ALBERT

Description: There is a continued need to develop new selective human monoamine oxidase (hMAO) inhibitors that could be beneficial for the treatment of neurological diseases. However, hMAOs are closely related with high sequence identity and structural similarity, which hinders the development of selective MAO inhibitors. “Three-Dimensional Biologically Relevant Spectrum (BRS-3D)” method developed by our group has demonstrated its effectiveness in subtype selectivity studies of receptor and enzyme ligands. Here, we report a series of novel C7-substituted coumarins, either synthesized or commercially purchased, which were identified as selective hMAO inhibitors. Most of the compounds demonstrated strong activities with IC50 values (half-inhibitory concentration) ranging from sub-micromolar to nanomolar. Compounds, FR1 and SP1, were identified as the most selective hMAO-A inhibitors, with IC50 values of 1.5 nM (selectivity index (SI) 〈 −2.82) and 19 nM (SI 〈 −2.42), respectively. FR4 and FR5 showed the most potent hMAO-B inhibitory activity, with IC50 of 18 nM and 15 nM (SI 〉 2.74 and SI 〉 2.82). Docking calculations and molecular dynamic simulations were performed to elucidate the selectivity preference and SAR profiles.

Description: Background: In β-thalassemia, imbalanced production of α and β globin chains in erythroid precursors inhibits late-stage erythroid differentiation, leading to ineffective erythropoiesis and anemia. Luspatercept is a modifiedactivin receptor type IIB-IgG Fc fusion protein that promotes terminal erythroid differentiation and mitigates ineffective erythropoiesis in murine β-thalassemia. In phase 2 studies, luspatercept led to long-term increases in hemoglobin (Hb) levels and reduction in transfusion burden in patients (pts) with β-thalassemia. Aims: To characterize the pharmacokinetics (PK) of luspatercept, and to explore the exposure-response relationship for efficacy and safety in pts with β-thalassemia, thereby informing selection of the starting dose for phase 3 studies of luspatercept in β-thalassemia. Methods : PK, safety, and efficacy data were collected from two phase 2 studies (base and extension). In the base study, luspatercept was administered once every 3 weeks by subcutaneous injection to sequential cohorts for up to 5 doses. The base study included a dose-finding phase (at fixed doses ranging from 0.2 mg/kg to 1.25 mg/kg) and an expansion cohort (at a starting dose of 0.8 mg/kg followed by individual dose titration up to 1.25 mg/kg). Pts completing the base study were eligible to enroll in an extension study, where they continued to receive luspatercept every 3 weeks for up to 24 months. Pts who had treatment interruption for ≥ 3 months before enrolling in the extension study received a starting dose of 0.8 mg/kg (followed by dose titration) and were treated as "new" pts in the exposure-response analysis. The main exposure endpoint was area under the luspatercept serum concentration-time curve (AUC). Clinical endpoints included Hb increase, transfusion reduction, and drug-related adverse events (AEs) in weeks 1-15. Responders were defined as non-transfusion-dependent (NTD) pts with a mean Hb increase ≥ 1.0 g/dL/15 weeks, and transfusion-dependent (TD) pts with a transfusion reduction ≥ 33%/15 weeks. Results : As of July 20, 2016, preliminary data were available for 89 pts, including 49 NTD pts (baselineHb 6.5-9.8 g/dL) and 40 TD pts (baseline transfusion burden 4-18 units/12 weeks). Median age was 37 years (range 20-62), and 47% were female. A total of 49 pts were eligible for individual dose titration, with ~71% having at least 1 dose escalation (to 1 mg/kg) and 27% having 2 dose escalations (up to 1.25 mg/kg) within the first 5 cycles. Luspatercept PK was adequately described by a 1-compartment PK model with linear absorption and elimination. The half-life of luspatercept in serum was~10 days across doses. Body weight positively correlated with luspatercept clearance and its volume of distribution. TD pts had~23% lower