We are international

Vaccine Strategies in Multiple Myeloma
By Håkan Mellstedt, M.D., Ph.D., Karolinska Institute


Myeloma cells express several tumor antigens that might be candidates for use in vaccine development. At present, focus is on idiotypic proteins. These proteins, coupled to carriers or loaded onto autologous dendritic cells, have been used for vaccination together with adjuvant cytokines. Immune responses as well as clinical effects have been noted in early trials. A low tumor burden and a well-preserved immune system may be prerequisites for success of the vaccination approach. In the future, improved vaccine preparations and protocols will facilitate the development of a therapeutic vaccination strategy. Large clinical trials are needed to establish the clinical efficacy of idiotype vaccination. Such trials, however, should not be initiated until a promising vaccination protocol has been developed. With this in mind, it is anticipated that by the end of this decade vaccination will be an established therapeutic approach in myeloma.


The tumor clone in multiple myeloma consists of monoclonal myeloma B cells (plasma cells and late B-lymphoid cells) synthesizing the immunoglobulin of the B cell of origin. The idiotypic determinants of the immunoglobulin can be considered a well-defined tumor antigen. Peptides of the idiotypic region, complementarily-determining regions (CDRs) and framework regions, can be presented on MHC class II and I molecules of myeloma cells and thus be recognized by both CD4 and CD8 lymphocytes.

Antigens that may induce spontaneous immunity might be used preferentially as vaccine candidates. The idiotypic immunoglobulin region of myeloma cells can induce natural T-cell immunity, including cytotoxic T lymphocytes that might lyse the autologous tumor cells. This indicates that the idiotype might be an immunodominant epitope and a good candidate for use in vaccine development. Furthermore, as different peptide sequences of the complete V region can be presented on MHC molecules, the whole idiotype should be used as vaccine and not only CDRs. 

The induction of idiotypic immunity is probably induced by the secreted myeloma protein taken up by dendritic cells (DCs), which present MHC class II- and I-restricted peptides to the immune system and induce CD4 and CD8 responses. The relative contribution of different T-cell subsets to the eradication of myeloma cells is not clear. Both CD4 and CD8 T-cells have been considered important to tumor rejection, and both subsets probably contribute. Dr. Freda Stevenson of the University of Southhampton in England proposed that there might be a CD4 tolerance for the idiotype induced by the long-term presence of tumor cells. However, in colon carcinoma with a high secretion of CEA, which might be analogous to myeloma, it had been claimed that there is CD8 T-cell tolerance. These contradictory results probably indicate that we do not have complete information and that there might be great variation, supporting the notion that a broad cellular immune response should be the goal of vaccination.

There are other antigens besides the idiotype that can be used as vaccine candidates, eg, mucins. At present, however, the focus is entirely on clone-specific idiotypic immunoglobulin.

Induction of immunity in myeloma patients, which should have the potential to eradicate tumor cells, is a great challenge. Dr. Mellstedt noted that the goal is not only to have a good antigen presented in a proper way but also to have a well-functioning immune system. There might be many hurdles to overcome in developing a vaccination strategy. Dr. Massaia of the University of Torino in Italy suggested that the antigen-presenting cells and DCs may be defective in myeloma. Patients with myeloma, even at a low stage, who are chemotherapy naive have immune T-cell dysfunction measured as a low capability to secret cytokines and a defect expression of T-cell–signaling molecules. This information is important to the further clinical development of effective vaccines, as measures have to be taken to correct defects in T-cell signaling. This might be achieved by simultaneous administration of cytokines and high doses of vitamin E, which have also been suggested to correct T-cell dysfunction.

Collectively, these prevaccination results indicate that the idiotype is an appropriate antigen, but it is important to select patients with well-preserved immune systems to be able to induce adequate immune responses. 

