ALBERT

Description: Abstract 4478 Allogeneic stem cell transplantation (AlloSCT) is the treatment of choice in advanced chronic myelogenous leukemia (CML) and Ph+ acute lymphoblastic leukemia (ALL). However, post transplant relapse rate is high and outcome is often poor in this setting. Reduction of tumor mass pre-transplant and maintenance therapy post alloSCT, may improve response rate and reduce relapse rate. We speculated that the second-generation tyrosine-kinase inhibitor (TKI) Nilotinib (Tasigna, Novartis Pharmaceuticals) would be effective in achieving these goals. In the current study Nilotinib, was administered as maintenance therapy post alloSCT in patients (pts) with advanced CML and Ph+ ALL (study CAMN107AIL03T). However, TKIs have been demonstrated in previously published literature to affect T cells proliferation and signal transduction and to potentiate LGL and NK cell activity. Furthermore, in recent studies TKIs have also been demonstrated to ameliorate chronic GVHD. We therefore assessed immunological reconstitution and function including flow analysis of lymphocyte subsets, T-mitogenic response to αCD3 and PHA, thymic activity as determined by the quantification of the T cell receptor excision circles (TREC), TCR repertoire and NK cells cytotoxic activity against K562 cell line. In all, the study included 24 pts. Patients engrafted in a median day +15 (range, 10–38) with 100% donor chimerism. Acute GVHD grade 3/4 was reported in 3 pts (14%) and the rate of extensive chronic GVHD at last follow up was 50%. At 6 months after alloSCT, 11 of 15 pts with advanced CML had attained CCyR, 11 of 15 pts with advanced CML had attained a MMR or better, and 5 of 7 pts with Ph+ ALL attained a CR. The median OS was 16 months, with predicted 1- and 2- year rates of 55% (95% CI, 32% – 72%) and 50% (95% CI, 28% – 68%), respectively. The median PFS was 11 months, with predicted 1- and 2- year rates of 50% (95% CI, 28% – 68%) and 38% (95% CI, 17% – 59%), respectively. Immunological testing was performed pre- and post Nilotinib maintenance therapy in 12 pts (advanced CML-8, Ph+ ALL-4) who received Nilotinib for at least 90 days following alloSCT. The median age was 34.5 years (range, 21–57) and 75% were males. Six pts underwent alloSCT from an HLA-matched sibling donor, 4 from matched unrelated and 2 from an alternative donor (cord blood-1, haploidentical-1). All had myeloablative conditioning. GVHD prophylaxis included CSA and MMF. The relative percentage of T- lymphocyte subsets (assessed by FACS) and total lymphocytes number were stable during Nilotinib maintenance administration after alloSCT, while a 7.8±2.5 fold increase in B cells was noted. T cell mitogenic response with αCD3 and PHA (stimulation index ratio) was sustained (2.5±1.0, vs. 2.8±1.05 and 3.3±1.3 vs. 5.3±2.9 stimulation, pre- and post Nilotinib therapy, respectively). Mean thymic output determined by TREC quantification pre-, during and post Nilotinib administration was 81.8±108, 81.2±90.3 and 142.8±197.4 copies per 0.5ug DNA indicating continuous thymopoiesis. Similarly, no significant change of the TCR repertoire was observed during Nilotinib treatment. Specifically, normal expression of the TCR repertoire was detected in 15.1±5.5 and 15.3±5.6 of the examined TCRs, clonal expression was detected in 2.5±2.2 and 2.9±3 of the examined receptors, while reduced expression was detected only in 6.4±4.3 and 5.8±4.5 of the examined receptors pre-and post Nilotinib treatment, respectively. NK cytotoxic activity against K562 expressed as fold of change from baseline, also remained stable during Nilotinib treatment (2.8±1.1 and 2.3±0.8, respectively). In summary, Nilotinib maintenance therapy post alloSCT in pts with advanced CML and Ph+ALL did not interfere or jeopardized immune reconstitution and function including the number of immune cell subsets, T cell mitogenic response, TCR repertoire, thymic output and NK cytolytic activity post alloSCT. Based on this immunological data we would further recommend Nilotinib maintenance therapy post alloSCT in pts with advanced CML and Ph+ALL. Disclosures: Nagler: Novartis: Honoraria, Research Funding.

