Antivector and Tumor Immune Responses Following Adenovirus-Directed Enzyme Prodrug Therapy for the Treatment of Prostate Cancer

Abstract

We have completed a phase I/II suicide gene therapy clinical trial in patients with prostate cancer, using an E1/E3-deleted replication-deficient adenovirus (CTL102) encoding the bacterial nitroreductase enzyme in combination with prodrug CB1954. This study has provided an opportunity to monitor and characterize vector- and tumor-specific adaptive immunity before and after single or repeat injections of adenovirus. Here we report robust vector-specific humoral and cellular immune responses in all patients monitored. However, we found no correlation between preexisting immunity or the magnitude of the immune response to vector and the clinical outcome as measured by changes in serum prostate-specific antigen (PSA) level. Increased frequency of T cells recognizing prostate-specific antigens PSA or prostate-specific membrane antigen (PSMA) was detected in 3 of 11 patients after therapy, suggesting that this direct cytotoxic strategy can also stimulate tumor-specific immunity.

Introduction

Recombinant adenoviruses (rAds) are one of the most promising class of vectors for human gene therapy (Young et al., 2006). They have been found to be safe in numerous clinical trials and a replication-deficient adenovirus vector expressing p53 is the first licensed cancer gene therapy (Peng, 2005). One potential limitation to the use of Ad vectors is the presence of preexisting cellular and humoral immunity in humans (Bessis et al., 2004). Replication-deficient adenoviruses have been shown to induce strong innate and adaptive immune responses in murine and primate animal models, where rapid inflammatory responses appear to define dose-limiting toxicity (Lieber et al., 1997; Schnell et al., 2001), T cell-mediated killing of infected cells can reduce transgene expression (Dai et al., 1995), and neutralizing antibodies may limit readministration (Chen et al., 2000). However, animal models of immunity induced by human adenoviruses are a poor surrogate for the study of adenovirus immunity in humans, because of the incomplete replication in animal cells (which can bias immunity toward early proteins), and because of the complexity of adenovirus pathogenesis in humans. Humans are typically exposed to acute adenovirus infection at an early age and will go on to be exposed to multiple serotypes throughout life (Horwitz, 2001). Most Ad infections are rapidly resolved, following the induction of serotype-specific neutralizing antibodies (nAbs) and cross-reactive nonneutralizing antibodies; however, there is evidence that the virus can persist for prolonged periods of time (Garnett et al., 2002). There is a memory T cell repertoire composed of CD4⁺ and CD8⁺ T cells which recognize highly conserved epitopes in the viral capsid (Smith et al., 1996, 1998; Heemskerk et al., 2003; Leen et al., 2004a,b; Onion et al., 2007). The magnitude of this memory T cell repertoire dwindles with age (Sester et al., 2002) and its composition is likely to be influenced by the presence of persistent viral antigen as well as the effect of virus modulators of the immune system. Consequently, it is important that the influence of preexisting immunity to Ad on the efficacy of rAd-based gene therapy vectors be assessed in the clinical setting

CTL102/CB1954 virus-directed enzyme prodrug therapy (VDEPT) functions through the nitroreductase (NTR)-mediated conversion of the relatively nontoxic prodrug CB1954 into a highly toxic alkylating agent that cross-links DNA (Friedlos et al., 1992) in cells transduced with the replication-deficient Ad vector CTL102. The activated CB1954 derivatives are able to cross cell membranes and kill surrounding tumor cells via a local bystander effect (Bridgewater et al., 1997). In addition to this localized tumor cell killing, it has been proposed that an “ideal” VDEPT approach for cancer therapy would include the stimulation of an immune bystander effect, which would have the ability to clear remaining tumor cells at the local site of therapy and potentially also tumor metastases (Vile et al., 2000). It has been demonstrated in animal models that cell killing via suicide gene therapy can stimulate the immune system and thus may have potential in generating antitumor immunity (Vile et al., 1997). In addition to the potentially proinflammatory effects of activated CB1954 tumor cell killing, the immunogenic Ad vector may act as an adjuvant in the generation of an antitumor immune response.

Despite numerous clinical trials utilizing adenovirus, few have investigated the immune response generated by the vector, and still fewer have correlated this with response to therapy. We have completed a phase I/II clinical trial of VDEPT in which patients received intraprostatic injection of CTL102 before radical prostatectomy (operable arm, single injection) or for the treatment of local recurrent prostate cancer, in combination with prodrug CB1954 (inoperable arm, single and repeat injection) (Patel et al., 2009). Immune responses to therapy were studied in a cohort of these patients in order to address the following questions: Does humoral immunity pose a barrier to transgene expression or to clinical outcome? What is the magnitude and nature of the T cell response to vector and does it affect outcome? Is VDEPT able to stimulate an antitumor T cell response?

We demonstrate that neutralizing antibodies did not limit gene expression (operable arm) or outcome (inoperable arm), as measured by serum prostate-specific antigen (PSA) levels. Vector-specific T cell frequency was monitored in real time, revealing the huge proliferative capacity of human Ad-specific CD4⁺ and CD8⁺ T cells in vivo and the immunodominance of hexon as a target antigen. High-frequency antivector T cell responses were detected in patients who responded to therapy as well as in those who did not. In a small cohort we demonstrate the transient presence of T cells specific for prostate tumor antigens (PSA and prostate-specific membrane antigen [PSMA]) following VDEPT.

