ALBERT

Description: Background: Classical Hodgkin lymphoma (HL) has a unique histopathology, with rare malignant Hodgkin/Reed Sternberg (HRS) cells surrounded by a strong inflammatory cellular component in the tumor microenvironment (TME). Although extensive studies describe the interdependence of the HRS cells and the TME, the impact of HL on systemic immunity has not been well described. Here, we develop a new approach, employing a recently commercialized single cell cytokine secretion platform (IsoLight) to assess, precisely and comprehensively, the function of peripheral blood mononuclear cells (PBMCs) in HL patients. Methods: Cells were selected from 4 HL patients: 2 newly diagnosed who had a complete response (CR) to therapy standard first line therapy, and 2 relapsed patients who progressed on second line chemotherapy (PD). Cryopreserved PBMCs from a pre-treatment and post-treatment time point for each patient were thawed, rested overnight, stimulated with PMA/ionomycin and loaded into the IsoLight single cell cytokine secretion system. IsoLight captures single cells in microwells; as cytokines are secreted, they are bound by antibodies lining the microwell cover. Bound cytokines are then revealed by fluorescent secondary antibodies and photos are taken at various time points to assess fluorescence intensity, which corresponds to the relative amount of each cytokine secreted. Twenty thousand cells can be assayed per sample simultaneously. Results: The percentage of cytokine-secreting cells varied dramatically by donor (12%-48%), with monofunctional cells making only TNFa, MIP1b, or IL-15 dominating the functional landscape. Polyfunctional cells, capable of making three or more cytokines simultaneously represented only 0.1-7% of the cells in each sample, but there were more of these cells, and each secreted higher levels of cytokines, in individuals who responded to therapy with a CR. Responders also secreted higher levels of IL2, Perforin, IL4, IL12, MIP1a, and TNFb (p values ranging from 0.005 to 0.03), and lower levels of IL9 and IL22 (p=0.0028 and 0.021, respectively), compared to non-responders at diagnosis. Responders lost expression of IL4, IL7, and MIP1a over the course of treatment (pre- vs post-treatment, p=0.01 to 0.05), while non-responders gained cells that expressed IL4, IL5, IL10, IL17, and TNFb from diagnosis to end of treatment (p=0.001 to 0.05). Conclusion: This work represents an important methodological advance in immune monitoring for hematologic malignancies. Single cell cytokine secretion technology measures more cytokines simultaneously than flow cytometry, providing a sample-sparing and comprehensive overview of the functional landscape of immune cells in a patient. Moreover, the technology provides cell-by-cell information about cytokine secretion, unlike Luminex. Our work represents the first application of this technology to HL, which we use to define, for the first time, the particular combinations of 32 cytokines that can be secreted by individual immune cells. We also identify candidate cytokines whose frequency at diagnosis may predict treatment outcome, and reveal changes in cytokine levels over treatment time that may distinguish patients destined to relapse. Immunotherapy may impact PBMC function differently, this may partially explain the high efficacy of this therapy in the relapsed population. The impact of immunotherapy on cytokine levels is currently under investigation by our group in a larger study. Other important questions which are under investigation include the impact of prior chemotherapy on cytokine profiles in relapsed patients, and whether certain cytokines which increase during treatment may be a surrogage for tumor bulk in patients with PD. Cytokines elevated in patients with poor responses to treatment include IL9, IL10, IL17, and IL22, which may present attractive drug targets if validated in our larger ongoing follow-up study. Disclosures Diefenbach: Bristol-Myers Squibb: Consultancy, Research Funding; MEI: Research Funding; Denovo: Research Funding; Genentech: Consultancy, Research Funding; Incyte: Research Funding; LAM Therapeutics: Research Funding; Merck: Consultancy, Research Funding; Seattle Genetics: Consultancy, Research Funding; Trillium: Research Funding; Millenium/Takeda: Research Funding. Hymes:Celgene: Consultancy. Martin:Janssen: Consultancy; Sandoz: Consultancy; Karyopharm: Consultancy; Celgene: Consultancy; Teneobio: Consultancy; I-MAB: Consultancy. Ruan:Celgene: Consultancy, Honoraria, Research Funding; AstraZeneca: Consultancy, Honoraria; Seattle Genetics: Consultancy, Honoraria, Research Funding; Janssen: Consultancy, Honoraria; Pharmacyclics LLC, an AbbVie company: Research Funding; Juno: Consultancy; Kite: Consultancy. Leonard:Miltenyi: Consultancy; Akcea Therapeutics: Consultancy; Sandoz: Consultancy; AstraZeneca: Consultancy; Bayer Corporation: Consultancy; Epizyme, Inc: Consultancy; BeiGene: Consultancy; Miltenyi: Consultancy; ADC Therapeutics: Consultancy; Akcea Therapeutics: Consultancy; Sandoz: Consultancy; Celgene: Consultancy; Epizyme, Inc: Consultancy; Karyopharm Therapeutics: Consultancy; AstraZeneca: Consultancy; Bayer Corporation: Consultancy; Celgene: Consultancy; Sutro Biopharma: Consultancy; Merck: Consultancy; Genentech, Inc./F. Hoffmann-La Roche Ltd: Consultancy; Gilead: Consultancy; Karyopharm Therapeutics: Consultancy; Sutro Biopharma: Consultancy; Nordic Nanovector: Consultancy; ADC Therapeutics: Consultancy; MorphoSys: Consultancy; Gilead: Consultancy; Nordic Nanovector: Consultancy; BeiGene: Consultancy; Merck: Consultancy; Genentech, Inc./F. Hoffmann-La Roche Ltd: Consultancy; MorphoSys: Consultancy.

