SEARCH

SEARCH BY CITATION

Keywords:

  • human;
  • tumor immunity;
  • viral;
  • vaccination

Abstract

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES

BACKGROUND

The route of administration and extent of helper T-cell activation are factors that are likely to be important for the development of effective cancer vaccines. In order to optimize CD8+ cytotoxic T-lymphocyte (CTL) responses, the immunologic effects of direct lymph node (LN) injections of canary pox virus (ALVAC) vectors (expressing the melanoma antigen, gp100) and immunogenic gp100 peptides, along with concomitant injections of the helper adjuvant, tetanus toxoid, were studied in high-risk HLA-A*0201+ patients.

METHODS

Forty-two patients were vaccinated using six different protocols. Twenty-three patients were ‘primed’ with ALVAC(2)-gp100m and ‘boosted’ with gp100 peptides, either subcutaneously or into an LN. Intranodal (IN) peptides, alone, were administered to six patients. Thirteen patients were given tetanus toxoid initially, and with each gp100 vaccination. Toxicity was recorded and immunologic responses were determined in 35 patients by enzyme-linked immunospot (ELISPOT) and gp100-tetramer binding assays and anti-ALVAC(2) enzyme-linked immunosorbent assays (ELISAs).

RESULTS

All vaccine protocols were tolerated well. Using stringent criteria for immunologic response, 8 of 18 patients responded to the viral vaccines, in striking contrast to peptides only (0 of 6 patients) or with help in trans from tetanus-reactive T-cells (1 of 11 patients). Changes in gp100-reactive CTL frequencies and ALVAC antibodies were greatest when viruses were injected directly into LNs.

CONCLUSIONS

IN injections of ALVAC(2)-gp100m viruses are feasible, safe, and may be a superior method of vaccination in humans. CTL responses to this vaccine were not enhanced by tetanus toxoid. Cancer 2006. © 2006 American Cancer Society.

Effective cancer vaccines need to increase the frequency and activity of tumor-specific cytotoxic T-lymphocytes (CTLs) for prolonged periods of time.1 In this study, we evaluated the route of administration, along with a strategy to increase T-helper cells, on the immunologic efficacy of a virus-based melanoma vaccine.

Defined tumor antigens2 (such as the melanoma antigen, gp1003) form the basis of a number of different vaccine strategies.4–6 Viruses (notably Adenoviruses7 and Poxviruses6), engineered to express tumor antigens, can potentially induce type 1 immune responses that are required for clearing tumor cells.1 Avian poxviruses, such as ALVAC, provide safety advantages over other poxviruses because they do not live long enough in mammalian cells to produce infectious progeny.8

ALVAC is an attenuated version of the canary pox virus Rentschler strain.8 ALVAC(2) is a second-generation virus derived from plaque-purified ALVAC. ALVAC-based recombinants have been found to be safe and immunogenic in several clinical situations.9–11 Repeated administration of ALVAC viruses, that expressed three human immunodeficiency virus genes, induced humoral and cellular responses that tended to plateau after the third injection.12 Subsequently, a ‘prime-boost’ strategy of two injections of ALVAC recombinants followed by two injections of recombinant protein and adjuvant gave the highest antibody and CTL responses.

Whereas the ‘prime-boost’ approach suggests a basis for using ALVAC recombinants, the optimal route of immunization in humans has not yet been defined. Viruses generally replicate at sites of infection and are taken up by dendritic cells (DCs) that induce immune responses after trafficking to draining lymph nodes (LNs). Localization of antigen in an LN is critical for the induction of an immune response. Indeed, it has been shown in mice that peptides alone, injected continuously into an LN, can induce an effective CTL response.13 Because ALVAC poxviruses do not survive for long in mammalian cells, they may be degraded too quickly (whereas DCs traffic to an LN) to provide sufficient antigenic loads to support a strong immune response. Consequently, immune responses may be enhanced by injecting ALVAC viruses directly into LNs.

