Fax: (011) 886-2-2792-9654
A pilot clinical trial of vaccination with dendritic cells pulsed with autologous tumor cells derived from malignant pleural effusion in patients with late-stage lung carcinoma
Article first published online: 6 JAN 2005
Copyright © 2005 American Cancer Society
Volume 103, Issue 4, pages 763–771, 15 February 2005
How to Cite
Chang, G.-C., Lan, H.-C., Juang, S.-H., Wu, Y.-C., Lee, H.-C., Hung, Y.-M., Yang, H.-Y., Whang-Peng, J. and Liu, K.-J. (2005), A pilot clinical trial of vaccination with dendritic cells pulsed with autologous tumor cells derived from malignant pleural effusion in patients with late-stage lung carcinoma. Cancer, 103: 763–771. doi: 10.1002/cncr.20843
- Issue published online: 3 FEB 2005
- Article first published online: 6 JAN 2005
- Manuscript Accepted: 27 OCT 2004
- Manuscript Revised: 13 OCT 2004
- Manuscript Received: 30 APR 2004
- Taichung Veterans General Hospital
- National Health Research Institutes
- dendritic cell;
- nonsmall cell lung carcinoma;
- pleural effusion
The authors conducted a pilot clinical trial to explore the vaccination of patients with late-stage lung carcinoma with dendritic cells (DCs) pulsed with necrotic tumor cells derived from malignant pleural effusion specimens, and to evaluate the antitumor immune response induced by this therapy.
Autologous DCs were generated by culturing adherent mononuclear cells with interleukin-4 and granulocyte-macrophage–colony-stimulating factor for 7 days. Day-7 DCs were cocultured overnight with autologous necrotic tumor cells derived from pleural effusion specimens to allow internalization of tumor antigens. DCs were then treated with tumor necrosis factor-alpha for 16 hours. The antigen-loaded DCs were injected into each patient's inguinal lymph nodes under sonographic guidance. Eight patients with late-stage nonsmall cell lung carcinoma were treated in this manner. Patients were vaccinated once weekly for 4 weeks and then boosted twice biweekly.
The authors found that there was no Grade II/III toxicity and autoimmune response in all patients after intranodal injection of the DC vaccine. Minor to moderate increases in T-cell responses against tumor antigens were observed after DC vaccination in six of eight patients. Five patients had progressive disease. One patient had minor tumor response and two patients had stable disease. The two patients who had longer disease control also had better T-cell responses.
The results indicated that it was feasible to immunize patients with lung carcinoma intranodally with DCs pulsed with necrotic tumor cells enriched from pleural effusion specimens, and this approach may generate T-cell responses and provide clinical benefit in some patients. Cancer 2005. © 2005 American Cancer Society.
The prognosis of patients with late-stage (AJCC Stage IIIB and Stage IV) nonsmall cell lung carcinoma (NSCLC) with malignant pleural effusion who do not respond to standard chemotherapy is very poor: the median survival time for these patients is 16 weeks.1 Novel therapeutic approaches are needed to prolong survival and improve quality of life. Among various approaches, immunotherapy has become one promising strategy to provide clinical benefits for these patients. It is postulated that tumor-associated antigens (TAAs) of tumor cells are not presented optimally to cytotoxic T lymphocytes (CTLs) by professional antigen-presenting cells (APCs) during the development of malignancy, which leads to inadequate T-cell responses and outgrowth of tumor.2 One possible approach to overturn such an immune system defect is to immunize patients with cancer with APCs pulsed with appropriate TAAs in vitro under optimal conditions. It is likely that such treatment could trigger a T-cell response against tumor cells in vivo and clinically inhibit tumor growth.3, 4
Dendritic cells (DCs) are the most potent APCs in the body, and they are responsible for the initiation of both innate and adoptive immune responses.5–7 Vaccination of DCs pulsed with TAAs induces protective immune responses in several animal models.8, 9 Several studies have shown that specific CTLs can be generated in vitro using human DCs pulsed with TAAs and T cells from peripheral blood of healthy donors or patients with cancer.10–12 Some studies employed preidentified TAA or synthetic peptides containing CTL epitopes of known TAA as the source of antigens.13–17 Alternative strategies have been explored in animal models and human clinical trials to expand DC-based therapy to cancers for which appropriate TAA or relevant CTL epitopes have not been determined. These strategies include pulsing DCs with peptides stripped from the surface of autologous tumor cells,18 tumor cell lysates,9, 19 tumor-derived RNA,20, 21 or necrotic and apoptotic tumor cells,22, 23 as well as fusing DCs with tumor cells.24–26 DCs loaded with tumor lysates were used in clinical trials for patients with parathyroid carcinoma, late-stage melanoma, and metastatic renal cell carcinoma.14, 22, 27, 28 No major complications after DC immunization have been reported to date.