Description: Phosphorylation of signal transducer and activator of transcription 3 (STAT3) is essential for cell survival, proliferation and differentiation. STAT3 phosphorylation results from signaling by cytokines and growth factors, and constitutive STAT3 activity is characteristic of a number of human malignancies, including Cutaneous T Cell Lymphoma (CTCL). Furthermore, we now know that STAT3 is also required for the initiation and maintenance of the Th17 differentiation program. Th17 cells are a subset of CD4 T helper cells that have been implicated in chronic inflammatory conditions like rheumatoid arthritis and psoriasis. Mycosis fungoides (MF) and the leukemic variant of this disease, Sezary syndrome (SS), are the most frequently encountered forms of CTCL and in both of these diseases, the cell of origin – as far as the type of Teffector cell involved, has not been defined. Recent results from our laboratory and that of our colleagues have lead us to believe that Th17 cells may either be the cells of origin in CTCL or may act as critical mediators of chronic inflammation that creates a favorable environment for tumor growth in the context of this malignancy. In an effort to elucidate the role of STAT3 as a transforming factor in T cell malignancies, we generated a mouse model wherein T cell specific expression of a hyper-active STAT3 mutant protein (STAT3C) leads to the development of a lymphoproliferative disease that is highly reminiscent of CTCL. We are now taking advantage of this unique mouse model, patient biospecimens and carefully characterized CTCL cell lines to dissect the role of STAT3 signaling cascade in the malignant transformation and maintenance of CTCL. Most recently, our attention has been focused on understanding the mechanism of action of epigenetic therapy in the form of histone deacetylase inhibitors (HDACi), which is highly effective in the treatment of CTCL. We hypothesize that HDAC inhibitors affect the STAT3 mediated Th17 differentiation and thus have clinical efficacy in this disease. In addition to the regulation of chromatin accessibility through the regulation of histone modifications, HDACi have also been implicated in a less conventional mode of protein regulation directly influencing STAT3 serine phosphorylation. To dissect the action of HDACi on malignant cells, we took advantage of CTCL cell lines and cultured these with and without Romidepsin, which is an effective HDAC inhibitor used in clinic. MyLa2059 and PB2B are MF cell lines with skin only phenotype whereas SeAx and SeZ4 are SS cell lines. The cells were cultured for 48 hours with no Romidepsin, 5nm and 50 nm Romidepsin. After 48 hours, cells were fixed and permeabilized using BD fix-perm protocol. Cells were then stained to assess Serine 727 STAT3 and Tyrosine Y705 STAT3 phosphorylation and analyzed using flowcytometry. We found that Romidepsin affected serine phosphorylation exclusively in CTCL cell lines (Figure 1). This leads us to believe that STAT3 serine phosphorylation might play an important role in lymphomagenesis and can act as a potential therapeutic target. The role of serine phosphorylation in the context of STAT3 signaling is hotly debated and we are now attempting to characterize the role of Serine STAT3 phosphorylation in the context of CTCL. We are also hoping to recapitulate these observations in patients' biospecimens collected before and after treatment with HDAC inhibitors. We will also study the role of serine phosphorylation in STAT3 activity in carcinogenesis using our mouse model with phenotypic and pathological characteristics similar to CTCL. We hope that these studies will advance our knowledge about the role that Stat 3 signaling plays in MF/SS malignant transformation and cancer progression and help us develop target specific treatment options for the clinical practice. Disclosures: Hymes: Celgene: Consultancy.

