Search

The unfavorable clinical evolution in indolent non-Hodgkin lymphomas suggests defective control of neoplastic growth by the immune system. To address this issue, we evaluated phenotype, function, and maturation profile of CD4+ and CD8+ T cells from peripheral-blood, lymph nodes, or bone marrow of patients with B-cell non-Hodgkin lymphoma (NHL) at diagnosis. T cells from these patients frequently showed an activated but apoptosis-prone phenotype with low frequency of tumor-reactive T cells showing a TH2/Tc2 functional profile in the response to autologous tumor. In peripheral blood or in lymph nodes and bone marrow, and, in comparison to healthy donors, patients' T cells showed a skewed differentiation toward Tnaive and Tcentral memory stages, with low expression of granzyme B and perforin. T-cell culture with autologous tumor in the presence of IL-2, IL-15, and autologous bone marrow–derived cells led to massive T-cell expansion and to differentiation of cytotoxic factor+ CD8+ T cells releasing IFN-γ and killing autologous B-cell tumor in an HLA-class I–restricted fashion. These results suggest impaired T-cell differentiation to effector stage in patients with B-cell NHL, but indicate that T-cell responsiveness to γc cytokines is retained, thus allowing to promote generation of antitumor T cells for immune intervention.

Eighteen relapsed patients with measurable indolent non-Hodgkin lymphoma (NHL) were vaccinated with dendritic cells (DCs) loaded with killed autologous tumor cells. Six patients had objective clinical responses including 3 continuous complete responses (CRs) and 3 partial responses (PRs), with a median follow up of 50.5 months. Eight patients had stable disease, whereas 4 had progressive disease. Clinical responses were significantly associated with a reduction in CD4+CD25+FOXP3+ regulatory T cells, an increase in CD3−CD56dimCD16+ natural killer (NK) cells, and maturation of lymphocytes to the effector memory stage in either postvaccination peripheral blood or tumor specimen samples. In partial responding patients, vaccination significantly boosted the IFN-γ–producing T-cell response to autologous tumor challenge. In one HLA-A*0201+ patient who achieved CR, IL-4 release by circulating T cells in response to tumor-specific IgH-encoded peptides was also documented. Immunohistochemical analysis of tumor biopsies using biotin-conjugated autologous serum samples revealed a tumor-restricted humoral response only in the postvaccination serum from responding patients. Collectively these results demonstrate that vaccination with tumor-loaded DCs may induce both T- and B-cell responses and produces clinical benefits in indolent NHL patients with measurable disease. This study is registered with the Istituto Superiore di Sanità: http://www.iss.it with protocol number 7578-PRE 21-801.

Introduction TNF-related apoptosis-inducing ligand (TRAIL) is a protein functioning as a ligand that induces the process of cell death. TRAIL has been shown to kill in vitro a wide variety of tumor cells with minimal effects on normal cells. Despite its in vitro activity, recombinant soluble TRAIL has so far shown limited efficacy in vivo. In contrast, recent reports have shown that significant apoptosis can be observed both in vitro and in vivo when TRAIL is expressed on the cell membrane (mTRAIL). A further innovation might be the delivery of bioactive proapoptotic TRAIL through its expression by extracellular vescicles (EVs), the nanovesicular organelles secreted by cells. In fact, EVs are viewed as an effective tool for intercellular cross-talk and receptor discharge. The trans-membrane expression of TRAIL ligand within the double layer exosomal membrane may induce a more potent death signal when compared with the soluble molecule. Material and Methods Mesenchymal Stromal Cells (MSC) from bone marrow were cultured in vitro and used for EVs production. Cultured MSC in 75 cm2 flasks, at 80% confluence were infected with a lentivector encoding TRAIL, maintained in culture, and cell-supernatants repeatedly collected over several days, ultracentrifugated, with EVs-containing pellet harvested in PBS. EVs were produced also from uninfected MSC as control (EVs-CTRL). EVs were characterized by flow cytometry for expression of MSC markers and mTRAIL, EV size was evaluated by NanoSight technology. Total protein concentration was used to quantify EVs, Western Blot analysis was performed to characterize membrane-bound TRAIL. In vitro