One of the major limitations in the development of DC-based vaccines is that proper tumor rejection antigens are not available for most human cancers. In the current study, we explored a strategy of using autologous cancer cells isolated from malignant pleural effusions as the source of TAAs. The antigen-loaded DCs were then injected directly into inguinal lymph nodes. We demonstrated that immunization of patients with DC vaccines prepared under this protocol is feasible and safe.
MATERIALS AND METHODS
The study protocol was approved by the institutional review board of Taichung Veterans General Hospital and by the Department of Health, Taiwan. Informed consent was obtained from each patient. Patients ≥ 18 years old with a cytologic or histologic diagnosis of NSCLC refractory to conventional chemotherapy, as well as patients who had no effective treatment options, were eligible for the study. Each patient in the study had an Eastern Cooperative Oncology Group performance status score ≤ 2. Patients who received chemotherapy or radiotherapy ≤ 4 weeks before the study were not eligible. All patients had adequate bone marrow, liver, and renal functions. Patients who had metastatic disease in the central nervous system, autoimmune disease, or active acute or chronic infection and patients who received steroid or biologic treatment within 4 weeks before study enrollment were excluded. Lesions noted at baseline were measured or evaluated by chest computed tomographic scan every 12 weeks or by chest X-ray every 4 weeks as a measure of response according to standard criteria.29 Immunologic surveys, including monitoring the levels of antinuclear antibody (Ab), rheumatoid factors, and anti-thyroid Ab, and immunoelectrophoresis were repeated 4 weeks after the 6 DC vaccine injections were received.
Tumor Cell Collection and Preparation of Tumor Lysate
Patients eligible for the protocol were selected and late taps of the pleural effusions were collected by thoracentesis under chest sonography. Pleural effusion specimens were used for cytologic examination and for collection of tumor cells. Tumor cells in the pleural effusion specimens were enriched by density gradient centrifugation with 75% Ficoll-Paque (Amersham Pharmacia Biotech, Uppsala, Sweden). Cells of low density (e.g., those at the Hank balanced salt solution [HBSS] —75% Ficoll interface) were collected as the source of tumor cells. These cells were usually large and clustered, with little lymphocyte contamination. The characteristics of enriched cells were monitored by staining with a fluorescein isothiocyanate (FITC)-labeled monoclonal antibody (MoAb) against human epithelial antigens (clone Ber-EP4; Dakocytomation, Glostrup, Denmark) or with an FITC-labeled MoAb against cytokeratins (clone MNF116; Dakocytomation) after permeabilization with a fluorescence-activated cell sorter permeabilizing solution (BD PharMingen, San Diego, CA).
Enriched tumor cells were suspended (1 × 106 cells/mL) in X-VIVO15 medium (BioWhittaker, Walkersville, MD) and lysed by 5 cycles of freezing and thawing. Aliquots of the necrotic cell mixture (crude cell lysate) were stored at −80 °C until use. One aliquot of the crude cell lysate was examined under the microscope and placed into culture to ensure the absence of live tumor cells. To prepare the tumor lysate for T-cell proliferation assay, the crude cell lysate was further centrifuged at 12,000 rpm to remove unbroken cells or cell debris, and the protein concentration of the supernatant (clear cell lysate) was determined.
