DEVELOPMENT OF ASSAY CALIBRATION MEHTHODOLOGY FOR CEPHEID XPERT®
BCR-ABL ULTRA P190 PROTOTYPE‡ ASSAY
Yuanyuan Liu, Tran Tran, Phat Nguyen, Vineela Kadiyala, Jessica Dosanjh, Adam Aslam, Huilin Wei and
Lin Yuan
Oncology R&D, Cepheid, Sunnyvale
Background: Currently, no International Standards are available for the quantitation of the
e1a2 p190 transcript for both CML and ALL, posing a challenge when developing a quantitative
assay to monitor the e1a2 p190 transcript levels.
Objectives: Implementation of a primary control material, containing in vitro transcribed RNA
(IVT-RNA) of ABL and BCR-ABL e1a2, for standardization of ABL and BCR-ABL e1a2
quantifications through copy number (CN) calibration was used to develop an assay
calibration methodology at Cepheid. This process allows generation of in-house RNA control
materials as secondary standard (SS) mimicking clinical samples, for calibration and
standardization of the Xpert BCR-ABL Ultra p190 Prototype Assay.
Method: Titration studies were performed with Cepheid’s IVT-RNA, containing e1a2 and ABL,
to derive CN calibration curves of the target genes relative to the corresponding Cycle
Threshold (Ct) for both e1a2 and ABL in the background of PAXgene human whole blood. The
calibrated standard curves assigned the BCR-ABL e1a2/ABL% values to a 4-member in-house
control material (SS) ranging from ~10% to 0.02% generated by spiking BCR-ABL e1a2 total
RNA and ABL total RNA (HL60) in PAXgene blood. A lot-specific curve of delta Ct vs BCR-ABL
e1a2/ABL% value was generated from the secondary standard for the BCR-ABL p190
Prototype Assay providing an Efficiency Value (E!Ct) and a lot-specific scaling factor (SF). The
BCR-ABL e1a2/ABL% value of any given clinical sample was calculated with E!Ct and SF. A 4-
assay comparison study was performed to validate the assay calibration methodology with
contrived clinical samples.
Results: Based on the observation of constant delta Ct between e1a2 and ABL from IVT-RNA,
the ABL standard curve was generated by BCR-ABL e1a2 standard curve. Through the BCR-ABL
e1a2 and ABL calibration curves generated from IVT-RNA PAXgene blood background, the
BCR-ABL e1a2/ABL % value of the SS was assigned. The E!Ct and SF of 4 lots of BCR-ABL Ultra
p190 kit were generated. BCR-ABL e1a2/ABL % values of contrived clinical samples in EDTA
whole blood and PAXgene whole blood from BCR-ABL Ultra p190 Prototype Assay were
consistent with those from three other comparator assays, RT-ddPCR, Asuragen, and Ipsogen
(Table 1).
Conclusion:
In conclusion, implementing a primary control material of IVT-RNA, allowed linkage of known
RNA CN to corresponding Ct, thus providing a methodology to assign BCR-ABL e1a2/ABL %
value to a human whole blood-based secondary standard. A lot-specific E!Ct and SF for BCR-ABL
Ultra p190 Prototype Assay was generated from the secondary standard and was utilized
to calculate the BCR-ABL e1a2/ABL % value for a given clinical sample. Data from 4-assay
comparison suggested that we achieved the goal of establishing a successful assay calibration
methodology with IVT-RNA in human blood background as primary standard and the BCR-ABL
e1a2/ABL % value reported by BCR-ABL Ultra p190 Prototype Assay was compatible with other
comparator assays.
POSTER 12