To the Editor:
We have evaluated the occurrence of the AML1/ETO fusion transcripts by reverse transcription-polymerase chain reaction (RT-PCR) analysis1 in 7 patients with t(8; 21)(q22; q22) acute myeloid leukemia (AML) who underwent allogeneic bone marrow transplantation (BMT; n = 6) or peripheral blood progenitor cell transplantation (PBPCT; n = 1). We found here that the AML1/ETO fusion transcripts were detectable in only 2 of 7 patients in the long-term follow-up posttransplant (12 and 112 months post-BMT). One of these two patients who were tested positive for the AML1/ETO fusion transcript at 1 month and at 12 months posttransplant remained also to be mixed chimeric by variable number of tandem repeats PCR (VNTR-PCR) analysis at 12 months post-BMT, but relapsed 5 months later and subsequently died. However, another patient who was AML1/ETO positive up to 112 months posttransplant remained in complete remission during the whole observation period. This patient had a complete chimerism status by VNTR-PCR at 3 months and at 48 months post-BMT.
Three of five patients who were initially AML1/ETO positive in the RT-PCR assay converted 6, 9, and 30 months posttransplant, respectively, and showed at least two consecutive negative PCR assays for the AML1/ETO fusion transcript. All patients who achieved a molecular remission by the AML1/ETO PCR assay remained in stable cytogentic remission.
Studies about the presence of the AML1/ETO fusion transcripts after allogeneic BMT or PBPCT are rare and results are discussed controversely.1-5 Recently, Jurlander et al2 reported that AML1/ETO fusion transcripts in patients treated with allogeneic BMT for t(8; 21) leukemia were still persistent post-BMT in all 9 of 9 evaluable patients.1 According to these investigators, persistence of the leukemic clone after BMT suggests that allogeneic BMT, like conventional chemotherapy and autologous BMT, is usually not sufficient to eliminate expression of AML1/ETO transcripts.2 Contrary to Jurlander et al,2 we showed here, using a similar sensitive method, that allogeneic transplantation leads to sustained suppression or elimination of the leukemic clone in most of the studied patients due to a combination of pretransplantation marrow-ablative conditioning regimen and allogeneic immune reaction (graft-versus-leukemia effect). Our results are supported by a study of Miyamoto et al,3 who reported that the AML1/ETO fusion transcript could not be detected in 4 patients who had been in maintaining remission for more than 30 months after allogeneic BMT. In three other studies,4-6 a total of 5 patients were analyzed by RT-PCR for the AML1/ETO fusion transcripts post-BMT. In 3 of 5 patients the AML1/ETO fusion transcript was consistently detected in the long-term follow-up, whereas 2 patients achieved a molecular remission in these studies.
These data may suggest that patients who achieved a molecular remission after allogeneic transplantation might have a better prognosis with respect to leukemic relapse than their counterparts with a persisting positive AML1/ETO PCR test. However, the detection of the AML1/ETO fusion transcript seems to have not that high of a risk of leukemic relapse after allogeneic BMT as, for example, the detection of BCR/ABL transcripts.
In absence of quantitative PCR assays, chimerism analysis using sensitive VNTR-PCR techniques might help to increase the possible predictive value of a positive AML1/ETO PCR assay in regards to leukemic relapse posttransplant. By having a lower sensitivity usually by a factor of 101 to 102, chimerism analysis by VNTR-PCR may detect a larger number of cells of host-type hematopoiesis. For example, it has thus been shown that, in patients with chronic myeloid leukemia after BMT, a mixed chimerism is associated with a higher incidence of MRD and risk of leukemic relapse.7 8
Response
To our knowledge there are now reports on a total of 23 patients who have been studied for expression of the AML1/ETO transcript after allogeneic BMT.1-1-1-7 Of these patients, 15 have been found to have persistent expression, 7 were or became negative, and 1 was not evaluable. The discrepancy most likely reflects the lack of consensus on when to score a given sample as negative. Our data suggested that the sensitivity of each assay, the amount of starting material, or the source of the material (ie, BM v blood) might all contribute to some of the negative results obtained.1-7 For a patient to be considered negative at a given timepoint in our study, samples had to (1) be amplified in three independent experiments using 2.0 μg of total cellular RNA per reaction, (2) be succesfully amplified for the β-actin in each reaction; (3) be performed simultaneously with an RT-PCR showing a sensitivity for detection of the AML1/ETO transcript of ≥1 × 105 in all three reactions, and (4) assays had to be performed on blood and BM. Taken together, these data suggest that, in the majority of patients, persistent expression of the AML1/ETO is compatible with continued clinical remission and, with the reported follow-up times of up to 10 years, even cure. Recently, similar results, although not after allogeneic BMT, were reported in childhood ALL.1-8
