ALBERT

Description: Background: Over the past two decades, peripheral blood stem cells (PBSC) have surpassed bone marrow as the preferred graft source for adult allogeneic transplantation due to its more rapid engraftment and potentially better graft-vs-tumor effects, and because PBSC collection is much less invasive for the donor. The optimal CD34+ PBSC dose is ≥ 4.0x106cells/kg, but doses ≥ 8.0x106cells/kg are suggested by some for reduced-intensity conditioning and haploidentical transplants. There is no established minimum CD34+ PBSC dose, but doses below 2.0x106cells/kg have been associated with a higher risk of engraftment delay and failure. There is significant inter-donor variability in the ability to mobilize PBSCs. Several factors have been identified as predictors of PBSC mobilization in healthy donors including: gender, age, weight, body mass index (BMI), and blood counts before and after mobilization. The impact of the donor’s comorbidities on mobilization is currently unknown. Patients/Methods: We performed retrospective chart review of 488 consecutive adult patients who underwent apheresis for allogeneic stem cell donation at Washington University School of Medicine from 2006 through 2013. Patients who received any collection regimen other than 10mcg/kg of G-CSF daily with 20 liters (+/-10%) apheresis on Day 5 were excluded. Patients who had undergone a previous apheresis for stem cell donation were excluded. Univariate analysis was performed to identify predictors of CD34+ PBSC collection in a single apheresis. Variables analyzed were: gender; age; weight; BMI; donor-to-recipient weight ratio; pre and post-mobilization blood counts (white blood count [WBC], hematocrit, platelets, neutrophils, lymphocytes, and monocytes); pre-mobilization glucose and triglyceride levels; post-mobilization peripheral blood (PB) CD34+count; and medical history significant for hypertension, hyperlipidemia, or diabetes mellitus. Subsequently, a linear regression multivariate analysis was performed with all variables found to be significant in the univariate analysis. 2-tailed tests for significance were used throughout the analysis. Results: 304 patients met the eligibility criteria for analysis. The median age was 53 years (range 18-76), 90% were Caucasian, and 50% were male. The median number of CD34+ cells collected was 7.4 x106/kg (range 0.8-27.1). 97% (295) collected ≥ 2.0x106 CD34+cells/kg, 81% (247) collected ≥ 4.0x106 CD34+cells/kg, and 44% (134) collected ≥ 8.0x106 CD34+ cells/kg. Post-mobilization PB CD34+ count (r= 0.841, p

Description: Background: Induction chemoimmunotherapy followed by autologous stem cell transplantation (ASCT) is a standard of care for transplant eligible (TE) patients (pts) with untreated mantle cell lymphoma (uMCL); however, there is no consensus on the optimal induction regimen. The addition of rituximab + high-dose cytarabine (RC) to an RCHOP-like regimen is associated with better outcomes. In addition, two randomized trials have demonstrated superior efficacy and tolerability for rituximab/bendamustine (RB) compared to RCHOP for uMCL. Based on this, we conducted a phase 2 trial of 3 cycles of RB followed by 3 cycles of RC in 23 TE pts with uMCL, with encouraging preliminary results (Armand, BJH 2016). Pts continued to be followed for relapse and survival. Meanwhile, RB/RC became the standard frontline regimen at Dana-Farber Cancer Institute (DFCI). Simultaneously, investigators at Washington University in St. Louis (WUSTL) initiated a similar study of alternating cycles of RB/RC for uMCL. Herein, we report the results of both phase 2 trials as well as the off-trial experience. Methods: In the DFCI trial, TE pts (age 18-69) with uMCL were treated with 3 cycles of RB (R 375 mg/m2 d1, B 90 mg/m2 d1-2) followed by 3 cycles of RC (R 375 mg/m2 d1, C 3gm/m2 BID d1-2 with dose reductions for age, renal dysfunction, or pre-existing neurotoxicity). Off-trial pts treated with RB/RC at DFCI or in consulting community practices were retrospectively identified using clinical and pharmacy databases. In the WUSTL trial, TE pts (age 18-65) received alternating cycles of RB (cycles 1, 3, 5) and RC (cycles 2, 4, 6) (same dosing as above). Response assessments were made using CT scans for the DFCI trial and PET/CT for the WUSTL trial and DFCI off-trial pts. Results: In total, 86 pts (23 DFCI trial, 49 DFCI off-trial, 14 WUSTL trial) were treated with RB/RC. The median age was 57 (range 30-72). Pts in the WUSTL cohort were more likely to be male, have a high MIPI score, and have blastoid variant (Table). 