Fax: (210) 916-6658
Use of booster inoculations to sustain the clinical effect of an adjuvant breast cancer vaccine†
From US Military Cancer Institute Clinical Trials Group Study I-01 and I-02
Article first published online: 15 SEP 2010
Published 2010 American Cancer Society
Volume 117, Issue 3, pages 463–471, 1 February 2011
How to Cite
Holmes, J. P., Clifton, G. T., Patil, R., Benavides, L. C., Gates, J. D., Stojadinovic, A., Mittendorf, E. A., Ponniah, S. and Peoples, G. E. (2011), Use of booster inoculations to sustain the clinical effect of an adjuvant breast cancer vaccine. Cancer, 117: 463–471. doi: 10.1002/cncr.25586
This article is a U.S. Government work and, as such, is in the public domain in the United States of America.
- Issue published online: 20 JAN 2011
- Article first published online: 15 SEP 2010
- Manuscript Accepted: 19 JUL 2010
- Manuscript Revised: 12 JUL 2010
- Manuscript Received: 28 APR 2010
- breast cancer;
- HER-2/neu peptide vaccine;
The authors are conducting clinical trials of the HER-2/neu E75-peptide vaccine in clinically disease-free breast cancer (BC) patients. Their phase 1-2 trials revealed that the E75 + granulocyte-macrophage colony-stimulating factor (GM-CSF) vaccine is safe and effective in stimulating clonal expansion of E75-specific CD8+ T cells. They assessed the need for and response to a booster after completion of primary vaccination series.
BC patients enrolled in the E75 vaccine trials who were ≥6 months from completion of their primary vaccination series were offered boosters with E75 + GM-CSF. Patients were monitored for toxicity. E75-specific CD8+ T cells were quantified using the human leukocyte antigen-A2:immunoglobulin G dimer before and after boosting.
Fifty-three patients received the vaccine booster. Median time from primary vaccination series was 9 months (range, 6-35 months), and median residual E75-specific immunity was 0.70% (range, 0-3.49%) CD8+ lymphocytes. Elevated residual immunity (ERI) (CD8+ E75-specific T cells >0.5%) was seen in 94.4% of patients at 6 months from primary vaccination series versus 48% of patients at >6 months (P = .002). The booster was well tolerated, with only grade 1 and 2 toxicity observed. Local reactions were more robust in patients receiving the booster at 6 months from primary vaccination series compared with those at >6 months (99.4 ± 6.1 mm vs 81.8 ± 4.1 mm, P = .01). In patients lacking ERI, 85% had increased ERI after vaccination (P = .0014).
The HER-2/neu E75 peptide vaccine E75 stimulates specific immunity in disease-free BC patients. However, immunity wanes with time. A vaccine booster is safe and effective in stimulating E75-specific immunity in those patients without ERI. These results suggest that the booster may be most effective at 6 months after completion of the primary vaccination series. Cancer 2011. Published 2010 by the American Cancer Society.
Peptide-based vaccines are being used in clinical trials either to treat1, 2 or to prevent recurrence of malignancy.3, 4 Peptide-based vaccination techniques involve inoculating patients with immunogenic epitopes from tumor-associated antigens, typically given with an immunoadjuvant, to stimulate proliferation of peptide-specific lymphocytes. Peptide-specific lymphocytes then identify and eliminate tumor cells presenting the vaccine-targeted epitope. Peptide-based vaccines are attractive immunotherapeutic options because they have no malignant potential, have low toxicity profiles, are simple and inexpensive, are easily monitored and studied, and eventually will be easily exportable to the community.
