| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Original Article |
Department of Interdisciplinary Oncology, H. Lee Moffitt Cancer Center & Research Institute, at the University of South Florida, Tampa, FL, USA
Correspondence: Address correspondence and reprint requests to: Sophie Dessureault, H. Lee Moffitt Cancer Center & Research Institute, 12902 Magnolia Drive, Tampa, FL 33612, USA; E-mail: dessursm{at}moffitt.usf.edu
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONPATIENTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Methods: Patients received three intradermal vaccine injections (irradiated autologous tumor cells plus GM.CD40L bystander cells) at 28-day intervals. Patients with no disease progression received three additional vaccines at 4, 12, and 24 months. Patients were monitored for toxicity, tumor response, and tumor-specific immune responses.
Results: Twenty-one patients received at least three vaccine injections, with no toxicity attributable to the vaccine. Immunohistochemistry of vaccine injection site biopsies with CD1a and CD86 antibodies confirmed recruitment and activation of dendritic cells. There was no tumor regression after vaccination, but many patients had stable disease, including six of ten melanoma patients. Four patients developed tumor-specific T-cell responses on ELISPOT testing. One patient, who had stable disease for 24 months, demonstrated an increase in MART-1-specific T-cells by tetramer analysis after re-immunization; biopsy of the tumor that progressed 2 years after the onset of vaccination revealed a massive peritumoral and intratumoral T-cell infiltrate.
Conclusions: Vaccination of cancer patients with autologous tumor cells and GM.CD40L bystander cells (engineered to express GM-CSF and CD40L) is safe, can recruit and activate dendritic cells, and can elicit tumor-specific T-cell responses. Phase-II trials are underway to evaluate the impact of bystander-based vaccines on melanoma and mantle cell lymphoma.
Key Words: cancer vaccine • immunotherapy • phase-I clinical trial • bystander • CD40 ligand • GM-CSF
Abbreviations: APC, antigen-presenting cell • CD40L, CD40 ligand • DC, dendritic cell • DTH, delayed-type hypersensitivity • ELISA, enzyme-linked immunosorbent assay • GM-CSF, granulocyte macrophage-colony stimulating factor • IL, interleukin • MIP-1, monocyte inflammatory protein-1 • MRD, minimal residual disease • PBMC, peripheral blood mononuclear cell • RECIST, Response Evaluation Criteria in Solid Tumors • TAA, tumor-associated antigen
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONPATIENTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Granulocyte macrophage-colony stimulating factor (GM-CSF) secretion at a vaccine site results in the recruitment and differentiation of DCs.10,11 and many GM-CSF-based vaccines are being developed to prime anti-tumor immune responses in clinical trials.12–16 GM-CSF-producing tumor cells serve as the source of TAAs that are processed by the recruited DCs, which subsequently migrate to lymph nodes where they activate TAA-specific T cells. Chiodoni et al.17 generated an improved immune response when CD40L was combined with GM-CSF in a murine model. Tumors that formed in mice transplanted with tumor cells that had been transfected with both the GM-CSF and CD40L genes were infiltrated with a larger number of DCs, and these tumor-infiltrating DCs could prime a tumor-specific CTL response in vivo. Murine C-26 colon carcinoma cells cotransduced with GM-CSF and CD40L showed reduced tumorigenicity and even regression after tumor take in some mice, resulting in 70% tumor-free mice.
The need for in vitro cell culture and individualized gene transfer is expensive, labor-intensive, and limited by variable levels of gene expression. Levitsky et al.18 described the use of a vaccine in which universal MHC-negative GM-CSF-producing "bystander cells" (secreting significant levels of GM-CSF in the local vaccine site microenvironment for several days after lethal irradiation) are mixed with irradiated but otherwise unmodified fresh autologous tumor cells (antigen source). With the bystander vaccine approach, there is no need to genetically manipulate autologous tumor cells. Not only does this strategy greatly simplify the clinical development of autologous tumor cell-based vaccines, but also it allows for the use of unmodified tumor cells, without selection for various antigenic determinants in culture.
Levitsky’s group demonstrated that immunization of mice with a mixture of autologous wild-type tumor and allogeneic GM-CSF-producing bystander cells could impart effective antitumor protection and eradicate pre-established lymphoma.18 Brenner et al. used a similar bystander approach and tested a CD40L-based tumor vaccine in a murine model of multiple myeloma.19 They admixed CD40L-transduced bystander cells with tumor cells as the source of tumor antigen. This vaccine was very effective in protecting mice from a tumor challenge by recruiting and activating professional APCs at the vaccine site.