94% of pts completed 6 cycles of RB/RC therapy. Off-trial pts were more likely to receive a lower starting dose (≤ 2gm/m2) of cytarabine (76%) compared to trial pts (38%). At the EOI, the overall response rate and CRR were 98% and 92%, respectively, with similar response rates across cohorts (Table). 73 pts (85%) subsequently underwent ASCT and 4 additional pts (5%) have ASCTs planned. 9 pts did not undergo ASCT because of persistent or PD (n=3), prolonged cytopenias (n=3), an incidentally identified ASXL1 mutation without cytopenias (n=1), pt preference (n=1), and inadequate stem cell collection (n=1). Delayed platelet engraftment after ASCT was seen for pts receiving alternating cycles of RB/RC compared to sequential RB/RC at day 30 (plts

Description: We sought to determine whether bexarotene can be combined with decitabine in elderly and relapsed AML patients. Both drugs have been shown to be well tolerated in acute myeloid leukemia (AML) patients as single agents, and these agents have non-overlapping mechanisms and side-effect profiles; bexarotene activates transcriptional effects of RXRA through hetero- and homodimers, while decitabine is thought to act through DNA hypomethylation. Furthermore, through Affymetrix expression array profiling of 111 AML patients and Nanostring analysis of 7 MDS and AML patients, we observed consistently elevated levels of RXRA relative to RARA, suggesting that a ligand specific for RXR may be more effective to induce AML differentiation than the RARA ligand ATRA. We treated 18 elderly (≥ 60 years old) or relapsed AML patients in 3+3 dose escalating bexarotene cohorts: 100 mg/m2/day, 200 mg/m2/day, 300 mg/m2/day. All patients were treated with decitabine 20 mg/m2IV on days 1-5 of 28 day cycles. All patients were monitored for hypertriglyceridemia and hypothyroidism, and treated accordingly. The average age was 73, the average performance status was 1, an adverse karyotype was observed in 9 patients, and 12 patients had relapsed after prior therapy. Only one patient experienced a dose limiting toxicity (grade 3 fatigue) and 8 patients were treated with the maximum dose (myelosuppression, infection, differentiation syndrome, hypertriglyceridemia, hyperlipidemia, hypothyroidism, nausea, weight loss and reversible electrolyte abnormalities were not considered dose limiting). The overall response rate was 22%: 1 patient achieved complete remission with incomplete count recovery (CRi) and 3 patients achieved blast reduction greater than 50% (partial response, PR). In addition, six patients achieved stable disease (SD). Patients with CRi, PR, or SD completed an average of 4.25 cycles, while other patients completed an average of 1.2 cycles. Of note, 3 patients successfully transitioned to allogeneic transplant following therapy (average age 68). We correlated ex vivo bexarotene sensitivity with clinical response. Bone marrow cells were collected on day 0 and day 3 of bexarotene therapy (during cycle 1, decitabine was administered on day 3 after bone marrow collection) and co-cultured with irradiated MS5 murine stromal cells for 72hrs with or without further bexarotene treatment. We used flow cytometry to compare CD11b expression in cells treated with and without bexarotene ex vivo, and compared expression between samples collected on day 0 vs day 3 (in vivo treatment). Bexarotene increased CD11b expression greater in the 4 responding patients vs non-responders (fold increase in CD11b: ex vivo average 2.1 ± 0.3 vs 1.1 ± 0.1 fold, p 〈 0.003; and in vivo 1.6 ± 0.3 vs 0.7 ± 0.2 fold, p 〈 0.03; increase in absolute percentage of CD11b+ cells: ex vivo average 24% ± 2.6% vs 0.7% ± 1%, p 〈 