ALBERT

Description: Chronic myelomonocytic leukemia (CMML) is a genetically heterogeneous hematopoietic stem cell disorder that combines features of a myelodysplastic syndrome and a myeloproliferative neoplasm and exhibits a strong bias towards older age. The prognosis of CMML is poor, with overall survival of less than 3 years in most studies, however recurrent somatic mutations explain only 15-24% of the clinical heterogeneity of CMML (Elena C. et al. Blood 128:1408-17, 2016). The extreme skewing of the CMML age distribution suggests that CMML reflects the malignant conversion of the myelomonocytic-biased differentiation characteristic of an aged hematopoietic system. We hypothesized that separating the contribution of the normal aging process from bona fide CMML-specific alterations will improve the molecular characterization and biological understanding of CMML. We decided to focus on monocytes as the phenotypic minimal common denominator of genetically heterogeneous diseases. CD14+ monocytes were sorted from the blood of untreated CMML patients (N=12, median age 77 years, range 61-90), age-matched healthy controls (old controls: N=12, median age 68 years, range 62-74) and young healthy controls (young controls: N=16, median age 29 years, range 24-44) and subjected to RNA sequencing and DNA methylation profiling. Differentially expressed genes in CMML monocytes compared to healthy controls were identified with DESeq2 using a 1% false discovery rate (FDR) and a fold-change cutoff set at 〉│2│ (Figure 1A). We identified the 2480 CMML-specific genes by subtracting all genes with significant differences in the young controls vs. old controls comparison from the CMML vs. old controls comparison. The top-25 most significantly upregulated genes (Figure 1B) included transcription factors, TNFα signaling genes, genes that regulate genomic stability, and genes involved in apoptosis. The most significantly downregulated transcripts were genes involved in response to DNA damage, RNA binding, monocyte differentiation and mediators of inflammatory process. To link these observations to function, we imputed the 2480 CMML-specific differentially expressed genes into the ingenuity pathway analysis (IPA) application. This analysis uncovered significant enrichment of pathways involved in: mitotic roles of Polo-like kinase, G2/M DNA damage checkpoint regulation, lymphotoxin β receptor signaling, IL-6 signaling and ATM signaling (Figure 1C). DNA methylation profiling revealed 909 differentially methylated regions (DMRs) between CMML and age-matched controls, with most regions being hypermethylated in CMML monocytes. Of these, 37% of the DMRs were intronic, 22% were exonic, 14 % were in the promoter region (Figure 1D), 10% were downstream, 10% were upstream, the remainder were 3' and 5'-overlaps. We also performed integrated analysis using the promoter DMRs and the gene expression profile to identify CMML-associated genes that are likely to be regulated by specific changes in methylation. We observed concomitant changes in CMML-specific mRNA transcripts and DNA methylation promoter regions in the CMML vs. old controls contrast for 10 genes (Figure 1E). AOAH, SERINC5, TAF3 and AHCYL1 were downregulated and hypermethylated; MS4A3, TNF, VCAM1, and IFT80, were upregulated and hypermethylated; TUBA1B was upregulated and hypomethylated and PITPNA was downregulated and hypomethylated. Our study is the first to combine transcriptional and methylation profiling for molecular characterization of CMML monocytes. Conclusions: (i) age-related gene expression changes contribute significantly to the CMML transcriptome; (ii) the CMML-specific transcriptome is characterized by differential regulation of transcription factors, inflammatory response genes and anti-apoptotic pathway genes; (iii) differences in promoter methylation represent only a small proportion of overall differences in methylation, suggesting that intragenic or intronic methylation is a major contributor to the leukemic phenotype; (iv) age-related changes may be necessary, but are not sufficient to realize the CMML phenotype. Figure 1. Figure 1. Disclosures Deininger: Pfizer: Consultancy, Membership on an entity's Board of Directors or advisory committees; Blueprint: Consultancy.

