The immunoglobulin receptor of B-cell lymphomas constitutes a specific tumor antigen (idiotype) and a target for active immunotherapy. Encouraging results have been reported in phase II trials after s.c. vaccination of follicular lymphoma patients during clinical remission with idiotype produced from eukaryotic cell lines and coupled to an immunogenic carrier macromolecule. We have developed a good manufacturing protocol for rapid expression of idiotype vaccines as recombinant Fab fragments in Escherichia coli. The objectives of this trial were to show safety and feasibility of intradermal immunization with this vaccine and to investigate whether immune responses were induced by this immunization route. Patients (n = 18) with advanced B-cell malignancies received repetitive intradermal vaccinations with 0.5 to 1.65 mg recombinant idiotype Fab fragment mixed with lipid-based adjuvant in combination with 150 μg granulocyte macrophage colony-stimulating factor s.c. at the same location. The patients' immune status was assessed by flow cytometry of peripheral blood lymphocytes and concomitant hepatitis B vaccination. Cellular and humoral immune responses to the vaccine were assessed by enzyme-linked immunospot and ELISA. Side effects of a total of 65 vaccinations were mild and did not affect the immunization schedule. No patient developed hepatitis B surface antibodies (anti-HBs) after two hepatitis B immunizations. Of 17 evaluable patients, five developed specific anti-vaccine antibodies, and eight developed anti-Fab T-cell responses. T-cell reactivity was independent of the cellular immune status and was idiotype specific as shown by statistical regression analysis (P = 0.0024) and epitope mapping studies. Intradermal administration of uncoupled recombinant idiotype with appropriate adjuvants may overcome profound clinical immunosuppression and induce specific immune responses. (Cancer Res 2006; 66(8): 4496-502)

B-cell non-Hodgkin's lymphomas (B-NHL) are clonal proliferations of cells expressing an identical antigen receptor. The hypervariable parts of this immunoglobulin form a unique idiotype of the malignant clone and represent a tumor-specific antigen. Immunization with lymphoma idiotype induces protective and therapeutic anti-lymphoma immunity in animal models (14). Depending on the tumor model and the vaccine formulation, both cellular and humoral immunity may play a role in tumor rejection (57).

Idiotype-specific immune responses and disappearance of minimal residual disease (MRD) have been observed after idiotype immunization in follicular lymphoma patients during clinical remission (811). Induction of immunity was associated with superior outcome (9, 12). Some objective tumor regressions upon idiotype vaccination have also been reported (8, 13, 14).

With the exception of multiple myeloma and Waldenstrom's macrogobulinemia, where soluble idiotype protein may by purified from the patient's serum (15, 16), idiotype vaccines from nonsecreting lymphomas must be manufactured. Viable lymphoma cells from a biopsy may be fused with a myeloma cell line by somatic cell hybridization followed by purification of idiotype from the culture supernatant of the resulting heterohybridoma (17). Several strategies for production of recombinant idiotype protein in eukaryotic cells with an appropriate expression system have been proposed (i.e., expression as an IgG3 in a murine lymphoma cell line, as IgG1 in sf9 insect cells, and as scFv fragment in tobacco plants; ref. 18). In almost all reported clinical trials with these vaccines, the weakly immunogenic idiotype protein was covalently coupled to the potent immunogenic carrier molecule keyhole limpet hemocyanin (KLH).

To ease manufacturing of individual idiotype vaccines and to shorten production times, we have adapted a bacterial expression system for recombinant immunoglobulin Fab fragments to the production of recombinant idiotype vaccines in Escherichia coli (19, 20). Stimulation of peripheral blood T cells from lymphoma patients with such recombinant, lymphoma-derived Fab fragment resulted in specific, MHC class I–restricted cytotoxic activity against autologous, idiotype-expressing target cells (21). Because coupling rates to KLH were too variable to fulfill good manufacturing (GMP) criteria,3

3

F. Simon, unpublished observations.

we decided to test alternative strategies to obtain high immunogenicity without further chemical vaccine modification. Intradermal vaccination with a reduced antigen dose has recently been shown to be equipotent to a conventional s.c. influenza vaccine, presumably due to antigen delivery to professional antigen-presenting cells (APC) of the skin (22). In addition, novel lipid-based adjuvants with superior immunogenicity, such as MF59, have been introduced into commercially available vaccines (23). Based on these advances in the design of vaccines against common infectious agents, we chose to apply a mixture of the recombinant idiotype with MF59 via an intradermal injection route. Because this trial tested to our knowledge for the first time a drug manufactured individually in E. coli, and because significant toxicities of the intradermal formulation could not be excluded, we conducted a clinical phase I trial to show feasibility and tolerability of such a vaccination and to investigate the immunogenicity of this idiotype vaccine.

Subject eligibility. In a typical phase I design, patients with B-cell malignancies in advanced stages that had relapsed or progressed after a full treatment course with anthracyclin- or fludarabine-containing or high-dose chemotherapy with autologous stem cell transplantation were enrolled after giving informed consent. Patients were at least 18 years old and had a life expectancy of at least 6 months. Exclusion criteria were severe concomitant illnesses or infections, active central nervous system involvement, and immunosuppressive or antineoplastic therapy (including irradiation) 4 weeks before vaccination. The trial was approved by the Institutional Review Board and was conducted according to the Declaration of Helsinki.