The idiotype can be delivered as a protein antigen, as protein loaded onto isolated autologous dendritic cells, or as DNA presented in a plasmid. There is debate regarding which is the best vaccine preparation, protein or plasmid DNA. Further clinical development will reveal which antigen preparation is the most powerful in inducing clinically effective immunity. When one uses the idiotypic protein as an antigen, it is clear that granulocyte-macrophage colony-stimulating factor (GM-CSF) is necessary; otherwise, CD4 and CD8 T-cell response will not be induced. In a preclinical model described by Dr. Stevenson, plasmid-expressing idiotype DNA alone (producing idiotypic protein in vivo) did not have a clinical effect. Fusion to fragment C of tetanus toxid was necessary; this construct induced protective immunity; CD4 cells were important. Plasmid-idiotype DNA and GM-CSF were not tested in this model. Idiotype fused to fragment C of tetanus toxid was necessary, but GM-CSF might have a similar effect. Fusion of fragment C of tetanus toxid to CEA was necessary to induce an anti-CEA response, whereas plasmid CEA alone had no effect. Dr. Mellstedt reported that vaccination with protein CEA alone induced no or only weak immunity, but protein CEA in combination with GM-CSF induced a strong humoral (IgG) response and CD4 as well as CD8 T-cell responses. It has also been shown that GM-CSF facilitated the MHC class I antigen-presentation pathway and, thereby, the induction of cytotoxic T lymphocytes, which may be important in rejection of tumor cells. An increasing number of reports from different experimental systems indicate that CD4 cells are extremely important for tumor rejection. The idiotypic protein isolated from serum should, from a practical point of view, facilitate the use of the idiotype as a vaccine preparation.

In most cases where ex vivo DCs have been used for loading the idiotypic protein, monocyte-derived DCs were selected. However, with a new technological platform, blood DCs, which are a more efficient immunogen than monocyte DCs, can be isolated. Dr. Hart stated that sufficient numbers of blood DCs can be isolated from one leucapheresis product to produce 30 doses of vaccines, which can be cryopreserved. This new purification platform may facilitate the use of DC-based vaccines.

Monitoring of an idiotype-specific immune response is not yet standardized. It is important that all research groups involved in vaccine development use the same technologies with standardized protocols. Proliferation assay (3H-thumidine incorporation) is used by all groups. However, to facilitate picking up idiotype-specific T-cells, several assays should be applied; these T-cells might be in different phases of the cell cycle as well as in maturation stages and of different subsets that may not be detected by proliferation assay only, which mainly detects CD4 T-cells. Cytokine assays should be added. ELISPOT estimates the frequency of cytokine-secreting cells, and both type I and II cytokines can be analyzed. A much more sensitive technique for cytokine assay is real-time polymerase chain reaction (PCR), which detects fewer cytokine-specific T-cells and requires fewer cells. Moreover, several cytokines can be assayed easily with real-time PCR, which gives a better prediction of the functional capability of idiotype-specific T-cells. A clinically sufficient T-cell response may not consist of more than one or two specific T-cells per 10,000, which cannot be detected by ELISPOT or flow cytometry. Therefore, it is recommended that proliferation assay, ELISPOT, and real-time-PCR be used as readout systems to have a good chance of adequately monitoring an idiotype-specific immune response. In addition, patients’ immune T-cell dysfunction should be evaluated and correlated to the capability of the patient to mount an idiotype-specific immune response. These results might yield more information regarding why patients may or may not mount idiotype-specific immune responses and form a basis for selection of patients for idiotype vaccination protocols.

Numerous studies were presented using idiotype loaded onto DCs together with GM-CSF. On average, 31% (CI95%: 19%-43%) mounted idiotype-specific proliferative responses. The patients were in advanced stage or in remission after chemotherapy. The induction of a proliferative response suggested a CD4 T-cell response that did not support an assumption of the presence of tolerant CD4 T-cells. Interesting preliminary clinical observations were also made. Patients who mounted immune responses had a longer time to disease progression than patients without immune responses. Those with preexisting idiotype responses had the longest time to disease progression. The last observation is in line with one by Dr. Mellstedt that patients with preexisting immune responses developed idiotype-specific responses more rapidly and that the magnitude was higher than in patients with no preexisting immune response. Dr. MacKenzie also reported that patients with minimal tumor burden after chemotherapy had a longer time to disease progression than those with "bulky" tumor, and the shortest time was noticed for patients who had relapsed after transplant. These data indicate that immune T-cell dysfunction had an impact on the clinical efficacy of idiotype vaccination. Dr. Lacy observed five clinical responses (three complete remission, one new complete remission, and one partial remission) and five patients with stable disease for a long time out of 17 patients with residual disease enrolled in a vaccination trial after transplantation. 

