Bacterial Pathogens

Streptococcus pneumoniae. The antibiotic-resistant pneumococcal problem is predominantly the result of the selective pressure of antibiotics and clonal dissemination. An evaluation of 363 isolates of S. pneumoniae with high-level penicillin resistance from 31 states found that most belonged to one of only five serogroups (23, 6, 19, 9, 14) or were nontypable. Approximately 40% were macrolide-resistant. Almost two-fifths belonged to one of two international epidemic clones, clone A and clone B. The former is indistinguishable from the 23F "Spanish/USA" clone while the latter belongs to serogroup 9. (Abstract C-18.) A study in Dallas failed to find an association of increased carriage of resistant pneumococci and antibiotic use, but the study did find increased carriage of penicillin-resistant Streptococcus viridans, which are believed to be the source of the resistance genes in S. pneumoniae. (Abstract C-30.)

A major reason for the spread of these clones is believed to be the selective pressure of excess antibiotic use, particularly in children with respiratory tract infections. A strong association between recent receipt of oral beta-lactams and oropharyngeal carriage of penicillin-resistant pneumococci was found in French children. (Abstract C-24.) Similarly, a CDC study found that counties in the United States with higher rates of use of beta-lactam and macrolide antibiotics had significantly higher proportions of S. pneumoniae which lacked susceptibility to these antibiotics. (Abstract C-29.)

The best way to deal with an infection is to prevent it. Administration of the Wyeth Lederle heptavalent conjugate pneumococcal vaccine to infants at 2, 4, 6, and 12-15 months of age was associated with 100% protection against invasive pneumococcal disease when compared to a control group given meningococcal vaccine. (Friday, Abstract LB-9.)

Staphylococcus. Administration of linezolid for five days eradicated S. aureus nasal colonization in 45 (94%) of 48 subjects, while the colonization persisted in all eight subjects given placebo. Nonetheless, the majority had become recolonized after 30 days. (Abstract A-4.)

Adherence of staphylococci to foreign material, such as IV catheters, is an important pathophysiologic mechanism in many infections. Subinhibitory concentrations of each of six fluoroquinolones were demonstrated to upregulate fibronectin-binding proteins gene transcription in quinolone-resistant S. aureus and, as a consequence, to promote adhesion of the organism. This may provide these strains with a survival advantage and allow its persistence in foreign body infections. (Abstract C-182.) Similarly, subinhibitory concentrations of ciprofloxacin were demonstrated to increase adherence of some strains of S. epidermidis. (Abstract K-7.)

MRSA. Most, but not all, studies have found that the virulence of MRSA is no greater than that of susceptible strains of S. aureus. However, a Korean study found that the prevalence of the gene for the toxic shock syndrome toxin, TSST-1, was 16.7% in MRSA and only 1.5% in MSSA (P < 0.01).

VSSA. The MIC of vancomycin-susceptible S. aureus was demonstrated to be increased in the presence of heavy inocula. This inoculum effect appeared to be due to "trapping" of vancomycin by staphylcococcal cell wall elements. (Abstract C-138.) This effect may account for the relatively slow killing of S. aureus by vancomycin.

VISA. The existence of strains of S. aureus with intermediate MICs to vancomycin raises many questions with regard to the choice of antibacterial therapy in the management of infections caused by VISA.

Vancomycin resistance of VISA is heterogeneous, with the majority of an inoculum killed by concentrations less than 8 mcg/mL. (Abstract E-173a.) Vancomycin was, nonetheless, ineffective in the treatment of experimental endocarditis caused by a vancomycin-intermediate strain of S. aureus (MIC = 8 mcg/mL) in a rabbit model. However, even a single intravenous dose of lysostaphin was effective in clearing bacteremia and frequently sterilized vegetations. (Saturday, Abstract LB-7.)

Vancomycin and gentamicin had synergistic activity in vitro against VISA isolates, while rifampin did not improve the antibacterial activity of vancomycin. (Abstract E-173a.) Despite this finding, the addition of gentamicin to vancomycin was not effective in an in vitro fibrin clot model of VISA infection. (Abstract A-136.) A combination of nafcillin and vancomycin appeared to be synergistic and was effective in the treatment of VISA endocarditis in the rabbit model. (Saturday, Abstract LB-8.)