Vaccine production. Lymphoma tissue was obtained by lymph node biopsy (n = 9), bone marrow aspiration (n = 7), or phlebotomy (n = 2). GMP-grade individual idiotype vaccines were produced commercially (CellGenix GmbH, Freiburg, Germany) by anchored reverse transcription-PCR cloning of the idiotype heavy (H) and light (L) chain expressed by the lymphoma. Lymphoma-derived variable immunoglobulin segments were inserted into pFab.γκ or pFab.γλ, and recombinant idiotype protein was expressed in E. coli as a recombinant IgG1 Fab fragment (20). Fab protein was purified by affinity chromatography and gel filtration. The success rate for Fab production is currently 89% in a continued series of 78 consecutive vaccine productions with a median purity of 99% heterodimeric Fab protein (range, 72-100%) and a median yield of 17.0 mg (range, 1.2-250 mg) from a single 14-liter fermentation run.4

4

Details to be published by Bertinetti et al., submitted.

Treatment schedule. One vaccination consisted of an intradermal injection of 10 μg/μL idiotype solution mixed 5:1 (v/v) with MF59 adjuvant (Chiron Behring, Marburg, Germany; ref. 23). The minimum dose of idiotype protein was 0.5 mg per vaccination and was planned to be increased according to the modified Fibonacci design after treatment of three patients at each dose level without relevant side effects. In addition, 150 μg granulocyte macrophage colony-stimulating factor (GM-CSF; molgramostim; Leucomax; Novartis, Nuernberg, Germany) were injected s.c. at the same anatomic location. Vaccinations were given in weeks 0, 3, 5, and 9.

Beginning with the fourth patient enrolled, anti-HBs-negative patients received a standard hepatitis B vaccine (Gen-HB-Vax; Aventis Pasteur, Frankfurt, Germany) at the first and third idiotype vaccination. Staging according to National Cancer Institute criteria (24) with bidimensional measurement of lymphomatous masses or by quantitative assessment of appropriate disease variables [e.g., chronic lymphocytic leukemia (CLL), lymphocyte counts; myeloma, immunoglobulin concentration] was done within 1 week before the first vaccination, at the third vaccination, 2 weeks after the fourth vaccination, and at weeks 16, 24, and 52. Disease progression at any time excluded the patient from further follow-up. Side effects were graded according to WHO criteria (25). Blood samples for routine clinical chemistry, quantitation of lymphocyte subsets by dual-color flow cytometry (antibodies from BD Biosciences, Heidelberg, Germany), assessment of cellular and humoral immunity against idiotype, and screening for autoantibodies were taken on the days of the staging examinations.

Detection of idiotype antibodies by ELISA. Microtiter plates (Nunc, Wiesbaden, Germany) were coated with 10 μg/mL recombinant Fab fragment in 50 mmol/L sodium carbonate buffer (pH 9.6) and blocked with 0.1 mol/L Tris-HCl (pH 7.6)/0.05% Tween 20. Serially diluted sera were added for 1 hour at room temperature. Horseradish peroxidase (HRP)–labeled goat F(ab′)2 anti-human IgM (Biosource, Camarillo, CA), anti-human IgG-Fc (Calbiochem, La Jolla, CA), or biotin-conjugated mouse anti-human κ or λ (BD Biosciences) served as secondary antibodies. After additional incubation with Streptavidin-HRP (BD Biosciences) for biotin-conjugated secondary antibodies, bound HRP was detected with ABTS solution (Roche, Mannheim, Germany) and absorbance measurement at 405 nm. According to previously described criteria (12), induction of an idiotype-specific in vivo antibody response was defined as a 4-fold titer increase compared with prevaccination serum and to a control Fab.

Generation of idiotype-presenting monocyte-derived dendritic cells. Peripheral blood mononuclear cells (PBMC) were adhered to culture flasks for 2 hours at 37°C in RPMI (Life Technologies/Invitrogen, Karlsruhe, Germany) supplemented with 0.5% HSA (Baxter Germany, Munich, Germany) followed by culturing in serum-free medium (CellGro DC; CellGenix) with 100 ng/mL GM-CSF (R&D Systems, Wiesbaden-Nordenstadt, Germany) and 50 ng/mL interleukin-4 (IL-4; CellGenix) for 5 days. At day 4, 100 μg/mL idiotype Fab protein was added, and final dendritic cell differentiation was induced with 20 ng/mL tumor necrosis factor-α (CellGenix) and 1 μg/mL prostaglandin E2 (ICN, Aurora, OH) for 48 hours. For peptide epitope mapping, matured dendritic cells were loaded with 20 μg/mL peptide. MHC class I–restricted potential peptide epitopes were identified by the SYFPEITHI algorithm.5