Description: Rates of obesity have substantially increased in recent years. Pharmacokinetics of drugs including chemotherapy is different in obese patients due to alteration in the clearance and volume of distribution. Thus, appropriate chemotherapy dosing for obese patients with malignant diseases is a significant issue. Limiting chemotherapy doses in overweight and obese patients may negatively influence the outcomes in these patients. ASCO has recently published clinical practice guidelines for conventional chemotherapy dosing for obese patients with cancer indicating that up to 40% of obese patients received reduced chemotherapy doses that are not based on actual body weight (ABW) [Grigg A, JCO 2012]. Concerns about toxicity or overdosing in obese patients, based on the use of ABW, are unfounded. Moreover, there is a paucity of information addressing the pharmacokinetics of high dose chemotherapy in obese patients undergoing hematopoietic stem cell transplantation (HSCT). A rather small international survey of drug dosing schemes among transplant centers revealed that there is no consensus regarding appropriate dose adjustment for obese patients [Grigg A, Leuk Lymphoma 1997]. Also, there is limited data on outcomes in obese versus non-obese patients in various small retrospective studies. For this reason, the ALWP of the EBMT constructed an electronic survey for assessing current practice of dose adjustment of chemotherapy in patients undergoing HSCT, in transplant centers and for planning retrospective analysis and prospective studies in the future. Fifty six EBMT centers from 27 countries filled the online survey. Among the 56 centers, the percentage of obese patients was less than 10% in 22 centers (40%), between 10 to 19% in 23 centers (42%) and more than 20% in 10 centers (17%). Forty five centers declared they adjust chemotherapy dose for obese patients (80.5%) and only 11 (19.5%) declared they do not adjust dose. Among centers which adjust dose, most uses BMI as the parameter for defining obesity (28 centers, 62%), others use percentage over the actual body weight (ABW) as the basis for defining obesity (11 centers, 24.5%), both BMI and ABW (3 centers, 6.7%) or other parameter (3 centers, 6.7%). Most of the centers that use BMI for adjusting dose define BMI 〉 30 kg/m2 as the cut-off value (formal definition for obesity), only one center uses morbid obesity (BMI 〉 40 kg/m2), and the remainder uses other cut-off values. Among 11 centers who use ABW, 9 use ABW more than 120% of ideal body weight for adjustment. Eighty four percents of the centers use one level of obesity for adjustment while the rest uses 2 levels. The method for determining the weight for chemotherapy calculation was actual body weight (ABW) in 16 centers, ideal body weight (IBW) in 10 centers, IBW + 25% of difference between IBW and ABW (IBW+0.25*(ABW-IBW)) in 16 centers and other methods in the rest. Among centers that use dose adjustment, 44% also cap the dose at 2 m2 for chemotherapy dose based on BSA while 56% do not cap. On the contrary, most of the centers (9/11) that do not adjust dose for weight also do not cap the BSA at 2 m2. Seventy nine percents of responding centers use the same approach to dose adjustments for myeloablative, reduced intensity (RIC) or non myeloablative (NMA) conditioning, while 21% reduce the dose less for RIC or NMA conditioning. For Busulfan dose only 7 centers monitor pharmacokinetics (pk). Eleven centers use ideal body weight for calculation, 17 centers use actual weight and 18 centers correct weight according to percentage over actual body weight. Conclusion This EBMT survey reveal large diversity among transplant centers regarding dose adjustment practice for high dose conditioning chemotherapy. Most of the EBMT centers use dose adjustment for obese patients and about half of them also cap BSA at 2 m2, while capping is uncommon in the centers that do not adjust dose. Thus, the range of the final dose is very wide. Even for Busulfan where dose is calculated normally according to ideal body weight, the diversity of dose given for obese patients is wide. Our next step is to analyze outcomes of transplantation according to dose adjustment practice and subsequently to formulate a methodology for future prospective studies. Disclosures: No relevant conflicts of interest to declare.

