Search

BACKGROUND: In 2014 we identified a new subset of DLBCL, defined as "IgM-secreting" (Cox MC & Di Napoli A , PLOS One 2014). This was characterised by poor prognostic features and outcome as well as frequent central nervous (CNS) system localizations. Furthermore, IgM-secretion, was an independent prognostic factor in multivariate analysis. Here we report on the largest series of IgM-secreting-DLBCL, from a multicentre Italian study. METHODS: The observational and biological study was approved by the Ethical Committee of the AUO Sant'Andrea, Italy. Enrolment criteria were: DLBCL with an associated IgM paraprotein diagnosed between 1st January 2010 and 31st December 2018 (IgM-secreting). Data were collected both prospectively and retrospectively from 17 Centres participating in the study. In addition, histopathology samples were centrally revised for immunohistochemistry (IHC) and FISH analyses. The control group (CTRL) consisted in a series of consecutive DLBCL, without an associated IgM-paraprotein (diagnosed between 01/01/2013 and 30/06/2016, enrolled in the Lymphoma Registry of the Lazio region (ReLLi Network). Last follow-up was carried out on 31st December 2019. RESULTS: 569 DLBCL cases were enrolled: 102 (17.9%) were IgM-secreting; 48 (8.4%) had a non-IgM paraprotein (IgA, IgG, or other), and 414 (72.7%) had no associated paraprotein (CTRL). IgM-secreting cases within the consecutive DLBCL patients enrolled in the ReLLi Registry were 41/466 (8.8%, 95CI 6.4-11.7%) while non IgM-paraprotein DLBCL cases were 11/466 (2.4%, 95CI 1.2-4.2%). The median level of IgM paraprotein was 17gr/L (range: 60 (p=.001); 2] advanced stage (pUNL (p=.008) ; 5] ≥2 Extra-nodal sites involved (p60, B-symptoms, bulky disease, IPI >low risk and IgM-secreting IgM showed a worse survival (all with p CONCLUSION: Our data confirm that IgM-secreting DLBCL: 1) represents a sizable proportion of non-DH DLBCL; 2) have poor prognostic features and 3) have mostly a non-GC phenotype. Furthermore, IgM secretion appears to be an independent prognostic factor for both PFS and OS. Studies to define the biological features of this new subset are ongoing. Disclosures Cantonetti: Mundipharma: Consultancy; Takeda: Consultancy; Vifor: Consultancy; Roche: Consultancy. Re:BerGenBio ASA: Research Funding. Abruzzese:Incyte: Consultancy, Membership on an entity's Board of Directors or advisory committees; Novartis: Consultancy, Membership on an entity's Board of Directors or advisory committees; Pfizer: Consultancy, Membership on an entity's Board of Directors or advisory committees; Bms: Honoraria.

Purpose: We have developed a novel DC population differentiated from human monocytes in the presence of GM-CSF and IFN-alpha (IFN-DC), which is highly efficient in internalizing tumour cell antigens and in the cross-priming of CD8+T cells, promoting efficient anti-lymphoma response. Preclinical results prompted us to start a clinical trial of IFN-DC combined with low-dose Rituximab(RTX) in patients with Follicular lymphoma (FL), which is an indolent, immune-responsive, but still incurable cancer. Patients and Methods: In October 2014 we started a phase I clinical trial (EudraCT: 2013-003158-25) testing safety, immunogenicity and systemic clinical responses of low-dose RTX combined with autologous IFN-DC, administered by intra-nodal injections. Patients features: age 18-75y, indolent FL, relapsed or refractory (R/R), stage III-IV low tumour burden, with at least 1 superficial pathologic lymph node. One leukapheresis was scheduled at study entry for harvesting mononuclear cells. Eight cycles of RTX (5-20mg) and IFN-DC (30±10x106) were planned: the first 4 every two weeks, followed by other 4 cycles administered monthly. All injections were delivered with the aid of an ultrasound guide. Blood test for immune-responses and autoimmunity markers were scheduled at 0,+72d, +132 days ; +9; +18;+24 months. Results: as of the 31st July 2018, 8 patients have completed treatment and have been evaluated for response (median follow-up=24 months; range=8-45 mo). Median age of the study population was 60 years (range 27-76). All subjects had disseminated disease and enlarged lymph nodes or lesions greater than 1.5 cm at multiple sites available for local injections and monitoring for distant response. All patients were R/R after previous systemic therapy (median lines=2, range1-4) and 3/8 (37.5%) after auto-transplantation. The treatment was very well tolerated and manageable. No relevant adverse events were observed. Four out of 8 (50%) patients had objective response (OR) with 3/8(37.5%) attaining complete remission (CR); 2/8 (25%) are in stable disease at +8 and +25 months; 2/8 (25%), who did not achieve OR, progressed at +8 and +14 respectively. One patient progressed 4 months after achieving CR (Figure1). In all patients who achieved OR or who are in stable disease, induction of a tumour specific immune response was observed. Conclusions: local RTX and IFN-DC combined therapy was effective in eliciting in 50% of R/R FL with low tumour burden systemic and durable responses, which were associated with induction of specific anti-FL T cell responses. Disclosures No relevant conflicts of interest to declare.

