An Assessment of a New Rapid Multiplex PCR Assay for the Diagnosis of Meningoencephalitis

The rapid and broad microbiological diagnosis of meningoencephalitis (ME) has been possible thanks to the development of multiplex PCR tests applied to cerebrospinal fluid (CSF). We aimed to assess a new multiplex PCR panel (the QIAstat-Dx ME panel), which we compared to conventional diagnostic tools and the Biofire FilmArray ME Panel. The pathogens analyzed using both methods were Escherichia coli K1, Haemophilus influenzae, Listeria monocytogenes, Neisseria meningitidis, Streptococcus agalactiae, Streptococcus pneumoniae, Enterovirus, herpes simplex virus 1–2, human herpesvirus 6, human parechovirus, varicella zoster virus, and Cryptococcus neoformans/gattii. We used sensitivity, specificity, PPV, NPV, and kappa correlation index parameters to achieve our objective. Fifty CSF samples from patients with suspected ME were included. When conventional methods were used, 28 CSF samples (56%) were positive. The sensitivity and specificity for QIAstat-Dx/ME were 96.43% (CI95%, 79.8–99.8) and 95.24% (75.2–99.7), respectively, whereas the PPV and NPV were 96.43% (79.8–99.8) and 95.24% (75.1–99.7), respectively. The kappa value was 91.67%. Conclusions: A high correlation of the QIAstat-Dx ME panel with reference methods was shown. QIAstat-Dx ME is a rapid-PCR technique to be applied in patients with suspected ME with a high accuracy.


Introduction
Central nervous system (CNS) infections can be caused by bacteria, viruses, or fungi and are clinical entities associated with high morbidity and mortality [1][2][3].Infections like these are frequently encountered in clinical and emergency department settings, presenting challenges due to their highly variable clinical manifestations, the significant differences in CSF characteristics between viral and bacterial infections, and the complexities involved in establishing a prompt and accurate diagnosis.
Therefore, to optimize directed therapy and hopefully improve patient outcomes, the early and accurate identification of the etiological agent is critical [3][4][5].This has been made possible through the development of multiplex PCR (M-PCR) tests to detect the most common microorganisms causing encephalitis, meningitis, or meningoencephalitis (ME) in cerebrospinal fluid (CSF) [6].At present, there are two M-PCR tests on the market: the Filmarray ME Panel, BioFire Diagnostics (Salt Lake City, UT, USA), launched in 2015, Diagnostics 2024, 14, 802 2 of 12 which is based on a nested PCR followed by a melt curve analysis in a microarray format, and the recently introduced QIAstat-Dx ME panel cassette (QIAGEN, Hilden, Germany), which is based on a multiplex real-time PCR platform.Both methods target potential ME pathogens in CSF, Escherichia coli K1, Haemophilus influenzae, Listeria monocytogenes, Neisseria meningitidis, Streptococcus agalactiae, Streptococcus pneumoniae, enterovirus, herpes simplex virus 1-2 (HSV-1 and 2), human herpes virus 6 (HHV-6), human parechovirus (HPV), varicella zoster virus (VZV), and Cryptococcus neoformans/gattii.The QIAstat-Dx ME panel has two additional bacterial targets which are for the detection of Mycoplasma pneumoniae and Streptococcus pyogenes, whereas the Filmarray ME panel has a target for CMV.The main purpose of this study was to assess the new QIAstat-Dx ME panel using comparisons with data obtained using a FilmArray ME panel, using conventional laboratory diagnostic methods and clinical diagnosis as a gold standard.

Study Design and Clinical Samples
This study was conducted using 50 consecutive CSF samples from patients with a clinically suspected central nervous system infection which were analyzed in parallel using Gram staining, a Filmarray ME panel (BioFire Diagnostics; Salt Lake City, UT, USA, and conventional methods, meaning culture, antigen detection, and rt-PCR methods (see Section 2.2), as indicated in algorithm Figure 1.The remaining volume of each sample was frozen at −20 • C to be thawed and analyzed later using the QIAstat-Dx ME panel (QIAGEN, Hilden, Germany).Positive samples were frozen at −20 • C for a period lasting between 1 and 6 months The motivation for this analysis stemmed from the potential advantages of the QIAstat-Dx ME test, particularly its capability to perform amplification curve analyses and assess the cycle threshold (Ct) value upon obtaining a positive result.
the Filmarray ME Panel, BioFire Diagnostics (Salt Lake City, UT, USA), launched in 2015, which is based on a nested PCR followed by a melt curve analysis in a microarray format, and the recently introduced QIAstat-Dx ME panel cassette (QIAGEN, Hilden, Germany), which is based on a multiplex real-time PCR platform.Both methods target potential ME pathogens in CSF, Escherichia coli K1, Haemophilus influenzae, Listeria monocytogenes, Neisseria meningitidis, Streptococcus agalactiae, Streptococcus pneumoniae, enterovirus, herpes simplex virus 1-2 (HSV-1 and 2), human herpes virus 6 (HHV-6), human parechovirus (HPV), varicella zoster virus (VZV), and Cryptococcus neoformans/gattii.The QIAstat-Dx ME panel has two additional bacterial targets which are for the detection of Mycoplasma pneumoniae and Streptococcus pyogenes, whereas the Filmarray ME panel has a target for CMV.The main purpose of this study was to assess the new QIAstat-Dx ME panel using comparisons with data obtained using a FilmArray ME panel, using conventional laboratory diagnostic methods and clinical diagnosis as a gold standard.