PBMC in vitro stimulation and enzyme-linked immunospot assay. Ninety-six-well HA S4510 plates (Millipore, Billerica, MA) were coated with 15 μg/mL anti-IFNγ antibody (BD Biosciences PharMingen, Heidelberg, Germany) overnight at 4°C and blocked with 10% AB serum (PAA, Pasching, Austria) in basal Iscove's medium (Biochrom, Berlin, Germany). PBMCs (5 × 104) were stimulated with 5 × 103 antigen-presenting dendritic cells or with 1 μg/mL pokeweed mitogen (viability control; Sigma-Aldrich, Taufkirchen, Germany) in RPMI with 5% AB serum. After a 48-hour incubation at 37°C, plates were washed with 0.05% Tween 20 in PBS and incubated overnight at 4°C with 0.5 mg/mL biotinylated anti-IFNγ antibodies (BD Biosciences PharMingen). IFNγ-secreting cells were detected with 1:1,000-diluted streptavidine-alkaline phosphatase and substrate solution (Bio-Rad, Muenchen, Germany). Spots were counted semiautomatically with an enzyme-linked immunospot (ELISPOT) reader (AID, Straßberg, Germany). To increase the sensitivity of the ELISPOT when necessary, 5 × 104 PBMCs were prestimulated with 5 × 103 autologous, Fab-loaded dendritic cells per well with 20 IU/mL IL-2 (Chiron, Ratingen, Germany) and 5 ng/mL IL-7 (Cell Concepts, Umkirch, Germany) for 1 week. Only experiments with 200 to 300 spots per well in the pokeweed mitogen controls and with more than three background spots (without antigen) per 105 cells were analyzed. A T-cell response was defined as a significant (P < 0.05) increase of the response to the vaccine idiotype by two-tailed t test. The specificity of T-cell responses was analyzed with patient and control idiotype as ELISPOT targets by a Poisson regression model accounting for repeated measurements using the generalized estimation equations method for variable estimation with SAS version 8 software (SAS Institute, Cary, NC; ref. 26). This model includes the different conditions under which measurements were taken as independent variables. The estimated effects in this model then represent the ratios that relate average counts under one condition to another. All other statistical calculations were done with GraphPad Prism software (GraphPad Software, San Diego, CA).

Treatment administration and side effects. Eighteen patients (13 male and 5 female) with various B-cell malignancies and a median age of 53.5 years (range, 32-75 years) were enrolled in the vaccination protocol (Table 1). A total of 65 vaccinations was given within the trial. The dose was 0.5 mg idiotype protein per vaccination in 13 patients, 1.0 mg in three patients, and 1.65 mg in two patients. Further dose escalation was abandoned because the accumulating production experience suggested that it would be unlikely to obtain sufficient vaccine without repeated fermentation runs. Localized erythema, hyperthermia, and transient skin induration of up to 12 cm, which tended to increase with each vaccination, developed at the vaccination site but resolved after a few days. Grade 3 anemia and thrombocytopenia developed in one case each who were already in grade 2 cytopenia immediately before vaccination. All other symptoms and side effects were mild (Fig. 1), and all vaccinations were given according to schedule.

Table 1.

Patient characteristics, vaccinations, immunologic responses, and time to progression

PatientDiagnosis, stage*GammopathyPrior treatment regimensPrior immunotoxic therapyTime last therapy/vaccination (mo)Remission status at vaccinationNo. vaccinationsVaccine dose (mg)T-cell response by t testAntibody response, titer increaseAuto-antibodyTime to progression (mo)
A05 FL, IV — F, R >2nd PR 0.5 NS Negative — NA 
A15 FL, IV — — 13 PD 0.5 P = 0.032 Negative Trace ANA ≥5 
A25 CLL, III — 21 PD 0.5 NS Negative — NA 
A03 LPL, IV — autoSCT, R PD 1.0 NS IgM 8×; IgG 64× RF = 17 IU/l§ 
A26 MCL, IV — — 73 PD 4 (+4) 1.0 P < 0.0001 IgM 128×; IgG 128× Trace ANA 
A27 FL, IV — autoSCT, R, F >2nd PR 1.0 NS IgM 6×; IgG 8× Trace ANA, RF = 20 IU/l§ 26 
A28 DLCL, III — >2nd PR 4 (+4) 1.0 (P = 0.052) Negative RF = 37 IU/l§ ≥36 
A51 MM, III IgGκ autoSCT PD 1.65 P = 0.0002 IgM 8×; IgG <2× — NA 
A83 CLL, IV — autoSCT PD 4 (+2) 1.65 P = 0.0014 IgM <2×; IgG 8× — 
A21 MCL, IV — PD 4 (+3) 0.5 P = 0.018 IgM <2×; IgG 6× (ANA, RF) ≥39 
A32 MCL, IV — — 43 PD 4 (+4) 0.5 NS Negative — 
A55 MM, II IgGκ autoSCT 2nd PR 0.5 NS Negative — 
A60 LPL, IV IgMκ 2nd PR 0.5 P < 0.0001 Negative Trace ANA 
A73 FL, III — autoSCT, F >2nd PR 4 (+3) 0.5 P = 0.0066 IgM 2×; IgG 32× — ≥27 
A75 MM IgGκ autoSCT 12 PD 0.5 NA NA NA NA 
A82 FL — PD 0.5 P = 0.0079 Negative — NA 
B07 CLL — 10 F, R, alloSCT PD 0.5 NS Negative — 
A88 CLL — 21 PD 0.5 NS Negative — 
PatientDiagnosis, stage*GammopathyPrior treatment regimensPrior immunotoxic therapyTime last therapy/vaccination (mo)Remission status at vaccinationNo. vaccinationsVaccine dose (mg)T-cell response by t testAntibody response, titer increaseAuto-antibodyTime to progression (mo)
A05 FL, IV — F, R >2nd PR 0.5 NS Negative — NA 
A15 FL, IV — — 13 PD 0.5 P = 0.032 Negative Trace ANA ≥5 
A25 CLL, III — 21 PD 0.5 NS Negative — NA 
A03 LPL, IV — autoSCT, R PD 1.0 NS IgM 8×; IgG 64× RF = 17 IU/l§ 
A26 MCL, IV — — 73 PD 4 (+4) 1.0 P < 0.0001 IgM 128×; IgG 128× Trace ANA 
A27 FL, IV — autoSCT, R, F >2nd PR 1.0 NS IgM 6×; IgG 8× Trace ANA, RF = 20 IU/l§ 26 
A28 DLCL, III — >2nd PR 4 (+4) 1.0 (P = 0.052) Negative RF = 37 IU/l§ ≥36 
A51 MM, III IgGκ autoSCT PD 1.65 P = 0.0002 IgM 8×; IgG <2× — NA 
A83 CLL, IV — autoSCT PD 4 (+2) 1.65 P = 0.0014 IgM <2×; IgG 8× — 
A21 MCL, IV — PD 4 (+3) 0.5 P = 0.018 IgM <2×; IgG 6× (ANA, RF) ≥39 
A32 MCL, IV — — 43 PD 4 (+4) 0.5 NS Negative — 
A55 MM, II IgGκ autoSCT 2nd PR 0.5 NS Negative — 
A60 LPL, IV IgMκ 2nd PR 0.5 P < 0.0001 Negative Trace ANA 
A73 FL, III — autoSCT, F >2nd PR 4 (+3) 0.5 P = 0.0066 IgM 2×; IgG 32× — ≥27 
A75 MM IgGκ autoSCT 12 PD 0.5 NA NA NA NA 
A82 FL — PD 0.5 P = 0.0079 Negative — NA 
B07 CLL — 10 F, R, alloSCT PD 0.5 NS Negative — 
A88 CLL — 21 PD 0.5 NS Negative — 

