|Home | About | Journals | Submit | Contact Us | Français|
Prostatic acid phosphatase (PAP) is a tumor antigen in prostate cancer and the target of several anti-tumor vaccines in earlier clinical trials. Ultimately, the goal of anti-tumor vaccines is to elicit a sustainable immune response, able to eradicate a tumor, or at least restrain its growth. We have investigated plasmid DNA vaccines and have previously conducted a phase 1 trial in which patients with recurrent prostate cancer were vaccinated with a DNA vaccine encoding PAP. In this study, we investigated the immunologic efficacy of subsequent booster immunizations, and conducted more detailed longitudinal immune analysis, to answer several questions aimed at guiding optimal schedules of vaccine administration for future clinical trials. We report that antigen-specific cytolytic T-cell responses were amplified after immunization in 7 of 12 human leukocyte antigen-A2-expressing individuals, and that multiple immunizations seemed necessary to elicit PAP-specific interferon-γ-secreting immune responses detectable by enzyme-linked immunosorbent spot assay. Moreover, among individuals who experienced a ≥200% increase in prostate-specific antigen doubling time, long-term PAP-specific interferon-γ-secreting T-cell responses were detectable in 6 of 8, but in only 1 of 14 individuals without an observed change in prostate-specific antigen doubling time (P=0.001). Finally, we identified that immune responses elicited could be further amplified by subsequent booster immunizations. These results suggest that future trials using this DNA vaccine, and potentially other anti-tumor DNA vaccines, could investigate ongoing schedules of administration with periodic booster immunizations. Moreover, these results suggest that DNA vaccines targeting PAP could potentially be combined in heterologous immunization strategies with other vaccines to further augment PAP-specific T-cell immunity.
Anti-tumor vaccines, or active immunotherapies, seek to elicit immune responses within the host that recognize individual or multiple tumor-associated antigens to eliminate tumor cells expressing these antigens. Over the last 2 decades, there has been considerable evaluation of various anti-tumor vaccine approaches in preclinical models and human trials. Prostate cancer is one malignancy in particular where there have been significant efforts, due to several factors: (1) the disease is highly prevalent and the second leading cause of cancer-related death in men in the United States, and therefore a significant health concern for which new therapies are needed; (2) there is typically a long natural history and consequently a potentially long interval of time for the development of a therapeutic anti-tumor immune response; (3) the availability of a serum marker [prostate-specific antigen (PSA)] as a biochemical indicator of minimal residual disease; (4) the prostate is an expendable organ, and consequently autoimmune responses against normal prostate tissue are not a significant concern; and (5) several tissue-specific antigens have already been identified that might serve as targets for vaccines. Within the last decade, several antigen-specific vaccine approaches have been investigated in clinical trials for patients with prostate cancer targeting PSA,1-5 prostatic acid phosphatase (PAP),6,7 prostate-specific membrane antigen,8-10 and multiple carbohydrate antigens.11
Despite the significant effort in developing and evaluating anti-tumor vaccines, there is little consensus as to the “best” antigens to target and the optimal means of targeting these antigens. The prioritization of anti-tumor vaccine antigens, in fact, has been the focus of recent efforts led by the National Cancer Institute.12 Most vaccine approaches targeting individual antigens have focused on eliciting CD8+ T cells, as these are part of the adaptive arm of the immune system with direct cytolytic activity. Consequently, to specifically elicit CD8+ T cells, most approaches have used antigen-loaded antigen-presenting cellular vaccines13-15 or genetic vaccines using viral vaccines or naked DNA plasmids.3-5,8,10 Of these particular approaches, DNA vaccines are generally believed to be a “weaker” immunization strategy given the absence of a concurrent inflammatory antiviral response, and the low level of in-vivo transfection of antigen-presenting cells that occurs after direct administration. However, we have been particularly interested in DNA vaccines as a simpler means of antigen-specific immunization given our experience in rodent models demonstrating that DNA vaccines can elicit antigen-specific CD8+ T-cell responses to an autologous antigen with repetitive immunization and without the need for heterologous immunization strategies.16,17 These findings suggest that repetitive immunization with DNA vaccines might lead to CD8+ anti-tumor T-cell responses as effectively as more cumbersome strategies. Moreover, several anti-tumor DNA vaccines have begun to enter clinical trials and have demonstrated immunologic efficacy and safety.18-20 However, with the continued clinical development of DNA vaccines many questions will need to be answered, such as the number of immunizations necessary to elicit effector CD8+ T cells, the durability of this response, and whether repetitive immunization continues to augment a productive anti-tumor immune response or ultimately leads to tolerant immune responses.