Description: Background: In Hodgkin lymphoma (HL) the malignant Hodgkin Reed-Sternberg (HRS) cells comprise only a small fraction of the total cellular tumor population. These HRS cells orchestrate an inflammatory microenvironment of reactive cells that propagate a permissive milieu for HL growth, contributing to an ineffective local anti-tumor immune response. The peritumoral CD4 and CD8 T cells in HL patients show high expression of the receptor programmed death-1 (PD-1), involved in the functional impairment and “exhaustion” of T cells. Growing data suggests that this HL-mediated immune suppression may have effects that extend beyond the tumor microenvironment. High systemic levels of inflammatory cytokines and chemokines in HL patients has been reported. We characterized the systemic immune profile of HL patients with both newly diagnosed (ND) and relapsed (R) disease. Methods: Informed consent for correlative blood testing was obtained from patients with ND (n=8) or R (n=5) HL treated at the NYU Perlmutter Cancer Center or NY Presbyterian/Weil Cornell since January of 2013. Blood samples were drawn pre-treatment, and at sequential timepoints during and after therapy. Peripheral blood mononuclear cells (PBMC) were isolated using Ficoll separation method and cells were frozen for subsequent analysis. The frozen PBMC were then stained with fluorescent-conjugated antibodies against T cell surface molecules in 10-color FACS analysis. The analyses were performed after gating live cells for CD4, CD8 and memory and effector T cell markers. Patient samples were compared to normal controls matched for age and sex (n=18). Results: The median HL patient age was 32 (22-72), and 8 subjects were male. All ND HL patients were treated with ABVD (range 4-6 cycles) +/- consolidative radiation; R patients had median of 3 prior therapies. One patient out of 5 had prior autologous stem cell transplant (SCT), and 1 had prior allogeneic SCT, but was not on immunosuppression. Eight patients (6ND, 2R) responded to therapy (8 CR); 5 patients (1ND, 4R) progressed on therapy or had stable disease. HL patients displayed a high frequency of the exhaustion marker PD-1 on CD4 central memory T cells (CD4+CD45RO+CD27+) compared to normal matched controls (NC): mean 41, standard error (SE) 4.8 for HL patients vs. mean 22.2, SE 1.3 for NC (p = 0.0002) (Figure 1A). PD-1 expression was similarly elevated on CD8 central memory T cells (CD8+CD45RO+CD27+) of HL patients: mean 55, SE 3.3 vs. NC: mean 40, SE 3.3 (p = 0.003) (Figure 1B). HL patients also displayed an increased frequency of PD-1 expression on CD27 negative CD4 effector T cells: mean 43, SE 4, vs. NC: mean 28.5, SE 2.4 (p = 0.003) (Figure 2). In 4 of the HL patients who responded to therapy, PD-1 expression on central memory CD4+ cells declined after therapy: mean 30.1 vs. mean increase of +2.67 in 3 patients who progressed on therapy (p〈 0.009). A higher number of subjects in prospective analysis is underway, to confirm whether a response to therapy may be correlated with a reversal of the suppressed phenotype of T cells in these patients. Conclusion: HL patients have evidence of chronic activation/exhaustion in their central memory and effector T cells, suggesting that ineffective immune clearance of the HRS cells may be a systemic rather than local phenomenon. In patients with progressive disease for whom this phenotype persists it is worthy of investigation whether this immune dysfunction is a cause or consequence of resistance to therapy. This may be rationale for immune targeted therapy in patients with relapsed or resistant disease. Figure 1. Evidence for increased levels of T cell exhaustion in central memory T cells of HL patients. PBMC were stained with specific fluorescent conjugated antibodies against T cell markers (CD3, CD4, CD8) together with differentiation markers (CD45RO, CD27) and PD1 and analyzed using FACS (LSR-II). The proportion of PD1+ T cells were determined in: A) CD4+CD45RO+CD27+ and B) CD8+CD45RO+CD27+ T cells. Figure 1. Evidence for increased levels of T cell exhaustion in central memory T cells of HL patients. PBMC were stained with specific fluorescent conjugated antibodies against T cell markers (CD3, CD4, CD8) together with differentiation markers (CD45RO, CD27) and PD1 and analyzed using FACS (LSR-II). The proportion of PD1+ T cells were determined in: A) CD4+CD45RO+CD27+ and B) CD8+CD45RO+CD27+ T cells. Figure 2. Evidence for increased levels of T cell exhaustion in effector memory CD 4+ T cells of HL patients. PBMC were stained with specific fluorescent conjugated antibodies against T cell markers (CD3, CD4) together with differentiation markers (CD45RO, CD27) and PD1 and analyzed using FACS (LSR-II). The proportion of PD1+ T cells was determined in CD4+RO+CD27- T cells Figure 2. Evidence for increased levels of T cell exhaustion in effector memory CD 4+ T cells of HL patients. PBMC were stained with specific fluorescent conjugated antibodies against T cell markers (CD3, CD4) together with differentiation markers (CD45RO, CD27) and PD1 and analyzed using FACS (LSR-II). The proportion of PD1+ T cells was determined in CD4+RO+CD27- T cells Figure 3 Figure 3. Disclosures No relevant conflicts of interest to declare.