The individual peptides used in cancer vaccine preparations are often major histocompatibility complex (MHC) class-restricted and human leukocyte antigen (HLA) type-specific, thus stimulating either CD4+ or CD8+ lymphocytes and sometimes limiting the applicability of peptides to patients with the correct HLA type. Peptide vaccination techniques have been tried with HLA class I-binding peptides (stimulating CD8+ lymphocyte activation and proliferation),2, 5, 6 HLA class II-binding peptides (stimulating CD4+ lymphocytes),7, 8 or a combination of both HLA class I and II-directed peptides.9, 10 Directly stimulating CD8+ lymphocytes with HLA class I peptides makes intuitive sense, because class I molecules are present on most nonhematopoietic tumor cells, allowing CD8+ cytotoxic lymphocytes to specifically identify and lyse these cells. However, it is becoming increasingly clear that CD4+ lymphocyte stimulation plays a vital role in mounting an effective antitumor immune response and maintaining long-term immune memory.11 Although there is evidence that memory CD8+ lymphocytes can be stimulated by a MHC class I peptide given together with granulocyte-macrophage colony-stimulating factor (GM-CSF), it is not clear if this alone is sufficient to generate an effective antitumor response or whether, as some advocate, concomitant direct stimulation with MHC class II peptides is necessary.12, 13
Regardless of whether MHC class I-directed peptides, MHC class II-directed peptides, or both are used in peptide vaccinations, the peptide-specific immunity must be maintained at sufficient levels over an adequate duration to accomplish the clinical objective. In patients with large-volume metastatic cancer, vaccines must induce sufficient immunity to overcome the tolerance/suppression inherent to the large tumor, a feat that has so far proved unachievable with peptides and current immunoadjuvants alone. For vaccines used in the adjuvant setting for patients who are clinically free of cancer but at high risk of recurrence, immunity must be maintained at levels to identify and destroy or suppress small groups of tumor cells before they can become established, clinically identifiable cancer recurrences. The level of immunity necessary to achieve this objective and the length of time that this immunity should be maintained is not yet clear and may be unique for each cancer and peptide used. However, it seems likely that for adjuvant cancer vaccination to be successful, immunity should be maintained at some threshold level for at least the post-treatment period, when the risk of recurrence is highest.
We have completed enrollment of a phase 1-2 trial with the immunogenic peptide E75 (HER-2/neu, 369-377, KIFGSLAFL), which is HLA-A2/A3 restricted. Administration of E75 intradermally stimulates antigen-specific cytotoxic T lymphocytes (CTLs) in both animals and humans. These CTLs lyse HLA-A2+, HER-2/neu-expressing breast cancer cells in culture, suggesting a role in the treatment of breast cancer.14, 15 We have treated lymph node-negative and lymph node-positive breast cancer patients who are rendered surgically disease-free and have completed standard adjuvant chemoradiation therapy but are at high risk for recurrence. Our initial results have showed that E75 + GM-CSF is well tolerated, induces antigen-specific immunity, and may prevent clinical recurrences.16
The initial results indicate that, in addition to the primary vaccination series, a booster may be necessary to maintain significant residual immunity and probably to maintain the clinical effectiveness of the vaccine. We have previously demonstrated that this vaccine induces activation of CD8+, with specific enhancement of peptide-specific CD8+ lymphocytes shortly after initial vaccination in these breast cancer patients.12 However, we also made the observation that peptide-specific CTLs tend to decline over time after completion of the primary vaccination series. More importantly, the combined results of our lymph node-positive and lymph node-negative trials initially showed a significant reduction in breast cancer recurrences (5.6% vs 14.2%) at 20 months median follow-up (P = .04).16 Unfortunately, late recurrences in the vaccination arms of the trials corresponding with observed decreases in peptide-specific CTL levels diminished the difference in recurrence rates.
Given the waning immunity and late recurrences, we have investigated the feasibility of administering a vaccine booster to patients after completion of the primary vaccination series. In this study, we describe the clinical and immunological response to vaccine booster in patients previously vaccinated with E75. The goals of the study were to further evaluate the duration of residual immunity in previously vaccinated patients, to determine the safety of, and monitor immunologic response to, a booster vaccine, and to determine efficacy and optimal timing of booster inoculations.
MATERIALS AND METHODS
The lymph node-positive and lymph node-negative trials were approved at the local institutional review boards and conducted at Walter Reed Army Medical Center, Washington, DC and the Joyce Murtha Breast Care Center, Windber, Pennsylvania. These clinical trials are being conducted under an investigational new drug application (BB-IND #9187) approved by the US Food and Drug Administration. Details have been provided previously.3, 16 Briefly, all patients had histologically confirmed breast cancer, had completed standard therapy, and were disease-free and immunocompetent at the time of initial enrollment. HLA-A2+ and A3+ patients were vaccinated with various doses of E75 and GM-CSF and on various schedules over a 6-month period. Patients were offered an optional booster dose of E75 1000 μg + GM-CSF 250 μg if they were ≥6 months from completion of their primary vaccination series.