volume of distribution than NTD pts. In NTD pts, higher luspatercept serum AUC correlated with a greater increase in Hblevels (P 〈 0.01). Most TD pts had a reduction in transfusion burden; however, an exposure-dependent trend was not apparent for relative or absolute reduction in units transfused, possibly due to the narrow exposure range tested. The median AUC was similar for NTD responders and TD responders (~100d·μg/mL vs ~120d·μg/mL). As AUC increased, the frequency of responders increased for NTD pts, TD pts, and the 2 groups combined (Figure). Population PK simulation predicted that a starting dose of 1.0 mg/kg luspatercept would result in 〉 90% of NTD pts and ~50% of TD pts achieving the median AUC concentration observed in responders. By contrast, 〈 50% pts (NTD or TD) were predicted to achieve the target AUC at the 0.8 mg/kg dose. Grade 2-3 drug-related AEs (all types) were more frequent with higher AUC (P 〈 0.05). In both NTD and TD pts, bone and muscle pain were the most common AEs; however, these AES did not have a significant relationship with luspatercept serum exposure. Conclusions: Higher luspatercept serum exposure correlated with greater erythroid hematopoietic response as well as more frequent grade 2-3 related AEs. Exposure-response modeling and PK simulation support a phase 3 starting dose of 1.0 mg/kg and intra-patient dose escalation up to 1.25 mg/kg according to pts' response. A phase 3 study of luspatercept in regularly transfused adults with β-thalassemia is ongoing (BELIEVE study; ClinicalTrials.gov NCT02604433). Disclosures Chen: Celgene Corporation: Employment, Equity Ownership. Laadem:Celgene Corporation: Employment, Equity Ownership. Wilson:Acceleron Pharma: Employment, Equity Ownership. Zhang:Acceleron Pharma: Employment. Sherman:Acceleron Pharma: Employment, Equity Ownership, Patents & Royalties. Ritland:Celgene Corporation: Employment, Equity Ownership. Attie:Acceleron Pharma: Employment, Equity Ownership.

Description: Abstract 2737 Introduction: ATL is prevalent in Japan and has the worst prognosis among T-cell malignancies. PTCL also has a poor prognosis with currently available chemotherapeutic regimens, and both would benefit from better treatment modality. Lenalidomide is an immunomodulatory agent with direct tumoricidal and antiproliferative activity, and is approved for multiple myeloma (MM) in combination with dexamethasone after at least 1 prior therapy and for transfusion-dependent anemia due to low- or intermediate-1-risk myelodysplastic syndromes associated with 5q deletion. We conducted a phase 1 study of lenalidomide in patients with relapsed ATL or PTCL to establish the recommended dose and schedule for a subsequent phase 2 study. Patients and Methods: This multicenter, phase 1, dose-escalation study assessed the safety, maximum tolerated dose (MTD), pharmacokinetics, and efficacy in patients with relapsed advanced ATL or PTCL. Dose-escalation was conducted according to the standard 3+3 design. Up to one PTCL patient was allowed to be included in each cohort of 3 patients. Patients in Cohort 1 received oral lenalidomide 25 mg daily on Days 1–21 of a 28-day cycle. Patients in Cohorts 2 and 3 received 25 and 35 mg/day, respectively, on each day of the 28-day cycle. Dose-limiting toxicity (DLT) was defined as febrile neutropenia lasting 5 or more days; thrombocytopenia (platelets

Description: Introduction Lenalidomide (LEN) is a weak substrate of P-glycoprotein (P-gp) in vitro and renal excretion of LEN is the primary elimination route following oral administration.  A P-gp inhibitor may have the potential to increase systemic exposure to LEN by reducing renal elimination at the tubular level and enhancing oral absorption at the gut level. Recently, a single uncontrolled phase 1 study (Hofmeister, 2011) and a case report (Takahashi, 2012) described that plasma exposure to LEN and temsirolimus was increased in multiple myeloma patients when lenalidomide was co-administered with a P-gp inhibitor (temsirolimus and intraconazole, respectively).  