CNS. Determination of the clinical significance of the coagulase-negative staphylococci detection in blood cultures is often difficult. Clinical findings and the repeated isolation of the organism from blood cultures is ordinarily considered to be good evidence for the presence of true bacteremia. In a study that shakes these beliefs, 95 CNS blood stream isolates were obtained from 42 patients. In 37 (88%) of these patients, the organism was recovered repeatedly from their blood within a five-day interval. However, clonality of isolates by pulsed field gel electrophoresis of DNA fragments correlated poorly with each of three different sets of clinical criteria used to determine significance, including the CDC surveillance definition for bacteremia. Discrepancies occurred commonly in both directions. This suggests the possibility of frequent unnecessary treatment, often with vancomycin, triggered by recovery of CNS from blood. (Abstract K-69.)


Independent risk factors for the development of enterococcal bacteremia in hospitalized patients were found to be neutropenia, presence of an indwelling urinary catheter, and prior receipt of cephalosporins and/or imipenem. (Abstract K-83.)

Genetic analysis suggests that colonization with VRE may persist for years, even during periods of time when VRE cannot be isolated from feces. (Abstract K-86.) This raises questions concerning the infection control management of VRE colonized inpatients whose cultures become negative.

Many VRE are susceptible to both doxycycline and chloramphenicol, but their efficacy as therapeutic agents against apparently susceptible VRE has been questioned. Doxycycline was ineffective, while chloramphenicol was commonly effective in protecting mice against infection with VSE or VRE. The two drugs were largely indifferent in their interaction both in vitro and in the murine model. (Abstract B-61.)

Six patients with CAPD-associated peritonitis due to VRE received quinupristin/dalfopristin intraperitoneally; five also received the drug IV. Intraperitoneal dosing ranged from 30 mg to 250 mg per dialysis bag q12h to q6h for 2-15 days. Treatment was well-tolerated, but assessment of efficacy was confounded by the administration of additional antibiotics. (Abstract MN-52.)

Prior receipt of a third generation cephalosporin or of IV metronidazole were found to be significant risk factors for acquisition of VRE in a multivariate analysis. (Abstract K-82.)


The increasing antimicrobial resistance among gastrointestinal pathogens just goes marching forward! A study performed in Spain found that 76% of strains of Campylobacter jejuni isolated from humans and 79% isolated from food (chicken) were resistant to ciprofloxacin. While only 3.9% of strains isolated from humans were resistant to erythromycin, 18.4% of those isolated from food were macrolide resistant, suggesting that resistance to this class of drugs will soon be widespread in humans in Spain. (Abstract C-183b.)


One hundred forty-six adults with brucellosis were randomized to receive doxycycline 100 mg q12h for either 30 or 45 days, in addition to all receiving gentamicin 240 mg qd for the first seven days. Of 73 patients in each group, three assigned to 30 days and one assigned to 45 days of treatment failed therapy, while relapses occurred in 13 (18%) and nine (12%), respectively (P = 0.48). (Abstract MN-25.) Thus, 30 days of treatment with this regimen appears to provide results similar to those obtained with 45 days of treatment, although the relapse rate was high with both regimens.

Relapse continues to be a problem in brucellosis, even with regimens that include rifampin or streptomycin with doxycycline for 45 days (Antimicrob Agents Chemother 1995;39:2061). Among the complications of brucellosis is epididymoorchitis; patients with this lesion may have a higher risk of relapse and those with necrotizing epididymoorchitis may require orchiectomy. (Abstract L-80.)


Helicobacter pylori DNA was detected by PCR in the water supplies of two Arctic communities with a high prevalence of seropositivity to this organism. (Abstract E-64.) Whether this represents viable bacteria, however, remains to be determined. H. pylori is killed by free chlorine (Appl Environ Microbiol 1997;63:4969). Consistent with previous evidence of clustering within families, genetic analysis of gastric isolates of H. pylori provided evidence of intrafamilial transmission, a finding previously reported (N Engl J Med 1990;322:359; Gut 1993;34:1348). (Abstract E-72.)

Garlic and omeprazole have a synergistic antibacterial effect against H. pylori in vitro. (Abstract E-65.) H. pylori is inhibited in vitro by catechins derived from tea. (Abstract E-71.) Thus, Helicobacter joins other organisms, including MRSA, which are inhibited by tea extracts (J Antimicrob Chemother 1998;42:211; Infect Dis Alert 1998;17:185).


Nine percent of 469 isolates of Clostridium difficile isolated in Madrid were resistant (MIC > 32 mcg/mL) to metronidazole. (Abstract E-173.)