Description: Abstract 2249 Poster Board II-226 BACKGROUND. Allogeneic hematopoietic stem cell transplantation (HSCT) is being used to treat a range malignant and nonmalignant conditions. However, it often causes potentially lethal Graft-versus-Host Disease (GVHD). Several studies revealed that mesenchymal stromal cells (MSCs) from human bone marrow can down regulate GVHD after HSCT. METHODS. MSCs were obtained from BM, expanded and characterized by their morphology, immunophenotype, immunosuppressive potential for autologous, partially and fully mismatched lymphocytes. Twenty five patients (pts) got 39 (range, 1 to 4) MSCs infusions for 27 episodes of acute GVHD, which was defined as steroid resistant grade IV aGVHD in all but one patient. RESULTS. GvHD characteristics. Acute GVHD started from day +6 to d +46 from HSCT (median d+18) excluding single pt in whom it occurred on d+150. In 2 pts disease manifested in a hyper acute form before the engraftment (on d+6 and d+8, respectively). In all but one pt GI GVHD was defined as grade IV with full clinical picture of it. Skin GVHD accompanied GI symptoms in 17 pts and 13 was 〉 grade II, in 4 grade IV. Liver symptoms presented in 13 pts, one pt was determinate as VOD. In 3 pts liver involvement was defined as grade IV. In all 24 of 25 pts had GVHD grade IV 4 and one grade III. Previous conventional anti GVHD therapies included: MP in all pts, MTX, various calcineurin inhibitors, MMF, ECP, serotherapy. Only 3 pts showed temporary limited partial response. Treatment with MSCs. The 1st infusion with MSC was given on day +32.5 (range, 8 to 74); d+50 (range, +28 to +180) post diagnosis of aGVHD and HSCT HSCT, respectively. In 22 evaluable pts we treated 24 separate episodes of GVHD: 22 first episodes and in 2 pts relapse of GVHD. In 24 of 39 cases treatment was performed with frozen MSC and in 15 with fresh cultured cells. In 37/39 cases passages 1 to 3 were used. Median number of infused cells was 1 (range, 0.3 to 3.1) x10 6 per kg of pt body weight in each treatment. Initial response was seen in 17/24 episodes of aGVHD (70.8%), in 8 partial (PR) and in 9 complete (CR). Two pts experienced GVHD relapse after achieving CR and very good PR: one case was treated successfully while another was resistant to MSCs. The latest pt with grade IV aggressive liver GVHD received MSCs injections intra hepatic arteries under radiological control following IRB approval with no anti GVHD effect. The procedure was uneventful with no evidence of microembolisation, no changes of blood flow or deterioration of liver tests (cytolysis). We observed a trend for better effect with higher MSC cell dose: total and per first infusion MSCs dose (1.93±1.28 vs. 1.23±0.50, p=0.078; 1.28±0.79 vs. 0.88±0.28, p=0.086). In all cases effect was seen after the first procedure. No difference was noted in the anti GVHD activity between fresh vs. frozen MSCs. There were no immediate or late toxicity or side effects. Overall survival. 14/25 patients are alive up to 20 months follow up. 8/11 pts, died from GVHD (4 within 1-10 d from MSCs infusion), and 3 from other unrelated causes including TTP-1, DAH-1 and disease progression-1. CONCLUSION. Treatment with MSCs seems to be a new novel type of therapy for steroids resistant GVHD with promising preliminary results and low toxicity. Obviously multicenter 2 arm randomized study is in need to confirm these encouraging. Disclosures: No relevant conflicts of interest to declare.