INTRODUCTION Follicular lymphoma (FL) is considered a chronic, immune responsive and still incurable tumour. Recently two papers have shown the efficacy of active immunotherapy by repeated administration of autologous dendritic cells (DC) in Follicular Lymphoma (FL). Notably, in both studies a third of patients achieved long lasting remission. A novel DC population (IFN-DC), differentiated from human monocytes in the presence of GM-CSF and IFN-alpha has been developed in the laboratory of the Istituto Superiore di Sanità (ISS). IFN-DC have been demonstrated highly efficient in internalizing tumour-cell antigens, recovering "in vitro" T-cell responses from FL patients, as well as in mediating NK cell activation and enhanced cytotoxic effector function toward autologous FL cells. We have recently started a phase I study (approved by EC in 2014, EudraCT: 2013-003158-25) of intranodal immunotherapy based on sequential injection of low-dose anti-CD20 antibody (Ab) and IFN-D, in patients with advanced FL. Here we report early results of the ongoing trial. STUDY DESIGN IFN-DC Trial (EudraCT: 2013-003158-25) is a phase I trial aiming at evaluating safety and tolerability as well as immune and clinical responses of a IFN-DC based therapy in combination with rituximab for the treatment of patients with advanced FL. · The Primary endpoints are: I) Evaluation of safety and tolerability of treatment II) Evaluation of tumor-specific immune responses. by determining: i) Tumor-specific immune response in peripheral blood; ii) Intratumoral infiltration of immune cells; iii) DTH test. · The secondary endpoint is clinical response. The study population consists of patients with relapsed/refractory FL, aged 18-75 years, not needing immediate retreatment basing on the GELF Criteria. The · Treatment: the regimen foresees eight injection cycles, each consisting of the combined administration of rituximab and IFN-DC. The first four administered every two-weeks and the remaining four administered monthly. Rituximab (5-10mg) and low or high doses of IFN-DC (20±5 x106 or 40±5x106cells respectively), are administered by intranodal direct injection (IDI). RESULTS Five patients have been enrolled so far. The treatment proved to be safe, feasible and well tolerated. Three patients have completed treatment and two experienced complete remission with disappearance of both proximal and distal lesions, as confirmed by PET scan. Notably in both patients the distal lymph nodes, which had not been directly treated by intranodal injections, reduced to normal size earlier than those treated by IDI (abscopal effect). One of these two is still in durable complete clinical and molecular remission after 17 months while the latter patient, who had progressed after completing eight cycles of treatment, experienced a complete clinical remission after a second course of immunotherapy ( Fig.1). Two out of three patients performed the DHT test and, despite one of the two achieved remission, they both resulted non-responsive. The characterization of IFN-DC and lymphoma nodes in FL patients have shown high expression of PD-1 and PD-L1 in FL microenvironment and on IFN-DC respectively. Immunomonitoring is currently underway. However, quantification of NF-kB (p65) nuclear translocation by multispectral image flow cytometry, as a marker of NK cell activation, showed the persistence of higher percentages of activated NK cells in the two responding patients (Fig.2). CONCLUSION Currently no definitive data prove the benefit of IDI treatment, however the trial is ongoing and other patients are under screening. Notably both the responding patients showed the abscopal effect. The possible set-up of an effective and comparatively economic treatment, that exploits patients' immune system against this chronic and immune responsive tumour, is worth of exploring. Furthermore it can be envisaged that the combination of immune checkpoint inhibitor antibodies to DC-based therapies, might improve clinical activity of DC-based strategies. Disclosures No relevant conflicts of interest to declare.