Description: We analyzed a cohort of 26 patients with chronic myeloid leukemia who had failed imatinib and a second tyrosine kinase inhibitor but were still in first chronic phase and identified prognostic factors for response and outcomes. The achievement of a prior cytogenetic response on imatinib or on second-line therapy were the only independent predictors for the achievement of complete cytogenetic responses on third-line therapy. Younger age and the achievement of a cytogenetic response on second line were the only independent predictors for overall survival (OS). At 3 months, the 9 patients who had achieved a cytogenetic response had better 30-month probabilities of complete cytogenetic responses and OS than the patients who had failed to do so. Factors measurable before starting treatment with third line therapy and cytogenetic responses at 3 months can accurately predict subsequent outcome and thus guide clinical decisions.

Description: Abstract 3290 Poster Board III-1 There is a great variability in the degree of molecular responses achieved by chronic myeloid leukemia (CML) patients treated with imatinib. These different levels of molecular response could reflect different degrees of adherence to therapy. We measured the adherence to imatinib therapy in 87 consecutive CML chronic phase patients who had received imatinib 400 mg day as first line therapy for a median of 59.7 months before enrolment (range 25–104) and therefore all them were in complete cytogenetic response. Adherence levels were monitored during a 3-month period using microelectronic monitoring devices (MEMS) and were related to levels of molecular response. MEMS consist of an electronic device fitted in the cap of a normal looking medication bottle that automatically records each time the bottle is opened. MEMS are considered as the ‘gold standard' for measuring adherence. We also measured the imatinib plasma level, the presence of TKD mutations and the following prognostic factors measured at diagnosis: hOCT1 transcripts level, polymorphism 1236C〉T in ABCB1, Sokal risk group, hemoglobin, leukocytes , BCR-ABL1 transcript type and BCR1-ABL1 ratio and demographic data. The study protocol was approved by the Research Ethics Committee and patients gave written informed consent to participate. The median adherence rate was 97.6% (range 22.6–103.8%). In 23 (26.4%) patients adherence was ≤90% (median 76%) and in 12 (13.8%) ≤80% (median 63%). We found a strong association between adherence rate (≤90% or 〉90%) and the 6-year probability of major molecular response (MMR) (28.4% vs 94.5%, p

Description: Background: Treatment free remission (TFR) is now a realistic goal of treatment for CML. Approximately 50% of patients (pts) who discontinue tyrosine kinase inhibitors (TKI) after achieving deep molecular responses (DMR) are able to remain off treatment without losing major molecular response (MR3). Data from the largest available TKI stopping trial, EURO-SKI, showed that the most important variable associated with prolonged TFR is the duration of DMR. However, to date no clinical tool exists to guide clinicians and patients in predicting the likelihood of success of TFR attempt. Methods: We performed a retrospective analysis of clinical data from 172 pts with CML in whom treatment was discontinued in 6 hospital centres in order to identify factors associated with TFR. Data analysis started with a training set (TS) derived from pts treated at a single centre which was then validated on a validation set (VS) derived from the 5 other centres. Eligibility criteria included diagnosis of CML in chronic phase, a minimum duration of treatment with TKI of 3 years and discontinuation of TKI after achievement of confirmed ≥MR4. Patients diagnosed in accelerated phase CML and/or who underwent prior allogeneic stem cell transplant were excluded. Kaplan-Meier method was used for univariate analysis, with log-rank test for group comparison. A Cox proportional hazards model was employed with the purpose of choosing the most influential prognostic predictors on the probability of TFR in MR3 (pTFR3) and TFR in MR4 (pTFR4) on the TS. Variables with a p-value ≤0.1 entered in the multivariate analysis (MVA). Proportional hazard assumptions were tested for the final model. A prognostic TFR score was built from the combination of the predictors identified by the Cox model and validated on the VS. Results: The TS included 118 pts, while the VS 54 pts (Table 1). In the TS, the 2-year pTFR3 was 67.4% (95% CI 66.5-68.3%) and the 2-y pTFR4 was 56.8% (95% CI, 55.9-57.7%). The median time to MR3 loss was 3.8 months (range 1-31), and for MR4 loss was 3.2 months (range 0.8-24.5). After loss of MR4, the 1-year probability of MR3 loss was 77% (95% CI, 70.8-73.2%). However, 10 pts (8.5%) resumed TKI after MR4 loss and were not evaluable for time to loss of MR3. In univariate analysis, the variables most significantly associated with higher pTFR3 and pTFR4 were age at diagnosis 〉40 years (p=0.029 and p=0.002), absence of previous TKI resistance (p=0.003 and p= 0.068), longer duration of MR4 (p=0.003 and p

