Comparison of the Rapid Yeast Plus Panel with the API20C Yeast System for Identification of Clinically Significant Isolates of Candida Species

Abstract

The RapID Yeast Plus system (Innovative Diagnostic Systems, Norcross, Ga.) is a qualitative micromethod employing conventional tests and single-substrate chromogenic tests and having a 4-h incubation period. This system was compared with the API20C (bioMerieux Vitek, Hazelwood, Mo.) system, a 24- to 72-h carbohydrate assimilation method. One hundred thirty-three clinical yeast isolates, including 57 of Candida albicans, 26 of Candida tropicalis, 23 of Candida glabrata, and 27 of other yeasts, were tested by both methods. When discrepancies occurred, isolates were further tested by the Automated Yeast Biochemical Card (bioMerieux Vitek). Germ tube production and microscopic morphology were used as needed to definitively identify yeast isolates. The RapID Yeast Plus system correctly identified 125 yeast isolates, with an overall accuracy of 94% (125 of 133). Excellent correlation was found in the recognition of the three yeasts most commonly isolated from human sources. The test was 99% (105 of 106 isolates) accurate with C. albicans, C. tropicalis, and C. glabrata. The RapID Yeast Plus system compares favorably with the API20C system and provides a simple, accurate alternative to conventional assimilation methods for the rapid identification of the most commonly encountered isolates of Candida species.

Recently, there has been an increase in fungal infections in immunocompetent as well as immunocompromised patients. Yeasts (particularly Candida albicans) are the most common fungi isolated from human infections. Various yeast species are inherently or potentially resistant to amphotericin B and the new azole agents (13, 14). Rapid identification of significant clinical yeast isolates is therefore imperative for prompt institution of appropriate antifungal therapy. Conventional methods for the definitive identification of clinically significant yeasts to species level are generally based upon the ability of the yeast to utilize sugars by fermentation or assimilation (9, 14). The API20C (20C) system consists of strips containing dehydrated substrates for carbohydrate assimilation reactions. Although the test is simple and accurate, results are not available for 24 to 72 h. Recently, the RapID Yeast Plus (RYP) panel, which requires only a 4-h incubation period, was introduced. This system of biochemical reactions, based upon single-substrate enzymatic test reactions involving preformed enzymes in the inoculum, greatly decreases the turnaround time for the identification of the most common yeasts isolated from human patients. The purpose of this study was to evaluate the performance of this rapid test compared to that of the older, slower API system.

A total of 133 clinical yeast isolates were examined in this study. Specimens were obtained from inpatients and outpatients seen in satellite clinics affiliated with Memorial Hospital of Rhode Island. All yeasts were tested with the RYP panel and the 20C test strip. Discrepancies were resolved with the Vitek (VIT) Yeast Biochemical Card and germ tube production, as well as microscopic morphologic examination for blastoconidia, chlamydospores, pseudohyphae, and true hyphae, following growth on cornmeal Tween 80 agar, after incubation at 25°C for 72 h. The 20C yeast strip was inoculated with each yeast suspended in basal medium, which was heated before use, following the manufacturer’s instructions. Following incubation at 30°C, growth in each well was recorded at 24, 48, and 72 h. Final identification was made with the Analytical Profile index, when the selection was described as excellent, very good, or acceptable. The RYP was inoculated with a yeast suspended in RapID inoculation fluid, according to the manufacturer’s instructions. Following incubation at 30°C for 4 h, color development was noted in each test cavity either directly or after adding reagents. A minimal microcode was constructed and interpreted by comparison of test results to reactivity patterns stored in a database.

The VIT Yeast Biochemical Card is designed for use with the automated Vitek system. Each card was inoculated with a yeast isolate suspended in 0.45% NaCl to a McFarland no. 2 standard, according to the manufacturer’s instructions. Suspensions were inoculated into the card with the filling module, sealed with the sealing module, and incubated at 30°C for 24 or 48 h. The cards were read by the reader module; biochemical reactions were analyzed and interpreted by matching biocodes with the database of the computer. Germ tube production was determined microscopically for all isolates after incubation of the yeast at 37°C in rabbit plasma for 2.5 h.