Study Design and Clinical Samples
This study was conducted using 50 consecutive CSF samples from patients with a clinically suspected central nervous system infection which were analyzed in parallel using Gram staining, a Filmarray ME panel (BioFire Diagnostics; Salt Lake City, UT, USA, and conventional methods, meaning culture, antigen detection, and rt-PCR methods (see Section 2.2), as indicated in algorithm Figure 1.The remaining volume of each sample was frozen at −20 °C to be thawed and analyzed later using the QIAstat-Dx ME panel (QI-AGEN, Hilden, Germany).Positive samples were frozen at −20 °C for a period lasting between 1 and 6 months The motivation for this analysis stemmed from the potential advantages of the QIAstat-Dx ME test, particularly its capability to perform amplification curve analyses and assess the cycle threshold (Ct) value upon obtaining a positive result.
CSF cytology and biochemistry results, the final diagnosis tools of this study, were collected retrospectively.The initial decision to use the Filmarray ME panel was made by the attending physician after consulting with specialists in infectious diseases and/or microbiology, guided by a clinical suspicion of ME.

Conventional Methods
The conventional microbiology protocol for CSF samples with a suspicion of ME includes the inoculation of blood and chocolate agar, thioglycolate broth, an S. pneumoniae antigen (BinaxNOW S. pneumoniae Antigen Card, BinaxNOW, Abbott, Chicago, IL, USA), and the detection of a Cryptococcus neoformans antigen (Remel™ Cryptococcus Antigen Test Kits, Thermo.Scientific, Lenexa, KS, USA).Complementary tests were those that were CSF cytology and biochemistry results, the final diagnosis tools of this study, were collected retrospectively.The initial decision to use the Filmarray ME panel was made by the attending physician after consulting with specialists in infectious diseases and/or microbiology, guided by a clinical suspicion of ME.

Conventional Methods
The conventional microbiology protocol for CSF samples with a suspicion of ME includes the inoculation of blood and chocolate agar, thioglycolate broth, an S. pneumoniae antigen (BinaxNOW S. pneumoniae Antigen Card, BinaxNOW, Abbott, Chicago, IL, USA), and the detection of a Cryptococcus neoformans antigen (Remel™ Cryptococcus Antigen Test Kits, Thermo.Scientific, Lenexa, KS, USA).Complementary tests were those that were performed additionally on the CSF samples or those in which discrepancies were observed.
The analytical limits of detection for viruses using conventional methods were 119 copies/mL for herpes viruses, 69 copies/mL for VZV, 183 IU/mL for HHV-6, and 3.2 copies/mL for enteroviruses.

Multiplex PCRs
A volume of 200 µL of CSF was used for both the Filmarray ME, (BioFire Diagnostics Salt Lake City, UT, USA), and QIAstat-Dx ME (QIAGEN, Hilden, Germany) analyses.To keep the processes as similar as possible, all samples were handled in the same biosafety hood in the microbiology laboratory with the necessary safety and hygiene measures for handling this type of sample, cleaning the hood after processing each sample.Both techniques were performed according to the manufacturer's instructions.
The M-PCR results were considered true (negative and positive) if they were consistent with the results obtained by conventional methods.
The detection limits for viruses for the Filmarray ME and QIAstat-Dx ME methods were 281 and 250 TCID50/mL for HSV-1, 28 and 50 TCID50/mL for HSV-2, 170 copies/mL and 1660 copies/mL for VZV, and 31,300 copies/mL and 10,000 copies/mL for HHV-6, respectively.For the enterovirus, both had a limit of detection of 5 TCID50/mL.This information was obtained from the manufacturers.

