Publications by Year: 2018

Microscopic interpretation of stained smears is one of the most operator-dependent and time-intensive activities in the clinical microbiology laboratory. Here, we investigated application of an automated image acquisition and convolutional neural network (CNN)-based approach for automated Gram stain classification. Using an automated microscopy platform, uncoverslipped slides were scanned with a 40× dry objective, generating images of sufficient resolution for interpretation. We collected 25,488 images from positive blood culture Gram stains prepared during routine clinical workup. These images were used to generate 100,213 crops containing Gram-positive cocci in clusters, Gram-positive cocci in chains/pairs, Gram-negative rods, or background (no cells). These categories were targeted for proof-of-concept development as they are associated with the majority of bloodstream infections. Our CNN model achieved a classification accuracy of 94.9% on a test set of image crops. Receiver operating characteristic (ROC) curve analysis indicated a robust ability to differentiate between categories with an area under the curve of >0.98 for each. After training and validation, we applied the classification algorithm to new images collected from 189 whole slides without human intervention. Sensitivity and specificity were 98.4% and 75.0% for Gram-positive cocci in chains and pairs, 93.2% and 97.2% for Gram-positive cocci in clusters, and 96.3% and 98.1% for Gram-negative rods. Taken together, our data support a proof of concept for a fully automated classification methodology for blood-culture Gram stains. Importantly, the algorithm was highly adept at identifying image crops with organisms and could be used to present prescreened, classified crops to technologists to accelerate smear review. This concept could potentially be extended to all Gram stain interpretive activities in the clinical laboratory.

Coccidioides immitis/posadasii presents in mycelial form with branching hyphae and arthroconidia when cultured in the laboratory. On histopathology, the presence of endospore-containing spherules is considered diagnostic of coccidioidomycosis. Here we report an unusual case of coccidioidomycosis with hyphae and arthroconidia in pulmonary tissue sections. A 49-year-old male patient with intermittently treated pulmonary coccidioidomycosis sought treatment for residual pulmonary complaints. A cavity in the left upper lobe was seen on computed tomographic scan. Due to minimal improvement of symptoms despite treatment with fluconazole, a left upper lobectomy was ultimately performed. Coccidioides mimmitis/posadasii was identified by culture and DNA probe from the lobectomy specimen. The histopathology showed a fibro-cavitary lesion, with arthroconidia and hyphal structures, but no typical endospore-forming spherules. While uncommon, C. immitis/posadasii may present with hyphae and arthroconidia on histopathology. Pathologists should be aware of this unusual presentation; culture remains the most reliable method for definitive diagnosis.

Apramycin, an aminocyclitol aminoglycoside, was rapidly bactericidal against Acinetobacter baumannii In a neutropenic murine thigh infection model, treatment-associated A. baumannii CFU reductions of >4 log10 per thigh were observed for all exposures for which area under the curve (AUC)/MIC ratio was >50 and maximum concentration of drug in serum (Cmax)/MIC was ≈10 or higher. Based on these findings, we suggest that apramycin deserves further preclinical exploration as a repurposed therapeutic for multidrug-resistant Gram-negative pathogens, including A. baumannii.

The observed MIC may depend on the number of bacteria initially inoculated into the assay. This phenomenon is termed the inoculum effect (IE) and is often most pronounced for β-lactams in strains expressing β-lactamase enzymes. The Clinical and Laboratory Standards Institute (CLSI)-recommended inoculum is 5 × 105 CFU ml-1 with an acceptable range of 2 × 105 to 8 × 105 CFU ml-1 IE testing is typically performed using an inoculum 100-fold greater than the CLSI-recommended inoculum. Therefore, it remains unknown whether the IE influences MICs during testing performed according to CLSI guidelines. Here, we utilized inkjet printing technology to test the IE on cefepime, meropenem, and ceftazidime-avibactam. First, we determined that the inkjet dispense volume correlated well with the number of bacteria delivered to microwells in 2-fold (R2 = 0.99) or 1.1-fold (R2 = 0.98) serial dilutions. We then quantified the IE by dispensing orthogonal titrations of bacterial cells and antibiotics. For cefepime-resistant and susceptible dose-dependent strains, a 2-fold increase in inoculum resulted in a 1.6 log2-fold increase in MIC. For carbapenemase-producing strains, each 2-fold reduction in inoculum resulted in a 1.26 log2-fold reduction in meropenem MIC. At the lower end of the CLSI-allowable inoculum range, minor error rates of 34.8% were observed for meropenem when testing a resistant-strain set. Ceftazidime-avibactam was not subject to an appreciable IE. Our results suggest that IE is sufficiently pronounced for meropenem and cefepime in multidrug-resistant Gram-negative pathogens to affect categorical interpretations during standard laboratory testing.