This clinical study assessed LEN-drug interactions via P-gp using two probe drugs under controlled conditions.  Quinidine, a P-gp inhibitor with high in vivo inhibition potential and proven effect on plasma exposure of the prototype P-gp substrate digoxin in humans, was used to maximize the likelihood of detecting drug-drug interactions via P-gp.  Temsirolimus, a P-gp inhibitor/substrate, was used to evaluate P-gp mediated interactions on either drug in comparison with the results reported in literature. Methods This was a phase 1, single-center, open-label, 2-part, fixed-sequence crossover study conducted in healthy men. Part 1 comprised of 2 treatment periods with LEN alone (25 mg single dose on Day 1) in period 1, followed by LEN (on Day 4) plus quinidine (300 mg twice daily [BID] on Day 1 and 600 mg BID on Days 2–5) in period 2.  Part 2 consisted of three treatment periods with LEN (25 mg single dose on Day 1) alone in period 1, temsirolimus (25 mg single dose intravenously [IV] on Day 1) alone in period 2, and LEN plus temsirolimus in period 3 (Day 1).  Treatment periods were separated by a washout of 7–10 days. Serial samples were collected to determine the plasma, whole blood or urine concentrations of LEN, quinidine, and temsirolimus (and its active metabolite, sirolimus [also a P-gp inhibitor/substrate]).  Safety was monitored throughout the study. Results A total of 31 healthy men, aged 22 to 62 years; were enrolled (14 in Part 1 and 17 in Part 2). Renal excretion of LEN was almost complete at 12 h post dose for all treatments (Figure 1). In the absence or presence of a P-gp inhibitor, the mean percentage LEN dose excreted in the urine (74% vs 70% in Part 1, respectively; 81% vs 80% in Part 2) and renal clearance (227 vs 245 mL/min in Part 1; 251 vs 229 mL/min in Part 2) were similar, demonstrating that the rate and capacity of LEN renal excretion were not reduced by P-gp inhibition. Both the median time (1 h) to reach the maximum concentration (Cmax) and the oral bioavailability (70–80% of the administrated dose, as indicated by the renal excretion of unchanged drug) of LEN, were comparable in the absence or presence of a P-gp inhibitor (0.5–1 h and 74–81% of the dose, respectively); therefore, the rate and extent of LEN oral absorption were also not altered by P-gp inhibition. Consequently, there was no significant change in the plasma exposure to LEN in the presence of a P-gp inhibitor (Figure 1). The 90% confidence intervals (CIs) for the ratio of geometric means between LEN alone and LEN plus a P-gp inhibitor were completely contained within the equivalence limits of 80–125% for Cmaxand area under the concentration-time curve (AUC) of LEN. In addition, LEN had no effect on blood exposure to temsirolimus and sirolimus, with the 90% CIs for the ratio of their geometric mean Cmax and AUC between temsirolimus alone and temsirolimus plus LEN between 80–125%. No significant safety findings were reported when LEN was given with quinidine or temsirolimus compared to LEN alone. Conclusions Co-administration of either the P-gp inhibitor quinidine or temsirolimus had no clinically relevant effect on the systemic exposure of LEN.  Similarly, co-administration of LEN had no clinically relevant effect on the systemic exposure of the P-gp substrates temsirolimus and sirolimus.  A single dose of LEN was well tolerated when co-administered with quinidine or temsirolimus in healthy men. Disclosures: Chen: Celgene Corporation: Employment, Equity Ownership. Weiss:Celegene Corporation: Employment, Equity Ownership. Reyes:Celgene Corporation: Employment, Equity Ownership. Liu:Celgene Corporation: Employment, Equity Ownership. Kasserra:Celgene: Employment, Equity Ownership. Wang:Celgene Corporation: Employment, Equity Ownership. Zhou:Celgene Corporation: Employment, Equity Ownership. Kumar:Celgene Corporation: Employment, Equity Ownership. Weiss:Celgene Corporation: Employment, Equity Ownership. Palmisano:Celgene Corporation: Employment, Equity Ownership.

