Optimizing dendritic cell-based immunotherapy in multiple myeloma: intranodal injections of idiotype-pulsed CD40 ligand-matured vaccines led to induction of type-1 and cytotoxic T-cell immune responses in patients

Authors


Qing Yi MD, PhD, Department of Lymphoma and Myeloma, The University of Texas M.D. Anderson Cancer Center, 1515 Holcombe Boulevard, Unit 0903, Houston, TX 77030, USA. E-mail: qyi@mdanderson.org

Summary

Vaccination with idiotype (Id) protein-pulsed dendritic cells (DCs) has been explored in multiple myeloma and the results have been disappointing. To improve the efficacy of DC vaccination in myeloma, we investigated the use of Id- and keyhole limpet haemocyanin (KLH)-pulsed, CD40 ligand-matured DCs administered intranodally. Nine patients with smouldering or stable myeloma without treatment were enrolled and DC vaccines were administered at weekly intervals for a total of four doses. Following vaccination, all patients mounted Id-specific γ-interferon T-cell response. Interleukin-4 response was elicited in two, and skin delayed-type hypersensitivity reaction occurred in seven patients. More importantly, Id-specific cytotoxic T-cell responses were also detected in five patients. Most if not all patients mounted a positive T-cell response to KLH following vaccination. At 1-year follow-up, six of the nine patients had stable disease, while three patients had slowly progressive disease even during the vaccination period. At 5-year follow-up, four of the six patients continued with stable disease. No major side effects were noted. In summary, intranodal administration of Id-pulsed CD40 ligand-matured DCs was able to induce Id-specific T and B-cell responses in patients. Current efforts are geared towards breaking tumour-mediated immune suppression and improving clinical efficacy of this immunotherapy.

Multiple myeloma (MM) is a B-cell neoplasia characterized by the infiltration of malignant plasma cells in the bone marrow of patients. Although high-dose chemotherapy has improved the clinical outcome, many patients still relapse (Barlogie et al, 1999). Thus, additional measures are needed after transplantation to eliminate minimal residual disease, a scenario in which immunotherapy has been explored.

Dendritic cells (DCs) are the most potent antigen-presenting cells (APCs) equipped with the necessary costimulatory, adhesion and major histocompatibility complex (MHC) molecules needed for the initiation of a primary immune response (Banchereau et al, 2000). DCs are able to prime naive T cells, stimulate CD8+ cytotoxic T lymphocytes (CTLs) directly (Young & Inaba, 1996) and, by secretion of interleukin (IL)-12, polarize the immune response towards a type-1 T-cell response (Zitvogel et al, 1996). These properties make DC ideally suited to serve as a natural adjuvant for the purpose of cancer immunotherapy (Young & Inaba, 1996). Animal studies have already demonstrated that tumour antigen-pulsed DCs are capable of inducing a protective and therapeutic anti-tumour immunity (Mayordomo et al, 1995; Celluzzi et al, 1996; Zitvogel et al, 1996), which has prompted clinical trials in human melanoma, colon, prostate and breast cancers (Hsu et al, 1996; Nestle et al, 1998; Kugler et al, 2000). However, DC-based vaccination trials have thus far failed to prove increased benefit compared with standard chemotherapy (Eubel & Enk, 2009). Nevertheless, a promising result (survival benefit) has been obtained from a phase III randomized clinical trial in patients with metastatic hormone-refractory prostate cancer who received antigen-loaded DCs (Provenge) (McKarney, 2007). These results indicate that DC-based immunotherapy may be clinically efficacious but its effects need to be improved.

The monoclonal immunoglobulin (Ig) (M-protein) secreted by myeloma cells carries unique antigenic determinants (idiotype; Id) (Yi, 2003a). Immunotherapy with Id-pulsed DCs has been explored in MM and the results have been disappointing. Less than 50% of patients mounted Id-specific immune responses, and clinical responses have rarely been observed (Lim & Bailey-Wood, 1999; Reichardt et al, 1999; Liso et al, 2000; Titzer et al, 2000). To improve the efficacy of DC vaccination in MM, we have investigated the use of intranodal administration of myeloma antigen-pulsed, CD40 ligand (CD40L)-matured DCs in smouldering or stable myeloma patients. In this study, we report the results of nine patients receiving Id-pulsed DC vaccines.

