Reactivity of Platelia Aspergillus Galactomannan Antigen with Piperacillin-Tazobactam: Clinical Implications Based on Achievable Concentrations in Serum

MATERIALS AND METHODS

A total of 10 antibiotics were tested for galactomannan by using the Platelia Aspergillus EIA, Bio-Rad Laboratories, Redmond, Wash. These included intravenous formulations of piperacillin, ampicillin-sulbactam, nafcillin, piperacillin-tazobactam, cefazolin, ceftazidime, gentamicin, erythromycin, and levofloxacin, in 0.9% saline or 5% dextrose (Table 1). Three separate lots of piperacillin-tazobactam (one bag from lot A, three bags from lot B, and five bags from lot C) were tested. Since amoxicillin powder is insoluble in water, it was tested as a solution reconstituted in a phosphate buffer (pH 6.0) to yield the same concentration of the drug that is present in oral suspension (50 mg/ml).

All drugs were initially tested at full strength or undiluted for the galactomannan assay. Drug diluents (0.9% saline, 5% dextrose, or a phosphate buffer [pH 6.0]) were used as controls, and all tests were conducted in duplicate wells. Briefly, 50 μl of the undiluted antibiotic sample was added to the wells containing 50 μl of the conjugate, and the plates were incubated at 37°C for 90 min. The plates were then washed with an automated washer, and 200 μl of the chromogen substrate solution was added. After incubation at 18 to 25°C for 30 min in the dark, stop solution was added. The plates were read at a wavelength of 450 nm by using a reference filter of 630 nm. The index for each sample was calculated by dividing its optical density (OD) by the cutoff value (mean OD) of the threshold control. Indices of >0.5 were considered positive per the cutoff values for serum samples noted in the manufacturer's package insert.

Drugs that tested positive for galactomannan as undiluted samples were further tested at achievable concentrations in serum. We also sought to determine the minimal concentration of the antibiotic in serum that yielded a positive test. For this experiment, the drug was diluted in serum which had been pretested and shown to be negative for galactomannan (index of <0.20). Antibiotic dilutions tested were based on the achievable peak concentration in serum as the target level for each drug, with 1 twofold dilution above and 2 serial dilutions below the target level. The serum sample (300 μl) was pretreated to remove immune complexes and proteins per the package insert instructions, boiled for 3 min, and centrifuged at 10,000 × g for 10 min. Fifty microliters of the supernatant was then added to the duplicate wells, and the experiment for each dilution was completed by following the steps outlined for assaying the undiluted samples. Sterile water and pretested negative serum were used as controls.

RESULTS

When tested full strength or undiluted, piperacillin-tazobactam and piperacillin tested positive in the galactomannan assay, whereas amoxicillin, nafcillin, ampicillin-sulbactam, cefazolin, ceftazidime, gentamicin, erythromycin, and levofloxacin tested negative. The index (the mean of two values when the drug was tested in duplicate wells) for piperacillin was 3.681 (Table 2). All lots of piperacillin-tazobactam and all bags within each lot yielded a positive result, with index values calculated at >5.168 (Table 2). The OD of the piperacillin-tazobactam wells was out of the absorbance range of the plate reader; i.e., it was >3.000. The index was therefore calculated using an OD value of >3.000.

Since piperacillin and piperacillin-tazobactam tested positive as undiluted samples, dilutions of these antibiotics above and below the achievable peak concentrations in serum were tested for galactomannan assay positivity (Table 3). Given that the peak concentration of piperacillin in the serum following an intravenous dose of 4.5 g of piperacillin-tazobactam or 4 g of piperacillin is 298 μg/ml (or ∼300 μg/ml), the dilutions tested were 600 μg/ml (1 twofold dilution higher than the peak serum drug concentration), 300 μg/ml (achievable peak serum drug concentration), 150 μg/ml (1 twofold dilution lower than the peak serum drug concentration), and 75 μg/ml (2 twofold dilutions lower than the peak serum drug concentration). Piperacillin yielded a negative result at all concentrations tested, i.e., 75, 150, 300, and 600 μg/ml. For piperacillin-tazobactam, one bag from each lot was tested. Piperacillin-tazobactam from lot A tested negative at concentrations of 75, 150, 300, and 600 μg/ml (Table 3). Piperacillin-tazobactam from lot B tested negative at concentrations of 75, 150, and 300 μg/ml but positive at 600 μg/ml (Table 3).

At the achievable peak concentrations in serum (300 μg/ml) and at 1 dilution higher (600 μg/ml), piperacillin-tazobactam from lot C yielded a positive test (Table 3). The index values for duplicate wells at a concentration of 300 μg/ml were 1.45 and 1.42, and at 600 μg/ml, they were 2.50 and 2.36. The test was also positive at concentrations of 150 μg/ml (indices of 0.84 and 0.78), approached negativity at 75 μg/ml (indices of 0.45 and 0.50), and was negative at concentrations of 10 μg/ml (index of 0.17 for both wells) and 5 μg/ml (indices of 0.21 and 0.20) (Table 3). Saline, dextrose, the phosphate buffer, and serum controls tested negative in all experiments.

