ALBERT

Description: Graft rejection (GR) following an allo-SCT occurs in 10-20% of patients with β thalassemia major (TM). The reported clinical outcome following second transplants have been associated with a high incidence of graft rejection, treatment related mortality, and graft versus host disease [Haematologica 2009; 94(9)]. There is limited data on the clinical profile and long term outcome of patients who have had a graft rejection. We undertook a retrospective analysis of patients who had a graft failure post allogeneic SCT for TM at our center. From October, 1991 to June 2016, 506 HLA matched related transplants for TM were done at our center. Of these 55 (11%) patients had a graft failure. An additional 7 patients with graft failure following an allo-SCT done at other centers were referred to us for a second transplant. Of the 62 patients with graft failure 32 (52.4%) were primary graft failures (PGF; 15 with aplasia and 17 with autologous recovery) while 30 (47.6%) were secondary graft failures (SGF; 5 with aplasia and 25 with autologous recovery). The median age of the patients who had graft failure was 8 years (range:1-19) and there were 38 (60.3%) males. On conventional risk stratification 40 (63.5%) were Class III. Eighteen (54.5%) cases with PGF and 16 (53.3%) with SGF did not receive a second transplant. From Oct 2009, at our center, a reduced toxicity myeloablative (MAC) treosulfan based conditioning regimen with a PBSC graft was offered to all high risk patients and for second transplants at the treating physician's discretion. Twenty nine (46%) of the patients with GR underwent a second allo-SCT. With the exception of one patient (first allo-SCT with an unrelated cord blood product) the donor for the second transplant was the same as the first transplant. Conditioning regimen for second SCT was busulfan based MAC in 7 (24%), treosulfan based MAC in 12 (41.3%) and the remaining received non-myeloablative conditioning regimens (fludarabine based, low dose TBI, OKT3, Cy-OKT3) often in view of pancytopenia and perceived inability to tolerate a MAC. All patients receiving a treosulfan based regimen had a PBSC graft. A BM graft was used in 7 (41%) of the remaining cases. None of the patients conditioned with a treosulfan based regimen had a graft rejection though one patient died of Grade IV acute GVHD. Of the remaining 17 patients 10 died following a second GR, 3 died of regimen related toxicity. Four are alive of which one has recurrent TM and the rest are well and transfusion independent at 55, 80 and 204 months from second transplant (all busulfan based MAC). The baseline characteristics and clinical outcomes of patients who received a treosulfan based MAC regimen versus the rest is summarized in table 1 and figure 1. On a univariate analysis a non-treosulfan based conditioning regimen and time from GR to second transplant of

Description: Among transplant related complications, graft versus host disease (GvHD) significantly affects survival among patients undergoing allogeneic hematopoietic stem cell transplantation (aHSCT). There is limited data on GvHD and its impact on outcomes of aHSCT in patients with thalassemia major (TM). We have reviewed the incidence and outcome of GVHD among patients with TM who underwent aHSCT at our center. All patients with TM undergoing their first aHSCT between January 2007 and December 2017 were included in this analysis. Till 2009, all patients received conditioning with busulfan (16mg/kg over 4 days) with cyclophosphamide (200mg/kg over 4 days). From 2010, most patients received treosulfan (42 G/m2 over 3 days) with thiotepa (8mg/Kg for one day) and fludarabine (160mg/m2over 4 