The challenges and feasibility of validating molecular diagnostic kits on multiple real-time PCR platforms

Abstract number: P867

Purpose: The MycAssay family of molecular diagnostic real-time PCR assays are designed to enhance the speed, sensitivity and specificity of fungal diagnosis and are initially validated on the Cepheid SmartCycler. However, for broad application across the fungal community, they need to run on multiple platforms, thus offering a commercial, standardised assay suitable for all potential users with their existing laboratory equipment. Although the chemistry of the assays are designed to work with the majority of real-time PCR instruments, the reality of conducting platform transfer studies for an in vitro diagnostic assay is not simple.

Methods: Common commercially available real-time PCR platforms were assessed and prioritised for platform transfer. The MycAssay Aspergillus assay was tested on a number of different platforms during assay optimisation. Each platform was assessed for suitability by determining efficiency, as well as the ability to reliably distinguish negative samples. Replicate runs were performed to check that threshold and other analysis settings were robust. Platforms satisfying these criteria were then used for validation studies to satisfy the requirements for in vitro diagnostic devices.

Results: The list of PCR platforms that can support the MycAssay family of products includes; AB7500, RotorGene Q, Stratagene MX3005P, Roche LightCycler 2.0 and Bio-Rad CFX96. Initial formulation testing with the MycAssay Aspergillus product showed acceptable results for both the Aspergillus and internal amplification control reactions on all platforms, allowing use in the research setting. However, the ability to consistently reproduce results, as required for a CE-marked diagnostic, varied between platforms. The source of this variation appears to be in the platform specific interpretative software. In addition, while very similar, Ct (or equivalents) results were not 100% comparable across platforms.

Conclusions: Initial tests with molecular diagnostic assays may indicate suitability on multiple real-time PCR platforms. However, the work required to ensure long-term robustness can be far more complex than expected. While raw data can look normal, analysis algorithms within each manufacturer's software programs can cause the generation of discrepant amplification curves and false positive results. Therefore, before a molecular diagnostic kit can reliably be used on multiple platforms it must undergo the meticulous testing required for CE-marking standards.