0.001; and in vivo 13.6% ± 4% vs -3.6% ± 2.2%, p 〈 0.002). Furthermore, all 4 responding patients demonstrated an equivalent or increased induction of CD11b when treated ex vivo with ATRA compared with bexarotene. These results show that bexarotene, a retinoid which selectively binds to and activates RXRs, but not RARs, can be safely combined with decitabine in relapsed and refractory AML patients. This combination leads to partial response in a subset of patients, is well tolerated, and can bridge elderly patients to allogeneic transplant. Because ex vivo bexarotene treatment identified all patients achieving a PR, further studies should focus on patients who display ex vivo sensitivity. Finally, the mechanism of RXRA-activated differentiation is likely to be through the RXRA/RARA heterodimer, as all 4 patients who responded to bexarotene also responded to ATRA when tested ex vivo. Disclosures: Welch: Eisai: Research Funding. Off Label Use: Bexarotene for the treatment of AML. Abboud:Ariad, Alexion, Novartis, Teva: Honoraria, Speakers Bureau. Stockerl-Goldstein:Celgene : Speakers Bureau; Millennium: Speakers Bureau.

Description: Background: High-dose chemotherapy with autologous stem cell transplantation (ASCT) is a widely used treatment strategy in lymphoma and myeloma; however, no standard approach for the mobilization of peripheral hematologic stem and progenitor cells (HSPCs) has been established. Levels of circulating CD34+ cells, a surrogate marker for mobilization efficiency, vary widely between pts, and may be influenced by disease state, prior therapy, and/or mobilization regimen. Methods: The Washington University (St. Louis, MO) transplantation database includes clinical parameters from 407 multiple myeloma (MM), 562 non-Hodgkin’s Lymphoma (NHL), and 164 Hodgkin’s disease (HD) pts who received an ASCT between 1995 and 2006. A retrospective analysis of this large (1133 pts) population was conducted to determine factors associated with mobilization efficiency. Mobilization failure was defined as collection of 〈 2 × 10^6 CD34+ cells/kg within 5 apheresis days. Statistical analysis included analysis of variance (ANOVA) with Scheffe Test to determine differences in mobilization between the various mobilization regimens (G-CSF, G-CSF/chemotherapy, G-/GM-CSF, G-CSF/AMD3100). Results: All pts were included in the analysis; 87% received G-CSF alone as the initial mobilization regimen. Mobilization failure rates are summarized in Table 1. NHL and HD pts had an approx. 4-fold higher failure rate than MM pts. The combination of G-CSF with chemotherapy increased the median CD34+ yield compared to G-CSF alone, although no obvious impact on the failure rate was noted in this relatively small group of pts. Remobilization was associated with high failure rates in NHL (79.2%), HD (77.1%), and MM (73.3%). Pooled collections were

Description: Background: No standard approach for the mobilization of peripheral hematologic stem and progenitor cells (HSPCs) has been established. High levels of circulating CD34+ cells, a surrogate marker for mobilization efficiency, are associated with less apheresis days. A higher dose of CD34+ cell transfused after high-dose chemotherapy decreases time to hematologic recovery. Consequently, a better understanding of variables associated with mobilization kinetics may further optimize stem cell collection and reduce complications associated with autologous stem cell transplants. Methods: The Washington University (St. Louis, MO) transplantation database includes clinical parameters from 407 multiple myeloma (MM), 567 non-Hodgkin’s Lymphoma (NHL), and 164 Hodgkin’s disease (HD) pts who received an ASCT between 1995 and 2006. A retrospective analysis of this large pt population was conducted to determine factors associated with the mobilization kinetics of CD34+ cells. Results: Figure 1 summarizes the mobilization kinetics as defined by number of days to reach a target of 2 × 10^6 CD34+ cells/kg. Overall, the median number of aphereses to reach the target were 1, 2, and 2 in MM, NHL, and HD, respectively. Daily median CD34+ yields in MM pts