Along with the route of antigen administration, helper CD4+ T-cells are also important for effective CTL responses.14 Unlike Class I human leukocyte antigen (HLA)-binding tumor antigens that stimulate CD8+ T-cells, specific helper epitopes are less well defined. However, help can be provided potentially in trans by CD4+ T-cells that make cytokines to support CTL proliferation and differentiation.15

In this trial, we compared the effects on antiviral and antitumor immunity of subcutaneous (s.c.) or intranodal (IN) injections of ALVAC(2)-gp100m recombinant viruses (which express a full-length gp100 protein containing two epitopes modified for enhanced binding to HLA-A*0201 (gp100m)16) and the modified epitopes. We also studied the effects of priming CD4+ responses with tetanus toxoid (to provide help in trans) on immune responses to ALVAC(2) and gp100.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES

Reagents

HLA-A*0201-binding influenza matrix protein (FLU) epitopes and gp100 epitopes modified to increase Class I major histocompatability complex (MHC) binding (gp100:209-2M and gp100:280-9V) were synthesized by Sanofi-Pasteur (Toronto, Canada) as described previously.17 CD8-FITC antibodies were from PharMingen (San Francisco, CA). Phycoerythrin-labeled tetramers of HLA-A*0201 complexed to gp100:209-2M, gp100:280-9V, or FLU were from ProImmune (Oxford, UK).17 Tetanus toxoid (nonadsorbed) concentrates at 100 or 500 Lf/mL were from Sanofi Pasteur.

HLA Typing

HLA-A2+ patients were identified as before.17

Patients

Between April 2000, and August 2002, 42 HLA-A*0201+ melanoma patients (Table 1) were recruited sequentially into one of six groups (Fig. 1). Eligibility criteria included a confirmed diagnosis of melanoma with a primary lesion greater than 1.5 mm deep, involved lymph nodes that had been resected surgically, or metastatic disease. Metastases greater than 5 cm wide had to be ‘stable,’ with less than 10% increase in size, based on radiologic assessments over 2 months. In addition, patients were required to be HLA-A*0201+, older than 18, European Cooperative Oncology Group (ECOG) performance status of 0 or 1, survive longer than 3 months, and give informed, written consent according to national and institutional guidelines. Patients with known central nervous system metastases were excluded.

Table 1. Patient Characteristics
 Group 1 (n = 6)Group 2 (n = 6)Group 3 (n = 8)Group 4 (n = 9)Group 5 (n = 7)Group 6 (n = 6)
  • a

    Stage I is localized primary lesion with thickness < 2 mm. Stage II is localized primary lesion between 2 and 4 mm in depth. Stage III is LN (lymph node; but not distant organ involvement) or in-transit/satellite metastases. Stage IV is involvement of distant skin, nodal metastases, and/or organ involvement.29 Note that Patient 11 in Group 2 was classified as Stage IV but had a solitary lung metastasis resected and was NED (no evaluable disease) before beginning vaccinations. All patients with Stage III or lower disease were NED at the time of study.

  • b

    European Cooperative Oncology Group (ECOG) 0 is fully active, able to carry on all predisease performance without restriction. ECOG 1 is restricted in physically strenuous activity but ambulatory and able to carry out light work.30

  • c

    Except for surgery to remove LNs.

Gender      
 Male445565
 Female223411
Age in yrs48.8 ± 12.456.2 ± 12.853.1 ± 15.555.1 ± 16.760.1 ± 10.561.5 ± 12.8
Clinical stagea      
 I000000
 II102130
 III422246
 IV144600
ECOGb      
 0668856
 1000120
Site of metastases      
 Liver 1  1 
 Lung 2 3  
 Nodes   31 
 Mediastinum 1    
 Skin11141 
Prior treatmentc      
 IFN521422
 Chemo001120
 Radiation230340
thumbnail image

Figure 1. Schematic diagram of trial design.

