Strengths and Limitations of T2Candida Implementation From an Antimicrobial Stewardship Perspective

Background and Introduction

Candidemia is the fourth leading cause of hospital-associated bloodstream infections in the United States and is associated with all-cause mortality rates of 25% to 40%.¹ The estimated total economic burden of invasive candidiasis (IC) infections, including candidemia, in the United States is upward of $1.2 billion per year.² Blood cultures (BCs) remain the gold standard for diagnosing IC but are positive in only 50% of cases.³ In cases in which BCs are positive, the time to positivity of Candida species can be significantly prolonged. Candida albicans, Candida tropicalis, and Candida parapsilosis tend to be detected within 24 to 36 hours, whereas cultures with Candida glabrata can take up to 80 hours to turn positive.⁴

Low BC sensitivity, prolonged time to detection, and poor-performing risk scores contribute to delayed initiation of antifungal therapy when it is needed.⁵ Delayed initiation of antifungal therapy has been shown to be an independent predictor of mortality in patients with candidemia.⁶ Additionally, these same factors may lead to inappropriate use of antifungal therapy.⁵ Both scenarios present different but significant challenges to antimicrobial stewardship programs (ASPs) when assessing which patients should or should not receive antifungal treatment.

T2Candida Panel

The T2Candida Panel, designed to be used with the T2Dx Instrument, is a novel, rapid diagnostic assay that uses polymerase chain reaction and magnetic resonance to identify the 5 most common Candida species from whole blood samples: C albicans, C tropicalis, C parapsilosis, C glabrata, and C krusei.⁷ The test is performed independent of BC and results are available in approximately 3 to 5 hours. Results are grouped based on antifungal resistance patterns: C albicans/C tropicalis, C parapsilosis, and C glabrata/krusei. A result of “positive” or “target not detected” is assigned to each group.⁷

The FDA approved the T2Candida Panel in 2014 based on the results of the DIRECT trial (NCT01525095), in which clinical BC samples were tested after being inoculated with less than 1 to 100 colony-forming units (CFUs)/mL of the 5 Candida species included in the panel. Overall sensitivity and specificity were 91.1% and 99.4%, respectively. Positive predictive value (PPV) and negative predictive value (NPV) were reported as 72% to 84% and 99%, respectively. Lower limits of detection previously determined were redemonstrated in the DIRECT trial: 3 CFUs/mL for C albicans and C tropicalis, 2 CFUs/mL for C glabrata and C krusei, and 1 CFU/mL for C parapsilosis.⁸

Strengths and Limitations

Since its approval and introduction to the market, T2Candida has been evaluated for its performance against clinical samples collected in real time from patients with proven or suspected candidemia. The sensitivity and specificity rates have ranged from 89% to 100% and 92% to 96%, respectively, whereas its PPV and NPV have ranged from 25% to 60% and 89% to 100%, respectively.^9,10 Unfortunately, T2Candida has shown suboptimal performance when identifying deep-seated IC in patients without candidemia. Recent studies showed T2Candidawas positive in only 27% to 33% of culture-proven IC cases in which BCs were negative.^11,12

When contemplating the implementation of the T2Candida Panel at an institution, ASPs must consider its strengths and limitations (Table). Its undeniable strengths include its independence from BC positivity, low limits of detection, rapid turnaround times, excellent NPVs, and modest PPVs. These factors have contributed to positive outcomes from several, real-world studies, the majority of which evaluated patients with suspected or proven IC. T2Candida implementation has consistently shown statistically significant reductions in time to detection of candidemia and time to appropriate antifungal therapy.^5,13,14 Similarly, significant reductions in total duration of therapy (DOT) have also been shown.⁵ One study evaluated the financial effect of T2Candida use and saw a $280 cost savings per patient as a result of reductions in overall duration of antifungal therapy.⁵ Another significant strength is that, unlike many other rapid diagnostic tests, the T2Candida Panel used on the T2Dx Instrument is fully automated and does not require manual, time-consuming DNA extraction.⁷ This reduces the risk of human error/contamination and overall burden on the microbiology department. These outcomes are incredibly important for ASPs to recognize when considering implementing the T2CandidaPanel as a tool for improved antifungal stewardship.

However, T2Candida has limitations that all clinicians should be aware of and consider when initially implementing the technology and subsequently interpreting results. Although early initiation of antifungal therapy has been shown to significantly reduce mortality, studies evaluating outcomes when using T2Candida have shown only non–statistically significant trends of reduced rates of death.^5,13,14 Significant reductions in time to appropriate therapy associated with T2Candida use may not translate to significantly reduced mortality because of small sample sizes and the inability to reach power to identify significant differences. Similarly, no significant reductions in overall length of stay (LOS) or intensive care unit LOS have been shown.^5,13,14 This negatively affects expected cost savings associated with T2Candida use.

