Carmen Scheibenbogen

Quantitation of antigen-reactive T cells in peripheral blood by IFNγ-ELISPOT assay and chromium-release assay: a four-centre comparative trial

The ELISPOT assay is increasingly being used for the monitoring of the induction of antigen-reactive T cells in cancer vaccination trials. In order to evaluate the reliability of T cell frequency analysis with the ELISPOT assay, a comparative study was performed in four European laboratories. Six samples from healthy subjects were analyzed for the frequency of influenza-reactive CD8+ T cells in peripheral blood mononuclear cells (PBMC) by IFNgamma-ELISPOT assay. In addition, one laboratory determined cytotoxic T cell precursor (CTL) frequencies in these samples by limiting dilution chromium-release assay (LDA), and three laboratories performed a variant of the LDA, the multiple microculture assay (MMA). Consistent frequencies of influenza peptide-reactive T cells were obtained with the ELISPOT assay in all four laboratories. The numbers detected by ELISPOT assay correlated closely with those determined by LDA. In contrast, the frequencies obtained with the MMA differed considerably and showed little correlation with the other two assays. This study shows that it is possible to use the ELISPOT assay to determine with reliability antigen-reactive T cells in a multicenter setting. We suggest that this assay may be suitable for monitoring cancer vaccine trials.

Quantification and characterization of specific T-cells by antigen-specific cytokine production using ELISPOT assay or intracellular cytokine staining.

The novel generation of sensitive T-cell assays facilitates the direct quantitation and characterization of specific T-cell responses. Functional T-cell assays such as the ELISPOT assay and the intracellular cytokine cytometry (ICC) employ the antigen-specific induction of cytokines to detect specific T-cells on a single cell level. ICC has the advantage that the simultaneous phenotypic characterization of the antigen-specific T-cells is possible. There is evidence now from clinical cancer vaccination trials, that there is a relationship between the detection of vaccine-induced T-cells by cytokine-based assays and clinical responses. As these assays become increasingly relevant in clinical practice to suggest issues of assay validation and quality control become of major importance.

Application of the IFN-γ ELISPOT assay to quantify T cell responses against proteins

The ELISPOT assay has been established for the direct ex vivo quantification of peptide-reactive T lymphocytes from peripheral blood mononuclear cells (PBMC). In this report we studied the ELISPOT assay conditions for the detection of T cell responses against protein antigens including Tetanus toxoid (TT), purified-protein-derivative (PPD) and a synthetic 30 mer peptide derived from influenza matrix protein (IMP) containing the HLA-A*0201-restricted 9 mer peptide epitope GILGFVFTL as a model for a viral protein. We found several aspects to be crucial for a sensitive detection of T cell responses against proteins including a pellet preincubation step and the monocyte concentration. Using optimized assay conditions specific CD4+ T cell responses against TT and PPD as well as CD8+ T cell responses against IMP can be quantified directly ex vivo from peripheral blood mononuclear cells.

Flow cytometric determination of intracellular or secreted IFNγ for the quantification of antigen reactive T cells

The detection of antigen-induced IFNgamma secretion at the single cell level can be used to identify and enumerate antigen-reactive T cells from peripheral blood. This study was performed to analyze the suitability of T cell enumeration by flow cytometry in comparison with the ELISPOT assay. Peripheral blood mononuclear cell (PBMC) samples from six HLA-A2+ healthy subjects were analysed for the frequency of influenza-reactive CD8+ T cells by flow cytometry detecting either intracellular IFNgamma (IC-FC) or secreted IFNgamma (S-FC). All samples were also analysed by IFNgamma ELISPOT assay. The frequency of influenza peptide-reactive T cells determined by IC-FC was 0.01 to 0.34% of CD8+ T cells and by ELISPOT assay 0.02 to 0.23% of CD8+ T cells (n=6 subjects) with a high inter-assay reproducibility and a close correlation between the assays (r=0.77, P<0.001). Little or no IFNgamma production was observed in unstimulated PBMC samples using either the IC-FC or the ELISPOT assay. In contrast, using S-FC large numbers of IFNgamma-secreting CD8+ T cells (0.37% to 5.55%, n=6 subjects) were detected in unstimulated PBMC. The frequency of influenza-reactive CD8+ T cells (0.57-5.19%, n=6 subjects) determined by S-FC did not correlate with the values from the IC-FC or ELISPOT assays. This comparative study shows the suitability of the determination of frequencies of antigen reactive T cells in PBMC by IC-FC. The advantage of IC-FC is the possibility to phenotype simultaneously antigen-reactive T cells.

