ALBERT

Description: Abstract 3720 Background: Dendritic cells (DCs) are professional antigen-presenting cells which display an extraordinary capacity to induce, sustain and regulate T cell responses. Recently, 6-sulfo LacNAc+ (slan) DCs (formerly termed M-DC8+ DCs) have been described as a major subpopulation of proinflammatory human blood DCs which are principal producers of tumor necrosis factor-alpha and interleukin-12. In addition, it has been demonstrated that slanDCs efficiently induce antigen-specific CD4+ and CD8+ T cells and direct the polarization of naïve CD4+ T lymphocytes into Th1 cells. In the present study, we investigated the reconstitution kinetics of slanDCs after allogeneic stem cell transplantation (aSCT) in comparision to CD1c+ myeloid DCs and plasmacytoid DCs representing two additional major human blood DC subsets. Material and Methods: The frequency of slanDCs, CD1c+ myeloid DCs and plasmacytoid DCs in the peripheral blood was quantified by flow cytometry in 70 patients following aSCT at different time points in early engraftment (

Description: Abstract 1915 Bone marrow (BM) has been associated with a decreased rate of Graft-versus-Host-Disease (GvHD) when compared to peripheral blood stem cells (PBSC) in the context of allogeneic stem cell transplantation (allo-SCT). It has been reasoned that this may relate in part to the differential quantitative and qualitative immunologic composition of both stem cell sources. It has, for example, been noted that BM contains a higher percentage of CD8+ T-cells, natural killer (NK) cells and recently also more regulatory T-cells than PBSC. It is further known that the total nucleated cell concentration (TNC) correlates negatively with the BM harvest volume, which could argue for an increasing dilution of the transplant with peripheral blood (pB) during the BM harvest. How this impacts on the different immunological compartments and how the BM at different time points of the harvest procedure compares to PBSC products has not been determined yet. During routine BM harvest procedures (n = 24) we subjected a part of the aspirate drawn at the beginning, after withdrawing half of the prescribed volume and at the end of the harvest procedure to detailed immunophenotyping using multicolor flowcytometry including intracellular staining for FoxP3 and IL17. The PBSC products of an age matched group (n = 20) were analyzed in parallel with the same flowcytometry protocols and used as a comparison. During the BM harvest the median TNC dropped rapidly from 53.2 Gpt/l in the beginning to 17.0 Gpt/l after half the volume had been collected and finally reached 6.2 Gpt/l at the end of the procedure. As expected, the first BM specimen contained a higher concentration of NK- (6.7 vs. 4.6 %, p = 0.001) and B-cells (14.1 vs. 10.2 %, p 〈 0.001) but less T-helper-(Th)-cells (36.0 vs. 53.7 %, p 〈 0.001) than pB. After collection half of the prescribed volume the NK-cell concentration had already dropped to 4.5 %, which was not significantly different from pB (p = 0.412) anymore. The B-cell concentration, however, remained at levels comparable to the first specimen (14.5 %) and significantly differed from pB (p 〈 0.001). The last BM specimen drawn still had a higher B-cell concentration than pB (10.9 vs. 10.2 %), although this was not statistically significant (p = 0.265). Th-cells demonstrated a steady increase in concentration during the harvest with a median concentration of 43.4 and 47.0 % in the middle and at the end of the procedure. Both the halfway value and the end of harvest concentration were significantly different from pB (p 〈 0.001 and p = 0.004, respectively). An overview on the NK-, B- and Th-cell-concentrations at the different time points analyzed can be found in Figure 1. Within the Th compartment we were unable to see a clear trend for Treg which had a concentration of 7.6 % in the beginning, 7.0% halfway and 7.1 % after all BM had been collected. None of these concentrations was significantly different from that found in pB (7.6 %). The same applied to Th17 cells which made up 0.51 % of Th in the beginning, 0.59 % in the middle and 0.50 % at the end of the procedure and 0.84 % in the pB. Moreover there was no significant difference between the first BM specimen and PBSC products with respect to Treg (7.6 vs. 6.7 %, p = 0.181) or Th17 (0.51 vs. 0.61 %, p = 0.100). In contrast, the NK- and Th-cell concentration in PBSC products was significantly higher than in the first BM specimen (NK: 10.7 % vs. 6.7 %, p =0.005; Th: 40.3 % vs. 36.0 %, p =0.038). We conclude that the composition of immune cells within the BM changes significantly during the harvest procedure probably due to an increasing dilution with peripheral blood. These changes variably affect different compartments and may have an impact on post-transplant immunological function and complications. Therefore BM harvest volumes may need to be considered when comparing BM and PBSC with respect to clinical outcomes. In contrast to previous reports, we found no indication that BM, no matter at which time point analyzed, contained a higher concentration of Treg when compared to PBSC. Disclosures: No relevant conflicts of interest to declare.