were 3.8, 1.2, and 0.5 × 10^6 on day 1–3, respectively. In NHL pts, yields were 1.4, 0.8, and 0.4 × 10^6 on day 1–3. In HD pts, yields were 1.8, 0.8, and 0.3 × 10^6 on day 1–3, respectively. The addition of chemotherapy increased the % of pts requiring only a single apheresis to reach the mobilization target. Figure 2 summarizes the mobilization kinetics for each re-mobilization regimen. In general, a limited number of cells was collected with each aphereses; 〉70% of pts failed to mobilize 2 × 10^6 CD34+ cells/kg. In contrast, remobilization with AMD3100 allowed the collection of sufficient CD34+ cells in 67% of pts; median number of apheresis to reach the target was 3. Conclusions: Factors associated with mobilization kinetics of CD34+ cells include disease state and mobilization regimen. Re-mobilization is associated with high failure rates, re-mobilization regimens including AMD3100 are more successful. Figure 1: Mobilization kinetics by disease state Figure 1:. Mobilization kinetics by disease state Figure 2: Mobilization kinetics by re-mobilization regimen Figure 2:. Mobilization kinetics by re-mobilization regimen

Description: Relapsed or refractory (rel/ref) classical Hodgkin lymphoma (cHL) remains a clinical challenge, with limited effective treatment options available after stem cell transplantation. In a multicenter phase 2 study, the efficacy of lenalidomide in rel/ref cHL patients was evaluated at a dose of 25 mg/d on days 1-21 of a 28-day cycle. Patients remained on lenalidomide until disease progression or an unacceptable adverse event (AE) occurred. Thirty-eight cHL patients were enrolled with a median of 4 (range, 2-9) prior therapies; 87% had undergone prior stem cell transplantation and 55% of patients did not respond to their last prior therapy. Of 36 evaluable patients, responses were 1 complete remission (CR), 6 partial remissions (PRs), and 5 patients with stable disease (SD) for ≥ 6 months resulting in an International Working Committee (IWC) objective overall response rate (ORR) of 19% and a cytostatic ORR of 33%. Decreased chemokine (CCL17 and CCL22) plasma levels at 2 weeks were associated with a subsequent response. The treatment was well tolerated, and the most common grade 3/4 AEs were neutropenia (47%), anemia (29%), and thrombocytopenia (18%). Four patients discontinued lenalidomide because of rash, elevated transaminases/bilirubin, and cytopenias. We provide preliminary evidence of lenalidomide's activity in patients with rel/ref cHL, and therefore exploration of lenalidomide in combination with other active agents is warranted. This trial is registered at www.ClinicalTrials.gov as NCT00540007.

Description: Abstract 32 Background: Plerixafor (Mozobil®) is a CXCR4 antagonist that was recently approved by the Food and Drug Administration for use in combination with granulocyte-colony stimulating factor (G-CSF) to mobilize hematopoietic stem cells (HSCs) in patients with non-Hodgkin's lymphoma and multiple myeloma undergoing autologous transplantation. As part of our ongoing clinical evaluation of plerixafor, we observed that the drug efficiently mobilized a CD34dim population of HSCs. Here, we characterize the CD34dim population and assess its relative frequency following mobilization with G-CSF and/or plerixafor in apheresis samples and in non-mobilized cord blood samples. Methods: Aliquots of apheresis products were obtained from 16 healthy donors following treatment with a single injection of plerixafor or 5 days of G-CSF (10 μg/kg/day). In a separate study, five patients with lymphoma were mobilized sequentially with plerixafor (240 μg/kg), G-CSF (10 μg/kg/day), and plerixafor + G-CSF. These patients received an injection of plerixafor on day −5. Twenty-four hours later, patients were treated with G-CSF for 4 days. On the fifth day (day 0), patients received a dose of G-CSF and a dose of plerixafor followed four hours later by apheresis. Samples were collected after treatment with plerixafor on day −5 as well as immediately before and 4 hrs after plerixafor administration on day 0. Cord blood units were obtained from the St. Louis