Singh, Preet Paul

Abstract Purpose A protocol was developed to evaluate the value of an NK-1 receptor antagonist for preventing nausea and vomiting resulting from highly emetogenic chemotherapy when an olanzapine-based antiemetogenic regimen was used. Materials and Methods A221602, a prospective double-blind, placebo-controlled clinical trial, was developed to compare 2 ­olanzapine-containing antiemetic regimens, one with an NK-1 receptor antagonist (aprepitant or fosaprepitant) and one without. Trial patients had a malignant disease for which they received intravenous highly emetogenic chemotherapy (single day cisplatin ≥ 70 mg/m2 or doxorubicin plus cyclophosphamide on 1 day). Patients on both arms received commonly administered doses of a 5-HT3 receptor antagonist, dexamethasone, and olanzapine. Additionally, patients were randomized to receive an NK-1 receptor antagonist (fosaprepitant 150 mg IV or aprepitant 130 mg IV) or a corresponding placebo. The primary objective was to compare the proportion of patients with no nausea for 5 days following chemotherapy between the 2 study arms. This trial was designed to test for the noninferiority of deleting the NK-1 receptor antagonist, with noninferiority defined as a decrease in freedom from nausea by less than 10%. Results A total of 690 patients were entered on this trial, 50% on each arm. The proportion of patients without nausea for the complete 5-day study period was 7.4% lower (upper limit of the one-sided 95% confidence interval was 13.5%) in the arm without an NK-1 receptor antagonist compared with the arm with an NK-1 receptor antagonist. Conclusion This trial did not provide sufficient evidence to support that deletion of the NK-1 receptor antagonist was as good as keeping it, as a part of a 4-drug antiemetic regimen for highly emetogenic chemotherapy (ClinicalTrials.gov Identifier: NCT03578081).

12107 Background: Chemotherapy-induced nausea and vomiting (CINV), a major adverse effect of cancer treatment, can attenuate life quality. Multiple international antiemetic guidelines have recommended a 4-drug regimen (corticosteroid, 5-HT3 receptor antagonist, NK-1 receptor antagonist, and the antipsychotic medication olanzapine) for decreasing highly emetogenic CINV. This raised a question regarding whether antiemetic therapy regimens for highly emetogenic chemotherapy could be de-intensified, decreasing cost and side effects from anti-emetogenic agents. Methods: Study A221602 was developed to compare two olanzapine-containing antiemetic regimens, one with the use of an NK-1 receptor antagonist (aprepitant or fosaprepitant) and one without. Patients (pts) received intravenous highly emetogenic chemotherapy as either 1) cisplatin, given on one day, at a dose of ≥ 70 mg/m2, with or without other chemotherapy agent(s) or 2) doxorubicin (60 mg/m2) plus cyclophosphamide (600 mg/m2) on one day. Pts in both arms received 1) a 5-HT3 receptor antagonist (palonosetron 0.25 mg IV or ondansetron 8-16 mg IV or 16-24 mg PO,) on day one, 2) dexamethasone (12 mg PO, day one followed by 8 mg PO, days 2-4), and 3) olanzapine (10 mg/day PO, days 1 to 4). On Day 1, all agents were given prior to chemotherapy, with the exception that olanzapine could be taken prior to chemotherapy or at bedtime. Additionally, pts were randomized to receive an NK-1 receptor antagonist (fosaprepitant 150 mg IV or aprepitant 130 mg IV; 346 pts) or a matching placebo (344 pts). The primary objective was to compare the proportion of patients with no nausea for 5 days following receipt of their chemotherapy between the two study arms. Results: While there was no suggestion of outcome differences on day 1 between the two study arms, fewer pts (7.4%; upper limit of the 95% confidence interval was 13.5%) without NK-1 receptor antagonists were without nausea for the complete 5-day study period. Conclusions: Per the study design, created to exclude a 10% benefit for NK-1 receptor antagonists, the results did not provide sufficient evidence to reject the hypothesis that the 4-drug regimen was superior to the protocol 3-drug regimen (p=0.24). Support: UG1CA189823; Clinical trial information: NCT03578081.