Abbreviations: FL, follicular lymphoma; MM, multiple myeloma; MCL, mantle cell lymphoma; LPL, lymphoplasmocytic lymphoma; DLCL, diffuse large cell lymphoma; F, fludarabine; R, rituximab; autoSCT, autologous hematopoietic stem cell transplantation; alloSCT, allogeneic SCT; PD, progressive disease; PR, partial remission; NS, not significant; ND, not done; NA, not applicable due to early disease progression; RF, rheumatoid factor.

*

Staging system applied: CLL, Rai; MM, Durie and Salmon; all other lymphoma types, Ann Arbor.

A complete course of radiation therapy was counted as one regimen.

The planned dose escalation was discontinued at the third level.

§

Normal range for RF = 0 to 16 IU/L.

>50% cross-reactivity with control Fabs.

Patient with preexisting seropositive rheumatoid arthritis.

Figure 1.

Side effects observed during idiotype vaccination. A, percentage and severity of side effects per total of 65 idiotype vaccinations. B, percentage and severity of side effects per total of 18 vaccinated patients. *, thrombocytopenia and anemia grade 3 were observed in one patient each. Both patients had formal grade 2 cytopenia before vaccination.

Figure 1.

Side effects observed during idiotype vaccination. A, percentage and severity of side effects per total of 65 idiotype vaccinations. B, percentage and severity of side effects per total of 18 vaccinated patients. *, thrombocytopenia and anemia grade 3 were observed in one patient each. Both patients had formal grade 2 cytopenia before vaccination.

Close modal

In six patients, testing for autoantibodies indicated possible subclinical autoimmune activity 11 to 24 weeks after start of idiotype vaccination (Table 1). One additional patient (A21) with follicular lymphoma and rheumatoid arthritis had an approximate doubling of rheumatoid factor to 586 IU/mL at 24 weeks after the first idiotype vaccination but no clinical exacerbation of his arthritis at the metacarpophalangeal joints.

T-cell status and cellular immune responses. Patients without circulating lymphoma cells had low normal to severely depressed circulating CD4+ T-cell counts before vaccination. Prior treatment with fludarabine was associated with a particularly poor T-helper cell status but had a lesser effect on cytotoxic T cells (Fig. 2A). In accordance with published data (27, 28), the CLL/lymphocytic lymphoma patients had normal to high CD4+ and CD8+ T-cell counts even when pretreated with fludarabine, but these cells are known to be functionally incompetent (29). Induction of T cell–mediated immunity to the vaccine was detected by ELISPOT in eight patients (Fig. 2B; Table 1). Induction of T cells responses seemed to be independent of T-helper cell counts before vaccination (Fig. 2C). Because four of these patients also had an increase of T-cell reactivity to a randomly chosen control Fab (Fig. 2B), we did a combined Poisson regression analysis of the ELISPOT results. This analysis proved predominant specificity of the T-cell response for the vaccine compared with randomly chosen control Fabs (Fig. 2D).

Figure 2.

Immune status and detection of vaccine-specific cellular immune responses by ELISPOT. A, T-cell subsets in the peripheral blood of vaccinated patients according to diagnostic subgroups and fludarabine pretreatment. Hatched areas represent normal ranges. F, fludarabine. *, a case with leukemic mantle cell lymphoma (A32). B, columns, mean of IFNγ spots obtained from individual patients against the vaccine (patient Fab) and a randomly chosen control Fab; bars, SD. Patient A60 has a strictly specific immune response, whereas patient A15 displays partial cross-reactivity. C, T-helper cell counts of patients according to detection or absence of a cellular immune response as determined by ELISPOT. *, leukemic non-Hodgkin's lymphoma (four CLL and one mantle cell lymphoma). Pos., positive; neg., negative. D, combined statistical analysis of all patients compared with prevaccination reactivity to control Fab (with 95% confidence interval) by a Poisson regression model. Pre, before vaccination; Post, after vaccination.

Figure 2.

