Roy, Vivek

Abstract Background: Previous studies have proposed combination of various factors that predict poor mobilization after GCSF stimulation. However, there is no single clinical or laboratory test that reliably correlates with bone marrow reserve and PBPC mobilization. Peripheral blood CD34+ cell count (pCD34) reliably correlates with PBPC mobilization and yield but only after GCSF administration. Methods: We have prospectively followed 36 autologous PBPC transplant candidates, from the time of initial evaluation to transplant, to evaluate factors the correlate with bone marrow reserve and autologous PBPC mobilization. Results: The patients median age was 63 (range 26 – 76). 21 of the 36 patients were diagnosed with dysproteinemia, primarily multiple myeloma, 10 patients had non-Hodgkin’s lymphoma while the remaining 5 had other form of hematological malignancies. Our evaluation shows baseline platelet count prior to growth factor administration significantly correlate with total CD34+ cell yield (spearman: r = 0.64, p <0.001, Figure 1). In addition, daily platelet count during PBPC harvest correlate with CD34+ cell yield for the day (spearman: r = 0.6, p <0.001). Using multiple linear regression model (F = 19.1, p <0.001, R2 = 0.64), we have determined patients’ age (p = 0.03) and type of disease (p <0.001), in addition to platelet count (p<0.001), significantly correlate with total CD34+ cell yield. Further analysis showed low baseline platelet count predicts poor PBPC mobilization. Using Receiver Operating Curve (ROC), we have determined that platelet count of 174,000/ul was the best cut-off for collecting at least 2 million/Kg CD34+ cells with sensitivity of 81% and specificity of 78% and estimated correct classification of 81%. Even though pCD34 counts were slightly more sensitive and were better at identifying poor mobilizers in our cohort (best cut-off 10/ul, sensitivity 96%, specificity 67%, correct classification 89%), their clinical utility were diminished since the values were only known after GCSF administration. Conclusion: Baseline platelet count is a sensitive predictor of bone marrow reserve and can be used prior to growth factor administration to predict poor mobilization. Figure 1: Correlation of baseline platelet count and total CD34+ cell yield Figure 1:. Correlation of baseline platelet count and total CD34+ cell yield

Abstract Most common abnormalities in MDS involve the loss of chromosomal material resulting from the loss of an entire chromosome or a part of it, due to deletions or unbalanced translocations. Among partial chromosome losses, del (5q) was found to be the most common followed by del (20q). MDS can also be associated with multiple morphologic RBC abnormalities including macrocytes, ovalocytes, dacrocytes, fragmented cells and basophilic stippling. Limited cases of MDS with elliptocytosis have been reported in the literature. Elliptocytosis encompasses a heterogeneous group of erythrocyte disorders. Although acquired cases of elliptocytosis have been reported, these are most often hereditary disorders with a variable clinical presentation from asymptomatic carriers to patients with severe hemolytic anemia. This disorder is quite rare in the Unites States (US). Mutations in protein 4.1, alpha spectrin, beta spectrin, glycophorin C and band 3 have been reported in hereditary elliptocytosis. Only 7 well reported cases of marked elliptocytosis associated with MDS have been reported in the literature. Of those 7 cases, 4 of them were associated to del(20q). All of them were elderly males and all of them had ≤ 5% bone marrow blasts. Anemia is usually moderate (median 9.4 g/dL; range 8.1–12) and thrombocytopenia is mild to moderate (median platelet count 116 x 109/L; range 46–149 x 109/L). All cases were described either in Europe (n= 1; Rummens et al. Acta Haematologica1986; 75:174–177) or in Asia (n= 3; Ideguchi et al. Br J Haematol.1993, 75:387–92; Ishida et al. Cancer Genetics and Cytogenetics1997, 108:162–165; Hur et al. Clin Lab Haematology, 2004, 26: 69–72). Of these 7 cases, 3 of them were associated with protein 4.1 deficiency. To the best of our knowledge