Description: Abstract 2934 Introduction: Len combined with Dexamethasone (Dex) is becoming a standard treatment in multiple myeloma (MM). As Len is mainly excreted by the kidneys, its dose should be adjusted according to renal function. Current dosing recommendations are based on a study conducted in non-malignant patients (pts) and on modelling/simulations. To assess whether these recommendations are actually valid in MM pts, we conducted a prospective study evaluating pharmacokinetics (PK), safety and efficacy of Len+Dex in pts with various degrees of renal impairment (RI). We also compared the glomerular filtration rate (GFR) estimated either by the Modification of Diet in Renal Disease (MDRD) dosage (GFR/MDRD) or by the Cockroft and Gault (CG) equation (GFR/CG) for determining Len dosage. Methods: 37 Caucasian pts (median age 65 yrs) with symptomatic MM who had received ≥ 1 previous line of treatment, were enrolled. All had stable renal function over 4 weeks prior to inclusion. They were divided in 5 groups according to GFR/CG at baseline: group 1, 〉 80mL/min (N = 10); group 2, ≥ 50 & ≤ 80 mL/min (N = 10); group 3, ≥ 30 & 〈 50 mL/min (N = 7); group 4, 〈 30 mL/min (N = 5); group 5, chronic hemodialysis (N = 5). Cast nephropathy and light chain deposition disease were documented by kidney biopsy as the respective cause of RI in 6 and 1 of the 17 pts from groups 3, 4 and 5. Pt characteristics were similar, except for pts in group 4 who were significantly older (p = 0.01). All pts received ≥ 3 cycles of oral Len+Dex (40 mg weekly) regimen from Days 1–21 of each 28-day cycle. Len starting dose was defined according to the current dosing guideline, i.e.: group 1 and 2: 25 mg/d; group 3: 10 mg/d; group 4: 15 mg/qod; group 5: 5 mg/d. Blood samples were collected on a dosing day in the first cycle for PK analyses, as follows: at pre-dose, 0.5, 1, 1.5, 2, 4, 6, 8, 10, 24 h (and 36 and 48 h in groups 3–5) after the Len dose on day 5, and at predose and 2 h (near tmax) after the Len dose on days 9 and 15. Results: Len clearance was highly correlated to GFR/CG (R2 = 0.86, p 〈 0.001). Mean Len clearance declined from 239 mL/min in group 1, to 160, 93, 54, and 41 mL/min, and mean terminal half life was prolonged from 3 h to 5, 7, 10, and 24h in groups 2, 3, 4 and 5, respectively. These findings were consistent with those reported for pts with RI due to non-malignant conditions. The average daily AUC values for groups 2–5 were 103–149% of that for group 1 (1794 h*ng/mL). No difference was found in mean plasma Len concentrations at 2 h post-dose between day 9 (Len alone) and day 15 (Len+Dex) across groups. GFR/MDRD and GFR/CG were highly correlated (R2 = 0.85, slope = 0.89) and similarly predicted Len clearance (slope = 2.38 and 2.15, respectively). When the same cutoff values were used for GFR/MDRD and GFR/CG, the % reduction in Len clearance in groups 3–5 compared with the combined groups 1 and 2 was very similar for GFR/MDRD- and GFR/CG-based renal function classifications. However, a dosage discordance between GFR/MDRD and GFR/CG would occur in 5/32 non-dialysis pts. All the 5 pts would be assigned to a lower starting dose according to GFR/MDRD. The estimated resulting AUC levels would have been reduced to 68–94% of the mean AUC in group 1 from the observed 91–238%. These estimated AUC levels would be considered to be in the therapeutic range. After 3 cycles of Len+Dex, hematological response rate (≥ PR) was 73% (VGPR 27%). Response rates were 70% in pts with 〈 50ml/min. Renal function remained stable in all pts. A total of 22 serious (≥ grade 3) adverse events (SAEs), including 16 hematological SAEs, occurred in 12 pts, leading to dose reduction in 7 cases. Of these, 2 pts would have been assigned to a lower starting dose according to GFR/MDRD. The frequency of SAEs was not significantly increased in group 5 compared to the other groups (60% vs 45%, p = 0.3). After a median follow-up of 10 months, pts had received a mean number of 7.6 ± 4.3 cycles, and 5 had died, because of MM progression in 4 cases. Conclusion: The study demonstrated that the effect of RI on the Len PK in MM patients receiving concomitant Dex was similar to that in non-malignant pts receiving Len alone. The recommended dose adjustments achieved the appropriate plasma exposure with similar efficacy and safety across different renal function groups in MM pts. GFR/MDRD and GFR/CG may be interchangeable for determining the Len dosage. Disclosures: Chen: Celgene Corporation: Employment. Alakl:Celgene Corporation: Employment. Neel:Celgene Corporation: Employment. Jaccard:Celgene Corporation: Consultancy.