Materials and methods

Patients

Table I lists the characteristics of the nine patients included in this study. At study entry, all patients had smouldering MM or stable disease requiring no treatment. The University of Arkansas for Medical Sciences Institutional Review Board-approved informed consent was obtained from all patients. Median patient age was 58·3 years (range 42·7–72·8), all had Southwestern Oncology Group performance status scores of 0–1, and none had signs of active infection or inflammatory disease. Clinical evaluation of the disease was done before, during and after vaccination, by examination of routine blood count, chemistry, electrophoresis of serum and urine immunoglobulins, and of bone marrow aspirates.

Table I.   Characteristics of patients.
PatientsSexAge (years)Disease status at enrollmentM-proteinM-spike (g/l)*β2M (mg/l)*BM PC (%)*Δ13AutoTxTime to vaccine from Tx (months)CD4 count (×109/l)*
  1. Δ13, chromosome 13 abnormalities; β2M, β2-microglobulin; BM PC, bone marrow plasma cells; AutoTx, autologous transplant; NA, not applicable; SD post AT, stable disease post autologous transplantation.

  2. *Value at study entry.

1F43SmoulderingIgG/λ351·9200NA0·659
2F52SmoulderingIgA/λ101·7130NA1·306
3M68SmoulderingIgG/λ442·3320NA0·684
4M50SmoulderingIgA/κ8·01·8100NA1·156
5F67SmoulderingIgG/222·2300NA1·072
6F75SmoulderingIgA/6·03·2220NA0·970
7M63SmoulderingIgG/κ132·8200NA0·509
8M54SmoulderingIgG/322·420+0NA0·605
9M45SD post ATIgG/171·452740·774

Purification of Id protein

Plasma was collected from patients by plasmapheresis. To remove other proteins, plasma was precipitated in 50% ammonium sulphate (Sigma, St Louis, MO, USA), resuspended in isotonic saline, and extensively dialysed against isotonic saline to remove ammonium sulphate. IgG Id proteins were purified by sterile MabTrapG columns® (Pharmacia Biotech AB, Uppsala, Sweden) (Bergenbrant et al, 1996), and IgA Id proteins were purified by affinity chromatography columns prepared with anti-human IgA (α-chain specific, Sigma) monoclonal antibody-conjugated agarose (Sigma). The purity of the monoclonal IgG and IgA fraction was confirmed by sodium dodecyl sulphate polyacrylamide gel electrophoresis (SDS-PAGE) to be >95%. The purified Id protein fraction was dialysed against sterile NaCl overnight, followed by filtration through a Millipore filter (0·22 μm). Samples to be used for vaccination were tested for endotoxins by the Limulus assay (QCL-1000, BioWhittaker, Walkersville, MD, USA), as well as for bacterial, fungal and mycoplasma contamination. Specimens containing more than 5 endotoxin units (EU)/ml endotoxin were not used.

Ex vivo generation of DCs

Peripheral blood mononuclear cells (PBMCs) harvested by leukapheresis were used to generate DCs, and the clinical-grade Id-preloaded DC vaccines were prepared as described previously (Szmania et al, 2005). Briefly, PBMCs (108) were added to 75 cm2 tissue culture flasks and allowed to adhere for 2 h. Non-adherent cells were removed by gentle washing and the remaining adherent cells were then cultured in AIM V medium (Gibco/Invitrogen, Grand Island, NY, USA) containing 1000 u/ml granulocyte-monocyte colony-stimulating factor (GM-CSF; Immunex-Amgen, Thousand Oaks, CA, USA) and IL-4 (CellGenix, Antioch, Il, USA). On days 3 and 5, 50% of medium was replaced by fresh DC medium containing twofold higher concentration of the cytokines (resulting in a final concentration of 1000 u/ml GM-CSF and IL-4). On day 6, the DC culture volume was reduced to half to conserve Id protein, which was added to the culture at a final concentration of 100 μg/ml. The DCs were incubated overnight in the presence of Id and keyhole limpet haemocyanin (KLH; 50 mg/ml) which provided T-helper epitopes and served as a neotracer adjuvant antigen (Shimizu et al, 2001). On day 7 the culture volume was brought back to 20 ml per flask with AIM V medium containing GM-CSF, IL-4, and 500 ng/ml trimeric CD40-ligand (Immunex-Amgen), to induce DC maturation for 48 h. On day 9, antigen-pulsed mature DCs were harvested, counted, and cryopreserved for infusion. All DCs were tested for bacteria, fungi, and endotoxin. DC release criteria comprised of negative microbial cultures, negative for mycoplasma testing, and endotoxin levels <5 EU/ml or 350 EU infused/ml/h per vaccine.