8600 Background: The combination of bortezomib, dexamethasone, thalidomide, cisplatin, doxorubicin, cyclophosphamide and etoposide (VDTPACE) was developed as an intense regimen for disease control prior to tandem transplantation for multiple myeloma (MM) in total therapy protocols. The regimen is very effective in this setting, and since has also been used in the relapsed setting. We examined the outcomes of a set of patients undergoing VDTPACE therapy for relapsed MM at our institution. Methods: We identified 71 patients who received VDTPACE for relapsed MM, at Mayo Clinic from 7/2006 to 7/2012. Plasma cell leukemia was excluded. All data was extracted from clinical records. Results: The median age of patients was 59 years (range, 39-80); 48 (67.6%) were male. The median time from diagnosis to initiation of VDTPACE was 38.2 months (range, 2-125). The median number of cycles given was 1 (range, 1-9). The overall response rate after one cycle was 57.1% (14.3% VGPR, 22.2% PR and 20.6% MR) in the 63 patients in whom the response was evaluable. The median overall survival (OS) post-VDTPACE was 8.2 months (95% CI, 5.7-10.9). Eighteen (25.4%) patients went on to autologous stem cell transplantation (SCT), and 7 (9.9%) received matched allogeneic SCT following VDTPACE, and the median OS post-VDTPACE was significantly longer for these groups compared to those who were not transplanted (15.3 and 20.5 months, respectively vs 5 months, p-value <0.001). Thirty eight of 66 (57.6%) patients were rehospitalized after initial admission for infusion therapy for a median duration of 6 days (range, 1-26). The median platelet and red cell transfusions were 4 (range, 0-21) and 5 (range, 0-22) units, respectively. Renal toxicity was seen in 13/62 (21%) patients and 27/65 (41.5%) patients developed neutropenic fever. The median duration to absolute neutrophil and platelet count recovery was 18 (range 12-42) and 27 (range, 12-42) days, respectively. Three (4.2%) patients died within 30 days and 11 (15.5%) within 8 weeks of initiating VDTPACE. Conclusions: VDTPACE is an effective therapy in relapsed MM but is associated with significant morbidity and short-term mortality. It appears to be more effective when followed by an autologous or allogeneic SCT.

8587 Background: Patients (pts) undergoing an auto-transplant (ASCT) for multiple myeloma (MM) have disease assessment approximately 100 days later. This result may direct decisions of further therapy versus observation. However, some pts continue to experience a decline in their serum or urine monoclonal (M) - protein after day 100 without more therapy. Little is known about the prevalence and significance of this phenomenon. Methods: We identified 903 MM pts who underwent ASCT within 12 months (mos) of diagnosis (Dx) at our institution. Their day 100 post-ASCT M-protein from serum and urine electrophoresis with immunofixation as well as serum free light chains were compared to subsequent values during follow-up. The IMWG criteria were used to evaluate response. Results: Of the pts included, 510 (56%) were male and median age at ASCT was 59 (range 28-76). Median follow up from Dx and ASCT was 82 mos (95% CI; 75 - 86) and 74 mos (95% CI; 70 - 79) respectively. There were 453 (50%) pts seen in follow-up who had not achieved a CR at Day 100 nor initiated on maintenance therapy. Of these pts, 167 (37%) had a further decrease in their M-protein after day 100 at a median of 9.4 mos (95% CI; 8 – 10) post-ASCT. Given the time taken for maximal response, we assessed patients’ clinical response at one year post-ASCT. Compared to patients who did not have further clinical response between day 100 and 1 year, pts experiencing further response had a longer time to next therapy (TTNT) (43 mos vs. 17 mos, P < 0.001) as well as overall survival (OS) (96 mos vs. 62 mos, P < 0.001). Positive predictors for continued response included having an IgG isotype, evolution from a pre-existing MGUS, smoldering myeloma or solitary plasmacytoma and a Day 100 bone marrow plasma cell < 3%. In a multivariate analysis, elevated creatinine at Dx and lack of continued response predicted for shorter TTNT and OS post-ASCT. Older age and high-risk MM by FISH also predicted a shorter OS. Conclusions: In MM pts undergoing ASCT, continued M - protein response was seen in a third of the pts lacking a CR at day 100. This phenomenon appears prognostic and must be considered when interpreting studies evaluating post-ASCT response and the need for further therapy.

