Experimental hematogenous endophthalmitis due to Cryptococcus neoformans.

New Zealand white female rabbits injected with 6 X 10(7) to 5.5 X 10(8) CFU of Cryptococcus neoformans into the right carotid artery developed three types of clinically detectable eye lesions. These eye manifestations included: (1) iritis that had distinctive 1-3 mm white nodules; (2) white vitreous lesions with ill-defined borders, and (3) focal and diffusely infiltrative lesions of the optic nerve radiations. Most of the eye lesions appeared between 1 to 3 weeks after intracarotid injection and resolved almost completely by 10 weeks. Their clinical appearance resembled some of the eye lesions described from human cases. Predominantly a mononuclear leukocyte response with scattered cryptococcal organisms were seen in the iris and optic nerve lesions. Serial quantitative cultures from eye tissues, brain, and kidney indicated that the clearance of viable cryptococci was slowest from the eye tissues. Among the three diagnostic techniques (India Ink mount, culture, and cryptococcal polysaccharide antigen (CPA) determination), the CPA determination was the most sensitive test to detect C. neoformans in the aqueous and vitreous humors of clinically involved eyes. This rabbit model may be useful for further studies on hematogenous cryptococcal endophthalmitis.

Comparative ocular pathogenicity of Cryptococcus neoformans, Candida glabrata, and Aspergillus fumigatus in the rabbit.

In a previous study, 88% of rabbits with disseminated infection caused by Candida albicans developed ophthalmoscopically visible, hematogenous endophthalmitis (chorioretinitis) over a 2 week period. To determine the incidence of this ocular complication in disseminated infection caused by Cryptococcus neoformans, Candida glabrata, and Aspergillus fumigatus compared with that caused by C. albicans, the first three species of fungi were injected intravenously (between 10(5) and 10(9) organisms per animal) into 36 New Zealand white rabbits. No chorioretinal lesions were seen by indirect ophthalmoscopy over a 2 week period. C. glabrata and A. fumigatus were not cultured from chorioretinas despite positive cultures from brains and kidneys at 1 and 2 weeks. In contrast, C. neoformans was cultured from 12 of 18 chorioretinas. (mean Log10 3.45 colony forming units/gm of tissue) as well as from the brains and kidneys. The less intense inflammatory cell response to C. neoformans compared with that 10 C. albicans seen on histopathologic examination most likely explains the nondetectability of the cryptococcal chorioretinitis by indirect ophthalmoscopy. These data suggest that C. glabrata. A. fumigatus, and possibly C. neoformans have less ocular pathogenicity than C. albicans in rabbits and correlate with the small number of documented human cases of ophthalmoscopically visible hematogenous endophthalmitis caused by fungi other than C. albicans.

Formulary evaluation of third-generation cephalosporins using decision analysis.

A structured, objective approach to formulary review of third-generation cephalosporins using the decision-analysis model is described. The pharmacy and therapeutics (P&T) committee approved the evaluation criteria for this drug class and assigned priority weights (as percentages of 100) to those drug characteristics deemed most important. Clinical data (spectrum of activity, pharmacokinetics, adverse effects, and stability) and financial data (cost of acquisition and cost of therapy per day) were used to determine ranking scores for each drug. Total scores were determined by multiplying ranking scores by the assigned priority weights for the criteria. The two highest-scoring drugs were selected for inclusion in the formulary. By this decision-analysis process, the P&T committee recommended that all current third-generation cephalosporins (cefotaxime, cefoperazone, and moxalactam) be removed from the institutions's formulary and be replaced with ceftazidime and ceftriaxone. P&T committees at other institutions may structure their criteria differently, and different recommendations may result. Using decision analysis for formulary review may promote rational drug therapy and achieve cost savings.

Combined in vitro effect of amphotericin B and rifampin on Cryptococcus neoformans.

The combination of amphotericin B and rifampin was synergistic in vitro in both inhibiting and killing seven strains of Cryptococcus neoformans by the checkerboard microtitration technique.

Cryptococcal intracerebral mass lesions: the role of computed tomography and nonsurgical management.

Cephalic computed tomography (CT) is a sensitive technique for defining certain intracerebral diseases. Four patients with cryptococcosis were evaluated with cephalic computed tomography over 3 years. All had focal, intracerebral "contrast-enhanced" lesions consistent with cryptococcal mass lesions, confirmed histopathologically in two. An analysis of 55 cases of cryptococcal intracerebral mass lesions from the literature showed that 18% of patients with these lesions associated with cryptococcal meningitis did not have specific symptoms or signs of focal intracerebral disease or increased intracranial pressure. We therefore recommend that patients with cryptococcal meningitis, regardless of localizing symptoms or signs, be considered for cephalic CT evaluation to ascertain the presence of mass lesions. Three patients in this report were treated with systemic antifungal medication without surgery. Decreased size or disappearance of these lesions was seen on sequential CT scan in all patients. We conclude that selected patients with cryptococcal intracerebral mass lesions may be managed successfully with systemic antifungal therapy alone.

Therapy of experimental Pseudomonas endocarditis with high-dose amikacin and ticarcillin.

Right-sided infective endocarditis due to Pseudomonas aeruginosa was induced in 130 rabbits. Animals received either: (1) no therapy (controls); (2) standard-dose amikacin (AMK) (15 mg/kg/day) plus ticarcillin (300 mg/kg/day), or (3) high-dose AMK (20 or 25 mg/kg/day) plus ticarcillin, for 20 days. Animals in each treatment group were evaluated at 10 days after therapy for bacteriologic relapse. Both standard- and high-dose AMK regimens significantly decreased mortality and Pseudomonas aeruginosa vegetation titers versus controls (p less than 0.01, p less than 0.05, respectively). Despite significantly higher serum AMK levels at 25 mg/kg/day, there was no significant difference in mean vegetation titers, percent of vegetations sterilized, or posttherapy bacteriologic relapse in the three treatment groups. AMK at 20 or 25 mg/kg/day (but not at 15 mg/kg/day) significantly reduced the incidence of pulmonary infarction versus untreated controls (p less than 0.01).

Detection of fungemia obscured by concomitant bacteremia: in vitro and in vivo studies.

Our recent clinical experience suggested that bacteremia may interfere with the detection of concomitant fungemia when standard blood culture methods are used. To determine the extent to which bacteria may interfere with fungal isolation from blood cultures, an in vitro model simulating blood cultures taken during concomitant fungemia and bacteremia was created. Each of six bacteria (Staphylococcus epidermidis, Staphylococcus aureus, Streptococcus faecalis, Escherichia coli, Klebsiella pneumoniae, and Pseudomonas aeruginosa) was combined with each of three pathogenic yeasts (Candida albicans, Candida tropicalis, and Torulopsis glabrata) in vented blood culture bottles containing enriched brain heart infusion broth and fresh normal human blood. Blood culture bottles were analyzed at 1, 2, and 7 days of incubation. Gram strains and subcultures onto chocolate and MacConkey agars failed to detect fungi in 37.0, 66.7, and 100% of samples, respectively. However, subcultures onto Sabouraud dextrose agar failed in only 13% of the samples (occurring only with P. aeruginosa). In a rabbit model of concomitant fungemia with C. albicans and bacteremia with P. aeruginosa, no yeasts were recovered from blood cultures despite 100% detection of P. aeruginosa. Therefore, the usual microbiological techniques may be inadequate to detect fungemia when concomitant bacteremia is present.