Vaccination schedule

Each patient received four intranodal DC vaccines on days 1, 14, 21, and 28. Intranodal injections of the vaccines into inguinal lymph nodes were performed by a radiologist under ultrasound guidance. To enhance the efficacy of vaccination, a low dose of recombinant IL-2 (0·2 × 106 iu/injection, Chiron, Emeryville, CA, USA) was given subcutaneously for 5 consecutive days following each DC vaccination (Shimizu et al, 1999). Prior to, during and after vaccination, patients were followed up at regular intervals and immune responses to vaccines monitored closely.

Immunofluorescence

A two-colour immunofluorescence assay was applied to all cultured cells and PBMCs from patients. All stains and washes were performed at 4°C. The following mouse monoclonal antibodies were used: fluroscein-isothiocyanate (FITC)-conjugated antibodies against CD14, CD40, CD80, CD86, HLA-ABC and -DR (Becton-Dickinson, Mountain View, CA, USA); phycoerythrin (PE)-conjugated antibodies against CD1a, CD3, CD4, CD8, CD16, CD19, CD25, CD83 (Becton-Dickinson). Cells were incubated for 30 min with FITC- and PE-conjugated antibodies, washed with phosphate-buffered saline, and analysed by a FACScan flow cytometer (Becton-Dickinson).

Proliferation assay

Fresh PBMCs were isolated from blood of patients and resuspended in complete medium. A 96-well round-bottomed microtitre plate was used. PBMCs (1 × 105 cells per well) were added to each well and incubated with 10 μg/ml of the autologous Id or an isotype-matched allogeneic protein for 6 d with 5% CO2 at 37°C. Cells incubated with medium only or with KLH (10 μg/ml) were used as controls. Eighteen hours before harvest, 37 kBq/well of 3H-thymidine (Amersham, Piscataway, NJ, USA) was added. The cells were collected using a Skatron combi cell harvester (Skatron A/S, Lier, Norway), and radioactivity was measured in a liquid scintillation counter. The results are expressed as mean count per min (cpm) of triplicates. Stimulation index (SI) was calculated by dividing the mean cpm of stimulated cells by that of unstimulated cells. SI exceeding 2 was considered positive, which is in accordance with our previous results (Yi et al, 1995).

Detection of anti-Id B cells

Details of an enzyme-linked immunospot (ELISPOT) assay used for the detection of B cells secreting IgM antibodies binding to Id were described earlier(Bergenbrant et al, 1996). PBMCs (2 × 105 cells per well) were added and incubated overnight. After incubation, cells were washed away and plates were incubated with biotinylated goat anti-human IgM antibody (Sigma), followed by avidin-biotin complex (ABC Vectastain-Elite kit) and peroxidase staining. All samples were tested in duplicate. The data are expressed as mean numbers of cells/106 PBMCs.

Detection of Id-specific interferon (IFN)-γ or IL-4 secreting T cells

The ELISPOT assay for the detection of Id-specific, IFN-γ or IL-4 secreting T cells was used as described earlier (Yi et al, 1993, 1995). Briefly, plates were coated with monoclonal anti-human IFN-γ or IL-4 antibodies (R&D Systems) at 4°C overnight, and 200 μl of fresh PBMCs (1 × 105 cells per well) were added. Cells were incubated with 10 μg/ml of the autologous Id or an isotype-matched allogeneic protein for 48 h at 37°C with 5% CO2 and high humidity. Cultures without additions were used to determine the spontaneous secretion of cytokines. Cells stimulated with KLH (10 μg/ml) served as positive controls. To visualize spots corresponding to cytokine-secreting cells, PBMCs were detached from the plates by washing. The wells were incubated with rabbit polyclonal anti-human IFN-γ or IL-4 (R&D Systems), followed by biotinylated anti-rabbit IgG (Vector Laboratories, Burlingame, CA, USA), avidin-biotin peroxidase complex (ABC Vectastain-Elite kit, Vector Laboratories), and peroxidase staining, using the substrate 3-amino-9-ethylcarbazol (Sigma). Spots corresponding to cytokine-secreting cells were enumerated blindly under a dissection microscope. All samples were tested in duplicate.