days) based conditioning regimen. All patients receiving busulfan conditioning received bone marrow (BM) as the graft while most patients receiving treosulfan conditioning received mobilized peripheral blood stem cells (PBSC). GvHD prophylaxis was with short-course methotrexate (10mg/m2 on day +1, and 7mg/m2 on days 3, 6 and 11) with cyclosporine. Thymoglobulin was added for matched unrelated donors (MUD). GvHD was prospectively recorded and graded according to the Glucksberg classification. Between January 2007 and December 2017, 363 first transplants were done for patients with TM with HLA identical donors. There were 12(3.3%) class 1, 105(28.9%) class 2 and 246(67.8%) class 3 (Pesaro risk stratification), with 115(46.7%) of the latter being high risk (Vellore risk stratification - BBMT, 2007; 13: 889). The median age was 8 years (range: 1-25) with a male predominance (60%). 331 (91.2%) patients had matched related donors (MRD) and 32 (8.8%) had MUDs. Donor gender was mismatched in 207 (57%) of which 129 (35.5%) were female to male transplants. The graft was obtained from the bone marrow in 137 (37.7%) of whom 53 (38.7%) were class III, and from mobilized peripheral blood in 226 (62.3%) of whom 193 (85.4% were class III. 149 (41%) patients developed GvHD - acute GvHD (aGvHD) in 115 (31.7%) and chronic GvHD (cGvHD) in 80 (22%). aGvHD was grade I in 32 patients (27.8%), grade II in 36patients (31.3%), grade III in 16 patients (13.9%) and grade IV in 25 patients (21.7%), while 6 patients (5.2%) had features of overlap GvHD only (oral lichen planus). First line treatment was with steroids in all patients with grade II and above aGvHD (n=83) with 43 (51.8%) of them responding adequately. There were 37 patients (44.5%) who required various second line agents for aGvHD with 20 (24%) receiving more than one immunosuppressive agent. 20 patients (24%) with persistent aGvHD went on to develop cGvHD. Out of the total of 80 patients with cGvHD, 13 (16.3%) had limited and 67 (83.7%) had extensive cGvHD. 26 patients (32.5%) developed de novo cGvHD, 8 (10%) of them after donor lymphocyte infusion (DLI) for potential rejection. The other 46 patients (57.5%) had chronic overlap GvHD following previous aGvHD. Among the different variables evaluated for association with aGvHD (patient/donor age, gender mismatch, MUD vs MRD), none were significant. Among those with MRD, aGVHd occurred in 36/135 patients (26.7%) of patients receiving BM grafts compared to 65/196 patients (33.2%) who received PBSC grafts (p=ns). cGvHD occurred in 23/106 patients (21.7%) in those receiving BM grafts vs 52/171 patients (30.4%) receiving PBSC (p=ns). 30 patients (8.3%) persisted to have cGvHD at last follow-up but only 20 (5.2%) required treatment. Mortality of the whole cohort was 66 (18.2%), out of which 32 (8.8%) were related to GvHD - 24 (6.6%) due to aGvHD and 8 (2.2%) due to cGvHD. At a median follow up of 41 months (range: 0-148), the 5-year and 10-year overall survival (OS) was 81.1±2.1% each for the whole cohort. The 5-year OS of those with grade 2-4 aGvHD was significantly lower than those with grade 0/1 aGvHD (65.7±5.3% vs 85.7±2.2%, p=0.000) [figure 1]. The 5 year OS of those with cGvHD was 88.9% ± 3.8% as compared to those without cGVHD was 96.9% ± 1.2% (P=0.009) [figure 2]. There was no significant difference in OS among those with limited and extensive cGvHD (90.9±3.6% vs 88.4±4.2% (p=ns). Our data shows that, as expected, severe aGvHD and extensive cGvHD significantly lowers survival in patients with TM undergoing aHSCT. However, PBSC graft did not result in higher acute or chronic GVHD compared to BM. Disclosures No relevant conflicts of interest to declare.