Materials and Methods

Antibody tests

Plasma samples were taken at pretreatment; 7, 14, 21, and 28 days; and 2 and 3 months after treatment, heat-inactivated at 56°C for 30 min, and analyzed by ELISA to quantify total immunoglobulin response against adenovirus as previously described (Stallwood et al., 2000; Palmer et al., 2004). In brief, ELISA plates were coated with heat-inactivated wild-type Ad5 virus at a concentration of 20 ng of protein per well and bound antibodies were detected with a horseradish peroxidase-conjugated rabbit anti-human antibody and o-phenylenediamine substrate (Sigma, St Louis, MO). The neutralizing activity of patient plasma against Ad5 was tested with an E1,E3-deleted replication-defective Ad5 virus encoding β-galactosidase under the control of the cytomegalovirus immediate-early (CMV IE) promoter in A549 cells (Stallwood et al., 2000). Aliquots (100 μl) of plasma dilutions were applied to 10⁴ A549 cells in a 96-well plate and 100 μl of virus (2 × 10⁵ particles/cell) was added. At 48 hr postinfection, cell lysates were analyzed for β-galactosidase activity, using a Tropix luminescence assay (Applied Biosystems, Foster City, CA) and a VICTOR X plate reader (PerkinElmer, Cambridge, UK). Results are expressed as the plasma titer giving 50% reduction in infectivity relative to a positive control of virus-infected cells.

Interferon-γ enzyme-linked immunospot assay for Ad-specific T cells

Interferon (IFN)-γ enzyme-linked immunospot (ELISpot) assays were performed as previously described (Onion et al., 2007). Briefly, peripheral blood mononuclear cell (PBMC) preparations were stimulated with heat-inactivated CsCl-banded CTL102 (1000 particles per cell) or mock infected for 1.5 hr (37°C, 5% CO₂). Cells were washed and plated in triplicate at 4 × 10⁵ and 1 × 10⁵ cells per well into polyvinylidene difluoride-backed 96-well plates (Millipore, Watford, UK) precoated with anti-IFN-γ mAb (Mabtech, Nacka Strand, Sweden) at 15 μg/ml. Alternatively, PBMCs were plated as described previously and peptide (5 μg/ml) or dimethyl sulfoxide (DMSO) was added directly to the wells. After overnight incubation (37°C, 5% CO₂) spots were developed and then counted with an automated ELISpot reader (AID, Strassberg, Germany) and are represented as spot-forming cells (SFCs) per 10⁶ cells in the test wells, minus the background response to mock-infected or DMSO-treated cells. All peptides were synthesized according to 9-fluorenylmethoxycarbonyl chemistry (Alta Bioscience, University of Birmingham, Birmingham, UK) and dissolved in DMSO, and their concentrations were determined by biuret assay. Peptides used were as follows: TYFSLNNKF, KPYSGTAYNSL, MPNRPNYIAF, TFYLNHTFKK, and TDLGQNLLY (Leen et al., 2004b), DEPTLLYVLFEVFDV (Olive et al., 2002), and GTAYNALAPKGAPNP, VDCYINLGARWSLDY, and QWSYMHISGQDASEY (Onion et al., 2007). Peptides are referred to in text by their first three N-terminal amino acid letters.

IFN-γ secretion assay

The IFN-γ secretion assay was performed with a MACS IFN-γ secretion assay kit (Miltenyi Biotec, Bergisch Gladbach, Germany) according to the manufacturer's instructions with the following antigenic stimulations: PBMCs were plated on 24-well plates at 10⁷/ml with 5 × 10⁶/well in RPMI 1640, 5% human serum (HS), 2 mM glutamine, penicillin (100 U/ml), and streptomycin (0.1 mg/ml). Wells were loaded with 300 μg of A549 cells infected with wild-type Ad5 lysate (or mock infected). Cells were incubated overnight at 37°C, 5% CO₂. A control sample was loaded with staphylococcus enterotoxin (SEB, 1 μg/ml) (Sigma, Poole, UK) and incubated at 37°C, 5% CO₂ for 3 hr. Following the IFN-γ secretion assay cells were either stained for flow cytometry or positively selected by adding 20 μl of anti-phycoerythrin (PE) microbeads (15 min, on ice) and passed through MACS MS columns, using an OctoMACS (Miltenyi Biotec); positively selected fractions containing IFN-γ-secreting cells were collected. For flow cytometric analysis after the IFN-γ secretion assay the following antibodies were used: CD4–FITC (fluorescein isothiocyanate) and CD8–PC5 (R-phycoerythrin covalently linked to Cy5). For analysis following positive selection CD4–PC5, CD8–PC5, CD45RA–FITC, CD45RO–FITC, and CCR7–FITC antibodies were used (all from Beckman Coulter, High Wycombe, UK). For all flow cytometric analysis propidium iodide (1 μg/ml; Sigma) was used to exclude dead cells and cells were analyzed on a EPICS XL flow cytometer (Beckman Coulter).

Isolation of clonal Ad-specific T cells

T cells positively selected following the IFN-γ secretion assay were cloned by limiting dilution at 0.3 and 3 cells per well with allogeneic γ-irradiated, phytohemagglutinin-treated PBMCs (10⁵ per well) in RPMI 1640 supplemented with interleukin (IL)-2 at 200 U/ml (Chiron, Amsterdam, The Netherlands), 10% fetal bovine serum (FBS), and 1% HS. After 2–3 weeks, growing microcultures were screened for Ad5 reactivity by IFN-γ ELISA, using Ad5-infected or mock-infected autologous EBV-transformed B-lymphoblastoid cell lines (LCLs) as antigen-presenting cells (APCs). Selected clones were expanded in 2-ml wells, using the culture conditions described previously for cloning.