The data are expressed as mean number of cytokine-secreting cells/105 PBMCs. The stimulated number of spots (calculated as the total number of spots minus the number of spots in culture medium without additions) was used for data presentation. The mean +2 standard deviations (SD) of stimulated number of spots in patients using isotype proteins was 9 (range 0–18) for IFN-γ-, and seven spots (0–14) for IL-4-secreting cells. A stimulated number of cytokine-secreting cells exceeding nine spots was considered a positive response to antigen stimulation.

Cytotoxicity assay

The standard 4-h 51Cr-release assay (Qian et al, 2005) was performed to measure cytolytic activity of the T cells against target cells including autologous DCs pulsed with or without Id protein and primary myeloma cells isolated from patients. Target cells were incubated with 3·7 MBq of 51Cr-sodium chromate for 1 h, washed extensively, seeded (1 × 104 cells per well) into 96-well U-bottomed plates in AIM V medium, and cocultured for 4 h with various numbers of patient’s PBMCs collected before or 1 week after the fourth vaccination and pre-stimulated with autologous Id protein (10 μg/ml) for a week in AIM V medium containing IL-2 (20 u/ml). All assays were performed in triplicates. Results are expressed as mean percentage of 51Cr release calculated as described previously (Qian et al, 2005). Spontaneous release was <20% of the maximum 51Cr uptake.

Delayed-type hypersensitivity (DTH) skin tests

DTH tests were performed with unpulsed DCs and DCs pulsed with autologous Id proteins or KLH before the vaccination and at 1 week after the fourth vaccination. DCs (5 × 105 cells) were injected intradermally into the forearm. A positive skin-test reaction was defined as >5 mm diameter erythema and induration 48 h after injection.

Statistical analysis

For analysis of the immune response, values before and after vaccination in the patients were used and compared by the Student t test. Generation or enhancement of a relevant immune response in each patient was defined as ≥1·5-fold increase over the values of pre-vaccination and observed in two consecutive follow-ups in Id-specific proliferation and cytokine-secreting cells in the ELISPOT assay. Significance was set at < 0·05.

Results

DC vaccines

DC vaccines were successfully prepared from all nine patients, and Table II lists the numbers of Id-pulsed DCs injected to the patients. Flow cytometric analysis confirmed that these cells were CD14/CD83+/CD40+/CD80+/CD86+/CD54+/HLA-DR+. Functionally, these cells were able to induce strong alloreactive T-cell responses as we described previously (Szmania et al, 2005).

Table II.   Numbers of DCs injected.
PatientsVaccine 1 (×106)Vaccine 2 (×106)Vaccine 3 (×106)Vaccine 4 (×106)
17·42·02·02·2
22·71·22·43·1
34·25·54·83·9
424·724·515·821·0
535·631·013·535·5
64·816·016·817·0
713·28·110·011·1
812·513·014·913·5
911·011·018·814·5

Intranodal injections of DC vaccines

To improve the chance for Id-pulsed DCs to interact with specific T cells, we explored an intranodal route of vaccine administration to patients. This was carried out by a radiologist under ultrasound guidance. For each vaccination, the cells were injected into one or two lymph nodes in the groyne.

KLH-specific immune responses

To improve the potency of the Id DC vaccines and to measure the immune competence of the patients, KLH was also used to pulse DCs together with Id proteins. Therefore, we monitored patient’s immune responses to KLH before and after DC vaccination. As shown in Table III and Fig 1, all patients mounted positive proliferative (Fig 1A) and/or IFN-γ (Fig 1B) T cell responses to KLH after the vaccination. Six of the nine patients also had a positive IL-4 T-cell response (Fig 1C). These data suggest that these myeloma patients were immune competent and able to mount antigen-specific immune responses after immunization.