Description: Management of acute myeloid leukemia (AML) in India remains a challenge. A major constraint is the cost of therapy. In a predominantly self paying system the majority of patients will not have the resources to manage a subsequent relapse. Hence, the choice of consolidation therapy has to be carefully considered to balance cost and efficacy. An allogeneic SCT (alloSCT) with a reduced intensity conditioning regimen (RIC) in first remission (CR1) is an attractive option to fulfill these requirements of relatively low cost without compromising efficacy. The need for consolidation chemotherapy prior to offering a RIC alloSCT for AML CR1 remains controversial. To evaluate these aspects we undertook a retrospective analysis of patients with AML CR1 who received a RIC alloSCT from multiple centers in India. Conventional criteria were used for definition of conditioning regimens to be considered RIC (CIBMTR). Data from 8 centers in India was collected between 2005 and 2013. A total of 138 patients fulfilled the criteria of AML CR1 having received an alloSCT with a RIC regimen. The median age was 34 years (range: 2 – 63) and 60% were males. The median time from diagnosis of AML to transplant was 99 days (range: 41 – 504). 123 (89%) were HLA matched related donors, 3 (2.1%) were MUD transplants and the rest were HLA mismatched related donors. The majority by cytogenetics (n=115) were intermediate risk (76%) followed by high risk (23%). 70 (51%) received chemotherapy consolidation prior to transplant, 61 (44%) did not and data was not available in 7 (5%). 68% of those that received consolidation received intermediate or low dose cytosine based regimens. 129 (94%) were CMV serology positive pre-transplant. Fludarabine with melphalan (140mg/m2) (128{93%}) was the most commonly used regimen and cyclosporine with short course low dose methotrexate (126{91%}) the most commonly used GVHD prophylaxis regime. All patients received a PBSC graft with a median CD34 cell dose of 9.1x106/kg (range: 1.3 – 43). With the exception of one, all patients engrafted. The median time to ANC 〉500/mm3 was 13 days (range: 7 – 22) and platelet count of 〉20,000/mm3 was 15 days (range: 0-33). Of those that engrafted, 97% achieved complete chimerism at one month post transplant (data not available in 4). Post transplant CMV reactivation was seen in 32% and a fungal infection (possible, probable or definitive) in 13%. Acute GVHD Grade 2-4 was seen in 29% and of patients evaluated 62% had chronic GVHD, the majority of these being limited (61%). The 100 day treatment related mortality (TRM) was 7.5% and the one year TRM was 25.6%. At a median follow up of 24 months the 5 year EFS and OS was 64.0±5.07 (Figure 1A) and 71.1±4.0 respectively. The 5 year cumulative incidence of relapse was 21.8% (Figure 1A). The baseline characteristics as mentioned above were not significantly different between the group that received consolidation and the group that did not. The use of consolidation therapy prior to alloSCT did not have a significant impact on EFS or OS (Figure 1B). On univariate analysis the factors that adversely impacted EFS were mismatched non sibling family donor (RR 8.1; P-value 0.001), CMV reactivation (RR 2.6; P-value 0.001), fungal infection post transplant (RR 6.8; P-value 0.000) and acute GVHD (RR 2.1; P-value 0.02). On a forward stepwise multivariate analysis adjusting for these and other conventional risk factors only CMV reactivation (RR 2.0; 95% CI 1.03-3.87; P-value 0.042) and fungal infection (RR 7.1; 95%CI 3.154-16.12; P-value 0.000) retained their adverse impact. There was no correlation between CMV reactivation and relapse of disease post transplant. The mean costs of induction chemotherapy for these patients was US$ 9239±3596 (n=74), for consolidation chemotherapy it was 5007±3490 (n=21) and for alloSCT it was 18138±13826 (n=118; costing up to 1 year post transplant). Induction chemotherapy followed by HLA matched RIC alloSCT is likely to be a cost effective and affordable treatment option for young adults with AML in CR1in an Indian context. With an average gross net income in India of US$3500/year (http://indiabudget.nic.in) the limitation still remains the cost of treatment and number of centers that can offer this therapy. Figure 1 Figure 1. Disclosures Srivastava: Octapharma: Consultancy, Other. Off Label Use: Bortezomib in the treatment of acute promyelocytic leukemia.