The E75 peptide was commercially produced in good manufacturing practices grade by NeoMPS (San Diego, Calif). Peptide purity was verified by high-performance liquid chromatography and mass spectrometry, and the amino acid content was determined by amino acid analysis. The peptide was purified to >95%. Sterility and general safety testing was carried out by the manufacturer. Lyophilized peptide was reconstituted in sterile saline at a concentration of 1000 μg in 0.5 mL. The peptide was mixed with GM-CSF (Berlex, Seattle, Wash) in 0.5 mL, and the 1.0 mL inoculation was split and given intradermally at 2 sites 5 cm apart. Booster vaccination was given in the same extremity as the primary series.
Patients were observed 1 hour postvaccination for immediate hypersensitivity reactions and returned at 48 to 72 hours for assessment of the injection site and to be questioned about toxicities. Toxicities were graded by the National Cancer Institute's Common Toxicity Criteria for Adverse Events, v3.0 and reported on a scale from 0 to 5. Patients who had previously had significant (grade 2 or 3) local or systemic toxicity received a reduced dose of GM-CSF at 125 μg.
Peripheral blood mononuclear cell isolation and cultures
Blood was drawn in Vacutainer CPT tubes (Becton Dickinson, Mountain View, Calif) before booster vaccination and 3 to 4 weeks after booster to isolate peripheral blood mononuclear cells (PBMCs) as previously described and used as a source of lymphocytes.3
HLA-A2: immunoglobulin dimer assay
The presence of CD8+ E75-specific cells in freshly isolated PBMCs from patients was directly assessed by using the dimer assay as previously described.17 Briefly, the HLA-A2:immunoglobulin (Ig) dimer (Pharmingen, San Diego, Calif) was loaded with the E75 or control peptide (folate binding protein peptide-E37 [25-33] RIAWARTEL) by incubating 1 μg of dimer with an excess (5 μg) of peptide and 0.5 μg of β2-microglobulin (Sigma Chemical Co, St Louis, Mo) at 37°C overnight, then storing it at 4°C until used. PBMCs were washed and resuspended in Pharmingen Stain Buffer and were added at 5 × 105 cells/100 μL/tube in 5 mL round-bottom polystyrene tubes (Becton Dickinson) and stained with the loaded dimers and antibodies as described previously. In each HLA-A2+ patient, the level of CD8+ E75-specific cells was determined before and after the vaccine booster. Because the dimer assay can currently only be run in HLA-A2+ patients, HLA-A3+ patients (n = 10) were excluded from the dimer evaluation.
Interferon-γ enzyme-linked immunospot assay
Interferon (IFN)-γ-producing cells were detected using an ex vivo enzyme-linked immunospot (ELISPOT) assay. PBMCs were washed and resuspended in culture medium consisting of RPMI with 5% heat-inactivated fetal calf serum (CM-FCS) or 5% heat-inactivated human AB serum (CM-AB) and penicillin, streptomycin, and L-glutamine. The PBMCs were incubated with 1 mL of CM-FCS or 1 mL of CM-AB, both containing 50 μg/mL interleukin (IL)-7 and 10 μg/mL IL-12 for 2 hours at 37°C. The cells were then washed by centrifuging and resuspended with CM-FCS or CM-AB (without cytokines) and plated into a 96-well round-bottomed plate at a concentration of 5 × 105 cells/100 μL/well in the presence or absence of E75 peptide at 50 μg/well. The plate was then incubated overnight at 37°C. At the end of the incubation, the plate was centrifuged to pellet the cells. After centrifugation, the supernatant was discarded, and the cells were resuspended in the appropriate culture medium (CM-FCS or CM-AB) and transferred to duplicate wells of a 96-well IFN-γ ELISPOT plate at 2.5 × 105 cells/200 μL/well (IFN-Pro Kit, Mabtech, Cincinnati, Ohio). The plate was incubated for 2 hours at 37°C and then washed and developed as per the manufacturer's instructions. Spots were analyzed and counted using the ImmunoSpot S4 Pro Analyzer and ImmunoSpot software.
Local reactions were measured as an in vivo functional assessment of the clinical immune response. Local reactions were measured 48 to 72 hours after vaccination, measured in 2 directions, and reported as an orthogonal mean ± standard error using the sensitive ball point method.18 Local reactions were compared with the patient's own previous local reactions to assess response to booster.
