Clinical performance evaluation of the BioFire Joint Infection Panel

ABSTRACT

The BioFire Joint Infection (JI) Panel offers a significant advancement in the rapid diagnosis of joint infections by facilitating the simultaneous detection of multiple bacterial and fungal pathogens, as well as resistance markers, directly from synovial fluid samples. An article published in the Journal of Clinical Microbiology by Moran et al. (J Clin Microbiol 62:e00182-24, 2024, https://doi.org/10.1128/jcm.00182-24) presents both prospective and retrospective analyses of the panel’s real-world clinical application. The study highlights the panel’s benefits, such as its rapid turnaround time and ability to identify challenging pathogens, while also discussing its limitations, particularly in detecting certain off-panel organisms.

COMMENTARY

The advent of multiplex syndromic panels has revolutionized the diagnostic landscape in clinical microbiology, enabling the simultaneous detection of multiple pathogens directly from clinical specimens. Since the first FDA-approved multiplex panel for respiratory viruses in 2008, the technology has expanded to include panels for gastroenteritis, meningitis, pneumonia, and more. These panels offer rapid turnaround times and high diagnostic accuracy, significantly impacting patient management by facilitating timely and appropriate treatment decisions.

Recently, the introduction of the BioFire Joint Infection Panel (bioMérieux, Marcy-l'Étoile, France) marks a significant advancement in the diagnosis of joint infections. This panel allows for the rapid identification of 29 bacterial targets, two yeasts, and eight resistance markers directly from synovial fluid samples. Joint infections, often challenging to diagnose due to the wide array of potential pathogens and the difficulty in obtaining adequate sample volumes, now benefit from the streamlined and comprehensive diagnostic approach offered by the multiplex PCR technology.

In a recent manufacturer-independent study by Moran et al. (1), the performance of the BioFire JI Panel was evaluated using both retrospective and prospectively collected synovial fluid specimens. The retrospective analysis involved 63 frozen specimens selected from the NorthShore University HealthSystem Clinical Microbiology Laboratory chosen based on positive culture results or the presence of white blood cells or organisms on Gram stain, indicating high levels of joint inflammation. Of these, 28 specimens were culture-positive, growing either Gram-positive or Gram-negative bacteria that are targets on the BioFire JI Panel. Additionally, 35 specimens were either culture-negative, with some showing leukocytes or organisms on Gram stain or growing organisms not included in the JI panel, such as Staphylococcus epidermidis (in two cases), Staphylococcus schleiferi, and Pasteurella multocida.

For the 28 culture-positive specimens, the BioFire JI Panel detected pathogens in 26 cases, resulting in a positive percent agreement (PPA) of 92.8% (95% confidence interval [CI]: 76.5–99.1%). Specifically, the panel correctly identified two methicillin-resistant Staphylococcus aureus (MRSA) cases by detecting the mecA/MREJ target. Among the 35 culture-negative or off-panel target specimens, the panel yielded a negative percent agreement (NPA) of 97.1% (95% CI: 85.1–99.9%), with 34 specimens testing negative on the panel.

In the prospective study, 104 fresh joint fluid specimens were tested with the JI panel within 1 week of collection. Out of these, 14 organisms were isolated from 12 specimens in routine culture, with the JI panel detecting pathogens in seven specimens. The detection of off-panel organisms, such as S. epidermidis and Pasteurella multocida, led to the classification of these cases as culture-negative for the purposes of JI panel analysis. This yielded a PPA of 71.4% (95% CI: 29.0–96.3%) and an NPA of 94.8% (95% CI: 88.4–98.3%).

The authors also conducted a thorough discrepancy analysis for specimens with results discordant between the BioFire JI Panel and routine culture. Out of the 167 total joint fluid samples tested, 12 specimens showed discrepancies. These were further investigated using repeated JI panel testing, 16S rRNA sequencing at Mayo Clinic Laboratories, joint fluid analysis, and review of other patient cultures collected around the same time. Among these, six specimens maintained the initial discrepancy upon retesting; one specimen remained partially discrepant; and four matched the routine culture results on retesting. The standardized criteria applied for discrepancy resolution included: favoring results consistent with other positive cultures within a week, matching culture or repeat BioFire results with sequencing data, and considering other discrepancies as inconclusive. The analysis revealed that culture missed significant pathogens, including Streptococcus species, Enterococcus faecalis, and Staphylococcus aureus. Conversely, three false positives identified by culture were confirmed as negative by BioFire, involving coagulase-negative staphylococci, Enterococcus faecalis, and a group D Salmonella. The BioFire JI Panel missed an MRSA detection, though it was caught on repeat testing, and showed false positives for Staphylococcus aureus, coagulase-negative staphylococci, and enterococci.