Cord Blood Bank. Human CD34+ cells were purified by positive selection with a Magnetic Affinity Cell Selection CD34 isolation kit. We used Affymetrix Human Genome U133+2 arrays to generate gene expression profiles for 24 samples of CD34+ cells collected following mobilization of healthy donors with either Plerixafor (n=12) or G-CSF (n=12). Results: Human CD34+ cells can be divided into three distinct subsets based on their cell surface expression of CD45RA and CD123 (IL-3Rα): (i.) CD34+CD45RA−CD123+/− primitive HSCs, (ii.) CD34+CD45RA+CD123+/− committed progenitors, and (iii.) CD34dimCD45RA+CD123hi cells. Table 1 shows the relative frequencies of each of these CD34+ cell subsets in apheresis products obtained from healthy donors mobilized with G-CSF or plerixafor as well as in cord blood units. Strikingly, we observed that each graft type was significantly enriched for one of the CD34+ cell subsets compared to the other two grafts. G-CSF mobilized grafts contained more CD45RA−CD123+/− primitive HSCs, cord blood units were enriched with CD45RA+CD123+/− committed progenitors and plerixafor mobilized products had significantly more CD34dimCD45RA+CD123hi cells. Table 1 also shows the CD34 cell subset distribution following sequential mobilization of lymphoma donors with plerixafor, G-CSF, and plerixafor + G-CSF (PL+G). This data agrees with the results we obtained from healthy donors and confirms that G-CSF grafts are enriched with CD34+CD45RA−CD123+/− cells while plerixafor preferentially mobilizes the CD34dimCD45RA+CD123hi subset. Extensive FACS and functional analyses determined that the CD34dim population represents a plasmacytoid pro-DC2 (for progenitor of pre-dendritic cell type 2) progenitor compartment as indicated by their CD45RA+CD123hiBDCA−2+BDCA−4+CD36+CXCR4hiCD4dimCD25−CD13− phenotype and inability to form CFU-GM colonies in vitro. Finally, we found that the gene signature of plerixafor-mobilized CD34+ cells is distinct from that of G-CSF-mobilized CD34+ cells. Plerixafor-mobilized CD34+ cells expressed significantly more of the specific transcriptional regulator of plasmacytoid DC (pDC) development, E2-2, as well as additional transcriptional factors (SpiB, IRF7, IRF8) and cell surface markers (BDCA-2, ILT7) that are specific to the pDC lineage. Conclusions: This data suggest that the CD34dimCD45RA+CD123hi cells preferentially mobilized by plerixafor are precursor pDCs. Further study is required to determine the impact these cells have on the engraftment and function of plerixafor mobilized grafts after hematopoietic stem cell transplantation. Disclosures: DiPersio: Genzyme Corp.: Honoraria.

Description: NK cells are a promising cellular immunotherapy for cancer. Cytokine-induced memory-like (ML) NK cells differentiate after activation with IL-12, IL-15, and IL-18, exhibit potent anti-tumor responses, and safely induce complete remissions in patients with leukemia. However, many cancers are not fully recognized via NK cell receptors. Chimeric antigen receptors (CAR) have been utilized to enhance tumor-specific recognition by effector lymphocytes. We hypothesized that memory-like differentiation and CAR-engineering would result in complementary improvements in NK cell responses against NK-resistant cancers. To test this idea, peripheral blood ML NK cells were modified to express an anti-CD19 CAR (19-CAR-ML), which displayed significantly increased IFN-γ production, degranulation, and specific killing against NK-resistant lymphoma lines and primary targets, compared to non-specific control CAR-ML NK cells or conventional CAR NK cells. The 19-CAR and ML responses were synergistic, CAR-specific, and required ITAM signaling. Furthermore, 19-CAR-ML NK generated from lymphoma patients exhibited improved responses against their autologous lymphoma. 19-CAR-ML NK cells controlled lymphoma burden in vivo and improved survival in human xenograft models. Thus, CAR engineering of ML NK cells enhanced responses against resistant cancers and warrant further investigation, with the potential to broaden ML NK cell recognition against a variety of NK cell-resistant tumors.