T2Candida has not shown as significant an effect on rates of appropriate antifungal de-escalation as one may expect. Two publications showed that only 27% to 47% of patients had their empiric antifungal therapy discontinued following a negative T2Candida result.^5,15 Despite consistently high NPVs and years of the T2Candida Panelbeing available commercially, providers may not feel comfortable discontinuing treatment based solely on a negative result, especially in patients with a high clinical suspicion of IC. One potential reason for this is T2Candida’s poor sensitivity for detecting deep-seated IC in patients without candidemia.^11,12 If clinical suspicion for noncandidemic IC is high, it would be prudent to continue antifungal therapy despite a negative T2Candida result.

Unfortunately, there is a widespread lack of knowledge of which populations may benefit most from T2Candida testing. This has led to significant variations in the type of patients represented in clinical studies and is likely responsible for the low PPVs that have been associated with T2Candida.^9,10 Although the majority of patients tested in these studies had suspected or proven IC, the inclusion criteria differed widely, using various combinations of the patient’s clinical status and present risk factors to determine eligibility. Taking this into account, a positive T2Candida may represent a false-positive result in cases in which BCs are negative and there is a low clinical suspicion for deep-seated IC. Positive results must be interpreted carefully depending on the patient’s clinical picture.

Another potential limitation is the time and cost required to obtain and validate equipment, adequately train personnel, and integrate testing into workflow. Although this may not significantly affect hospitals with more available resources, it would likely be challenging for small, community institutions. Also, T2Candida has shown poor performance, reported as high rates of invalid results, when tested on previously frozen samples. This may be a challenge for institutions that want to offer T2Candida testing offsite, as samples must be refrigerated and tested within 72 hours to avoid the need for freezing.⁷

Unfortunately, the T2Candida Panel can detect only 5 species of Candida. This should not be a significant limitation because these 5 species are responsible for 95% of IC infections in the United States.¹⁶ However, each institution should evaluate its own etiologic distributions prior to implementation. Also, like the majority of commercially available rapid diagnostic tests, T2Candida does not provide susceptibility data. Although Candida species tend to have stable resistance patterns and susceptibility can often be inferred based on the species identified, only positive BCs can be used to determine true susceptibilities. Lastly, T2Candida is validated for testing on only whole blood samples. Samples from alternative sources (i.e. peritoneal fluid) have been tested in small studies and this will likely continue to be an area of investigation in the future.¹⁷

Future Directions

Hope exists that with continued use and research, current utilization can be optimized and discoveries of new uses can be made. Future studies should evaluate and identify the patient populations that will most benefit from testing. Until this population is defined, the medical community will continue to see low PPVs and have to consider the possibility of false-positive results in patients with negative BCs. Another area of research could be focused on the performance of the T2Candida Panel as a monitoring tool for the clearance of candidemia. Thus far, studies have identified that T2Candida results can remain positive long after BCs have cleared.¹⁸ The significance of this and whether antifungals should be continued until a negative result is obtained are unclear. Lastly, a new panel, the T2Cauris Panel, is currently being evaluated for the detection of Candida auris, Candida duobushaemulonii, and Candida haemulonii in skin and blood samples.¹⁹ As C auris is a serious, global, public health threat, its rapid and accurate identification is imperative because identifying and isolating colonized and/or infected patients in a timely manner is key to slowing the spread.

Conclusions

ASPs must carefully consider all strengths and limitations associated with T2Candida implementation and use. Despite its limitations, T2Candida use positively affects common antifungal stewardship initiatives and should be considered a breakthrough technology for the detection of candidemia.

