Isheeta Shah1*, Natasha Browne-Marke2, Sarah Holt2, Cinzia Pultrone3, Davide Rigamonti3, Alessandro Di Cosimo3,
Roberto Abis3 and Gavin Wall4
1 QIAGEN Sciences Inc., 19300 Germantown Rd, Germantown, MD 20874, United States;
2 QIAGEN Manchester Ltd, Citylabs 2.0. Hathersage Road, Manchester, M13 0BH, UK
3 Sentinel Diagnostics, Via Robert Koch, 2, 20152 Milano MI, Italy;
4 QIAGEN GmbH, QIAGEN Straße 1, 40724 Hilden, Germany
* Presenting author
HAdV is a common virus from the Adenoviridae family and is known for causing self-limiting infections in the conjunctiva, respiratory and gastrointestinal tracts. HAdV can remain latent and in immunocompromised individuals may reactivate and spread to multiple organs via the bloodstream. The importance of appropriate diagnostic HAdV monitoring in blood is underlined by the fact that the morbidity and mortality in immunocompromised patients with invasive infection can be very high, both in the pediatric and adult settings. Quantitative DNA measurements can contribute to the diagnosis of disseminated infection and act as surrogates that correlate with clinical response to therapy. qPCR may also be an effective screening modality to identify asymptomatic patients at risk for progressive adenovirus-associated disease.
Performance of the NeuMoDx® HAdV Quant Assay was characterized in plasma, at a specimen input volumes of 550 μl. Performance was demonstrated across key analytical metrics including analytical sensitivity (LoD), linearity, precision and specificity. Analytical sensitivity was characterized by testing a dilution series: 42 replicates at each dilution level were processed (positive samples) and 8 replicates for negative samples, per day. To determine LLoQ and ULoQ, a TAE ≤1.0 criterion was used for each of the HAdV target levels >95% detection. Analytical specificity was demonstrated by screening 23 organisms commonly found in plasma/serum, as well as species phylogenetically similar to HAdV for cross-reactivity. Precision was determined by testing two replicates of a 5-member panel of HAdV specimens twice daily, using one NeuMoDx 96 System for 20 days.
Analytical sensitivity of the NeuMoDx HAdV Quant Assay was characterized by testing a dilution series of the EDX AdV Verification Panel (Exact Diagnostics) in HAdV-negative plasma/serum samples to determine LoD on NeuMoDx Systems. LoD was defined as the closest target level above the concentration determined by Probit-style analysis with 95% CI. The study was performed over 3 days with multiple lots of NeuMoDx reagents. 42 replicates at each dilution level were processed (positive samples) and 8 replicates for negative samples, per day.
LLoQ and ULoQ
Lower limit of quantitation (LLoQ) and upper limit of quantitation (ULoQ) are defined as the lowest and upper target levels at which >95% etection is achieved AND the total analytical error* (TAE) ≤1.0.
To determine LLoQ and ULoQ, TAE was calculated for each of the AdV target levels >95% detection.
*TAE: Bias + 2*SD (Westgard Statistic)
Analytical specificity was demonstrated by screening 23 organisms commonly found in plasma/serum, as well as species phylogenetically similar to AdV for cross-reactivity. Organisms were prepared in pools of between 5/6 organisms and tested at a high concentration. Two organisms (E. coli and HCV) were analyzed in silico. No cross-reactivity was observed with any of the organisms tested, confirming 100% analytical specificity of the NeuMoDx HAdV Quant Assay.
The NeuMoDx HAdV Quant Assay was evaluated in the presence of typical exogenous and endogenous interfering substances encountered in HAdV clinical plasma/serum specimens. Each substance was added to screened HAdV-negative Base Matrix 53 spiked with 2.5 Log10 copies/ml HAdV and samples were analyzed for interference. None of the exogenous and endogenous substances affected the specificity of the NeuMoDx HAdV Quant Assay.
Precision was determined by testing two replicates of a 5-member panel of AdV specimens prepared with HAdV plasmid twice a day, using one NeuMoDx 96 System for 20 days. The within-run, between-run, within-day and between-day precisions were characterized, and the overall standard deviation determined to be ≤0.30 Log10 copies/ml. Excellent precision was demonstrated across days and runs.
Determined using a 5-member panel of HAdV prepared with HAdV plasmid was used to assess performance on one NeuMoDx 96 Molecular System across three separate runs. Maximum overall bias was 0.39 Log10 copies/ml. Equivalent performance was demonstrated across lots as quantitation of all panel members was within tolerance specification.
Determined using three different systems (two NeuMoDx 288 Molecular System and one NeuMoDx 96 Molecular System). A 5-member panel of HAdV prepared with HAdV plasmid was used to assess performance. Testing was performed in parallel on the systems for 5 days.
The variation within-day and between systems was
characterized, and the overall standard deviation was ≤0.30 Log10 copies/ml. Equivalent performance was demonstrated across systems as SD in quantitation of all panel members was within tolerance specification.