Description: Introduction Disease recurrence is the major cause of treatment failure after allogeneic hematopoietic stem cell transplantation (alloHSCT) for acute myeloid leukemia (AML) and myelodysplastic syndrome (MDS). Graft-versus-host disease (GVHD) is the major cause of non-relapse morbidity and mortality after alloHSCT. Decitabine (DAC) is a hypomethylating agent that irreversibly binds to and inhibits DNA methyltransferase-1, leading to loss of DNA methylation. DAC maintenance may help eradicate minimal residual disease and facilitate a graft-versus-leukemia effect. Lower DAC doses are expected to be better tolerated after alloHSCT and equally effective in promoting hypomethylation. Additionally, DAC maintenance may have a favorable effect on the incidence of GVHD by enhancing the effect of T-regulatory lymphocytes (Choi et al, Blood, 2012). Methods Patients (pts) with AML/MDS in complete remission (CR) after alloHSCT, with ANC〉 1,500/mm3, platelets〉 50,000/mm3, and without grade III-IV acute GVHD were eligible to receive DAC, starting between day +50 and +100 after alloHSCT. We investigated 4 DAC doses: 5, 7.5, 10 and 15 mg/m2/day IV x 5 days of a 6-week cycle, for a total of 8 cycles. Each cohort contained 4-8 evaluable patients. The Maximum Tolerated Dose (MTD) was defined as the maximum dose at which〈 20% of patients experience hematologic or non-hematologic dose limiting toxicities (DLT) during the 1st cycle of treatment. GVHD prophylaxis was at the physician discretion. Results 19 pts were enrolled to date; the median age was 60 y (22-66); 14 pts had AML and 5 MDS. All conditioning regimens were myeloablative; 14 donors were unrelated and 5 related. 3 cohorts have been completed and a final 4th cohort is currently enrolling. Median follow-up from alloHSCT is 24 mo (7-36). 8 pts (44%) completed all 8 cycles: 7 pts remain in CR with stable counts and full donor chimerism and 1 pt developed CNS-only relapse 26 mo after alloHSCT. 9 pts went off study before cycle 6: 1 pt for poor compliance after 6 cycles, 3 pts for relapsed disease (after 1, 2 and 5 cycles, respectively), 2 pts for sepsis, and 2 pts after physician decision. 6 pts have died: 3 from relapse, 2 from sepsis after 3 cycles of DAC (they were not neutropenic at a time), and 1 form sepsis 〉1 y after getting off study. 2 pts are still on study passed 3rd cycle. DAC maintenance was well tolerated. Associated hematological toxicities were mostly grade I/II leukopenia and thrombocytopenia. There was one occurrence of hematological DLT. No MTD has been reached. Non-hematological toxicities were grade I/II nausea, fatigue, neuropathy, and transaminase elevation. 2 pts had grade I-II acute GVHD prior to starting DAC and both resolved while on DAC; 1 pt developed grade IV gut GVHD coinciding with first cycle of DAC that completely resolved on DAC; 1 pt developed late acute GVHD of skin and liver around 6th cycle of DAC that resolved after few wks. 2/8 pts who completed 8 cycles of DAC developed very mild skin and oral chronic GVHD not requiring any systemic therapy, 1/8 pt developed late acute GVHD responding to therapy, and 1 pt developed overlap GVHD syndrome. 4/7 pts who went off study prior cycle 6, and did not have an early relapse, developed severe chronic GVHD requiring intensive immunosuppressive therapy. Conclusion To our knowledge this is the first report of DAC as maintenance therapy after alloHSCT. DAC at the dose of 15 mg/m2 for 5 days every 6 weeks is safe and can be administered in heavily pretreated pts in the post-alloHSCT setting. Approximately 43% of pts were able to receive all 8 cycles. The lack of toxicities and low incidence of GVHD indicate that a longer period of administration should be investigated. Although there is a trend of increased FOXP3 expression, results were not statistically significant. Further correlative studies, including genome-wide methylation studies, are ongoing. Disclosures: Off Label Use: Decitabine maintenance after alloHSCT.