Table III.   Patient’s post-vaccination immune and clinical responses at 5-year follow-up.
PatientsB cellsProliferationIFN-γIL-4CTLDTH5 year follow up
IdKLHIdKLHIdKLHIdIdKLHIdDisease statusTherapy
  1. PR, partial response; SD, stable disease; PD, progressive disease; ND, not done; Chemo + Tx, chemotherapy with autotransplantation; Chemo, chemotherapy without Tx.

  2. *At 1-year follow-up.

1++++++++PR/SD*None
2+++++++SDNone
3+++++++PDChemo + Tx
4+++++++++SDNone
5++++SDNone
6++++++++SDNone
7++++++++PDChemo
8++++PDChemo + Tx
9+++++++NDNDPDChemo
Figure 1.

 KLH-specific T cell responses following DC vaccination. (A) Proliferative; (B) IFN-γ-secreting, or (C) IL-4-secreting T cell responses in nine patients before (week 0) and at different time points after the first DC vaccination. Shown are stimulated numbers of IFN-γ or IL-4 spots per 105 PBMCs.

Id-specific B-cell response

Consistent with our previously published results(Bergenbrant et al, 1996; Yi et al, 2002), low numbers of B cells secreting IgM antibodies reacting to Id protein (anti-Id B cells) were detected in most of the patients before the treatment (Fig 2A). DC vaccinations induced a significant increase in the number of anti-Id B cells in the nine patients. The differences compared with pre-vaccination values in the nine patients were statistically significant (< 0·05 or < 0·01) at weeks 4, 10, 20 and 40. The number of B cells secreting antibodies binding to control isotype protein was low (<3 cells per 106 PBMCs in most patients), and no difference was observed in such B cells during and after DC vaccination (data not shown).

Figure 2.

 Id-specific B and T cell responses following DC vaccination. (A) Numbers of anti-Id B cells per 106 PBMCs; (B) Id- and (C) isotype-induced proliferative response (stimulation index) of T cells in nine patients before (week 0) and at different time points after the first DC vaccination.

Id-specific proliferative response

We measured Id-specific proliferative response of T cells by testing PBMCs from the patients before and after each vaccination. No patient had a positive proliferative response to the Id protein pre-vaccination. After the treatment a positive proliferative response was generated in four patients (Patients 1, 4, 7 and #; Fig 2B), which was specific for autologous Id and was MHC restricted (data not shown). In Patients 1 and 4, T cells also responded weakly to isotype-matched allogeneic Id proteins at weeks 10 or 50, respectively (Fig 2C). Thus, DC vaccination induced an Id-specific T-cell proliferative response in four out of nine patients vaccinated.

Id-specific cytokine-secreting T cells

To examine the types of T-cell responses induced by Id-pulsed DC vaccines, Id-specific IFN-γ- (type-1) and IL-4- (type-2) secreting cells were enumerated in patients before and after the vaccination. Prior to the treatment, all patients had low numbers of Id-induced IFN-γ- and IL-4-secreting cells. This is consistent with our previous studies showing that Id-specific T cells could be detected in many myeloma patients (Yi et al, 1993, 1995, 2002). DC vaccination induced or enhanced an IFN-γ response in all nine patients although the numbers of T cells became lower by weeks 40 and 50 after the vaccination (Fig 3A). The differences comparing with pre-vaccination values in the nine patients were statistically significant (< 0·05) at weeks 4, 10, and 20. The induced IFN-γ response was specific for Id because isotype-matched allogeneic Id protein-induced responses were low (Fig 3B). These results indicate that intranodal vaccination with Id-pulsed mature DCs was able to enhance Id-specific IFN-γ T-cell responses in all vaccinated patients but the response gradually faded away after vaccination.

Figure 3.

 T-cell cytokine response of vaccinated patients. Shown are stimulated numbers of spots per 105 PBMC representing (A) Id- and (B) isotype-induced IFN-γ-; and (C) Id- and (D) isotype-induced IL-4-secreting cells in nine patients before (week 0) and at different time points after the first DC vaccination. The mean + SD of stimulated numbers of IL-4 spots for isotype controls were 2·8 + 2·1, and mean + SD of stimulated numbers of IFN-γ spots for isotype controls were 4·1 + 2·3 in all patients.

