The following summary of selected abstracts from 3 meetings will be published in multiple parts. The 43rd Interscience Conference on Antimicrobial Agents and Chemotherapy (ICAAC) met in Chicago September 14-17, 2003. The Infectious Disease Society of America (IDSA) met in San Diego October 9-12, 2003. The American Society of Tropical Medicine and Hygiene met in Philadelphia December 3-7, 2003. — Stan Deresinski, MD, FACPGram-Positive Pathogens
Antibacterials, Staphylococcus, Enterococcus, Group B Streptococcus, Anthrax
Dalbavancin, an investigational glycopeptide with once-weekly dosing, is active against many Gram-positive organisms, including the Pennsylvania vancomycin-resistant S aureus (VRSA) isolate. In a phase 2 randomized trial, dalbavancin, given as a 1-gram IV dose followed by a 500-mg dose 1 week later, appeared to be at least as effective as the standard-of-care comparator agents chosen by the treating physician (IDSA 172, 173, 299).
Daptomycin is a cyclic lipopeptide antibiotic for intravenous administration whose activity is limited to Gram-positive organisms but includes methicillin-resistant Staphylococcus aureus (MRSA) and vancomycin-resistant enterococci (VRE). It recently received US FDA approval for the treatment of complicated skin and skin-structure infections due to susceptible Gram positives. Its mechanism of action has remained incompletely defined, although it appears to bind to the bacterial cell membrane via its fatty acid tail, followed by a decrease in membrane potential (Antimicrob Agents Chemother. 2003;47:1318). Transmission electron micrography was consistent with the cell membrane as being the target of daptomycin and with bactericidal activity occurring without gross lysis of the bacterial cell (ICAAC C1-2135).
Daptomycin was active in vitro against enterococci and S aureus resistant to either linezolid or quinupristin/dalfopristin. It was more rapidly and completely bactericidal against susceptible staphylococci and streptococci than was either cefazolin or vancomycin in peritoneal dialysate fluid in vitro and is bactericidal against vancomycin-intermediate S aureus (VISA) strains. Daptomycin was also bactericidal at 8 mg/mL against 5 of 6 strains of Enterococcus faecium resistant to both vancomycin and linezolid. The postantibiotic effect of daptomycin against staphylococci and pneumococci lasted 1.0-6.2 hours (ICAAC C1-1643, A-1152, C1-1641, C1-1638, A-1172).
When compared in clinical trials to either a semisynthetic penicillin or vancomycin, daptomycin was as successful as these agents in the treatment of complicated skin and skin-structure infections due to S aureus, including both MRSA and methicillin-susceptible S aureus (MSSA), as well as against streptococcal infections (IDSA 296, 297).
The introduction and use of an antibiotic is inevitably followed by the emergence of bacterial resistance to the agent. At one institution, all vancomycin-resistant E faecium (VREF) were susceptible to linezolid at concentrations < 2 mg/mL prior to the introduction of this antibiotic into use. In the period immediately after its introduction, 2 of 12 strains had a linezolid MIC = 4 mg/mL. One of these 2 strains was examined and found to have the typical G2576U mutation in domain V of 1 of 6 23S rRNA gene copies. Elsewhere, an outbreak of linezolid-resistant E faecium bloodstream infections involved 6 patients in an oncology unit. Resistance to linezolid may also be observed among staphylococci. An S aureus isolate with reduced susceptibility to linezolid (MIC = 8 mg/mL) associated with mutations in 2 of 5 copies of the 23S rRNA gene was the cause of fatal ventilator-associated pneumonia (IDSA 216, ICAAC K-1112, K-1405).
Linezolid, which is bacteriostatic against staphylococci and enterococci, was less effective than vancomycin in the treatment of endocarditis due to either MRSA or glycopeptide-intermediate S aureus (GISA) in a rabbit model. Linezolid was, however, synergistic with imipenem in a rabbit model of MRSA endocarditis (ICAAC B-316, B-1093).
Oritavancin, like dalbavancin, a second-generation glycopeptide, was active in vitro against VRE with either VanA or VanB phenotypes. Three to 7 days of treatment with oritavancin was as effective as a similar duration of treatment with vancomycin, followed by cephalexin to complete a total of 14 days in patients with complicated skin/skin-structure infections. Oritavancin, which was given for a mean of 5.3 days vs 10.9 days in the comparator group, was significantly better tolerated (ICAAC C1-1640, L-739a).