Our laboratory created a universal MHC-negative bystander cell line (GM.CD40L), which secretes GM-CSF and expresses CD40L on its cell surface.5 Pre-clinical experiments have shown that the GM.CD40L bystander cell line enhances the activation of anti-tumor T-cell responses in an ex vivo human mixed autologous tumor cell/lymph node cell model. Flow cytometric analysis of DCs and cytokine profiling of culture supernatants confirm that the GM.CD40L effect is mediated through DC activation. DCs co-cultured with the GM.CD40L bystander cells show an increase in surface maturation markers (CD83, CD86, HLA-DR) and secrete high levels of interleukin (IL)-12 and monocyte inflammatory protein-1 (MIP-1). IL-12 is also detected in our mixed autologous tumor cell/lymph node cell model, further confirming that the GM.CD40L effect is mediated through DC activation. These preclinical findings in a human model provided a rationale for testing this approach in a phase-I clinical trial.
|
PATIENTS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONPATIENTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Eligibility Criteria
Patients over the age of 18 who had a tissue diagnosis of malignancy and who were deemed incurable and not suitable for standard forms of therapy were considered for enrollment. Only patients with a resectable primary tumor or metastatic lesion were eligible for study. Patients were required to have an Eastern Cooperative Oncology Group (ECOG) performance status of 0, 1, or 2, along with adequate bone marrow, hepatic, and renal function. Patients receiving immunosuppressive doses of glucocorticoids were not eligible for the study. Patients with symptomatic brain metastases were excluded. Prior chemotherapy, immunotherapy, or radiotherapy were allowed if the patient had recovered from all toxic effects and treatment had been completed at least 4 weeks before vaccination.
Generation of the GM.CD40L Bystander Cells
The GM.CD40L bystander cell line was established in our laboratory by transfecting the human erythroleukemia K562 cell line20 with the genes for hGM-CSF and hCD40L. A detailed description of the methods was published earlier.5 Like the K562 parent cell line from which it is derived, the GM.CD40L bystander cell line is an MHC-negative cell line that grows in cell suspension. A Master Cell Bank (MCB) was generated by serial subculture and expansion of the original GM.CD40L clone until 4 x 108 cells were available for simultaneous harvest and cryopreservation. This created a uniform population of cells which was divided equally into 19 vials (2 x 107 cells per vial) and stored in the vapor phase of liquid nitrogen. The Manufacturer’s Working Cell Bank (MWCB) was generated from two ampoules of the MCB. The MCB source cells (passage 8) were thawed and expanded by serial subculture in AIM-V serum-free medium (Life Technologies, GIBCO BRL) containing hygromycin B (500 mg/mL) (Sigma Aldrich, St. Louis, MO, USA). Cells were removed from hygromycin B-containing medium and returned to fresh AIM-V serum-free medium 48 h prior to final harvest for the MWCB. Cell viability of these harvested cells, as determined by trypan blue exclusion, was 83%. A fraction of the cells was dispensed into 48 individual ampoules (2 x 107 cells per ampoule), and cryopreserved to form the MWCB. Another fraction of the cells was irradiated (15,000 rad) and dispensed into 81 ampoules (5 x 106 cells per ampoule), and cryopreserved to form the first lot (L001) of the biological vaccine product. All subsequent lots (L002, L003, L004, and so on) were generated from single ampoules of the MWCB.