All isolates of C. albicans produced germ tubes; all non-C. albicans isolates were germ tube negative. Eighty percent of the yeasts isolated were identified as either C. albicans (43% [57 of 133]), Candida tropicalis (20% [26 of 133]), or Candida glabrata (27% [23 of 133]). The accuracy of the RYP with this commonly isolated group of yeasts was 99% (105 of 106). Of this group, one isolate of C. tropicalis was misidentified as Candida stellatoidea by RYP (Table 1). There was initial agreement between the two methods among 92% (122 of 133) of the isolates tested. The 11 isolates yielding discrepant results with RYP and 20C were further studied by morphological examination and with the VIT card. These studies confirmed results obtained with RYP in three cases, giving the RYP an overall accuracy of 94% (125 of 133). Two of the three isolates were identified as C. glabrata, producing blastoconidia but no pseudohyphae on cornmeal agar. One isolate was identified as C. tropicalis, showing robust blastoconidia arranged in small groups along pseudohyphae and true hyphae microscopically. These three isolates gave excellent identification (99%) on the Vitek instrument. The eight remaining yeasts (misidentified by RYP) were confirmed morphologically (see Table 2). Seven of these isolates were correctly identified by the Vitek instrument; one isolate of Candida parapsilosis was misidentified by Vitek as C. tropicalis. Microscopic morphology confirmed 20C results in all eight cases. Certain isolates were problematic (Table 2). Several yeasts consistently gave no acceptable codes; these yeasts were Trichosporon beigelii (two isolates) and Rhodotorula rubra (one isolate), rarely isolated from our patient population. A cost analysis of each system including materials and labor showed $7.85, $7.40, and $7.00 for an RYP test, a 20C test, or a Vitek assay, respectively.

TABLE 1.

Identification of yeast isolates by the RYP system

Species identification	No. tested	No. of isolates identified
		Correctly	Incorrectly
Candida albicans	57	57	0
Candida tropicalis	26	25	1
Candida glabrata	23	23	0
Candida parapsilosis	10	8	2
Candida krusei	3	3	0
Candida lusitaniae	2	0	2
Cryptococcus albidus	2	2	0
Saccharomyces cerevisiae	2	2	0
Trichosporon beigelii	2	0	2
Blastoschizomyces capitatus	1	1	0
Candida guilliermondii	1	1	0
Cryptococcus neoformans	1	1	0
Cryptococcus laurentii	1	1	0
Rhodotorula rubra	1	0	1
Torulopsis candida	1	1	0
Total	133	125	8

TABLE 2.

Yeasts misidentified by the RYP system

Species identification	No. of isolates misidentified/tested	Identification by RYP	Microscopic morphology
Candida parapsilosis	2/10	Candida lusitaniae	Curved pseudohyphae, with blastoconidia; large giant cells
Candida lusitaniae	2/2	Torulopsis candida	Rare pseudohypha with blastoconidia
		Candida parapsilosis	Rare pseudohypha with blastoconidia
Candida tropicalis	1/26	Candida stellatoidea	Pseudohyphae with single blastoconidia
Trichosporon beigelii	2/2	No code	Pseudohyphae, true hyphae, blastoconidia, arthroconidia
Rhodotorula rubra	1/1	No code	Blastoconidia (red pigmentation macroscopically)

Conventional methods for the identification of clinically significant yeasts are time-consuming and labor-intensive. Rapid screening methods, including the ability to produce germ tubes (8, 13) and production of the enzymes beta-galactosaminidase and l-proline aminopeptidase (2, 4, 7, 11), may be used to presumptively identify C. albicans. Other rapid screening methods involve the use of differential culture media (1, 6, 10, 12). Although cellular fatty acid analysis by gas-liquid chromatography has been successfully used to identify mycobacteria and other fastidious microorganisms, the current database is considered to be inadequate for the routine identification of clinically significant yeasts (3). The 20C system was one of the first standardized commercial products used for rapid yeast identification and is frequently used as a reference method (5, 8). Although the VIT Yeast Biochemical Card has been reported to be inadequate for the rapid identification of uncommon yeast isolates, it has been shown to provide accurate identification of the most commonly isolated yeasts (5). Although both the 20C and VIT tests yield accurate identification of most yeast isolates from clinical sources, a minimum of 24 and often 48 h of incubation is required, especially for slowly growing yeasts (8). Results of our study show that the RYP system is highly accurate in identifying the most commonly isolated yeasts. The cost per test for the RYP system is comparable to that of the API20C system; the RYP assay is an accurate, user-friendly method for same-day identification of Candida species.