Citations

4109 Background: Prognostic nomograms have been developed in various cancers, including ovarian, breast, and gastrointestinal; however, there is limited information on nomograms in MPC. The large, phase 3 MPACT study of nab-P + Gem vs Gem alone for the treatment of MPC provides a robust database for the development of a nomogram to predict OS using baseline patient variables. Methods: A multivariable Cox model was created from MPACT data using factors that were significantly predictive of OS in univariable analysis or considered clinically important (stepwise selection to remain in model). From the Cox model, a nomogram was derived that assigned points equal to the weighted sum of relative significance of each variable. The nomogram was internally validated using bootstrapping, a concordance index (c-index), and calibration plots. Results: Data from all 861 pts were used. Seven of the 34 considered variables were retained in the multivariable analysis (Table; all factors significant at the P < 0.01 level, except for analgesic use [ P = 0.07]). The resulting nomogram was able to distinguish low (n = 216), medium (n = 430), and high (n = 215) risk groups (c-index: 0.69; CI: 0.67-0.71) with median OS values of 12.9, 8.2, and 3.7 months, respectively. Calibration curves showed that the nomogram’s predicted probabilities were mostly consistent with observed probabilities for 6-, 9-, and 12-month OS. Conclusions: Treatment arm, Karnofsky performance status (KPS), neutrophil-to-lymphocyte ratio (NLR), albumin level, sum of longest tumor diameters (SLD), and presence of liver metastasis were the key predictors of OS. This nomogram, which will be presented in visual format in the final presentation, may help physicians and pts make informed treatment decisions. Clinical trial information: NCT00844649. [Table: see text]

TPS3102 Background: Multiple myeloma (MM) is largely incurable despite available therapies, including immunomodulatory drugs (IMiDs), proteasome inhibitors (PIs), and monoclonal antibodies (mAbs). As most MM patients (pts) eventually have relapsed/refractory (RR) disease, there is an unmet need. Myeloma cells upregulate PD-L1 [Liu et al, 2007]. Nivolumab (nivo), an immuno-oncology mAb, binds PD-1 on T cells and natural killer cells and inhibits signaling by PD-L1–expressing tumor cells, thus augmenting antitumor immunity. Nivo monotherapy has shown acceptable safety and modest clinical activity in RRMM [Lesokhin et al 2016]. Daratumumab (dara) is a cytolytic mAb that targets CD38+ myeloma cells, inducing antibody‐dependent cellular cytotoxicity, complement-dependent cytotoxicity, antibody-dependent cellular phagocytosis, and apoptosis, and may have an immunomodulatory role via depletion of CD38+ immune-suppressor cells [Dimopoulos et al 2016]. Dara is approved as monotherapy (US: after ≥3 prior lines of therapy; EU: after a prior PI and IMiD and progression on last therapy), and with lenalidomide/dexamethasone (dex) or bortezomib/dex (US: after ≥1 prior line of therapy for both combinations). Combining PD-1 and CD38 mAbs, immunotherapies with different mechanisms of action, may overcome resistance and improve outcomes. This phase 1 safety study (NCT01592370) includes multiple cohorts of nivo as monotherapy or in combination regimens across RR hematologic malignancies; 2 MM cohorts will evaluate nivo plus dara, with or without pomalidomide and dex. Methods: Eligible pts are aged ≥18 y, with RRMM after ≥2 prior therapies. Pts are RR to their last regimen, RR to prior IMiD and PI therapy, and agreed to bone marrow aspiration. Primary outcome of safety/tolerability will be measured by incidence of drug-related adverse events (AEs), serious AEs, and laboratory test abnormalities. Secondary endpoints include minimal residual disease, overall response rates and duration of response, and progression-free survival. Study funding: BMS. Writing support: C Tomas, Caudex, funded by BMS. Clinical trial information: NCT01592370.