Effects of granulocyte-macrophage colony-stimulating factor and foreign helper protein as immunologic adjuvants on the T-cell response to vaccination with tyrosinase peptides

Immunologic adjuvants are used to augment the immunogenicity of MHC class I–restricted peptide vaccines, but this effect has rarely been systematically evaluated in a clinical trial. We have investigated, in a phase I study, whether addition of the 2 adjuvants GM‐CSF and KLH can enhance the T‐cell response to MHC class I peptide vaccines. Forty‐three high‐risk melanoma patients who were clinically free of disease received 6 vaccinations with MHC class I–restricted tyrosinase peptides alone, with either GM‐CSF or KLH or with a combination of both adjuvants. The primary end point was induction of tyrosinase‐specific T cells, and serial T‐cell monitoring was performed in unstimulated peripheral blood samples before and after the second, fourth and sixth vaccinations by ELISPOT assay. Tyrosinase‐specific IFN‐γ‐producing T cells were detected as early as 2 weeks after the second vaccination in 5 of 9 patients vaccinated with tyrosinase peptides in combination with GM‐CSF and KLH but not in any patient vaccinated with tyrosinase peptides without adjuvants or in combination with either adjuvant alone. After 6 vaccinations, tyrosinase‐specific T cells were found in patients immunized with peptides either without adjuvants (3 of 9 patients) or in combination with the single adjuvant GM‐CSF (4 of 9 patients) but not with KLH (0 of 10 patients). Our results suggest that addition of either GM‐CSF or KLH as a single adjuvant has little impact on the immunogenicity of tyrosinase peptides. The combined application of GM‐CSF and KLH was associated with early induction of T‐cell responses.

High frequencies of circulating melanoma‐reactive CD8+ T cells in patients with advanced melanoma

To determine whether circulating tumor‐reactive T cells are present in melanoma patients, unstimulated T cells from peripheral blood were tested for recognition of HLA‐A2‐ or HLA‐A1‐matched melanoma cell lines using the ELISPOT assay. Eleven out of 19 patients with metastatic melanoma had a T‐cell response with up to 0.81%, 0.78%, 0.53%, 0.12%, 0.10%, 0.09%, 0.07%, 0.06%, 0.06%, 0.04%, and 0.04% of peripheral blood mononuclear cells (PBMC) secreting IFNγ upon exposure to various HLA‐A2‐ or HLA‐A1‐matched melanoma cell lines. These T‐cell responses were mediated by CD8+ T cells and could specifically be blocked by an anti‐HLA‐A2 antibody in HLA‐A2‐positive patients. Separation experiments performed in one melanoma patient showed tumor‐reactive T cells in both the CD8+ effector T cell (CD45RA+/IFNγ+) as well as the CD8+ memory T‐cell compartment (CD45RO+/IFNγ+). In 3 out of 5 patients, in whom autologous cell lines were available, similar frequencies of T cells in response to HLA‐A1‐ or HLA‐A2‐matched allogeneic and autologous tumor cells were observed, while 2 patients had a T‐cell response restricted to either the autologous or the allogeneic cell lines. These results give evidence for the presence of tumor‐reactive CD8+ T cells in more than half of melanoma patients tested. Although some of these patients have clinical evidence for an immunological‐mediated tumor control, several patients have growing tumors suggesting presence of escape mechanisms. Int. J. Cancer 87:659–664, 2000.

Identification of known and novel immunogenic T-cell epitopes from tumor antigens recognized by peripheral blood T cells from patients responding to IL-2-based treatment.

In previous studies CD8+ T cells specific for melanocyte antigens have been frequently found in melanoma patients responding to interleukin‐2 (IL‐2)‐based therapies. In our study we analyzed the suitability of using circulating T cells from melanoma patients with clinical response after IL‐2‐based therapy to identify novel T‐cell epitopes from defined tumor antigens. Using unstimulated peripheral blood mononuclear cells and the interferon‐γ (IFN‐γ) ELISPOT assay, we studied CD8+ T‐cell responses against 5 peptides from the tumor antigen tyrosinase (Tyr) selected by epitope prediction using an HLA‐A1‐binding computer algorithm. T cells specifically secreting IFN‐γ in response to 3 of these 5 peptides, namely, Tyr (454–463), Tyr (146–156) and Tyr (243‐251), could be detected in 4 of 4 HLA‐A1‐positive patients with clinical response. In contrast, no T‐cell responses against these peptides were seen in 6 HLA‐A1‐positive melanoma patients with progressive disease and in 8 healthy subjects. We could generate specific cytotoxic T lymphocytes (CTL) against Tyr (454–463) using peptide‐pulsed autologous dendritic cells as antigen‐presenting cells. The induced CTLs efficiently killed melanoma cells that express HLA‐A1 and tyrosinase. The peptides Tyr (146–156) and Tyr (243‐251) had recently been identified as CTL epitopes by other groups. Further ex vivo characterization of the T cells reactive against the novel epitope Tyr (454–463) in 1 patient by multicolor flow cytometry showed specific CD3+/CD8+/IFN‐γ+ T cells with frequencies of up to 0.41% of the CD3+/CD8+ T‐cell population. Most of this T‐cell population also expressed granzyme B. Our data confirm that in patients with tumor regressions induced by immunotherapy or chemoimmunotherapy circulating T cells reactive with tyrosinase epitopes can frequently be detected. Peripheral blood T cells from such patients are a valuable source for screening peptides selected by epitope prediction This strategy facilitates the rapid identification of immunogenic T‐cell epitopes that are probable targets of immune‐mediated tumor rejection.