Description: Abstract 518 Targeted dosing of busulfan (Bu) has been shown to improve outcome of allogeneic HSCT (aHSCT) in patients with leukemia. There is limited data on correlation of Bu PK with outcome in patients with thalassemia major (TM)) undergoing aHSCT. We have previously shown that first dose trough level of Bu (Cmin1) predicts graft rejection (Chandy et al, BMT 2005), and Bu Css is significantly lower in patients with hepatic veno-occlusive disease (HVOD) (Srivastava et al, Blood 2004). The aim of the present study was to evaluate the correlations of Bu PK with outcome of HSCT in a larger cohort of patients and to evaluate the pharmacogenetic basis for the differences. We retrospectively analyzed oral Bu PK after 1st and 13th doses of Bu in 255 patients out of the 291 thalassemic patients who underwent aHSCT from matched related donors between 1991 till February 2010 at our centre. All patients received busulfan (at a dose of 14 or 16 mg/kg/day or 600mg/m2/day) in combination with Cy (at a dose of 160mg/kg for those 〉15 years or 200mg/kg for all others) as part of the conditioning regimen. Bu levels were measured by HPLC as previously described (Poonkuzhali et al, 1999). We also analyzed GSTA1*1B, GSTM1 and GSTT1 deletion polymorphisms in these patients. Based on Lucarelli's risk stratification, 18/291 patients belonged to class I, 121/291 were class II and 151/291 were class III. The class III patients were further risk stratified into class 3 high risk and low risk based on age and liver size (high risk: age 〉7 years and liver size 〉5cm; and the rest as low risk; Mathews et al, 2007). None of the Bu PK parameters were significantly different between Lucarelli classes as well as between class III high and low risk patients. For the entire group, EFS was 77%, OS 81%, NRM (non rejection mortality)15% and graft rejection 8.6%. Class III patients had a significantly lower EFS (p=0.0007) and OS (p=0.0051) compared to class I and II. Bu Cmin1 (p=0.007), but not Bu Css1 was significantly lower in those who rejected their graft compared to those who did not. On quartile analysis, patients with Cmin1 156ng/ml (RR= 9.8; p=0.0001). Those with Bu Css1 in the lowest quartile also had significant risk of rejection (14/57 with Css1 490 ng/ml; RR= 3.8, p=0.027) but the correlation was not as strong as that with Cmin1. Upon multivariate analysis of all the variables that were significantly influencing aHSCT outcome in univariate analysis, only Lucarelli class III high risk (p=0.034), SGOT level (p=0.036) and Bu Cmin1 (p=0.0001) were significantly influencing graft rejection. In addition, GSTA1*1B homozygous variant genotype was significantly associated with higher Bu Cmin1 (p=0.008) and Css1 (p=0.009). Since Bu Cmin1 was significantly influenced by GSTA1*1B genotype, we compared the combined risk of Cmin1

Description: Acute myeloid leukemia (AML) is a clinically and biologically complex and heterogeneous hematopoietic neoplasm. Recent advances in acute myeloid leukemia (AML) biology have lead to prognosticate and predict treatment outcome in AML based on molecular markers. Mutations in NPM1, CEBPA are considered good prognosis and High BAALC, ERG & MN1 expression associate with worse outcome in AML patients treated with standard chemotherapy. Although many efforts have been made to identify genetic mutations and modulated gene expression levels that can be used to predict outcomes in patients with AML, the association between these prognostic markers has not been evaluated. We have reported previously that the NPM1 mutated patients have significantly high dCK and hENT1 gene expression (involved in cytarabine metabolism) and low ABCG2 and ABCB1transporter expression (Abraham et al, ASH abstracts; Nov 2011; 118: 3481 and Nov 2012; 120: 143), suggesting that the good prognostic nature of this mutation is possibly due to the better metabolism and transport of the chemotherapeutic drugs used in induction therapy.  We extended this study to look for association between NPM1/FLT3 mutation status and the RNA expression of other good or poor prognostic markers in patients with AML. We prospectively included 274 adult patients with AML in this study. The median age was 42 years (range 16-74y). AML was diagnosed according to the FAB and WHO classifications. There were 238 patients with de novo AML; Secondary AML -6; Therapy related AML- 2 and Relapsed AML-28. Bone marrow cytogenetics and immunophenotyping analysis was available for all patients at diagnosis and/or relapse. Diagnostic bone marrow MNCs were isolated by ficoll- density gradient centrifugation and stored in trizol reagent for RNA expression and mutation detection. RT-PCR was used to screen AML-ETO and Inv 16, and the expression of BAALC, ERG1, MN1, CXCR4 and WT1were analyzed using RQ-PCR. NPM1-c, FLT3 ITD and TKD were screened using DNA PCR followed by gene-scan, sequencing or RFLP methods. The basic demographics and the frequency of the markers are listed in Table 1. When analyzed separately in normal karyotype AML (NK-AML), the frequencies of the mutations were: NPM1: 52.2%; FLT3-ITD: 24%; TKD: 4.3%. When the RNA expression of BAALC, WT1, ERG1, CXCR4 and MN1 was compared in patients with NPM1 or FLT3 mutations, we noticed that patients with NPM1 had significantly low expression of BAALC, MN1 and ERG1 while those with FLT3 mutations (ITD or TKD) had higher expression of these genes (Figure1). There was no significant association with CXCR4 or WT1 expression and these mutations. When analyzed separately in the normal karyotype AML, these associations were still significant. In addition, the relapsed patients had significantly higher expression of BAALC, MN1, and ERG1 RNA compared to de-novoAML cases (data not shown). To conclude, we show that NPM1 or FLT3 mutations acquire the prognostic significance due to several factors including BAALC, ERG1 and MN1expression levels in addition to drug metabolizing enzymes’ and drug transporter expression. These factors must be taken into consideration when attempting to personalize chemotherapy in AML. Disclosures: No relevant conflicts of interest to declare.