6614 Background: CDKN2A locus on chromosome 9p21 is implicated in altered molecular pathways leading to leukemogenesis as a tumor suppressor by inhibiting the proliferative kinases CDK4/CDK6 and blocking the cell-cycle division during G1/S phase. Deletion of CDKN2A locus is frequently seen in ALL, although prognostic significance remains unclear. Methods: Retrospective chart review at Mayo Clinic between 8/2005-11/2011 was performed on adult B-ALL patients (pts) who underwent fluorescent in situ hybridization (FISH) studies at diagnosis. Survival estimate was calculated using Kaplan-Meier statistics. Event-free survival (EFS) was defined as time between diagnosis and induction failure, relapse, or death, while overall survival (OS) as time between diagnosis and death Results: Out of 27 pts who had FISH studies for CDKN2A, 15 (56%) pts had CDKN2A deletions (8 homozygous/7 hemizygous), while 12 (44%) pts had diploid cytogenetics (none had abnormalities of MLL, bcr/abl fusion or CDKN2A deletion on FISH). There was no significant difference in median age, gender, WBC, hemoglobin, platelet count, LDH or complete remission (CR) rate (p=0.924) between the two groups. CDKN2A-deleted pts had higher circulating blasts (34% vs 16%, p=0.017). Of 13 CDKN2A-deleted pts, chromosome 9p was intact in 9 (69%) pts using routine cytogenetics. In the CDKN2A-deleted group, allogeneic stem cell transplant (ASCT) was performed in 7 (47%) pts at first CR and 1 at second CR. Four of 15 (27%) CDKN2A-deleted pts also had bcr/abl fusion. Five-year EFS was poorer in CDKN2A-deleted pts compared to normal FISH group (22% vs 60%, p=0.042), while 5-year OS was 72% vs 60% (p=0.308), respectively. Excluding the 4 bcr/abl fusion pts, 5-year EFS was still statistically significant (p=0.008). In the CDKN2A deleted group, both 5-yr EFS (80% vs 0%, p=0.006) and OS (100% vs 42%, p=0.016) were significantly superior in pts who received ASCT. Conclusions: CDKN2A deletions predict earlier relapse in pts with B-ALL. Majority of 9p abnormalities were only found by FISH testing and missed by conventional cytogenetics. In CDKN2A deleted pts, ASCT gave overall survival advantage. Further investigation through larger cohorts of pts is needed to validate these findings.