IFN-γ ELISA

Cloned T cells were incubated in V-bottomed microtest plate wells with autologous LCLs that were either prepulsed for 2 hr with purified Ad5 hexon or fiber at 5 μg/ml (Onion et al., 2007) or peptide at 5 mg/ml (or DMSO as a control), or infected with Ad2, Ad4, Ad5, or Ad11 (or mock infected) for 1.5 hr and then washed. The supernatant medium harvested after 18 hr was assayed for IFN-γ by ELISA (Pierce Biotechnology, Cramlington, UK) in accordance with the manufacturer's recommended protocol.

Construction of DNA expression vectors

Total RNA was extracted from LNCap cells (CRL-1740; American Type Culture Collection [ATCC], Manassas, VA), using an RNeasy kit (Qiagen, Crawley, UK), and cDNA was synthesized with a high-capacity reverse transcription cDNA synthesis kit (Applied Biosystems). Full-length PSA (Entrez accession no. M26663) and PSMA (Entrez accession no. M99487) were amplified by polymerase chain reaction (PCR) and inserted into the multiple cloning site of the pShuttle-CMV plasmid (Stratagene, La Jolla, CA), as was the cDNA encoding enhanced green fluorescent protein (eGFP; kindly provided by P. Searle, University of Birmingham). Protein expression following nucleofection of PBMCs was confirmed by Western blot and immunohistochemistry (antibodies: anti-PSA clone ER-PR8 from Abcam [Cambridge, UK] and anti-PSMA clone 3C9 from Northwest Biotherapeutics [Bethesda, MD]) and nucleofection efficiency was determined by measurement of GFP expression with the EPICS XL flow cytometer (Beckman Coulter).

Nucleofection

Cryopreserved PBMCs were thawed and then incubated overnight at 37°C, 5% CO₂ in AIMV medium (Invitrogen, Paisley, UK). Cells were nucleofected with 5 μg of plasmid, using a human dendritic cell Nucleofector kit and Nucleofector device I (Amaxa Biosystems, Cologne, Germany) according to the manufacturer's instructions. Cells were washed and viability was assessed by trypan blue exclusion before plating in triplicate for the IFN-γ ELISpot assay as described previously.

Results

Humoral responses to replication-deficient adenovirus

Patients were monitored for both neutralizing and total Ad-specific antibody responses, pre- and postintraprostatic injection with increasing doses of CTL102. All patients had preexisting Ad antibodies (total immunoglobulin) and Ad5 neutralizing activity as shown by a sensitive neutralization test using an Ad5 vector expressing β-galactosidase (Stallwood et al., 2000; Patel et al., 2009). In most patients there was a robust response to CTL102 injection, with an increased level of antibody detected in both assays. In the operable arm following prostatectomy, serial sections of the prostate were immunostained for the presence of encoded enzyme NTR, and the level of expression was quantified as the percentage of NTR-positive slides (Patel et al., 2009). There was no indication that high levels of preexisting neutralizing Ad antibodies correlated with low levels of gene expression either as a whole cohort or within given dose levels (Fig. 1A). In the inoperable arm there was evidence of treatment response as measured by serum PSA level; Fig. 1B shows that those responding to treatment (stable disease, partial response, and good response) did not have significantly higher preexisting neutralizing antibody than those who failed to respond to treatment. Individual patient details are shown in Table 1.

FIG. 1.

Pretreatment Ad neutralizing antibody levels. Patient plasma was serially diluted and incubated with A549 cells before infection with Ad5β-gal (2 × 10⁵ particles/cell). At 48 hr postinfection, cell lysates were analyzed for β-galactosidase activity. Results are expressed as the plasma titer giving 50% reduction in infectivity relative to a positive control of virus-infected cells. (A) Antibody titer in inoperable patient compared with NTR expression in resected prostate. Virus dose: solid squares, 10¹⁰; solid triangles, 5 × 10¹⁰; solid inverted triangles, 10¹¹; solid diamonds, 5 × 10¹¹. (B) Pretreatment antibody titer in operable patients in comparison with response to treatment (assessed 6 months posttreatment). Responders: those patients with stable disease, partial response, and good response. Nonresponders: those with progressive disease. Lines represent mean values.

Table 1.

Treatment Outcome and Adenovirus-Specific Immunity

Virus dose (particles)	Patient no.	Outcome	Treatment round	Preexisting neutralizing antibody ^a	Peak adenovirus-specific T cell response ^b
5 × 10¹⁰	16XX	GR	First	−5.205	47
			Second	−4.111	187
	19XX	SD	First	−1.542	277
			Second	−3.136	3
	20XX	PD	First	−2.997	207
			Second	−3.757	NA
1 × 10¹¹	21X	SD	Single	−4.663	30
	22X	PD	Single	−4.494	290
	24XX	PR	First	−3.235	47
			Second	−4.103	587
5 × 10¹¹	27XX	SD	First	−2.341	277
			Second	−3.089	207
	29XX	PD	First	−2.189	633
			Second	−4.753	672
	34XX	PR	First	−1.977	NA
			Second	−4.664	NA
	35XX	PR	First	−1.926	580
			Second	−4.283	NA
	36X	PD	Single	−2.653	NA
	37XX	PD	First	−1.021	NA
			Second	−3.121	NA
	38XX	GR	First	−1.059	NA
			Second	−2.379	NA
1 × 10¹²	30XX	PD	First	−2.972	244
			Second	−4.707	NA
	31XX	PD	First	−3.488	447
			Second	−4.316	53

Abbreviations: GR, good response; PR, partial response; SD, stable disease; PD, progressive disease; NA, not available.

log₁₀ of serum dilution giving 50% neutralization.