Download figure to PowerPoint

Study Design

Vaccines consisted of ALVAC(2)-gp100m recombinants (that expressed a full-length gp100 gene encoding two epitopes modified for enhanced HLA Class I binding16) and the two modified gp100 peptide epitopes. Four groups of six to nine HLA-A*0201+ melanoma patients were vaccinated by different routes (Fig. 1). Group 1 patients received s.c. ALVAC(2)-gp100m (0.5 mL, 0.5 × 107.09 cell culture infectious doses 50% [CCID50]) on Day 1 of a 21-day cycle for three cycles followed by 2 mL of a mixture of the modified gp100 peptides (500 μg/mL of each peptide) s.c. on Day 1 of a 21-day cycle for two cycles. Group 2 patients received 0.4 mL of the gp100 peptide mixture IN daily for 5 days every 21 days for two cycles. Group 3 patients received s.c. ALVAC(2)-gp100m on Day 1 of a 21-day cycle for three cycles followed by two cycles of IN peptides. Group 4 patients received IN ALVAC(2)-gp100m (0.5 mL) on Day 1 of a 21-day cycle for three cycles followed by two cycles of IN peptides. In Groups 5 and 6, patients received one injection of tetanus toxoid (to prime helper T-cells) 4 weeks before the ALVAC injections and were then immunized against gp100 using the Group 1 schedule. With each gp100 injection, either low doses (50 μL, 100 LF/mL) (Group 5) or high doses (50 μL, 500 LF/mL) (Group 6) of tetanus toxoid were also administered to provide exogenous ‘help’ for the anti-gp100 responses (Fig. 1).

The peptide dose of 1 mg per treatment cycle was based on results of other investigators.18, 19 The dose of 0.5 × 107.09 CCID50 per injection of ALVAC(2)-gp100m was based on previous studies with poxvirus vaccines.9, 10, 12

Vaccine Administration, Clinical Follow-up, and Response Assessment

Patients were treated as out-patients at the Toronto-Sunnybrook Regional Cancer Center (Groups 1, 2, 3, 4, and Patients 37 and 38 in Group 6), the Ottawa Regional Cancer Center (Group 5), and the Royal Victoria Hospital (remaining patients in Group 6). After each vaccination, patients were observed for adverse reactions. At each visit a physical examination was performed and blood was drawn to evaluate potential vaccine toxicities and for immunologic monitoring. Plasma and peripheral blood mononuclear cells (PBMCs) were separated in a central laboratory the next day and stored at −70 °C or in liquid nitrogen, respectively. Radiologic studies depended on the clinical situation. Clinical responses (complete, partial, or minor response, and stable or progressive disease) were evaluated by standard criteria.20

Vaccines were injected into unaffected inguinal nodes using ultrasound guidance and the images were used to confirm that the injections had been successful.

T-Cell Stimulation In Vitro

PBMCs were thawed and incubated overnight in AIM-V medium (Gibco, Burlington, Ontario). Cells (2–3 × 106/mL in AIM-V plus 5% AB serum; Sigma, St. Louis, MO) were plated with FLU or both gp100 peptides added at previously optimized final concentrations of 10 or 25 μg/mL, respectively. IL-2 (50 IU/mL) (Chiron, Emeryville, CA) was added 3 and 6 days later and cells were harvested after 8–9 days for enzyme-linked immunospot (ELISPOT) and tetramer assays.17

IFNγ ELISPOT Assays

ELISPOT assays were performed as described,17 with minor modifications. Capture and detection antibodies were from the 1-DIK and 7-B6-1 clones, respectively (MABTECH, Stockholm, Sweden). Cultured T-cells (5 × 105/mL or 1 × 105/mL) were reactivated in triplicate wells with FLU or both modified gp100 peptides (final concentrations of 10 and 25 μg/mL, respectively) and IL-2 (100 IU/mL).

