We read with interest the letter of Emilia et al1concerning the absence of BCR-ABL rearrangement in essential thrombocythemia (ET). They studied 112 patients with ET with nested reverse transcriptase–polymerase chain reaction (RT-PCR) and found only one BCR-ABL+ patient who progressed to myelofibrosis and blast crisis 12 years after diagnosis. Based on their findings, the authors conclude that the presence of BCR-ABL rearrangement in ET identifies a subset of patients at risk for progression to acute leukemia and suggest that these patients probably represent chronic myelogenous leukemia (CML) of thrombocythemic onset. We do not share their opinion on the basis of our own experience.
We used single and nested RT-PCR and interphase fluorescence in situ hybridization (FISH) to prospectively evaluate 47 patients with ET who fulfilled the criteria of the Polycythemia Vera Study Group (32 women; median age 49 years, range 14-80 years). Ten patients were untreated, 31 were treated with anagrelide, and 6 were treated with hydroxyurea. Median platelet count was 1300 × 109/L (range 608-3742 × 109/L), 20 patients had mild leukocytosis, none had basophilia, 3 had a borderline low leukocyte alkaline phosphatase (LAP) score. Twenty patients had mild splenomegaly and 15 had focal reticulin fibrosis. None of the 40 cytogenetic studies performed revealed the Ph chromosome.
One-step and nested RT-PCR for BCR-ABL transcripts (b3a2 and b2a2 subtypes) were performed on peripheral blood samples using methodology described by Kawasaki et al2 and Cross et al,3 respectively, with slight modifications, with a sensitivity of 10−4 and 10−6 determined by K562 cell dilution. Assays were carried out with appropriate controls and 20 healthy individuals were BCR-ABL−. FISH analysis was performed on peripheral blood smears as described.4 In 5000 bone marrow cells from 25 controls the BCR-ABL extra signal (ES) probe (Vysis, Downers Grove, IL) demonstrated 1 yellow (fusion) signal, 1 green signal, and 2 red signals (BCR-ABL fusion positive pattern) in 0.14% (0.3% [3 SD = 0.9%]), with a sensitivity of 98.5%. For each patient, 200 nuclei were scored.
Three Ph chromosome–negative patients were found to harbor b3a2 BCR-ABL fusion transcripts by RT-PCR (6.4 %). Both one-step and nested PCR results were positive in 1 patient; only nested PCR yielded positive results in the other 2 patients. BCR-ABL transcripts were detected in the first patient at diagnosis and several times during follow-up while on treatment with anagrelide. The other 2 BCR-ABL+ patients were first evaluated during treatment with anagrelide. Repeated samples remained positive in the second patient while negative PCR results were obtained on sequential analysis in the third patient. Interphase FISH results on these patients showed lack of fusion signal. Follow-up was 73, 68, and 116 months, respectively. Clinical and laboratory features in patients with positive results did not differ from the BCR-ABL−group and they followed an uneventful course. Median follow-up in the BCR-ABL− group was 60 months (range 8-169 months); one of these patients developed myelofibrosis with myeloid metaplasia 7 years after diagnosis.
Studies on BCR-ABL expression in ET have yielded discordant results. While Blickstein et al5 and Singer et al6reported high frequency (48% and 63%, respectively) of BCR-ABL signals, Hackwell et al7 and Emilia et al1were unable to corroborate these findings. Our results confirm the presence of BCR-ABL positivity by PCR in some patients with ET, albeit at a lower frequency than that reported by Blickstein et al. We analyzed peripheral blood samples, but higher frequencies have been reported in bone marrow. Although the presence of the Ph chromosome in ET carries a poor prognosis,8 the clinical significance of BCR-ABL positivity in Ph-negative ET at various levels of sensitivity remains unclear. The finding of this molecular abnormality by PCR in our patients doesn't seem to imply a diagnosis of CML, as suggested by the absence of clinical features of CML, their long-term uneventful follow-up, and the spontaneous disappearance of BCR-ABL transcripts in one of them. Interphase FISH in peripheral blood samples was negative in our 3 patients with positive PCR tests, thus excluding the existence of a masked t(9;22) in the background of a hidden CML. The presence of the BCR-ABL rearrangement detected by means of PCR in these patients doesn't seem to exclude the diagnosis of ET nor to influence the course of the disease. In our opinion, therapeutic decisions should not be based on PCR results.