GM.CD40L Cell Bank Qualification and Identity Testing
The MCB was tested for adventitious agents (including bacteria, fungi, mycoplasmas, and viruses) under Good Laboratory Practices (GLP)-compliant conditions. Identity testing of the MCB, also under GLP-compliant conditions, included cytogenetic (karyology) and isoenzyme analysis (carried out by Applied Genetics Laboratories, Inc., Melbourne, FL), as well as tests needed to establish all significant properties of the cells (including negative MHC cell surface expression, positive CD40L cell surface expression, and GM-CSF secretion into the local microenvironment) and the stability of these properties throughout the manufacturing process (including irradiation and a freeze/thaw cycle). Absence of MHC expression by the GM.CD40L cell line was confirmed by flow cytometry using pan-antibodies [PE-conjugated murine anti-human HLA-A,B,C (MHC class I) antibody, Pharmingen Catalog # 555553, Clone G46-2.6; FITC-conjugated murine anti-human HLA-DR,DP,DQ (MHC class II) antibody, Pharmingen Catalog # 555558, Clone TY39]. CD40L expression was quantified by flow cytometry using the mouse FITC-conjugated anti-human CD40L (CD154) monoclonal antibody (Catalog # 353-040; Immunology Research Products, Ancell Corp., Bayport, MN). GM-CSF expression was quantified by ELISA. Briefly, 1 x 106 cells were grown in 5 mL of AIM-V serum-free medium, and the supernatant was collected 24 h later. Human GM-CSF was quantified with ELISA kits (Catalog # DGM00; R&D Systems, Minneapolis, MN, USA) as per the manufacturer’s instructions. The level of GM-CSF produced by 1 x 106 cells over 24 h (300–700 ng/1 x 106 cells/24 h) was used as an indication of cell purity over time. The MWCB, being derived from the MCB and propagated for a maximum of 150 passages in tissue culture, was spot checked for contaminants that may have been introduced from the culture medium. Tests, including testing for sterility, screening for mycoplasma, in vitro and in vivo testing for viral contamination, and checking for cell line cross-contamination, were completed under GLP-compliant conditions and showed the MWCB to be free of contaminants. Lot-to-lot characterization of the GM.CD40L irradiated vaccine product and routine monitoring for adventitious agents is carried out by the Cell Therapy Core Facility at the HLMCC under Good Manufacturing Practices (GMP)-compliant conditions. This includes testing of production cell cultures and unprocessed and processed cell culture fluids. Quality control testing is performed on each bulk lot. Testing for bacterial and fungal sterility is performed on the unprocessed bulk lot, the final bulk lot, and the final product. The unprocessed bulk is the pooled harvests of cell culture fluids that constitute a homogeneous mixture for manufacture into a unique lot of product. Testing for adventitious agents is performed prior to further processing, such as spinning, washing, and resuspending of cells. The final bulk is the concentrated, washed GM.CD40L cell suspension prepared for irradiation, distribution into separate vials or aliquots, and final freezing for vaccine purposes. The final bulk is subjected to a variety of lot release tests which will include sterility testing. The final product—GM.CD40L cells which have been irradiated, frozen, and thawed—is tested for sterility and endotoxin. Routine testing for mycoplasma and in vitro and in vivo testing for adventitious viruses is performed on every lot using production cells and unprocessed bulk fluids. All lots used for the phase I study were fully tested and passed all sterility and identity testing, including confirmation of negative MHC cell surface expression (by flow cytometry), positive CD40L cell surface expression (also by flow cytometry), and GM-CSF secretion into the local microenvironment (by ELISA).
Autologous Tumor Cells
Resected tumors were rinsed in betadine for 60 s and then transported immediately from the operating room to the Cell Therapy Core Facility at the HLMCC in sterile RPMI medium. Tumors were minced under sterile conditions into small pieces using two crossed #10 scalpels. Tissue fragments were collected, rinsed with phosphate-buffered saline (PBS), and suspended in a collagenase-based enzymatic cocktail (RPMI with collagenase, 2 mg/mL; pronase E, 0.2 mg/mL; DNase, 600 U/mL), and stirred at 150–200 rpm using a sterile magnetic stir bar for 2–3 h, until tissue fragments were fully digested.5,21 Cell suspensions were filtered through a 40-mm nylon mesh, washed twice with PBS, and resuspended in serum-free X-VIVO medium to a concentration of 107 cells/mL. An aliquot of the cell suspension was subjected to sterility testing (Microbiology Department, HLMCC). Then, 2 x 107 cells were frozen and stored in liquid nitrogen as a reserve; 8 x 106 cells were frozen and stored for subsequent in vitro assays of immune function; 3 x 106 cells were irradiated (15,000 rad) and stored as two aliquots in liquid nitrogen for use in the delayed-type hypersensitivity (DTH) skin testing. All cells were frozen in 70% Plasmalyte, 20% human serum albumin, and 10% DMSO. The remaining cells were resuspended in seven parts plasmalyte:two parts human serum albumin, irradiated (15,000 rad), and then one part DMSO was added. Cells were then frozen using a programmed stepdown freezer and stored in the vapor phase of liquid nitrogen (<–170°C) in aliquots of 5 x 106 cells/1.5 mL for subsequent vaccine preparations. All cryopreservations were carried out in a Forma Cryomed model 1010-controlled rate freezing apparatus.