Generation of Dendritic Cells
Peripheral blood mononuclear cells (PBMCs) from a 200-mL blood specimen were enriched by density gradient centrifugation with Ficoll-Paque. The PBMCs were incubated for 2 hours at 37 °C in AIM-V medium (Invitrogen, Carlsbad, CA), and the adherent cells were cultured in X-VIVO15 medium containing 2% heat-inactivated autologous plasma, 1000 U/mL human interleukin (IL)-4 (GMP-grade, Strathmann Biotec AG, Hannover, Germany), and 500 U/mL granulocyte-macrophage–colony-stimulating factor (Leukomax; Novartis International AG, Basel, Switzerland). On Day 7, loosely attached or floating cells were collected as immature DCs.
Preparation of the Vaccine and the Vaccination Protocol
Day 7 immature DCs (1.5 × 106 cells in 1.5 mL) were cultured with 1.5 mL crude tumor cell lysate (from 1.5 × 106 tumor cells) at 37 °C for 16 hours. On Day 8, antigen-pulsed DCs were collected and matured by culturing in X-VIVO15 medium containing 2% heat-inactivated autologous plasma and 1000 U/mL tumor necrosis factor-alpha (TNF-α; Strathmann). On Day 9, cells were collected as antigen-pulsed, matured DCs. Collected DCs were washed 5 times with HBSS (BioWhittaker), and 1 × 106 DCs were suspended in 0.3 mL phosphate-buffered saline supplemented with 1% heat-inactivated autologous plasma. To remove any cell cluster, the cell suspension was passed slowly through a 25-gauge needle. The cell suspension was then injected into one inguinal lymph node of the appropriate patient with cancer under the guidance of real-time sonography. Patients were treated once weekly for 4 weeks and then boosted twice biweekly. Thirty-milliliter blood specimens were collected from patients 2 weeks after the second, the fifth, and the last vaccination. PBMCs were purified and cryopreserved for evaluation of immune responses against the clear tumor cell lysate. No bacteria, fungus, mycoplasma, and endotoxin contamination was detected in the tumor lysate and DC vaccine.
Flow Cytometry Analysis of Dendritic Cells
Cells to be analyzed for expression of surface markers were stained with different fluorescence-labeled MoAbs and then analyzed using a flow cytometer (EPICS XL-MCL, Beckman Coulter, Fullerton, CA). The MoAbs used in the current study included the following: FITC-anti-HLA-DR (Immunotech, Marseille Cedex, France), phycoerythrin (PE)-anti-CD86 (Immunotech), FITC-anti-CD80 (Immunotech), FITC-anti-CD40 (Serotec, Oxford, UK), PE-anti-CD14 (Immunotech), anti-CCR7 (BD PharMingen), and PE-anti-CD83 (BD PharMingen). Isotype-matched control MoAbs were obtained from BD PharMingen and Caltag (Burlingame, CA). For the analysis of DCs cocultured with the tumor lysate, Day 7 immature DCs were generated from PBMCs from a normal donor and pulsed for 16 hours with control PBMC lysate obtained from a different normal donor or with the tumor lysate from a patient with lung carcinoma. DCs were cultured further with 1000 U/mL TNF-α for 16 hours and collected for flow cytometry analysis. Subsequently, some DCs initially cocultured with the tumor lysate were cultured with a CD40 ligand (CD40L)-transfected cell line or with a vector-transfected control cell line at a 1:1 ratio for 48 hours as described previously.30 Then, these DCs were collected for flow cytometry and mixed leukocyte response analysis.
T-Cell Proliferation Assay and Interferon-Gamma Production Analysis
PBMCs collected before and after vaccination were thawed at the same time for immunologic analysis. Three replicates of 1 × 105 PBMCs were cultured in 96-well culture plates for 6 days in RPMI-1640 supplemented with 5% human type-AB serum (BioWhittaker) in the presence of either 10 or 50 μg/mL autologous clear tumor lysate or a control PBMC lysate. Cellular proliferation was determined using a bromodeoxyuridine (BrdU) incorporation enzyme-linked immunosorbent assay (ELISA) kit (Roche Diagnostics GmbH, Mannheim, Germany). Data were expressed as stimulation index, which was obtained by dividing the mean optical density reading with the tumor lysate by the mean optical density reading with the PBMC lysate. Production of interferon-gamma (IFN-γ) in the culture supernatant by stimulated PBMCs was determined using a human cytometric bead array kit (BD PharMingen). Data from cultures with the control PBMC lysate were subtracted from those with the tumor lysate.
