Members of these two groups could be distinguished by their minor antigens.