Description: Introduction Greiner et al. demonstrated data suggesting a promising link between CD8+ T cell responses with a specificity of NPM1mut and graft versus leukemia effect (GvL). In patients with acute myeloid leukemia (AML) who underwent hematopoietic stem cell (HSC) transplantation, effector cells of the newly established immune system are thought to elicit a potent graft versus leukemia (GvL) effect eradicating residual leukemic cells. Cytotoxic T-lymphocytes (CTLs) specific for leukemia associated MHC-presented peptides may play a crucial role in the anti-tumor reaction and in achieving complete molecular remission. Leukemic cells are detectable after allogeneic HSCT in AML at least at the level of minimal residual disease (MRD). Malignant cells could therefore be identified and targeted by effector cells of the new immune system and stimulation of CTLs by antigen is assured. Detectable leukemia-specific CTLs will allow studies which address the correlation of the course of CTLs and clinical events. Further, it could allow for an assessment of the functional properties of these cells with respect to GvL effects. Methods We screened HLA-A0201 positive patients with AML and chronic lymphocytic leukemia (CLL) biweekly from day 28 after transplantation until day 112 for antileukemic CTLs applying streptamer technology. For AML a panel of 7 MHC peptide complexes was used to label CTLs from peripheral blood: phycoerythrin-(PE)-labeled streptamers refolded with peptides (number of different peptides) derived from the amino acid sequence of the proteins WT1 (1), PR3 (1), RHAMM (1), mutated NPM1 (2), and survivin (2). For CLL peptides derived from RHAMM (1), survivin (2) and fibromodulin (4) were chosen for screening. Peptides from cytomegalovirus (CMV) protein pp65 and from Influenza matrix protein M1 were used as positive controls. A peptide of the HIV reverse transcriptase served as negative control. Results Up to now neither for AML (n = 9) nor for CLL (n = 1), leukemia specific CTLs were detected according to the respective panel at any time point. However, it was possible to detect CMV-specific CTLs in most of the CMV seropositive patient-recipient pairs and frequencies increased when CMV reactivation occurred. Increasing frequencies of CMV-specific CTLs were even measured in CMV seropositive patients who received allogeneic HSC grafts from CMV seronegative donors, proving the adequacy of the method and reflecting the stimulation and proliferation of CMV-specific CTLs. Discussion Different reasons may account for our negative results so far: the limited numbers of selected peptides, which may not be immunodominant or the small number of screened patients are two of them. It must be considered that the CTL immune response to virus infection and concomitant CMV clearance may not represent a proper model for the GvL effect, which possibly is characterized only by a weak proliferation of CTLs resulting in cell counts below the detection limit of the applied flowcytometric measurements. We will further extend our screening approach to a minimum of 18 individuals per specificity. If frequencies above the limit of detection will not be detected in at least one out of 18 patients the 95 % confidence interval of the true percentage of patients with measurable CTLs of this specificity should be 〈 15 %. Conclusions This study highlights the difficulties to study leukemia-specific CTLs. So far, leukemia-specific CTLs have only been measured in single patients after allogeneic HSCT. Information on cohorts of patients who have been monitored longitudinally for specific CTLs is very rare. The kinetics of leukemia-specific CTLs which mediate GvL effects are almost unknown. Further research on delineating T-cell mediated GvL-effects and realistic estimates for frequencies of specific CTLs at well defined time-points are urgently needed. Disclosures: No relevant conflicts of interest to declare.