Immune status and detection of vaccine-specific cellular immune responses by ELISPOT. A, T-cell subsets in the peripheral blood of vaccinated patients according to diagnostic subgroups and fludarabine pretreatment. Hatched areas represent normal ranges. F, fludarabine. *, a case with leukemic mantle cell lymphoma (A32). B, columns, mean of IFNγ spots obtained from individual patients against the vaccine (patient Fab) and a randomly chosen control Fab; bars, SD. Patient A60 has a strictly specific immune response, whereas patient A15 displays partial cross-reactivity. C, T-helper cell counts of patients according to detection or absence of a cellular immune response as determined by ELISPOT. *, leukemic non-Hodgkin's lymphoma (four CLL and one mantle cell lymphoma). Pos., positive; neg., negative. D, combined statistical analysis of all patients compared with prevaccination reactivity to control Fab (with 95% confidence interval) by a Poisson regression model. Pre, before vaccination; Post, after vaccination.

Close modal

A comparison of ELISPOT data of patient A26 with two different control Fabs provided further insight into the molecular basis for a specific anti-idiotype T-cell response (Fig. 3). The T-cell response induced by the vaccination was stronger to the vaccine Fab than to a control Fab with low VH sequence homology (A14). In contrast, a highly homologous control Fab (A21) with the same VH gene as the vaccine Fab contained numerous shared potential epitopes as revealed by epitope prediction algorithms. Accordingly, no preferential T-cell reactivity to the vaccine could be detected.

Figure 3.

Specificity analysis of anti-idiotype T cell response in patient A26. Top, position of MHC-restricted peptide epitopes identified by the SYFPEITHI algorithm. Bottom, cellular immune response from patient A26 against A14 Fab (left) and A21 Fab (right) as specificity controls. Since the comparisons were done in separate experiments, numbers of IFNγ-reactive spots are given as percentage of prevaccination reactivity to A26 idiotype. Restriction elements and binding scores (bottom right).

Figure 3.

Specificity analysis of anti-idiotype T cell response in patient A26. Top, position of MHC-restricted peptide epitopes identified by the SYFPEITHI algorithm. Bottom, cellular immune response from patient A26 against A14 Fab (left) and A21 Fab (right) as specificity controls. Since the comparisons were done in separate experiments, numbers of IFNγ-reactive spots are given as percentage of prevaccination reactivity to A26 idiotype. Restriction elements and binding scores (bottom right).

Close modal

Availability of sufficient T cells permitted a formal epitope mapping in patient A83 (Fig. 4). T-cell epitopes derived from various positions in the idiotype H chain and predicted to be presented by the patient's HLA haplotype by reverse immunology were tested separately and compared with a peptide from a control Fab with similar HLA-binding scores. Recognition of C region peptides was significantly inferior to the control peptide. In contrast, a significantly higher response was detected against a peptide from the CDR2 region (an HLA-restricted CDR3 peptide could not be identified in this case).

Figure 4.

T-cell peptide epitope mapping of idiotype of patient A83. ELISPOT analysis of post-vaccination PBMC against synthetic peptides identified by the SYFPEITHI algorithm in the patient's lymphoma idiotype IgH chain as indicated. A73: control epitope from the CDR3 of patient A73 (KNQTGPFDY). Restriction elements and binding scores (bottom right).

Figure 4.

T-cell peptide epitope mapping of idiotype of patient A83. ELISPOT analysis of post-vaccination PBMC against synthetic peptides identified by the SYFPEITHI algorithm in the patient's lymphoma idiotype IgH chain as indicated. A73: control epitope from the CDR3 of patient A73 (KNQTGPFDY). Restriction elements and binding scores (bottom right).

Close modal

B-cell status and humoral immune responses. Pretreatment effects on the B-cell compartment were less pronounced than on T cells: nonleukemic patients who had received rituximab before vaccination did not have significantly less B cells than rituximab-naive patients (Fig. 5A). None of 10 patients who were evaluable for anti-HBs responses after two hepatitis B vaccinations developed any detectable anti-HBs antibodies. This subgroup did not differ from the remaining patients with respect to fludarabine (P = 1.0, Fisher's exact test) or rituximab pretreatment (P = 0.64), CD4+ T-cell counts (P = 0.7, two-sided Mann-Whitney test), or B-cell counts (P = 0.6).

Figure 5.

Detection of vaccine-specific antibody responses. A, B cells in the peripheral blood of vaccinated patients according to diagnostic subgroups and rituximab pretreatment. Hatched area, normal range. Ritux., rituximab. B, B-cell counts of patients according to detection or absence of a specific humoral immune response as determined by ELISA. *, leukemic NHL (four CLL and one mantle cell lymphoma). C, serial dilutions of sera from vaccinated patients were analyzed for binding to Fab fragment by ELISA with an anti-Fcγ detection antibody. Top, longitudinal analysis of the anti-vaccine antibody titer during and after the vaccination course. Bottom, specificity analysis of preimmune and immune sera with vaccine and control Fab targets.

Figure 5.

Detection of vaccine-specific antibody responses. A, B cells in the peripheral blood of vaccinated patients according to diagnostic subgroups and rituximab pretreatment. Hatched area, normal range. Ritux., rituximab. B, B-cell counts of patients according to detection or absence of a specific humoral immune response as determined by ELISA. *, leukemic NHL (four CLL and one mantle cell lymphoma). C, serial dilutions of sera from vaccinated patients were analyzed for binding to Fab fragment by ELISA with an anti-Fcγ detection antibody. Top, longitudinal analysis of the anti-vaccine antibody titer during and after the vaccination course. Bottom, specificity analysis of preimmune and immune sera with vaccine and control Fab targets.