Description: We sought kinase domain (KD) mutations at the start of treatment with dasatinib in 46 chronic myeloid leukemia (CML) patients resistant to or intolerant of imatinib. We identified BCR-ABL mutant subclones in 12 (26%) cases and used pyrosequencing to estimate subsequent changes in their relative size after starting dasatinib. Four patients lost their mutations, which remained undetectable, 3 patients retained the original mutation or lost it only transiently, 3 lost their original mutations but acquired a new mutation (F317L), and 2 developed another mutation (T315I) in addition to the original mutation within the same subclone. This study shows that expansion of a mutant Ph-positive clone that responds initially to a second generation tyrosine kinase inhibitor may be due either to late acquisition of a second mutation in the originally mutated clone, such as the T315I, or to acquisition of a completely new mutant clone, such as F317L.

Description: Pleural effusions (PE) are a relatively common side effect of dasatinib (an oral multi-targeted kinase inhibitor) and have been reported in 20–30% of CML patients. The underlying mechanism is unclear but may be related to inhibition of the PDGFR gene. We report a series of 43 patients, who received dasatinib while in first chronic phase, after imatinib resistance (n=31) or intolerance (n=12). Twenty two patients had previously received interferon. Patients were treated with dasatinib 70 mg twice daily. The median follow-up was 149 days (range 21–730). Of the 43 patients, 14 developed PE at a median of 150 days (range 21–698) after starting dasatinib. The dasatinib was interrupted in all cases and not resumed until the PE had resolved completely, which occurred in all cases. Diuretics were administered in some cases. The median duration of discontinuation was 14 days (range 7–65). In 8 of the 14 patients the drug was re-started at reduced dose following the first episode of PE. PE recurred in 4 patients (3 patients had two and 1 patient 5 episodes). The dose was further reduced in these cases and eventually abandoned in 2 patients. We performed univariate and multivariate analysis to identify prognostic factors for the development of PE. Significant variables were: prior skin rash on imatinib therapy, [21% non PE vs 57% PE (p=0.035)], skin rash on dasatinib [(before the development of PE vs any time during the follow up) 4% non PE vs 47% PE (p=0.002)], and previous history of autoimmune disease, [non PE 7% vs 50% PE (p=0.006)]. The documented autoimmune diseases were: hyperthyroidism (n=1), hypothyroidism (n=3), systemic lupus erythematosus (n=1), Sweets syndrome (n=1) and auto-immune hepatitis (n=1). We did not find correlation between the previous history of autoimmune disorders and having been treated with interferon. The dose of dasatinib prior to the onset of PE was compared with the dose in patients who did not develop PE at 6 months; patients still receiving 70 mg twice a day were more likely to develop PE than those for whom the dose had been reduced, with a relative risk (RR) for the development of PE of 4.7 (95CI 1.2–18.4, p=0.02). The only variables that were predictive in the multivariate analysis were a history of autoimmune disease and the dose of dasatinib [RR 13 (95CI 1.6–103) and RR 7.4 (95CI 1.2–44.3) respectively]. Interestingly the presence of generalised fluid retention was not found to be significant (p=0.54) nor the previous therapy with interferon (p=1.0). Our results are in agreement with recent reports which suggest that dasatinib-induced PE does not correlate with generalised fluid retention and may be mediated by an immune mechanism. Moreover dasatinib-induced PE are usually exudates and are believed to respond to steroids. In summary in this series PE occurred more frequently in patients with a previous history of autoimmunity and imatinib-related dermatological side effects; both observations support the notion that PE secondary to dasatinib therapy may have an auto-immune pathogenesis.