We have recently reported the results of a phase 1 trial in which patients with PSA-recurrent prostate cancer were immunized 6 times at biweekly intervals using a DNA vaccine encoding PAP.21 We demonstrated that after 6 immunizations, PAP-specific T-cell responses were detectable in 10 of 22 individuals. In this report, we have conducted more detailed longitudinal immune analysis designed to answer the following specific questions: (1) Were immune responses elicited early in the course of immunization, or only after multiple immunizations, suggesting that continued immunization might be necessary? (2) Did delayed immune responses occur months after an initial immunization course? (3) Were immune responses, once elicited, durable? and (4) Were the development of immune responses associated with possible clinical responses as determined by changes in PSA doubling time? In addition, 2 patients who previously developed PAP-specific CD8+ T-cell responses were further vaccinated to answer whether PAP-specific T-cell immune responses, once elicited, could be further augmented by subsequent booster immunizations.
We have previously reported the results of a phase 1 trial using a plasmid DNA vaccine, pTVG-HP, in patients with biochemically (serum PSA) recurrent prostate cancer after definitive surgery and/or radiation therapy not receiving androgen deprivation therapy.21 Cryopreserved samples from that trial were used for the analysis described. In addition, that trial was amended to permit the evaluation of booster immunizations using the same pTVG-HP vaccine. Specifically, 2 subjects who had previously been vaccinated, had evidence of a PAP-specific CD8+ T-cell response defined by interferon (IFN)γ enzyme-linked immunosorbent spot (ELISPOT) or T-cell proliferation after immunization, had no significant adverse events during the initial treatment, and who had otherwise stable disease not requiring additional therapy for prostate cancer, were permitted to participate. Subjects were treated at monthly intervals with 100 μg pTVG-HP admixed with 200 μg granulocyte macrophage colony-stimulating factor (GM-CSF; Leukine, sargramostim, Berlex Oncology, Bergenfield, NJ), and coadministered intradermally with a 28-guage needle and syringe as described earlier.21 Patients were evaluated for potential toxicity after every 2 vaccinations by examination and blood tests as reported earlier.21 Serum PSA was evaluated monthly for disease monitoring, and blood was obtained after every 2 booster immunizations for immunologic monitoring. The study protocol was reviewed and approved by all local (including Institutional Review Board of the University of Wisconsin and Biosafety Committee), sponsor (Human Subjects Review Board of the US Army), and federal (Food and Drug Administration, NIH Recombinant DNA Advisory Committee) entities. Both patients gave additional written informed consent for participation.