Close modal

A specific increase of anti-idiotype antibodies was detected in five patients (Table 1). These humoral responses were confirmed in all cases with anti-“opposite” L chain detection antibodies (i.e., anti-λ detection antibody for idiotypes with a κ IgL chain and vice versa; data not shown). Two additional patients developed a humoral anti-Fab response with substantial cross-reactivity to different human Fabs but not to a murine control Fab (Table 1). Humoral immune responses were not correlated to T-helper or B-cell counts (Fig. 5B). Anti-idiotype IgG responses persisted for at least 48 weeks in two patients (A21 and A26) without disease progression during this period (Fig. 5C).

Clinical outcome. Six patients were taken off study before the fourth immunization due to early disease progression. Six patients received one to four additional compassionate use immunizations in monthly intervals (Table 1). Six patients have died from lymphoma 4 to 38 months after start of vaccination therapy.

Patients with an anti-idiotype immune response tended to have a better progression-free survival than immunologic nonresponders (Table 1), although the heterogeneity of the patient population in this phase I trial design does not permit meaningful statistical correlations. A quantitative analysis of disease variables in multiple myeloma and CLL patients (multiple myeloma, paraprotein; CLL, leukocyte counts) showed no evidence for any alteration of the natural disease history during the vaccination for multiple myeloma and CLL (data not shown). Two patients (A27 and A28) who were vaccinated during partial remission after their last chemotherapy experienced further shrinkage of their residual masses and reached a complete remission during the vaccination. The patient with diffuse large cell lymphoma (A28) was immunized after an abbreviated course of chemoimmunotherapy of his second relapse. Although his first and second remission had only lasted 3 and 6 months, respectively, he remains in complete remission for >3 years after initiation of immunization.

This first clinical application of a novel idiotype vaccine differs in several aspects from conventional idiotype vaccination approaches: technologically, the production strategy is based on A-PCR, and minute amounts of tissue suffice as starting material. The expression strategy as a recombinant Fab fragment in E. coli (20) has a success rate of ∼90% and permits substantially reduced manufacturing times. Although lack of glycosylation of these bacterially expressed vaccines could potentially affect their efficacy, experimental evidence suggests that immunization with unglycosylated antigen may not impair but rather enhance cellular immune responses against naturally glycosylated peptides (30, 31).

With respect to vaccine administration, intradermal vaccination at a GM-CSF-conditioned site was chosen to deliver the antigen to professional APCs of the skin. GM-CSF is a potent adjuvant for induction of antigen-specific T cells through stimulation of bone marrow–derived APCs (6, 32). This immunization route was adopted explicitly to induce cellular immunity because cellular immune responses induced by idiotype vaccination with GM-CSF was associated with clearance of MRD (10), and because our own in vitro data showed efficient induction of MHC-restricted, idiotype-specific CD8+ CTL by idiotype-presenting dendritic cells, the most potent APC type known (21).

Finally, our vaccine was not conjugated to KLH as in most trials (9, 10, 18) because conjugation rates were highly variable and failed to fulfill GMP criteria. Instead, the vaccine was formulated with the potent lipid-based adjuvant MF59 (23), which has not yet been evaluated for antitumor vaccination.

Because this trial tested, to our best knowledge for the first time, the parenteral application of proteins expressed individually in E. coli, a phase I design with feasibility and safety as primary end points was adopted. Although the vaccination was generally well tolerated, the main disadvantage of the trial design was the inclusion of heavily pretreated patients with a strongly impaired immune system as indicated by low numbers of circulating T-helper cells and refractoriness to hepatitis B vaccination. In contrast to our study population, reported response rates to two hepatitis B vaccinations, defined as anti-HBs titers of ≥10 IU/mL, vary between 86% in a healthy control group (mean age, 57 years), and 41% in breast carcinoma patients in remission at 8 weeks (33).

Despite this profound immunosuppression, humoral and/or cellular immune responses against idiotype developed in 10 of 17 evaluable patients and seemed to occur independently of the patient's cellular immune status. This experience is encouraging because the humoral immune response to a KLH-conjugated s.c. vaccine seems to be dependent on B-cell recovery after passive immunotherapy with anti-CD20 antibody (34). The strong immunogenicity of our vaccine formulation is further emphasized by the fact that seven of the immune responders failed hepatitis B vaccination.

The development of trace amounts of autoantibodies in some patients cautions against the use of even stronger adjuvants to avoid the induction of autoimmune disease. On the other hand, no disease exacerbation was noted in a patient with concomitant rheumatoid arthritis.

A predominant specificity of the cellular immune response towards the individual idiotype was shown by stringent analysis of ELISPOT results with a statistical regression model. This model assesses induction of increased specific reactivity by appropriately accounting for actual reactivity against control targets at every time point. Thus, this calculation performs an objective comparison between pre- and post-vaccination samples without requiring prior definition of thresholds to correct for numbers of spots in negative (no target) and positive (lectin induced) control wells. A lesser increase of nonspecific T-cell reactivity observed against random control targets may be explained by reverse immunology through shared epitopes in vaccine and control target and seems to be dependent on their degree of similarity. Therefore, the specificity of an anti-idiotype immune response may be “relative” and to depend critically on the control target analyzed.

Further probing into the nature of idiotype specificity by epitope mapping revealed that the only peptide that was recognized statistically better than the control was derived from a CDR region. This assay with peptide targets also confirms the idiotype reactivity with a different antigenic format. The inferior response observed against C region peptides is consistent with the assumption that individual rather than shared idiotype epitopes are better targets for cellular immune responses because strong tolerance can be expected for these ubiquitous C region epitopes.