Mixed Leukocyte Response
DCs (2 × 104 cells per well) after various treatments were 2-fold serially diluted, irradiated (30 gray), and cultured with allogeneic, nonadherent PBMCs (2 × 105 cells per well) in 96-well plates for 6 days as described previously.30 Cellular proliferation was determined using a BrdU incorporation ELISA kit (Roche Diagnostics) as suggested by the manufacturer.
Toxicity and Clinical Evaluation
The grading of toxicity was determined according to the National Cancer Institute Common Toxicity Criteria (Version 2.0). The clinical responses were defined as follows: 1) a minor response, i.e., a 25–50% decrease in lesions lasting ≥ 1 month or a > 50% decrease in lesions lasting < 1 month; 2) stable disease, i.e., a < 25% change in size with no new lesions for 6 weeks; and 3) disease progression, i.e., the appearance of new lesions or a > 25% increase in the area of existing lesions.
Quality Control of Denditic Cells and Tumor Lysates
Eight patients were treated in the current study. Six patients received six DC injections. Two patients received only one or two DC injections before withdrawal from the study due to rapid disease progression with intractable pleural effusion. The characteristics of all patients are listed in Table 1. For every vaccine preparation, the tumor lysate, the DC culture supernatant on Day 7, and the supernatant fluid after the last washing on Day 9 were examined for contamination with endotoxin, mycoplasma, fungus, or bacteria. All samples analyzed were negative for microorganism contamination and the endotoxin level was always < 0.15 Endotoxin units [EU]/mL (data not shown). The average yield of DCs obtained on Day 7 was approximately 8.7% of input PBMCs. The recovery rate of DCs on Day 9 after pulsing with the tumor lysate overnight followed by incubation with TNF-α for 16 hours was approximately 80%. Day 9 DCs displayed an immature DC staining profile: HLA-DR+, CD86+, CD80low, CD40+, CD14dim, CCR7−, and CD83− (Table 2). It is likely that the presence of the tumor lysate may interfere with the up-regulation of certain surface markers by DCs treated with TNF-α. To examine this possibility, we have shown (Fig. 1A) that the proportion of cells expressing CD83 is higher in DCs cocultured with TNF-α and the control PBMC lysate (42.5 ± 3.3%) than in those cocultured with TNF-α and the tumor lysate from a patient with lung carcinoma (23.8 ±1.5%). This result suggests that a coculture with the tumor lysate may inhibit the maturation of DCs. However, these DCs could be induced to further maturation, because 52.5 ± 16.0% of DCs initially cultured with the tumor lysate expressed CD83 after subsequent coculturing with cells expressing surface CD40L.30 These reactivated DCs can efficiently stimulate the proliferation of allogeneic PBMCs (Fig. 1B). These results suggest that DCs cultured with the tumor lysate can still be activated by CD40L-expressing T cells.