Peptide-specific T-cell lines were generated by repetitive weekly in-vitro stimulations with peptide-loaded autologous dendritic cells (DC). Briefly, peripheral blood mononuclear cells (PBMC) from human leukocyte antigen (HLA)-A2-expressing individuals were plated at 108 cells/flask in T150 tissue culture flasks for 2 hours at 37°C. Nonadherent cells were then removed, and the adherent cells were cultured for 6 days in the presence of 20 ng/mL rhGM-CSF (R&D Systems, Minneapolis, MN) and 10 ng/mL rhIL-4 (R&D Systems). The resulting DC were then “matured” for 24 to 48 hours by culture in the presence of 150 ng/mL interleukin (IL)-6, 10 ng/mL IL-1β, 10 ng/mL tumor necrosis factor-α, and 1 μg/mL prostaglandin E2. Autologous DC were then pulsed with 10 μg/mL peptide [p18-26, p112-120, or p299-307 derived from PAP, or GILGFVFTL derived from the influenza A matrix protein (pFlu)] for 2 hours, and cultured with autologous T cells selected from PBMC (T-cell negative isolation kit, Dynal, Carlsbad, CA), and obtained before or 2 weeks after immunization, in T-cell medium (Roswell Park Memorial Institute 1640 medium supplemented with l-glutamine, penicillin/streptomycin, β-mercaptoethanol, and 10% human AB serum (Valley Biomedical, Winchester, VA). 20 U/mL rhIL-2 (R&D Systems) and 10 U/mL rhIL-7 (R&D Systems) were added after 24 hours. Cells were restimulated at weekly intervals with irradiated (3000 cGy) peptide-loaded antigen-presenting cells (autologous DC or HLA-A2-expressing B-cell lines) in similar fashion, and T-cell lines were characterized for cytolytic function after 2 to 5 weeks. Cytolytic activity was measured by lactate dehydrogenase release from target cell lines (Cytotox 96 Assay kit, Promega, Madison, WI). Specifically, effector cell lines were plated in 96-well plates at various effector-to-target (E:T) cell ratios. Targets used were T2 cells pulsed with specific peptide or an irrelevant HLA-A2-binding peptide. After 4- to 6-hours at 37°C, plates were centrifuged and 50 μL of culture supernatant was assessed for lactate dehydrogenase concentration spectrophotometrically, according to the manufacturer’s instructions. Controls included wells with effector cells only, wells with media only, target cells only (minimum release), and target cells with 1% Triton X-100 (maximum release). The optical density (OD) signal contributed by the media alone was subtracted from all values. The percent specific lysis was then calculated as: (ODexperimental well – ODeffector only–ODtarget minimum release)/(ODtarget maximum release–ODtarget minimum release). A “positive” cytotoxic T-lymphocyte (CTL) response was defined as the detection of a titratable lysis of peptide-specific T2 cells and significantly higher relative to irrelevant peptide-loaded T2 cells. The first week at which peptide-specific CTL were detected is recorded. Of note, pFlu-specific cultures and CTL analysis were only conducted with pretreatment PBMC specimens from selected patients.
Wells of nitrocellulose 96-well microtiter (ELISPOT) plates were coated with an anti-IFNγ capture monoclonal antibody (Endogen, Rockford, IL), according to the manufacturer’s instructions. Cryopreserved PBMC, obtained at various times before or after vaccination, were then thawed and cultured directly (without in-vitro stimulation) in T-cell medium in these nitrocellulose microtiter (ELISPOT) plates at 2×105 cells/well. Cultures were allowed to proceed for 48- to 72-hours in the presence of media only (no antigen), 2 μg/mL PAP protein (Research Diagnostics Inc, Flanders, NJ), 2 μg/mL PSA protein (Research Diagnostics Inc, negative control), 250 ng/mL tetanus toxoid (Calbiochem, San Diego, CA), or 5 μg/mL phytohemaglutinin (positive mitogenic control, Fisher, Pittsburgh, PA) in 3- to 6-well replicates. ELISPOT plates were then washed and probed for 1.5 hours with a biotinylated anti-IFNγ antibody (Endogen). After incubation, wells were washed, incubated with streptavidin-labeled alkaline phosphatase for 1 hour, and then developed with 5-bromo 4-chloro 3-indoylphosphate/nito blue tetrazolium chloride colorimetric substrate (BioRad, Hercules, CA) for 15 to 30 minutes. The number of spots per well was determined with an automated ELISPOT reader and normalized to 106 starting PBMC. The mean number of spots detected under media-only conditions at each time point was subtracted from the antigen-specific conditions to enumerate antigen-specific IFNγ spot-forming units±SD. Comparison of experimental wells with control no antigen wells was performed using a 2-tailed t test, with P-value less than 0.05 used to define a significant T-cell response. Data from samples for which the phytohemaglutinin-positive control was not significantly positive were not considered viable or adequate for interpretation; however, these data are included for completeness.
CD4+ and CD8+ T-cell proliferation in response to antigen stimulation was determined by a 5-day bromodeoxyuridine (BrdU) incorporation assay, using the same methodology and individual lots of test antigens as described earlier.21 For analysis, a proliferation index was determined as the percentage of BrdU+ events for each antigen-stimulated condition divided by the percentage of BrdU+ events with the media-only control. BrdU+ events were assessed at each time point by gating 0.05% events in the no antigen (media only) control group, and applying that gating to all antigen-stimulated conditions.