Description: The erythropoietic effects of lenalidomide are cytokine dependent, suggesting that the erythroid hematologic improvement (HI-E) rate may be augmented by combined treatment (CT) with recombinant human erythropoietin (rhu-EPO) in myelodysplastic syndrome (MDS). In the present study, we explored the benefits of CT and the relationship between lenalidomide pharmacokinetics and hematologic toxicity in transfusion-dependent patients with low- to intermediate-1–risk MDS who failed prior rhu-EPO. In stage I, patients received 10 or 15 mg/d of lenalidomide monotherapy. At week 16, erythroid nonresponders (NRs) were eligible for CT with rhu-EPO 40 000 U/wk. Among 39 patients, HI-E response rate to monotherapy was 86% (6 of 7) in del(5q) and 25% (8 of 32) in non-del(5q) patients (10 mg, 17.7%; 15 mg, 33.3%). Twenty-three patients proceeded to CT, with 6 (26.0%) achieving HI-E. In 19 non-del(5q) patients, 4 (21.1%) showed HI-E. Mean baseline serum EPO in non-del(5q) patients was lower in monotherapy and CT responders than in NR (not statistically significant). Thrombocytopenia was significantly correlated with lenalidomide area under the plasma concentration-time curve (P = .0015), but severity of myelosuppression did not. The benefits of lenalidomide plus rhu-EPO are currently under investigation in a phase 3 Eastern Cooperative Oncology Group (ECOG)–sponsored intergroup study. This study is registered at www.clinicaltrials.gov as NCT00910858.

Description: BACKGROUND: There is an unmet need for effective therapies for relapsed/refractory primary and secondary central nervous system (CNS) and intraocular lymphomas (IOL), complications that are increasing in frequency among patients age 〉60. Whole brain irradiation, broadly utilized in the relapsed setting, yields insufficient efficacy and considerable morbidity, particularly in older patients. Implementation of new agents that selectively target survival pathways upregulated in refractory CNS lymphoma would have significant impact. Lenalidomide (CC-5013) has established activity as a single agent in aggressive NHL, particularly in ABC-type DLBCL. We recently demonstrated the activity of lenalidomide, at modest doses, in the treatment of recurrent/refractory intraocular and CNS lymphoma (J. Clin Oncol, 2011). We also provided evidence of cereblon-dependent efficacy of lenalidomide in preclinical models of patient-derived CNS DLBCL (ASH, 2013). These observations are the basis for this first trial of IMiD® immunomodulatory compound therapy in CNS NHL, as monotherapy, and in patients with inadequate responses to lenalidomide, in combination with combined intravenous plus intraventricular rituximab. (NCT01542918). METHODS: The primary objective of this phase I study is to evaluate the safety and efficacy of lenalidomide at three dose levels (10, 20, and 30mg) in relapsed/refractory CD20+ CNS NHL, with staging evaluations involving brain, CSF and intraocular compartments. Key secondary endpoints include: (1) determination of extent of CSF penetration by lenalidomide; (2) feasibility and activity of combined intraventricular and intravenous rituximab plus lenalidomide in patients with recurrent CNS lymphoma, not responsive to lenalidomide monotherapy; (3) evaluation of the effects of lenalidomide on the tumor microenvironment, including effects on tumor metabolism and immune cell infiltration and phenotypes. RESULTS: Thus far, seven subjects with relapsed CNS DLBCL (6 PCNSL, 1 SCNSL) have been treated on protocol at UCSF (median age 63, range 46-77). Each had methotrexate-resistant disease and 4 had lymphoma