BACKGROUND Ponatinib is a potent oral tyrosine kinase inhibitor of unmutated and mutated BCR-ABL, including BCR-ABL with the tyrosine kinase inhibitor-refractory threonine-to-isoleucine mutation at position 315 (T315I). We conducted a phase 2 trial of ponatinib in patients with chronic myeloid leukemia (CML) or Philadelphia chromosome-positive acute lymphoblastic leukemia (Ph-positive ALL). METHODS We enrolled 449 heavily pretreated patients who had CML or Ph-positive ALL with resistance to or unacceptable side effects from dasatinib or nilotinib or who had the BCR-ABL T315I mutation. Ponatinib was administered at an initial dose of 45 mg once daily. The median follow-up was 15 months. RESULTS Among 267 patients with chronic-phase CML, 56% had a major cytogenetic response (51% of patients with resistance to or unacceptable side effects from dasatinib or nilotinib and 70% of patients with the T315I mutation), 46% had a complete cytogenetic response (40% and 66% in the two subgroups, respectively), and 34% had a major molecular response (27% and 56% in the two subgroups, respectively). Responses were observed regardless of the baseline BCR-ABL kinase domain mutation status and were durable; the estimated rate of a sustained major cytogenetic response of at least 12 months was 91%. No single BCR-ABL mutation conferring resistance to ponatinib was detected. Among 83 patients with accelerated-phase CML, 55% had a major hematologic response and 39% had a major cytogenetic response. Among 62 patients with blast-phase CML, 31% had a major hematologic response and 23% had a major cytogenetic response. Among 32 patients with Ph-positive ALL, 41% had a major hematologic response and 47% had a major cytogenetic response. Common adverse events were thrombocytopenia (in 37% of patients), rash (in 34%), dry skin (in 32%), and abdominal pain (in 22%). Serious arterial thrombotic events were observed in 9% of patients; these events were considered to be treatment-related in 3%. A total of 12% of patients discontinued treatment because of an adverse event. CONCLUSIONS Ponatinib had significant antileukemic activity across categories of disease stage and mutation status. (Funded by Ariad Pharmaceuticals and others; PACE ClinicalTrials.gov number, NCT01207440 .).

Allogeneic hematopoietic SCT is an effective treatment in accelerated (AP) or blast phase (BP) CML. Imatinib (IM) has transient but significant activity in advanced phases of CML, which may permit early allografting for responding patients. To identify prognostic factors in allograft recipients previously treated with IM, we analyzed 449 allogeneic hematopoietic SCTs performed from 1999 to 2004 in advanced-phase CML, using the data reported to the Center for International Blood and Marrow Transplant Research. CML patients in second chronic phase (CP2, n =184), AP ( n =185) and BP ( n =80) received HLA-identical sibling (27%), related (3%), or matched or mismatched unrelated donor (70%), peripheral blood (47%) or BM (53%) hematopoietic SCT after myeloablative (78%) or non-myeloablative (22%) conditioning. In all, 52% in CP2, 49% in AP and 46% in BP received IM before hematopoietic SCT. Disease-free survival was 35–40% for CP2, 26–27% for AP and 8–11% for BP. Cumulative incidence of acute and chronic GVHD and TRM were not affected by the stages of CML or pre-hematopoietic SCT IM exposure. Multivariate analyses showed that conventional prognostic indicators remain the strongest determinants of transplant outcomes. In conclusion, there are no new prognostic indicators of the outcomes of allogeneic hematopoietic SCT for advanced-phase CML in the IM era.

Natural killer (NK) cells are expanded in chronic myeloid leukemia (CML) patients on tyrosine kinase inhibitors (TKI) and exert cytotoxicity. The inherited repertoire of killer immunoglobulin-like receptors (KIR) may influence response to TKI. We investigated the impact of KIR-genotype on outcome in 166 chronic phase CML patients on first-line imatinib treatment. We validated our findings in an independent patient group. On multivariate analysis, KIR2DS1 genotype (RR=1.51, P=0.03) and Sokal risk score (low-risk RR=1, intermediate-risk RR=1.53, P=0.04, high-risk RR=1.69, P=0.034) were the only independent predictors for failure to achieve complete cytogenetic response (CCyR). Furthermore, KIR2DS1 was the only factor predicting shorter progression-free (PFS) (RR=3.1, P=0.03) and overall survival (OS) (RR=2.6, P=0.04). The association between KIR2DS1 and CCyR, PFS and OS was validated by KIR genotyping in 174 CML patients on first-line imatinib in the UK multi-center SPIRIT-1 trial; in this cohort, KIR2DS1(+) patients had significantly lower 2-year probabilities of achieving CCyR (76.9 vs 87.9%, P=0.003), PFS (85.3 vs 98.1%, P=0.007) and OS (94.4 vs 100%, P=0.015) than KIR2DS1(−) patients. The impact of KIR2DS1 on CCyR was greatest when the ligand for the corresponding inhibitory receptor, KIR2DL1, was absent (P=0.00006). Our data suggest a novel role for KIR-HLA immunogenetics in CML patients on TKI.