Description: Donor lymphocyte infusions have been effective in patients with chronic myeloid leukemia (CML) relapsing after allogeneic stem cell transplantation, but their use is associated with the risk of graft-versus-host disease. We investigated the effects of prophylactic infusion of in vitro-generated donor T cells reactive against peptides derived from CML-associated antigens. Fourteen CML patients received conditioning therapy followed by CD34+-selected peripheral blood stem cells from matched siblings (n = 7) or unrelated (n = 7) donors. Donor-derived mature dendritic cells generated in vitro from CD14+ monocytes were loaded with human leukocyte Ag-restricted peptides derived from PR1, WT1, and/or B-cell receptor–ABL and used to repetitively stimulate donor CD8+ T cells in the presence of IL-2 and IL-7. Stimulated T cells were infused 28, 56, and 112 days after transplantation. Thirteen patients are alive and 7 remain in molecular remission (median follow-up, 45 months). Interestingly, all 4 patients receiving CD8+ T cells displaying marked cytotoxic activity in vitro and detectable peptide-reactive CD8+ T cells during follow-up have not experienced graft-versus-host disease or relapse. Our study reveals that prophylactic infusion of allogeneic CD8+ T cells reactive against peptides derived from CML-associated antigens is a safe and promising therapeutic strategy. This trial was registered at www.clinicaltrials.gov as #NCT00460629.

Description: Background Lenalidomide (LEN) results in hematopoietic improvement in the majority of treated patients with myelodysplastic syndrome and isolated deletion of the long arm of chromosome 5 (MDS-del5q). The drug is believed to exert antiangiogenetic, antiproliferative, proerythropoietic and immunomodulatory effects. The precise nature especially of the latter mode of action is incompletely understood. While some reports in multiple myeloma argue for an immunosuppressive effect, others proposed immunostimulatory properties. This prompted us to study the effects of LEN on the adaptive immune system in a cohort of patients with MDS-del5q. Patients, Material and Methods The composition of the T-cell compartment in peripheral blood (pB) and bone marrow (BM) of 11 patients with MDS-del5q treated within a multicenter, single-arm, phase II study of the German MDS study group (LEMON-5 trial) was studied by flow cytometry. Patients received LEN 10 mg per day for 21 days of each 28 day cycle. The median age was 71 years and 9 patients were female. Regulatory T-cells (Treg) were defined as CD3+CD4+CD25++CD127dimFoxP3+ while Th17 cells had to be CD3+CD4+ with positive intracellular staining for IL17 and negative intracellular staining for IFNgamma. Peripheral blood samples were taken prior to initiation of each new LEN cycle and BM specimens were obtained prior to treatment initiation and after 5 to 7 cycles of LEN. The results obtained in MDS patients were compared to a cohort of 27 healthy volunteer donors. Results After a median of 6 cycles all but one patient had responded with a hematologic improvement of erythrocytes, while 6 individuals achieved a complete morphologic and 5 a complete cytogenetic remission. In comparison to healthy donors, MDS patients had significantly lower frequencies of Treg in pB (5.0 vs. 7.4 %, p = 0.008) and BM (3.4 vs. 7.4 %, p = 0.006). After initiation of LEN treatment we observed a rapid increase in Treg frequencies in pB (Figure 1A) and after a median of 6 cycles the Treg were significantly more abundant in pB (9.6 % p = 0.001, Figure 1B) and BM (8.1 %, p = 0.016, Figure 1C) compared to baseline. The median Treg frequency reached the level observed in normal donors in BM and surpassed it in pB (Figure 1B+C). Fluorescence-in-situ hybridization demonstrated that the expanded Treg were not part of the malignant clone. Interestingly, although the median absolute lymphocyte count increased during LEN treatment (pre: 1.10 Gpt/l vs. post: 1.43 Gpt/l p = 0.013) there were no significant changes with respect to relative lymphocyte frequency (pre: 68.3 %, post: 75.61 %, p=0.499) or the percentage of T-effector cells (pre: 48.0 %, post: 51.5 %, p=0.520). Moreover, the expansion of Treg was not accompanied by a counter-regulatory increase of pro-inflammatory Th17-cells (pre: 2.3 % vs. post: 1.33 %, p = 0.328). To address a possible mechanism of the observed Treg expansion we analyzed the concentration of transforming growth factor beta (TGFbeta), a key factor contributing to the development of Treg, in pB and BM plasma. However, LEN treatment was not associated with an increase of TGFbeta levels in these two compartments. On the contrary, the TGFbeta levels after 6 cycles of LEN were significantly lower than pre-treatment values in pB and BM (pB 4.9 vs. 2.3 ng/ml, p = 0.039; BM 4.5 vs. 0.8 ng/ml, p = 0.023). Discussion Our results show that LEN treatment corrects an MDS-del5q-inherent Treg deficit. This Treg expansion is not accompanied by a counter-regulatory increase in pro-inflammatory Th17 or other T-effector cells and therefore possibly results in an immunosuppressive milieu, which might contribute to the beneficial effects of the drug on myelodysplastic hematopoiesis but may also facilitate leukemic progression. Figure 1 Figure 1. Disclosures Platzbecker: Celgene: Research Funding.