Vaccine Production, Release Criteria, and Administration to Patients
One vial containing 5 x 106, 10 x 106, or 15 x 106 (depending on patient cohort) radiated autologous tumor cells and one to three vials (depending on patient cohort) containing 5 x 106 radiated GM.CD40L cells each were thawed rapidly by immersion in a 37°C waterbath, diluted in 10 mL of sterile saline for 15–30 min at 37°C, centrifuged, and resuspended in a final volume of 1.1 mL of sterile saline. The reconstituted vaccine was drawn up in a 1-cc syringe with an 18G needle. The needle was removed and 0.1 mL of aliquot was drawn from the syringe with a TB syringe for microbiological testing (including Gram stain and cultures for aerobes, anaerobes, and fungi) and immunofluorescence analysis after staining with DAPI (to eliminate the possibility of mycoplasma contamination). A 
" 25G needle with an intradermal bevel was added to the syringe and all air bubbles were expressed. If testing revealed an absence of microbial contamination, the vaccine lot was considered adequate for clinical use and was then transported on ice to the Clinical Research Unit at the HLMCC. This cell suspension was injected intradermally into eight separate injection sites (two intradermal injections in each of four nodal basins: bilateral axillary and bilateral inguinal nodal basins). Each injection consisted of approximately 0.13 mL of cell suspension. Basins that had undergone prior lymph node dissections were not injected, so injections originally planned for those sites were distributed evenly amongst the remaining naïve nodal basins.
Treatment Schema
Treatment consisted of intradermal vaccine injections at 28-day intervals for a total of three immunizations. Injections were performed on days 1, 29, and 57. The vaccine, consisting of 1 mL of the cell suspension (GM.CD40L bystander cells admixed with an equivalent number of thawed autologous tumor cells), was administered into eight separate injection sites, as described earlier. The dose of the vaccine formulation was escalated in successive cohorts of patients by increasing the number of autologous tumor cells and the number of GM.CD40L cells administered: cohort 1 patients received 5 million tumor cells +5 million bystander cells; cohort 2 patients received 10 million tumor cells +10 million bystander cells, except for the first injection, which consisted of 10 million tumor cells +5 million bystander cells; cohort 3 patients received 10 million tumor cells +10 million bystander cells; and cohort 4 patients received 15 million tumor cells +15 million bystander cells. Three patients were treated in each cohort. All patients in a cohort were observed for a minimum of 4 weeks from the time of the first vaccine injection before any patient was treated at the next dose level. Patients were monitored for evidence of toxicity, the development of a specific immune response, and objective tumor responses. Patients who had no disease progression after the first three vaccines and who had no toxicity attributable to the vaccine were eligible for an additional three vaccines administered every 28 days, starting on day 85, as long as sufficient material was still available. Patients who still had no evidence of disease progression at 12 months were eligible for three booster vaccines at 28-day intervals.
Toxicity Assessment
Patients were observed in the Clinical Research Unit at the HLMCC for 1 h after each vaccine. Vital signs were recorded, local and systemic effects were noted, and toxicity was scored using the National Cancer Institute Common Toxicity Criteria, version 2.0 (CTC v 2.0). Dose-limiting toxicity (DLT) was defined as any grade 3 toxicity. Serial blood samples were taken for measurement of hemoglobin, white blood count, platelet count, electrolytes, urea, creatinine, liver enzymes, and bilirubin.
Immune Response Testing
Immunogenicity assays were carried out prior to the first vaccine and after the third vaccine. Vaccine immunogenicity was measured by in vitro testing of peripheral blood mononuclear cells (PBMCs) for cytokine-secreting T cells in ELISPOT assays, DTH skin testing with irradiated autologous tumor cells, and biopsy of the DTH site for immunohistochemical analysis of infiltrating lymphocytes 48 h after intra-dermal injection of autologous tumor cells. Biopsy of the second vaccine injection site was used to explore the local response to vaccination.
ELISPOT Analysis
Exactly 60 cc of heparinized blood (five green top tubes) were obtained from each patient on three separate occasions during the 2-week period prior to the first vaccine injection (to establish baseline T-cell precursor frequency), on day 71 (2 weeks after administration of the third vaccine injection), on day 141 (3 months after the third vaccine injection), at 6, 9, and again at 12 months (to assess for the development/persistence of memory T-cell populations). PBMCs were isolated by Ficoll gradient centrifugation. PBMCs from each sample were suspended in 90% human autologous serum and 10% warm DMSO, and frozen in liquid N2 using a programmed stepdown freezer until all samples for each patient were collected. For testing, all of an individual’s PBMC samples were thawed at the same time, along with a frozen aliquot of the individual’s autologous tumor cells. PBMCs were washed twice with PBS containing 5% human serum and then cultured in AIM-V medium at a density of 3 x 106 PBMCs in 2 mL of the medium. The PBMCs were restimulated in vitro with 105 radiated autologous tumor cells. After 48 h in culture, the cells were recovered and washed. These restimulated PBMCs were then tested in 20-h ELISPOT assays using Cytokine ELISPOT kits from BD Biosciences (Pharmingen, San Diego, CA, USA). Restimulated PBMCs were transferred into the wells of an ELISPOT plate that had been precoated with a capture antibody specific for human IL-2 or 
-interferon. Autologous tumor cell lysate, generated by subjecting tumor cells to six cycles of freezing and thawing, was added to each well as a fresh source of tumor antigens. 50 µL of lysate made from 2 x 105 tumor cells was added to each well of the ELISPOT plates. After an 18-h incubation in a humidified 37°C, 5% CO2 incubator in AIM-V medium containing 10% human serum, the wells were washed and the biotinylated detection antibody specific for human IL-2 or -interferon was added to the wells. Unbound biotinylated antibody was washed away and an avidin horseradish peroxidase conjugate was added. Following four washes to remove the unbound enzyme, the substrate solution (AEC) was added. A colored precipitate formed and appeared as spots at the sites of cytokine localization. The number of spots was counted using an automated CTL ELISPOT reader. An assessment of immune competence and a positive control for the ELISPOT assays was obtained by immunizing patients with tetanus toxoid at the time of enrollment. All ELISPOT assays included reactions to test for antitetanus immune responses: PBMCs were restimulated in vitro with tetanus toxoid in replicate ELISPOT assays.