analysis was performed on SU-DHL-4 (human B cell lymphoma) and MEL-1300 (human melanoma) cell lines, exposed for 24 hours to 20-100 μg/ml EVs-TRAIL or EVs-CTRL. Annexin/propidium iodide assay was used to quantify apoptotic/necrotic cells. For the in vivo assessments, SU-DHL-4 and MEL-1300 cells were transduced with Luc-Lentiviral particles to obtain Luciferase positive cell lines. These cells were used to engraft NOD scid gamma (NSG) mice (2x106 SU-DHL-4 and 3x105 MEL-1300 cells for each subcutaneous injection point). To visualize tumor cells, mice were injected intraperitoneum with luciferin and analyzed with the Xenogen system. Mice bearing subcutaneous tumor nodules received single intravenous injections of 100, 200, 300 µg or multiple (x 3) 200 µg injections of either EVs-TRAIL or EVs-CTRL. Results FACS analysis showed strong TRAIL expression on EVs from TRAIL-infected MSC compared to EVs-CTRL, with a high proportion of positive particles (median 85%, range 78-93). In addition, EVs-TRAIL displayed MSC membrane markers, i.e. CD 105, CD 90, CD73 and CXCR4. Western Blot analysis under non-reducing conditions showed the presence of TRAIL ligand, with strong prevalence of dimeric TRAIL isoform (barely detectable the trimeric isoform, undetectable monomeric isoforms). NanoSight analysis revealed that EVs had a variable size, up to approximately 400 nm in diameter, with a predominant peak at 273 nm. A strong and dose-dependent cytotoxic effect was observed on SU-DHL-4 cells exposed to EVs-TRAIL (annexin/PI+ve cells: up to 87% for 100 μg/ml EVs-TRAIL), compared to EVs-CTRL exposure (15% Annexin/PI+ve cells for 100 μg/ml EVs-TRAIL). A similar, albeit less pronounced in vitro cytotoxic effect of EVs-TRAIL was observed on the melanoma MEL-1300 cell line. The anti-tumor effect was remarkably strong when EVs-TRAIL were injected in vivo in mice bearing either SU-DHL-4 or MEL-1300 nodules. A marked reduction of the tumor luminescence from 1.2x1010 photon/sec to Conclusion EVs isolated from genetically engineered TRAIL-expressing MSC: i. do express mTRAIL; ii. display potent antitumor activity, inducing extensive apoptosis/necrosis both in vitro and in vivo in animal models bearing lymphoma and melanoma nodules. Thus, EVs-TRAIL may represent a promising strategy for delivering pro-apoptotic signals to tumor cells. Moreover, the Results could pave the way to the use of EVs for therapeutic purposes. Disclosures: No relevant conflicts of interest to declare.

Abstract 3711 INTRODUCTION: Patients with refractory or relapsed classical Hodgkin Lymphoma (cHL) represent an unmet medical need and would benefit from the development of new therapies. Histone deacetylases (HDACs) and the RAF/MEK/ERK pathway are aberrantly controlled in cHL and influence a broad repertoire of tumor processes, suggesting a rationale for therapeutically targeting these pathways. We targeted these pathways using the HDAC inhibitor Givinostat (Italfarmaco S.p.A., Milan, Italy), and the RAF/MEK/ERK inhibitor Sorafenib (Nexavar, Bayer, Germany, EU) in order to investigate in vitro and in vivo the activity and mechanism(s) of action of this two-drug combination. METHODS: Three cHL cell lines, including HDLM-2, L-540 and HD-MyZ, were used to investigate the effects of Givinostat and Sorafenib, used alone or in combination, by means of in vitro assays evaluating cell growth and cell survival. Additionally, live cell imaging was used to asses the production of reactive oxygen species (ROS), and Western blotting (WB) to assess modulating effects of the two-drug combination on MAPK, PI3K/AKT, HDACs as well as the apoptotic pathways. The efficacy of Givinostat/Sorafenib combination was finally confirmed in NOD/SCID mice with cHL cell line xenografts. RESULTS: While Givinostat and Sorafenib as single agents exerted a limited activity against cHL cells, the combined Givinostat/Sorafenib treatment was associated with potent dephosphorylation of MAPK and PI3K/Akt pathways and significantly increased H3 and H4 acetylation due to a nearly complete inhibition of class I and II HDACs. Furthermore, these events were associated with a time-dependent synergistic cell growth inhibition (70% to 90%) in all Givinostat/Sorafenib-treated cHL cells. Upon Givinostat/Sorafenib exposure, HDLM-2 and L-540 cell lines showed significantly (P ≤.0001) increased levels of apoptosis (90 ± 2% and 96 ± 1%, respectively) and mitochondrial dysfunction (up to 70%, P≤.0001), as compared with single