Increasing numbers of second allogeneic hematopoietic stem cell transplantation (allo-SCT) are being performed for patients who have previously undergone and failed an allogeneic or autologous stem cell transplant. We investigated whether the European Group for Blood and Marrow Transplantation (EBMT) risk score, previously established for chronic myeloid leukemia, could be used to predict outcome after second allo-SCT. We analyzed prognostic factors in 124 consecutive patients who underwent second transplants using allogeneic donors at our institution between 10/1985 and 02/2010. Patients with either a first autologous (N = 64) or first allogeneic (N = 60) SCT were included. Age of patient, disease stage, time interval from diagnosis to transplant, donor type, and donor-recipient gender combination were used to establish a score from 0 to 7 points, from which 3 groups were identified. The survival probability at 5 years decreased from 51.7% (95% Confidence interval [CI] 33-70%) for risk scores 0-3 (low, n = 25), to 29.3% (95% CI, 17%-47%) for risk score 4 (medium, n = 40), and only 10.4% (95%CI, 4%-24%) for risk scores 5-7 (high, n = 57). Other factors investigated in univariate analysis included previous transplant type, myeloablative conditioning (Y/N), and interval between first and second SCT (median 20 months). In multivariate analysis, two factors were found to be important. An increased risk of death was associated with: a high risk score (Relative Risk 2.89, p = 0.001) when compared with the low risk group and with a shorter interval between 1st and 2nd SCT (< 20months), RR 1.55, p = 0.05. Given that relapse after transplant can be a lethal situation and that select patients can remain alive and disease-free after second allogeneic transplant, a recent evidence-based guideline published by the American Society of Hematology has given the use of second allogeneic transplant after relapse a recommendation score of 1B. We propose that the five well-defined pre-transplant patient and donor characteristics that make up the EBMT risk score give an excellent risk estimate of outcome for second allogeneic HSCT. Taking into account both the interval between first and second SCT's and the risk score, can provide an accurate estimate of outcome and help in the selection of patient who are most likely to benefit from a second transplant. Increasing numbers of second allogeneic hematopoietic stem cell transplantation (allo-SCT) are being performed for patients who have previously undergone and failed an allogeneic or autologous stem cell transplant. We investigated whether the European Group for Blood and Marrow Transplantation (EBMT) risk score, previously established for chronic myeloid leukemia, could be used to predict outcome after second allo-SCT. We analyzed prognostic factors in 124 consecutive patients who underwent second transplants using allogeneic donors at our institution between 10/1985 and 02/2010. Patients with either a first autologous (N = 64) or first allogeneic (N = 60) SCT were included. Age of patient, disease stage, time interval from diagnosis to transplant, donor type, and donor-recipient gender combination were used to establish a score from 0 to 7 points, from which 3 groups were identified. The survival probability at 5 years decreased from 51.7% (95% Confidence interval [CI] 33-70%) for risk scores 0-3 (low, n = 25), to 29.3% (95% CI, 17%-47%) for risk score 4 (medium, n = 40), and only 10.4% (95%CI, 4%-24%) for risk scores 5-7 (high, n = 57). Other factors investigated in univariate analysis included previous transplant type, myeloablative conditioning (Y/N), and interval between first and second SCT (median 20 months). In multivariate analysis, two factors were found to be important. An increased risk of death was associated with: a high risk score (Relative Risk 2.89, p = 0.001) when compared with the low risk group and with a shorter interval between 1st and 2nd SCT (< 20months), RR 1.55, p = 0.05. Given that relapse after transplant can be a lethal situation and that select patients can remain alive and disease-free after second allogeneic transplant, a recent evidence-based guideline published by the American Society of Hematology has given the use of second allogeneic transplant after relapse a recommendation score of 1B. We propose that the five well-defined pre-transplant patient and donor characteristics that make up the EBMT risk score give an excellent risk estimate of outcome for second allogeneic HSCT. Taking into account both the interval between first and second SCT's and the risk score, can provide an accurate estimate of outcome and help in the selection of patient who are most likely to benefit from a second transplant.

