Performance of the clarus Aspergillus galactomannan enzyme immunoassay prototype for the diagnosis of invasive pulmonary aspergillosis in serum

Serum galactomannan (GM) testing is essential for diagnosing invasive aspergillosis (IA), particularly in immunocompromised individuals. The global lack of on‐site GM testing capacities necessitates cost‐effective alternatives, such as .the clarus Aspergillus GM enzyme immunoassay prototype (clarus AGM prototype).

Invasive aspergillosis (IA) is a severe fungal infection caused by Aspergillus species. 1 It primarily affects immunocompromised individuals, posing a significant risk to those with haematological malignancies, solid organ malignancies and transplant recipients. 2,3[6] Early and accurate diagnosis of IA is crucial for improving survival rates 7 .][10] Hence, every available diagnostic method, including fungal cultures, measurement of biomarkers, Aspergillus-PCR and histopathological examination, is needed.
Aspergillus serum galactomannan (GM) is a polysaccharide considered one of the mycological diagnostic criteria 10 for diagnosing IA, according to the EORTC/MSG 11 and FUNDICU 12 consensus definitions.Despite its efficacy, many labs lack on-site GM testing, highlighting the need for more accessible alternatives. 13,14er the past two decades, significant progress has been made in developing alternative detection tests for Aspergillus-specific antigens, following the pioneering Platelia AGM assay by Bio-Rad in Marnes-la-Cocquette, France. 15Among these advancements is the innovatively designed clarus Aspergillus GM Enzyme Immunoassay prototype (clarus AGM prototype, Ref AGM101) by IMMY, headquartered in Norman, Oklahoma.7][18] The clarus AGM prototype assay utilises a unique combination of monoclonal antibodies, featuring the mouse-derived ME-A5 human immunoglobulin G monoclonal antibody alongside a proprietary antibody targeting an undisclosed Aspergillus GM antigen, 19 similar to the approach adopted by the IMMY Lateral Flow Assay. 20In contrast, the Platelia AGM assay relies on a rat monoclonal immunoglobulin M antibody, EB-A2, specifically designed to target the Aspergillus GM antigen. 15 this single-centre, cross-sectional study, the main objective was to assess and compare the analytical performance of the new clarus AGM prototype assay with the well-established Platelia AGM assay.

| Study design and study cohort
The primary objective of this study was to assess the sensitivity and specificity of serum samples from patients at risk for IA.Secondary The IA classification followed modified 2020 EORTC/MSG criteria, 11 additionally including solid cancer and COPD gold E as host factors, and 2024 FUNDICU criteria, 12 as originally proposed by authors and indicated in the respective publications supplement (including multi-organ failure as host factor).

| Galactomannan assays
During routine clinical procedures, all prospectively collected samples underwent AGM testing.Additionally, samples from IA patients collected retrospectively were retested using Platelia AGM to ensure testing within a 4-week period for both the clarus AGM prototype and Platelia AGM.For the clarus AGM prototype procedure, as recommended by the manufacturer, each sample was mixed with 100 μL of pre-treatment buffer (4% EDTA solution; 0.2% ProClin) and heated in a dry heat block (Grant QBD2, Grant Instruments, Amsterdam, NL) at 120°C for 7 min.After centrifugation (14,000 × g), 100 μL of the sample were transferred to a microtiter plate and incubated at 37°C for 60 min.The plate was washed (20× Enzyme Immunoassay wash buffer; contains 0.4% Tween20, 0.2% ProClin, 300 μL) using an automated microplate washer (HydroFlex microplate washer, TECAN Grödig, Austria) and incubated with conjugate (HRP-conjugated Aspergillus antibodies, 100 μL).This was followed by additional washing steps and the addition of substrate (Tetramethylbenzidine, 100 μL).
Corrected optical density values were calculated to determine enzyme immunoassay units.

| Statistics
Statistical analyses were carried out using IBM SPSS Statistics Version 29 (SPSS Inc., Chicago, IL, USA).Sensitivity and specificity for proven/probable versus non-IA with possible excluded were calculated.Calculations were based on manufacturer-recommended cut-offs: ≥0.2 optical density index (ODI) for clarus AGM prototype and ≥0.5 ODI for serum for Platelia AGM.An optimal cut-off value for clarus AGM prototype was determined using Youden's index.
Receiver operating characteristic (ROC) curve analyses were performed for Platelia AGM and clarus AGM prototype, with AUC values and 95% confidence interval (CI) calculated for the outcomes of proven/probable IA versus non-IA in the overall study cohort.
Spearman's rho correlation analysis was utilised to compare the two assays.
The ethical committee of the Medical University of Graz (35-204 ex 22/23) approved the study protocol and all study-related procedures.
Among the non-IA samples, one non-IA sample from a patient categorised as possible IA who did not receive mould-active antifungal prophylaxis was included as a control.Seven percent (17/232) of the patients and 10% (30/300) of the serum samples were obtained from individuals receiving mould-active antifungal prophylaxis at time of sampling.Patients baseline characteristics are displayed in Table 1.
Spearman's correlation analysis showed a weak but significant correlation between the two assays, with a rho value of 0.382 (p < .001; Figure 2).