Background: Peripheral blood lymphopenia at diagnosis and after treatment has been widely reported as a negative prognostic factor in DLBCL. This phenomenon, which reflects host systemic immunity, is still poorly characterized. Very recently, we have described a profound phenotypic and functional alteration of the peripheral NK cell compartment in newly-diagnosed DLBCL patients at diagnosis and upon R-CHOP chemoimmunotherapy, and suggested that the complex dynamics of the immune system plays a pivotal role in the response to chemoimmunotherapy. Since the role of T cell-dependent memory and effector capabilities in long-term cure of chemoimmunotherapy-treated cancer patients is increasingly appreciated, we have focused our studies on their modifications in DLBCL. Aims: To analyze the phenotypic and functional profile of peripheral blood T lymphocyte compartment in DLBCL patients at diagnosis, and to assess the long-term impact of R-CHOP chemoimmunotherapy on T cell populations and functional competence. Patients and Methods: We prospectively compared 32 consecutive newly diagnosed DLBCL patients with 27 healthy, age- and sex-matched controls for: 1) absolute number (/mcl) and percentage (over PBMC) of the main T cell subsets: "conventional CD4+ and CD8+, double positive (CD4+CD8+), double negative (CD4-CD8-), CD56+ innate-like, and FOXP3+CD25bright regulatory T cells (Treg), measured by blood cell count and multi-parameter flow cytometric (FACS) analysis; 2) functional capability of T cell subsets, evaluated as the frequency of IFN gamma (IFNg)-producing cells upon a short-term (6-hr) stimulation with PMA/ionomycin, and the percentage of GrzB+ (cytotoxic granule marker) cells; and 3) plasma concentration of selected cytokines. Patients were analyzed at diagnosis, during and at different timepoints after chemoimmunotherapy completion. Results: DLBCL patients at diagnosis showed lower lymphocyte count (p The phenotypically skewed T cell profile of DLBCL patients at diagnosis was associated with severe functional alterations, as the frequencies of IFNg-producing and Granzyme B-expressing cells were increased in both CD4+ (p=.007 for IFNg+ and p=.005 for GrzB+) and CD8+ (p=.002 for IFNg+ and p Finally, IL-6 and IL-10 plasma levels were significantly elevated in DLBCL at diagnosis (p Interestingly, CD4+ and CD8+ T cell capability to promptly produce IFNg was significantly higher in GC, but not in non-GC DLBCL patients at diagnosis, suggesting that patients T cell functional features may depend on DLBCL phenotype. Moreover, a higher percentage of IFNg+ T cells was observed in patients attaining a long-term remission, but not in resistant or relapsing ones, hinting to the association of immune competence with therapy outcome. Conclusions: T cell-dependent memory and effector capabilities might significantly contribute to the success of chemoimmunotherapy strategies, as suggested by the recently discovered vaccinal effect. The novel information provided by this study may reveal useful for innovative therapeutic approaches more focused on exploiting the dynamics of host immune cells in fighting cancer. *MC Cox and S Battella contributed equally to this work Disclosures No relevant conflicts of interest to declare.