ALVAC(2) Antibodies

Plasma levels of ALVAC(2) antibodies were determined in an indirect enzyme-linked immunosorbent assay (ELISA). Wells of a 96-well flat bottom plate (Maxisorp, Nalge Nunc International, Rochester, NY), coated with 100 μL of heat-inactivated ALVAC(2) antigen in 0.05 M carbonate-bicarbonate buffer (2.5 μg/mL), were washed extensively with 0.1% Tween 20 (Sigma) in phosphate-buffered saline (PBS) and blocked with 5% instant skim milk powder in PBS. After incubation with serial dilutions (100 μL) of thawed plasma or reference/standard serum, the wells were washed and 100 μL of peroxidase-conjugated rabbit antihuman IgG (1/20,000 dilution in 5% BSA) (Jackson Immunoresearch, Jackson Laboratories, Bar Harbor, ME) was added for 1.5 hours at room temperature. The plates were developed with 0.2% tetramethylbenzidine (TMB) and 0.004% hydrogen peroxide (1:9) (Sigma). Reactions were stopped after 10 minutes with 1 M sulfuric acid and read with a Spectra Max 190 ELISA reader (Molecular Devices, Sunnyvale, CA) using Soft Max Pro ELISA reader software (Test wavelength 450 nm; reference wavelength 540 nm). From the standard curve the concentrations of each specimen, and of internal quality controls, were calculated and the final values adjusted according to the dilution factor. The final titer of a sample was equal to the mean titer of at least two dilutions, satisfying the criteria that the adsorbances were within the limits of the standard curve and the variation coefficients of individual optical density values were lower than 20%.

Immunofluorescence

Staining was performed and analyzed at the end of the culture period as described.17

Statistical Analysis

All patients were included in the toxicity analysis. Patients had to receive all viral and/or peptide injections to be included in the immunologic response assessments. Because the study was primarily descriptive (and each arm essentially a small Phase I trial without dose escalations), samples were arbitrarily fixed at six subjects per group. The results from individual wells of the ELISPOT and ELISA assays were used to determine averages and tested for statistical significance using a paired t-test. Response in the ELISPOT assay required both a peak response greater than fourfold over baseline values, and more than 10 specific spots/105 cells that were greater than 3 standard deviations (SDs) above the average of the media control wells.

RESULTS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES

Patient Characteristics

Forty-two HLA-A*0201+ patients (Table 1) with intermediate and high-risk melanoma (based on deep primary lesions, LN involvement, or metastatic disease) were assigned sequentially to six different groups (Fig. 1). The effects of IN injections of ALVAC(2)-gp100m and the modified gp100 peptide epitopes on subsequent immunologic responses to gp100 and ALVAC(2) were studied in Groups 1–4. The role of ‘help’ from CD4+ T-cells (delivered in trans by tetanus toxoid injections) was examined in Groups 5 and 6.

Immune monitoring was performed in 35 patients. However, 13 of 42 patients did not complete the study, including the followup period. Twelve withdrew due to progressive disease, but none because of vaccine-associated toxicity. All IN injections of viruses and peptides were successful, as confirmed by ultrasound (not shown).

Toxicity

Vaccination was generally well tolerated. Injection site reactions occurred mainly with s.c. viral injections (Groups 1, 3, 5, 6). IN peptides (Groups 2, 3, 4) or viruses (Group 4) did not cause significant local discomfort. However, hematologic and biochemical abnormalities (mainly reversible anemia or increased liver function tests, respectively) were seen more frequently in the latter. Vitiligo (associated with antimelanoma immunity21) was not observed.

Assessment of Anti-gp100 Responses

Immunologic monitoring was performed before each injection and for 3 months after the injections (Fig. 1). IFNγ-ELISPOT assays were used to assess anti-gp100 responses because of the association of IFNγ production with effective antitumor T-cell immunity.1 As an example, gp100-specific responses in Patient 18 from Group 3 increased after the second ALVAC(2) injection (bleed 3), peaked approximately 6 weeks after the final ALVAC(2) injection (bleed 5), and then decreased despite subsequent IN peptide injections (Fig. 2A). Responses to control FLU peptides were generally constant (Fig. 2A, bottom panel). Frequencies of gp100-reactive CD8+ T-cells after vaccination were also measured by flow cytometry, with HLA*0201 tetramers bound to the gp100 peptides, after the in vitro culture period. Increased tetramer-binding T-cell numbers were also seen at bleed 5 (Fig. 2B).