No significant changes were observed in Id- (Fig 3C) and isotype- (Fig 3D) induced IL-4-secreting cells in these patients. None of the patients had detectable (above the cut-off level) Id-specific IL-4-secreting T cells in blood before vaccination, and DC vaccination slightly enhanced the numbers of Id-specific IL-4-secreting cells in Patients 7 and 9. However, no statistical difference was found in the patients in terms of IL-4-secreting cells before and after vaccination. Thus, these results indicate that DC vaccination induced or enhanced predominantly Id-specific IFN-γ, but not IL-4, responses in the patients.

Id-specific cytotoxic T-cell response

We also examined whether Id-pulsed DC vaccines induced an Id-specific CTL response in vaccinated patients. As the frequency of such T cells would be very low, we restimulated PBMCs collected from the patients before vaccination and 1 week after the fourth vaccination with autologous Id protein in AIM V medium in the presence of IL-2 for 1 week. The T cells were then collected, washed, and assayed for their capacity to lyse autologous primary myeloma (if available) or Id-pulsed autologous DCs. As shown by the representative results from Patient 1, restimulated PBMCs collected post- but not pre-vaccination (data not shown) showed specific T-cell proliferative response to Id-pulsed but not to isotype-pulsed or unpulsed autologous DCs (Fig 4A). More importantly, the T cells effectively lysed autologous primary myeloma cells and Id-pulsed DCs but not allogeneic primary myeloma cells or unpulsed or isotype-pulsed autologous DCs (Fig 4B). Similar results were also obtained with T cells from Patient 2, from whom primary myeloma cells were available (Fig 4C). Figure 4D shows the cytolytic activity of restimulated T cells from the nine patients against Id-pulsed autologous DCs. It is evident that an induction of or enhanced Id-specific CTL responses were observed in five out of nine patients.

Figure 4.

 Induction of Id-specific CTL responses in vaccinated patients. (A) Proliferative response of restimulated T cells in PBMCs from Patient 1 induced by unpulsed autologous DCs or DCs pulsed with Id or isotypic proteins. Different T:DC ratios were used; Cytotoxic activity of restimulated T cells from (B) Patient 1 and (C) Patient 2 against autologous primary (Auto-PC) or allogeneic (Allo-PC) myeloma cells, or unpulsed autologous DCs or DCs pulsed with Id or isotypic proteins. Two different effector:target (E:T) ratios were used; PBMCs were collected 1 week post the 4th DC vaccination. Primary myeloma cells were available only from two patients for the test. (D) Cytotoxic activity of restimulated T cells from nine patients in PBMCs collected before (pre-vaccination) and post-vaccination (1 week after the 4th vaccination). An E:T ratio of 10:1 was used. (E) The representative image of skin DTH reaction to unpulsed autologous DCs or DCs pulsed with Id protein or KLH from patient #7.

DTH skin tests

To examine in vivo priming or enhancement of Id-specific immune response to DC vaccines in patients, DTH skin test by injecting DCs pulsed with Id or KLH proteins was performed. Injection of unpulsed, autologous DCs was used as a control. None of the patients showed a positive skin reaction to Id- or KLH-pulsed or unpulsed DCs before vaccination. Figure 4E depicts the representative skin DTH reaction to Id- and KLH-pulsed DCs but not to unpulsed DCs (Patient 7) after DC vaccination. Among eight patients examined, seven mounted a positive skin DTH reaction to Id-pulsed DCs or to KLH-pulsed DCs. However, among seven patients with a positive skin reaction to Id-pulsed DCs, five also showed a positive but weaker (compared with that of Id-pulsed DCs) skin reaction to unpulsed DCs (data not shown).

Clinical follow-up

DC vaccination was well tolerated with no significant side effects observed in any of the patients throughout the treatment period. All patients had measurable M-proteins in serum when this study was initiated. At year-1 follow up from the start of DC vaccine, the level of M-protein in Patient 1 was gradually reduced, from 35 to 25 g/l (Fig 5). No significant change was observed in Patients 2, 4, 5, 6 and 7. However, Patients 3, 8 and 9 showed sign of disease progression during or after DC vaccination. Disease status, updated at January 1 2010, which was more than 5 years following vaccination, was as follows; Patients 2,4,6,7 continued with stable disease and were off treatment whereas Patients 3, 5, 8, 9 had progressive disease starting at 8, 88, 5, or 15 months respectively, after enrollment (Table III). No significant change was noted in circulating CD4+ T cells or CD19+ B cells, uninvolved Ig and percentages of bone marrow plasma cells (data not shown). All patients were maintained without other treatment.