Abstract Background The presence of circulating plasma cells (PCs) in multiple myeloma (MM) is a known poor prognostic marker. Prior studies have utilized a technically challenging slide-based immunofluorescence technique to detect and quantify the presence of circulating PCs which was not widely adapted in clinical practice. More recently, the routine use of flow cytometry has provided the opportunity to quantitatively assess circulating PCs in MM patients with relative ease. We report the prognostic value of quantifying circulating clonal PCs using multiparametric flow cytometry in MM patients with relapsed disease. Methods We evaluated all MM patients seen at the Mayo Clinic, Rochester from 2009 to 2011 with previous or ongoing relapsed disease and who had their peripheral blood samples evaluated by flow cytometry. Each blood sample had its peripheral blood mononuclear cells isolated by ficoll gradient and stained with antibodies to CD45, CD19, CD38, CD138 and cytoplasmic Kappa and Lambda Ig light chains. A six-color multi-parameter flow cytometer (Becton Dickinson FacsCantos II) was used to examine each sample with a target of detecting 150,000 events (cells) that was then analyzed using the Facs Diva Software. Plasma cells were selectively analyzed through combinatorial gating using light scatter properties and CD38, CD138, CD19, and CD45. Normal PC's were then separated from clonal plasma cells based on the differential expression of CD45, CD19 and polytypic Ig light chains. The clonal plasma cells detected were reported out as the number of clonal events/150,000 collected total events. For those samples where less than 150,000 events were gated or examined, the number of final clonal events was adjusted to 150,000 events. Survival analysis was performed by the Kaplan-Meier method and differences in survival assessed using the log rank test. Results There were 647 consecutive patients with a history of treated MM who had a peripheral blood flow cytometry as part of their routine clinical evaluation. The median patient age was 62 years (29-88) and 55% were male. The median time since diagnosis was 12 months (range: 1-363) and the median number of lines of treatment received was 2 (1-11). There were 81 (13%) patients with clonal circulating PCs with a median of 368 cells (4 – 133,464). The 2-year overall survival (OS) for the 81 (13%) patients with any circulating PCs was 17% compared with 65% for those with none (P<0.001). The presence of circulating clonal PCs was associated with high-risk disease by FISH (P <0.001) as well as higher PCLI (P <0.001). We then correlated the presence of circulating clonal PCs with the disease status at flow cytometry assessment. Among the study patients, only 145 (22%) had actively relapsing disease with the remaining 502 (78%) in a plateau including patients in CR. Circulating clonal PCs were more likely to be detected in patients with actively relapsing disease compared with plateau phase (43% vs. 4%, P < 0.001); none of the CR patients and <5% of the remaining plateau phase patients had any detectable circulating clonal PCs. We then restricted the analysis to the actively relapsing patients; the best cutoff predicting 1-year mortality by ROC analysis was 100 events. Based on this, we defined >=100 events as a cutoff for defining the prognostic role of circulating clonal PCs in actively relapsing patients. The median OS for those with >=100 clonal PC events was 12 months compared with not reached for those with <100 events (p<0.001; Figure 1 ). Among patients with actively relapsing disease, >= 100 circulating PCs was associated with a higher ISS stage, plasma cell labeling index and bone marrow PC% compared with those with <100 circulating PCs. In a multivariable model, only LDH > 222 (HR: 2.08, P = 0.045) and >= 100 circulating PCs (HR: 2.48, P = 0.048) were found to adversely affect OS in actively relapsing patients. Conclusion The utilization of flow cytometry to quantify circulating clonal PCs in patients with relapsed MM appears to have significant prognostic relevance. In patients with actively relapsing disease, >= 100 circulating PCs predicted for worse OS with a median of 12 months. Future studies are needed to determine if this would allow an opportunity to develop a more risk adapted approach for patients with relapsed disease. Disclosures: Kumar: Celgene: Consultancy, Research Funding; Millennium: Consultancy, Research Funding; Onyx: Consultancy, Research Funding.