DTH Skin Testing and Immunohistochemical Analysis of the Biopsy Site
DTH testing was performed within 2 weeks prior to the first vaccine, and again on day 78, after all the three vaccines had been administered. Cryopreserved aliquots containing 106 previously irradiated but viable tumor cells were thawed, washed, and resuspended in 0.2 mL of plasmalyte A. These cells were injected intradermally in the forearm and marked. Forty-eight hours later, the injection site was inspected and the diameter of the induration and erythema was recorded. Each patient underwent a punch biopsy of the DTH injection site 48 h after administration. Immunohistochemical staining of paraffin-embedded sections for CD3, CD4, and CD8 (T-cell markers); CD1a (APCs); and CD68 (monocyte marker) was performed to study the mononuclear cell infiltrate. For staining, deparaffinized tissue sections were treated with methanol containing 0.3% H2O2 to block endogenous peroxidase activity, and antigen retrieval was performed by microwaving. The primary monoclonal antibodies were then applied, and visualization was achieved using a biotin-IgG/strep-tavidin-horseradish peroxidase immunodetection kit and diaminobezidine chromogen. All sections were scored by a pathologist (JM) in a blinded fashion. Infiltration of lymphocytes was quantified by the relative density of infiltration (+, low; ++, intermediate; +++, high).
Measurement of GM-CSF Serum Levels
Serum was collected immediately before the first vaccine and then 24 and 96 h later. GM-CSF levels were determined by ELISA (R&D Systems, Minneapolis, MN, USA) according to the manufacturer’s instructions. In brief, serum was separated from whole blood by centrifugation at high speed for 10 min and frozen in 1-mL aliquots at –80°C until the day of testing. All samples were run simultaneously. Each assay plate included a GM-CSF control standard and sample levels were determined using linear regression analysis.
Biopsy of the Vaccine Site
A 3-mm punch biopsy of the second vaccine site (injected on day 29) was taken prior to injection of the third vaccine and examined histologically in order to assess for persistence of irradiated bystander and autologous tumor cells, to document recruitment and activation of DCs, and to analyze the profile of any lymphocytic infiltrate. Tissue sections were stained for CD3, CD4, and CD8 (T-cell markers); CD1a (APCs); CD68 (monocyte marker); CD56 (NK cell marker); and CD86 (activated APCs).
Tetramer Staining Analysis
PBMCs from one HLA-A2 positive patient (#5) were obtained before and after her 2-year anniversary booster vaccine for better characterization of the immune response demonstrated on earlier ELISPOT assays. PBMCs were resuspended in PBS containing 2% FCS and 0.1% sodium azide, aliquoted into wells of a 96-well plate, and stained with tetrameric MHC class I/peptide staining reagent consisting of HLA-A*0201 Mart-1 (ELAGIGILTV) conjugated to PE (Beckman Coulter Immunomics, San Diego, CA, USA) along with an anti-CD8 monoclonal antibody conjugated to FITC. After staining and washing, cells were analyzed using two-color flow cytometric analysis on a FACScan instrument (Becton Dickinson, San Jose, CA, USA). For analysis, a dual gate was drawn around the lymphocyte subset in the forward scatter versus side scatter histogram and the CD8 (FITC)-positive cells. The number of cells that were double-positive for FITC and PE was determined in the FITC versus PE histograms of the dual-gated cells. This was taken as the proportion of tumor antigen-specific T cells that was present in the peripheral blood.