Cryopreserved PBMC from multiple time points were thawed, resuspended in T-cell medium, and cultured for 1 week with HLA-A2-derived epitopes p112-120 (TLMSAMTNL) and p299-307 (ALDVYNGLL), as well as an HLA-A2-binding PAP-derived nonspecific control peptide p135-143 (ILLWQPIPV). IL-2 was added to 10 IU/mL on the second day of culture. After 1 week, cell cultures were stained for CD3, CD8, and pentamers (ProImmune, Inc, Bradenton, FL) specific for each peptide (versus isotype control) according to the manufacturer’s instructions. The percentage of pentamer-labeled cells at each time point was determined as: [(% CD3+ CD8+ Pentamer+) – (% CD3+ CD8+ Isotype control+)]/(% CD3+ CD8+).
We have recently reported the results of a phase 1 trial of a DNA vaccine encoding PAP in patients with biochemical recurrence of prostate cancer. In that report, we identified that 10 of 22 patients had evidence of a T-cell response specific for PAP at the conclusion of the 12-week immunization period as evidenced by antigen-specific T-cell proliferation and/or CD8+ IFNγ ELISPOT. To further characterize the resulting immune response, and determine whether cytolytic T cells were specifically amplified after immunization, we analyzed the 12 HLA-A2+ subjects for the presence of PAP epitope-specific T cells22. PBMC were cultured in the presence of 3 HLA-A2-specific peptides derived from the amino-acid sequence of PAP (p18-26, p112-120, and p299-307) and evaluated after weekly in vitro stimulations with peptide for the presence of peptide-specific CTL (Fig. 1). CTL could be detected specific for pFlu after as few as 2 weekly in-vitro stimulations in the majority of patients before immunization, demonstrating the viability of the cells and robustness of the methodology (Fig. 1D). As demonstrated, CTL specific for p18-26 (Fig. 1A), p112-120 (Fig. 1B), and p299-307 (Fig. 1C), previously identified as HLA-A2-specific PAP epitopes, could be detected in multiple individuals after as few as 2 weekly in-vitro stimulations from PBMC obtained after immunization but not from PBMC obtained before immunization. Although these methods are not strictly quantitative, they suggest that the frequency of these cells were augmented after DNA immunization, given that fewer in-vitro stimulations were necessary for detection after immunization.
In the earlier clinical trial, we observed that approximately half of the patients vaccinated did not have evidence of PAP-specific T cells at the completion of the immunization series. We consequently questioned whether this was potentially because of an inadequate number of immunizations, reasoning that if immune responses were detectable only later in the course of immunization, that potentially 6 immunizations with this DNA vaccine was an insufficient number of immunizations for the majority of patients. PBMC were available pretreatment, and after 2, 4, or 6 immunizations from multiple individuals, and PAP-specific T cells from these time points were assessed by IFNγ ELISPOT without in-vitro stimulation. By this methodology, 4 subjects had evidence of significant PAP-specific IFNγ-secreting T cells after beginning the immunization series and not at baseline (Fig. 2). As demonstrated in Figure 2, responses tended to increase with the number of immunizations, and the majority of significant PAP-specific T cells were detectable only after 6 immunizations.
We next questioned whether immune responses once elicited were durable, or initially detectable, over a year of follow-up. We reasoned that potentially beneficial immune responses, those that might be associated with a better clinical outcome, should be durable. PBMC were available after 3, 6, 9, and/or 12 months from all 22 subjects, and the presence of PAP-specific IFNγ-secreting T cells were assessed by IFNγ ELISPOT as shown above. Of 8 individuals previously reported to have experienced a ≥200% increase in PSA doubling time in the year of follow-up after vaccination (Table 1),21 6 were found to have PAP-specific IFNγ-secreting immune responses at least twice in this year period (Figs. 3A–H). Although some immune responses detectable immediately after immunization seemed durable (Fig. 3B), others seemed to wane over time (Fig. 3I), and others seemed to develop months after the immunization was completed (Figs. 3D, E). Of the 14 individuals who did not experience an increase in PSA doubling time, only 1 was found to have detectable PAP-specific IFNγ-secreting immune responses at least twice in the year of follow-up (Fig. 3J). The association of durable immune responses, defined here as responses detectable at least twice in the 4 quarterly PBMC specimens, was highly associated with a ≥200% increase in PSA doubling time over this same period compared with preimmunization (P=0.001, χ2 test).
