a.) Objectives and Postulates The primary objective of the second phase of the standardization project was to test the hypothesis that use of common primers and fluorescent probe in the qRT-PCR assay with a common lot of master mix and a Standard Operating Protocol (SOP) will reduce the variability in BCR-ABL transcript copy number per g RNA at MMR between labs compared to the situation of each lab using its own home brew method for clinical qRT-PCR assays. Further, it will be determined whether predicted maximum magnitude of difference between labs can be reduced to the level of less than 1 log10 for samples in the MMR range. A secondary objective of this project is to determine the source(s) of variability within the assay and their contribution to the overall variance in the assay. Although the samples will be limited to mixtures of cell lines that contain either of the two major variants of BCR-ABL in CML (b3a2 and b2a2) in a negative cell line, the three phases of the assay will be assessed using the cells, their total RNA and cDNA. This will assess the errors generated during the three phases of the assay: RNA isolation, reverse transcription, and qPCR itself. Due to budgetary constraints, the assessment of all three phases of the assay (i.e., assay of RNA and cDNA in addition to cells) will occur only for the b3a2 variant. b.) Development of Calibrators The Jessup lab prepared Standard Reference Materials (SRM) of ssRNA in control (HL-60 total RNA) to act as calibrators. The two major variants of BCR-ABL are b3a2 and b2a2 in which the exons of BCR are fused to Abl as part of the Ph1+ translocation t (9, 22) (q34, q11) e13a2 = b2a2;e14a2 = b3a2. The sources of SRM were the following cell lines: 1) for the b3a2 fusion gene the K562 cell line (CB Lozzio and BB Lozzio Human chronic myelogenous leukemia cell-line with positive Philadelphia chromosome. Blood 45: 321-334, 1975;Gabert J, Beillard E, van der Velden VH, et al. Standardization and quality control studies of 'real-time'quantitative reverse transcriptase polymerase chain reaction of fusion gene transcripts for residual disease detection in leukemia - a Europe Against Cancer program. Leukemia 17:2318-57, 2003) and 2) for the b2a2 fusion gene the KCL-22 cell line (I Kubonishi, Y Ohtsuki, S Yoshimoto and I Miyoshi. Heterotransplantation and maturation of a chronic myelogenous leukemia cell line (KCL-22) in vivo. Int. J. of Cell Cloning, 2, 243-253, 1984;Saussele S, Weisser A, Mller MC, Emgi M, La Rose P, Paschka P, Kuhn C, Willer A, Hehlmann R, Hochhaus A. Frequent polymorphism in BCR exon b2 identified in BCR-ABL positive and negative individuals using fluorescent hybridization probes. Leukemia. 14:2006-10, 2000). The genes were cloned and sequenced. After confirmation of identity T7 promoters were added to the primers and ssRNA created. Analysis of products of RT-PCR indicated that the the two calibrators were the correct size and that the fusion genes were absent from HL-60 cells that lack either BCR-ABL fusion gene. Test samples of the calibrators were sent to each of the seven participating consortium labs who confirmed their ability to detect both BCR-ABL variants in the test calibrators. The Jessup lab then proceeded with creating log dilutions from 4 40,000 variant copies in a standard amount of HL-60 RNA to be used as the calibrator samples for creation of standard curves for the assays. c.) Study design Seven laboratories will assess the BCR-ABL transcript level of samples from cell mixtures, samples of RNA and cDNA mixtures. The cell mixtures will contain mixtures, in various proportions, of the cell lines HL-60 (negative for BCR-ABL fusion transcripts) and K562 (positive for b3a2 fusion variant) or KCL-22 (positive for the b2a2 fusion variant). The RNA and cDNA mixtures also contain mixtures of either the b3a2 or b2a2 variant added to total RNA or cDNA from HL-60 in proportions that mimic the concentrations in the cell mixtures. A manufacturer of a commercial kit has provided the set of primers and probes for their qRT-PCR BCR-ABL clinical assay under an approved UBMTA and will remain anonymous. However, they are the source of the standard primers to be tested in the assay as a means of reducing interlaboratory variance. One other aspect of this provider is the use of a Standard Operating Protocol (SOP) and the use of a standard set of reagents with which to perform the assay under the SOP. All of these were provided to the consortium members. Additionally, each laboratory will use its own CLIA-based laboratory assay as well as the assay involving the standard primers so that a comparison can be made between each laboratorys Home Brew assay and the Standard Primers assay. In order to define the sources of error, it will be important to assess the quality of RNA isolated from the cell mixtures. This quality assessment will include both 260/280 ratio and the RIN number on the extracted or prepared total RNA before it is used in the 1-step and/or the labs qRT-PCR assay. In addition, SRM ssRNA calibrators will be prepared to establish standard curves for both variants. These calibrators will consist of a standard amount of essentially normal RNA that does not contain BCR-ABL transcripts spiked with varying amounts of either b3a2 or b2a2 ssRNA. This will allow for assay results produced under either the home brew protocols or the common protocol to be calibrated to standard units (# fusion transcript copies per mug RNA) for comparison across labs. The standard curve will be created for each BCR-ABL fusion gene variant and then high and low concentrations of each calibrator will be included in each assay run to assure that each run is within acceptable limits for the calibration curve. Finally, when the data are returned, the data will be subjected to audits for accuracy and then Dr. Lisa McShane of the Biometric Research Branch, NCI will perform appropriate statistical analysis to assess whether interlaboratory comparability has been improved and where the major sources of error exist. This will inform whether there is a need to continue this standardization project further. b.) Progress and Future Directions: The data have been returned and demonstrate that use of common primers and reagents harmonize the assay so that the transcript numbers were within a log of each other across all the labs. Use of the calibrator with the home brew assay only marginally improved assay performance. Unfortunately, we have been stymed for half a year in getting these data submitted for publication because the statistician has not gotten to the project.