These results are noteworthy because T cells with specificity for FR-derived peptides are present in the blood of lymphoma patients and can be activated in vitro (35). In contrast, little information is available on the precise specificity of in vivo–induced immune responses to idiotype vaccines. Our data indicate that CDR-derived rather than FR epitopes can represent preferential targets of MHC class I–restricted T cells after activation in vivo. In addition, MHC class II–restricted CDR peptides have been identified as cellular immune targets upon vaccination with KLH-conjugated idiotype immunoglobulin and GM-CSF (36), thereby strengthening the conclusion that individual rather than shared idiotype features are preferential targets for vaccination-induced immune responses. Whether cellular immunity indeed represents the most relevant anti-idiotype effector mechanism is not yet clear because a correlation between FcγR III genotype and outcome of follicular lymphoma patients after idiotype vaccination suggests an important role for humoral anti-idiotype immunity (12). Nevertheless, it may be assumed that a T-helper cell response may be required for efficient induction of anti-idiotype antibody responses.

Because the A-PCR-based production strategy of the recombinant idiotype Fab fragment requires only small lymphoma biopsies, we generally obtained insufficient material for in vitro testing for cellular or humoral immune reactivity against lymphoma cells. Killing of autologous lymphoma cells by idiotype vaccination-induced immune cells has, however, already been described (10). Because our own in vitro experiments have also established MHC class I–restricted, specific killing of idiotype-expressing target cells by T cells stimulated with recombinant Fab idiotype (21), it may be assumed that the cellular immunity induced by our vaccine should also be able to attack actual lymphoma cells.

In conclusion, the results of this trial indicate feasibility, good tolerability, and potent immunologic activity of the recombinant idiotype vaccination strategy tested. Immune responses to the vaccine were induced irrespective of the cellular immune status or lack of response to a conventional anti-infection vaccine, indicating that our vaccine formulation may overcome even profound clinical immunodeficiency. These results warrant further clinical testing in earlier disease stages and more homogeneous patient cohorts to assess clinical efficacy. Possible induction of autoimmunity has been identified as a potential hazard of strong vaccine immunogenicity and calls for monitoring of autoimmunity in clinical trials testing potent novel cancer vaccines.

Grant support: Deutsche Krebshilfe grant 70-2698-Ve I (H. Veelken).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

We thank D. Herchenbach for expert technical assistance, Roland Mertelsmann, Director of the Department of Hematology, for assigning departmental staff and research funds to this project, and T. Boon, Brussels, and C. Ottensmeier, Southampton, for critical reading of the manuscript.