thumbnail image

Figure 2. Example of an immunologic response to vaccination. (A) PBMCs from a Group 3 patient (Patient 18) were cryopreserved before each vaccination and monthly during followup (bleeding points 1–8). The samples were thawed simultaneously and stimulated with a gp100 peptide mixture (top panel) or FLU peptides, as a control that the culture conditions could reveal the presence of memory T-cells (bottom panel). As described in Materials and Methods, cells were reactivated after 8 days on ELISPOT plates with the gp00 (solid bars) or FLU peptides (hatched bars), or simply cultured without peptides (open bars). The average and standard error of the number of spots from three replicate wells are shown for each bleeding point. For this patient, the results indicate that at bleed 5 the number of spots after reactivation with gp100 peptides was much greater than the spots from cells alone, or stimulated with the FLU peptide (top panel). In the bottom panel, the results indicate that FLU-specific CTL frequencies were fairly constant throughout the course of vaccination. (B) Cells stimulated with gp100 peptides for 8 days were stained with CD8-FITC antibodies and the two phycoerythrin-labeled gp100 peptide tetramers. The percentages of CD8+tetramer+ cells (representing gp100-reactive T cells) were then determined by flow cytometry and indicated in the boxes in each dot-plot.

Download figure to PowerPoint

ELISPOT reactivity

Immunologic responses were defined as: more than a fourfold increase over baseline any time after vaccination, and more than 10 specific spots (i.e., difference between average spot number in samples stimulated with gp100 peptides and average number in samples cultured alone). Using these stringent criteria, 9 of 34 patients responded to the vaccinations, for an overall response rate of 26% (Table 2). However, no responses were observed in the absence of virus (Group 2), even though murine studies predicted that daily IN peptides would induce strong CTL responses.13

Table 2. Summary of Clinical and Immunologic Responses
GroupPatient no.ELISPOT (/105 cells)aanti-ALVAC (mEU/mL)Immune responseClinical response
BaselineMaxFinalbBaselineMaxFinalELISPOTcanti-ALVACd
  • a

    ELISPOT results are expressed as number of specific spots per 105 cells. Specific spots are average spots of gp100 samples–average spots of control (no gp100).

  • b

    If the number of specific spots was less than 3 SD (standard deviation of the negative control triplicates) or the number of specific spots was less than 10, a value of < 10 is listed.

  • c

    ELISPOT responders had a peak response ≥ 4-fold over baseline value (prevaccine) and ≥ 10 specific spots/105 cells. 1 is responder; 0 is nonresponder.

  • d

    Antibody responders had a peak ALVAC antibody concentration ≥ 10-fold over the initial concentration. 1 is responder; 0 is nonresponder.

  • e

    NED, no evaluable disease; PD, progressive disease; SD, stable disease

  • f

    Patient 103 from Group 5 did not complete treatment and was not evaluable.

  • g

    NS, cells not sufficient for analysis.

1253.2207.4< 10100.039477.12100.701NEDe
 3< 1063.3< 10476.919282.94001.011PD
 5< 1030.6< 10968.328080.52632.001NED
 6< 1039.0< 1098.13290.41316.301PD
 7< 1048.9< 10258.55786.81545.711PD (died)
 8< 1017.813.4224.96581.22695.301NED
Total       2/66/6 
211< 10< 10< 101572.61630.71630.700NED
 12< 10< 10< 101026.81374.41374.400NED
 13< 10< 10< 101308.11994.71351.500PD (died)
 15< 10< 10< 10958.7941.5684.200PD
 17< 10< 10< 10324.9422.6340.700NED
 9< 10< 10< 102212.32531.02467.100PD
Total       0/60/6 
318< 10450.0< 10468.716874.96455.011PD
 21< 1040.712.4138.16754.32041.211SD
 22< 10< 10< 102636.618161.13780.400NED
 23452.077.8< 10278.630769.014811.601PD
 31104.764.0< 10203.410974.51840.101NED
 32< 10182.7< 10467.06615.02245.611NED
Total       3/65/6 
424< 10< 10< 10174.710960823590.001PD
 30< 10429.350.0724.224408763527.811NED
 33< 10175.0< 10143.517065119453.311NED
 3473.395.3< 10374.052299.26541.801PD (died)
 35< 10< 10< 10595.319676.48332.801NED
 36< 101061< 10755.218492829953.311NED
Total       3/66/6 
5f101NSg77.3< 10310.95285.0654.81PD
 104< 10< 10< 1097.3740.450.600NED
 20220.097.3< 10118.4803.9449.210NED
 205< 10< 10< 10501.56662.4807.301PD
 210< 10< 10< 10275.32300.9532.600PD
Total       1/42/5 
6112< 10< 10< 10554.23117.71418.800NED
 116< 10< 10< 10439.012212.44304.401PD
 212< 10< 10< 10340.92402.8477.500NED
 215< 10< 10< 10368.7527.7363.500NED
 3725.027.0< 10248.22491.7638.201PD
 38NS< 10< 10NS1498.0795.1SD
Total       0/51/5 