SFCs/10⁶ PBMCs.

Antivector T cell responses

To determine whether gene therapy induced antivector T cell immunity, cryopreserved PBMCs from patients in the inoperable arm of the trial were tested in an IFN-γ ELISpot assay, after stimulation with heat-inactivated CTL102 virus as antigen. Ad-specific T cells were at low to undetectable frequencies in the peripheral blood of patients before gene therapy, in keeping with the known decline in Ad-specific T cells with age (Sester et al., 2002). Figure 2 shows that the level of Ad-specific T cells was boosted in all patients tested after gene therapy. Posttreatment peak frequencies of Ad-specific T cells are shown in Table 1; there was no indication that high peak T cell responses correlated with poor treatment outcome. To investigate these responses in more detail, fresh PBMCs from three patients were used in a flow cytometry-based IFN-γ secretion assay with wild-type Ad5-infected cell lysate as antigen. All three patients had low frequencies of Ad-specific T cells before treatment. Injection of CL102 induced proliferation of both CD4⁺ and CD8⁺ Ad-specific T cells with increased frequencies detected within the first 28 days postinjection (Fig. 3). Frequencies then returned toward pretreatment levels and rose again following retreatment. In patient 38XX Ad-specific T cells accounted for nearly 2% of the total CD8⁺ T cell pool at 1 month postretreatment. For the three patients monitored there were three different clinical outcomes as assessed by measurement of serum PSA level: for patient 37XX disease continued to progress, whereas for patient 35XX disease stabilized, and patient 38XX responded well to treatment with PSA levels declining to less than 50% of pretreatment levels by 120 days posttreatment. Although no statistically significant correlations can be drawn from this the small number of patients, it can be seen that for patient 38XX, and to a lesser extent for 35XX, robust antivector T cell responses were not a barrier to efficacy of therapy.

FIG. 2.

Ad-specific T cell immunity. Patient PBMCs were used in an IFN-γ ELISpot assay pretreatment and at weekly and monthly intervals after initial treatment. Number of spot-forming cells in the control (A549 lysate) was subtracted from the test wells (CTL102) and results are represented as SFCs per 10⁶ PBMCs. Pretreatment and peak frequency posttreatment are shown.

FIG. 3.

CD4⁺ and CD8⁺ T cell response to Ad vector. IFN-γ secretion assay was done with PBMCs from (A) patient 35XX, (B) patient 37XX, and (C) patient 38XX following 16 hr of stimulation with Ad5 or mock-infected cell lysate. The dotted line in each panel represents time of retreatment.

To determine the specificity of the antivector T cells, an IFN-γ ELISpot assay was used to enumerate responses to five HLA class I-restricted and four HLA class II-restricted, hexon-derived T cell epitopes (Olive et al., 2002; Leen et al., 2004b; Onion et al., 2007). As patient HLA type was not available, all three patients were screened with all peptides. Notable responses were detected to the HLA-A1-restricted T cell epitope TDL, but not to the other epitopes used, in patients 37XX (Fig. 4) and 35XX; PB38XX did not respond to any of the peptides tested (data not shown). Ad-specific T cells from patient 37XX at 1 month postretreatment were selected with magnetic beads (Fig. 5A–D) and immunophenotyped. Figure 5E and F shows that the CD4⁺ Ad-specific T cells were effector memory cells, that is, CCR7^–CD45RA^–CD45RO⁺, and that the CD8⁺ cells were effector cells, that is, CCR7^–CD45RA⁺CD45RO⁺.

FIG. 4.

Immune response to conserved Ad hexon epitope TDL. PBMCs from patient 37XX were stimulated with peptide TDL (5 μg/ml) and DMSO as control in an IFN-γ ELISpot assay. Results are expressed as SFCs per 10⁶ PBMCs with background response to DMSO subtracted. The dashed line indicates the time of retreatment; error bars indicate the standard deviation of triplicate wells.

FIG. 5.

Phenotype and specificity of Ad-specific T cells. An IFN-γ secretion assay was performed with PBMCs from patient 37XX 1 month postretreatment (A and B); cells were selected with anti-PE magnetic beads (C and D). Selected cells were also stained with anti-CD45RA, anti-CD45RO, or anti-CCR7 antibodies and analyzed by gating on IFN-γ-positive cells (E and F). IFN-γ-selected cells were cloned by limiting dilution and antigen recognition was determined in IFN-γ ELISAs. (G) Two CD8⁺ T cell clones generated from patient 37XX 2 months postretreatment; (H) two CD4⁺ T cell clones generated from patient 37XX 1 month posttreatment.