Abstract 2659 Background: Gene-expression-profiling defined at least two main groups within Diffuse-large-B-cell Lymphoma (DLBCL) patients who have substantially different outcomes: Activated-B-cell (ABC-type) and Germinal-Center-B-cell (GCB-type). The translation of gene-expression-profiling arrays into robust algorithm useful for clinical purposes is still in progress. The detection of IgM monoclonal component (IgM MC) in DLBCL has been previously described in a few reports, mainly because it was associated with autoimmune hemolytic anaemia. To our knowledge this is the first report describing the incidence and prognosis of a series of DLBCL with IgM MC. Aims: In this report we compared clinical and biological features of DLBCL patients with and without secretory IgM MC . Patients & Methods: Within a consecutive series of 132 patients, diagnosed between September 2004 and April 2012 with conventional DLBCL, 16 cases (12%) with a IgM MC were identified. We selected a set of 95 consecutive DLBCL patients, treated with 6–8 cycles of RCHOP-like for comparison of histological features and survival. Only cases with a follow up time >24 months were included, unless a DLBCL–related event (i.e. primary refractoriness or relapse) had occurred earlier. Biological material was obtained after receiving patient's consent. This study was approved by our Institutional-Review-Board. Immunohistochemistry and FISH: Paraffin sections were immunostained for CD3, CD5, CD20, CD10, CD30, CD79a, CD138, ALK-1, MUM1, BCL2, BCL6, IgM, Kappa and Lambda immunoglobulin light chains, using an automated immunostainer (DAKO, Denmark). The Hans algorithm was used in order to classify cases as GCB-type and non GCB-type. FISH with Vysis break-apart probe was used to assess c-MYC gene abnormalities in tissue sections (Abbott Molecular Inc. US). Statistics: univariate comparisons between groups were carried out with appropriate non parametric test. Survival analyses were done by the Kaplan-Meier method, the analyses of factor predicting survival were carried out by the log-rank test. Cox's regression was used for multivariate analyses. The SPSS19 package (SPSS Inc.Chicago IL) was used for elaborations. Histology, immunohistochemistry and FISH Results: In 14 out of 16 cases (87.5%) the IgM MC was related to the DLBCL clone. This was ascertained by immunostaining of cytoplasm for IgM, Kappa and Lambda immunoglobulin light chains. All the 14 cases were classified as non GCB-type. FISH analysis detected no c-MYC gene rearrangements in all the cases. Clinical Results: The incidence of bone marrow involvement, two or more extranodal sites, female sex, IPI score 3–5 and failure to achieve CR on RCHOP treatment were significantly more frequent in the IgM MC group. Noteworthy four out of 14 patients had central nervous system involvement at diagnosis or at relapse. All but one, with a previous diagnosis of marginal zone lymphoma, were de novo DLBCL. Twelve patients (85.7%) presented a DLBCL related event compared to 35 patients (37%) without IgM MC (p=.001). Seven patients (50%) died with primary refractory or relapsed-chemoresistant disease, another one died of an adverse event during chemotherapy. Two are alive on salvage treatment, two are in PFS at +30 and +13 months after salvage treatment with Bortezomib-RDHAP followed by high dose therapy. Only two patients are in PFS after first line RCHOP at +56 and +29 months respectively. Survival analysis: The estimated two-year EFS, PFS and OS were significantly worst for IgM MC group (22% Vs 70%, p Conclusions: A secretory IgM MC related to the neoplastic clone, was detected in more than 10% of newly diagnosed DLBCL. IgM MC in DLBCL might be easily missed, given its negligible entity, the rarity of associated clinical signs and the rapid MC fading during treatment. This group showed homogeneous morphologic and immunohistochemical features consistent with the non GCB-type. FISH analysis was negative for c-MYC gene rearrangements. IgM MC in DLBCL patients, might be related to a very poor non GCB-type subset who should be given upfront intensified therapies. Disclosures: No relevant conflicts of interest to declare.

Natural killer (NK) activity against K-562 tumor cells was evaluated in peripheral blood leukocytes (PBL) obtained from untreated patients affected by hairy cell leukemia (HCL). NK activity present in PBL from 10 HCL patients was at least six-fold lower (p less than 0.01) than that present in PBL from 15 healthy donors. Decreased NK activity in HCL PBL was not due to dilution of the NK effector cells by the neoplastic cells; in fact, NK activity of PBL from 4 HCL patients with less than 5% circulating neoplastic cells was still five-fold lower (p less than 0.01) than that present in normal PBL. Partial characterization of the NK defect in HCL patients indicated that: (A) defective cytotoxicity was not dependent on the duration of the assay; (B) HCL PBL added to normal PBL during the assay did not exert suppressor activity; (C) the NK activity of HCL PBL could be potentiated in vitro by interferon; and (D) low levels of NK activity were associated with reduced numbers of circulating monocytes (p less than 0.01) and of large granular lymphocytes (LGL) (p less than 0.01). In conclusion, our results indicate that the low levels of NK activity present in the peripheral blood of HCL patients may be related to reduced numbers of circulating effector cells.