Reverse Transcription with a Random Pentadecamer Primer Increases the Sensitivity of Quantitative PCR for BCR-ABL.

Real-time quantitative reverse transcriptase PCR for BCR-ABL (RQ-PCR) is used to monitor treatment response in chronic myeloid leukaemia (CML). BCR-ABL levels continue to decline over several years of imatinib treatment and increasing numbers of patients have BCR-ABL levels at or below the limit of detection. The sensitivity of current RQ-PCR assays limits our ability to identify patients who have a continuing decline in BCR-ABL levels; or to assess the effects of novel therapies to improve outcome in patients who have a good response to ABL kinase inhibitors, but harbour minimal residual disease. Improvements in therapy for these patients will be hard to assess without more sensitive monitoring of BCR-ABL. BCR-ABL levels are usually reported relative to a control gene. The control gene value gives an indication of the sensitivity achieved within each RNA sample; this varies with sample quality and efficiency of reverse transcription (RT). Control gene copy number below a defined value indicates that the analysis may be unreliable. We investigated the use of a random pentadecamer (RP; 15mer) primer in the RT reaction to improve the sensitivity of RQ-PCR. The RP primer is reported to be more efficient than the random hexamer (RH; 6mer) which is commonly used for RQ-PCR. BCR-ABL and BCR control gene transcripts were measured in 30 peripheral blood samples. After Trizol extraction 2μg RNA and 400U Superscript II reverse transcriptase were added to two RT reactions using RP or RH at a final concentration of 25μM. Each RT and quantitative PCR was performed 2–3 times and results compared using the Wilcoxon signed rank test or paired t-test. A more efficient RT is indicated by higher transcript levels. The table shows that BCR-ABL and BCR control gene values were significantly higher with RP primers. There was a proportionate increase in all transcripts so that the change in BCR-ABL/BCR% was not significant.

We tested samples with undetectable BCR-ABL including 10 from CML patients on imatinib, and 10 from BCR-ABL negative control subjects. Each RNA was tested in two independent RTs and Q-PCRs. If the results were discordant a third RQ-PCR was performed and a consensus result determined. Using RP primers 6/10 patient samples were positive; 2/10 were positive with RH. None of the BCR-ABL negative control samples was positive with either primer. Sensitivity was calculated using the lower limit of detection and the standardised baseline value for untreated CML patients.The median calculated sensitivity increased from 4.5-log with RH to 5.0-log with RP. The use of pentadecamer primer made possible the detection of BCR-ABL in a small number of patients who had undetectable BCR-ABL with random hexamer RQ-PCR. This initial analysis suggests that the sensitivity of BCR-ABL detection may be improved 2–3-fold with a simple modification to RT methodology. This also has the potential to reduce the number of samples deemed unsatisfactory due to low control gene levels.