Description: Introduction: Allogeneic stem cell transplantation (SCT) is the best form of therapy for a young patient (〈 50 years) with severe aplastic anaemia. In developing countries, there is a big time interval between diagnosis and SCT leading to increased transfusions and increased risk of infections, both of which adversely affects transplant outcome. This retrospective analysis is aimed at studying the outcomes of SCT among Indian patients with aplastic anaemia. Methodology: The Indian Stem cell transplant registry (ISCTR) is a group of transplant physicians representing about 30 active transplant centres in India. This retrospective analysis was done on data reported on 634 patients by 20 centres who reported outcomes of SCT for aplastic anaemia. Data was collected from individual medical records and databases. Analysis was done using SPSS software version 16.0 Results: Six hundred and thirty four patients [445 males and 189 females] with a median age of 21 years (range: 2 - 65) underwent allogeneic SCT between 1990 and March 2015. There were 209 children (age 〈 15 years). The median time from diagnosis to SCT was 5 months (range: 1 - 120) while the median number of transfusions was 20 (range: 1 - 150). All donors were HLA identical sibling or family donors; matched unrelated and haplo-identical donor transplants were excluded from this analysis. Conditioning regimen was Cyclophosphamide based (Cy/ Cy+ ATG/ Cy+ TBI/TLI) in 78 patients (12.3%) while majority received Fludarabine with Cyclophosphamide (n = 481; 75.8%) and 75 received other conditioning regimens (Flu/TBI, Flu/Bu, Bu/Cy etc). Graft source was bone marrow [BM] in 124 (19.5%) and peripheral blood stem cells in 510 patients (80.5%). Graft versus host disease (GVHD) prophylaxis predominantly consisted of Cyclosporine and methotrexate in 543 patients (85.6%). Engraftment was seen in 572 patients (90.4%) while 19 (2.9%) had primary graft failure and 43 (6.7%) expired prior to engraftment due to infection or bleeding. The median time to neutrophil engraftment was 13 days (range: 8 - 21) while platelet engraftment occurred at 13 days (range: 5 - 37). Grade II - IV acute GVHD occurred in 29.3% while grade III-IV was seen in 14.1%. Chronic GVHD was seen in 41% of evaluable patients which was limited in most patients. At a median follow up of 43 months (range: 1 - 264), 431 patients are alive. The 5 yr OS for the entire group is 66.3 + 2.0%. The OS was higher in children compared to adults (73.4 + 3.3% vs 62.8 + 2.5%; p = 0.006), better for Flu/Cy compared to Cy based conditioning (69.8 + 2.2% vs 57.8 + 5.6%; p = 0.002) and better for PBSC compared to BM (68.9 + 2.2% vs 56.1 + 4.8%; p = 0.020). There has been significant improvement in outcomes over the past 15 years [3 yr OS of 41.9 + 1.3% for 1984-1995, 40.9 + 1.2% for 1996-2000, 70.6 + 5.0% for 2001 -2005, 70.3 + 3.1% for 2006-2010 and 68.8 + 3.0% from 2011 onwards]. Conclusion: Outcomes of patients with aplastic anaemia are improving and patients have a 70% chance of getting cured with a HLA identical sibling donor transplant. The use of PBSC as graft source is not associated with inferior outcomes. Disclosures No relevant conflicts of interest to declare.