The probability of achieving an optimal response to treatment in CML is most likely to be influenced by the variable capacity of a given patient's leukemia cells to accumulate intracellular imatinib (IM). The transport of IM into the cells is an active process mediated by the human organic cation transporter (hOCT1) protein, a membrane associated ATP-dependent protein channel.1 Recently we reported that higher levels of hOCT1 transcript levels in blood derived white cells (WBC) collected at diagnosis predicted for a 3-log reduction in BCR-ABL1 transcript levels (major molecular response, MMR) and also for greater degrees of transcript reduction in patients who had achieved complete cytogenetic responses (CCyR).2 Investigators in Liverpool also showed that higher hOCT1 transcript levels in diagnostic WBC predicted for a better response to IM.3 The mean hOCT1 mRNA level in pre-IM bone marrow mononuclear cells (MNC) was also reported to be lower in the patients who failed to achieve CCyR by 12 months of therapy.4 However a recent study of total WBC collected from newly diagnosed patients before starting IM failed to confirm any correlation between hOCT1 transcript levels and response to IM.5 These conflicting reports on the predictive value of hOCT1 transcript levels2, 3, 4, 5 might be due to the different target cells studied, to different primers and probes used to measure hOCT1 mRNA or to differing phases of the disease when the hOCT1 was measured. We isolated polymorphonuclear cells (PMNC) and mononuclear cells (MNC) from the peripheral blood of 21 normal individuals and measured hOCT1 transcript levels in both populations by Taqman real time PCR as described before.2 hOCT1 was expressed at a significantly higher level in the PMNC population than in the MNC population (Figure 1a). This suggests the investigation of hOCT1 transcript level as a prognostic factor should be consistent with cell type. We also measured hOCT1 in WBC from paired samples collected at diagnosis and at attainment of complete cytogenetic response (CCyR) in 60 patients treated with IM. The hOCT1 transcript level was significantly higher (P<0.0001) in samples collected in CCyR (range: 0.15–8.88, median: 2.2) compared to the matching diagnostic samples (range: 0.02–3.511, median: 0.14) (Figure 1b). The obvious difference at these two time points is that cells collected from blood at the time of CCyR are derived from normal rather than leukemic haemopoiesis. The relative increase in the hOCT1 transcript level in the CCyR samples could be due to an inhibitory effect of the BCR-ABL1 oncoprotein on the expression of hOCT1 in leukemic cells. As shown in Figure 1c, the level of hOCT1 transcripts in total WBC in CCyR samples (n=60) from patients on IM was not significantly different (P=0.156) from the level of hOCT1 in total WBC from normal individuals (n=21), which supports this explanation. This hypothesis is indirectly supported by the observation of a reduction in hOCT1 transcript levels in CML blast crisis samples5 which in turns is associated with increased level of BCR-ABL1 transcripts.6 In addition to the prognostic value of hOCT1 transcript levels, some investigators have focused on the possible predictive value of hOCT1 single nucleotide polymorphisms (SNPs)7 which have been previously shown to affect hOCT1 function in metformin transport.8 To study the association between the polymorphisms with significant influence on the function of hOCT1 and response to IM we investigated the frequency and association of the 4 SNPs (R61C (rs12208357), P341L (rs2282143), G401S (rs34130495) and rs622342) in 132 CML patients treated with IM in chronic phase. SNP P283L, R287G and C88R were not analysed for this association as the variation of SNP genotypes was too small to be used for any comparison (Table 1). We were unable to identify an association between the transcript level of hOCT1 and any of the individual SNPs except for SNP G401S (rs34130495) which showed a correlation with IM response (Table 1). Patients with 401(G/A) genotype (n=6) had a higher probability of achieving MMR (P=0.015). However, because patients carrying this polymorphism are rare, this result should be interpreted with caution and a larger cohort should be investigated for validation. In summary, the transcription of hOCT1 is a function of the leukocyte lineage in peripheral blood. Its level is significantly higher in PMN (granulocytes) than in MNC (which are largely composed of lymphocytes in normal individuals). Our data suggest that the BCR-ABL1 oncoprotein might reduce the transcript level of hOCT1. Based on these findings we suggest that hOCT1 as a prognostic factor for response to IM should be investigated using granulocyte-derived cDNA, prior to the start of therapy. The predictive value of hOCT1 expression in our cohort was not improved by consideration of the genotype of candidate SNPs within this gene. However a larger study may be required to validate this observation. D Marin, JF Apperley, JM Goldman, D Milojkovic and K Rezvani receive honoraria and/or research funds from Novartis Pharmaceuticals. However, Novartis had no role in the design of the study, collection of the data or the decision to publish.