refractory to high-dose chemotherapy. In general, lenalidomide has been well-tolerated in the CNS lymphoma population with one DLT (non-hematologic) noted at the 20 mg dose level: fatigue and memory loss. Clinical benefit has been noted in 3 out of 4 patients with IOL, with 1 PR 〉 6 months duration and 1 SD 〉 10 months duration, each to lenalidomide monotherapy. Brain parenchymal responses to lenalidomide monotherapy (20 mg) have been demonstrated at restaging MRI, including 1 CR and 1 PR. There has been 1 CR in the leptomeninges with resolution of B-cell lymphoma in CSF, quantified by serial flow-cytometry. Combined intraventricular (20 mg) plus intravenous infusion of rituximab was well-tolerated in 3/3 patients and, with 10 mg lenalidomide, resulted in 1 PR in highly refractory IOL. Using GC/MS, we demonstrated lenalidomide penetration in ventricular CSF (0.6-7.9 ng/ml) in each of 4 patients, 12-15 hours after dosing at the 20 mg level, but trough lenalidomide concentrations 〉0.5 ng/ml were detected in only 1 of 3 patients receiving 10 mg lenalidomide. Serial metabolomic profiling revealed that CSF lactate correlated with clinical response to lenalidomide. Finally, we demonstrated that lenalidomide is reproducibly associated with a rapid and reversible effect on peripheral blood macrophage polarization to an M1, iNOS+ phenotype. CONCLUSIONS: These preliminary results provide evidence that lenalidomide is well-tolerated at 10 and 20 mg dose levels in CNS lymphoma. We demonstrate for the first time lenalidomide penetration in ventricular CSF, with higher trough concentrations detected at 20 mg compared to the 10 mg dose level. Encouraging evidence of lenalidomide efficacy in relapsed CNS DLBCL has been demonstrated in intraocular, CSF and brain compartments. Combined intraventricular and intravenous infusion of rituximab is feasible and represents a novel strategy to the potentiation of immunotherapeutic strategies in brain tumors. Studies are in progress to further elucidate the safety, pharmacokinetics and mechanisms involved in this biological approach to the treatment of refractory CNS lymphomas. Supported by NCI, Leukemia and Lymphoma Society, UCSF Helen Diller Comprehensive Cancer Center, Celgene and Genentech Disclosures Off Label Use: Intrathecal administration of rituximab. Wang:Celgene: Employment. Chen:Celgene: Employment.

Description: Abstract 3454 Introduction: Sotatercept (ACE-011), a recombinant fusion protein consisting of the extracellular domain of the human activin receptor type IIA linked to the human IgG1 Fc domain, is a ligand trap for multiple TGF-β super family ligands including activin A/B and growth differentiation factor-11 (GDF-11). Data from in vitro studies and animal models suggest that sotatercept may act on late erythroblasts to promote erythropoiesis. The relationship between serum sotatercept exposure and erythropoietic responses was studied in healthy postmenopausal women (HPMW) and cancer patients (pts). Methods: Data were obtained from a total of 157 subjects in 5 clinical studies, including79 HPMW, 30 pts with multiple myeloma (46% were anemic, hemoglobin [Hgb] 〈 11 g/dL), and 48 anemic pts with solid tumors. In HPMW, sotatercept was given as either a single intravenous (IV) infusion over 60 min (0.01 to 3 mg/kg), or as multiple subcutaneous (SC) injections (0.03 to 1 mg/kg Q4W). In pts, sotatercept was given at 0.1 to 0.5 mg/kg Q4Wor 15 and 30 mg Q6W. Serial blood samples for pharmacokinetic analysis were collected for up to 24 weeks (wks) after the first dose. Hgb was monitored weekly or more frequently in most studies. The