Using a ‘prime-boost’ approach (where viral injections were followed by two peptide cycles), 8 of 18 patients responded to vaccination (Groups 1, 3, 4). Somewhat surprisingly, exogenous T-cell help from tetanus toxoid (which was predicted to enhance anti-gp100 responses15) appeared to suppress them (Groups 5, 6).

The magnitude of the responses appeared to be greater with IN virus injections (Group 4) and tended to occur earlier than with s.c. virus injections (Groups 1, 3) (Fig. 3A). However, the responses were not sustained, and increased gp100-reactive CTL numbers were never observed 3 months after completing vaccination (Fig. 3A).

thumbnail image

Figure 3. Time courses of anti-gp100 T cell reactivity. Cryopreserved PBMCs from the different bleeding points were thawed and stimulated with gp100 peptides. After 8 days the cells were analyzed by (A) IFNγ ELISPOT assays or (B) with gp100 tetramers. Specific gp100-reactive spots were calculated by subtracting the background number (obtained by culturing the cells on the ELISPOT plates without additional stimulation) from the number obtained by reactivating the cells with gp100 peptides and expressed per 105 cells. Unless the number of specific spots exceeded both a value of 10 and 3 SDs above the average of media control wells, the values were set to 0. Control cultures, stimulated by the FLU peptide recall antigen, were used to ensure the adequacy of the culture conditions in each case (not shown).

Download figure to PowerPoint

Tetramer binding

Tetramer-binding T-cell numbers correlated generally with the ELISPOT assay results (Fig. 3B). However (especially in Groups 5 and 6), increases in CD8+tetramer+ T-cells were sometimes noted and not seen in the ELISPOT assays.

Delayed type hypersensitivity (DTH) and LN enlargement

Despite in vitro evidence for vaccine-induced gp100-reactive T-cells, no positive DTH reactions to intradermal injections of gp100 peptides were observed. Similarly, persistent lymphadenopathy after IN viral or peptide injections was also not seen, by both direct physical examination and ultrasound.

ALVAC Antibodies

Whereas enhanced cell-mediated immunity is the desired effect of a cancer vaccine,1 antibodies are important for antiviral immunity.22 Because we could not purify enough recombinant gp100m for ELISAs, ALVAC antibody titers were used to evaluate the different vaccine protocols. As expected, Group 2 patients (who did not receive virus) did not increase their ALVAC antibody levels (Fig. 4; Table 2). Group 1 and 3 patients (who received s.c. virus) mounted anti-ALVAC responses that plateaued around 105 mEU/mL after the third viral injection (Fig. 1) and diminished by ∼1 log over the followup period. Remarkably, antibody levels in Group 4 patients (who received IN virus) were nearly an order of magnitude greater than in Group 1 and 3 patients. In contrast, exogenous help from tetanus toxoid (Groups 5 and 6) appeared to suppress ALVAC antibody production (Fig. 4), analogous to its effects in ELISPOT assays (Fig. 3).

thumbnail image

Figure 4. Time courses of ALVAC antibody titers. The concentration of ALVAC(2)-binding IgG antibodies at each bleeding point was determined by ELISA as described in Materials and Methods. The averages of the results from at least two different dilutions are reported for each time point.