3551 Background: Cutaneous toxicity is a known adverse effect of multikinase inhibitors and has been associated with clinical outcomes (Granito 2016). In the phase 3 CORRECT trial (NCT01103323), the multikinase inhibitor regorafenib significantly improved overall survival (OS) vs placebo in patients with mCRC (hazard ratio [HR] 0.77, 95% CI 0.64, 0.94; 1-sided P = 0.0052). This retrospective analysis explored whether HFSR was associated with outcomes in CORRECT. Methods: Patients randomized to receive regorafenib 160 mg/day during the first 3 weeks of each 4-week cycle were divided into subgroups based on whether or not they had HFSR. Estimates of OS and progression-free survival (PFS) (95% CI) were calculated using the Kaplan–Meier method. Patients who were randomized, but not treated, were included in the no HFSR group for the analysis of survival. Results: Of the 505 randomized patients, 500 received at least one dose of regorafenib. Among the treated patients, 47% (n = 235) had HFSR of any grade and 17% (n = 83) had HFSR grade 3. Of the patients who had HFSR, 69% (n = 162) had their first HFSR event (any grade) during the first treatment cycle. There was some imbalance in baseline characteristics between groups (Table). Survival was improved in patients who had HFSR at any time vs those who did not (Table). The OS benefit was also observed in patients who had the first HFSR event during Cycle 1 vs those who did not (median OS 7.2 vs 5.7 months; HR 0.66, 95% CI 0.51, 0.87). Conclusions: This post-hoc exploratory analysis suggests that patients who had HFSR had a greater treatment benefit from regorafenib. Since HFSR and survival are post-baseline assessments, results may be confounded by baseline factors or other unknown factors. Clinical trial information: NCT01103323. [Table: see text]

3090 Background: The correlation of immune response (IR) and clinical benefit in patients (pts) who receive active immunotherapy remains controversial. CMB305 is an active immunotherapy designed to generate and expand anti-NY-ESO-1 T cells, consisting of LV305 (a DC targeting lentiviral vector encoding NY-ESO-1) and a boost (NY-ESO-1 recombinant protein plus GLA-SE, a TLR-4 agonist). Phase 1 studies of LV305 and CMB305 evaluated the IR and survival of pts with NY-ESO-1 positive tumors. Methods: 62 pts with recurrent NY-ESO-1 expressing (by IHC) solid tumors were enrolled in 2 Phase 1 trials (31 pts each: 43 sarcomas, 12 ovarian, 4 melanomas, and 3 NSCLC). Peripheral blood was assessed pre and post therapy for anti-NY-ESO-1 T cell responses (IFNγ ELISPOT) and antibodies (Ab) (ELISA) (n = 62), TCR-β CDR3 repertoire and conserved (public) TCR (n = 28), and antigen spreading (ELISA, ELISPOT) (n = 23). Relationship between progression-free survival (PFS), overall survival (OS) and up to 15 biomarkers were analyzed by the Kaplan-Meier method. Results: Preexisting anti-NY-ESO-1 T cells and Ab were identified in 34% and 40% of pts. LV305 and CMB305 induced (de novo or increased) anti-NY-ESO-1 T-cells in 51% and 65% pts, and anti-NY-ESO-1 Ab responses in 13% and 68% pts, respectively. LV305 and CMB305 induced antigen spreading in 17% and 36% pts. TCR clonality increased after CMB305 as compared to LV305. The TCRβ-CDR3 amino acid sequences of three NY-ESO-1 specific TCR clones in a patient with a > 2-year response were fully conserved in 20/28 patients. Biomarkers showing the strongest association with OS and PFS were pre-existing anti-NY-ESO-1 T-cell response and Ab, integrated IR (T cells and antibodies), and NY-ESO-1 expression level. Presence of public TCRs was associated with longer survival, particularly in pts with high ( > 50%) NY-ESO-1 expression. Conclusions: While LV305 and CMB305 are both immunogenic, CMB305 resulted in a stronger and broader integrated IR, including antigen spreading. The three identified survival variables, preexisting and treatment-induced NY-ESO-1 IR and the presence of public TCRs, warrant validation in randomized studies of CMB305.