these are the first 2 cases of patients with marked elliptocytosis and MDS with del 20q abnormality reported in the US: Patient 1 Patient 2 Age/sex 82/male 79/male FAB classification RA RA Year of diagnosis of MDS 1998 1999 WBC 3.0 x 109/L 11.5 x 109/L Platelets 82 x 109/L 79 x 109/L Hemoglobin 11.8 g/dL 10.7 g/dL Hematocrit 34.4% 31.9% Reticulocyte 4.3% NA LDH 350 u/L NA Direct Coombs Negative NA BM Cytogenetics 46 XY, del (20)(q11.2) 46 XY, del (20)(q11.2) Prior Therapy supportive care only Erythropoeitin Evaluation for protein 4.1 deficiency in one of these 2 patients is currently pending. The elliptocytosis noted in these MDS patients were most likely derived from acquired clones with abnormal development of the erythroid lineage and were associated with deletion of chromosome 20q. Patients with acquired elliptocytosis and associated MDS with del 20q chromosomal abnormality seem to have an indolent course with prolonged history of mild to moderate blood cytopenias and low risk to transformation to acute leukemia. Further research is needed to deduce the clinical significance of this subtype of MDS with this unusual RBC morphology

Abstract Transfusion of ABO-mismatched platelets is common practice. Although a unit of single donor apheresis platelet (SDP) contains as much plasma and passive anti-A or/and anti-B antibodies (allo-agglutinins) as a unit of fresh frozen plasma, ABO- mismatched platelet transfusions are considered safe and hemolytic transfusion reactions have only rarely been reported. We report a case of a 67-year old male, blood group A, with mantle cell lymphoma who received 1 SDP for chemotherapy induced thrombocytopenia. Several prior platelet transfusions had been uneventful. After 150 ml of irradiated, leukoreduced, SDP from blood group O donor had been infused, he developed nausea, light-headiness, chills, and back pain. He became bradycardic and hypotensive. Platelet transfusion was discontinued and intravenous dexamethasone and rapid saline infusion was administered. Later, his hemoglobin dropped from 10.6 to 7.4 g/dl and total serum bilirubin increased by 9 fold. A direct antiglobulin test (DAT) on post-transfusion sample demonstrated 1+ reactivity at room temperature. An elution test performed on the patient’s circulating red blood cells contained anti-A antibodies. These results are consistent with a severe acute hemolytic transfusion reaction. The donor anti-A and anti-B antibodies titers at room temperature and at 37°C were 512 and 2048 with A1 cells and 128 and 256 with B cells, respectively.This case illustrates the rare possibility of hemolytic transfusion reaction occurring as a result of ABO-mismatched platelets from group O donor with high-titer anti-A antibodies. SDP (in contrast to pooled platelets) offers the advantages of reduced risk of infection transmission and alloimmunization by limiting exposure to one donor only. However, it increases the opportunity for hemolytic transfusion reaction to occur if the donor has high-titer antibodies which would have been diluted if pooled platelets were used. As SDP use becomes more widespread, the risk of clinically significant hemolytic reactions is likely to increase, especially in seriously ill patients who tolerate these reactions particularly poorly. Moreover, increasingly, apheresis platelets are being donated by a small group of highly motivated donors, raising the chances of repeated transfusions from the same donors. We propose that the safety of ABO-mismatched platelet transfusions be re-evaluated. Effectiveness (and costs) of potential preventive strategies such as screening all SDP units for high-titer antibodies and/or plasma volume reduction of all mismatched SDP units should be studied.