As some immune responses were observed to wane over time, we next questioned whether immune responses, once elicited by DNA immunization, could be augmented with subsequent “booster” immunizations. The clinical protocol, described by McNeel et al,21 was modified to permit subsequent booster immunizations of subjects who had previous evidence of a PAP-specific CD8+ T-cell response after immunization, no evidence of adverse events, and otherwise stable prostate cancer not requiring other therapies. These immunizations were given at a dose of 100 μg plasmid DNA, intradermally, with 200 μg GM-CSF as adjuvant, at monthly intervals for 6 additional immunizations. Blood was collected after every 2 immunizations for immunologic monitoring. Two subjects chose to pursue this, of whom 1 (number 13) began 7 months after completing the initial immunization course. This particular subject, although he had evidence of a CD8+ immune response specific for PAP immediately after the initial immunization course by one assay (CD8+ ELISPOT), had no evidence of an immune response at any other posttreatment time point, and did not demonstrate an increased immune response after the booster immunizations (data not shown). The other subject (number 3), had evidence of PAP-specific T cells after the initial immunizations by multiple methods (Figs. 1A, C, Fig. 2B).21 This subject received monthly booster immunizations beginning 19 months after the initial immunization series. Immediately before receiving this second course, no PAP-specific T cells could be detected by antigen-specific T-cell proliferation (Fig. 4A). By as few as 2 booster immunizations, however, PAP-specific CD4+ and CD8+ T cells were detectably augmented (Fig. 4A). As this particular subject was HLA-A2+, pentamer staining was performed to directly assess for PAP-specific CD8+ T cells. As shown in Figure 4B, PAP-specific CD8+ T cells were clearly augmented shortly after the initial immunization series and the booster immunizations months later. This particular individual remains in long-term follow-up, and to date has required no other therapy for prostate cancer (Fig. 4C). No subject experienced any adverse events greater than grade 1 during the 6-month booster immunization period.
We have previously reported the results of a clinical trial in which patients with biochemically recurrent prostate cancer were immunized 6 times at 2-week intervals with a DNA vaccine encoding PAP.21 Ultimately, the goal of antitumor vaccines is to elicit a sustainable immune response able to eradicate or restrain a tumor. Hence, most clinical vaccine trials have used schedules of immunization, informed by animal studies, conducted over a limited period of time with the hope of eliciting durable anti-tumor immune responses. This approach is in contrast to most current targeted molecular pharmacologic approaches to the treatment of cancer in which it is generally assumed that chronic therapy is necessary. In addition, as most preclinical immunization studies are performed in inbred rodent strains, estimations of appropriate vaccine schedules may not be uniformly applicable to human trials. Moreover, the means of immunization may greatly affect optimal schedules of administration. To date, despite anti-tumor vaccines having entered phase 3 clinical trials, there has been very little evaluation of whether the number and schedule of immunizations is sufficient to elicit an immune response that is potentially therapeutic and durable. This leaves questions about whether prolonged schedules of administration should be considered, or booster immunizations should be used. In addition, there have been concerns that repetitive immunization might lead to counterproductive tolerant immune responses.23 In an earlier vaccine trial, patients were immunized 6 times at 2-week intervals with a DNA vaccine encoding PAP, a schedule that we had investigated earlier in rats.16 The current analysis was conducted to answer several questions aimed at guiding the schedule of vaccine administration to be used in future clinical trials with this particular DNA vaccine. Although the number of samples evaluable was relatively small, we sought to determine the following:
Using IFNγ ELISPOT as a measure of PAP-specific T cells, we found that in those individuals in whom PAP-specific CD4+ and/or CD8+ T-cell responses were elicited over the course of immunization, the majority of responses were elicited late in the course of DNA immunization. This was similar to what we observed earlier in rats in which although immune responses to the human PAP xenoantigen were elicited after as few as 2 vaccinations, more vaccinations were necessary with a vaccine encoding the rat PAP self-antigen to elicit responses.17 This finding suggests that it could be advantageous to continue immunization beyond 6 initial vaccinations, using DNA immunization alone without a heterologous immunization strategy. Alternatively, it is possible that some individuals were tolerant to PAP at baseline and might not develop detectable PAP-specific T cells. Future studies will explore continued immunization, beyond 6 initial vaccinations, to evaluate these possibilities. However, as suggested earlier, these types of studies must be performed with some caution given that genetic vaccines delivered at frequent intervals over prolonged periods may elicit tolerant responses, and have in fact been investigated by other groups as a means of inducing T-cell tolerance.24
We have identified earlier that rats immunized multiple times with the same DNA vaccine encoding human PAP developed PAP-specific memory CD4+ and CD8+ T cells detectable by antigen-specific T-cell proliferation for at least 1 year after immunization (data not shown). Although in that system the human PAP is a xenoantigen, we similarly identified that several patients had long-lived responses detectable many months after immunization (Fig. 3). Although such findings complicate clinical trial evaluation, as it is difficult to define an appropriate time to measure immune responses, they suggest that immune monitoring at static time points may be insufficient to characterize what is likely a more dynamic process. This may be one reason why many vaccine trials to date have failed to demonstrate an association between the development of an immunologic response and a clinical response. In our previous trial, patients were enrolled with biochemical (serum PSA) recurrence only, and we identified that 8 patients experienced a ≥200% increase in PSA doubling time from pretreatment to the 1-year period of time after treatment.21 We also identified that 10 of 22 patients had a PAP-specific response detectable immediately after immunization, although this was not associated with changes in PSA doubling time.21 Although changes in PSA doubling time have to be interpreted with caution, because this has not been validated prospectively as a clinical trial endpoint, changes in PSA kinetics remain the only marker of possible clinical response in this population of patients with early recurrent disease. In this study, we identified that changes in PSA doubling time were highly associated with the development of late/durable immune responses. Future clinical trials will evaluate whether the durability of PAP-specific T cells, and effector and memory T-cell immune responses in particular, are positively associated with clinical measures such as time to disease progression.
Finally, our studies demonstrate that immune responses to PAP can be subsequently augmented with the same DNA vaccine encoding PAP. In one individual, PAP-specific responses were detected after as few as 2 booster immunizations in terms of proliferative T-cell responses and specific CD8+ T-cell populations. Of note, this particular individual was treated with only 100 μg of plasmid DNA for each immunization. Although the immunization of a single patient is anecdotal, it does provide evidence that comparable doses of plasmid DNA as used in rodent studies, not dosed per weight, can be immunologically effective in human immunization. A similar observation that higher plasmid DNA doses are not necessarily immunologically superior to lower doses was recently reported in a phase 2 DNA vaccine trial for patients with multiple sclerosis.25 Our results suggest that ongoing boosters may have led to decreased antigen-specific proliferative responses, however, this again remains anecdotal. In any case, our findings demonstrate that antigen-specific T-cell immune responses, having been elicited earlier, can be augmented with a DNA vaccine, although optimal schedules of vaccination may need to be evaluated. The specific finding that this DNA vaccine encoding PAP can augment PAP-specific T-cell immunity may be of particular relevance given that a vaccine being evaluated by Dendreon Corporation, sipuleucel-T, is targeting the same PAP antigen and has demonstrated clinical benefit in terms of improving patients’ overall survival.7,26 An “off-the-shelf ” DNA vaccine targeting the PAP antigen could potentially be investigated in combination with that vaccine to maintain or augment long-term PAP-specific T-cell immunity.
This study was supported by NIH (K23 RR16489), by the NIH National Gene Vector Laboratory Program, by the NIH NCRR Clinical and Translational Science Award (CTSA) program (1UL1RR025011), and by the US Army Medical Research and Materiel Command Prostate Cancer Research Program (W81XWH-05-1-0404).
All authors have declared that there are no financial conflicts of interest in regards to this study.