Description: Introduction CMV-infection is a serious complication in patients after allogeneic stem cell transplantation (SCT) where immunosuppressive therapy and impaired T cell reconstitution result in a high risk for viral infections. Monitoring of CMV-virus load by PCR and preemptive therapy are important tools to prevent CMV disease. However, CMV specific cytotoxic T cells (CMV-CTLs) are needed to successfully control CMV-infections. CMV-specific multimers composed of the patients HLA Class I molecule bound to CMV pp65 epitopes give the possibility to monitor CMV-CTLs. Here, we present the case of CMV-reactivation following SCT for AML. Methods The percentage of CMV-specific CD8+ T cells was determined by flow cytometry and mapped to clinical and laboratory parameters of the patient. CD8+ T cells were detected using CD8-fluorescein isothiocyanate (FITC, Beckman Coulter) antibody and CD3 as a T-cell marker was labeled with CD3-allophycocyanin (APC, MACS Miltenyi Biotec) antibody. CMV-specific CD8+ T cells were detected using the CMV major histocompatibility complex (MHC) with Strep-Tactinphycoerythrin (PE) conjugate (Streptamers, IBA GmbH). Case A 60 years old male patient was diagnosed with acute myeloid leukemia (AML) with 95% myeloid blasts in the bone marrow and extramedullary AML manifestations at the time of diagnosis. Following induction therapy the patient was transplanted from a matched unrelated donor. The stem cell recipient as well as his donor had been tested sero-positive for CMV prior to SCT. Within the first month following transplantation, the patient developed an effective CMV specific immunity as seen by high levels of CMV-specific T cells (Figure 1). About three months following transplantation the patient was diagnosed with intestinal GVHD requiring high-dose glucocorticoid treatment. Following steroid exposure, levels of CMV-CTLs dropped and shortly thereafter rising CMV-copy numbers were observed which was accompanied by clinical signs of CMV enteritis. With the administration of antiviral treatment the CMV specific virus load decreased. However, levels of CMV-CTLs remained low, presumably as a result of ongoing steroid exposure. Discussion High levels of CMV-CTLs appeared to control CMV, as seen by a non-detectable virus load in standard PCR testing. The close correlation between the drop in CMV-CTL count and CMV activation highlights the potential of this method to monitor and understand immune responses to CMV following SCT. Of note, early presence of high frequencies of CMV-CTLs did not guarantee CMV-control under steroid exposure as seen in our case. Previous reports have suggested that high dose glucocorticoids may impact CMV-CTLs survival. This is supported by our case, where we see a rapid drop in CMV-CTLs following glucocorticoid exposure. However, the exact molecular mechanisms and more importantly, the predictive value of this finding remain elusive. Furthermore, these data suggest, that patients with ongoing high steroid exposure may not benefit from a transfer of CMV-specific T-cells to control CMV disease. Conclusion Further investigations to clarify the potential of CMV-CTL measurements and to understand the effect of steroid exposure at the functional level are warranted. Studies to correlate CMV-CTL counts with the level of immunosuppression and their influence on controlling CMV-disease will follow. In future, this tool could provide a chance to select patients at high risk of CMV reactivation who could profit from an individualized monitoring and early treatment. Disclosures: No relevant conflicts of interest to declare.