Abstract Melphalan (M), Prednisone (P) and Lenalidomide (R) are active in the treatment of multiple myeloma (MM). We evaluated the toxicity and efficacy of MPR in a phase I/II trial. Results of phase I part were previously reported (Blood2006, 108(11) abst 3558). M 5 mg/m2 on days (d) 1–4, 60 mg/m2 d 1–4 and R 10 mg d 1–21 of a 28 day cycle was determined as MTD for the phase II study. Methods: Patients with previously untreated MM who were not candidates for stem cell transplantation, age ≥18 yr, ECOG PS 0–3, evaluable or measurable disease, ANC >1500, platelets > 100,000 and creatinine ≤ 3.0 mg/dL were eligible. All patients received aspirin 325 mg/d as DVT prophylaxis. Routine antibacterial prophylaxis was not used. G/GM-CSF was used at the discretion of the treating physician for neutropenia. Response rate (RR), confirmed on two consecutive determinations was the primary end-point. The study had 90% power to detect a RR of at least 45%. Results: Twenty-six patients were accrued between July ’06 and Feb ’08. The median age was 74 y (64 – 87y), 17 (65%) were men. Six, 13, 6, 1 (23%, 50%, 23%, and 4%) patients had ECOG PS 0, 1, 2, and 3, respectively. Three (12%), 11 (42%) and 10 (38%) patients were ISS stage 1, 2 and 3. A total of 106 cycles were administered; median 4.5 (1–13). Seventy-seven % patients experienced at least 1 grade 3 or 4 toxicity. Neutropenia (42%/8%) and thrombocytopenia (20%/8%) were the most common grade 3/4 adverse events followed by anemia (16%/0%), fatigue (8%/4%), rash (11%/0%), and hyperglycemia (4%/0%). There were 20 treatment delays in 9 patients, and 18 dose reductions (9 in R, 11 in M and 1 P) occurred in 7 patients, most commonly because of neutropenia or thrombocytopenia. Eleven patients (42%) required G-CSF treatment. Eleven patients are currently receiving treatment; 15 have gone off study (4 completed per protocol, 3 refused, 3 adverse effects, 3 unrelated illnesses, 2 other). Median follow up of survivors is 8 m (0 – 22m). Eighteen (69 %; 95%CI 48–86) patients had a PR or better according to International Myeloma Working Group Uniform Response criteria; 3 (12 %) CR, 5 (19 %) VGPR and 10 (38%) PR. Median number of cycles before achieving a response was 2 (1 – 11). Median Progression Free and Overall Survival is 16.7m (95% CI 11.2–23.0), and 23 m (95% CI: 22 – 23), respectively. Conclusions: MPR combination is feasible with a manageable toxicity profile and has significant activity in the treatment of patients with MM who are not candidates for stem cell transplant, with an overall response rate of 69%, and a CR plus VGPR rate of 31%. Grade 3 or 4 neutropenia and thrombocytopenia was the most common toxicity, seen in up to 50% of patients in this cohort of predominantly elderly patients. Dose reduction in R and/or M was required in 27% patients. No patient developed neuropathy or thromboembolic complications.

Abstract Background: mTOR inhibition has produced responses in mantle cell lymphoma as well as other non-Hodgkin lymphomas. This phase II study tested the oral mTOR inhibitor everolimus (RAD001, Novartis Pharmaceuticals) in simultaneous two-stage phase II lymphoma studies, including one arm involving Hodgkin lymphoma. The goals were to learn the toxicity profile and to assess the anti-tumor response. A total of 17 patients with Hodgkin lymphoma were enrolled in the uncommon arm at completion of enrollment. Methods: Patients (pts) received 10 mg PO daily for each 28 day cycle (up to 12, with a possible extension in responders) and restaged after 2, 6, and 12 cycles. The primary endpoint is the confirmed response rate, including CR, CRu or PR. Results: The median age of the 17 pts with Hodgkin lymphoma was 37 yrs (range: 27–68), with a median of 6 (range, 4–14) prior therapies. Fourteen pts (82.4%) had a prior stem cell transplant (SCT). Pts completed a median of 6 (range, 1–13) cycles of therapy. Fifteen of 17 patients were evaluable for response as of this analysis. The overall response rate was 47% (7/15), all partial responses. Ten patients are continuing on study while 6 have gone off due to disease progression and 1 due to other reasons. Common grade 3 adverse events (AEs) include thrombocytopenia (5 pts), anemia (5 pts) and alkaline phosphatase elevation (1 pts). 1 patient was reported to have grade 4 neutropenia. Conclusions: Oral everolimus has promising activity with acceptable toxicity in Hodgkin lymphoma. These results provide the rationale for additional studies with this novel class of agents and to integrate mTOR inhibitors into salvage treatment regimens for Hodgkin lymphoma.