We describe a new rapid and accurate immunoassay-based technology capable of counting single target molecules using digital imaging without magnification. Using the technology, we developed a rapid test for Clostridium difficile toxin B, which is responsible for the pathology underlying potentially fatal C. difficile infections (CDI). There are currently no tests for CDI that are rapid, sensitive, and specific. The MultiPath C. difficile toxin B test images and counts complexes of target-specific magnetic and fluorescent particles that have been tethered together by toxin B molecules in minimally processed stool samples. The performance characteristics of the 30 minute test include a limit of detection of 45 pg/mL, dynamic range covering 4-5 orders of magnitude, and coefficient of variation of less than 10%. The MultiPath test detected all toxinotypes and ribotypes tested, including the one most commonly occurring in the US and EU; shows no cross reactivity with relevant bacterial species; and is robust to potential interferants commonly present in stool samples. On a training set of 320 clinical stool samples, the MultiPath C. difficile toxin B test showed 97.0% sensitivity (95% CI, 91.4-99.4%); 98.3% specificity (95% CI, 96.8-99.2%); and 98.2% accuracy (95% CI, 96.7-99.0%) compared to the cellular cytotoxicity neutralization assay (CCNA) reference method. Based on these compelling performance characteristics, we believe the MultiPath technology can address the lack of rapid, sensitive, specific, and easy-to-use diagnostic tests for C. difficile.

We evaluated the in vitro activity of apramycin against clinical strains of vancomycin-intermediate and methicillin-resistant and -susceptible Staphylococcus aureus. Apramycin demonstrated an MIC50/MIC90 of 8/16 μg/mL. No strains had an MIC above the epidemiological cutoff value of 32 μg/mL, suggesting apramycin resistance mechanisms are rare in this strain population. The mounting evidence for broad-spectrum in vitro activity of apramycin against S. aureus and other bacterial species suggests that further exploration of apramycin or derivatives as repurposed human therapeutics is warranted.

Resistance to colistin, a polypeptide drug used as an agent of last resort for the treatment of infections caused by multidrug-resistant (MDR) and extensively drug-resistant (XDR) Gram-negative bacteria, including carbapenem-resistant Enterobacteriaceae (CRE), severely limits treatment options and may even transform an XDR organism into one that is pan-resistant. We investigated the synergistic activity of colistin in combination with 19 antibiotics against a collection of 20 colistin-resistant Enterobacteriaceae isolates, 15 of which were also CRE. All combinations were tested against all strains using an inkjet printer-assisted digital dispensing checkerboard array, and the activities of those that demonstrated synergy by this method were evaluated against a single isolate in a time-kill synergy study. Eighteen of 19 combinations demonstrated synergy against two or more isolates, and the 4 most highly synergistic combinations (colistin combined with linezolid, rifampin, azithromycin, and fusidic acid) were synergistic against ≥90% of strains. Sixteen of 18 combinations (88.9%) that were synergistic in the checkerboard array were also synergistic in a time-kill study. Our findings demonstrate that colistin in combination with a range of antibiotics, particularly protein and RNA synthesis inhibitors, exhibits synergy against colistin-resistant strains, suggesting that colistin may exert a subinhibitory permeabilizing effect on the Gram-negative bacterial outer membrane even in isolates that are resistant to it. These findings suggest that colistin combination therapy may have promise as a treatment approach for patients infected with colistin-resistant XDR Gram-negative pathogens.