Description: Background: The CALGB 10603 RATIFY trial demonstrated that midostaurin, a multi-targeted small molecule FLT3 inhibitor, when given in combination with standard 3+7 induction consisting of daunorubicin 60mg/m2 for 3 doses and cytarabine 200mg/m2/day for 7 days significantly prolonged overall survival (OS) compared to placebo plus standard induction in FLT3 positive acute myeloid leukemia (AML) patients. The safety and efficacy of the other anthracycline commonly utilized in standard AML induction, idarubicin, has not been evaluated to date. Methods: A single-center retrospective review from May 2017 to July 2018 was performed. Patients were included if they had a diagnosis of FLT3 positive AML and received induction chemotherapy with idarubicin. The primary outcome was incidence of adverse effects attributed to midostaurin. Grading of adverse effects was in accordance with the Common Terminology and Criteria of Adverse Effects (CTCAE) version 5. Additional outcomes included OS, relapse-free survival (RFS), similar as in the Ratify trial, as well as detection of FLT3 on subsequent bone marrow biopsies. Results: Ten patients were included. All patients received induction therapy with idarubicin 12mg/m2 for 3 doses and cytarabine 100mg/m2/day for 7 days. Median age was 53 years (range: 33 to 66) and 6/10 were male. Eight of 10 patients exhibited internal tandem duplication (ITD) on diagnosis; two had FLT3 tyrosine kinase domain (TKD) D835. Eight patients had diploid cytogenetics; the other two patients had core-binding factor AML. Midostaurin was initiated on day 8 of induction in all but 2 patients, who started on day 11 and 12, respectively. Nine of ten patients completed all 28 planned doses of midostaurin. All patients received antifungal prophylaxis with micafungin throughout the course of midostaurin. The median time from day 1 of induction to neutrophil (〉500/µl) and platelet (〉100,000/µl) recovery was 23 days and 26 days, respectively. Granulocyte colony stimulating factor (G-CSF) was initiated for all patients after day 14 bone marrow biopsy, as per institutional procedure. Four of 10 patients experienced an adverse event attributed to midostaurin. Maculopapular rash was observed in 3 patients a median of 5 days after midostaurin initiation: 2 of 3 patients had a grade 2 rash and continued therapy with topical steroids; one patient had a grade 3 rash and discontinued midostaurin after 17 of 28 planned doses. A grade 1 drug fever was attributed to midostaurin in a fourth patient. Fevers persisted despite neutrophil recovery and subsequently dissipated after completion of the final midostaurin dose. Persistent FLT3 mutation was detected in 4/9 (1 not reported) day 14 bone marrow biopsies but was negative in 9/10 pts on day 28. The lone positive FLT3 result on day 28 occurred in a patient with refractory disease (〉5% blasts) necessitating salvage therapy. Notably, this patient only completed 17 of 28 planned doses. All other patients exhibited a complete response (CR) on day 28. The median follow-up time was 243 days (range: 57 to 394). Nine patients are alive at the time of reporting. Six patients proceeded to allogeneic transplantation -one death was attributed to transplant-related complications, occurring in the same patient needing salvage and reduced duration midostaurin. Two patients relapsed a median of 184 days after start of induction -both FLT3-ITD positive and neither having undergone allogeneic transplant prior to relapse. Conclusions: Midostaurin in combination with idarubicin-based induction 3+7 therapy in this first case series appears to be safe. While the incidence of rash was higher (30%) than reported in RATIFY, this only resulted in discontinuation of therapy in one patient. Although patient numbers are limited, 90% achieved a FLT3 negative CR after completion of induction therapy and six patients proceeded to allogeneic transplantation. A confounding variable includes the routine use of G-CSF, which likely contributed to the shorter duration from induction to neutrophil recovery observed compared to RATIFY (23 days vs 26 days). The influence of G-CSF use on outcome is uncertain, however represents an interesting observation. A randomized, prospective trial comparing midostaurin in combination with idarubicin versus daunorubicin at both 60mg/m2 and 90mg/m2 is warranted to establish the optimal anthracycline induction therapy for FLT3 mutated AML patients. Disclosures Al-Homsi: Celyad: Membership on an entity's Board of Directors or advisory committees. Diefenbach:Denovo: Research Funding; Merck: Consultancy, Research Funding; Acerta: Research Funding; Incyte: Research Funding; Trillium: Research Funding; Millenium/Takeda: Research Funding; Bristol-Myers Squibb: Consultancy, Research Funding; Genentech: Consultancy; Seattle Genetics: Consultancy, Research Funding.