Description: A toxicity reduced conditioning regimen containing Treosulfan (Treo), fludarabine (Flu), thiotepa for high risk Thal Major (TM) has been used since 2009 at our centre that has significantly improved transplant outcomes of these patients compared to the historical cohort of patients receiving busulfan/ cyclophosphamide based myeloablative regimen (Mathews et al, 2013). Limited knowledge is available on the pharmacokinetics (PK), pharmacogenetics (PG) and pharmacodynamics of fludarabine and treosulfan, especially in non-malignant hematological disorders like TM. We describe here the PK of Flu and Treo in patients with TM undergoing HSCT, the factors influencing the inter-individual variability in PK and the role of these factors on HSCT outcome. Seventy one patients diagnosed with TM undergoing HSCT with Flu/Treo based conditioning regimen between January 2012 and January 2015 were included (Table: Patient demographics). Selected functional polymorphisms in the NT5E, DCK, hENT1 and GST genes that are involved in fludarabine or treosulfan metabolism were screened. All patients received Flu 40mg/m2/day x 4 days as an 1hr infusion on days 1 and 4 and Treo as 14g/m2/day x 3 days at the rate 5g/hr. Plasma was separated from the peripheral blood collected at predetermined time points after the infusion of Flu and Treo PK analysis. Plasma Flu was analyzed using a LC-MS/MS method and the concentration was expressed as mMole/ml while Treo was analyzed using a HPLC-RI method and concentration was expressed as mg/L. Flu and Treo PK was estimated using nonlinear mixed effects modeling via Monolix 4.3.3. The covariates tested for both PK were: age, sex, body weight, BSA, ferritin, and polymorphisms in NT5E, hENT1, dCK and GST genes. The PK parameters AUC, CL, V and k were estimated on day 1 for Treo and on day 1 and day 4 for Flu (Table). The influence of Flu and Treo PK and PG on graft rejection, early transplant related mortality (TRM) & chimerism status was estimated using logistic regression analysis. Wide inter-individual variation in Flu and Treo PK was noted (7 and 9 fold Vs 5 and 8 fold respectively for Day 1 & 4 Flu AUC & Cl; 33 & 31 fold variation in Treo AUC and Cl) (Table). Flu CL was significantly higher on day 4 compared to day1 (Figure A). The variation in Flu PK was explained by genetic variants in NT5E and dCK. Patients having variant genotype for the SNPs in NT5E (rs2295890) and dCK (rs11544786) showed significantly lower plasma Flu clearance compared to those with wild type genotype (p=0.006 & p=0.05 respectively) (Figure B). This is consistent with our previous report in patients with aplastic anemia undergoing HSCT (Mohanan et al. 2014; Blood: 124 (21)). None of the genetic variants in the GST genes explained the variation in Treo PK. Day21 mortality was seen in 6/71 patients (8.5%) and graft rejection in 3/66 evaluable patients (4.5%). Analysis of the influence of PK and PG variables on transplant outcome showed significantly high first dose Flu AUC to be associated with D21 mortality upon Univariate analysis (median 42.5, range 32.1-63.7 compared to 31.8, range 15.2-111 mMole*h/mL, in those with and without TRM respectively; p=0.043); none of these parameters were significantly associated with graft rejection or mixed chimerism. There was no association between Treo PK parameters and graft rejection or TRM. The influence of Flu and Treo PK on regimen related toxicity is yet to be evaluated. The lack of the influence of PK on transplant outcome could be due to lower incidence of rejection and TRM in this cohort. Further analysis in a larger cohort of patients will be done once we enroll more patients for PK analysis. Our results demonstrate that Flu PK is influenced by genetic variants in NT5E and dCK, the enzymes involved in Flu biotransformation. The relationship between high-plasma Flu exposure and TRM and given the fact that multiple factors influence TRM, we can extrapolate that the plasma Flu AUC may be a surrogate marker of overall preparative regimen intensity as reported previously (Long-Boyle et al, Bone Marrow Transplant, 2011). The lack of association of genetic variants in GST genes in explaining the inter-patient variability in treosulfan exposure suggests the involvement of other drug metabolizing genes on treosulfan PK. We are currently evaluating the role of genetic variants in a large panel of drug metabolizing genes on explaining this inter-individual variability in Treo PK. Disclosures No relevant conflicts of interest to declare.