Description: Abstract 2260 Background: BKT140 is a high affinity CXCR4 inhibitor with an extended K off-rate. Pre-clinical studies in animal models with BKT140 showed a robust mobilization of white blood cells (WBC) and hematopoietic stem cells (HSC). Furthermore, BKT140 also showed a direct anti-tumor effect against human-derived multiple myeloma (MM), lymphoma and primary leukemia cells and cell lines in vitro and in vivo, causing significant apoptosis. Aims: To assess BKT140 toxicity (primary endpoints), the mobilization capacity of CD34+ hematopoietic progenitors and CD138 MM cells, and pharmacokinetic (PK) and pharmacodynamic (PD) (secondary endpoints). Methods: 16 MM patients in first CR/PR were included in a phase I/IIa study, in which escalating doses of BKT140 (30, 100,300,900 μg/kg) were administered together with a high-dose cyclophosphamide (Cy) (2 g/m2) and G-CSF (5 μg/Kg) for stem cell mobilization. G-CSF was started on day 5 post Cy and BKT140 was injected subcutaneously (SC) once on day 10. Toxicity, PK, and mobilization capacity (assessed by serial measurements of number of WBC and CD34+ and CD138+ cells) were measured pre- and post BKT140 administration. Results: BKT140 was well tolerated at all doses and none of the patients developed grade II-IV toxicity. BKT140 was rapidly absorbed with no observed lag time, with peak plasma concentrations occurring 0.5h after administration. Clearance was rapid, with a median terminal half-life of 0.69h. BKT140 administration resulted in a significant dose-dependent increase in the number of peripheral blood neutrophils, monocytes, lymphocytes, and CD34+ cells compared to the G-CSF/Cy individual patient baseline. The maximum increase in the number of WBC from baseline was observed within 8h following BKT140 injection, 2.5-, and 3.0-, 4.1- and 4.8-fold, for the 4 BKT140 doses, respectively. Furthermore, BKT140 administration resulted in a significant increase in the mean absolute PB CD34+ cells mobilized (6.6, 7.5, 11.2 and 20.6 ×106/kg) for the 4 BKT140 administered doses, respectively. Moreover, the number of aphaeresis was reduced from 2.25 procedures at the first two BKT140 doses to 1.25 and 1 aphaeresis at the highest BKT140 doses, respectively. An increase in the number of CD138+ cells was observed in 6 out of 6 pts that had CD138+ cells in their blood and were treated with lower doses of BKT140 (30 and 100 μg/kg). Interestingly, in pts that were treated with the highest doses of BKT140 (300 and 900 μg/kg) a reduced number of CD138+ cells was observed in 3 out of 7 pts that had CD138+ cells in their blood, whereas in 4 pts, an increase in the number of CD138+ cells was shown. Three pts who did not have CD138+ cells in their blood were not affected by BKT140. The BKT140 mobilized grafts were used for AutoSCT following 200 mg/m2 melphalan conditioning. Pts received an average of 5.3×106 CD34+ cells/kg. All transplanted pts rapidly engrafted (n=15). The median day for neutrophil (〉500/mm3) and platelet (〉20,000/mm3, 〉50,000/mm3,) was on day 11 (range, 0–13), day 11 (range, 0–14), and day 14 (range, 0–23), respectively. Conclusions: The current data suggests that BKT140 can safely be added to G-CSF-based harvesting regimens, can increase CD34+ cell mobilization and reduce the number of collection days. Furthermore, due to its ability to release MM cells from the bone marrow and stimulate their cell death, additional studies are warranted to further evaluate the effect of BKT140 as an anti-MM agent. Disclosures: Nagler: Biokine Therapeutics Ltd: Consultancy. Abraham: Biokine Therapeutics Ltd: Employment, Equity Ownership, Patents & Royalties. Wald: Biokine Therapeutics Ltd: Employment. Shaw: Biokine Therapeutics Ltd: Consultancy, Equity Ownership, Membership on an entity's Board of Directors or advisory committees. Eizenberg: Biokine Therapeutics Ltd: Employment, Equity Ownership, Patents & Royalties. Peled: Biokine Therapeutics Ltd: Consultancy, Equity Ownership, Membership on an entity's Board of Directors or advisory committees, Patents & Royalties, Research Funding.