HLA:IgG dimer and ELISPOT values are reported as medians, and P values were calculated using the Wilcoxon test. For proportional comparisons, Fisher exact test was used. Comparison of local reactions was made with paired or unpaired Student t test as appropriate. Correlation of local reaction to time of boosting was evaluated with analysis of variance.
Fifty-three patients received a booster vaccination. Overall, about half (47.2%) of patients were lymph node positive. The median time from prior vaccination was 9 months (range, 6-35 months). Patients were evaluated as either early booster patients (n = 22) if they received the booster at 6 months after completion of the primary series or late booster patients (n = 31) if they were >6 months from the primary series (median, 17 months; range, 7-35 months). The groups were demographically similar except that early booster patients were significantly more likely to be estrogen receptor/progesterone receptor negative (Table 1). In general, early booster patients were enrolled later in the trial and thus received higher total doses of peptide (5.6 mg vs 3.8 mg) and GM-CSF (1.5 mg vs 1.2 mg) in the dose escalation protocol than the late booster group (Table 1).
|Characteristic||Total||EB (n=22)||LB (n=31)||P|
|Median age, y||58||60||57||.64|
|Lymph node positive||47.2%||50.0%||48.3%||.86|
|Mean peptide dose, μg||4566||5591||3839||.01a|
|Mean GM-CSF dose, μg||1340||1500||1226||.01a|
|Median No. of inoculations||6||6||6||NA|
The booster dose was very well tolerated, with primarily grade 1 local (a desired effect) and systemic toxicity. Almost half (46%) of the patients had no systemic complaints. There were no grade 3 or 4 toxicities. There were no significant differences in toxicity levels between the early booster and late booster groups (Fig. 1). Two patients experienced grade 2 local toxicity (inflammation), and 1 patient experienced grade 2 systemic toxicity (headache). The most common systemic toxicities were fatigue (18%), headache (16%), and malaise (11%). Only 2 (approximately 4%) of 53 patients had a higher grade local toxicity, and approximately 4% (2 of 53) had a higher grade systemic toxicity during the booster than during the primary vaccination series.
In vivo (local reactions)
local reactions were measured as an in vivo clinical immunologic response. Patients who received the booster 6 months from their primary vaccination series had significantly larger local reactions than patients >6 months from their primary vaccination series (98.5 ± 5.4 mm vs 76 ± 4.7 mm; P = .005). There was no difference in the 2 groups when comparing the last local reactions of the primary vaccination series in early booster and late booster groups (80.5 ± 6.1 mm vs 81.5 ± 4.1 mm; P = .93) (Fig. 2). Furthermore, linear regression of the booster local reactions reveals a moderate correlation (R2 = 0.326) with the time of boosting. The slope of trend line indicates that the local reaction decreases by approximately 1.5 mm per month for each month the booster is delayed after 6 months (P < .001; Fig. 3).
Ex vivo (dimer)
Residual E75-specific immunity
Residual E75-specific immunity measured by the HLA-A2:IgG dimer assay declined over time from the primary vaccination series. The residual immunity was measured 6 months after the completion of the primary vaccination series and before and after booster inoculations in all HLA-A2 patients (n = 42). The median E75-specific dimer level 6 months after completion of the initial vaccine series was 0.95% (range, 0.0%-3.24%) for all patients. There was a difference in the 6-month postprimary vaccination series residual immunity in those boosted at 6 months (early booster) and those boosted later (late booster): 1.10% (range, 0.29%-3.43%) versus 0.7% (range, 0.0%-3.24%) (P = .03) for the early booster and late booster groups, respectively. Although this indicates a difference in the comparison groups largely because of the difference in vaccine doses given (total doses of peptide early booster = 5.6 mg vs late booster = 3.8 mg), the immunity for the late booster group continued to decline to a median of 0.47% (range, 0-2.67%) at the time the booster was given (Fig. 4A). In addition, the number of patients with elevated residual immunity (defined as antigen-specific CD8+ T cells ≥0.5%) was significantly different, at 93.5% (15 of 16) in the early booster patients compared with 48% (11 of 23) of late booster patients at a median of 6 months and 17 months after completion of the primary vaccination series, respectively (P = .02).