Description: Abstract 787 The interaction between leukemic blasts and the bone marrow microenvironment is postulated to be an important mediator of chemoresistance in AML. Although many candidate receptor/ligand pairs have been implicated, the CXCR4 / SDF-1 axis functions as the principal regulator of homing and retention of both normal and malignant hematopoietic cells in the marrow. We hypothesized that disruption of the CXCR4 / SDF-1 axis with plerixafor (Mozobil®), a small molecule inhibitor of CXCR4, may sensitize AML to the effects of chemotherapy. We conducted an open label, phase I/II study in patients with relapsed or refractory AML in which plerixafor was administered prior to salvage chemotherapy. A test dose of plerixafor was administered SQ followed by a 24 hour observation period to analyze its effects on AML blasts in the absence of chemotherapy. Plerixafor was then given 4 hours prior to MEC chemotherapy (mitoxantrone 8 mg/m2/d, etoposide 100 mg/m2/d and cytarabine 1,000 mg/m2/d) daily for 5 days. Forty pts have been enrolled in the study with median age of 49 yrs (range 19-71). Baseline characteristics include 6 pts (15%) with secondary AML, 4 (10%) with prior transplant, 24 (60%) with intermediate and 10 (25%) with poor risk cytogenetics. Thirty-six pts (90%) received plerixafor + MEC as their 1st salvage regimen for relapsed disease with 21 (53%) having a CR1 duration of 〈 12 months and 9 pts (6%) for primary refractory disease. The remaining four pts (10%) received the regimen as their 2nd salvage regimen. Three dose levels of plerixafor: 80 mcg/kg, 160 mcg/kg and 240 mcg/kg were tested in the phase I dose escalation. In the phase II, a total of 34 patients have been treated at the 240 mcg/kg dose level. Common grade ≥ 3 adverse events consisted primarily of cytopenias and infections. No evidence of hyperleukocytosis or significant delays in neutrophil recovery (ANC 〉500/mm3, median 27d, range 21-37) or platelet recovery (plt 〉50k/mm3, median 26d, range 20-40d) were observed. Of the 32 pts currently evaluable for response at the 240 mcg/kg dose level, a complete remission (CR+CRi) has been achieved in 50% of pts (CR=13, CRi=3) which compares favorably to historical CR rates of 25-35%. Treatment failure was due to persistent disease in 14 pts (44%) and early death due to complications from infection in 2 pts (6%). One year KM estimate of overall survival is currently 56%. Correlative studies demonstrate that plerixafor mobilizes AML blasts (mean 2.5-fold increase, range 0.9-7.3 fold) into the peripheral circulation peaking at 6-8 hours after administration. FISH performed in pts with informative cytogenetic abnormalities indicates that mobilization occurs equally in both non-leukemic and leukemic populations. Higher baseline surface CXCR4 expression correlated with increased mobilization of AML blasts (Pearson's r=0.53, p=0.023) into the PB at 6 hrs post-plerixafor. In addition, a strong correlation was also observed between baseline CXCR4 expression and % SDF-1 migration in transwell assays (r=0.84, p=0.0013). Furthermore analysis of AML PB blasts at 6 hrs post-plerixafor demonstrate increased CXCR4 expression as well as increased chemotaxis in response to an SDF-1 gradient in transwell assays compared to baseline (64% vs 38%, p=0.0006). We conclude that plerixafor can be safely administered in combination with cytotoxic chemotherapy in patients with AML. Based on encouraging preliminary evidence of efficacy and in vivo evidence of CXCR4/SDF-1 blockade, confirmatory randomized studies of plerixafor for chemosensitization in AML are being planned. Disclosures: Uy: Genzyme: Consultancy, Speakers Bureau. Off Label Use: Plerixafor for AML. Abboud:Genzyme: Consultancy. Vij:Genzyme: Consultancy. DiPersio:Genzyme: Consultancy, Honoraria.