Description: Abstract 2445 Constitutive activation of signal transducer and activator of transcription 3 (STAT3) correlates with drug resistance and a poor prognosis in many cancers. STAT3 signaling is mediated by phosphorylation at tyrosine-705 (p STAT3Y705), dimerization, and nuclear transactivation. In chronic myeloid leukemia (CML), pSTAT3Y705 is demonstrable under two distinct resistance scenarios: (1) extrinsic resistance, in which BCR-ABL1 kinase-independent survival signals originating from the bone marrow (BM) microenvironment activate pSTAT3Y705 in a JAK2- or TYK2-dependent manner, and (2) intrinsic resistance, in which BCR-ABL1 kinase-independent signals activate pSTAT3Y705 in response to kinase inhibition. Based on these observations, we identified TKI-resistant CML as an excellent model for developing and optimizing pharmacologic STAT3 inhibitors. Using K562 and AR230 CML cells that are resistant to 1 μM imatinib (K562R and AR230R; intrinsic resistance) and primary CML CD34+ progenitor cells exposed to BM stromal-derived conditioned medium (CM; extrinsic resistance), we examined the effects of direct pharmacologic inhibition of STAT3 in TKI-resistant CML. Here, we report the design and validation of next-generation STAT3 inhibitors identified through computational modeling and screening in AR230R CML cells expressing high levels of pSTAT3Y705. We initially examined the effects of an established STAT3 inhibitor, S3I-201.1066 (SF1–066). K562R or AR230R cells were treated with 1 μM imatinib and/or 10 μM SF1–066, followed by culture in methylcellulose medium and scoring after 14–16 days. Combination treatment reduced the clonogenicity of K562R and AR230R cells to 31.4% (p

Description: Abstract 3515 With the advent of tyrosine kinase inhibitors (TKI), allogeneic stem cell transplantation (allo-SCT) is largely reserved for patients with CML who do not achieve durable cytogenetic responses to TKIs or patients with advanced phase (Adv) disease. Data relating to the outcome of transplant in Adv-CML is limited. We have allografted 43 patients (median age 40.8 yrs) for Adv phase disease who had received prior treatment with one or more TKI. The indications for allo-SCT included progression from CP to accelerated phase (AP) (n=16) or blast crisis (BC) (n=11) on TKI and presentation in accelerated phase (AP) (n=9) or blast crisis (BC) (n=7). The median duration of TKI therapy prior to transplantation was 5.5 months (range 1–51 months); 42 patients received imatinib, 9 received dasatinib (8 following imatinib failure), 2 received nilotinib (following imatinib and dasatinib failure). 35 patients were transplanted from HLA-identical siblings and 36 patients received myeloablative conditioning. The status at transplant was CP〉1 in 17 patients, AP in 24, and BC in 2. In patients in whom CP was restored prior to transplant (n=17), this had been achieved using a TKI only in 6 and with combination chemotherapy in 11. There was no difference in disease-free survival (DFS) or overall survival (OS) between the TKI only group and the group that received chemotherapy in addition. Among the 43 patients in the TKI-treated cohort, 13 died without relapse, 3 from graft versus host disease (GVHD), 8 from sepsis, pneumonitis and multiorgan failure, and one each from graft failure and VOD. The estimated probabilities of non-relapse mortality (NRM) at 100 days and 1 year were 17.3% and 43.3%, Grade 2–4 acute GVHD was seen in 24% and extensive cGVHD in 54%. The estimated 1- and 3-year DFS rates were 23% and 16%. The 1 year and 3 year estimates of overall survival according to disease stage at allo-SCT were as follows: AP (54.2% and 50%), CP 〉1 (49.4% and 29.6%) and BC disease 0% and 0%. The impact of maximal disease stage was examined, documented as either AP (23/43 patients) or BP (20/43 patients) at any time prior to allo-SCT. The probability of 3 year OS for