8033 Background: MLN9708 is an oral, reversible 20S proteasome inhibitor. The feasibility of combining a proteasome inhibitor with an immunomodulatory drug and a steroid in previously untreated MM has been demonstrated with the RVD regimen. This is the first study of MLN9708 in combination with lenalidomide and dexamethasone (NCT01217957). Here we report the phase (Ph) 1 MTD and preliminary Ph 2 results. Methods: Pts with previously untreated MM, aged ≥18 yrs with measurable disease received oral MLN9708 (phase 1: 1.68–3.95 mg/m2) days 1, 8, and 15, lenalidomide 25 mg days 1–21, and dexamethasone 40 mg days 1, 8, 15, and 22, for up to twelve 28-day cycles. Primary objectives were determination of safety, MTD, and recommended phase 2 dose (RP2D) (Ph 1), and CR+VGPR rate (Ph 2). Results: At data cut-off (Dec 1, 2011), 29 pts had been enrolled (15 Ph 1, 14 Ph 2). Median age was 64 yrs (range 40–82); 69% ISS stage II/III. In Ph 1, the MLN9708 MTD was determined as 2.97 mg/m2 and the RP2D as 2.23 mg/m2; for Ph 2, the RP2D was converted to a 4.0 mg fixed dose based on population PK results. Ph 1 pts have received a median of 6 treatment cycles (range 1–11), 8 received ≥6 cycles; 6 stopped to receive ASCT, 7 are ongoing. Ph 2 pts received a median of 1 (range 1–2), all are ongoing. Grade ≥3 hematologic toxicity was reversible and included anemia (n=2) and thrombocytopenia (n=1). Grade ≥3 nonhematologic toxicity included erythematous rash, syncope, and vomiting (2 pts each). All-grade drug-related peripheral neuropathy was seen in 6 pts (21%), including grade 2 with pain in 2 (both Ph 1 at doses above the MTD). Two pts discontinued due to AE; there were 5 pts who had serious drug-related AE (all Ph 1). Of 19 response-evaluable pts (Ph 1 + Ph 2), all achieved ≥PR, including 5 CR (1 sCR), 4 VGPR, and 10 PR; all remain in response with duration of confirmed response of up to 9.5 months. Of 4 response-evaluable Ph 2 pts, 1 has achieved VGPR and 3 PR to date. Conclusions: Oral MLN9708 plus lenalidomide and dexamethasone appears well tolerated with manageable toxicity. These data show antitumor activity at the RP2D in pts with previously untreated MM, with ≥PR in all pts to date.

1048 Background: Previous trials with p plus gem show significant antitumor activity in MBC. Nab-p has lower toxicity compared to p with similar therapeutic effects. This study assessed proportion of confirmed tumor responses and toxicity profile of this combination. Methods: Single stage phase II study enrolled patients (pts) with measurable MBC eligible for 1st line chemotherapy (chemo). Pts treated (tx) with nab-p 125 mg/m2 followed by gem 1,000 mg/m2 days 1 and 8 of 21 day cycle. Dose modifications allowed for both drugs. Primary endpoint was proportion of confirmed responses; complete response (CR), partial response (PR) by RESIST criteria on two consecutive evaluations at least 6 weeks apart. 41 evaluable pts required to test the null hypothesis that the confirmed response rate was at most 40% against the alternative of at least 60%. Results: 50 pts enrolled 11/05 to 5/06. Median age 56 (29–86). 40 pts (80%) visceral disease, 34 (68%) ER positive, 28 (56%) PR positive, 1 (2%) HER2 positive, 1 (2%) prior trastuzumab treatment (tx). 25 (50%) had prior adjuvant chemo, including 15 (30%) prior taxane. Median follow-up 7 mos (3–12). Median of 6 cycles of tx completed (1–15). 41 pts discontinued tx, due to progression (41%), pt refusal (27%), adverse events (17%), alternate tx (7%). Confirmed response rate 48% (95% CI: 34–63%): 4 CR, 20 PR. Median duration of response 6.6 mos (95% CI: 5.7, not reached), median progression-free survival (PFS) 7.9 mos (95% CI: 5.3–9.3). Rate of overall survival (OS) at 6 mos 91% (95% CI: 84–100%), 6 month PFS 58% (95% CI: 45–74%). Median OS not reached. 30 pts required dose delays on total of 52 cycles, mostly due to hematologic AEs. 9 pts (19%) had grade (gr) 4 adverse event (AE), 32 (67%) had maximum gr 3 AE. Most common gr 3/4 AEs: neutropenia (52%), fatigue (26%), anemia (14%), dyspnea (14%), thrombocytopenia (12%). Four pts had gr 3 neuropathy. Conclusions: Nab-paclitaxel in combination with gem has clear activity in patients with MBC. Manageable hematologic, but no other significant toxicities. Study ongoing to correlate efficacy with caveolin-1 levels with results expected Spring 2007. Follow up study of this combination with an anti-angiogenesis agent is in development. No significant financial relationships to disclose.

Extramedullary multiple myeloma is defined by the presence of plasma cell infiltration outside of the bone marrow. It is associated with a poor prognosis and resistance to therapy and is often associated with high-risk cytogenetics. Aggressive relapsed and refractory extramedullary multiple myeloma is often treated with salvage infusional chemotherapy to achieve rapid disease control. Commonly used regimens include DCEP, CVAD, and VTD-PACE. While VTD-PACE contains bortezomib and thalidomide which have potent antimyeloma activity, the advent of novel agent therapy with proteasome inhibitors and immunomodulatory agents being used in the first-line setting has resulted in many patients being refractory to bortezomib by the time they are treated with VTD-PACE. Herein, we discuss two cases of aggressive relapsed, high-risk, bortezomib-refractory extramedullary multiple myeloma treated with KRD-PACE and review the available clinical data on salvage chemotherapy regimens used in relapsed refractory myeloma.