agents. Apoptosis induced by Givinostat/Sorafenib combination failed to induce processing of caspase-8, −9, −3, or cleavage of PARP, and was not reversed by the pan-caspase inhibitor Z-VADfmk, suggesting the occurrence of caspase-independent apoptosis. Besides downregulating the expression of the anti-apoptotic protein Mcl-1 and ERK1/2 phosphorylation, Givinostat/Sorafenib strongly increased expression of the BH-3 only protein Bim, compared to single treatments. These findings were dependent on a potent, early and time-dependent ROS generation (up to 60%, P≤.0001) that was synergistically induced by Givinostat/Sorafenib treatment. Additionally, pretreatment of cHL cells with the ROS inhibitor YCG063 prevented the generation of ROS as well as mitochondrial membrane depolarization along with cell death induced by the two-drug combination, suggesting that ROS generation is the triggering event in Givinostat/Sorafenib induced-cell death. In vivo Givinostat/Sorafenib treatment significantly reduced the growth of L-540 and HD-MyZ nodules, resulting in an average 35% to 65% tumor growth inhibition (P ≤.0001) compared to single treatments, in the absence of any toxicity. Interestingly, as compared to controls or treatment with single agents, the combined Givinostat/Sorafenib treatment significantly increased in vivo Bim expression (7- to 21-fold increase, P ≤.0001), resulting in a marked tumor necrosis (3- to 5-fold increase, P ≤.0001). CONCLUSIONS: The combined Givinostat/Sorafenib treatment demonstrates a potent preclinical in vitro and in vivo activity against cHL cell lines by targeting aberrant expression of HDACs and MAPK. Antitumor activity of this combination involves ROS generation and Bim upregulation and provides a rationale for clinical studies using this combination in refractory/relapsed cHL patients. Disclosures: No relevant conflicts of interest to declare.

Abstract 1653 Introduction: A significant proportion of Hodgkin lymphoma (HL) patients refractory to first-line chemotherapy or relapsing after autologous transplantation are not cured with currently available treatments and require new treatments. The PI3K/AKT and RAF/MEK/ERK pathways are constitutively activated in the majority of HL. These pathways can be targeted using the AKT inhibitor perifosine (Æterna Zentaris GmBH, Germany, EU), and the RAF/MEK/ERK inhibitor sorafenib (Nexavar®, Bayer, Germany, EU). We hypothesized that perifosine in combination with sorafenib might have a therapeutic activity in HL by overcoming the cytoprotective and anti-apoptotic effects of PI3K/Akt and RAF/MEK/ERK pathways. Since preclinical evidence supporting the anti-lymphoma effects of the perifosine/sorafenib combination are still lacking, the present study aimed at investigating in vitro and in vivo the activity and mechanism(s) of action of this two-drug combination. METHODS: Three HL cell lines (HD-MyZ, L-540 and HDLM-2) were used to investigate the effects of perifosine and sorafenib using in vitro assays analyzing cell growth, cell cycle distribution, gene expression profiling (GEP), and apoptosis. Western blotting (WB) experiments were performed to determine whether the two-drug combination affected MAPK and PI3K/AKT pathways as well as apoptosis. Additionally, the antitumor efficacy and mechanism of action of perifosine/sorafenib combination were investigated in vivo in nonobese diabetic/severe combined immune-deficient (NOD/SCID) mice. RESULTS: While perifosine and sorafenib as single agents exerted a limited activity against HL cells, exposure of HD-MyZ and L-540 cell lines, but not HDLM-2 cells, to perifosine/sorafenib combination resulted in synergistic cell growth inhibition (40% to 80%) and cell cycle arrest. Upon perifosine/sorafenib exposure, L-540 cell line showed significant levels of apoptosis (up to 70%, P ≤.0001) associated with severe mitochondrial dysfunction (cytochrome c, apoptosis-inducing factor release and marked conformational change of Bax accompanied by membrane translocation). Apoptosis induced by perifosine/sorafenib combination did not result in processing of caspase-8, -9, -3, or cleavage of PARP, and was not reversed by the pan-caspase inhibitor Z-VADfmk, supporting a caspase-independent mechanism of apoptosis. In responsive cell lines, WB analysis showed that anti-proliferative events were associated with dephosphorylation of MAPK and PI3K/Akt pathways. GEP analysis of HD-MyZ and L-540 cell lines, but not HDLM-2 cells indicated that