The study also included an analysis of off-panel organisms, which are pathogens not detected by the JI panel. Initially, these off-panel organisms were classified as culture-negative, consistent with the panel’s limitations. However, acknowledging that these organisms could still cause true joint infections, the researchers reanalyzed the data, treating these cases as “misses“ by the JI panel (three coagulase-negative staphylococci cases and a Pasteurella multocida case). This approach aimed to offer a fuller evaluation of the panel’s overall clinical performance. In this reanalysis, the PPA for the pre-selected and enriched frozen specimens, which mainly included organisms detectable by the panel, decreased to 81.2% (95% CI: 63.6–92.8%). This drop reflected the panel’s inability to detect clinically significant off-panel pathogens. The NPA remained high at 96.8% (95% CI: 83.3–99.9%). For the prospective specimens, where the selection was not biased toward panel-detectable organisms, the adjusted PPA was markedly lower at 41.7% (95% CI: 15.2–72.3%), with the NPA remaining consistent at 94.6% (95% CI: 87.8–98.2%).

The BioFire JI Panel has been evaluated across multiple different clinical studies (Table 1), each contributing insights into its diagnostic performance. The largest U.S.-based study (2) was a multicenter investigation funded and designed by BioFire Diagnostics. This study analyzed 1,544 synovial fluid samples to assess the JI panel’s diagnostic accuracy. Using sequencing as a comparator and conducting discrepancy analysis at BioFire without clinical history data, the study reported an overall sensitivity of 90.5% and a specificity of 99.6% for pathogen detection. Among the 1,544 samples, 202 were positive by standard culture for at least one on-panel organism, while the BioFire JI Panel detected pathogens in 242 specimens. Discrepancy investigations revealed 20 false negatives (FNs) and 70 false positives (FPs), with further PCR testing confirming the FP results in 76 (96.2%) cases. The study identified 75 off-panel organisms, including 53 coagulase-negative staphylococci and eight instances of Cutibacterium acnes. The inability to detect these off-panel organisms, which are significant in chronic infections, remains a limitation of the BioFire JI Panel. The study underscored the high specificity of the BioFire JI Panel, particularly for antimicrobial resistance genes, achieving a PPA and an NPA of 100% and 98.8%, respectively, for detecting resistance markers, such as mecA/C and MREJ (MRSA).

TABLE 1.

Biofire JI Panel evaluation studies

Study	Study type	Location	n	Sensitivity/PPA	Specificity/NPA	Key findings
Esteban et al. (2)	Retrospective multicenter study	USA, Canada, France	1,544	90.9%+ for most organisms, 100% for AMR genes	98.5%+ for all organisms, 95.7%+ for AMR genes	Demonstrated high sensitivity and specificity; noted limitations in detecting coagulase-negative staphylococci and Cutibacterium acnes
Berinson et al. (3)	Retrospective single- center study	Germany	123	100% for on-panel organisms	100%	Highlighted high specificity and turnaround time; limited by few positives and exclusion of important pathogens (coagulase-negative staphylococci and Cutibacterium acnes for prosthetic joint infections)
Azad et al. (4)	Retrospective single-center study	USA	60	91% for on-panel organisms, 56% overall	100%	tMGS showed higher overall sensitivity (93% vs. 56%); highlighted exclusion of significant pathogens (e.g., Staphylococcus epidermidis)
Hoffman et al. (5)	Retrospective analysis of a prospective single-center validation study	Israel	57	56% overall, 92% for on-panel organisms	100%	Emphasized rapid results; noted limited sensitivity
Gaillard et al. (6)	Prospective multicenter study	France	308	84.9%	100%	Highlighted high positive predictive value; addressed missing target species as a limitation
Schoenmakers et al. (7)	Prospective single-center study	Netherlands	45	83% (native septic arthritis), 73% (late acute PJI), 30% (early acute PJI)	100%	Identified clear clinical benefit for native septic arthritis and late acute PJI
Salar-Vidal et al. (8)	Prospective multicenter study	Spain, Portugal	262	69% agreement with culture	91.9%	Highlighted detection of fastidious organisms like Kingella kingae and Neisseria gonorrhoeae; however, JI panel missed five coagulase-negative staphylococci and the following: Escherichia coli (2), Citrobacter (1), Neisseria gonorrhoeae (1), Staphylococcus aureus (1), Streptococcus agalactiae (1), and Streptococcus pyogenes (1)
Gardete-Hartmann et al. (9)	Retrospective single-center study	Austria	268	41.4%	91.10%	Demonstrated utility in unclear cases; identified additional microorganisms
Pascual et al. (10)	Retrospective multicenter study	Europe and Middle East	1,527	88.4% (native joints), 85% (PJI) overall agreement (no discrepancy analysis)		Largest European study; emphasized increased detection of pathogens compared to synovial fluid cultures
Saeed et al. (11)	Retrospective multicenter study	UK and Ireland	399	91.6%	93%	Demonstrated higher diagnostic yield with the BioFire Joint Infection Panel (BJIP) compared to culture; highlighted the panel’s detection of resistant markers and additional organisms like Neisseria gonorrhoeae and Kingella kingae. Addressed limitations in detecting coagulase-negative staphylococci and Cutibacterium acnes