Clinical Responses
Tumor response rates were estimated in those patients with measurable disease. CT scans were obtained within 2 weeks before the first vaccine and again 3 weeks after the third vaccine was administered. Size of clinical lesions (e.g., skin nodules and palpable lymph nodes) were measured at the time of entry into the study (within 2 weeks of the first vaccine) and on day 78 (3 weeks after all the three vaccines were administered). Response and progression were evaluated using the international criteria proposed by the Response Evaluation Criteria in Solid Tumors (RECIST) Committee. For patients who had stable disease, partial response to treatment, or complete response to treatment, radiographic or clinical measurements of target lesions were made every 3 months until there was evidence of disease progression. Time to disease progression was estimated from available clinical and radiographic assessments.
Statistical Considerations
This phase-I trial was designed to examine the safety of the GM.CD40L bystander vaccine formulation administered to cancer patients with stage IV disease. The MTD of the GM.CD40L bystander cell concentration was not reached in this study. Due to the heterogeneity of the patient population, most other results were descriptive in nature.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONPATIENTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. Thirteen (50%) patients had melanoma; four (15%) had renal cell carcinoma; four (15%) had mantle cell lymphoma; two had sarcoma; one had small cell lung cancer; one had non-small cell lung cancer. All patients had metastatic disease at the time of enrollment, but after surgical resection for vaccine production, 7 (27%) did not have measurable disease; 19 (73%) had measurable disease by RECIST criteria. Four patients (# 3, 8, 13, and 24) died early of rapidly progressive metastatic disease before receiving all three vaccine doses; one patient (# 19) moved away for personal reasons and aborted vaccine therapy after his second vaccine. The remaining 21 patients received at least three vaccine doses and, therefore, were considered evaluable for toxicity and response to treatment. Most patients (19 of 26, or 73% of the study population) had received and failed to respond to prior systemic therapy.
|
). All signs of local inflammation resolved within 48 h.
GM-CSF Is Detected in the Serum after Vaccine Treatment
Serum GM-CSF was measured before and after the first vaccine injection. The peak in serum GM-CSF was detected 24 h after vaccination, and serum levels dropped over time, with a return to baseline within 5 days (Fig. 1). The mean peak serum level 24 h after vaccination was 0.56 pg/mL (range 0–2.4 pg/mL).
|
). CD1a staining revealed, in addition to normal numbers of Langerhans cells in the epidermis, successful recruitment of APCs in the dermis of all 20 (100%) of 20 vaccine sites tested (Fig. 2B). Staining with CD86 revealed weak (+), moderate (++), and strong (+++) levels of activated APCs in the dermis of 15 (79%) of 19 vaccine sites tested (Fig 2C). Results are summarized in Table 1. Resident DCs in the epidermis (not exposed to the bystander cells) remained CD86-negative, consistent with the hypothesis that direct contact with the bystanders was necessary for CD40L-induced activation.22 Twelve of 17 patients had occasional CD56-positive natural killer (NK) cells in biopsies of the vaccine injection sites, but this was not associated with NK cell-mediated killing of the MHC-negative K562-derived bystander cells.
|
). Figure 3 shows photomicrographs of a DTH site biopsy from a patient with a moderate DTH T cell response. Immunohistochemical staining shows that most DTH site infiltrates had a 2:1 ratio of CD4 to CD8 T cells (data not shown). Since naïve T cells do not have the capacity to extravasate, these most likely represent responding anti-tumor T cells.
|
-interferon secretion. Nineteen patients had samples before and after vaccine available for evaluation (Table 1). Of these, ten (53%) had no evidence of tumor-specific T cells either before or after vaccine, but these patients also had few (# 21 and 26) or no (# 1, 2, 9, 11, 17, 18, 22, and 23) cytokine-producing T cells identified in control wells containing tetanus toxoid, indicating general immune suppression rather than a failure of the cancer vaccine to generate a tumor-specific immune response.
Results from the other nine (47%) patients are shown in Fig. 4. Before beginning treatment, the majority of these patients had no detectable autologous tumor-reactive T cells in the peripheral blood. However, after receiving vaccine injections, four of these nine patients demonstrated an increase in the number of autologous tumor-reactive T cells. This increase occurred at different times for different patients, with a significant increase in one patient (# 5) occuring 6 months after the first vaccine injection. Of note is the fact that three of these four patients (who demonstrated an increase in the number of spots after vaccination) had stable disease after three vaccines and had a median time to tumor progression of 24 months (patients # 4, 5, and 7). On the other hand, two of the other four patients (who demonstrated no increase in the number of spots after vaccination) had stable disease after three vaccines, with a time to tumor progression of 14 months (patients # 10 and 14).