Dr. Nikhil Munshi of the University of Arkansas presented interesting data: He used monocyte-derived DCs loaded with the idiotypic protein and given intravenously. Only a few patients developed immune responses, and there was no clinical response. Those who developed immune responses had received the highest dose of DCs. In a subsequent study by the same group, patients received the highest dose of DCs (1 x 107) used in the previous study, but DCs were given subcutaneously. Dr. Qing Yi, also of the University of Arkansas, reported that one patient out of five entered OR (objective remission) (>50% reduction of the M component), and the remaining four had stable disease. This indicates the importance of the dose and route of administration of DCs.

Dr. Massaia and colleagues had vaccinated 15 patients in remission after chemotherapy using idiotype conjugated with KLH in combination with GM-CSF. Eighty-five percent mounted idiotype-specific delayed-type hypersensitivity reaction (reflecting a CD4 T-cell response). No proliferative response was recorded. The reason for this discrepancy is not clear, but it favors the notion that CD4 T-cells may not be tolerated by myeloma patients. These patients had a median time to relapse after start of vaccination of 35 months versus 24 months for a matched population. The difference was not significant. These patients also had abnormal T-cell repertoires, suggesting abnormal T-cell function and indicating that the population was not optimal for vaccination. Drs. Munshi and Kwak also showed that the idiotypic protein conjugated with KLH and given together with GM-CSF induced an idiotype-specific T-cell response (CD4/CD8) in about 75% of the patients, and those patients had significantly longer relapse-free survival than patients who did not mount idiotype-specific T-cell responses.
Another adjuvant cytokine of importance in vaccine development is interleukin (IL)-12. This cytokine leads to a type I T-cell response and amplifies the magnitude of the immune response. In studies performed by Dr. Mellstedt and colleagues, about 90% of the patients mounted idiotype-specific immune responses when the idiotypic protein (absorbed onto alum) was given together with GM-CSF and IL-12; 45% mounted immune responses when IL-12 was given alone and 70% when GM-CSF was given as the only adjuvant cytokine. The highest amplitude of immune response was seen in patients receiving the combination of IL-12 and GM-CSF. Only one major response (>50% reduction of the M component) was seen in this population of vaccinated stage I myeloma patients. However, in 50% of the studied patients, circulating tumor cells, measured by real-time PCR and VH-CDRIII specific primers and probes, gradually disappeared. The kinetics of the tumor response speak against a redistribution phenomenon of tumor cells in blood circulation. The result indicates that the tumor-specific immune response may not be able to reduce a large tumor burden but may affect minimal disease. The study also showed that repeated immunizations were necessary, as it may take up to 8 months (eight immunizations) before an idiotype-specific immune response can be evoked. Moreover, preliminary data indicate that there seems to be no tolerance induced by repeated immunizations. This is not to be expected when one uses a self-antigen as an immunogen. Self-antigens are usually low-avidity antigens, and it is necessary to do repeated immunization when using them.

Dr. Kwak and colleagues showed that sibling donors to recipients of allogeneic transplants, immunized with the patient’s idiotype, developed idiotype-specific CD4 cells, which could be transferred during the transplant to the recipient. This mode of idiotype immunization should circumvent CD4 tolerance. Studies are in progress by Drs. Kwak and Stevenson. However, this approach is only available for a restricted number of myeloma patients. 

We are at the beginning of development of a therapeutic vaccine strategy in myeloma. At present, focus is on the idiotype as target structure, but other antigens are also available. Induction of idiotype-specific CD4 and CD8 T-cells is necessary. The idiotypic protein coupled to KLH or alum in conjugation with GM-CSF may induce idiotype-specific CD4 and CD8 T-cells. Idiotype loaded onto dendritic cells also induce an immune response and clinical effects. Data regarding the clinical efficacy of DNA vaccines in humans will come in the near future. It is very important that research groups test various concepts with regard to vaccine preparations, adjuvant cytokines, and scheduling. As soon as we have optimized the vaccination procedure, extended clinical trials are urgently needed, but it is important to select optimal patients, who should have a minimal disease burden and well-preserved immune systems. 

Since the first results on idiotype immunization were reported in 1994,a great deal of progress has been made. The progress may seem slow, but within this decade, we should have a therapeutic vaccine concept in clinical use in multiple myeloma.

 related articles