patients who were in AP at maximal disease stage was 61% compared to 33% of patients who had at one time been in BC (p=0.04). Post allo-SCT, patients were monitored for relapse by RQ-PCR. Eleven patients received TKIs, 5 for molecular relapse, 1 for cytogenetic relapse, 4 for hematological relapse and 1 for GvHD. Three of the 11 remain alive, 2 of whom received a TKI for molecular relapse.We compared the outcome of these 43 patients with that of 158 patients who were transplanted for Adv-CML but who had been treated before TKI became available. The disease status at time of transplant was AP (n=90), CP〉1 (n=41) and BC (n=27). The two groups were matched for type of donor, conditioning regimens and time from diagnosis to transplant but the historical group were younger at allo-SCT with a median age of 33.3 yrs (p=0.001). There were no significant differences in the incidences of acute and chronic GvHD, NRM, DFS or OS between the two groups. The 1 year and 3 year estimates of OS for the historical cohort were 46.4% and 38.5% in AP, 53.7% and 24.3% for CP 〉1 and 7% and 0% in BC. For the total group of 201 patients the outcome of transplant defined as 3yr OS was 40.9% for AP 25.7% for CP〉1 and 0% for BC. In conclusion, we found that patients receiving transplant for advanced phase disease after prior treatment with a TKI have similar outcomes to a historical group of advanced phase patients transplanted prior to the advent of TKI therapy. Our data strongly support the influence of disease stage in prediction of allo-SCT survival. Allo-SCT may be valuable for CML patients who have never progressed to BC. Overt BC is a predictor of poor allo-SCT outcome, so attempts should be made to restore CP prior to allo-SCT. Close monitoring of patients still classifiable as AP who are responding poorly to TKI should permit identification of those who may do well if offered allo-SCT before their disease has progressed further. Disclosures: Marin: Novartis: Consultancy, Honoraria, Research Funding; Bristol Myers Squibb: Consultancy, Honoraria, Research Funding.

Description: Abstract 3397 The tyrosine kinase inhibitor (TKI), imatinib mesylate (IM), induces remissions in most chronic phase chronic myeloid leukaemia (CML-CP) patients, but emergence of drug resistance attributable to critical mutations in the BCR-ABL1 TK domain remains a significant clinical problem. We previously demonstrated that primitive (CD34+) CML-CP cells are both intrinsically insensitive to IM and genetically unstable. BCR-ABL1 TK domain mutations are, furthermore, readily detectable in CD34+ CML-CP cells even prior to IM treatment. However, the mechanisms by which they arise are unknown. In this study, we characterized the spectrum of mutations at each of the 3 codon positions in the BCR-ABL1 TK domain of IM-naïve CD34+ cells and used a mathematical model to compare these with those found in the same region of DNA in IM-resistant cells. A total of 460 TK mutations were identified in cells from 15 IM-naïve and from 316 IM-resistant patients. The mutations were non-randomly distributed across the 3 codon positions, with very few mutations at position 3 (1 of 136 mutations, 0.7%, in IM-naïve patients and 31 of 329, 9%, in IM-resistant patients). The remainder showed no bias in their distribution across codon positions 1 and 2 in IM-naïve patient cells but, in IM-resistant cells, a significant positive bias was observed at codon position 2 (p=0.042). Conversely, BCR-ABL1 TK domain sequences from IM-naïve patient cells had significantly more transitions relative to unselected regions of the genome across all codon positions, a significant positive A-to-G mutational bias (A〉G, 1.12×10−10) and an under-representation of C-to-T mutations (C〉T, 5.34×10−5) when these mutations were compared with unselected regions of the genome. In addition, we observed a T-to-C mutational hotspot (T〉C, 1.23×10−4) at