|
shows that patient # 5 had minimal residual disease (MRD) at the onset of vaccine therapy. She underwent wide local excision and sentinel lymph node biopsy in April 2001 for a primary melanoma of the left lower extremity. All sentinel lymph nodes were negative for malignancy. She was well until August 2002, when she developed multiple satellite lesions in the distal aspect of the left leg. In September 2003, she underwent a hyperthermic isolated limb perfusion with high-dose melphalan, along with a complete superficial groin lymph node dissection which revealed one of nine lymph nodes positive for metastatic disease. Shortly thereafter, she underwent surgery for metastatic melanoma in the inferior aspect of the left breast and left flank. She was started on systemic GM-CSF therapy and was well until June 2003, when she developed gross metastatic disease in the left iliac fossa. She underwent an iliac lymph node dissection, and tissue from an iliac lymph node replaced with metastatic melanoma was submitted to the Cell Therapy Core Facility for vaccine production purposes. The patient received a course of vaccine injections and had no progression of disease for 2 years. (Her sole site of disease at the time of first vaccination was a metastatic nodule in the right breast.) ELISPOT data demonstrate an increase in the number of autologous tumor-reactive T cells over the first 6 months, from less than 20 spots at baseline, to 32 spots on day 118, and 125 spots on day 181 (Fig. 5). At the time of her first annual booster, the number of spots had returned to baseline. After her second annual booster, there was a weak but tumor-specific response (Fig. 5: day 705, more spots in wells with tumor cell lysate than wells without). The patient did not have measurable disease by RECIST criteria when she started vaccine treatment, but she did have a subcentimeter lesion in the right breast which was positive for metastatic melanoma by fine needle aspiration. This lesion was followed closely by CT scan and by mammography, and it did not increase in size for 24 months. Progression was seen on a scan in June 2005, when the lesion increased in size from 0.7 to 1.5 cm. The patient had no other evidence of disease. The breast lesion was excised and submitted to the Cell Therapy Core Facility for production of a new vaccine. Histologic examination of a sample of tumor revealed a significant lymphocytic infiltrate which was not present in the contralateral malignant breast lesion resected prior to vaccine therapy in 2002 or in the inguinal lymph node resected for vaccine production in 2003 (Fig. 6). The patient continues to remain without any evidence of disease, 1 year after receiving the new vaccine.
|
|
) showed significantly higher frequencies of MART-1-positive CD8 T cells (0.45%) than in nonmelanoma donors (always less than 0.1%) on day 678, prior to the booster injection, and an expansion of melanoma-reactive T cells (0.68% and 0.92% on days 728 and 805, respectively) after the booster injection. Restimulation of PBMCs with tumor cell lysate resulted in an increase in MART-1-specific CD8 T cells (0.73% and 1.66% for days 678 and 728, respectively). These data, along with ELI-SPOT data and histologic examination of metastatic lesions before and after vaccination, support the conclusion that a tumor-specific immune response was successfully generated in patient # 5 as a result of GM.CD40L bystander-based vaccine therapy.
|
). One patient (#3) died of disease progression after his second vaccine; three patients completed the planned three initial doses of vaccine therapy, and one patient (#1) is still alive 3 years since initial vaccine treatment. Three patients (cohort 2) received 5 million bystander cells and 10 million tumor cells when they received their first vaccine injection; all subsequent doses for these patients were at cohort 3 levels. One of these patients had progressive disease at the time of initial reassessment (after three vaccines), but the other two patients are still alive and well, more than 2 years after their first vaccine. Three patients completed the planned three initial vaccine doses at the cohort 3 level; a fourth patient (#8) died of disease progression after her first vaccine. Two of the three treated patients are still alive, and neither one has had any other treatment. Fourteen patients received at least one vaccine injection at the cohort 4 level. One patient (#19) moved away after his second vaccine and was therefore taken o3 the study protocol. One patient (#13) developed a symptomatic brain metastasis after the second vaccine and was also taken o3 the study. One patient (#24) died before receiving her third vaccine. Three patients (# 12, 16, and 20) completed the three vaccine injections, but they were unable to return for their second DTH assay due to progressive disease; two of these patients (# 24 and 16) died shortly after their third vaccine. Because these three patients were not evaluable from an immunological point of view (no follow-up DTH or ELISPOT assay), three other patients were enrolled, for a total of nine evaluable patients in cohort 4.