Description: Abstract 2746 Introduction: Newly diagnosed and relapsed acute promyelocytic leukemia (APL) patients respond to therapy with arsenic trioxide (ATO) based regimens. Significantly more patients with relapsed APL have disease recurrence after ATO based therapy than newly diagnosed cases. We undertook a series of experiments to evaluate the potential mechanisms to explain this increased recurrence rate in patients with relapsed APL. Patients and methods: From April 2007 to March 2009 bone marrow samples from newly diagnosed and relapsed cases admitted at our center were utilized for these studies. If required the bone marrow blasts and promyelocyte component was enriched to above 90% using a lineage depletion cocktail and VarioMACS (Miltenyi Biotec, Germany). For in-vitro intracellular ATO concentration measurement, 2 × 107 cells were washed and suspended in RPMI media with 0.5 μM concentration of ATO and incubated for 24 hours. Cells were then washed, made into a pellet and solubilized with HNO3 and H2O2 and ATO content measured using an atomic absorption spectrophotometer. An in-vitro sensitivity assay of malignant cells as previously reported was standardized using an MTT assay system. The impact of co-culture of mesenchymal stromal cells (MSC) and malignant promyelocytes on ATO induced apoptosis was studied with 7AAD and Annexin staining using a flowcytometer. A gene expression array using 44k human microarray chip analysis (Agilent technologies) was done on 8 newly diagnosed and 8 relapsed cases. Results: Sixty five patients were included in this study. Of these 47 (72%) were newly diagnosed and 18 (28%) were relapsed cases. On immunophenotyping, CD34 was positive (〉20%) in 3.6% of newly diagnosed and 50% of relapsed cases (P=0.001). The mean MFI of the relapsed cases for expression of CD38, VLA-5 and CD13 was significantly lower in the relapsed group. The ability of both newly diagnosed and relapsed primary APL cells to concentrate ATO intracellular was not significantly different (15.2±9 nG/107 cell Vs. 16.3±9.7 nG/107 cell). Similarly the in-vitro IC-50 assay was not significantly different between the two groups (5.5±3.8 Vs. 4.7±4.5 μM). Neither of these assays correlate with clinical parameters such as relapse, event free (EFS) or overall survival (OS). Evaluation of the impact of MSC on ATO induced apoptosis demonstrates a protective effect in newly diagnosed and relapsed cases (Fig 1). This effect was mediated partly by the MSC conditioning media and could not be overcome by addition of VLA-4 or VLA-5 blocking antibodies (data not shown). Gene expression studies comparing the two cohorts revealed 1744 genes that were differentially expressed (〉2 fold) between samples at diagnosis and at relapse. Quantitative Real-time RT-PCR using SYBR- Green detection system was done to confirm the gene expression results obtained from microarray analysis. Using ΔΔCT method the fold difference was calculated for five selected genes which validated the microarray data. Conclusion: Relapsed patients have significant immuno-phenotypic differences from newly diagnosed cases. Mechanisms of resistance to ATO are probably multi-factorial but are unlikely to be related to intra-cellular ATO concentration. In-vitro IC-50 does not appear to predict clinical outcomes. Stromal interaction protects malignant promyelocytes from the apoptotic action of ATO which appears more pronounced in relapsed than in newly diagnosed cases. Further evaluation of parameters that enhance such stromal interaction and protect malignant promyelocytes from the apoptotic effect of ATO along with evaluation of dysregulated genes and pathways are required. Disclosure: No relevant conflicts of interest to declare.