Abstract Abstract 1756 Background: We hypothesized that surrogate markers of host immune response to clonal myeloproliferation predict survival in myeloid malignancies. Because of immediate applicability to current practice, we chose plasma immunoglobulin free light chain (FLC) concentration (B-cell activation marker) as the biomarker of interest. Patients and Methods: Two independent cohorts of patients with primary myelofibrosis (PMF) or myelodysplastic syndromes (MDS) were studied. Both groups were fully annotated for karyotype. Kappa (κ) and lambda (λ) FLC were measured by a quantitative nephelometric assay. Patients with monoclonal FLC were excluded. Results: Above upper normal-limit values for κ or λ FLC was documented in 33% of 240 study patients with PMF and 46% of 74 patients with MDS. Multivariable analysis revealed significant associations between increased FLC and elevated creatinine, as well as advanced age in PMF (p=0.0004) and hemoglobin <10 g/dL in MDS (p=0.005). Increased FLC predicted shortened survival in both PMF and MDS, independent of age, creatinine, and other conventional risk factors. Receiver-operating characteristic analysis-based cutoff levels for κ plus λ total FLC delineated risk groups with highly significant differences in overall survival (Figure); International Prognostic Scoring System-adjusted HR (95% CI) in PMF was 1.9 (1.3-2.7) and in MDS 6.3 (2.7-16.6). No correlations were seen with leukemia-free survival, karyotype or JAK2, MPL or IDH mutations. Conclusions: Elevated plasma FLC concentration predicts inferior survival in both PMF and MDS. Its lack of correlation with leukemia-free survival and tumor-specific genetic markers suggests a primarily host-driven biological phenomenon that might be applicable to other malignant and non-malignant conditions. Disclosures: No relevant conflicts of interest to declare.

Abstract Background Prior studies have shown that presence of increased circulating plasma cell (PCs) identified using a slide-based immunofluorescence method is an adverse prognostic marker for overall survival in multiple myeloma (MM), and increases the risk of progression in patients with MGUS and smoldering MM. However, utility in clinical settings has been limited by the cumbersome nature of the test and lack of widespread availability. We studied the prognostic value of circulating PCs using sensitive multiparametric flow cytometry that enable us to quantitatively assess circulating PCs in patients with MM. Methods We evaluated all newly diagnosed MM patients seen at the Mayo Clinic, Rochester from 2009 to 2011 who had their peripheral blood samples evaluated by flow cytometry prior to therapy. Patients with plasma cell leukemia were excluded. Each blood sample had its peripheral blood mononuclear cells isolated by ficoll gradient and stained with antibodies to CD45, CD19, CD38, CD138 and cytoplasmic Kappa and Lambda Ig light chains. A six-color multi-parameter flow cytometer (Becton Dickinson FacsCantos II) was used to examine each sample with a target of detecting 150,000 events (cells) that was then analyzed using the Facs Diva Software. PCs were selectively analyzed through combinatorial gating using light scatter properties and CD38, CD138, CD19, and CD45. Normal PC's were then separated from clonal PCs based on the differential expression of CD45, CD19 and polytypic Ig light chains. The clonal PCs detected were reported out as the number of clonal events/150,000 collected total events. For those samples where less than 150,000 events were gated or examined, the number of final clonal events was adjusted to 150,000 events. Survival analysis was performed by the Kaplan-Meier method and differences assessed using the log rank test. Results There were 158 consecutive patients with newly diagnosed MM who had their peripheral blood evaluated via flow cytometry as part of their routine clinical evaluation. The median age of this group was 66 years (39-95) and 59% were male. At the time of this analysis, 25 patients had died and the 2-year OS rate for the cohort was 83%. The 2-year OS for the 89 (55%) patients with any circulating PCs was 76% compared with 91% for those with none (P=0.02). The median number of circulating clonal PCs was 33 (range, 0-46,413)/ 150,000 gated events. Using a ROC analysis the best cutoff predicting 1 and 2-year mortality were 435 and 376 events, respectively. Based on this, we defined >=400 events as a cutoff for defining patients with high-risk disease. The median time-to-next-treatment (TTNT) for patients with circulating clonal PCs >=400 (n=37, 23%) was 14 months compared with 26 months for those with <400 events (n=121, 77%) (p<0.001; Fig 1a). The median OS for those with >=400 clonal PC events was 32 months compared with not reached for those with <400 events (p<0.001; Fig 1b). Patients with >= 400 circulating PCs had higher ISS stage, creatinine, LDH, PC labeling index and bone marrow PC% compared with the rest. In a univariate analysis examining the presence of circulating PC >=400, LDH, labeling index, marrow PC% and FISH risk status, only circulating PCs were prognostic for overall survival (P = 0.0011). Among the 89 patients with any circulating PCs, 26 (30%) had CD45+ clonal PCs. The TTNT in patients with a clonal CD45+ population was inferior to those with no CD45+ clonal PCs in circulation (median 11 vs. 19 months, P = 0.034) as was the OS (P = 0.039). Conclusion Quantitative estimation of circulating clonal PCs in patients with newly diagnosed MM is a powerful predictor of early relapse from therapy and mortality. A cutoff of >=400 clonal events/150,000 gated mononuclear events predicts for a median TTNT of 14 months and OS of 32 months. This parameter is more powerful than the current high-risk parameters such as FISH risk status as well as traditional high-risk markers such as ISS stage and LDH levels, and is able to identify a group of patients with particularly poor outcome. Disclosures: Kumar: Celgene: Consultancy, Research Funding; Millennium: Consultancy, Research Funding; Onyx: Consultancy, Research Funding.