Description: B-chronic lymphocytic leukemia (BCLL) is a lymphoproliferative disease that is characterized by clonal expansion of CD5+ B cells. BCLL is associated with secondary immunodeficiency and hypogammaglobulinemia. It has been suggested that T-cell dysregulation may play a role in the hypogammaglobulinemia and in the increased incidence of autoimmunity in BCLL patients. We attempted to transfer human peripheral blood mononuclear cells (PBMC) from BCLL patients in different stages of the disease into immunodeficient mice. PBMC from BCLL patients in stage 0, stages I to II, and stages III to IV were transplanted into the peritoneal cavity of lethally irradiated Balb/c or beige/nude/Xid (BNX) mice radioprotected with bone marrow (BM) from severe combined immunodeficiency (SCID) mice. Different engraftment profiles were found in the chimeric mice 2 weeks after transplantation of PBMC according to the disease stage of the BCLL donors. Infusion of PBMC from donors in stage 0 led to marked engraftment of human T cells, whereas the human tumor cells could hardly be detected. In contrast, chimeric mice receiving PBMC from patients in stage III to IV disease exhibited engraftment with a dominance of tumor cells, compared with a miniscule level of T cells. The ability of the engrafted cells to produce human Ig was also found to be correlated with the disease stage of the donor, although all donors had the same magnitude of hypogammaglobulinemia. Total human Ig production in the chimeric mice was normal in mice receiving PBMC from donors in stage 0, whereas in chimeric mice engrafted with PBMC from donors in stages III to IV almost no human Igs could be detected. This differential reconstitution of antibody production in the mouse model according to the stage of the patient's disease will allow further studies on possible cellular interactions between malignant and immune cells in BCLL.

Description: Background: The fully human IgG4 anti-PD-1 monoclonal antibody, nivolumab, has shown to be highly efficacious in relapsed and refractory (RR) classical Hodgkin lymphoma (cHL). Initial results of a single-arm nivolumab monotherapy in heavily pretreated young adults with RR cHL, previously treated with autologous hematopoetic stem cell transplantation (autoSCT) and brentuximab vedotin are promising. However, demonstrating similar results in real-life experience in patients not on Pharma clinical studies are of major clinical importance. Aim: The aim of the current study was to evaluate the efficacy and safety of nivolumab in previously heavily pretreated patients with RR cHL not on Pharma clinical studies. Methods: The records of patients with RR cHL who were treated with nivolumab in seven medical centers were reviewed. The regimen consisted of nivolumab 3 mg/kg once every 2 weeks until disease progression or unacceptable toxicity. Results: Between May 2015 and April 2016, 30 patients with RR cHL received nivolumab monotherapy. The median age was 35 years (range: 21 to 88). Patients were extremely heavily pretreated with median of 6 prior therapies (range: 3 to 11); 73% of them received ≥ 5 lines of therapy. Specifically, 29/30 (97%) received ABVD, 7/30 (23%) escBEACOPP, 20/30 (67%) gemcitabine based regimens, 30/30 (100%) brentuximab vedotin, 8/30 (27%) bendamustine, 3/30 (10%) lenalidomide, 28/30 (93%) other combinations of chemotherapy and 22/30 (73%) radiotherapy. Twenty patients (67%) previously underwent autoHSCT and 4/30 (13%) allogeneic transplantation (alloSCT), two before and two after receiving nivolumab. Median number of nivolumab doses received was 8 (range 1 to 29). Twenty three patients (77%) are still on treatment. Seven patients (23%) discontinued therapy due to disease progression 3/30 (10%), adverse events 2/30 (7%) and referral to alloSCT 2/30 (7%). Overall response rate among 25 evaluable patients was 76%, complete response (CR) - 7/25 (28%), partial response - 12/25 (48%), stable disease - 3/25 (12%) and only 3/25 (12%) progressed, one of them with cHL type of Richter's transformation. Two patients underwent alloSCT in remission following nivolumab treatment. One patient was transplanted from a matched sibling donor and the second patient from a T replete haploidentical donor with post transplantation cyclophosphamide as anti-graft versus host disease prophylaxis. Both patients are in CR, ten and three months post transplantation, respectively. The median overall survival (OS) and progression free survival (PFS) for the cohort were not reached (Figure 1). At 24 weeks the PFS was 74%. Adverse events reported in 9/30 patients (30%) were usually mild. Severe nivolumab related adverse events in 3/30 (10%) included immune related pneumonitis, myelitis and uveitis. No treatment related death occurred. Conclusions: The check point inhibitor nivolumab is a novel therapy for RR cHL patients, previously with unmet need following relapse or refractory to autoHSCT and brentuximab vedotin. Our real life experience confirms and highlights the efficacy and safety of nivolumab in very heavily pretreated young and elderly patients with cHL eligible as well as ineligible for autologous HSCT. Figure 1 Figure 1. Disclosures No relevant conflicts of interest to declare.