T cells were successfully cloned by limiting dilution from patient 37XX at 1 month posttreatment and 1 month postretreatment, and from patient 38XX at 1 month posttreatment. In total six CD4⁺ T cell clones and three CD8⁺ T cell clones were isolated. Two clones (one CD4⁺, one CD8⁺) were specific for the hexon protein and were also serotype and species cross-reactive (Ad2, Ad4, Ad5, and Ad11 tested). One CD4⁺ clone was Ad5 fiber specific and did not recognize virus from other serotypes, including the closely related Ad2. The remaining six clones did not recognize hexon or fiber, but were specific for antigen conserved between Ad2, Ad4, and Ad5 but not Ad11 (examples are shown in Fig. 5D and E).

Antitumor T cell responses

To monitor antitumoral immunity in the gene therapy recipients, plasmid expression vectors encoding prostate tumor-associated antigens PSA and PSMA were constructed and used to nucleofect PBMCs, and responding T cells were detected in an IFN-γ ELISpot assay. Cryopreserved PBMCs from 11 gene therapy recipients before and after single or repeat treatment were tested for tumor-specific T cells. As a control, PBMCs from five healthy donors were also tested. No responses were seen to PSA and PSMA in the healthy donors and six of the gene therapy recipients, above that seen with control vector expressing GFP (patients 16XX, 19XX, 20XX, 21XX, 22X, and 24XX). Figure 6 shows that of the remaining five patients, patients 30XX and 35XX had detectable PSA and PSMA responses pre- and posttreatment. Patients 27XX, 29XX, and 31XX had low/undetectable PSA- or PSMA-specific T cells pretreatment, which were increased after gene therapy. Patient 27XX had a modest increase in PSA-specific T cells 1 month after the first round of gene therapy, but no PSMA response was detected. PSA-specific T cells returned to undetectable levels by month 2. One month postretreatment PSA-specific T cells were again detected, this time at a higher frequency, and PSMA-specific T cells also rose to detectable levels. For patient 29XX no pretreatment or day-of-injection samples were available, and at 1 month after initial treatment no tumor-specific responses were observed; however, at month 2 a notable response to PSA was detected, exceeding the slightly elevated response to the GFP control vector. No further tumor-specific T cells were detected for patient 29XX, tested 6 months after initial treatment and 1 month postretreatment. Patient 31XX had the most sustained levels of both PSA- and PSMA-specific T cells. PSA- and PSMA-specific T cells were undetectable before gene therapy but were present 2 and 3 months posttreatment. Two months following retreatment, frequencies of PSMA-specific T cells were maintained. However, PSA responses declined toward background levels 3 months postretreatment. Frequencies of PSMA-specific T cells also fell back to undetectable levels 3 months postretreatment.

FIG. 6.

PSA- and PSMA-specific T cell responses following gene therapy. Patient PBMCs were nucleofected with plasmids expressing full-length PSA (diamonds), PSMA (triangles) or eGFP (inverted triangles) and entered into an IFN-γ ELISpot assay. SFCs were enumerated and expressed as SFCs per 10⁶ PBMCs. (A) Patient 30XX; (B) patient 35XX; (C) patient 27XX; (D) patient 29XX; (E) patient 31XX. The dotted lines indicate day of treatment or retreatment.

Discussion

The purpose of this study was to investigate the role of the human immune system in the treatment of prostate cancer with the replication-deficient Ad CTL102 in combination with prodrug CB1954. The presence of preexisting adaptive immunity is often cited as a major drawback to the use of Ad vectors. Despite this, detailed analysis of adaptive immune responses in the setting of clinical trials and correlation with efficacy are scant. Farace and colleagues reported reduced efficacy of a model rAd-based vaccine in one of four patients who had the highest preexisting neutralizing antibodies of the cohort (Gahery-Segard et al., 1997). In a phase I clinical trial of suicide gene therapy using an rAd encoding herpes simplex virus thymidine kinase (HSV-tk) in combination with the prodrug ganciclovir for treatment of mesothelioma, 12 of 21 patients showed Ad5-specific proliferation of PBMCs, indicative of a cellular immune response (Molnar-Kimber et al., 1998). In these patients there was an indication that when higher preexisting nAbs were present this led to lower gene expression seen in tumor biopsies. However, there was no indication that immune responses monitored correlated with clinical response.

In our study all patients had previously been exposed to Ad5, as shown by virus neutralization assays. Both total adenovirus-specific immunoglobulin and neutralizing serotype-specific antibody titers were raised after injection of vector, and were reboosted following repeat administration of virus. In the operable arm of the trial prostate tumors were surgically resected 48 hr following administration of CTL102 and tumors were sectioned and stained for NTR expression. There was no correlation between preexisting levels of Ad neutralizing or total immunoglobulin and levels of gene expression. Within this study there was a step-wise increase in the amount of virus delivered (5 × 10¹⁰−1 × 10¹² particles), with the suggestion of a trend toward increased gene expression. There was no indication in any dose cohort that high antibody level correlated with lower gene expression. This may be because the virus was delivered by direct intraprostatic injection and thus exposure to circulating antibodies was minimal before contact with target tumor cells. In addition, the physiology of the prostate gland may be important; it is an immunologically privileged site (Leibovitz et al., 2004), hence the level of neutralizing antibodies in the peripheral blood may be higher than in the prostate. In the inoperable arm of the trial there was evidence of treatment efficacy with two having good response, three partial response, three stable disease, and seven progressive disease. Higher preexisting nAb did not correlate with failure to respond to treatment as measured by serum PSA level.