Vaccine-induced E75-specific immunity
Antigen-specific CD8+ T cells were quantified before and 3 to 4 weeks after booster vaccination. Overall, the median residual immunity increased but not significantly from before to after boosting: 0.71% (range, 0-3.49%) to 0.99% (range, 0-3.23%) (P = .91). However, among those patients lacking elevated residual immunity at the time of boosting (n = 14), there was a significant increase in E75-specific CD8+ T cells from 0.3% (range, 0-0.49%) to 0.99% (range, 0-1.58%; P = .004) (Fig. 4B). Only 2 (14.3%) of 14 remained without elevated residual immunity after boosting. Patients with pre-existing immunity (n = 28) showed no change in their median pre- to maximum postdimer level (1.14% vs 1.11%, P = .27) (Fig. 4B).
Vaccine-induced E75-specific IFN-γ secretion
A majority of patients had IFN-γ-producing cells detected (quantified as spot-forming cells by ELISPOT) before and after booster vaccination in response to E75, GP2 (an immunogenic epitope from the transmembrane portion of HER-2/neu), or influenza antigen. There were 45 paired samples for E75, 30 pairs for GP2, and 43 pairs for influenza antigen. Overall, mean IFN-γ-producing cells measured in AB serum increased from 28 ± 15 to 63 ± 30 spots/106 cells in response to E75. The amount increased in both the early booster and late booster groups from before to after boosting. Overall, mean IFN-γ-producing cells measured in AB serum increased from 44 ± 21 to 129 ± 112 spots/106 cells in response to GP2. Mean IFN-γ-producing cells increased in both the early booster and late booster groups in response to GP2. Mean IFN-γ-producing cells measured in AB serum increased from 53 ± 37 to 59 ± 30 spots/106 cells overall, increasing in the early booster group and decreasing in the late booster group in response to flu antigen. The differences between the number of IFN-γ-producing spots/106 cells between the early booster and late booster groups and from before to after vaccination were not statistically significant in response to any of the peptides under any of the assay conditions tested (Fig. 5).
In this study, we show that breast cancer patients previously vaccinated with the E75 vaccine demonstrate declining immunity over time and that a single vaccine booster can safely and effectively increase immunity in these patients. A single booster vaccination of E75 peptide with GM-CSF immunoadjuvant is as safe and well tolerated as primary vaccination, with little systemic toxicity. The booster inoculation caused toxicity greater than previously experienced in only 7.5% (4 of 53) of patients. Importantly, early boosting (at 6 months) did not increase the local or systemic toxicity to the vaccine compared with those boosted later. This time point appears to be the optimal time to boost based on the waning immunity and the responsiveness as a function of time.
The observed decrease in E75-specific CD8+ lymphocytes after completion of the primary series is not surprising. After an antigenic challenge, levels of antigen-specific lymphocytes initially increase dramatically during the expansion phase of the immune response, only to drop substantially in a short period of time during the death phase. The resulting memory phase is composed primarily of antigen-specific central and effector memory CD8+ lymphocytes that persist for a variable amount of time depending on the vaccine dosing as well as the nature of the antigen. Although it is not clear whether memory CD8+ lymphocytes can persist without re-exposure to antigen,19-25 levels of antigen-specific CD8+ lymphocytes have been observed to decline in this and other peptide vaccine studies.26-28
At present, it is not known what threshold level of antitumor CD8+ lymphocytes, if one exists, is needed to prevent recurrence. In certain infectious diseases, there are antibody titers and even virus-specific CD8+ levels that are associated with protective immunity.25 However, similar data do not exist for cancer patients. It is doubtful that such data will be obtained using current immunologic assays, because these assays have not correlated well with clinical response in cancer vaccine trials to date. The lack of association has primarily been observed in trials treating advanced stage cancer where vaccination produced marked clonal expansion of antigen-specific lymphocytes but resulted in only modest and inconsistent clinical benefit as defined by tumor regression.29 This may be because of the multiple immunosuppressive mechanisms inherent to large tumor burden, which include lymphocyte dysfunction secondary to chronic antigen exposure.30 However, correlating clinical response to immunologic assays could also be problematic in the adjuvant setting, because it is not possible to distinguish patients who are truly deriving a clinical benefit because of vaccination from those who were cured with surgery, adjuvant chemotherapy, and radiation therapy.