magnitude of the initial Hgb response was defined as the maximum Hgb increase (g/dL) from baseline within 4 wks of Dose 1. Erythropoietin (EPO) and reticulocyte data were obtained mainly from HPMW. Results: Increases in the maximum serum concentration (Cmax) and area under concentration-time curve over 28 days (AUC28d) of sotatercept were proportional to dose for both IV and SC administration. The half-life in serum was 3 to 4 wks, independent of the dosing route or study population. Cmax was ∼3-fold higher for IV doses than for the same SC doses, and it occurred at the end of IV infusion as compared to ∼7 days after SC injection. Body weight and gender were not found to significantly affect sotatercept exposure. In HPMW, a rapid initial Hgb increase was observed near the time of sotatercept Cmax (Tmax). With single IV doses, a rise in Hgb was evident on Day 2 (mean increase = 0.46 g/dL at 0.1 mg/kg and 0.56–0.84 g/dL at 0.3 to 3 mg/kg). With SC doses, a rise in Hgb was evident 1 wk after Dose 1 (mean increase ≥ 0.68 g/dL at 0.1 to 1 mg/kg). AUC28d and Cmax correlated with the magnitude of initial Hgb response up to 200 day•μg/mL and 10 μg/mL, respectively. Increasing Cmax from 10 to 100 μg/mLwith single IV doses did not further elevate the initial Hgb increase, but it prolonged the duration of Hgb increase (≥1 g/dL) from ∼8 to 〉16 wks. A concentration-dependent increase in serum EPO was observed in HPMW, with the peak level ranging from 25 to 67 mIU/mL at the SC dose of 1 mg/kg (baseline EPO = 11 to 23 mIU/mL). An increase in reticulocytes was also observed at ≥ 1 mg/kg, reaching the peak 1–2 wks postdose, consistent with the time for the formation of reticulocytes from early erythroid progenitors. The effect of sotatercept on reticulocytes was more evident with IV doses (∼2-fold higher than the baseline), suggesting this effect may be related to Cmax. EPO and reticulocytes responses were not apparent at lower SC doses or concentrations of sotatercept (≤ 0.3 mg/kg or 〈 5 μg/mL), though these doses induced an initial rapid Hgb increase. A rapid rise in Hgb was also observed in cancer pts at all studied doses, again with the peak increase occurring ∼1 wk postdose (near Tmax). Exposure-response analysis suggests that the magnitude of the initial Hgb increase in cancer pts was comparable to that in HPMW. However, the duration of Hgb response was less sustained in cancer pts receiving chemotherapy, likely due to lower exposure levels studied and myelosuppression. Alternative dosing regimens to increase exposure by increasing dose or shortening dosing interval are currently being explored. Conclusion: The clinical data suggests that sotatercept acts through a mechanism distinct from that of EPO. The rapid Hgb increase at all dose levels supports the hypothesis that sotatercept acts essentially at a late stage of erythropoiesis to induce erythroid maturation. High concentrations of sotatercept are also associated with EPO release, reticulocyte production and sustained Hgb increase, possibly through influencing earlier stages of erythropoiesis. These findings define a relationship between inhibition of TGF-β superfamily ligands and erythropoiesis in humans and support the development of sotatercept in anemia and diseases of ineffective erythropoiesis. Disclosures: Chen: Celgene: Employment. Laadem:Celgene: Employment. Sherman:Acceleron Pharma: Employment, Equity Ownership. Zhou:Celgene Corporation: Employment. Sung:Celgene: Employment, Equity Ownership. Palmisano:Celgene: Employment, Equity Ownership. Chopra:Celgene Corporation: Employment, Equity Ownership.