Description: Abstract 4167 Acute graft-versus-host disease (GvHD) following allogeneic hematopoietic cell transplantation (HCT) has been classically assumed to be mediated by T helper cells type 1 (Th1), characterized by the production of interferon-γ (IFN-γ). Recently, Interleukin 17A (IL-17)-producing CD4+ T helper type 17 (Th17) cells have also drawn attention as possible effector cells of acute GvHD in murine models. Their role following allogeneic HCT in humans is unknown. We hypothesized that IFN-γ/IL-17-production and quantity of T helper cells might depend on the time-point after HCT, immune responses to allo-antigens (GvHD) or pathogens (e.g. bacterial infection, CMV reactivation) and the presence of T cell depleting antibodies (e.g. ATG). To explore this hypothesiswe initiated a prospective study to investigate the reconstitution of Th1, Th1/17 and Th17 cells in patients after HCT. 80 consecutive patients with various hematologic disorders undergoing allogeneic human leukocyte antigen (HLA)-matched HCT at our center between 12/2009 and 9/2010 were included into the study. Blood samples were collected once in the 1st month, 2nd month and 3rd month after HCT. To quantify IL-17- and IFN-γ-producing T helper cells, we used surface staining for CD3 and CD4 followed by intracellular cytokine staining for IFN-γ and IL-17 in PBMCs. T helper cells producing both IFN-γ and IL-17 (IFN-γ+IL-17+) were termed Th1/17 cells, T helper cells producing only either cytokine alone are indicated as Th17 cells (IFN-γ−IL-17+) or Th1 cells (IFN-γ+IL-17−). For each time period patient cohorts were defined according to the subsequent criteria: (i) bacterial infection (C-reactive protein 〉50 mg/L, positive blood culture and/or fever in the absence of viral or fungal infection), (ii) CMV reactivation (positive CMV-specific PCR), (iii) acute GvHD (according to the Seattle criteria) and (iv) ATG in the conditioning regimen (dose of 20mg/kg on day -3, -2 and -1 before HCT). As a reference group we chose time-matched and age-matched patients that did not meet any of the criteria under investigation. Student′s t test (two sided, unpaired) was used for statistical evaluation. In all patients with no relevant complication (absence of bacterial infection, CMV reactivation, acute GvHD) and no ATG in the conditioning regimen Th1, Th1/17 and Th17 cells were detectable within the first month after HCT. However, these T helper cell subsets did not reconstitute to levels of healthy controls within the first 3 month after HCT. In contrast to Th1 cells, no further expansion of Th1/17 and Th17 cells was observed following the 1st month after HCT. ATG during conditioning significantly reduced the frequency of Th17 cells at all time-points analyzed (median decrease: 1st month, 71.5%, P=0.0049; 2nd month, 82.5%, P=0.0002; late engraftment, 71.4%, P=0.0011). Th1/17 cells were also suppressed in patients with ATG, although this reduction was less prominent and reached no significance following the 2nd month after HCT (median decrease: 1st month, 76.19%, P=0.012; 2nd month, 70.11%, P=0.054; late engraftment, 50.7%, P=0.69). Finally, Th1 cells were not significantly reduced in patients receiving ATG compared to time-matches controls (median decrease: 1st month: 89.18%, P=0.34; 2nd month: 62.7%, P=0.21; late engraftment: 19.9%, P=0.8), indicating that the suppressive effect of ATG is less pronounced on Th1 and TH1/17 cells, compared to Th17 cells. Acute GvHD°I was not associated with significant changes in the size of the Th1, TH1/17 or Th17 cell subsets. In patients with GvHD°II-IV a tendency towards increased counts of Th1, TH1/17 and TH17 cells in the peripheral blood was observed. However, these changes were not statistically different compared to time-matched controls. CMV reactivation triggered the expansion of all T helper subsets and Th1 cells showed the strongest increase (median increase: Th1, 449.1%, P=0.00075; Th17, 74.9%, P=0.00069; Th1/17, 97.1%, P=0.00012). In contrast, no significant changes were found in the T helper cell compartment of patients with bacterial infection compared to time matched controls. In conclusion, quantitative reconstitution of Th1, Th1/17 and Th17 cells is impaired within the first 3 months after HCT, especially when ATG is administered during conditioning. CMV reactivation, but not bacterial infection, triggered the absolute expansion of these T cell subsets. Disclosures: No relevant conflicts of interest to declare.

Description: Current approaches for treatment of late-stage breast cancer rarely result in a long-term cure. In part this is due to tumor stroma that prevents access of systemically or intratumorally applied therapeutics. We propose a stem cell gene therapy approach for controlled tumor stroma degradation that uses the pathophysiologic process of recruitment of inflammatory cells into the tumor. This approach involves genetic modification of hematopoietic stem cells (HSCs) and their subsequent transplantation into tumor-bearing mice. We show that inducible, intratumoral expression of relaxin (Rlx) either by transplanting tumor cells that contained the Rlx gene or by transplantation of mouse HSCs transduced with an Rlx-expressing lentivirus vector delays tumor growth in a mouse model of breast cancer. The antitumor effect of Rlx was mediated through degradation of tumor stroma, which provided increased access of infiltrating antitumor immune cells to their target tumor cells. Furthermore, we have shown in a human/mouse chimeric model that genetically modified HSCs expressing a transgene can access the tumor site. Our findings are relevant for cancer gene therapy and immunotherapy.