Definitions and Final Diagnosis Assignment
ME was defined by the presence of an inflammatory process of the brain in association with clinical evidence of neurologic dysfunction and/or signs of meningeal irritation.The final diagnosis of an episode was made by the investigators (G.C., P.P-A., and J.V.) after a thorough evaluation of the microbiological and radiological results, clinical evolution, response to treatment, and the presence or absence of an alternative diagnosis.

Statistical Analysis
The sensitivity of the test was calculated as (true positive, TP)/(TP + false negative (FN)), and the specificity was calculated as (true negative (TN))/(TN + false positive (FP)).The positive likelihood ratio (pLR) was calculated as sensitivity/(1-specificity), and the negative likelihood ratio (nLR) was calculated as (1-sensitivity)/specificity.The positive predictive value (PPV) was calculated as TP/(TP + FP) and the negative predictive value (NPV) was calculated as TN/(TN + FN).Accuracy was calculated as (TN + TP(TP + FP + FN + TN) [7].Cohen's kappa coefficient test was used to assess the level of agreement between the different assessment methods, Filmarray ME and QIAstat-Dx ME, and the conventional methodology.Classification of kappa values included "poor" (0.00), "slight" (0 to 0.20), "fair agreement" (0.21 to 0.40), "moderate agreement" (0.41 to 0.60), "substantial" (0.61 to 0.80), and "complete agreement" (>0.8).Data were analyzed using Stata Statistical Software Release 18 (StataCorp, College Station, TX, USA).A p of less than 0.5 was considered a statistical significance level.The sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and kappa correlation index were calculated, considering the conventional methods as the gold standard.Both techniques yielded a result of "Not detected" or "Invalid" when the result was negative or invalid, respectively.FilmArray showed the positive results as "detected" and provided melting curves.QIA/ME showed positive results as "detected" and provided a cycle threshold (Ct).

Results
Fifty CSF samples were analyzed.Table 1 presents the results, including the CSF cytological and biochemical characteristics.Pathogens were detected by conventional methods in 29 samples (58%): 12/29 (41%) were bacterial (five S. pneumoniae, three L. monocytogenes, two S. agalactiae, one N. meningitidis, and one Acinetobacter baumannii), and 17/29 (59%) were viral cases of ME (seven HSV-1, four VZV, three HSV-2, two HHV-6, and one enterovirus).Twelve samples were determined to be negative by conventional methods.In 32 (64%) cases, there was concordance between all three methods: conventional methods and Filmarray ME and QIAstat-Dx ME panels.
The Filmarray ME panel reported seven CSF samples with single-pathogen false positive results and five CSF samples with polymicrobial false positive results.Four were false negative results.The QIAstat-Dx ME panel reported only one false positive and one false negative result.The false positives reported by Filmarray ME were as follows: nine HSV-1, two H. influenzae, two S. agalactiae, two S. pneumoniae, one E. coli K1, one CMV, and one HSV-2.The false negatives included three HSV-1 and two S. pneumoniae results.The only false positive result reported by the QIAstat-Dx ME panel was a VZV result, while the only false negative was an HSV-1 result.Table 3 shows the discrepancies observed between the M-PCR diagnostic techniques and conventional methods.Abbreviations.HSV = herpes simplex virus; HHV-6 = human herpesvirus 6; VZV = varicella zoster virus; CMV = cytomegalovirus; CAR-T = chimeric antigen receptor T-cell; MELAS = mitochondrial encephalopathy, lactic acidosis, and stroke-like episodes.Missing results have been denoted with hyphens.In some cases, with low white blood cells, neutrophil and lymphocyte percentages were not determined.Pleocytosis (>10 leukocytes/mm 3 ) and high protein CSF levels (>600 mg/dL) are commonly found in encephalitis and meningitis.However, normal CSF cells and protein levels can be found, particularly in early viral cases.