DISCUSSION

Patients at risk for or being evaluated for invasive aspergillosis are likely to have been on broad-spectrum antibiotics. A false-positive galactomannan test in this setting may lead to unnecessary diagnostic procedures or employment of antifungal therapy (11). On the other hand, a positive test may be attributed to false EIA reactivity and may therefore delay the appropriate investigations for aspergillosis. Thus, reactivity of the Platelia Aspergillus EIA with antibiotics is of potentially significant clinical relevance.

Aspergillus fumigatus galactomannan is a polysaccharide comprised of a linear mannan core with α-(1-2)-linked mannotetraose units attached with α-(1-6) linkage (6). The side chains, consisting of an average of four to five β-(1-5)-galactofuranose units, are linked to C-6 and C-3 positions of α-(1-2)-linked mannose units of the mannan core (6). Galactomannan is widely distributed among Aspergillus and Penicillium species (6, 14). Although subtle chemical differences exist, the galactomannan of Aspergillus is strikingly similar in structure to that of Penicillium.

We tested several antibiotics commonly used in clinical practice, including those of fungal and nonfungal origins and those from nonmicrobial and synthetic sources. Penicillins and cephalosporins, with the exception of cephamycins (which are produced by actinomycetes rather than fungi), are of fungal origin (8). Penicillins are derived from Penicillium spp.; the drug originally isolated by Alexander Fleming in 1929 was from a strain of Penicillium notatum (2). The piperacillin component of piperacillin-tazobactam is a semisynthetic acylaminopenicillin, and tazobactam is a synthetic penicillinate sulfone. Gentamicin and erythromycin are naturally occurring compounds of nonfungal origin; gentamicin is derived from Micromonospora purpurea, and erythromycin is derived from Streptomyces erythraeus, formerly known as Saccharopolyspora erythraea. Finally, levofloxacin, a chiral fluorinated carboxyquinolone, is synthetically produced.

Undiluted samples of piperacillin-tazobactam and piperacillin in our study tested positive for galactomannan, whereas other antibiotics, i.e., amoxicillin, nafcillin, ampicillin-sulbactam, cephalosporins, gentamicin, levofloxacin, and erythromycin, tested negative. All batches of piperacillin-tazobactam and all bags within each batch tested strongly positive (index, 5.168). However, at achievable concentrations in serum, piperacillin-tazobactam, but not piperacillin, resulted in a positive test. Given that an index of 1.0 equals approximately 1 ng of galactomannan/ml (10), piperacillin-tazobactam at its achievable peak concentration in serum (300 μg/ml) had a level of reactivity that indicated the presence of ∼1.4 ng of galactomannan per ml.

The precise reasons for the reactivity of piperacillin-tazobactam in the galactomannan assay are not known. It remains to be determined if the basis of reactivity is galactomannan or galactofuranose from either Aspergillus or a non-Aspergillus source such as Penicillium, a chemical reaction with the drug itself, or another compound found in piperacillin-tazobactam. Whether the reactivity is amenable to elimination in the manufacturing process or is part of the antibiotic molecule that cannot be modified is also unknown.

Following a 4.5-g dose, the peak levels (at 30 min) of piperacillin-tazobactam range from 155 to 298 μg/ml (Zosyn [piperacillin-tazobactam] product information, Wyeth Pharmaceuticals, Inc., Philadelphia, Pa.). Levels at 1, 2, 3, and 4 h are 141, 46.6, 16.4, and 6.9 μg/ml, respectively, and decline to <1.4 μg/ml 6 h after the dose or at a trough (Zosyn product information; Wyeth Pharmaceuticals, Inc.). Whereas piperacillin-tazobactam at concentrations of 75, 150, 300, and 600 μg/ml yielded false-positive reactivity in the galactomannan assay in our study, lower concentrations (10 and 5 μg/ml) tested negative for galactomannan. These data therefore suggest that while achievable levels of piperacillin-tazobactam in the serum at 30 min and 1 h following a 4.5-g dose may potentially yield a false-positive galactomannan test result, concentrations in serum that approximate the trough level are unlikely to result in a false-positive test.

Clinicians evaluating the results of the galactomannan test should be aware that achievable concentrations of piperacillin-tazobactam in serum can result in a positive galactomannan test in patients receiving this antibiotic. Alternatively, since the dilutions of piperacillin-tazobactam in serum that mimicked the levels achievable at trough concentrations did not yield a positive test, the samples for the galactomannan test in patients receiving piperacillin-tazobactam could be timed so as to be collected prior to the administration of a dose or to coincide with a trough level of piperacillin-tazobactam. These findings warrant validation in future investigations with tests conducted on sera from patients receiving piperacillin-tazobactam.