Description: Allogeneic Hematopoietic SCT (HSCT) is a curative treatment for high risk AML. However, search continues for novel conditioning regimens that will reduce SCT-related toxicity while retaining maximal anti-leukemia effect. Treosulfan (L-threitol-1, 4-bis-methanesulfonate; dihydroxybusulfan) was shown in preclinical studies to have a myeloablative effect on committed and non-committed stem cells. In addition, it has potent immunosuppressive characteristics, which makes it an attractive candidate for use in conditioning regimens before HSCT. We therefore preformed a retrospective analysis of the data for HSCT with Treosulfan based conditioning using the ALWP data set. We analyzed 795 HSCT in which Treosulfan was used in the pre-transplantation conditioning in adult patients with AML that were transplanted between Jan 2000 and Dec 2011. Out of the 795 AML patients 375 were transplanted in CR1, 197 in CR2 and 223 with active disease. 258 (32%) were transplanted from HLA matched sibling donors, 465 (58%) from unrelated donor and 72 (9%) from family mismatch donor. Mobilized peripheral blood stem cells (PBSC) were used as the graft source for 700 (89%), and bone marrow (BM) for 95 (12%) of the transplants. Conditioning was myeloablative (MAC) in 505 (64%), and reduced toxicity regiments (RTC) in 290 (36%) of the HSCT. Most frequently the Treosulfan based conditioning was combined with Fludarabine (Flu/Treo) and was used in 67% of the MAC and 86% of RTC. In 96 transplants Treosulfan was combined with Cyclophosphamid (Cy) and Etoposide (VP16). Other combinations included Melphalan (n=56), ARA-C (n=19) and Clofarabin (n=7). The median follow-up was 14 mo (1-116). Engraftment was achieved in 96% of the HSCT with no difference between MAC and RTC 95% and 97%, respectively. NRM were 17±4%, 21±2% and 33±3% for patients in CR1, CR2 and advance disease, respectively. Acute GVHD ≥ II occurred in 21% of the HSCT and was the same for MAC and RTC HSCT, while 2 year incidence of chronic GVHD (2y) was seen in 43±2% of HSCT (MAC-45%, RTC-38%). Incidences of relapse were 31±2%, 36±4% and 43±4% for patients transplanted in CR1, CR2 and advance disease, respectively. 2y OS and DFS were 61±3% and 51±3% for patients in CR1, 53±4% and 43±4% in patients with CR2 and 32±4% and 24±3% in patients with advanced disease. LFS was similar for MAC and RTC HSCT 60±5% and 50±8% for patients transplanted in CR1 from sibling donors while 43±6% and 50±6% for those transplanted from unrelated donors, respectively. Relapse rate were also similar 30±4% and 39±8% vs. 35±6% and 28±6% for transplants from siblings and unrelated donors, respectively. While NRM was 9±3% and 11±5% vs. 22±4% and 21±5% for HSCT from siblings vs. unrelated with MAC vs. RTC, respectively. In conclusion: Treosulfan mostly in combination with Fludarabin for conditioning prior to HSCT is associated with good engraftment, acceptable rates of NRM, GVHD, OS and LFS for patients in both CR1 and CR2. Further studies analyzing the use of Treosulfan in pre-HSCT conditioning regiments are in need. Disclosures: No relevant conflicts of interest to declare.