| DISCUSS ION
We conducted a comparative analysis of the clarus AGM prototype assay and the Platelia AGM assay using serum samples from a mixed patient cohort, combining prospective and retrospective data.Our Although the performance of the Platelia AGM was influenced by its use in clinical classification of IA, rendering it outside the scope of this study for evaluation, our primary findings highlight a low sensitivity of the clarus AGM prototype at the manufacturer's recommended cut-off, coupled with nearly perfect specificity.This sensitivity could be enhanced by establishing an optimised cut-off.
However, this improvement was accompanied by a reduction in specificity.Compared to Platelia AGM, the clarus AGM prototype showed only a weak correlation, outlining the differences in antigen targets between the two assays.Studies suggest that different antigens may predominate in different stages of IA, 21 implying that tests targeting distinct antigens may offer varying advantages and disadvantages across disease stages.This could explain the limited correlation observed between Platelia AGM and clarus AGM prototype results, as they may recognise similar but not identical antigens.The variability of antibodies poses a challenge in immunological assays, emphasising the significance of considering the species of origin of the antibody, as it can affect the test's sensitivity and specificity.
Regarding improving sensitivity at the expense of specificity, an improved lower cut-off for the clarus AGM prototype would enhance sensitivity but would reduce specificity.However, it may be worthwhile to explore this approach, as detecting even one additional patient with IA could be clinically significant.Further investigation is warranted to evaluate the balance between sensitivity and specificity and its impact on patient outcomes.potential for increased specificity at an optimised cut-off. 22is could be attributed to various reasons.Our study population was very heterogeneous with patients included with diverse underlying diseases.Since serum GM testing is known to have limited sensitivity compared to BALF GM testing in IA, 28,29 particularly in non-neutropenic patients due to primarily airway invasive growth, 23 this may have driven limited sensitivities observed with the clarus AGM prototype. 24The IMMY Aspergillus GM LFA has been the subject of several studies assessing its utility in both serum and BALF.
In a two-centre mixed cohort study, 25 the IMMY LFA, when used with a cube reader, demonstrated promising sensitivity (78.6%) and specificity (80.5%) in serum.Similarly, in a multicentre study 26 evaluating BALF samples, the IMMY LFA demonstrated a sensitivity of 74% and specificity of 83% at an ODI of 1.5.These findings are comparable with our data for the optimised clarus AGM prototype assay ODI cut-off (sensitivity of 78% and specificity of 79%).
The performance of the clarus AGM prototype appears to be less influenced by mould-active prophylaxis compared to the Platelia AGM assay.This finding suggests a potential advantage in clinical settings where such prophylaxis is common.However, further investigation is warranted to elucidate the implications of this finding.
A limitation of our study is the storage and freezing of serum samples before testing to supplement positive serum samples.However, it is unlikely that this significantly affected test performance, 27 as Platelia AGM was retested in those samples during the study to confirm this assertion.Furthermore, the advantage of Platelia AGM over the clarus AGM prototype may be influenced by selection bias, given the inclusion of Platelia AGM results in disease classification.
objectives included the examination of the impact of mould-active antifungal prophylaxis on serum clarus AGM prototype levels and exploring the correlation between serum GM values obtained using the clarus AGM prototype and Platelia AGM assay.Preselection of samples focused on true positives and true negatives.The inclusion criteria encompassed the submission of serum samples for IA testing within the specified time period and the inclusion of all positive samples with adequate volume, while exclusion criteria involved insufficient volume and improper storage.Prospective samples that were sent for Platelia AGM testing in clinical routine were collected at the Medical University of Graz, Austria, from May 2023 to December 2023 and after routine Platelia AGM testing stored at −80°C until further testing.Additionally, we retrospectively included 25 serum samples stored at −80°C from seven patients diagnosed with probable IA between 2012 and 2015 to augment the number of probable IA cases.All samples underwent clarus AGM prototype testing within 4 weeks of being assessed together with routine Platelia AGM testing.If clarus AGM prototype testing was conducted 4 weeks after the routine GM testing with the Platelia AGM assay, we repeated the Platelia AGM test to ensure consistency.
results showed a weak correlation between the two assays, with clarus AGM prototype displaying inferior sensitivity and specificity compared to Platelia AGM.Upon adjusting the cut-off values, clarus AGM prototype demonstrated enhanced diagnostic accuracy, suggesting the necessity for additional exploration into its clinical performance.Several factors may contribute to the discrepancy in optimisedcut-off values, including variances in assay antigens, patient demographics and disease severity, all of which can influence the diagnostic accuracy of the assay.Additionally, general factors such as sample quality/pre-analytics and conditions could also affect test accuracy.

F I G U R E 1
Receiver operating curve analysis of Platelia AGM (≥0.5 ODI) and clarus AGM prototype (≥0.2 ODI) for differentiation proven/probable versus non-IA in serum per sample.F I G U R E 2 Scatterplot of Platelia AGM and clarus AGM prototype in serum per sample.
Study cohort characteristics at patient level.
Abbreviations: AGM, Aspergillus galactomannan; IA, invasive aspergillosis; PCR, polymerase chain reaction; ODI, optical density index.a One proven and 26 probable IA cases.b One possible IA and 204 non-IA cases.c One missing patient age.d AsperGenius (PathoNostics).
Sensitivity and specificity for the Platelia AGM and the clarus AGM prototype for the diagnosis of proven/ probable versus non-IA (possible IA excluded) in serum.
TA B L E 2