Ad-specific memory T cells in healthy individuals are predominantly CD4⁺ with low-frequency CD8⁺ memory T cells, often detectable only with the aid of highly sensitive HLA class I tetramer technology. IFN-γ ELISpot data from 11 patients in the inoperable arm showed low preexisting total Ad-specific T cells in the peripheral blood, consistent with a previously reported decline in frequency with age (Sester et al., 2002) (mean age of patients was 67.6 years). In all patients tested, administration of CTL102 resulted in expansion of adenovirus-specific T cells, which were reboosted in those patients receiving a repeat injection of CTL102. Three patients' responses were studied in detail in IFN-γ secretion assays, IFN-γ ELISpot assays with known Ad-derived T cell epitopes, and T cell cloning to further study antigen specificity. From this it was clear that administration of CTL102 resulted in rapid expansion of both CD4⁺ and CD8⁺ T cells, with peak responses for patient 38XX accounting for 2% of total CD8⁺ T cells. These results indicate the high proliferative capacity of low-frequency memory Ad-specific T cells carried by the majority if not all the population and to our knowledge is the first report of the kinetics of T cell responses to Ad or Ad-based vector. The hexon-derived epitope TDL was shown to be an important target for Ad-specific T cells in two of three patients studied and challenge with Ad was shown to rapidly increase these cells from very low frequency. T cells recognizing TDL are known to be serotype cross-reactive (Leen et al., 2004b) and indeed all hexon-specific T cell clones isolated from patients in our study recognized all serotypes tested. There were serotype-specific responses to the Ad fiber protein and also responses to other as yet undetermined viral protein(s). Response to treatment was seen in patients with high-frequency T cell responses, indicating that preexisting T cell immunity to adenovirus is not a barrier to Ad-delivered suicide gene therapy.

We have shown that direct cytotoxic killing of prostate tumor cells by the CTL102–CB1954 system resulted in increased levels of tumor antigen-specific T cells in 3 of 11 patients. Although these numbers are small, they are encouraging in the development of tumor-killing strategies that simultaneously induce an immune bystander effect. It has been shown that chemotherapy agents such as bortezomib and anthracyclins can mediate immunogenic apoptosis whereas others such as mitomycin C do not (Zitvogel et al., 2004; Casares et al., 2005; Obeid et al., 2007; Spisek et al., 2007). Key to the apoptosis being immunogenic is the cell surface expression of heat shock proteins (Spisek et al., 2007; Obeid et al., 2007) and the stimulation of dendritic cells to mature, enabling immunogenic rather than tolerogenic antigen presentation (Apetoh et al., 2007). Cell killing caused by activated CB1954 occurs via a caspase-dependent apoptotic mechanism (Palmer et al., 2003); whether this is an immunogenic form of apoptosis is the subject of current investigation. Replication-deficient Ad vectors are known to induce the maturation of monocytes and immature dendritic cells (Rea et al., 1999; Morelli et al., 2000; Rouard et al., 2000; Korst et al., 2002; Lyakh et al., 2002; Miller et al., 2003; Molinier-Frenkel et al., 2003) and in this system may provide an appropriate danger signal to help promote antigen presentation, leading to the generation of tumor-specific immunity.

Taken together these data suggest that adenovirus-mediated NTR expression in combination with CB1954 for the treatment of prostate cancer is not inhibited by preexisting adaptive immunity to the vector and that tumor cell killing in this context may be immunogenic.

Footnotes

Acknowledgments

The clinical trial was sponsored by Innovata (formerly ML Laboratories), and the authors acknowledge the support of the clinical and nursing teams involved.

Author Disclosure Statement

No competing financial interests exists.

References

Apetoh

, Ghiringhelli

, Tesniere

, Obeid

, Ortiz

, Criollo

, Mignot

, Maiuri

M.C.

, Ullrich

, Saulnier

, Yang

, Amigorena

, Ryffel

, Barrat

F.J.

, Saftig

, Levi

, Lidereau

, Nogues

, Mira

J.P.

, Chompret

, Joulin

, Clavel-Chapelon

, Bourhis

, Andre

, Delaloge

, Tursz

, Kroemer

, Zitvogel

2007. Toll-like receptor 4-dependent contribution of the immune system to anticancer chemotherapy and radiotherapy. Nat. Med., 13:1050–1059.

Bessis

, GarciaCozar

F.J.

, Boissier

M.C.

2004. Immune responses to gene therapy vectors: Influence on vector function and effector mechanisms. Gene Ther, 11,Suppl. 1:S10–S17.

Bridgewater

J.A.

, Knox

R.J.

, Pitts

J.D.

, Collins

M.K.

, Springer

C.J.

1997. The bystander effect of the nitroreductase/CB1954 enzyme/prodrug system is due to a cell-permeable metabolite. Hum. Gene Ther., 8:709–717.

Casares

, Pequignot

M.O.

, Tesniere

, Ghiringhelli

, Roux

, Chaput

, Schmitt

, Hamai

, Hervas-Stubbs

, Obeid

, Coutant

, Metivier

, Pichard

, Aucouturier

, Pierron

, Garrido

, Zitvogel

, Kroemer

2005. Caspase-dependent immunogenicity of doxorubicin-induced tumor cell death. J. Exp. Med., 202:1691–1701.

Chen

, Kovesdi

, Bruder

J.T.

2000. Effective repeat administration with adenovirus vectors to the muscle. Gene Ther., 7:587–595.

Dai

, Schwarz

E.M.

, Gu

, Zhang

W.W.

, Sarvetnick

, Verma

I.M.