perifosine/sorafenib treatment induced upregulation of genes involved in amino acid metabolism and downregulation of genes regulating cell cycle, DNA replication and cell death. In addition, in responsive cell lines, perifosine/sorafenib combination strikingly induced the expression of tribbles homologues 3 (TRIB3) both in vitro and in vivo. Silencing of TRIB3 prevented cell growth reduction induced by perifosine/sorafenib treatment. In vivo, the combined perifosine/sorafenib treatment significantly increased the median survival of NOD/SCID mice xenografted with HD-MyZ cell line as compared to controls (81 vs 45 days, P ≤.0001) as well as mice receiving perifosine alone (49 days, P ≤.03) or sorafenib alone (54 days, P ≤.007). In mice bearing subcutaneous nodules generated by HD-MyZ and L-540 cell lines but not HDLM-2 cell line, perifosine/sorafenib treatment induced significantly increased levels of apoptosis (2- to 2.5-fold, P ≤.0001) and necrosis (2- to 8-fold, P ≤.0001), as compared to controls or treatment with single agents. CONCLUSIONS: Perifosine/sorafenib combination resulted in potent anti-HL activity both in vitro and in vivo. These results warrant clinical evaluation in HL patients. Disclosures: No relevant conflicts of interest to declare.

B-cell malignancies represent a potential target for anti-cancer vaccination programs due to the expression of tumor-specific antigens. Although immunization with tumor-derived idiotype protein is a frequently used procedure, vaccination with DCs loaded with killed tumor cells may activate response to a much wider range of antigens, without requiring prior molecular identification of such determinants. Furthermore, such DC-based vaccines could be available to all patients, irrespective of the HLA type. To evaluate the safety and tolerability of this approach, 18 patients with measurable relapse/refractory follicular (FCL; n= 12) and lymphoplasmocytoid (n= 6) lymphoma have been enrolled in a phase I study. Median prior number of treatment regimens was 2 (range 1–5) comprising 4 patients treated with high-dose chemotherapy supported by autologous stem cell transplantation. The vaccination was started after at least 6-months from the last chemotherapy treatment. All patients were evaluable for toxicity and 16/18 patients for efficacy with a median follow-up of 12.5 months (range 3–29 months). Each patient received 4 intradermal/subcutaneous injections at 2-weekly intervals of 50x10e6 tumor-loaded DCs. Immature DCs were generated by 5-days culture of autologous monocytes in the presence of IL-4 and GM-CSF. After selection by immunomagnetic technique, autologous CD19+ tumor cells, harvested from lymph nodes (n= 12) and/or peripheral blood (n= 6), were heat shocked and then irradiated by UVC. DCs were loaded for 48 hrs with killed tumor cells and then, to induce their maturation, were cultured for 12 hrs in the presence of TNF-alfa. Overall, vaccinations were well tolerated and no autoimmune reactions were observed. Mild erythema in the site of injection developed in the majority of patients (12/18), but only in 2 cases induration and extended erythema was observed. Six of 16 (37.5%) evaluable patients had objective responses. Two patients had partial responses (PR). One is still in PR and the other had a PR lasting 7 months. Four patients had complete remission (CR). Two patients are still in CR and the other 2 patients had a mean CR duration of 14.5 months. The remaining 10 patients had stable disease (n=5) or progressive disease (n=5). The overall monitoring of immune responses is ongoing. However, in one patient in PR, we evaluated the frequency of anti autologous tumor-specific T cells, by ELISPOT assay for IFN-gamma, on pathologic lymph nodes harvested before and after 2 months from the last vaccination. A significant increase of specific T-cell frequency was observed in the post-vaccination lymph node, compared to the tissue sample taken before vaccination. Moreover, evaluation of CD8+ T cell maturation markers, by analysis for CCR7 and CD45RA expression, indicated a shift of tumor-infiltrating T cells towards memory and effector stages in the lymph-node isolated after vaccination. In conclusion, injection of DCs loaded with killed tumor cells is a well-tolerated procedure achieving clinical and immunological responses also in the presence of significant tumor burden. However, further strategies, following DC-vaccination, are needed to ensure durable immune and clinical responses.