A major drawback to efficient management of MRSA-infected patients is the delay, usually of at least 48 hours, in the identification of methicillin resistance by classical microbiological techniques. The IDI-MRSA, a real-time PCR assay with same-day turn around, had a sensitivity of 98% and specificity of 100% for the detection of MRSA directly on nasal and rectal swabs in 2 studies. The total assay time to process the samples was 3-4 hours (ICAAC K-1394, 1748).
Eighty-five percent of cases of bacteremic S aureus pneumonia were hospital acquired. Of the total of 55 cases, 34 were due to MRSA and 21 to MSSA. Only 48% of patients infected with MRSA infection received effective antibiotic therapy within 48 hours of admission (ICAAC K-765).
In a prospective study of S aureus bacteremia, 158 of 245 patients had blood cultures drawn every 2-3 days after the start of therapy, until they were negative. The median time to blood culture positivity was 15.2 hours (range, 4-53 hours). The mean duration of bacteremia was 4.0 ± 6.4 days (range, 8-59 days). In 65% of patients, bacteremia had cleared by day 3, while it persisted for 4-7 days in 23% and for > 7 days in 12%. Risk factors for persistence for 4-7 days, on multivariate analysis, were the presence of an orthopedic prosthesis and rapid growth in blood culture (< 15 hours to positive culture). Predictors of persistence for > 7 days were an endovascular source of infection and treatment with vancomycin, especially in endocarditis (ICAAC K-763).
In patients with S aureus bacteremia, a time to blood culture growth detection of 16 hours or less was associated with an increased likelihood of an endovascular source (including endocarditis and IV catheter infection), persistence of bacteremia for more than 4 days, and the development of metastatic foci of infection (IDSA 307).
Twenty-eight patients with bacteremic pneumonia due to S aureus were identified; 43% were community acquired. The pneumonia was believed to be secondary to the bacteremia in two-thirds of cases. Nine patients had endocarditis, 5 had osteomyelitis, 5 had septic arthritis, and 2 had epidural abscess. The mortality was 18% (IDSA 270).
Of 320 prospectively identified adult inpatients with S aureus bacteremia, 45 (14%) had soft-tissue infection as the source. Among 29 patients in this group with repeated blood cultures, the mean duration of bacteremia was 2.5 ± 2.9 days (range, 1-13 days); bacteremia persisted for > 4 days in 17%. The duration of bacteremia was inversely related to the incubation time to initial blood culture positivity. Nine patients received predominantly oral therapy (IDSA 271).
It was estimated by the CDC that approximately 120,000 hospitalizations with a diagnosis of MRSA infection occur annually in the United States. These include approximately 30,000 septicemias and 27,000 pneumonias. In a network of 8 community hospitals, only 18% of cases of MRSA infection or colonization were nosocomially acquired, but most of the rest were health care associated (IDSA 563, 562).
Analysis of the SCOPE-MMIT Antimicrobial Network data revealed a persistent and significant association between hospital levofloxacin use and the prevalence of MRSA. This is consistent with a number of other studies that have identified fluoroquinolone use as a risk factor for nosocomial acquisition of MRSA (Emerg Infect Dis. 2003. www.cdc.gov/ncidod/EID/vol9no11/03-0284.htm). It is also consistent with the observation that fluoroquinolones select MRSA from heteroresistant populations in vitro. Relevant to this finding, it was reported that garenoxacin exposure was associated with lesser selection of oxacillin-resistant clones from a population of mecA-positive MRSA in vitro than were ciprofloxacin, levofloxacin, or gatifloxacin. In addition, it was reported that a 95% reduction in fluoroquinolone use in a French hospital was associated with a trend toward a reduced frequency of MRSA isolation (ICAAC K-1743, K-1399, C1-64).
But not all reports agree with this putative role of fluoroquinolones in the epidemiology of MRSA infections. One group reported that neither the use of fluoroquinolones nor antimicrobials of other classes was associated with an increased risk of subsequent MRSA bloodstream infection in a matched case-control study (ICAAC K-1742).
When tested 3 or more months (median, 16 months) after successful MRSA decolonization, only 3 of 37 (8%) patients were again colonized with their original strain, and 1 individual had acquired a new MRSA strain. One potential reason for this failure may be continued gastrointestinal tract colonization—38 of 107 (36%) MRSA carriers had positive rectal swabs for MRSA (ICAAC K-1749, K-1401).
An MRSA outbreak in a burn unit occurred despite standard burn-wound precautions and targeted isolation but was terminated after implementation of preemptive barrier precautions with all patients in the unit. Separately, the use of intranasal mupirocin was believed to have terminated an MRSA outbreak, not affected by other control measures, in a neonatal ICU (ICAAC K-582, K-1744).