Description: Background: Over-expression of Aurora A has been shown in many subtypes of non-Hodgkin lymphoma (NHL) and is a negative prognostic factor in mantle cell lymphoma (MCL), where expression correlates with tumor proliferative signature (Ki67), and may contribute to cell cycle dysregulation. Aurora A regulates chromosomal stability through effects on mitosis and cell cycle regulation. Bortezomib increases both pro-apoptotic proteins and cell cycle dependent kinase inhibitors. Thus, there is a strong scientific rationale to combine Aurora A kinase and proteasome inhibition. We evaluated the combination of alistertib, an oral Aurora A kinase inhibitor, with bortezomib and rituximab in a CTEP sponsored phase 1 study of patients with relapsed /refractory (RR) low grade NHL or MCL. Methods: Patients with RR low grade NHL or MCL, who had received at least one line of immuno-chemotherapy were treated according to a 3 + 3 design with 3 dose escalation cohorts (DL1, DL2, and DL3) to identify the recommended phase 2 dose (RP2D) with an expansion cohort at the RPTD. Patients received alisertib 30mg BID (DL1 and DL2) or 40mg BID (DL3) on days 1-7, bortezomib subcutaneous 1mg/m2 (DL1) or 1.3 mg/m2 (DL2 and DL3) on days 1, 8, and 15, and rituximab 375mg/m2 on day 1 for the first 8 cycles, and subsequently q 3 months. An expansion cohort was treated at DL3. Treatment cycles were 28 days. Dose Limiting Toxicity (DLT) was defined within the first cycle of therapy. Results: A total of 24 patients at 3 sites were enrolled between 06/2013 and 02/2018, including DL1: 3, DL2: 7; DL3: 14. Median age was 64 (range 46-87). Twelve patients (50%) were male. Nineteen patients had low grade NHL, and 5 patients had MCL. Performance status was 0-1 in 22 (92%) patients, and 2 in 2 (8%) patients. Dose Escalation and Safety: The recommended phase II dose (RP2D) was DL3, with the only DLT in DL3 (prolonged grade 4 neutropenia). The most common grade 3-4 AEs at all dose levels were neutropenia (34%), thrombocytopenia (13%), anemia (8%), and fatigue (8%); grade 2 alopecia occurred in 37% (Table 1). Other grade 4 AEs included hypertension & hypotension, respiratory failure, hyperkalemia and hyperuricemia (n=1 (4%) each); one patient died of GI hemorrhage deemed unrelated to study treatment (NSAID overdose). Grade 3 neutropenia and leukopenia was seen in 4 patients and thrombocytopenia in 1. There was one grade 3 AE each of: hypertension, pulmonary hypertension, pneumonitis, fatigue, diarrhea, infusion reaction, lung infection, and febrile neutropenia. Efficacy: A total of 18 patients were evaluable for response. Five patients were not evaluable for response for the following reasons: withdrawal of consent (3, 1 each DL1, 2, 3), non-compliance (1 DL2), and death unrelated to study drug (1, DL2). For patients who were treated with at least 3 cycles of therapy, and had a restaging CT scan, the median duration of therapy was 142 days, (range 29-590). Response was seen in 7/18 evaluable patients for an overall response rate (ORR) of 38%; 4 of 18 patients (22%) had a complete response (CR), and 4 of 18 patients (22%) had a partial response (PR). An additional 8/18 patients (44%) had stable disease for a clinical benefit rate of 88%. With a median follow-up of 30.9 months, the median PFS was 6.9 months (95%CI 4.1-28.0) and median OS not reached. OS at 3 years was 64.9% (95% CI 44.1-95.3%; Fig. 1). Median duration of response was 14.1 mo (95% CI 2.63-NR). Conclusion: Alisertib in combination with bortezomib and rituximab is a well-tolerated regimen with significant and durable activity in heavily pretreated low grade NHL and MCL. The RP2D is DL3 (R + bortezomib 1.3mg/m2 and 1.8 alisertib 40mg bid). A correlation of Aurora Kinase tumor expression with clinical outcome is planned. Disclosures Barta: Janssen: Membership on an entity's Board of Directors or advisory committees; Merck, Takeda, Celgene, Seattle Genetics, Bayer: Research Funding. Diefenbach:Merck: Consultancy, Research Funding; Genentech: Consultancy; Incyte: Research Funding; Millenium/Takeda: Research Funding; Bristol-Myers Squibb: Consultancy, Research Funding; Acerta: Research Funding; Seattle Genetics: Consultancy, Research Funding; Trillium: Research Funding; Denovo: Research Funding.

Description: Abstract 4961 Background: Romidepsin, a histone deacetylase (HDAC) inhibitor has demonstrated overall response rates in CTCL of 35% in 2 separate phase 2 studies [enrolling 71 patients (pts) at 8 centers and enrolling 96 pts at 33 institutions worldwide] and is approved by the United States FDA for the treatment of CTCL in patients with progressive, persistent, or recurrent disease on or following at least 1 prior systemic therapy. Reports of clinical efficacy and toxicity outside the context of clinical trial are absent. Here, we report the largest single center experience of romidepsin in the treatment of relapsed or refractory CTCL following approval of the drug. Methods: Between June 2010 and June 2011, 11 pts with Mycosis Fungoides (MF) and Sezary syndrome (SS), the two predominant subtypes of CTCL, who had adequate organ function, ECOG PS 0 or 1 and who had failed 1 prior systemic therapy, were treated with single agent, romidepsin. Patients received romidepsin at a dose of 14 mg/m2 on days 1, 8, and 15 every 28 days. Response was measured with a quantitative composite which included skin involvement [severity-weighted assessment tool (SWAT) or erythroderma scale], lymph node involvement (CT or MRI scans), and blood involvement (circulating Sézary cells assessed by flow cytometry). Pruritus in pts with CTCL was assessed using a 10-point scale; relief defined as reduction of pruritus score of 3 points in pts with baseline score 3. ECGs were obtained (pre-/post-infusion on all treatment days in cycle 1, and on day 1 of subsequent cycles) to evaluate for potential cardiac toxicity. Serum Mg, Ca and K levels were measured with each treatment and supplementation given as necessary. Results: Ten pts (5 men