Abstract Abstract 2848 Background: Massive splenomegaly (MS) (>1500 grams) is a rare and overtly symptomatic manifestation of myelofibrosis. While splenectomy may provide palliation, the procedure may be technically difficult and lead to high operative risks. We retrospectively analyzed myelofibrosis patients who underwent splenectomy for MS to analyze complications and outcomes. Methods: Retrospective chart review of all patients with myelofibrosis at our institution who underwent splenectomy for MS between 1998 and 2006 was performed. MS was defined as post-operative spleen weight measured to be more than 1500 grams. A total of 63 patients with MS had splenectomy for palliation of pressure symptoms or transfusion requiring cytopenias during that time period. Kaplan Meier statistics and log rank test were used for survival analysis. Results: Mean age of the study patients was 64.9 ± 9.5 years and 56% of those were males. Commonest pre-splenectomy symptoms were left upper quadrant abdominal discomfort (85.7%), fatigue (81%), early satiety (76.2%), anorexia (54%) and weight loss (39.7%). Packed red blood cell and platelet transfusions were required pre-operatively in 36/63 (57.1%) and 10/63 (15.9%) of patients, respectively. Median spleen weight was 2805 (range 1540 – 13085) grams. Perioperative complications occurred in 15 patients (23.8%) including infection (= 4 patients [6.3 %]), thrombosis (= 6 patients [9.5%]), or bleeding (= 9 patients [14.3%]), 2 of which (3.2% of all patients) were fatal. Median post-splenectomy survival (PSS) of these patients was 16.4 (0–110) months and 5 year PSS was 20%. Excluding patients lost to follow up, 43/55 (78.2%), 33/55 (60%) and 26/54 (48.1%) patients were alive at 6 months, 1 year and 2 years after splenectomy, respectively. At 1 year, out of 33 alive patients, 24 (72.7%) patients had relief of pressure symptoms. Out of 36 patients who were red cell transfusion dependent pre-splenectomy, 13 were alive and 6/13 (46.2%) patients were transfusion independent at 1 year. Similarly, out for 10 patients requiring platelet transfusions pre-operatively, 5 were alive and 3/5 (60%) were transfusion independent after 1 year follow up. PSS after splenectomy for MS was decreased in patients that required pre-operative packed red cell transfusions (median 9.3 months vs. 32.1 months, p= 0.033) (Fig 1), but not affected by other factors including age (>64 years), pre-operative thrombocytopenia (<100 × 109/uL), prolonged operative time (>90 minutes) or larger spleen size (>2800 grams). Conclusions: Splenectomy is relatively safe and effective palliative treatment for pressure related symptoms and refractory cytopenias related to MS in patients with myelofibrosis. Pre-operative anemia requiring red cell transfusions may predict advanced disease and is associated with inferior PSS. Disclosures: No relevant conflicts of interest to declare.