Although ex vivo immunologic data have not correlated well with clinical benefit in clinical trials, they may in the end be a useful surrogate marker in adjuvant cancer vaccine trials. On the basis of dimer immunologic data alone, boosting patients who already have elevated residual immunity (>0.5% E75-specific CD8+ lymphocytes) did not increase peptide-specific CD8+ lymphocyte levels. These data support the practice of monitoring peptide-specific CD8+ lymphocyte levels and boosting if/when the patient's level drops below the 0.5% threshold. However, an important limitation of the dimer assay is that it does not reveal anything about the functionality of the lymphocytes that are counted. In contrast, the ELISPOT and in vivo local reaction data show increased response in patients boosted early independent of percentage E75-specific CD8+ lymphocyte level. These data argue for beginning booster inoculations at 6 months after completion of the primary series in all patients.
In our experience, in vivo response measured with delayed-type hypersensitivity (DTH) and local reaction has been the most reproducible and reliable predictor of immune response to vaccination. Accordingly, we have dosed the vaccine to produce large local reactions and then reduced the dose as necessary to prevent toxicity. In phase 1-2 trials with 3 different peptide vaccines, our group has observed that patients who required dose reduction because of large local reactions had significantly fewer recurrences than unvaccinated patients, suggesting a correlation between large local reaction and clinical response to vaccination.31 Patients boosted at 6 months had significantly greater local reactions to the booster than those boosted later, indicating greater immune response to early boosting, presumably because of greater numbers or function of memory lymphocytes to vaccination. However, a limitation of local reactions is that they are generated in response to both the peptide and GM-CSF immunoadjuvant, unlike DTH measurements in response to small doses (100 μg) of the peptide alone performed prevaccination and at the completion of the primary series.
In the current study, comparing the patients who were boosted early to patients who received boosters later is an imperfect comparison. The early booster patients were enrolled later in the dose-escalating trial, and therefore received greater doses of peptide and GM-CSF compared with the late booster group. This difference is especially relevant given that the magnitude and quality of long-term specific-CD8+ memory lymphocyte response could be effected by the amount of antigen and the size of the initial clonal burst generated by immunization.32, 33 However, despite the differences in primary series doses, we believe this is still a useful comparison, especially given that early booster and late booster patients had similar local reactions at completion of their primary vaccination series, indicating similar levels of in vivo immune response.
At present, it is difficult to determine the appropriate frequency of initial and subsequent boosting for cancer vaccines in the adjuvant setting without a reliable surrogate marker for an effective immune response. Existing immune assays may prove more reliable in the disease-free adjuvant setting than they have in therapeutic cancer vaccines. On the basis of immunologic assays, and more importantly, delayed recurrences seen in our experience with the primary vaccination series, re-exposure to the vaccine antigens through boosting may be necessary to maintain effective immunity to prevent recurrences in the long term. Although it appears that boosting at 6 months is advantageous compared with a less frequent schedule, boosting too frequently could lead to loss of functional and proliferative capability in memory CD8+ lymphocytes associated with chronic antigen stimulation or to activation-induced cell death.21, 34, 35
Although many variables remain, our experience with a single peptide adjuvant cancer vaccine demonstrates important immunologic and clinical data for dosing and maintaining peptide-specific immunity in isolation or in combination with other peptides. The ideal dosing and boosting strategy for adjuvant cancer vaccines may vary widely depending on the vaccination technique, immunogens selected, cancer being treated, and host. Indeed, as the variable responses seen in our experience demonstrate, the dosing and boosting may even need to vary for individual patients based on observed or measured immune responses. Ultimately, the development of adjuvant cancer vaccines will be improved though experience with individual vaccines by correlating immunologic data with clinical response.
CONFLICT OF INTEREST DISCLOSURES
Supported by the United States Military Cancer Institute, Department of Surgery, Uniformed Services University of the Health Sciences, and the Department of Clinical Investigation, Walter Reed Army Medical Center. Dr. Peoples has inventor rights. The E75 vaccine is licensed to Apthera, Inc.
- 2Clinical and immunologic results of a randomized phase II trial of vaccination using 4 melanoma peptides either administered in granulocyte-macrophage colony-stimulating factor in adjuvant or pulsed on dendritic cells. J Clin Oncol. 2003; 21: 4016-4026., , , et al.