|Patient no.||Gender||Age (yrs)||PS||Cell type||No. of previous treatments||Name of regimen||Clinical status/metastasis on enrollment|
|1||M||39||2||A||2||G + C, P||Lung, pleural effusion, retroperitoneum|
|2||M||77||2||A||1||G + C||Lung, pleural effusion, bone|
|3||M||67||1||S||3||G + C, P, TT||Lung, pleural effusion, retroperitoneum, bone|
|4||M||45||1||A||6||G + C, TT, V, U, CPT-11, I||Lung, pleural effusion|
|5||F||71||2||A||1||G + C||Lung, medastinal lymphadenopathy, bone, pleural effusion|
|6||F||49||2||A||5||G + C, TT, V, CPT-11.1||Lung, pleural effusion|
|7||F||75||1||A||3||T + C, TT, 1||Lung, pleural effusion|
|8||F||68||2||A||4||G + C, TT, V, CPT-11||Lung, liver, pleural effusion|
|Marker||Mean ± SD|
|HLA-DR||98.7 ± 1.4|
|CD86||96.4 ± 3.9|
|CD80||4.4 ± 3.9|
|CD83||5.2 ± 2.6|
|CD40||99.0 ± 0.7|
|CD14||20.7 ± 19.2|
|CCR7||4.7 ± 1.6|
The presence of tumor cells in the pleural effusion specimens was verified initially by cytologic examination (data not shown). The characteristics of enriched tumor cells were monitored by staining with a MoAb against human epithelial antigens (clone Ber-EP4) or with a MoAb against cytokeratins (clone MNF116). Both Ber-EP4 and MNF116 are used commonly to detect the presence of malignant epithelial cells in body fluids31 and tissue sections.32 Cell specimens from Patient 8 were not available for flow cytometry analysis due to a technique problem. Our previous study (data not shown) indicated that the majority of tumor cells enriched from pleural effusion specimens obtained from most patients with NSCLC stained positively with an Ab against human carcinoembryonic antigen and with a mixture of two antihuman cytokeratin MoAbs (clones AE1 and AE3). This AE1/AE3 MoAb mixture reacts with the majority of the 19 human cytokeratins except cytokeratins 9, 12, 17, and 18.33–36 We have found that different proportions of cells (7–49%) from individual patients stained positively with the MNF116 MoAb (Fig. 2 and Table 3). The MNF116 MoAb reacts with cytokeratins 5, 6, 8, 17, and 19.37 This result indicates that tumor cells enriched from pleural effusion specimens obtained from patients with lung carcinoma have different cytokeratin expression profiles as reported previously.38 The Ber-EP4 MoAb shows a broad pattern of reactivity with human cells of epithelial origin.39 A significant number of tumor cells from Patients 3 and 4 stained positively with Ber-EP4 (33% and 58%, respectively). The expression of this epithelial cell membrane antigen was low in cells from the remaining five patients examined (Fig. 2 and Table 3). The significance of this observation remains to be elucidated.
|Patient no.||Cytokeratins (%)a||EMA (%)||No. of DC injections||T-cell responseb||Clinical response||Survival period (wks)c|
|3||49.2||33.4||6||++++||SD (48 wks)||64|
|4||27.0||58.2||6||+++||SD (36 wks)||53|
|6||7.5||0.0||6||+||MR (24 wks)||31|
Enriched tumor cells were induced to necrosis by repeated freezing and thawing treatment. This crude lysate of tumor cells was free from contamination of microorganisms and was used to pulse DCs. Some tumor cells remained intact in shape but stained positively by trypan blue dye after the freeze/thaw treatment. These were all necrotic cells that did not grow in subsequent cultures. It has been suggested that exogenous particulate antigens are preferentially targeted to the major histocompatibility complex class I antigen presentation pathway by DCs and macrophages.40, 41 Therefore, we purposely used a crude cell lysate containing necrotic whole tumor cells and large cell debris as the source of tumor antigen.
Evaluation of Adverse Effects and Clinical Outcomes
No Grade II–IV toxicity was observed in any patient after intranodal injection of the DC vaccine. There was no evidence of treatment-related autoimmune response as determined by the examination of antinuclear Ab, rheumatoid factor, and antithyroid Ab (data not shown). These results confirm that it is feasible and safe to inject antigen-pulsed DCs into the inguinal lymph nodes of patients with late-stage NSCLC with our protocol. Six patients received all six DC injections. Of these 6 patients, 3 patients had progressive disease, 1 patient (Patient 6) had a minor response, and 2 patients (Patients 3 and 4) had stable disease (Table 3). The most substantial reduction in the volume of tumor lesions in the lung was observed in Patient 6 (Fig. 3). Unfortunately, this patient had disease progression when reevaluated 24 weeks after treatment.
