The t(8;21)(q22;q22) translocation is observed in about 40% of pediatric cases of AML M2. Although patients with t(8;21) AML have a relatively good prognosis, relapses still occur in about 30% cases. To predict relapse, several studies examining sequential quantitation of minimal residual disease (MRD) by real-time PCR after chemotherapy or BMT have been attempted. However, these studies have shown that only a very small number of AML1/ETO transcripts are detectable in most t(8;21) AML patients who maintain complete long-term remission (CR). A recent study clearly demonstrated the existence of nonmalignant stem cells with leukemia specific genes. In another study, the suppression of AML1/ETO by siRNA in vitro resulted in the increased expression of genes associated with myeloid differentiation as well as antiproliferative genes in both leukemic blasts and cell lines. This suggests that cells expressing lower level of AML1/ETO might be well-differentiated malignant cells with slower proliferation rates compared to those expressing higher levels of AML1/ETO. Taken together, we hypothesized that the translocated gene detected by PCR, even in long-term remission patients, might be derived from nonleukemic stem cells or well-differentiated leukemic cells. If we can estimate AML1/ETO transcripts in single cells possessing leukemic translocation genes in BM in CR, we can predict, with high reliability, whether early relapse would occur or not. To test this hypothesis, we simultaneously assessed MRD by FISH with locus-specific probes with higher sensitivity and by quantitative real-time PCR in BM samples obtained from seven pediatric patients with AML1/ETO during and after chemotherapy/BMT. We then evaluated the value which was calculated by dividing normalized AML1/ETO copies/μgRNA by the percentage of t(8;21)-positive cells by FISH (RQ-PCR/FISH), which ideally reflected actual number of AML1/ETO transcripts in single cell with leukemic translocation in BM cells. RQ-PCR/FISH of non-relapse patients ranged from 5.5 × 102 to 5.5 × 103 (mean 2.5 × 103 ) in BM in CR. In marked contrast, 8.0 × 103 to 1.8 × 105 (mean 5.9 × 104 ) was observed from the patients who finally relapsed. Indeed, we demonstrated that in some cases showing low value of RQ-PCR/FISH, the AML1/ETO was detected only in terminally differentiated basophils (CD34− CD203c+ ), not in stem cell fractions (CD34+ CD203c± ). In conclusion, we propose that RQ-PCR/FISH can be a powerful predictor of relapse, especially in patients with lasting MRD positive after chemotherapy/BMT.
Disclosure: No relevant conflicts of interest to declare.