Description: Abstract 518 Allogeneic stem cell transplantation (SCT) is a potentially curative therapy for patients (pts) with AML and MDS. SCT is associated with substantial mortality during the first 2 years after treatment due to relapse and non-relapse causes, whereas after 2 years survival curves often reach a plateau. However, late mortality and late events continue to cause treatment failures through the late post-transplant course. The pattern of late events has been reported following myeloablative conditioning (MAC) but is not well defined in the RIC setting. To explore late outcomes we retrospectively analyzed SCT results in a cohort of 298 pts with AML/MDS given allogeneic SCT with various regimens. As a general role, pts meeting standard eligibility criteria were given MAC, consisting of high dose intravenous busulfan (Bu) and cyclophosphamide (BuCy). Pts considered at excessive risk for non-relapse mortality (NRM), mostly due to advanced age, were given a reduced toxicity regimen such as fludarabine with high-dose Bu (FB4) or treosulfan (FT) or a RIC regimen of fludarabine and reduced doses of Bu (FB2). The 2-year and 5-year overall survival (OS) for this cohort was 48% (95CI, 42-54) and 38% (95CI, 31-44), respectively. We identified 93 pts who were alive and leukemia-free 2 years after SCT. The median age at SCT was 52 years (17-72), 52 male, 41 female. The donors were HLA-matched siblings (n=57) or unrelated donors (n=36. The conditioning regimen was FB2 (n=28), FB4 (n=21, FT (n=15) or BuCy (n=29). At the 2-year time-point, 28 pts had a history of acute GVHD, 68 had a history of chronic GVHD, 43 had active chronic GVHD which still required immune suppressive therapy (IST) in 37. Among pts surviving leukemia-free 2 years after SCT, the probability of remaining alive and leukemia-free, for the next 5 years, was 81% (95CI, 71-91) and 79% (95CI, 69-88), respectively. The probability of OS was 75%, 84%, 92% and 79%, after the various regimens, respectively (p=NS). There were 14 deaths beyond 2 years, 8 deaths due to relapse and 6 due to NRM. NRM included 3 deaths due to solid cancers, 2 due to breast cancer (both in pts transplanted for therapy related AML relapsing with the primary malignancy) and one due to colon cancer. There were 3 additional secondary cancers; breast cancer, squamous cell skin cancer and Kaposi sarcoma (1 pt each; all currently alive). Two pts died of myocardial infarction and one of chronic GVHD. In all, the cumulative incidence of late NRM was 10% (4-24). Interestingly, unlike the early post-transplant period, with current supportive care regimens, late deaths from GVHD and infections are rare. Twelve pts relapsed, 24-59 months after SCT, cumulative incidence 12% (6-23); 8 died, 4 are alive following further therapies, 3 long-term. The kinetics of late relapses was similar with MAC and RIC. Advanced age (〉50) was the most significant predicting factor for shortened survival. OS was 64% and 92%, in the older and younger groups, respectively (p=0.03). All deaths due to NRM, and all secondary cancers occurred in the older group (p=0.01). Multivariate analysis (MVA) identified advanced age as the only independent predicting factor for OS; HR 3.1 (1-11.3). The conditioning regimen, disease status at SCT and donor type were not predictive. A history of acute or chronic GVHD predicted improved leukemia-free survival. Age was the most important predicting factor for NRM Active chronic GVHD was the most important factor predicting for reduced relapse risk. The cumulative probability of stopping IST by 7 years post SCT, for the entire cohort, was 73% (63-82). This probability was related to the conditioning regimen and was 80%,76%,80% and 63%, after FB2,FB4,FT and BuCy, respectively (p=0.06 for BuCy versus others). MVA identified conditioning with BuCy (HR 1.8, p=0.06) and male gender (HR 1.7, p=0.04) as predicting for prolonged need for IST. Among the 37 pts who were still on IST 2 years after SCT, the cumulative probability of stopping IST during the following 5 years was only 35%. It was higher in younger pts (59% Vs 15%, p=0.05) and in pts given FB2 (42% Vs. 31%, p=NS). In conclusion the pattern of late outcome is similar after MAC and RIC. Younger pts with AML/MDS who are leukemia-free 2 years after SCT can expect a very good outcome, while older pts are at higher risk for NRM and second cancers, which may not always relate directly to SCT. Pts given RIC are more likely to be able to stop IST, which may be associated with a better quality of life. Disclosures: No relevant conflicts of interest to declare.