and 5 women) with MF (Stages IB-IVA, 6 pts had advanced stage disease i.e ≥IIB) and 1 pt with SS with median age of 73 (59-77), median number of prior therapies 3 (1-8) were treated with a median of 4 cycles (1-12) of romidepsin. Nine pts received at least 2 cycles of therapy. Overall response rate (ORR) [CR (complete response) plus PR (partial response)] was 70% by composite assessment in the 10 assessable pts including 4 CR's and 3 PR's with 3 pts demonstrating disease progression. Treatment in 1 pt was discontinued due to concern for cardiovascular toxicity that was probably not treatment related. For these 10 pts, the median duration of response was 32 weeks (20-46) with ongoing responses in 4 of them (20-44 weeks).Of the 7 responders, 4 continue to maintain their response despite discontinuation of the drug with a mean of 21 weeks (4-42), 2 pts relapsed within 8 weeks of discontinuation of treatment and were retreated with romidepsin with subsequent achievement of similar response as with initial treatment and 1 pt continues to receive treatment. Three of the 7 responders, had received prior treatment with other HDAC inhibitors including vorinostat, panobinostat and belinostat. All 7 responders had prutitus relief. Median time to response was 6 weeks (3.5-8) and median time to progression was 38 weeks (27-42).The most frequent drug-related adverse events (AEs) (all grades, all cycles) were generally mild and included: nausea (50%; none with grade 3), fatigue (50%; none with grade 3) and decreased platelets (40%; none with grade 3). On review of 166 ECGs performed on the 11 pts, small clinically insignificant prolonged QT interval and corrected prolonged QTc interval was observed in only 1 pt. No pts had QTcF values〉480 milliseconds and there was no change from baseline of 〉 20 milliseconds. The ECG values were not associated with functional cardiovascular changes or symptoms and did not lead to discontinuation of drug in any patient. Conclusion: In our single-center experience, romidepsin has greater ORR (70%, including 40% CR and 30% PR) than what has been reported in the phase II studies (ORR of 35%, including 6% CR and 28% PR) with comparable patient populations. Drug-related AEs were generally mild and consisted of GI disturbances and thrombocytopenia none of which were ≥ grade 3 compared to the 12% grade 3 and 4 toxicity observed in the phase 2 studies. We noted no clinically significant cardiovascular AEs in our pts in contrast with the previous phase 2 trials. Our experience also suggests that resistance to one HDAC inhibitor should not preclude treatment with romidepsin in CTCL pts. Disclosures: No relevant conflicts of interest to declare.

Description: Background: In Hodgkin lymphoma (HL) the malignant Hodgkin Reed-Sternberg (HRS) cells comprise only a small fraction of the total cellular tumor population, which orchestrate an inflammatory microenvironment of reactive cells that propagate a permissive milieu for HL growth. This HRS cell mediated immune suppression has effects that extend beyond the direct tumor microenvironment. Previously we have shown that the circulating CD4 and CD8 T cells of relapsed (R) HL patients show high expression of the receptor programmed death-1 (PD-1), and that the central memory T cells (TCM) of both newly diagnosed (ND) and R HL patients are perturbed towards chronic activation. Here we describe immune activation and perturbation in the myeloid compartment of 20 HL patients with newly diagnosed (ND) or relapsed (R) disease. Methods: Informed consent obtained from patients with ND (n= 14) or R (n=6) HL treated since January of 2013. Blood samples were drawn pre-treatment, and at sequential time-points during and after therapy. Cryopreserved PBMCs were thawed then evaluated with multi-color flow cytometry. Cells were stained with fixable viability dye (eBioscience); then stained with fluorescent-conjugated antibodies. For intracellular staining, cells were fixed and permeabilized using FOXP3 staining kit (eBioscience) then stained with intracellular antibodies. Stained cells were analyzed using LSRII flow cytometer (BD Bioscience) and FlowJo software (Tree Star). The following anti-human antibodies were used to analyze myeloid subsets: CD3, CD19, CD16, HLA-DR, CD14, CD303, CD141. Monocytes are defined as CD14+HLA-DR+CD3-CD19-, plasmacytoid dendritic cells (pDCs) as: CD303+HLA-DR+CD14-CD3-CD19-CD16-, and dendritic cells (DCs) as: HLA-DR+CD141+CD14-CD303-CD3-CD19-CD16-(Biolegend). Singlet cells were gated based on forward and side scatter properties. Dead cells were excluded based on viability dye. Patient samples were compared to normal controls matched for age and sex (n=20). Results: The median HL patient age was 38 (range: 20-90 years). Twelveof the HL subjects were male and 8 were female. All but 1 of the ND HL patients were treated with upfront ABVD +/- consolidative radiation. Two out of 6 R patients had prior allogeneic stem cell transplant (alloSCT); they were not on immunosuppression. Thirteen patients (12 ND, 1 R) responded to therapy (11 CR and 2 PR); 4 patients (1 ND, 3 R) progressed on therapy or had stable disease, 2 patients do not have disease status confirmed. We observed a decrease in the proportion of systemic circulating plasmacytoid dendritic cells (pDCs), but not monocytes or DCs, in HL patients at baseline compared to healthy controls (Figure 1). Interestingly, the ratio of both DCs and pDCs to monocytes were greatly disturbed in HL patients compared to healthy subjects (Figure 2). After a single cycle of chemotherapy we noted an increase in pDCs, and a decrease in the ratio of monocytes to both DCs and pDCs and of DCs to pDCs in treated patients. Together these findings suggest a systemic imbalance within the monocyte-DC-pDC axis in HL patients compared to normal healthy controls, which appears to normalize after one cycle of treatment with chemotherapy. Conclusion: HL patients have evidence of perturbation in the systemic myeloid compartment, suggesting that