References

^{1. Tsay SV, Mu Y, Williams S, et al. Burden of candidemia in the United States, 2017. Clin Infect Dis. 2020;71(9):e449-e453. doi:10.1093/cid/ciaa193}

^{2. Benedict K, Jackson BR, Chiller T, Beer KD. Estimation of direct healthcare costs of fungal diseases in the United States. Clin Infect Dis. 2019;68(11):1791-1797. doi:10.1093/cid/ciy776}

^{3. Clancy CJ, Nguyen MH. Finding the "missing 50%" of invasive candidiasis: how nonculture diagnostics will improve understanding of disease spectrum and transform patient care. Clin Infect Dis. 2013;56(9):1284-92. doi:10.1093/cid/cit006}

^{4. Kim SH, Yoon YK, Kim MJ, Sohn JW. Clinical impact of time to positivity for candida species on mortality in patients with candidaemia. J Antimicrob Chemother. 2013;68(12):2890-2897. doi:10.1093/jac/dkt256}

^{5. Patch ME, Weisz E, Cubillos A, Estrada SJ, Pfaller MA. Impact of rapid, culture-independent diagnosis of candidaemia and invasive candidiasis in a community health system. J Antimicrob Chemother. 2018;73(suppl 4):iv27-iv30. doi:10.1093/jac/dky046}

^{6. Garey KW, Rege M, Pai MP, et al. Time to initiation of fluconazole therapy impacts mortality in patients with candidemia: a multi-institutional study. Clin Infect Dis. 2006;43(1):25-31. doi:10.1086/504810}

^{7. T2Candida panel: faster targeted therapy and reduction in empiric antifungal costs. T2 Biosystems. Accessed July 26, 2021, https://www.t2biosystems.com/products-technology/t2candida-panel/}

^{8. Mylonakis E, Clancy CJ, Ostrosky-Zeichner L, et al. T2 magnetic resonance assay for the rapid diagnosis of candidemia in whole blood: a clinical trial. Clin Infect Dis. 2015;60(6):892-899. doi:10.1093/cid/ciu959}

^{9. Giannella M, Paolucci M, Roncarati G, et al. Potential role of T2Candida in the management of empirical antifungal treatment in patients at high risk of candidaemia: a pilot single-centre study. J Antimicrob Chemother. 2018;73(10):2856-2859. doi:10.1093/jac/dky247}

^{10. Clancy CJ, Pappas PG, Vazquez J, et al. Detecting infections rapidly and easily for candidemia trial, part 2 (DIRECT2): a prospective, multicenter study of the T2Candida Panel. Clin Infect Dis. 2018;66(11):1678-1686. doi:10.1093/cid/cix1095}

^{11. Zurl C, Prattes J, Zollner-Schwetz I, et al. T2Candida magnetic resonance in patients with invasive candidiasis: strengths and limitations. Med Mycol. 2020;58(5):632-638. doi:10.1093/mmy/myz101}

^{12. Lamoth F, Clancy CJ, Tissot F, et al. Performance of the T2Candida panel for the diagnosis of intra-abdominal candidiasis. Open Forum Infect Dis. 2020;7(3):ofaa075. doi:10.1093/ofid/ofaa075}

^{13. Steuber TD, Butler L, Sawyer A, Chappell R, Edwards J. Comparison of blood cultures versus T2 Candida panel in management of candidemia at a large community hospital. Eur J Clin Microbiol Infect Dis.2021;40(5):997-1001. doi:10.1007/s10096-020-04144-8}

^{14. Wilson NN, Alangaden, G, Tibbetts RJ, et al. T2 magnetic resonance assay improves timely management of candidemia. J Antimicrobial Stewardship. 2017;1(1):12-18.}

^{15. Steuber TD, Tucker-Heard G, Edwards J, Sawyer A, Thottacherry E, Hassoun A. Utilization and impact of a rapid candida panel on antifungal stewardship program within a large community hospital. Diagn Microbiol Infect Dis. 2020;97(4):115086. doi:10.1016/j.diagmicrobio.2020.115086}

^{16. Ricotta EE, Lai YL, Babiker A, et al. Invasive candidiasis species distribution and trends, United States, 2009-2017. J Infect Dis. 2021;223(7):1295-1302. doi:10.1093/infdis/jiaa502}

^{17. Kouri AM, Kieffer TW, Nailescu C, et al. Evaluation of T2Candida panel for detection of candida in peritoneal dialysates. Perit Dial Int. 2020;40(1):96-99. doi:10.1177/0896860819879874}

^{18. Mylonakis E, Zacharioudakis IM, Clancy CJ, Nguyen MH, Pappas PG. Efficacy of T2 magnetic resonance assay in monitoring candidemia after initiation of antifungal therapy: the serial therapeutic and antifungal monitoring protocol (STAMP) trial. J Clin Microbiol. 2018;56(4). doi:10.1128/jcm.01756-17}

^{19. T2Cauris panel: direct detection of emerging superbug in patient skin and blood samples. T2 Biosystems. Accessed July 26, 2021. https://www.t2biosystems.com/products-technology/pipeline/t2candida-auris-panel-ruo/}