• © 2011 by American Society of Clinical Oncology

Identifying Patients With Follicular Lymphoma Who Are Likely to Benefit From an Idiotype Vaccine

B lymphocytes express surface immunoglobulin with a unique molecular structure that is based on the heavy- and light-chain variable regions. The unique molecular structure of the antigenic determinants in the hypervariable region of the antibody is called the idiotype. When B cells undergo malignant transformation, the idiotype has the potential to be recognized as an antigen and used to develop a tumor-specific vaccine. Previous studies that have been performed predominantly in patients with follicular lymphoma have shown that the use of idiotype vaccines can elicit a specific immune response, induce antitumor effects, and impact disease-free survival (DFS) and overall survival.¹ Although the results of idiotype vaccines have seemed promising in small single-arm studies, larger randomized placebo-controlled trials were needed to confirm the initial results. Two previous randomized trials in patients with untreated follicular lymphoma failed to show a benefit for patients who received idiotype vaccination (Table 1).^2,3

In the article that accompanies this editorial, Schuster et al⁴ report the results of a third randomized trial that compared an idiotype vaccine with a placebo in patients with follicular lymphoma who had a complete response after 6 to 8 months of combination chemotherapy. Overall, the authors found no significant difference in DFS among the 177 patients (of the 375 planned) who were randomly assigned (hazard ratio for vaccine v placebo, 0.92; 95% CI, 0.51 to 1.65). In the subgroup of patients who remained in remission long enough after random assignment for the vaccine to be made and administered (66% of those who were randomly assigned), the use of a patient-specific hybridoma-derived idiotype vaccine prolonged the DFS. Among the patients with an immunoglobin M (IgM) vaccine isotype (34% of those randomly assigned), the authors also found benefit from vaccine therapy.

Although the results of this study suggest that this idiotype vaccine may have a therapeutic effect in a subset of patients, the study has limitations that suggest caution in interpretation of the results. Accrual was halted after fewer than half of the planned number of patients had been randomly assigned; this severely limits the statistical power of the study. Moreover, 60 of the randomly assigned patients (approximately 33% of each vaccine arm) did not receive any vaccinations; 55 patients relapsed while the vaccine was being made. There was no evidence of a difference in DFS among the 177 patients who were randomly assigned, but interestingly, a separation of the DFS curves was seen at 20 to 25 months after random assignment. In a secondary analysis of patients who received at least one vaccination, however, the difference in DFS achieved statistical significance. The study's approach is problematic in that the patient had to remain disease-free while awaiting manufacture of the vaccine—an average of 8 months after random assignment—and thus, patients who received the vaccine may have been less chemotherapy resistant, had less aggressive disease, or both. Imbalances between the vaccine arms in terms of the stratification factors (number of chemotherapy doses and prognostic index) may also have been introduced. From Table 1 in the article by Schuster et al,⁴ wherein the characteristics of randomly assigned patients who received vaccination are listed, the patients in the placebo vaccine arm tended to be men, more commonly expressed an IgM isotype, and more commonly had stage IV disease. It is unclear, however, how these imbalances may affect the efficacy findings in this subgroup. In an unplanned subgroup analysis of the 60 patients with an IgM isotype who received at least one vaccination, the DFS was prolonged for those who received the idiotype vaccine. A dramatic drop in DFS was again seen at 20 months after random assignment in the control group. Although the DFS in the IgM-idiotype vaccine group was clearly improved, there was no evidence of a plateau in the curve to suggest that some patients may remain disease-free for the long term.

These findings are intriguing, but a significant issue that makes additional testing necessary and integration of this approach into clinical practice challenging is the omission of rituximab from the initial treatment regimen. Although the study design called for use of exactly the same approach as was used in the initial phase II trial,⁵ rituximab-containing regimens are now the standard of care for follicular lymphoma. Furthermore, the chemotherapy regimen used in this study, albeit similar to more standard regimens, is uncommonly used in clinical practice. Although the addition of rituximab may not affect the immunologic response to the vaccine, the addition of rituximab to chemotherapy has clearly increased response rates and duration of response in other studies. This effect of rituximab would therefore be expected to significantly impact the number of patients in remission and free from relapse at the time of a planned vaccination approach and may influence the potential benefit of an idiotype vaccine. It is even possible that the population of patients who benefited from the idiotype vaccine in this study is the same group of patients who would have benefited from rituximab should it have been included. The clinical benefit may have resulted from the stimulation of the immune system by antibody-coated tumor cells, whether the antibody was host-generated in response to the vaccine and directed at the idiotype or passively infused and directed at CD20, as is the case with rituximab. It is also common practice that patients with follicular lymphoma in remission are treated with rituximab maintenance therapy. It may therefore not be easy for the patient to have a gap in therapy while waiting for the vaccine to be developed, given that maintenance approaches are now commonly used. Although the use of rituximab maintenance may decrease the number of patients who progress before vaccination, this added intervention would make integration of the results of this study into clinical practice difficult.

However, this study,⁴ when compared with the two previous randomized studies^2,3 that evaluated idiotype vaccines in follicular lymphoma, does provide a number of valuable insights with respect to the clinical use of idiotype vaccine approaches. First, the previous studies included patients with partial responses or even stable disease, and this study suggests that it is important for patients to have a complete response to treatment before vaccination, and that this complete response should be durable. The persistence of significant numbers of malignant cells may make the immune environment unfavorable for the development of an adequate immune response to the vaccine. Second, the vaccine for this study was generated from a hybridoma, whereas the other studies used molecular approaches; this suggests that manufacturing procedures may be important. Finally, the suggestion that patients in whom an IgM-idiotype vaccine was generated had a prolonged DFS indicates that this may be especially important in identifying the population who may benefit from this approach. Previously, the patients who subsequently had a measurable antibody or cellular immune response were those with improved clinical outcomes. Determining the idiotype isotype may facilitate early identification of patients who are more likely to benefit from a vaccine approach.

The results of this randomized study reported by Schuster et al⁴ provide insights with respect to a subgroup of patients with follicular lymphoma who may benefit from receiving an idiotype vaccine. However, in the face of multiple other treatment options for patients with follicular lymphoma that include maintenance therapy with rituximab, radioimmunotherapy, or novel, small-molecule approaches, it is difficult to determine where a vaccine approach fits into the therapy of newly diagnosed patients.