Discussion
The early and accurate identification of the etiological agent causing ME is crucial for patient management [1,2].A meta-analysis of 13 articles [8] showed high sensitivity (90%) and specificity (97%) for Filmarray ME results; however, there are still doubts about the reliability of certain M-PCR results in clinical practice [9][10][11].Trujillo-Gomez et al. conducted another meta-analysis including 19 studies and found high specificity for the technique.However, the sensitivity varied from 89.5% to 93.5% depending on the reference method used [12].
Recently, Humisto et al. [13] published a paper comparing Filmarray ME and QIAstat-Dx ME techniques for the early diagnosis of ME, reporting that Filmarray ME was more reliable than QIAstat-Dx ME, with 0% and 6.5% error rates, respectively, and concluded that the BioFire FilmArray meningitis/encephalitis panel produced more positive results than the QIAstat-Dx meningitis/encephalitis panel in herpesvirus analyses.However, no clinical data were taken into consideration in Humisto's study.Our study was in agreement with the above-mentioned results, showing a higher number of positive samples for herpesvirus in the Filmarray panel than in the QIAstat-Dx panel; however, we considered them false positives since our specific PCR results for herpesvirus were negative and, in addition, the final clinical diagnosis in all these samples but one was not herpetic encephalitis.In contrast, another recent publication showed comparable performance between both panels without significant differences [14].
Some previous studies have reported similar results concerning the use of the Filmarray ME panel.Johan Lindström et al. [9] analyzed 4199 CSF samples and obtained a sensitivity for HSV-1 of 82.4%.Amy L. Leber et al. [11] reported low sensitivities for some viruses, especially for HHV-6 (85.7%).False positives can be explained by two different reasons: (i) the CSF collection by lumbar puncture may have been traumatic and we are actually detecting traces of pathogenic genetic material present in the blood and not in the CSF [15], as may be the case for herpesviruses; however, if this scenario occurs, it should still be detected by conventional methods; and (ii) accidental contamination has occurred at some point in the process.In patients with suspected ME, findings of a false positive, especially for herpes simplex, are problematic because they may lead to unnecessary treatment with consequent drug toxicity [16].In addition, false positive findings may complicate the search for other possible explanations for the clinical picture.
With the FilmArray ME panel, the most frequent false positive we obtained was for HSV-1, which has also been reported in other series [17].In contrast, HSV-1 was the main pathogen involved in false negative results in other studies [8,9,18].The microbiological ability to interpret a positive FilmArray ME result is very limited since we can only observe the melting curve, making it difficult to differentiate between true positives and false positives or contaminations [18].It should be noted that the microorganisms responsible for most of the false positives obtained in our study coincide with those of other studies [9,11,[19][20][21].In a review by Trujillo-Gomez et al. [12] including 7090 CSF samples, the respective sensitivities and specificities were 87.5% and 98.5% for S. pneumoniae; 71.5% and 99.5% for S. agalactiae, and 75% and 99% for HSV-1.In addition to the better overall performance of the QIAstat-Dx ME panel in this study, one additional advantage of this method is that once a microorganism is detected, the amplification curve and Ct value can be assessed (Figure S1).This can be helpful in clinical interpretation.In our cohort, only one false positive was obtained by the QIAstat-Dx ME panel for VZV.In this case, the QIAstat-Dx ME panel showed a correct sigmoidal curve with a Ct of 38.5.The patient had already been diagnosed with VZV ME 15 days earlier in another hospital.This clinical picture, together with the high Ct value obtained, allowed this case to be interpreted as the detection of remnants of past ME.The patient was finally diagnosed with VZV-associated vasculitis, which responded properly to treatment with methylprednisolone, acetylsalicylic acid, and cyclophosphamide.Figure S1 shows the difference between the Ct of this patient and that of a case of active VZV encephalitis.This case leads to the need to point out that despite G.C., P.P.-A.and J.V.; resources, J.V. and M.Á.M.; data curation G.C.; writing-original draft preparation, G.C., P.P.-A.and C.C.-P.; writing-review and editing, A.S. and J.V.; supervision, J.V.; project administration, J.V. All authors have read and agreed to the published version of the manuscript.Informed Consent Statement: Informed consent was waived, as no intervention was involved, and no patient-identifiable information was included.

Funding:
ISGlobal (CEX2018-000806-S) is recipient of a Severo Ochoa Award of Excellence from MINECO (Government of Spain).ISGlobal acknowledges support from the Generalitat de Catalunya through the CERCA Program.This work was also supported by the Instituto de Salud Carlos III [Carlos III Institute of Health] (PI20/00766) and award 2021SGR01569 from the Agència de Gestió d'Ajuts Universitaris i de Recerca of the Generalitat de Catalunya [Agency for Management of University and Research Grants of the Catalan Government].Institutional Review Board Statement: This study was conducted in accordance with the tenets of the Declaration of Helsinki and was approved by the Hospital Clinic's Ethics Committee (HCB/2022/0943).

Table 1 .
Summary of cases analyzed.

Table 2 .
Comparison of sensitivity, specificity, PPV, NPV, and kappa correlation index for both M-PCR diagnostic techniques compared to the conventional methods.

Table 3 .
Discrepancies observed between M-PCR diagnostic techniques and conventional methods.