Figure 5.

 Clinical status of vaccinated patients at 1-year follow-up. Shown is the change of serum M-protein (g/l) in these patients before and at different time points since the first DC vaccination.

Discussion

DCs are the most potent APCs and exist in two main stages of maturation. Immature DCs are effective in taking up and processing native protein antigens but are less potent at activating T cells; conversely, mature cells lose antigen-capturing capacity but are more effective in stimulating resting CD4+ and CD8+ T cells to grow and differentiate (Banchereau et al, 2001). CD40 activation, induced by antibodies to CD40 or CD40L binding, has been shown to induce DC maturation (Banchereau & Steinman, 1998), condition DCs to activate CD8+ CTLs directly without CD4+ T cells (Ridge et al, 1998), break T-cell tolerance (Diehl et al, 1999), and provoke immunity to tumours in mice (French et al, 1999). Thus, to induce an optimal immune response after DC vaccination, antigen-pulsed mature DCs are necessary to utilize both the capacity of immature cells to take up antigens and the ability of mature cells to activate specific T cells. The use of CD40L to mature and condition DCs may be important in this study because in cancer patients, such as those with MM, the function of T-helper (Th) cells may be impaired (Frassanito et al, 1998; Van den Hove et al, 1998) and tumour-specific T cells may be anergized (Bogen et al, 2000). Vaccination with CD40L-conditioned DCs may overcome these problems.

The aim of this study was to improve the efficacy of DC vaccination in MM by utilizing Id- and KLH-pulsed, CD40L-matured DCs. To ensure that the infused DCs would home to lymph nodes and interact with specific T cells, we injected the DC vaccines intranodally. To ensure that patients had an equate immune system, patients with smouldering MM or with stable disease were recruited, which is different from previous studies in which patients with minimal residual disease after high-dose chemotherapy and transplantation were enrolled (Lim & Bailey-Wood, 1999; Reichardt et al, 1999; Liso et al, 2000; Yi et al, 2002). To support newly activated T cells induced by the vaccination, low doses of IL-2 were injected subcutaneously. We do not believe that the low doses of IL-2 would affect myeloma growth because it was injected subcutaneously. Furthermore, previous studies injecting subcutaneously much higher doses of IL-2 for a long period of time in myeloma patients did not show clinical antimyeloma efficacy (Peest et al, 1995). Compared to our previous studies of DC vaccination with maximal doses of Id-pulsed, cytokine (TNF-α and IL-1β)-matured DCs injected subcutaneously (Yi et al, 2002), the present study demonstrated that intranodal injection of Id-pulsed, CD40L-mature DCs was more effective at inducing Id-specific immune response in myeloma patients. This is attested to by the development of a specific, MHC-restricted T-cell response in all nine vaccinated patients, measured as increased numbers of Id-induced IFN-γ-secreting cells in all, a positive T-cell proliferation in 4, DTH in 7, and IL-4 in two of the nine patients. Based on the cytokine secretion pattern (high IFN-γ and low IL-4 secretion), the induced T-cell response was mainly the type-1 T-cell response (Mosmann & Sad, 1996). An anti-Id B-cell response was induced or enhanced in these patients. More importantly, Id-specific CTL response was also induced in five of the nine patients. Thus, these results indicate that the vaccination induced both cellular and humoral immune responses in all vaccinated patients, although the cohort size is small.

Although the primary endpoint of this study was the induction of Id-specific immune responses, clinical response in terms of partial or complete remission or stable disease in vaccinated patients was the secondary endpoint. As shown in Table III, at 5-year follow-up, four of the patients had stable disease. Patient 1 had a partial response with serum M-protein reduced by 30% at 1 year follow-up, and the patient was unfortunately lost to follow-up thereafter. Although four of the patients had progressive disease, three of the four patients had already shown signs of disease progression during DC vaccination. We did not find a correlation between immune responses and clinical outcome. Interestingly however, DC vaccination failed to induce an Id-specific CTL response in all three patients whose disease was progressing during vaccination, while only one of six patients with stable disease did not generate the response following the vaccination. This may suggest that the development of Id-specific CTL responses following DC vaccination may be relevant to a clinical anti-tumour response. This notion is supported by our and other studies demonstrating that Id-specific CTLs are able to lyse autologous primary myeloma cells in vitro (Li et al, 2000; Wen et al, 2001).