codon position 2. IM-naïve cells exhibited a transitional bias at position 2 and overall (1.99×10−4 and 6.19×10−5, respectively; at position 1, P=0.052). We observed a similar pattern of TK codon position mutations at position 1 and overall in sequences derived from IM-resistant cells. Interestingly, in IM-resistant cells, A-to-T (A〉T) transversions were over-represented at position 2 and overall (1.41×10−11 and 3.77×10−5). These cells also had a profound transversional bias (1.45×10−4) at codon position 3, suggesting that positive selection occurs at a position where transversions are typically non-synonymous. However, a skewed transition:transversion ratio and transition and transversion frequency at TK codon positions 1 and 2 was observed in both IM-naïve and IM-selected cells with mutations distributed across these 2 codon positions in a significantly uneven fashion. In IM-resistant cells, the frequency of C-to-T mutations (C〉T, 3.82×10−11) indicated a bias affecting codon position 2 more often than position 1, and A-to-T transversions occurred significantly more frequently at position 2 compared to position 1 (A〉T, 7.45×10−9). This was also the case for T-to-C transitions in TK sequences of IM-naïve cells (T〉C, 1.13×10−4). Frequencies of these TK mutations in IM-naïve and IM-resistant cells were also higher than in unselected regions of the genome, providing further evidence that the TK domain mutations in CML-CP cells have a distinct mutational profile. For example, the clinically observed M244V and D276G mutations result from A-to-G transitions and the F359L mutation arises from a T-to-C transition, both of which would be predicted by the activity of a CML-CP mutator. However, the most clinically important mutation, T315I, that confers resistance to most currently available TKIs, is generated by a C-to-T transition, suggesting a high mutational rate generating mutational escape around the principal mutator pattern and profound selection thereafter. Notably, the CML-CP mutational signature is distinct from that of the activation-induced cytidine deaminase (AID)-induced hypermutation reported in CML blast crisis, both in targeting specific codon positions and in the overall mutational pattern. This strongly favors the hypothesis that the genomic instability of primitive CML-CP cells is generated by a different, specific mutational process. The results also suggest that monitoring BCR-ABL mutational hotspots may be clinically useful in anticipating TKI resistance. Disclosures: No relevant conflicts of interest to declare.

Description: Abstract 31 In CML, TKI resistance in the absence of BCR-ABL1 kinase mutations is mechanistically unclear. Since extrinsic signals from the bone marrow (BM) microenvironment protect CML cells from TKIs in a STAT3-dependent manner (Bewry et al. 2008; Traer et al. 2011), we hypothesized that overt resistance may occur when CML cells maintain intrinsic STAT3 activation in the absence of extrinsic signals. We also asked whether combined targeting of STAT3 and BCR-ABL1 could produce synthetic lethality to overcome TKI resistance. To model extrinsic BCR-ABL1 kinase-independent resistance, we grew CML cell lines and CMLCD34+ progenitor cells from newly diagnosed patients in the presence of conditioned medium (CM) derived from HS-5 BM stromal cells. To model intrinsic resistance, we used the imatinib-resistant K562R and AR230R cell lines. Our models of TKI resistance proliferate in 1.0–2.5 μM imatinib and exhibit pSTAT3Y705 activation despite suppression of BCR-ABL1 kinase activity. To investigate the role of pSTAT3Y705 in TKI resistance, we used shRNA-mediated knockdown (shSTAT3), a dominant-negative mutant (dnSTAT3), and pharmacologic STAT3 inhibitors developed in our laboratories to block STAT3 activity. Compared to scrambled controls in the presence of HS-5 CM, shSTAT3 reduced colony formation (33.8% reduction, p