1995. Cellular and humoral immune responses to adenoviral vectors containing factor IX gene: Tolerization of factor IX and vector antigens allows for long-term expression. Proc. Natl. Acad. Sci. U.S.A., 92:1401–1405.

Friedlos

, Quinn

, Knox

R.J.

, Roberts

J.J.

1992. The properties of total adducts and interstrand crosslinks in the DNA of cells treated with CB 1954: Exceptional frequency and stability of the crosslink. Biochem. Pharmacol., 43:1249–1254.

Gahery-Segard

, Molinier-Frenkel

, Le Boulaire

, Saulnier

, Opolon

, Lengagne

, Gautier

, Le Cesne

, Zitvogel

, Venet

, Schatz

, Courtney

, Le Chevalier

, Tursz

, Guillet

J.G.

, Farace

1997. Phase I trial of recombinant adenovirus gene transfer in lung cancer: Longitudinal study of the immune responses to transgene and viral products. J. Clin. Invest., 100:2218–2226.

Garnett

C.T.

, Erdman

, Xu

, Gooding

L.R.

2002. Prevalence and quantitation of species C adenovirus DNA in human mucosal lymphocytes. J. Virol., 76:10608–10616.

10.

Heemskerk

, Veltrop-Duits

L.A.

, van Vreeswijk

, ten Dam

M.M.

, Heidt

, Toes

R.E.

, van Tol

M.J.

, Schilham

M.W.

2003. Extensive cross-reactivity of CD4⁺ adenovirus-specific T cells: Implications for immunotherapy and gene therapy. J. Virol., 77:6562–6566.

11.

Horwitz

M.S.

Fields

B.N.

, Knipe

D.M.

2001. Adenoviruses. Fields Virology. Lippincott Williams & Wilkins: Philadelphia, 2301–2326.

12.

Korst

R.J.

, Mahtabifard

, Yamada

, Crystal

R.G.

2002. Effect of adenovirus gene transfer vectors on the immunologic functions of mouse dendritic cells. Mol. Ther., 5:307–315.

13.

Leen

A.M.

, Sili

, Savoldo

, Jewell

A.M.

, Piedra

P.A.

, Brenner

M.K.

, Rooney

C.M.

2004a. Fiber-modified adenoviruses generate subgroup cross-reactive, adenovirus-specific cytotoxic T lymphocytes for therapeutic applications. Blood, 103:1011–1019.

14.

Leen

A.M.

, Sili

, Vanin

E.F.

, Jewell

A.M.

, Xie

, Vignali

, Piedra

P.A.

, Brenner

M.K.

, Rooney

C.M.

2004b. Conserved CTL epitopes on the adenovirus hexon protein expand subgroup cross-reactive and subgroup-specific CD8⁺ T cells. Blood, 104:2432–2440.

15.

Leibovitz

, Baumoehl

, Segal

2004. Increased incidence of pathological and clinical prostate cancer with age: Age related alterations of local immune surveillance. J. Urol., 172:435–437.

16.

Lieber

, He

C.Y.

, Meuse

, Schowalter

, Kirillova

, Winther

, Kay

M.A.

1997. The role of Kupffer cell activation and viral gene expression in early liver toxicity after infusion of recombinant adenovirus vectors. J. Virol., 71:8798–8807.

17.

Lyakh

L.A.

, Koski

G.K.

, Young

H.A.

, Spence

S.E.

, Cohen

P.A.

, Rice

N.R.

2002. Adenovirus type 5 vectors induce dendritic cell differentiation in human CD14⁺ monocytes cultured under serum-free conditions. Blood, 99:600–608.

18.

Miller

, Lahrs

, Shah

A.B.

, DeMatteo

R.P.

2003. Optimization of dendritic cell maturation and gene transfer by recombinant adenovirus. Cancer Immunol. Immunother., 52:347–358.

19.

Molinier-Frenkel

, Prevost-Blondel

, Hong

S.S.

, Lengagne

, Boudaly

, Magnusson

M.K.

, Boulanger

, Guillet

J.G.

2003. The maturation of murine dendritic cells induced by human adenovirus is mediated by the fiber knob domain. J. Biol. Chem., 278:37175–37182.

20.

Molnar-Kimber

K.L.

, Sterman

D.H.

, Chang

, Kang

E.H.

, ElBash

, Lanuti

, Elshami

, Gelfand

, Wilson

J.M.

, Kaiser

L.R.

, Albelda

S.M.

1998. Impact of preexisting and induced humoral and cellular immune responses in an adenovirus-based gene therapy phase I clinical trial for localized mesothelioma. Hum. Gene Ther., 9:2121–2133.

21.

Morelli

A.E.

, Larregina

A.T.

, Ganster

R.W.

, Zahorchak

A.F.

, Plowey

J.M.

, Takayama

, Logar

A.J.

, Robbins

P.D.

, Falo

L.D.

, Thomson

A.W.

2000. Recombinant adenovirus induces maturation of dendritic cells via an NF-κB-dependent pathway. J. Virol., 74:9617–9628.

22.

Obeid

, Tesniere

, Ghiringhelli

, Fimia

G.M.

, Apetoh

, Perfettini

J.L.

, Castedo

, Mignot

, Panaretakis

, Casares

, Metivier

, Larochette

, van Endert

, Ciccosanti

, Piacentini

, Zitvogel

, Kroemer

2007. Calreticulin exposure dictates the immunogenicity of cancer cell death. Nat. Med., 13:54–61.

23.