A metaanalysis of 10 studies found that the use of mupirocin was associated with a substantial reduction in S aureus infections in dialysis patients (ICAAC K-1746).
Acquisition of MRSA in the community has exploded in frequency. One-half of MRSA isolated from patients with systemic infections in 2002-2003 was community acquired (CA-MRSA) and carried the staphylococcal cassette chromosome type IV (SCCmec IV) in a Houston study. Consistent with the usual lack of multidrug resistance in CA-MRSA isolates containing SCCmec IV, all were susceptible to trimethoprim-sulfamethoxazole. Similarly, the proportion of community-acquired S aureus bacteremic episodes in Darwin, Australia, due to MRSA increased from 9% in 1998 to 20% in 2001. Empiric antimicrobial therapy was inappropriate in 80% of episodes. The isolates contained type IV SCCmec and were not multidrug resistant (IDSA 258, ICAAC L-619).
A retrospective review of 60 children with community-acquired S aureus infection admitted to a Houston hospital found that 45% of the isolates were methicillin resistant. MRSA was more likely to be associated with deep-skin infections than was MSSA, while the latter was more often associated with respiratory infection. The length of stay was 3 days longer in MRSA-infected patients, who were also more likely to have received inappropriate initial antimicrobial therapy (IDSA 799, 801).
Four outbreaks of CA-MRSA were identified in Los Angeles County in 2002. Those involved a group of 5 infants recently discharged from a newborn nursery, 2 members of an athletic team, a number of men who have sex with men (MSM), and 934 county jail inmates. Almost 10% of the 934 LA county inmates with CA-MRSA infections were hospitalized for a mean of 10 days. Pulsed-field gel electrophoresis revealed the same predominant clone in each outbreak (IDSA 263, 264).
Community-onset MRSA infection was detected in 35 adult outpatients—30 males, 4 females, and 1 transgender male-to-female. Compared to the general clinic population, MRSA patients were significantly more likely to be MSM. Among the MSM, the site of the skin abscess was consistent with the possibility of sexual transmission, involving the groin, genitals, buttocks or rectum, mouth, or face (ICCAC L-1602a).
Examination of MRSA isolates from jail inmates in both San Francisco and Los Angeles determined that they are SCCmec type IV and are of a single clone bearing the Panton-Valentine leukocidin (PVL) virulence factor (Infectious Disease Alert. 2002;22(6):44; Emerg Infect Dis. 2003;9:978-984). Separately, analysis of national MRSA isolates found that PVL is common only in community-onset isolates, while toxic shock syndrome toxin (TSST) was detected primarily in one hospital-associated lineage. Other novel MRSA strains that could not be typed by SCCmec analysis are also circulating in San Francisco. These strains have antibiotic-resistance patterns similar to the strains carrying SCCmec type IV (ICAAC K-1393, C2-1980, IDSA 259).
Another study in San Francisco found that S aureus containing SCCmec type IV, which has been associated with community-onset MRSA, was widely detected in both hospital and community settings. Furthermore, these strains are capable of capturing multiple antibiotic resistance determinants. In a separate study in which 6% of 308 homeless youth in that city had nasal colonization with CA-MRSA containing SCCmec type IV, emerging resistance to macrolides and fluoroquinolones was found (ICAAC C2-1983, IDSA 255).
The optimal choice of antibiotic for treatment of CA-MRSA remains a matter of discussion. However, many subcutaneous abscesses caused by community-onset MRSA were effectively treated with incision and drainage in the absence of effective antibiotic therapy (ICAAC G-1541).
Fifty-six percent of 161 erythromycin-resistant, clindamycin-susceptible S aureus isolates in Baltimore demonstrated inducible clindamycin (MLS) resistance. Elsewhere, 79% of S aureus and S epidermidis with a clindamycin-susceptible, erythromycin-susceptible phenotype exhibited inducible resistance to clindamycin. The authors conclude that laboratories should either test for inducible clindamycin resistance or report these strains as being resistant or potentially resistant to clindamycin. This approach may have prevented a reported case of failure of clindamycin therapy in an infection due to an isolate with inducible clindamycin resistance (IDSA 485, ICAAC C2-87, IDSA 485).