Background: HIT has the potential to improve the health of individuals and the performance of providers, yielding improved quality, cost savings, and greater engagement by patients. Minorities face a variety of financial and nonfinancial barriers to obtaining appropriate and timely health care. Studying the behavior of minority patients will help understand the specific challenges to HIT adoption and engagement in patients and their caregivers. We undertook a patient-reported survey to explore these factors and understand any underlying disparities. Methods: A 21-question paper-based anonymous questionnaire addressing the responder's attitudes, understanding and needs of the electronic medical record (EMR) was used. This was administered to adult patients with lymphoid malignancies and their caregivers at two campuses of Mayo Clinic in Florida (San Pablo and inner-city; Riverside). Categorical and continuous variables between white and minority responders as well as minority responders from the two campuses were compared using the Chi-square test and U Mann Whitney with a significance level of 0.05. Results: The survey was completed by 1004 patients and caregivers. Average time taken to complete the survey was approximately 11 minutes. Of the responders, 716 (71%) were whites and 271 (27%) were minorities, with race-ethnicity not reported by 17 (2%). Among minorities, 30% (n=80) were seen at San Pablo and 64% (n=173) at the Riverside campus, with 6% at a multiple myeloma patient-caregiver symposium. Whites were older (median age 66 vs. 62 years, p<0.001) and had higher education (undergraduate or higher 60% vs. 52%, p=0.023) as compared to minorities. Minority survey responders had lesser access to a computer (84% vs. 92%, p<0.001) or a smartphone (85% vs. 89%, p=0.08). (Figure 1a) Exploring HIT awareness showed that only 51% of minorities were registered to use the EMR as compared to 72% whites (p<0.001) and a lesser number of minorities even knew that they could access their medical records electronically (81% vs. 92%, p<0.001). even among those registered to use the EMR, 13% minorities reported never using the EMR as compared to 4% whites (p<0.001). Encouragingly, a higher willingness to learn about and use the EMR was noted among minorities who were more interested in knowing how to access the EMR (43% vs. 36%, p=0.04), would increase the use of EMR if someone explained how to use it (17% vs. 10%, p=0.002), and if someone explained about test results (14% vs. 6%, p<0.001). (Figure 1b) Comparing minorities among the two campuses, those from the San Pablo campus were older (67.5 yrs. vs. 57 yrs., P<0.001), had higher level of education (undergraduate or higher in 66% vs. 47%, p=0.005) and a higher number had access to a computer (94% vs. 78%, P=0.001) as compared to those from the inner-city Riverside. (Figure 2a) Minorities from the Riverside campus were less likely to be aware of the current HIT EMR (75% vs. 93%, p<0.001), were significantly less likely to have registered for the patient EMR portal (38% vs. 76%, p<0.001) and never used the EMR even if registered for it (22% vs. 7%, p=0.01) as against those from the San Pablo campus. Minorities at the Riverside campus had a higher interest in knowing how to access the EMR (47% vs. 33%, p=0.03). Linguistic barriers were more striking in the inner-city Riverside campus where a higher proportion of minority survey responders noted that it would help if the EMR was in their primary language (66% vs. 48%, p=0.006). (Figure 2b) Conclusions: Minorities continue to struggle with suboptimal utilization of healthcare system including that of HIT. There are barriers related to access, awareness, language and understanding of HIT that prevent the integration and participation of racial-ethnic minorities in the healthcare system to a full extent. It is important to qualify the minorities better, since there exist HIT utilization disparities even within the minorities from different areas representing socioeconomic and demographic differences. Efforts need to be made at several levels to help the minorities overcome educational, access and linguistic barriers so that they can benefit from advancements in the healthcare system. Disclosures Ailawadhi: Celgene: Consultancy; Amgen: Consultancy, Research Funding; Janssen: Consultancy, Research Funding; Pharmacyclics: Research Funding; Cellectar: Research Funding; Takeda: Consultancy. Manochakian:Novocure: Membership on an entity's Board of Directors or advisory committees; Takeda: Membership on an entity's Board of Directors or advisory committees; AstraZeneca: Membership on an entity's Board of Directors or advisory committees; Guardant Health: Membership on an entity's Board of Directors or advisory committees. Chanan-Khan:AbbVie: Research Funding; Xencor: Research Funding; Pharmacyclics: Research Funding; Merck: Research Funding; Jansen: Research Funding; Mayo Clinic: Employment; Ascentage: Research Funding; Millennium: Research Funding.