Abstract Background Conflicting results have emerged, especially with respect to the impact on overall survival (OS), from trials evaluating lenalidomide maintenance (LM) therapy after induction therapy alone or post-autologous stem cell transplant (ASCT) in multiple myeloma (MM). We performed a systematic review and meta-analysis of existing outcome data from LM trials to evaluate role of lenalidomide as maintenance strategy in MM. Patients and methods A comprehensive search of electronic databases and abstracts through June 2013 was performed to identify randomized controlled trials (RCTs) that compared LM vs. placebo/no maintenance. Single arm studies were excluded. Pooled hazard ratio (HR) or odds ratio (OR) estimates with 95% confidence intervals (CIs) were calculated using the random-effects model for clinical endpoints of progression free survival (PFS), OS, response rate (RR) and adverse events (AEs), including second primary malignancies (SPMs). Analyses were performed using Comprehensive Meta-Analysis Software Version 2. We assessed between-study heterogeneity with the Cochran Q test and quantified its extent with the I2 statistic. Results Overall, five RCTs, with data extractable from four phase III trials (3 publications and 1 abstract) were identified (n= 1935). All studies were RCTs with an adequate randomization. MRC MM XI study was excluded from analyses as survival data are not available. Two placebo controlled trials (IFM 05-02, CALGB 100104) addressed the role of LM post-ASCT, one placebo-controlled trial (MM-015) studied LM therapy in the non-transplant setting and the remaining trial (RV-MM-PI209) had a 2 X 2 design comprising of both ASCT and non-transplant randomized arms followed by a second randomization of LM versus no maintenance. There was no heterogeneity for estimate of PFS results (Cochran Q, p=0.68; I2=0%), but considerable heterogeneity for estimate of OS (Cochran Q, p=0.09; I2= 55%), among the studies. There was significant prolongation of both PFS (HR 0.49, 95% CI, 0.41–0.58, p<0.001) and OS (HR 0.77, 95% CI, 0.62–0.95, p=0.013) with LM vs. placebo/no maintenance (Figure 1). Best response during maintenance was reported only in 2 studies and odds of responding (very good partial response or better) were not significantly different with LM (OR 1.28, p=0.3). Grade 3-4 AEs data were available from 3 trials for calculation of pooled OR with LM compared with placebo. We observed a nearly two-fold increase in the risk of SPMs with LM (OR 1.99; 95% CI, 1.31–3.04; p=0.001). Patients on LM were more likely to have grade 3-4 AEs than placebo: neutropenia (OR 4.9, p<0.001), thrombocytopenia (OR 2.7, p<0.001), fatigue (OR 2.3, p=0.01) and venous thromboembolism (OR 3.2, p=0.02). Odds of discontinuing treatment were also significantly higher in patients on lenalidomide (OR 2.9, p<0.001). Conclusions Meta-analysis of RCTs demonstrates significant improvement in PFS and modest improvement in OS with LM. There is an increased risk of grade 3-4 adverse effects, including SPMs with LM. Substantial heterogeneity for estimate of OS among protocols is a limitation of this analysis. Lack of uniform access to lenalidomide upon disease progression in the placebo/no maintenance arms of the constituent studies should be taken into account while interpreting aggregate effect estimates for OS in this meta-analysis. OS: Cochran Q p=0.09, I2=55%, substantial heterogeneity PFS: Cochran Q p=0.68, I2=0%, minimal heterogeneity Disclosures: Off Label Use: Lenalidomide for maintenance therapy in multiple myeloma. Kumar:Merck: Consultancy, Honoraria; Celgene: Consultancy, Research Funding; Millennium: The Takeda Oncology Company: Research Funding; Novartis: Research Funding; Genzyme: Research Funding. Dispenzieri:Celgene, Millenium, Jansenn, Pfizer: Research Funding. Bergsagel:Onyx: Consultancy. Lacy:Celgene Corporation: Research Funding.