Leukemia-associated fusion transcripts are most conveniently detected by RT-PCR, involving initial conversion of RNA to complementary DNA (cDNA), followed by PCR. For molecular screening, a single round of PCR may suffice; however, to optimize specificity and sensitivity, two rounds of PCR using external and internal primer sets may be used (nested RT-PCR). Some laboratories avoid use of nested RT-PCR methods due to concern about increased risk of PCR contamination. Standardized methods for detection of AML-associated fusion genes by RT-PCR, most notably PML-RARA, AML1-ETO, CBFB-MYH11, and BCR-ABL, have been developed by the European BIOMED1 group.28 The majority of studies using nested RT-PCR for MRD monitoring purposes have focused on patients with fusion genes associated with the favorable-risk cytogenetic abnormalities. Maximal sensitivities for nested RT-PCR lie between 1 in 104 and 1 in 106. Evaluation of primary diagnostic material by RQ-PCR has revealed that, as might be expected, the key determinant of the sensitivity of each respective assay is the level of the fusion gene expression in leukemic blasts in relation to that of endogenous control genes.29 Indeed, RQ-PCR has revealed that PML-RARA is typically expressed at a lower level than AML1-ETO, thereby accounting for the difference in sensitivity observed with the respective qualitative RT-PCR assays (1 in 104 vs 1 in 105).
As discussed above, detection of MRD using conventional nested RT-PCR provides an independent prognostic factor among patients with a defined molecular lesion, in the form of PML-RARA in APL, who could not otherwise have been distinguished on the basis of pretreatment characteristics. Therefore, MRD monitoring is now firmly established as a means of shaping risk-adapted therapy in this subset of AML. A limitation to this approach relates to false-negative results, whereby relapses occur after negative RT-PCR tests. This could reflect relative insensitivity of the assay and insufficient frequency of MRD assessment with respect to the kinetics of relapse, but in some instances could be due to poor quality of RNA or relative inefficiency of the RT step.
For the CBF leukemias, the role of RT-PCR for molecular monitoring purposes has been more uncertain (reviewed in Reference 24). This reflects the detection of AML1-ETO and CBFB-MYH11 transcripts in some patients in long-term remission using conventional nested RT-PCR, which is likely to be related to the greater sensitivity of these assays (as compared to those for detection of PML-RARA). Furthermore, patients with CBFB-MYH11-associated disease are often slow to clear disease-related transcripts, with PCR positivity detected as long as 24 months following achievement of CR, therefore making it difficult to reliably distinguish patients likely to be cured from those destined to relapse. These potential shortcomings of nested RT-PCR have led to considerable interest in quantitative RT-PCR approaches. Indeed, studies using competitor-based assays have shown that patients with t(8;21)-associated leukemia can be distinguished in terms of prognosis on the basis of transcript levels detected in remission, with patients testing negative or with low transcript levels maintaining CR, while those with high or rising transcript numbers progressing to relapse. These competitor-based assays are highly labor-intensive and therefore not readily applicable to evaluation of large numbers of patients in clinical trials. However, over the last few years a variety of semiautomated quantitative PCR approaches have been developed that provide rapid, sensitive, and reproducible methods, combined with a capacity for high throughput. These attributes afford considerable advantages over conventional endpoint nested RT-PCR assays for MRD detection in the context of large-scale clinical trials as well as in individual institutions.
Was this article helpful?