There was no tumor regression after vaccine therapy, but many patients (most of whom had steady progression of disease prior to vaccine therapy) had stable disease, including six of ten patients with metastatic melanoma; three patients had stable disease for over 18 months. Median survival for all patients who received at least three vaccines (n = 21), including many heavily pretreated patients, was 24.4 months. One patient (#5), who had stable disease for 24 months, developed tumor-specific T-cell responses on ELISPOT testing over the course of the study. She also had an increase in MART-1-specific T cells by tetramer analysis. Biopsy of the tumor that progressed 2 years after the onset of vaccine therapy revealed a massive peritumoral and intratumoral inflammatory infiltrate, including both CD4 and CD8 T lymphocytes—in contrast to two prevaccine tumors that had few T cells.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONPATIENTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
These findings support the hypothesis that, when mixed with irradiated but otherwise unmodified autologous tumor cells (antigen source), these GM.CD40L bystander cells (1) help recruit DCs by secreting GM-CSF in the vaccine site microenvironment, and (2) activate those recruited DCs by expression of CD40L. We hypothesize that these activated DCs can then endocytose apoptotic bodies from the irradiated tumor cells, migrate to the regional lymph nodes, and present tumor antigens in the context of MHC class I and II, ultimately leading to tumor-specific T cell activation and systemic immunity. This phase I study showed no systemic toxicity and only mild local and self-limited irritation at the vaccine injection site. Correlative studies were used to provide "proof of principle" evidence that this approach is worthy of further investigation. Serum samples confirmed the presence of GM-CSF in the blood of some patients after vaccine injection. Immunohistochemistry of vaccine injection site biopsies showed recruitment of CD1a-positive APCs in the dermis of all patients tested, and activated (CD86-positive) APCs were present in 15 of 19 (79%) patients. Resident DCs are normally CD86-negative, so the presence of CD86-positive DCs in the dermis of vaccinated patients is further evidence of a bystander-induced effect. ELISPOT assays revealed that many patients were unable to generate T-cell responses, either to tetanus toxoid or to tumor antigen, suggesting a diffuse immune suppression, possibly as a result of advanced disease (many patients had significant measurable disease and 35% of patients had an ECOG performance status of 1 or 2). Tumor-specific T cells were generated in five (55%) of the remaining nine patients, and these patients had unusually long median times to tumor progression. Tetramer analysis of one relatively long-term survivor HLA-A2-positive patient with melanoma showed a high frequency of MART-1-reactive T cells before her second annual booster vaccine, and a further expansion of MART-1-reactive T cells after the booster.
These results support the conclusion that the GM.CD40L bystander-based vaccine can activate tumor-specific T-cell responses and potentially lead to prolonged patient survival. Follow-up for this group of patients is still short, but it appears that even patients who progressed after vaccine treatment had relatively indolent courses. One obvious possibility is the intrinsic bias of patient selection for such clinical trials, but another possibility is that the vaccine may improve the ability of patients to response to subsequent therapy.25–27 Phase-II studies are currently underway to study the impact of this bystander-based vaccine on antitumor T-cell responses, tumor progression, and overall survival of patients with melanoma and mantle cell lymphoma. This bystander-based therapeutic approach, like vaccine therapy in general, will likely benefit that group of patients with a high risk of recurrence but a good performance status and minimal tumor burden (e.g., resected stage IV disease). Studies designed to test this hypothesis will require larger sample sizes and longer follow-up.
|
ACKNOWLEDGMENTS |
|---|
Received for publication July 10, 2006. Accepted for publication July 13, 2006.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONPATIENTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
 |
|
 |
|
  K. L. Hippen, P. Harker-Murray, S. B. Porter, S. C. Merkel, A. Londer, D. K. Taylor, M. Bina, A. Panoskaltsis-Mortari, P. Rubinstein, N. Van Rooijen, et al. Umbilical cord blood regulatory T-cell expansion and functional effects of tumor necrosis factor receptor family members OX40 and 4-1BB expressed on artificial antigen-presenting cells Blood, October 1, 2008; 112(7): 2847 - 2857. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Kobayashi, C. Pulitano, G. Liddo, and N. Ohkohchi The Growing Evidence of the Role of T-Cells in the Treatment of Solid Tumors Ann. Surg. Oncol., April 1, 2008; 15(4): 1254 - 1254. [Full Text] [PDF]
|
|
|
|
 |
|
 A. A. Ryan, T. M. Wozniak, E. Shklovskaya, M. A. O'Donnell, B. Fazekas de St. Groth, W. J. Britton, and J. A. Triccas Improved Protection against Disseminated Tuberculosis by Mycobacterium bovis Bacillus Calmette-Guerin Secreting Murine GM-CSF Is Associated with Expansion and Activation of APCs J. Immunol., December 15, 2007; 179(12): 8418 - 8424. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. S. Neelapu and L. W. Kwak Vaccine Therapy for B-Cell Lymphomas: Next-Generation Strategies Hematology, January 1, 2007; 2007(1): 243 - 249. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