Download figure to PowerPoint

Statistical Analysis of Immune Responses Between Groups

We compared anti-ALVAC and -gp100 responses between the different groups (Fig. 5). For statistical analysis, the maximal positive changes in gp100-reactive CTL frequencies, from baseline frequencies (for each patient at any time from Table 2 and Fig. 3) were log-transformed (Fig. 5A). Increases in gp100-reactive T-cells for Groups 1, 3, and 4 were statistically different from Group 2 (P < 0.01, 0.05, and 0.05, respectively) and Group 6 (P < 0.01, 0.05, and 0.05, respectively). The greatest average increase in gp100-reactive CTL frequencies was in Group 4.

thumbnail image

Figure 5. Comparison of immune responses to gp100 and ALVAC(2) induced by the different vaccination protocols. (A) For each patient the logarithm of the number of specific gp100 spots before vaccination was subtracted from the logarithm of the largest number of specific gp100 spots at any time point after vaccination (from Table 2) to determine the logarithm of the maximal change in gp100-reactive CTL responses after vaccination. For Patients 23 and 31 in Group 3, this number was a negative value and was set to zero for statistical analysis. The individual results are shown as dots and the numbers beside the small horizontal lines are the average for the patients in each group. (B) The maximal concentration of anti-ALVAC(2) antibodies, at any time after vaccination, was divided by the baseline concentration (Table 2) to determine the maximal change in antibody titers for each patient. The average and standard error of the results for each group are shown (closed bars). Differences between the groups were tested for statistical significance and P-values are shown on the graphs.

Download figure to PowerPoint

The ratio of the maximal ALVAC antibody titer for each patient at any time after vaccination, compared with baseline, was also used for statistical analysis. Antibody titers in Group 4 were found to be significantly higher than the other groups (Fig. 5B). Taken together, the results suggested that the largest increases in gp100-reactive CTL cells and anti-ALVAC antibodies were achieved by IN virus injections.

Clinical Responses

Whereas the primary aim of the trial was to evaluate the immunologic efficacy and toxicity of the different vaccine strategies, relatively long follow-up information was available for the 31 patients treated at the Toronto site. No objective responses were seen in patients treated only with the vaccines. Melanoma progression developed in 11 of 15 AJCC Stage IV patients and 2 of 12 Stage III patients with median times to progression of 8.7 ± 2.9 and 8.0 ± 2.8 months, respectively. The remaining Stage II and III patients are still free of disease, in some cases approaching 5 years after vaccination. Of the four Stage IV patients who have not progressed, Patient 11 had been rendered NED (no evaluable disease) after surgical removal of a lung metastasis before entering the study. The other three Stage IV patients (Patient 21 [subcutaneous mass], Patient 30 [skin, lung, and lymph node metastases], and Patient 33 [neck mass]) were enrolled in a previously reported study of high-dose IFNα2b after vaccination (that also included Patient 11, as well as Patients 8, 17, and 38 [with Stage III disease] and Patients 9 and 18 [with Stage IV disease]).17 Patients 21, 30, and 33 developed transient gp100-reactive T-cell responses to vaccination with no change, or progression, in the size of metastases. However, as reported previously,17 gp100-reactive CTL responses were recalled after high-dose IFNα2b, with disappearance of the metastases. These patients remain disease-free with an average followup of 51.5 ± 3.4 months.17

DISCUSSION

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES

In this study, we found that ALVAC(2)-gp100m viruses increased gp100-reactive CTL frequencies in 8 of 18 patients. The response rate was even higher if less stringent criteria for responsiveness (i.e., twofold increases over prevaccination levels23) are used.

Immunologic monitoring of the trial was based on IFNγ-ELISPOT assays, performed after 8 days of restimulation in vitro. Although a concern is that the culture period may introduce artifacts, we suggest that positive responses were caused by exposure to gp100 antigens in the vaccine (rather than being selected in culture), because 29 of 33 patients (Table 2) had no or low baseline reactivity to gp100 peptides. However, the responses seen were weaker, and less prolonged, than reported in other melanoma vaccine trials of peptides and adjuvants.18, 19, 23 For example, other investigators have been able to demonstrate increases in circulating tumor antigen-reactive T-cell frequencies after vaccination directly ex vivo, using flow cytometric tetramer-based assays,28 which we were unable to do. The reasons for the weaker responses in our study are not entirely clear but may be related to the viral vectors, lack of an additional adjuvant, or possibly methodologic differences.