1
Campbell MJ, Esserman L, Levy R. Immunotherapy of established murine B cell lymphoma. Combination of idiotype immunization and cyclophosphamide.
J Immunol
1988
;
141
:
3227
–33.
2
Kaminski MS, Kitamura K, Maloney DG, Levy R. Idiotype vaccination against murine B cell lymphoma. Inhibition of tumor immunity by free idiotype protein.
J Immunol
1987
;
138
:
1289
–96.
3
George AJ, Tutt AL, Stevenson FK. Anti-idiotypic mechanisms involved in suppression of a mouse B cell lymphoma, BCL1.
J Immunol
1987
;
138
:
628
–34.
4
Sakato N, Eisen HN. Antibodies to idiotypes of isologous immunoglobulins.
J Exp Med
1975
;
141
:
1411
–26.
5
Campbell MJ, Esserman L, Byars NE, Allison AC, Levy R. Idiotype vaccination against murine B cell lymphoma. Humoral and cellular requirements for the full expression of antitumor immunity.
J Immunol
1990
;
145
:
1029
–36.
6
Kwak LW, Young HA, Pennington RW, Weeks SD. Vaccination with syngeneic, lymphoma-derived immunoglobulin idiotype combined with granulocyte/macrophage colony-stimulating factor primes mice for a protective T-cell response.
Proc Natl Acad Sci U S A
1996
;
93
:
10972
–7.
7
Lauritzsen GF, Weiss S, Dembic Z, Bogen B. Naive idiotype-specific CD4+ T cells and immunosurveillance of B-cell tumors.
Proc Natl Acad Sci U S A
1994
;
91
:
5700
–4.
8
Kwak LW, Campbell MJ, Czerwinski DK, Hart S, Miller RA, Levy R. Induction of immune responses in patients with B-cell lymphoma against the surface-immunoglobulin idiotype expressed by their tumors.
N Engl J Med
1992
;
327
:
1209
–15.
9
Hsu FJ, Caspar CB, Czerwinski D, et al. Tumor-specific idiotype vaccines in the treatment of patients with B-cell lymphoma-long-term results of a clinical trial.
Blood
1997
;
89
:
3129
–35.
10
Bendandi M, Gocke CD, Kobrin CB, et al. Complete molecular remissions induced by patient-specific vaccination plus granulocyte-monocyte colony-stimulating factor against lymphoma.
Nat Med
1999
;
5
:
1171
–7.
11
Nelson EL, Li X, Hsu FJ, et al. Tumor-specific, cytotoxic T-lymphocyte response after idiotype vaccination for B-cell, non-Hodgkin's lymphoma.
Blood
1996
;
88
:
580
–9.
12
Weng WK, Czerwinski D, Timmerman J, Hsu FJ, Levy R. Clinical outcome of lymphoma patients after idiotype vaccination is correlated with humoral immune response and immunoglobulin G Fc receptor genotype.
J Clin Oncol
2004
;
22
:
4717
–24.
13
Timmerman JM, Czerwinski DK, Davis TA, et al. Idiotype-pulsed dendritic cell vaccination for B-cell lymphoma: clinical and immune responses in 35 patients.
Blood
2002
;
99
:
1517
–26.
14
Hsu FJ, Benike C, Fagnoni F, et al. Vaccination of patients with B-cell lymphoma using autologous antigen-pulsed dendritic cells.
Nat Med
1996
;
2
:
52
–8.
15
Osterborg A, Yi Q, Henriksson L, et al. Idiotype immunization combined with granulocyte-macrophage colony-stimulating factor in myeloma patients induced type I, major histocompatibility complex-restricted, CD8- and CD4-specific T-cell responses.
Blood
1998
;
91
:
2459
–66.
16
Massaia M, Borrione P, Battaglio S, et al. Idiotype vaccination in human myeloma: generation of tumor-specific immune responses after high-dose chemotherapy.
Blood
1999
;
94
:
673
–83.
17
Carroll WL, Thielemans K, Dilley J, Levy R. Mouse x human heterohybridomas as fusion partners with human B cell tumors.
J Immunol Methods
1986
;
89
:
61
–72.
18
Hurvitz SA, Timmerman JM. Current status of therapeutic vaccines for non-Hodgkin's lymphoma.
Curr Opin Oncol
2005
;
17
:
432
–40.
19
Skerra A. A general vector, pASK84, for cloning, bacterial production, and single-step purification of antibody Fab fragments.
Gene
1994
;
141
:
79
–84.
20
Osterroth F, Alkan O, Mackensen A, et al. Rapid expression cloning of human immunoglobulin Fab fragments for the analysis of antigen specificity of B cell lymphomas and anti-idiotype lymphoma vaccination.
J Immunol Methods
1999
;
229
:
141
–53.
21
Osterroth F, Garbe A, Fisch P, Veelken H. Stimulation of cytotoxic T cells against idiotype immunoglobulin of malignant lymphoma with protein-pulsed or idiotype-transduced dendritic cells.
Blood
2000
;
95
:
1342
–9.
22
Kenney RT, Frech SA, Muenz LR, Villar CP, Glenn GM. Dose sparing with intradermal injection of influenza vaccine.
N Engl J Med
2004
;
351
:
2295
–301.
23
Ott G, Barchfeld GL, Chernoff D, Radhakrishnan R, van Hoogevest P, Van Nest G. MF59. Design and evaluation of a safe and potent adjuvant for human vaccines.
Pharm Biotechnol
1995
;
6
:
277
–96.
24
Cheson BD, Horning SJ, Coiffier B, et al. Report of an international workshop to standardize response criteria for non-Hodgkin's lymphomas. NCI Sponsored International Working Group.
J Clin Oncol
1999
;
17
:
1244
.
25
WHO. Reporting the results of cancer treatment.
Cancer
1981
;
47
:
207
.
26
Liang KY, Zegel SL. Longitudinal data analysis using generalized linear models.
Biometrika
1986
;
73
:
13
–22.
27
Totterman TH, Carlsson M, Simonsson B, Bengtsson M, Nilsson K. T-cell activation and subset patterns are altered in B-CLL and correlate with the stage of the disease.
Blood
1989
;
74
:
786
–92.
28
Kimby E, Mellstedt H, Nilsson B, Bjorkholm M, Holm G. Differences in blood T and NK cell populations between chronic lymphocytic leukemia of B cell type (B-CLL) and monoclonal B-lymphocytosis of undetermined significance (B-MLUS).
Leukemia
1989
;
3
:
501
–4.
29
Scrivener S, Goddard RV, Kaminski ER, Prentice AG. Abnormal T-cell function in B-cell chronic lymphocytic leukaemia.
Leuk Lymphoma
2003
;
44
:
383
–9.
30
Bacik I, Snyder HL, Anton LC, et al. Introduction of a glycosylation site into a secreted protein provides evidence for an alternative antigen processing pathway: transport of precursors of major histocompatibility complex class I-restricted peptides from the endoplasmic reticulum to the cytosol.
J Exp Med
1997
;
186
:
479
–87.
31
Wood P, Elliott T. Glycan-regulated antigen processing of a protein in the endoplasmic reticulum can uncover cryptic cytotoxic T cell epitopes.
J Exp Med
1998
;
188
:
773
–8.
32
Levitsky HI, Montgomery J, Ahmadzadeh M, et al. Immunization with granulocyte-macrophage colony-stimulating factor-transduced, but not B7–1-transduced, lymphoma cells primes idiotype-specific T cells and generates potent systemic antitumor immunity.
J Immunol
1996
;
156
:
3858
–65.
33
Rosen HR, Stierer M, Wolf HM, Eibl MM. Impaired primary antibody responses after vaccination against hepatitis B in patients with breast cancer.
Breast Cancer Res Treat
1992
;
23
:
233
–40.
34
Neelapu SS, Kwak LW, Kobrin CB, et al. Vaccine-induced tumor-specific immunity despite severe B-cell depletion in mantle cell lymphoma.
Nat Med
2005
;
11
:
986
–91.
35
Trojan A, Schultze JL, Witzens M, et al. Immunoglobulin framework-derived peptides function as cytotoxic T-cell epitopes commonly expressed in B-cell malignancies.
Nat Med
2000
;
6
:
667
–72.
36
Baskar S, Kobrin CB, Kwak LW. Autologous lymphoma vaccines induce human T cell responses against multiple, unique epitopes.
J Clin Invest
2004
;
113
:
1498
–510.