Olive

, Eisenlohr

, Flomenberg

, Hsu

, Flomenberg

2002. The adenovirus capsid protein hexon contains a highly conserved human CD4⁺ T-cell epitope. Hum. Gene Ther., 13:1167–1178.

24.

Onion

, Crompton

L.J.

, Milligan

D.W.

, Moss

P.A.

, Lee

S.P.

, Mautner

2007. The CD4⁺ T-cell response to adenovirus is focused against conserved residues within the hexon protein. J. Gen. Virol., 88:2417–2425.

25.

Palmer

D.H.

, Milner

A.E.

, Kerr

D.J.

, Young

L.S.

2003. Mechanism of cell death induced by the novel enzyme-prodrug combination, nitroreductase/CB1954, and identification of synergism with 5-fluorouracil. Br. J. Cancer, 89:944–950.

26.

Palmer

D.H.

, Mautner

, Mirza

, Oliff

, Gerritsen

, van der Sijp

J.R.

, Hubscher

, Reynolds

, Bonney

, Rajaratnam

, Hull

, Horne

, Ellis

, Mountain

, Hill

, Harris

P.A.

, Searle

P.F.

, Young

L.S.

, James

N.D.

, Kerr

D.J.

2004. Virus-directed enzyme prodrug therapy: Intratumoral administration of a replication-deficient adenovirus encoding nitroreductase to patients with resectable liver cancer. J. Clin. Oncol., 22:1546–1552.

27.

Patel

, Young

J.G.

, Mautner

, Ashdown

, Bonney

, Pineda

R.G.

, Collins

S.I.

, Searle

P.F.

, Hull

, Peers

, Chester

, Wallace

D.M.

, Doherty

, Leung

, Young

L.S.

, James

N.D.

2009. A phase I/II clinical trial in localized prostate cancer of an adenovirus expressing nitroreductase with CB1984. Mol. Ther., 17:1292–1299.

28.

Peng

2005. Current status of Gendicine in China: Recombinant human Ad-p53 agent for treatment of cancers. Hum. Gene Ther., 16:1016–1027.

29.

Rea

, Schagen

F.H.

, Hoeben

R.C.

, Mehtali

, Havenga

M.J.

, Toes

R.E.

, Melief

C.J.

, Offringa

1999. Adenoviruses activate human dendritic cells without polarization toward a T-helper type 1-inducing subset. J. Virol., 73:10245–10253.

30.

Rouard

, Leon

, Klonjkowski

, Marquet

, Tenneze

, Plonquet

, Agrawal

S.G.

, Abastado

J.P.

, Eloit

, Farcet

J.P.

, Delfau-Larue

M.H.

2000. Adenoviral transduction of human “clinical grade” immature dendritic cells enhances costimulatory molecule expression and T-cell stimulatory capacity. J. Immunol. Methods, 241:69–81.

31.

Schnell

M.A.

, Zhang

, Tazelaar

, Gao

G.P.

, Yu

Q.C.

, Qian

, Chen

S.J.

, Varnavski

A.N.

, LeClair

, Raper

S.E.

, Wilson

J.M.

2001. Activation of innate immunity in nonhuman primates following intraportal administration of adenoviral vectors. Mol. Ther., 3:708–722.

32.

Sester

, Sester

, Alarcon

S.S.

, Heine

, Lipfert

, Girndt

, Gartner

, Kohler

2002. Age-related decrease in adenovirus-specific T cell responses. J. Infect. Dis., 185:1379–1387.

33.

Smith

C.A.

, Woodruff

L.S.

, Kitchingman

G.R.

, Rooney

C.M.

1996. Adenovirus-pulsed dendritic cells stimulate human virus-specific T-cell responses in vitro. J. Virol., 70:6733–6740.

34.

Smith

C.A.

, Woodruff

L.S.

, Rooney

, Kitchingman

G.R.

1998. Extensive cross-reactivity of adenovirus-specific cytotoxic T cells. Hum. Gene Ther., 9:1419–1427.

35.

Spisek

, Charalambous

, Mazumder

, Vesole

D.H.

, Jagannath

, Dhodapkar

M.V.

2007. Bortezomib enhances dendritic cell (DC)-mediated induction of immunity to human myeloma via exposure of cell surface heat shock protein 90 on dying tumor cells: Therapeutic implications. Blood, 109:4839–4845.

36.

Stallwood

, Fisher

K.D.

, Gallimore

P.H.

, Mautner

2000. Neutralisation of adenovirus infectivity by ascitic fluid from ovarian cancer patients. Gene Ther., 7:637–643.

37.

Vile

R.G.

, Castleden

, Marshall

, Camplejohn

, Upton

, Chong

1997. Generation of an anti-tumour immune response in a non-immunogenic tumour: HSVtk killing in vivo stimulates a mononuclear cell infiltrate and a Th1-like profile of intratumoural cytokine expression. Int. J. Cancer, 71:267–274.

38.

Vile

R.G.

, Russell

S.J.

, Lemoine

N.R.

2000. Cancer gene therapy: Hard lessons and new courses. Gene Ther., 7:2–8.

39.

Young

L. S.

, Searle

P.F.

, Onion

, Mautner

2006. Viral gene therapy strategies: From basic science to clinical application. J. Pathol, 208:299–318.

40.

Zitvogel

, Casares

, Pequignot

M.O.

, Chaput

, Albert

M.L.

, Kroemer

2004. Immune response against dying tumor cells. Adv. Immunol., 84:131–179.