Late Relapse in Leukemia: Does the Niche Matter?

Abstract 4793

Very late recurrence (LR) of AML, defined as relapse after ≥5 years of complete remission (CR), has an incidence of 1.1–3% (Verma et al, 2010; Medeiros et al, 2010) and may be due to the re-growth of the original dormant leukemic clone which may have acquired additional genetic changes or to the emergence of a new leukemic clone favoured by previous drug exposure. In CML LRs after allogeneic haematopoietic stem cell transplantation (allo-HSCT) are also very rare (1.5% per year) and it is unknown whether rare BCR-ABL transcripts are true manifestations of the original leukemic clone or the results of a “benign” BCR-ABL rearrangement seen in healthy subjects (Sobrinho-Simòes et al, 2010). In order to highlight LR biology, most studies have focused on alterations of cellular dormancy, whereas very few on the disruption of bone marrow (BM) microenvironment.

The present study, which reports three patients (two AML and one CML) who experienced LR 5–16 years after CR achievement, was aimed at establishing the incidence of LR, the origin of the leukemic clone, and a potential association between LR and infective episodes. In our AML series the incidence of LR was 0.8% (2/245). The first patient, a 23 year-old male who carried a t(1;3)(p36;q26), was diagnosed as AML M5 and achieved a I CR after four courses of Etoposide (VP16) followed by one course of Daunorubicine (DNR) and a 2-years maintenance with VP16. He remained in CR for eight years when he developed a BM relapse after a viral infection. The leukemic clone presented the same chromosomal defect revealed at the onset of the disease. A II CR was achieved after a re-induction with VP16+Mitoxantrone (M). Two courses of this same regimen were used as consolidation and were followed by an auto-HSCT preceded by a conditioning with BCNU-VP16-Ara-C. Fifteen months later a second relapse occurred. Leukemic cells presented the same original t(1;3)(p36;q26). The patient did not respond to various chemotherapy schedules and died of sepsis. The second patient, a 46 year-old male who carried a t(3;11)(q26;q23), was diagnosed as AML M5 and achieved a I CR after three courses of DNR+Ara-C. Subsequently, he received a HSCT from his daughter after a conditioning with Busulfan+Cyclophosphamide (BU+CY). Take occurred at day +14 without any sign of acute GvHD, but on day +183 the patient developed a chronic extensive GvHD. Two months post-transplant FISH with sex chromosome probes (XY-FISH) revealed 0.5% of male cells, but from day +100 onwards it revealed a complete chimera. Five years later the patient complained of fever and a chest X-ray showed a left pulmonary infection. A BM biopsy revealed the almost exclusive presence of leukemic cells which karyotype was 50,XY,t(3;11)(q26;q23),+4,+8,t(12;?)(p21;?),-14,+mar1,+mar2, +mar3[13]. XY-FISH showed 60% of male cells. The patient died of sepsis. The CML patient was the only one who relapsed 16 years after a sibling allo-HSCT in I chronic phase. HSCT was performed as the patient was intolerant to alpha-interferon which had been replaced by hydroxyurea. Pre-transplant conditioning consisted of BU+CY, take occurred at day +17 without any sign of acute GvHD, but on day +329 the patient developed a chronic limited GvHD. On day +100 XY-FISH revealed a complete chimera, a chromosomal analysis a male karyotype and a BM nested RT PCR the absence of BCR-ABL transcripts. Subsequent analyses including quantitative RT PCR confirmed these findings. However, 16 years post-transplant the patient complained of otitis and fever. Her blood count showed: Hb 11.6g/dl, WBC 8.1×10⁹/l with 22% myelolasts, Plts 5×10⁹/L. A BM aspiration revealed the almost exclusive presence of CD34+, TdT+, CD10+ lymphoblasts having the following karyotype: 46,XY[4]/45,X,t(9;14;22)(q34;p10;q11)[6]. A XY-FISH showed 27% of female cells and a quantitative RT PCR a BCR-ABL/ABL ratio =214,2 (International scale). Tyrosine kinase inhibitors were immediately started.

In our series LR was i) a very rare event, ii) always due to the re-growth of the original leukemic clone, iii) always preceded by an infective episode which might have interrupted leukemic cells dormancy through a transient angiogenic burst within the BM microenvironment (Indraccolo et al, 2006). This suggestion is further strengthened by recent data which support a strict interaction between cancer cell metabolism and niche metabolism (Yusuf et al, 2012).