Defining the “Absence” of the CBFβ/MYH11 Fusion Transcript in Patients With Acute Myeloid Leukemia and Inversion of Chromosome 16 to Predict Long-Term Complete Remission: A Call for Definitions

To the Editor:

Inversion of chromosome 16, commonly associated with acute myeloid leukemia (AML) French-American-British (FAB) subtype M4Eo, identifies a group of patients with a high complete remission (CR) rate and long-term disease-free survival.1 At the molecular level, the inversion creates two novel fusion genes, CBFβ/MYH11 and MYH11/CBFβ, but only the first appears to be critical for leukemic transformation.2 Amplification of the CBFβ/MYH11 fusion transcript by the reverse transcription-polymerase chain reaction (RT-PCR) has been used to monitor patients in CR for evidence of minimal residual disease (MRD),3,4 but the predictive value of this assay remains uncertain. To date, detection of the fusion transcript has been reported by RT-PCR in bone marrow (BM) or blood (PB) of 51 AML patients evaluated during CR in nine different studies.3-11 Thirty-one of 51 were positive at the last follow-up for the CBFβ/MYH11 transcript in BM or PB, including 1 patient after allogeneic bone marrow transplantation (BMT), with only 12 of the 31 (40%) subsequently relapsing. The fusion transcript was undetectable by RT-PCR at the last follow-up in the remaining 20 patients, including 3 patients after allogeneic BMT, with 3 of the 20 (15%) subsequently relapsing.8,11 Results of RT-PCR for the CBFβ/MYH11 fusion transcript in patients with ≥3 years of follow-up have been reported in only 2 cases (both in CR), at 60 and 108 months, with undetectable and detectable transcript by nested RT-PCR, respectively.4,6 In our laboratory, we analyzed follow-up BM samples from 3 patients with AML and inv(16). Two patients were in first CR 36 months after chemotherapy alone, whereas 1 patient was in CR for 12 months after allogeneic BMT performed during first relapse. Diagnostic material from all 3 patients demonstrated inv(16)(p13q22) by conventional cytogenetics, and the CBFβ/MYH11 fusion transcript was detected in diagnostic samples by RT-PCR. The postremission samples were analyzed by nested RT-PCR. Each remission sample was negative for the fusion transcript by nested RT-PCR with a sensitivity of 1:10⁵ to 1:10⁶, determined by performing a limiting dilution of the CBFβ/MYH11-positive ME1 cell line into blood mononuclear cells from a normal donor. Rigorous criteria previously published by Jurlander et al12 and used in the current analysis were applied to define RT-PCR conditions under which an undetectable fusion transcript was considered acceptable: (1) no amplification of the CBFβ/MYH11 fusion transcript in three independent experiments using 2.0 μg of total RNA per reaction; (2) simultaneous demonstration of a sensitivity of at least 1:10⁵ in all three reactions; and (3) successful amplification of the β-actin transcript in each reaction. Both patients treated with chemotherapy alone continue to be in CR at 55 and 41 months from diagnosis, whereas the transplanted patient died, in apparent CR, of complications resulting from graft-versus-host disease 15 months after BMT.

With our 2 patients described here, a total of 4 patients with AML, inv(16), and CR with chemotherapy alone have now been assessed for the CBFβ/MYH11 fusion transcript by RT-PCR and reported with ≥3 years of CR duration. Three of the 4 patients have an undetectable CBFβ/MYH11 fusion transcript. We also describe 1 more case in addition to the 4 patients previously reported with inv(16)-associated AML evaluated for MRD after allogeneic BMT, 4 of whom had an undetectable CBFβ/MYH11 fusion transcript. Additional patients with long-term follow-up are needed to determine the predictive value of the CBFβ/MYH11 transcript detected by RT-PCR in patients with AML, inv(16), and CR. Furthermore, because 60% of the patients reported with a detectable CBFβ/MYH11 fusion transcript during CR continue to be in remission at the last follow-up, a prospective multiple time-point analysis is necessary to determine whether late (ie, ≥8 months)7 conversion to RT-PCR negativity occurs and how this ultimately impacts patient prognosis. In a recent study by Evans et al,13 3 of 8 patients with CBFβ/MYH11-positive AML analyzed at different time-points postremission using a competitive quantitative RT-PCR assay became RT-PCR negative at 24, 51, and 80 weeks.13 In the same study, the investigators reported that the amount of fusion transcript detected during morphological remission was approximately 3 to 4 logs lower than that at presentation, but they were unable to define a cut-off level of transcripts below which cure is likely and above which relapse occurs. Newer methodologies are currently being developed to quantitate such fusion transcripts with single round RT-PCR in real time and will hopefully assist in defining such quantities for prognostic value in this disease.14 

Similarly unresolved is the prognostic significance of an undetectable CBFβ/MYH11 transcript in patients with AML, inv(16), and CR. Achievement of RT-PCR negativity has not been proven yet to be predictive of cure, considering that 15% of the patients with an undetectable CBFβ/MYH11 fusion transcript during CR will subsequently relapse. Some of the difficulty in interpreting these negative data for predictive clinical value may result from the lack of standardized criteria for defining the absence of a fusion transcript by nested RT-PCR in leukemia. To more efficiently elucidate the potential clinical significance of a fusion transcript deemed undetectable by nested RT-PCR, we propose that, as a minimum definition, the criteria of Jurlander et al12 be adopted.