Long‐term freedom from recurrence in 2 stage IV melanoma patients following vaccination with tyrosinase peptides

We report here on 2 patients who received adjuvant vaccination with an HLA‐A2– or HLA‐A24–restricted tyrosinase peptide, respectively, and GM‐CSF for frequently relapsing stage IV melanoma. Following resection of metastases and irradiation of brain metastases in 1 patient, both patients were without evidence of disease when receiving the first vaccination. While the patients had had 9 and 12, respectively, mostly s.c., relapses during the 3 years before vaccination, they experienced freedom from relapse for more than 2 years after vaccination. We found a T‐cell response to the vaccine peptide in both patients in the peripheral blood by ex vivo IFN‐γ ELISPOT assay. The T‐cell population could be further characterized by 4‐color flow cytometry in 1 patient, showing that the majority of the peptide‐specific CD3+CD8+IFN‐γ+ T cells were granzyme B‐positive and CCR‐7‐negative, characterizing them as effector T cells with the ability to mediate cytotoxicity and migrate to inflamed tissues. In this patient also, augmentation of the T‐cell response to autologous tumor cells by vaccination could be detected. A single‐site postvaccination relapse occurred in both patients, showing downregulation of tyrosinase expression in 1 patient, while normal expression levels for tyrosinase, MHC class I antigens and components of the antigen‐processing machinery were found in the other patient. These results suggest that peptide vaccination resulted in a prolonged relapse‐free interval in these high‐risk patients.

Induction of tyrosinase-reactive T cells by treatment with dacarbazine, cisplatin, interferon-alpha +/- interleukin-2 in patients with metastatic melanoma.

We have shown the presence of tyrosinase-reactive T cells in the peripheral blood of melanoma patients, who had been in remission after treatment with IL-2-containing regimens. In this consecutive study, we analyzed the T-cell response to various peptides derived from tyrosinase in serial blood samples obtained from 7 stage-IV melanoma patients before, during and following treatment. All patients were treated within a randomized trial (EORTC 18951) with cisplatin (CDDP), dacarbazine (DTIC), interferon-alpha (IFN-alpha) +/- interleukin-2 (IL-2). Using an ELISPOT assay detecting peptide-specific IFN-gamma release, we measured the T-cell response to 4 different HLA class I-binding peptide epitopes derived from tyrosinase containing an HLA-A2.1-, HLA-A24- or HLA-B44-binding motif in peripheral-blood mononuclear cells (PBMC). In one patient, tyrosinase-reactive T cells were detected before therapy. In 4 out of 7 patients, tyrosinase-reactive T cells against both HLA-A2.1-binding peptides and the B44-binding peptide became detectable at frequencies of up to 30 in 5 x 10(5) lymphocytes following treatment. These patients received CDDP, DTIC and IFN-alpha, 2 of them without IL-2 and 2 with IL-2, resulting in one complete remission and 3 partial remissions. Two patients relapsed 8 and 9 months after treatment. At the time of relapse, no T cells reactive with tyrosinase were detectable. Our results show that high frequencies of tyrosinase-reactive T cells in the peripheral blood of melanoma patients can be induced by chemotherapy in combination with IFN-alpha, regardless of concomitant IL-2 administration.

Brain metastases following interleukin-2 plus interferon-alpha-2a therapy: a follow-up study in 94 stage IV melanoma patients.

This study analyses the frequency and therapy of brain metastases in 94 stage IV melanoma patients after treatment with high-dose interleukin-2 (IL-2) and interferon-alpha (IFN-alpha) within three subsequent trials between 1990 and 1995. Central nervous system (CNS) metastases occurred in 28 patients (30%) during the potential follow-up period of 6 years. Time to occurrence of brain metastases varied between 1 and 53 months, with a median of 10 months. Of 28 patients, 19 had < 5 metastases, which were treated with stereotactic radiosurgery (SR) in 9 patients. In 2 patients, SR was followed by resection. 9 patients had multiple metastases, of which 4 received whole brain irradiation (WBI). Median survival after the detection of CNS metastases was 6 months (95% Confidence Interval (CI) 1-11 months). SR plus resection was associated with a prolonged survival of 34 and 35 months in 2 patients, 1 patient survived for 41 months after WBI, demonstrating the efficacy of these therapeutic strategies.