530 Background: Therapeutic efficacy of H involves activation of the immune system.Age-dependent progressive deterioration of the immune response is referred as immunosenescence. In HER2+ breast cancer, the effects of aging and the ability of H to activate endogenous anti-tumor immunity is unknown. A previous report from HERA trial showed no significant increase in risk of early recurrence with age (≤ or > 40). We evaluated the long-term outcome of HER2+ patients (pt) related to age and immune landscape. Methods: 1,392 pt from N9831 trial were evaluated. Stromal tumor infiltrating lymphocytes (sTIL) were evaluated in H&E slides. Lymphocyte predominant breast cancer (LPBC) was defined as sTIL ≥ 50%. Molecular TIL (mTIL) and immune subset signatures were evaluated using NanoString research CodeSets. Cox proportional hazard ratio (HR) was used for analysis. Results: There were 1,111 (79.8%) pt > 40 years old (yo) and 281 (20.2%) pt ≤ 40 yo. Younger age was significantly associated with hormone receptor positivity (p 0.00011) and luminal B subtype (p 0.011). With a median follow up of 10.6 years, there was no significant difference in long-term outcome among pt ≤ 40 vs. > 40 yo who received H (HR = 0.88 (0.62-1.24), p = 0.45). Similar findings were observed when age was dichotomized at 50 and 60 yo. While there was no association between sTIL and age, a small but significant increase in mTIL CD45 expression with age (p 0.003) was observed. Similar small increases in cytotoxic cell (p 0.007) and T cell (p 0.015) immune scores were also observed with increasing age. Among pt who received chemotherapy alone, pt > 40 yo with LPBC (n 55) had excellent outcome with 92.4% recurrence free survival (RFS) at 10 years. However, there was no significant difference in RFS among pt ≤ 40 yo with or without LPBC. Conclusions: Among pt treated with H, there was no significant difference in outcome related to age. In contrast to a decline expected from immunosenescence, we observed small but significant increases in mTIL signatures for total lymphocytes, cytotoxic, and T cell subsets. Among patients who received chemotherapy alone, our data suggested that pt > 40 yo with LPBC had excellent prognosis, compared to pt > 40 yo without LPBC.

3010 Background: To test the safety and efficacy of the CAR T cell modality in relapsed/refractory multiple myeloma (MM), we have designed a second-generation CAR construct targeting B cell maturation antigen (BCMA) to redirect T cells to MM. bb2121 consists of autologous T cells transduced with a lentiviral vector encoding a novel CAR incorporating an anti-BCMA scFv, a 4-1BB costimulatory motif and a CD3-zeta T cell activation domain. We will report updated safety and efficacy following initial results (Berdeja et al, ENA 2016). Methods: CRB-401 (NCT02658929) is a multi-center phase 1 dose escalation trial of bb2121 in patients with relapsed and/or refractory MM who have received ≥ 3 prior regimens, including a proteasome inhibitor and an immunomodulatory agent, or are double-refractory, and have ≥ 50% BCMA expression on plasma cells. Peripheral blood mononuclear cells are collected via leukapheresis. Patients undergo lymphodepletion with Flu (30 mg/m 2 ) / Cy (300 mg/m 2 ) daily for 3 days then receive 1 infusion of bb2121. The study follows a standard 3+3 design with planned dose levels of 5, 15, 45, 80 and 120 x 10 7 CAR+ T cells. Results: As of November 18, 2016, 11 patients had been infused with bb2121 in the first 4 dose cohorts, and 9 patients had reached at least 1 month of follow-up. As of data cut-off, no dose-limiting toxicities and no > Grade 2 neurotoxicities or cytokine release syndrome (CRS) had been observed. Grade 1-2 CRS had been reported in 8/11 (73%) treated patients. All patients treated with doses of 15.0 x 10 7 or higher remained on study and the overall response rate (ORR) in the 6 evaluable patients at these doses was 100%, including 2 sCRs and 2 MRD-negative responses (1 sC, 1 VGPR). CAR+ T cell expansion has been demonstrated consistently. An additional 6 months of follow up on previously reported results and initial data from an additional ~10 patients will be presented. Conclusions: bb2121 shows promising efficacy at dose levels above 5 x 10 7 CAR+ T cells, including 2 sCRs and ongoing clinical responses at 6 months, with mild and manageable CRS to date. These initial data support the potential of CAR T therapy with bb2121 as a new treatment paradigm in MM. Study sponsored by bluebird bio. Clinical trial information: NCT02658929.