the impact of HRS cells on their microenvironment may have systemic as well as local effects that extend to the myeloid compartment and to antigen presentation, as well as to T cell phenotypes. This underscores the rationale for understanding the role of systemic immune dysfunction in HL, and for the use of immune targeted therapies in HL. Correlation with clinical data, and functional studies to further characterize this immune dysregulation are ongoing. Disclosures Diefenbach: Merck: Consultancy, Research Funding; Genentech: Consultancy, Research Funding; Gillead: Equity Ownership; Seattle Genetics: Consultancy, Honoraria, Research Funding; BMS: Consultancy, Research Funding. Martin:Gilead: Consultancy, Other: travel, accommodations, expenses; Janssen: Consultancy, Honoraria, Other: travel, accommodations, expenses; Celgene: Consultancy, Honoraria; Novartis: Consultancy; Teva: Research Funding; Acerta: Consultancy. Ruan:Seattle Genetics: Consultancy, Research Funding, Speakers Bureau; Celgene: Consultancy, Honoraria, Research Funding, Speakers Bureau; Pharmacyclics, LLC, an AbbVie Company: Research Funding, Speakers Bureau; Janssen: Research Funding.

Description: Zanolimumab (previously referred to as HuMax-CD4) is a fully human monoclonal IgG1k antibody, targeting the CD4 molecule on T-cells. It exhibits cytotoxic and anti-proliferative effects and has previously shown efficacy in cutaneous T-cell lymphoma. We report the early safety and peripheral CD4+ cell depletion results from the open-label, dose escalation part of a US phase III efficacy study. So far, 21 patients have been recruited and the results from the 4 mg/kg dose group (9 patients) and 8 mg/kg dose group (6 patients) are available. Zanolimumab was administered iv, once weekly for 12 weeks. In total, 2 Serious Adverse Events have been reported. Of these, 1 case of large cell transformation in a patient with large cells present prior to inclusion was judged possibly related to treatment by investigator. No increase in toxicity was seen upon dose escalation. Marked depletion of CD4+ T-cells was observed after just 1 infusion of zanolimumab at both dose levels. One week following the last dose, the median (range) peripheral blood CD4+ T-cell count compared to baseline values was decreased from 831 per μL (167–1928) to 24 per μL (4–141) in 4 mg/kg dose-group and from 638 per μL (182–1024) to 9 per μL (4–19) in the 8 mg/kg dose-group. These initial data indicate that zanolimumab has an acceptable tolerability profile at the doses tested and results in a rapid and pronounced decrease in peripheral CD4+ counts in MF CTCL patients.

Description: Abstract 882 Background: PLD is approved for treatment of Kaposi's sarcoma and concentrates highly in the skin. Overall response (ORR: complete response [CR] + partial response [PR] rates) ranging from 56% to 88%, and CRs from 20% to 44% have been reported with PLD in CTCL patients (pts) without strictly defined response criteria (Cancer 2003; 98: 993–1001; Arch Dermatol 2008; 144: 727–33) Bex is a synthetic retinoid with an ORR of approximately 50% in relapsed/refractory CTCL (J Clin Oncol 2001; 19: 2456–71; Arch Dermatol 2001; 137: 581–93). To clarify the true ORR for PLD and to assess whether the ORR and remission durations can be improved by sequential Bex following PLD, a multi-institutional phase II trial was undertaken for pts with advanced stage/refractory CTCL. Methods: Pts were treated with PLD 20mg/m2 IV every 2 weeks for 8 doses (16 weeks) followed by 16 weeks with Bex 300mg/m2 orally. Response assessments were performed after 8 (PLD) and 16 weeks (Bex). The Severity-Weighted Assessment Tool (SWAT) (Arch Dermatol 2002; 138: 42–8) was the primary measure of skin response with the Composite Assessment of Index Lesion Severity (J Clin Oncol 2001; 19: 2456–71) and a self-reported pruritus scale as secondary measures. For stage IV pts with baseline lymphadenopathy, CT of chest, abdomen and pelvis was repeated. Sézary cell count and/or flow cytometry were repeated for patients with Sézary syndrome. Pts with complete disappearance of skin lesions on examination were deemed to have a clinical CR (CCR) in skin. A repeat skin biopsy without evidence of disease changed CCR to full CR. Results: All of the planned 37 pts with stages IV (21 including 7 Sézary syndrome), IIB (10) or refractory earlier stage (6) CTCL have been enrolled and have completed 32 weeks of treatment. Median age was 56 years (27–81). There were 20 males and 17 females. All pts had prior topical treatment with or without irradiation. Fourteen pts had no prior systemic therapy; while23 pts had prior systemic therapies (median 2, range 1–11). Median follow-up for surviving pts is 7.54 months (0.7–41.9). Thirty-four pts were assessable for response (2 withdrew consent, 1 died of disease progression 5 days after initiation of 1st cycle). For 34 assessable pts: ORR 14/34 (41%: PR 12, CCR 2). Maximum responses were all seen after 16 weeks PLD. Median progression free survival (PFS) is 4.82 months. To date, 18 pts have died of disease progression and 1 patient died of CHF which occurred 3 months after last study intervention. That patient had a pretreatment left ventricular ejection fraction of 60%. Treatment-related grade 3/4 serious adverse events occurred in 9 pts: 4 tumor pain, 2 (grade 3) hand-foot syndrome, 1 infection- unknown ANC- skin (cellulitis), 1 infection- normal ANC- skin (cellulitis), 1 neutropenia. Conclusions: With strict criteria, PLD ORR is among the highest reported for single agents in CTCL but lower than previously reported. The population contained a high proportion of patients with advanced disease as reflected in the initial stage and the poor survival. Sequential bexarotene did not increase the response rate or duration. Disclosures: Horwitz: Celgene: Consultancy, Research Funding; Allos: Consultancy, Research Funding; Seattle Genetics: Consultancy; Novartis: Consultancy; Merck: Honoraria; Millennium: Consultancy; Genzyme: Research Funding.