Background: Survival of patients with MM has improved significantly over recent years due to therapeutic advancements. Population-level data has shown that a substantial proportion of MM patients may die early after their diagnosis and real-world evidence regarding early mortality (≤6 months after MM diagnosis) has been conflicting. We evaluated the trends, causes and predictors of early mortality among newly diagnosed MM (NDMM) patients in the real world. Methods: We used Surveillance Epidemiology and End Results (SEER) database, to identify all adult patients (age ≥ 18 years) who were diagnosed of MM from 1975 to 2015. Data on age, gender, geographical regions, year of diagnosis, survival time, vital status at 6 months after diagnosis and cause of death (COD) were extracted. Ages were categorized as quartiles while period of diagnosis for multivariate analysis were categorized by running a joinpoint analysis separately for the all-cause and myeloma-specific mortality. To look at trends in early mortality, we estimated proportion of deaths for each year and plotted it as a scatterplot against the year of diagnosis and fitted it in the join point regression model by applying the NCI's Join point Regression Program, Version 4.5.0.1. By this method we identified the year(s) when a trend change was produced, calculated the annual percentage change (APC) in rates between trend-change points, and the average annual percentage change (AAPC) in the whole period studied. To correct for known errors with COD attribution, we applied a special COD variable which has been recently developed by the SEER program to indicate if the death was due to the primary cancer diagnosis or other causes. A multivariate analysis was conducted to analyze the predictors of all-cause and myeloma-specific mortality within 6 months by fitting into logistic regression model. Results: Of 90,975 NDMM patients, early mortality due to any cause was noted in 18,810 (20.7%). For the whole cohort, median age was 68 years, majority of the patients were males (53.1%), of non-Hispanic white (NHW) race/ethnicity (65.5%) and were diagnosed in SEER registries in the western region (53.9%). An overall survival (OS) of ≤6 months was seen for approximately 22% of NHW but 18% of non-Hispanic blacks (NHB) and Hispanics. OS of ≤6 months was seen in 10% of patients in the first quartile of age (<60 years) as compared to 36% of patients in the 4th quartile (≥78 years). By geographical regions, 23.3% of MM patients in the Midwest had ≤6-month OS as compared to only 18.4% in the Northeast. The most common early mortality COD was MM in 74.7% of the patients, followed by cardiovascular (11.9%), other (8.8%) and infectious causes (3.1%). The overall pattern of COD at 6 months was similar to COD at any time in the whole cohort. Joinpoint regression analysis of trends data resulted in 1 joinpoint for all-cause 6-month mortality while 2 joinpoints were noticed for myeloma-specific mortality (Figure 1). A steep decline in mortality was noticed in the more recent time periods for both, all-cause (2006-2015) and myeloma-specific (2003-2015) mortality rates with an AAPC of 3.93 (95%CI p<0.001) for all-cause and 3.95 (95%CI p<0.001) for myeloma-specific mortality. Prior to 2006, the decrease in all-cause mortality was only modest whereas decrease in myeloma-specific mortality witnessed a sharp decline from 1975-1987 as well. On MVA, advanced age (p<0.001), NHW race/ethnicity and Midwest geographic region (p<0.001) were associated with higher likelihood of mortality while female gender (p<0.001), more recent period of diagnosis (p<0.001), and "other" races (not NHW, NHB or Hispanics) (p<0.001) were associated with lower risk of mortality. Conclusion: Our results show that early mortality due to MM remains high despite therapeutic improvements and there are significant gender, racial and geographic disparities for this phenomenon. However, significant declines in all-cause and MM-specific early mortality in the more recent times are promising. The risk of early mortality is higher among men, older patients, NHW and Midwest region. A focus on interventions to reduce early mortality, including early diagnosis, use of efficacious anti-MM therapy and aggressive supportive care should be directed at groups with high-risk for early mortality from MM. Disclosures Chanan-Khan: Millennium: Research Funding; Ascentage: Research Funding; AbbVie: Research Funding; Xencor: Research Funding; Pharmacyclics: Research Funding; Merck: Research Funding; Jansen: Research Funding; Mayo Clinic: Employment. Ailawadhi:Celgene: Consultancy; Pharmacyclics: Research Funding; Amgen: Consultancy, Research Funding; Takeda: Consultancy; Cellectar: Research Funding; Janssen: Consultancy, Research Funding.