Six of 21 (29%) S haemolyticus isolates demonstrated vancomycin heteroresistance, but this did not appear to adversely affect therapeutic outcomes. The presence of the accessory gene regulator (agr) in MRSA has been associated with glycopeptide-intermediate susceptibility (J Infect Dis. 2003;187:929). Persistence of MRSA infection despite therapy with vancomycin was associated in 4 of 5 patients with the presence of agr group II MRSA. Treatment with quinupristin/dalfopristin plus vancomycin was successful in 3 of the 4, with eradication of the organism in 1.5, 4, and 4.5 days (IDSA 218, 486).
Biofilm plays a critical role in many infections, particularly those involving foreign material such as vascular access catheters. Minocycline, rifampin, and quinupristin/dalfopristin were significantly more active than either vancomycin or linezolid against MRSA and VRE embedded in biofilm. In addition, ramoplanin was bactericidal against S epidermidis and S aureus growing in biofilm (ICAAC C1-118, C1-119).
Confirming the results of previous studies, a metaanalysis found that vancomycin resistance is an independent predictor of death in patients with enterococcal bloodstream infection (IDSA 491).
In further evidence of the potential for antibiotic-fed food animals to be the source of antibiotic-resistant bacteria in humans, interspecies transfer of the vanA operon from poultry-derived E faecium to human-derived E faecalis occurred at high frequency in the GI tract off human flora-associated mice (ICAAC C1-16).
Hand washing is, of course, critical to prevention of VRE transmission. VRE carriage on the hands of health care workers was transient or intermittent, but not persistent. Environmental contamination may also be important in the epidemiology of VRE. Thus, VRE contamination by patient carriers in an outpatient hemodialysis unit was frequently detected on chairs or couches, the gowns of health care workers, and on patient hands. But bathing ICU patients with a 2% chlorhexidine-impregnated washcloth system was associated with decreased patient skin, environmental, and health care worker hand VRE contamination when compared with soap and water bathing (ICAAC K-1105, K-1107, K-1108, K-1107).
Some studies have suggested that the use of piperacillin/tazobactam is associated with a decreased risk of colonization with VRE when compared to cephalosporin or ticarcillin/clavulanate use. This observation was apparently confirmed by one group that reported that VRE acquisition decreased in association with a formulary change from ticarcillin/clavulanate to piperacillin/tazobactam. Two other papers did not confirm this finding. Thus, VRE colonization in an ICU was detected in 24% of patients after receipt of piperacillin/tazobactam and 25% after cefepime. In one study, previous receipt of vancomycin, piperacillin-tazobactam, advanced generation cephalosporins, and gatifloxacin were each associated with an increased risk of acquisition of VRE (ICAAC K-1429, IDSA 490, ICAAC K-1417).
In a setting in which linezolid and quinupristin/dalfopristin (QD) have been used for 5 years, 2.6% of VREF were nonsusceptible to linezolid, and 15.5% were resistant to QD (ICAAC K-1409).
VREF are susceptible to ramoplanin, an orally administered, nonabsorbable glycolipodepsipeptide (IDSA 167).
Adults 65 years and older are colonized with group B streptococci at rates similar to younger populations but are more likely to carry type V, the leading cause of invasive infection with this organism in the elderly. Furthermore, healthy elderly individuals often lack type V-specific IgG in serum. A group V vaccine conjugated to tetanus toxoid was found to be safe and immunogenic in healthy elderly adults (IDSA 529, 519).
In vitro studies demonstrated that chloroquine and plasma inter-alpha inhibitor each rescued murine peritoneal macrophages from anthrax lethal toxin-induced cytoxicity (IDSA 709, 712).
Levofloxacin and linezolid were each effective in the treatment of inhalational anthrax in murine models of infection. In a primate model, both ciprofloxacin and levofloxacin were therapeutically effective. However, 3 of 10 monkeys became ill and died after the 30-day period of treatment, presumably as the consequence of germination of residual spores (ICAAC A-306, B-3300, B-331a).
Two hundred individuals potentially exposed to anthrax in a US bioterror event received both antibiotics and anthrax vaccine. There were no serious adverse events attributable to the vaccine. However, most of 15 anthrax survivors assessed 1 year after infection reported significant functional impairment and evidence of psychological distress (IDSA 831, 833).
Hyperimmune human serum is being explored as a potential therapeutic in anthrax. The antiprotective antigen IgG concentration measured by ELISA in plasma obtained from individuals vaccinated against anthrax correlated with its functional activity in a toxin-neutralization assay (IDSA 832).
Dr. Deresinski, MD, FACP Clinical Professor of Medicine, Stanford; Associate Chief of Infectious Diseases, Santa Clara Valley Medical Center.