Description: It has been proposed that bacteria play a direct role in progression of cutaneous T cell lymphoma (CTCL), although definitive evidence is missing, and the underlying mechanism of how microbes contribute to disease progression remains unknown. The skin of CTCL patients is frequently colonized with Staphylococcus aureus (S. aureus) strains and infections with hospital and community associated strains of S. aureus are a frequent cause of morbidity and mortality among patients with advanced CTCL. Here we provide a comprehensive analysis of the association between CTCL and S. aureus colonization, and use our unique pre-clinical animal model of CTCL to determine the cause-effect relationship between skin-associated S. aureus and CTCL progression. To understand the relationship between bacterial colonization and CTCL we collected skin swabs from active lesions, unaffected skin and nares of CTCL patients to perform S. aureus cultures and 16S rRNA gene sequencing. Skin swabs of psoriasis patients and healthy donors served as controls. The frequency of S. aureus colonization determined by culture based techniques revealed that 〉65% of advanced stage patients had S. aureus present at lesional/tumor sites, while corresponding sites in patients with psoriasis and in healthy controls rarely had detectable S. aureus. Colonization rates correlated positively with the disease stage. Unbiased, 16s sequencing based analysis of the skin microbiome from advanced CTCL patients revealed that the overall skin microbiome of these patients is distinct from that of healthy individuals and patients with psoriasis. A lower phylogenetic diversity and significantly higher relative abundance of Staphylococcus species was found in CTCL patients. To determine the causal relationship between skin flora and progression of CTCL we used our mouse model of CTCL and assessed disease progression in both conventionally housed specific-pathogen-free (SPF) conditions and in germ free (GF) isolators using a standardized clinical score. The CD4CreSTAT3stopfl/+ mice express a hyper-active STAT3C mutant protein selectively in T lymphocytes and virtually all mutant mice develop T cell infiltration in the epidermis causing skin lesions resembling CTCL, by eight months of age. In contrast to the SPF housed animals, GF mice remained disease free or developed only a mild phenotype (clinical score 1 out of 5) after 11 months of follow-up. Notably, when GF CD4CreSTAT3stopfl/+ mice were transitioned to SPF conditions they all developed advanced disease. Finally, we examined the role of T cell antigen receptor (TCR) signaling in mediating the transformation of T lymphocytes. R26STAT3Cstopfl/+CD4Cre Rag2KO OTII mice express only OVA-specific TCRs. T cells from R26STAT3Cstopfl/+ CD4Cre Stim1fl/fl mice express a normal TCR repertoire, but exhibit defective T cell receptor signaling due to compromised calcium influx. Both strains failed to develop typical skin lesions, suggesting an essential role for TCR interaction with tumor microenvironment and microbial antigens in the pathogenesis of CTCL. In conclusion, we demonstrate a strong correlation between CTCL staging and rates of S. aureus colonization. Our study supports a cause-effect relationship between skin flora and CTCL oncogenesis. We propose that CTCL represents an antigen driven malignancy. Further studies using mono-colonization with single bacterial strains are needed to further interrogate the role of specific bacteria. Disclosures Hymes: Celgene: Consultancy. Odum:Micreos human Health B.V: Consultancy. Geskin:Merck: Other: Supported/Contracted Research; UpToDate: Patents & Royalties: Royalty, Receipt of Intellectual Property Rights / Patent Holder; Mallinckrodt: Consultancy, Honoraria, Other: Supported/Contracted Research; Helsinn: Consultancy, Honoraria, Other: Supported/Contracted Research; Stratpharma: Other: Supported/Contracted Research; Medivir: Consultancy, Honoraria; Medscape: Speakers Bureau; Actelion: Other: Supported/Contracted Research.