Correction to: Leukemia (2009) 23, 1054–1061; doi:10.1038/leu.2009.38 Since the publication of this paper, the authors have noticed that the name of an Investigator from the IRIS trial has been omitted from the Appendix. The details of this Investigator are shown below. Italy—Università degli Studi di Firenze, Florence: V Santini.

A significant minority of chronic myeloid leukaemia patients eventually develop resistance to imatinib, often as a result of point mutations within theBCR-ABLkinase domain. Second-line tyrosine kinase inhibitors (TKIs) are effective against mutations that confer imatinib resistance; however, the T315IBCR-ABLmutant has proved resistant to all available TKIs. An assay facilitating early identification ofBCR-ABLT315Iwould therefore aid in identifying high-risk patients who may benefit from alternative therapy. This report describes the development of a sensitive T315I mutation detection methodology based on real-time PCR with self-probing fluorescent primers. The technique demonstrated complete concordance with direct sequencing, correctly identifying 34 T315I-positive samples from a total of 61 samples screened. In a limiting dilution assay, the mutated clone was detectable to a level of 1% of total cells. The data show that Scorpions PCR enables rapid screening forBCR-ABLT315Iin chronic myeloid leukaemia patients and is appropriate for use in a clinical setting.

The cure of chronic myeloid leukemia (CML) patients following allogeneic stem cell transplantation (SCT) is attributed to graft-versus-leukemia (GVL) effects targeting alloantigens and/or leukemia-associated antigens (LAA) on leukemia cells. To assess the potential of LAA-peptide vaccines in eliminating leukemia in CML patients, we measured WT1, PR3, ELA2 and PRAME expression in CD34+ progenitor subpopulations in CML patients and compared them with minor histocompatibility antigens (mHAgs) HA1 and SMCY. All CD34+ subpopulations expressed similar levels of mHAgs irrespective of disease phase, suggesting that in the SCT setting, mHAgs are the best target for GVL. Furthermore, WT1 was consistently over-expressed in advanced phase (AdP) CML in all CD34+ subpopulations, and mature progenitors of chronic phase (CP) CML compared to healthy individuals. PRAME overexpression was limited to more mature AdP-CML progenitors only. Conversely, only CP-CML progenitors had PR3 overexpression, suggesting that PR1-peptide vaccines are only appropriate in CP-CML. Surface expression of WT1 protein in the most primitive hematopoietic stem cells in AdP-CML suggest that they could be targets for WT1 peptide-based vaccines, which in combination with PRAME, could additionally improve targeting differentiated progeny, and benefit patients responding suboptimally to tyrosine kinase inhibitors, or enhance GVL effects in SCT patients.

Donor lymphocyte infusions (DLI) are an effective treatment for patients with chronic myeloid leukemia (CML) in relapse after allografting but the optimal cell dose has yet to be identified. To address this question, we investigated the factors affecting the dose required to achieve remission (effective cell dose, (ECD)) in 81 patients treated with an escalating dose regimen. The overall proportion of patients who achieved a molecular remission was 88%. The cumulative proportion of remitters increased significantly at each dose level. With a CD3(+) cell dose < or =10(7)/kg, 56% of patients in molecular/cytogenetic relapse obtained molecular remission, whereas only 20% of those in hematologic relapse did so. At the same cell dose, 58% of patients who received lymphocytes from volunteer unrelated donors achieved remission, as compared to 29% of those who received DLI from sibling donors. We conclude that the response to DLI is dose-dependent and that the ECD is influenced by the quantity and phase of CML at relapse and degree of donor/recipient histocompatibility.

We investigated the risk factors for graft-versus-host disease (GVHD) in 82 patients treated with donor lymphocyte infusions (DLI) using an escalating dose regimen for chronic myeloid leukaemia in relapse following conventional allografting. Two factors emerged as predictors of both acute and chronic GVHD: the infusion of male recipients with lymphocytes from a female donor and the interval between transplant and last DLI, but only the first remained significant at multivariate analysis. Surprisingly, lymphocyte dose did not influence the incidence of GVHD. Our results suggest that DLI can be given in large cell doses without increasing the risk of GVHD.