For the past decade, Id-based immunotherapy has been actively explored in B-cell malignancies, such as follicular B-cell lymphoma and MM, for the purpose of developing an additional therapy that can be used to control or eradicate the minimal residual disease after high-dose chemotherapy in patients(Kwak et al, 1992; Bendandi et al, 1999; Massaia et al, 1999; Neelapu et al, 2005) (Bergenbrant et al, 1996; Osterborg et al, 1998). The results from these studies demonstrated clearly that Id-specific immunity can be generated in many patients, including those with smouldering disease (Bergenbrant et al, 1996; Osterborg et al, 1998) and also those with advanced disease after high-dose chemotherapy (Lim & Bailey-Wood, 1999; Massaia et al, 1999; Reichardt et al, 1999; Liso et al, 2000). However, despite the promising results obtained in B-cell lymphoma (Hsu et al, 1996, 1997; Bendandi et al, 1999; Neelapu et al, 2005), a clinical anti-tumour response has only been anecdotally observed in vaccinated myeloma patients (Osterborg et al, 1998; Yi et al, 2002). This may raise a question as to whether myeloma patients may be less responsive to Id vaccination, which can possibly be attributed to the inhibitory effect of high levels of circulating Id proteins on specific T cells (Bogen, 1996) and the fact that Id protein is a weak antigen (Yi, 2003b). More importantly, myeloma cells are protected in vivo by the bone marrow microenvironment, which consists of matrix, stromal cells, and cytokines, many of which are immune suppressive (Anderson, 2007). Therefore, our future studies will focus on combinational immunotherapies targeting both myeloma cells and the suppressive tumour microenvironment.

Kyle and Greipp (1980) first described smouldering MM as a distinct clinical entity by the presence of a serum M-protein value higher than 30 g/l, bone marrow clonal plasma cells involvement of 10% or higher, and no bone lytic lesions or clinical manifestations attributed to the monoclonal plasma-cell proliferative disorder. In 2003, the International Myeloma Working Group agreed on a new definition of smouldering MM consisting of a serum M-protein of ≥30 g/l and/or ≥10% bone marrow plasma cells with no evidence of end-organ damage (hypercalcemia, renal insufficiency, anaemia or bone lesions) (Blade et al, 2010). The large majority of patients with smouldering MM will evolve into symptomatic MM and require treatment. The median time to progression has ranged between 2 and 3 years (Wisloff et al, 1991; Dimopoulos et al, 1993; Facon et al, 1995). We recruited these patients for our study because they represent an early stage of the disease, require no treatment, and their immune systems are almost intact. Based on the clinical data presented in Table III, it is possible that these patients might have benefited from Id-pulsed DC vaccination because 5-year long stable disease has been observed in four of the patients. However, as this is a single arm study with a small number of patients, definitive data may be obtained from two-arm randomized studies that should be performed in future.

To conclude, the present study demonstrated that immunotherapy with Id-pulsed, CD40L-matured DCs administered intranodally is efficient at inducing specific immune responses in myeloma patients. Future immunotherapy in MM must take into consideration the selection of patients who can mount proper immune responses against antigen challenge, include other myeloma antigens to recruit a broader repertoire of specific T cells, combine different immunization strategies such as more potent immune adjuvant CpG, as well as develop a new generation of immunotherapies that target other aspects of tumour cells and/or its microenvironment so that it will become a part of routine therapeutic intervention for malignancies.

Acknowledgements

This work was supported by grants from the National Cancer Institute (PO1 CA55819, R01 CA96569, R01 CA103978), Multiple Myeloma Research Foundation, the Leukemia and Lymphoma Society, and Commonwealth Foundation for Cancer Research.

Authors’ contributions

QY and FVR designed the study, analysed data, wrote the manuscript and gave final approval for the submitted manuscript. SS, JF, JQ, and NAR performed the experimental research. SV performed intranodal injection of the vaccines, MCF supervised GMP production of the vaccines. BB and GT helped recruiting patients and provided critical suggestions.

Ancillary