Because cells required for the immunologic assays were limiting, both of the gp100 peptides were included in the priming cultures. Accordingly, it is not clear if the responses observed were specific to one of the peptides. In addition, cytotoxicity assays were not performed routinely, because most of the effector cells were needed for ELISPOT assays and autologous tumor cells were not generally available. However, we reported previously that gp100-reactive T-cells generated after vaccination in some patients were only weakly lytic to peptide-coated targets, but acquired strong cytotoxic function after the patients were treated with high-dose IFNα2b.17

Importantly, we found that IN injections of ALVAC(2)-gp100m viruses are feasible, safe, and are a superior method to induce antibody responses in humans, although larger patient numbers will be required to make more definite conclusions about CTLs. To our knowledge, this is the first study of direct viral injections into human LNs. Previously, IN or intralymphatic injections of DC,24 autologous tumor cell,25 and DNA vaccines26 were shown to improve CTL responses in cancer patients. Although localization of viral antigens in LNs (as a result of trafficking of antigen-bearing DCs from the periphery) is thought to be required for the initiation of effective immunity,13 it remains uncertain whether immunity to viruses that are more virulent than ALVAC (such as adenoviruses or other poxviruses) would also be enhanced by direct LN inoculation.

In mice, frequent administration of peptide antigens into the immunogenic LN environment (without additional adjuvants) can induce protective CTL responses.13 Our results suggest that this is not true for humans, because repeated IN peptide injections alone did not result in effective CTL or antibody responses (Figs. 3–5; Table 2).

A ‘prime-boost’ design (i.e., three viral injections followed by two courses of peptides) was used because it yielded enhanced CTL responses in other clinical trials.12 The role of the peptide ‘boost’ in this vaccine study is unclear. Although Patients 2 and 202 in Groups 1 and 5, respectively, appeared to develop increased anti-gp100 CTL responses later on in their vaccination course (i.e., at bleed 7) (Fig. 3), enhanced anti-gp100 CTL responses were generally related temporally to the viral injections and not enhanced further by the peptide boosts.

IN viral injections produced ALVAC antibody titers that were an order of magnitude greater than the other routes of immunization (Fig. 4). Some of these antibodies could have neutralizing capabilities,22 and limit responses to subsequent injections of ALVAC(2)-gp100m. We suggest that this was not the case, because gp100-reactive T-cell responses were still enhanced in the presence of high titers of ALVAC antibodies (Fig. 3). In addition, previous work has shown that neutralizing antibodies are not usually generated after vaccination with ALVAC (unpubl. obs.).

Concomitant administration of tetanus toxoid (Groups 5 and 6) led to inferior anti-gp100 CTL and ALVAC antibody responses (Figs. 3–5). The reasons for the failure to enhance vaccine responses are not clear. Perhaps the initial induction of antitetanus responses biased the LN milieu to favor TH2/TC2 responses at the time of injection of the ALVAC(2)-gp100m viruses. Competition between virus- and tetanus-reactive B-cells may also have suppressed ALVAC antibody production, as reported when different immune responses occur simultaneously.27 Regardless of the mechanism, the current study suggests that priming with tetanus toxoid does not enhance the effects of cancer vaccines that employ ALVAC vectors.

Despite small numbers and patient heterogeneity in the vaccine groups, and the use of restimulated assays, this study obtained important preliminary information that requires confirmation in future trials. First, ALVAC(2) can increase antitumor immunity without peptide boosting, suggesting that it can be used in patients regardless of HLA-haplotype (because peptide epitopes are only known for some haplotypes). Second, IN administration appears to be a simple, clinically relevant method to enhance the immunogenicity of ALVAC(2) vectors.

Acknowledgements

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES

The authors thank X. Sheng-Tanner, H. Zheng, H. Lima, and J. Sarrazin for excellent technical assistance and R. Uger and B. Barber for help in designing the trial.

REFERENCES

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES