Biological and clinical features of T‐biphenotypic acute leukaemia: report from a single centre

We read with interest the work from Rubio et al (2003) on adult T-biphenotypic leukaemia, reporting the clinical and biological characteristics of this rare clinical entity. We have reviewed our file of acute leukaemias, both lymphoblastic and myeloid, to identify the biphenotypic acute leukaemias (BAL) and, among them, the cases co-expressing T-lymphoid and myeloid markers. In the last 10 years (1993–2003), we have diagnosed 280 acute leukaemias, 202 myeloid (AML) and 78 lymphoblastic (ALL). Using the diagnostic criteria recently proposed (Bene et al, 1995), 26 (9%) patients had a BAL. In accordance with other reports (Legrand et al, 1998; Killick et al, 1999), the most common BAL phenotype consisted of the co-expression of myeloid and B-lymphoid markers, but three of these patients (11Æ5%) had a T-BAL. Clinical and biological characteristics of the patients are listed in Table I. Two of the patients were female and one was male. The median age at diagnosis was 58 years (range 26–60 years). According to the French–American–British (FAB) classification, all the patients had a type 2 ALL, with co-expression of myeloid markers, with a score of at least two for both T-lymphoid and myeloid markers. Cytogenetic analysis was available for two patients. Both showed a deletion of chromosomes 7 and 12, plus other abnormalities; no Philadelphia chromosome (Ph+) metaphases were documented. When karyotype was missing, BCR-ABL transcripts were not detected by molecular biology. We analysed the multidrugresistance (MDR) proteins and found an overexpression of P-glycoprotein with a mean of 7Æ2 (range 6Æ6–8Æ1), while lung resistance-related protein and multidrug resistance-related protein 1 were normally expressed. Two patients (patients 2 and 3) received an ALL-designed induction therapy (vincristine, idarubicin, prednisone and asparaginase). Patient 2 died during induction due to a haemolytic-uraemic syndrome and pneumonia. Patient 3 achieved a complete remission (CR) that was consolidated with two courses of therapy (high-dose cytarabine, then vincristine, methotrexate, cyclophosphamide and adriamicin). She relapsed after 5 months but attained a second CR with salvage therapy (liposomal daunorubicin and high-dose cytarabine). The remission lasted for another 4 months, then the patient relapsed again and died of cerebral haemorrhage. Patient 1 displayed a more undifferentiated morphology, with two blast populations. He initially received an AML-like course with idarubicin and low-dose cytosine arabinoside. Day +14 bone marrow was still completely blastic, but with more distinctive lymphoid features. He therefore was switched to our ALL protocol, attaining a CR and subsequently underwent an autologous bone marrow transplant. He is alive in CR, 9 years after transplant. Our small experience partially confirms the findings of Rubio et al (2003). In addition, our patients were morphologically classified as L2 and presented superficial adenopathies at diagnosis, but none of them had mediastinal involvement. Leucocytosis and peripheral blast count were generally low or moderate. None of our patients had the Philadelphia chromosome or BCR-ABL rearrangement, but two cases displayed unfavourable cytogenetics, with deletions of chromosome 7. In a previous report, four of six patients with T-BAL had a complex karyotype and none was Ph+ (Carbonell et al, 1996). Treatment of BAL remains controversial and these patients have generally a bad prognosis. Except in patient 2, who died during induction, we observed a good response when a mixed or ‘sequential’ AML/ALL induction therapy was used, possibly followed by intensification with transplantation. When an ALL-type course was used alone (patient 3), a CR was obtained but it was short-lived, and also a salvage therapy with high-dose cytarabine and liposomal daunorubicin was only