Another large study that included manufacturer authors (10) conducted across 34 clinical sites in 19 European and Middle Eastern countries included 1,527 samples. The study demonstrated an overall agreement of 88.4% for native joints and 85% for prosthetic joint infections (PJIs) when compared to synovial fluid cultures. The JI panel detected more positive samples and microorganisms than traditional synovial fluid cultures, particularly identifying pathogens like Staphylococcus aureus, Streptococcus species, Enterococcus faecalis, Kingella kingae, and Neisseria gonorrhoeae.

Notably, a study by Azad et al. (4) involving 60 samples compared the BioFire JI Panel with 16S rRNA gene-based targeted metagenomic sequencing (tMGS). This study revealed a lower overall sensitivity of 56% for the JI panel compared to 93% for tMGS, underscoring the importance of comprehensive diagnostic tools, especially for detecting key pathogens like Staphylococcus epidermidis.

Several studies have reported on the sensitivity of the BioFire JI Panel, particularly for detecting on-panel target organisms. Notably, these studies (Table 1) have demonstrated sensitivities generally ranging from 80% to 100% for on-panel organisms. For instance, Saeed et al. (11) reported an overall PPA of 91.6% and an NPA of 93%, highlighting the panel’s effectiveness in identifying pathogens, including difficult-to-detect organisms like Neisseria gonorrhoeae and Kingella kingae. Similarly, Gaillard et al. (6) reported a PPA of 84.9% and an NPA of 100%, reinforcing the panel’s reliability for detecting acute arthritis pathogens. These findings are consistent with those of Hoffman et al. (5), who found a sensitivity of 92% for on-panel organisms in their study of 57 samples. Salar-Vidal et al. (8) also highlighted the panel’s utility in identifying fastidious organisms, particularly those challenging to culture traditionally.

However, the overall performance of the BioFire JI Panel is significantly lower when including organisms that are not on the panel. For example, Hoffman et al. (5) reported a sensitivity of just 56% for all detected organisms when considering the broader diagnostic context, including organisms not covered by the panel. This discrepancy underscores a critical limitation: the panel’s inability to detect certain pathogens, such as coagulase-negative staphylococci and Cutibacterium acnes, which are significant in chronic infections.

In one study (7), results were stratified based on clinical scenarios, revealing varying sensitivities: 83% for native septic arthritis, 73% for late acute periprosthetic joint infections (PJIs), and a notably lower sensitivity of 30% for early acute PJIs. The study concluded that while the BioFire JI Panel offers clear clinical benefits for native septic arthritis and late acute PJI, its utility is limited in early acute PJI due to the exclusion of relevant microorganisms like Staphylococcus epidermidis.

Similarly, Gardete-Hartmann et al. (9) categorized results into different scenarios, such as unexpected-negative cultures, unexpected-positive cultures, single-positive intraoperative cultures, and clearly septic and aseptic cases. The study reported an overall sensitivity of 41.4% and a specificity of 91.1%, with the JI panel proving particularly useful in resolving cases with unclear microbiological results. However, it also highlighted the panel’s limitations in not identifying some causative organisms like Staphylococcus epidermidis and Cutibacterium acnes, emphasizing the need for complementary traditional culture methods.

In summary, the BioFire JI Panel offers rapid and specific pathogen detection, particularly excelling at identifying fastidious organisms, such as Neisseria gonorrhoeae and Kingella kingae. The studies by Esteban et al. and Pascual et al., notable for their large sample sizes in American and European cohorts, respectively, reported sensitivities of 90.9% and 88.4% compared to standard culture methods. These studies highlight the panel’s effectiveness, though they lack detailed clinical specimen-level data for discrepancy analysis, limiting a comprehensive understanding of the panel’s performance, especially with mixed or off-panel organisms. In contrast, the Moran et al. study (1), an independent U.S.-based investigation, provided a thorough discrepancy analysis using both retrospective and prospectively collected specimens. This study underscored the challenges in detecting off-panel organisms, such as coagulase-negative staphylococci, highlighting the panel’s limitations. Despite these challenges, the panel’s rapid turnaround time supports quicker therapeutic decisions, which can significantly improve patient outcomes. The consistent recommendation across all studies for complementary use with traditional culture methods emphasizes the necessity of a holistic diagnostic strategy. This combined approach is critical for the accurate diagnosis and effective management of joint infections, particularly in complex orthopedic cases.

The views expressed in this article do not necessarily reflect the views of the journal or of ASM.