Description: Abstract 2785 Targeted therapy in CML with tyrosine kinase inhibitor (TKI) imatinib has resulted in significant improvement in outcome. However, resistance and intolerance to imatinib have seriously limited the success of this therapy, with only a proportion of patients achieving major molecular response (MMR). It is important to identify predictors of response to imatinib, since early switch to second generation TKI has been shown to improve outcome in non-responders. Overexpression of efflux transporters, decreased expression of influx transporters and polymorphisms in these transporters and drug metabolizing enzymes have been shown to influence imatinib therapy. We evaluated the role of RNA expression and polymorphisms in imatinib influx and efflux transporters in 86 newly diagnosed patients with CML at our centre receiving imatinib from December 2009 till April 2012. Total RNA was extracted and RNA expression of the transporter genes ABCA3 (sequesters imatinib; increased expression shown to decrease imatinib exposure), ABCB1, ABCG2 (both efflux transporters) and SLC22A1 (influx transporter) were analyzed with GAPDH as housekeeping gene using Taqman based gene expression assays. The dCT values (where dCT = CT value of target gene – CT value of housekeeping gene; lower dCT corresponds to higher expression and vice versa) were used for comparing expression between groups. Polymorphisms in ABCB1 (T1236C, T2677A, and C3435T), ABCG2 (Ex2 G34A and Ex5 C421A, promoter SNP -15994C-〉T), SLC22A1 (all coding exons) that were associated with the expression of these transporters were analyzed from genomic DNA samples by PCR followed by direct sequencing or RFLP. BCR-ABL transcript levels were monitored at diagnosis, 3, 6, 9, 12, 18, 24 and 30 months after start of imatinib. Molecular response to imatinib after a minimum of 12 months was calculated using the ratio of BCR-ABL to ABL transcript expressed in International scale. Patients were classified to have MMR (BCR-ABL/ABL 0.01; but complete or partial cytogenetic response [CCR or PCR] based on FISH analysis at around 12 months from the start of imatinib therapy), and resistant (BCR-ABL/ABL 〉10; no cytogenetic response even after imatinib dose was increased), or intolerant (severe cytopenia or skin toxicity requiring frequent dose reductions). The RNA expression was compared between different groups by ANOVA and the effect of SNPs on outcome was compared by Fisher's test. Out of the 86 patients, 34 achieved MMR, 22 did not achieve MMR; 14 were resistant and 12 were intolerant to imatinib; 4 patients progressed. RNA expression of ABCB1, ABCG2, ABCA3 and SLC22A1 showed wide variation among CML patients. RNA expression of SLC22A1 was significantly higher in those with MMR compared to no MMR; but not significantly different from resistance or intolerant groups (Fig 1a). ABCA3, ABCG2 and ABCB1 expression, though lower in the MMR group compared to no MMR, did not reach statistical significance. We then compared the ratio of influx to efflux transporters: dCThOCT1/dCTABCB1*dCTABCG2, referred to as resistance index (RI), in these groups; patients with MMR had significantly lower RI compared to those who did not achieve MMR (Fig 1b). We then compared the influence of SNPs in candidate genes with molecular response to imatinib. There was significantly increased frequency of variant alleles of MDR1 C3435T and T1236C in those who achieved MMR compared to those who did not achieve MMR/ resistant (p=0.03 and 0.01 respectively). We identified nine coding nonsynonymous variants including a novel exon2 variant Gly-165-Cys and 2 intronic variants in SLC22A1 upon sequencing. Among these SNPs, Pro341Leu in exon6 (p=0.015) and –TGA del polymorphism resulting in frame shift in exon 7 (p value=0.02) were significantly associated with decreased molecular response. None of the screened SNPs were particularly different in the intolerant patients. In conclusion, we show for the first time a predictive model with RNA expression pattern and genotype that could identify CML patients who would achieve MMR on imatinib therapy. The group that did not achieve MMR, but partial or complete cytogenetic response, could be made to achieve MMR by increasing the dose of imatinib or modulating the candidate ABC transporter expression. However, this model could not differentiate patients who develop therapy resistance or intolerance, which needs to be evaluated further. Disclosures: No relevant conflicts of interest to declare.