Description: Background: Luspatercept is a modified ActRIIB-IgG Fc fusion protein that corrects ineffective erythropoiesis. In phase 2 studies, luspatercept treatment led to long-term increases in hemoglobin (Hb) levels and reduction in transfusion burden in patients (pts) with IPSS Low- or Intermediate-1-risk MDS. Aims: To characterize the pharmacokinetics (PK) of luspatercept and explore the exposure-response relationship for efficacy and safety in pts with MDS, thereby informing selection of the starting dose for phase 3 studies of luspatercept in MDS. Methods : PK, safety, and efficacy data were collected from two phase 2 studies (base and extension). In the base study, luspatercept was administered once every 3 weeks by subcutaneous injection to sequential cohorts for up to 5 doses. The base study included a dose-finding phase (at fixed doses from 0.125 to 1.75 mg/kg), and an expansion cohort (at a starting dose of 1.0 mg/kg followed by individual dose titration up to 1.75 mg/kg). Pts completing the base study were eligible to enroll in an extension study, where they continued to receive luspatercept every 3 weeks for up to 24 months. Pts who had treatment interruption for ≥ 3 months before enrolling in the extension study received a starting dose of 1.0 mg/kg (followed by dose titration) and were treated as "new" pts in the exposure-response analysis. The main exposure endpoint was area under the luspatercept serum concentration−time curve (AUC). Clinical endpoints included Hb increase, transfusion reduction, and drug-related adverse events (AEs) in weeks 1-15. Responders were defined as pts achieving erythroid hematologic improvement (HI-E) per IWG criteria, i.e. Hb increase ≥ 1.5 g/dL for 8 weeks in low transfusion burden (LTB) pts, and transfusion reduction ≥ 4 RBC units/8 weeks in high transfusion burden (HTB) pts. Results : As of July 20, 2016, preliminary data were available for 66 pts: 22 LTB pts (baseline Hb 6.4-10.1 g/dL) and 44 HTB pts (baseline transfusion burden 4-18 units/8 weeks). Median age was 72 years (range 27-90); 41% were female. A total of 39 pts were eligible for individual dose titration; of these, ~49% had ≥ 1 dose escalation (to 1.33 mg/kg) and ~15% had 2 dose escalations (to 1.75 mg/kg) in the first 3 months. Luspatercept PK was adequately described by a 1-compartment PK model with linear absorption and elimination. Half-life of luspatercept in serum was ~10-14 days across doses. Body weight positively correlated with luspatercept clearance and its volume of distribution. Baseline transfusion burden (LTB vs HTB) and erythropoietin (EPO) level (10-4,752 U/L) had no significant effect on luspatercept PK. In LTB pts who were transfusion-free on treatment, higher luspatercept AUC correlated with greater Hb increase (P 〈 0.01). In HTB pts, AUC correlated with reduced transfusion units in pts with baseline EPO ≤ 500 U/L (P〈 0.01) but not in pts with baseline EPO 〉 500 U/L. Median AUC was 148 d·µg/mL in LTB responders and 185 d·µg/mL in HTB responders. Luspatercept AUC also correlated with frequency of IWG HI-E responders for LTB pts, HTB pts (baseline EPO ≤ 500 U/L), and the 2 groups combined. In pts requiring transfusion (≥ 2 units/8 weeks) with baseline EPO ≤ 500 U/L, baseline transfusion burden was a significant predictor of achieving transfusion independence (TI) ≥ 8 weeks, and higher luspatercept AUC was associated with greater TI rate after accounting for baseline transfusion burden. Thus, individualized dosing based on baseline transfusion burden may increase the likelihood of achieving TI in HTB pts. Population PK simulation predicted that the starting dose resulting in 90% of LTB pts and 50% of HTB pts achieving efficacious AUC for HI-E would be 1 mg/kg and 1.1 mg/kg, respectively; higher doses would result in a higher proportion of pts achieving efficacious AUC. There was no significant relationship between severity and frequency of drug-related AEs and luspatercept serum exposure. Conclusions: Higher luspatercept serum exposure correlated with greater erythroid hematopoietic response for both LTB and HTB pts. Exposure-response modeling and PK simulation support a phase 3 starting dose of 1.0 mg/kg and intra-patient dose escalation up to 1.75 mg/kg according to erythroid hematopoietic response. A phase 3 study of luspatercept in regularly transfused ring sideroblast positive patients with lower-risk MDS according to IPSS-R criteria is ongoing (MEDALIST study; ClinicalTrials.gov NCT02631070). Disclosures Chen: Celgene Corporation: Employment, Equity Ownership. Laadem:Celgene Corporation: Employment, Equity Ownership. Wilson:Acceleron Pharma: Employment, Equity Ownership. Zhang:Acceleron Pharma: Employment. Sherman:Acceleron Pharma: Employment, Equity Ownership, Patents & Royalties. Ritland:Celgene Corporation: Employment, Equity Ownership. Attie:Acceleron Pharma: Employment, Equity Ownership.