Evaluation of T-Cell Proliferation against the Tumor Lysate
We found that, in the six patients evaluated, there was a minor to moderate increase in the ability of T cells to respond to the tumor lysate after DC vaccination (Fig. 4 and Table 3). Additional blood samples were obtained from Patients 3 and 4 due to stabilization of their disease. A substantial increase in tumor-specific T-cell proliferation was observed for Patient 3. The T-cell response of this patient was detectable 120 days after the first vaccination (Fig. 4), which was in agreement with stable clinical status. A clear T-cell response against the tumor lysate (10 μg/mL) was observed in Patient 4 after 90 days (Fig. 4A). The response declined on Day 120, although this patient experienced stable disease up to that time point. An increased T-cell proliferation against the tumor lysate (10 μg/mL) was also observed in Patients 7 and 8 after 60 days (Fig. 4A). The fluctuations in T-cell response may have resulted from periodic injections of the DC vaccine. We detected a substantial increase in the production of IFN-γ by PBMCs stimulated with the tumor lysate at several time points after DC vaccination in Patients 3 and 4 (Fig. 5). These two patients also experienced a longer period of stable disease and survival after the initiation of DC vaccination (Table 3).
Both apoptotic and necrotic tumor cells have been evaluated as the source of antigens.14, 27, 42, 43 Vaccination of patients with cancer using DCs pulsed with lysates of necrotic cells has been reported in several clinical studies.14, 27 However, obtaining enough tumor tissue specimens or cells for the preparation of the DC vaccine is still difficult for many cancers. In the current study, we evaluated the feasibility of using tumor cells isolated from malignant pleural effusion specimens of patients with lung carcinoma to pulse DCs. Removing pleural effusion is required for many patients with late-stage lung carcinoma to relieve clinical symptoms. A substantial amount of tumor cells can be enriched from pleural effusion specimens obtained from patients with cancer. Therefore, malignant pleural effusion specimens may be a good source of tumor cells. Our results indicate that tumor cells enriched from different patients were heterogeneous. It is not clear whether such differences influence the generation of immune responses and clinical outcomes after vaccination. A better and thorough characterization of tumor cells may help in the selection of appropriate patients for DC-based treatment in the future. We also cannot rule out the possibility that there are contaminated normal cells in our tumor cell preparation, as in the case of using tumor cells from surgical specimens.
In this pilot trial, we also observed that incubation with TNF-α did not induce fully maturation of tumor lysate-pulsed DCs. However, these DCs expressed high levels of CD40. It is possible that they can be activated through CD40 after interaction with CD40L expressed on activated CD4+ T cells in the lymph nodes of patients. These DCs also expressed low levels of CCR7. The expression of CCR7 is important for DCs to migrate to the lymph nodes. Thus, a lack of CCR7 expression could hinder the ability of antigen-pulsed DCs to migrate to the draining lymph nodes if injected subcutaneously or intradermally. Our strategy of direct intranodal injection of the DC vaccine might circumvent this shortcoming.
We observed that T-cell responses in most patients with cancer after DC vaccination appeared to be moderate at best. Patients 3 and 4, who had a longer period of disease control (48 and 36 weeks, respectively), also had better T-cell responses. Although it is hard to draw conclusions from data for only a few patients, it is likely that a longer period of disease control might be associated with a more lasting T-cell response. Further studies with more patients will be needed to better resolve this correlation. It is reasonable to expect that T cells will respond to specific antigens in a dose-dependent manner. However, because we used unfractionated tumor lysates as the source of tumor antigens in our T-cell proliferation assay, other cellular materials in the lysate may interfere with the T-cell response at high concentrations. Attempts to maintain a prolonged T-cell response in DC-vaccinated patients have been explored by some researchers. Current efforts have focused on the supplement of cytokines such as IL-2 or IL-12 after vaccination to increase the lifespan of T cells.44–47 We are currently investigating the potential of such combined immunotherapy.
- 21Induction of carcinoembryonic antigen (CEA)-specific cytotoxic T-lymphocyte responses in vitro using autologous dendritic cells loaded with CEA peptide or CEA RNA in patients with metastatic malignancies expressing CEA. Int J Cancer. 1999; 82: 121–124., , , , , .