High-Risk Disease
The high-risk group consists mainly of patients with stage 4 disease who are greater than 1 year old at diagnosis, but also includes the following subgroups: stage 3 children greater than1 year old with either MYCN amplification or unfavorable Shimada features; stage 2 patients greater than 1 year with MYCN amplification; and stage 3, 4, and 4s infants with MYCN amplification. These patients continue to respond poorly even to very dose-intensive therapy, although incremental advances have certainly been made over the past decade.
As shown in Figure 3, the 4-year survival rate among the 507 stage 4 patients greater than 1 year old who were treated in the CCG studies from 1978 to 1985 was 9%, as compared with a rate of 30% among the 675 patients treated from 1991 to 1995 (P < .001).[24] Some of this improvement may be attributable to increases in the dose intensity of therapy; a meta-analysis by Cheung and co-workers has shown an effect of dose intensity in neuroblastoma similar to that reported in other cancers.[25] The improvement may also be the result of the increasing use of high-dose myeloablative therapy with autologous or allogeneic bone marrow transplantation (BMT).[26-33]
Nonrandomized Studies—Between 1988 and 1993, the CCG completed three concurrent studies of high-risk neuroblastoma: CCG-321P1, which tested allogeneic BMT; CCG-321P2, 1 year of chemotherapy; and CCG-321P3, autologous purged BMT. All three studies treated patients with five to six cycles of induction chemotherapy consisting of cisplatin (Platinol), doxorubicin, cyclophosphamide, and etoposide (VePesid). After undergoing surgery for local control and receiving irradiation to residual disease sites, patients received ablative chemotherapy and total-body irradiation (TBI) with BMT or were continued on four-drug induction chemotherapy for a total of 13 cycles.
Several important conclusions can be drawn from these nonrandomized studies:
- Allogeneic BMT was not shown to be superior to autologous purged BMT and, in fact, had a higher toxic death rate and an equal relapse rate.[29]
- With a median follow-up of 5 years, the event-free survival at 5 years from the time of autologous BMT for 147 patients is 37%.
-
In a retrospective nonrandomized comparison of all of the stage 4 patients who continued to receive chemotherapy vs those who went on to receive purged autologous BMT, the event-free survival rate was significantly better for those receiving autologous BMT (40% vs 19%; Figure 4).[18] This retrospective analysis involved 207 children greater than 1 year at diagnosis with stage 4 neuroblastoma treated on CCG protocol 321P2 (described above). By the end of five to seven courses of chemotherapy, 159 patients were event-free; 67 of these patients went on to receive autologous BMT, while 74 continued chemotherapy for a total of 1 year. The relative risk of an event after autologous BMT was only 58% of that with chemotherapy (P = .01). An even more significant advantage for autologous BMT was seen among patients who were only in partial, as opposed to complete, remission at the end of induction and for those who with tumor MYCN amplification.
Other Comparisons of Autologous BMT and Chemotherapy—Previous smaller studies comparing autologous marrow transplantation to chemotherapy had yielded conflicting results, with Pinkerton reporting the only randomized trial showing a significant improvement with autologous BMT.[32] This report included a total of only 65 patients, however.
A retrospective comparison by the POG of 116 patients achieving complete or partial remission did not show any significant difference in outcome for the 32 patients who underwent BMT prior to progression.[34] The Study Group of Japan reported a 50% event-free survival rate in a nonrandomized group of patients given myeloablative therapy with autologous BMT, as compared with a 39% rate in those who received chemotherapy.[31]
Overall, autologous BMT appears to be at least as effective as intensive chemotherapy and may provide an advantage in some extremely high-risk subgroups. However, a definitive conclusion cannot be made until the results of a recently completed CCG randomized prospective trial comparing outcomes among unselected high-risk patients from the time of diagnosis become available.
Another ongoing phase I dose-escalation study of a non-TBI-containing myeloablative regimen has thus far shown event-free survival comparable to that seen in the previous 321P3 study (55% at 2 years from autologous BMT). This phase I study has further verified the earlier suggestion from Kushner and co-workers[28] that higher-dose local irradiation to tumor sites may prevent primary site relapse, a common problem in previous CCG protocols. However, the relapse rate for patients transplanted after first disease progression has continued to be high; the current event-free survival rate at 2 years post-transplantation is only 15%. This is similar to the poor outcome observed in European studies, as reported by Ladenstein and colleaguesl.[27]
Recently Completed and Ongoing Trials
CCG Trial in High-Risk Disease
The most recently completed CCG trial in high-risk disease was designed to answer, in a prospective, randomized fashion, the relative importance of ablative therapy with purged autologous BMT vs intensive consolidation. The study accrued 550 high-risk patients in 4-1/2 years, with approximately one-third of patients randomized to each arm and the other one-third nonrandomly assigned to consolidation chemotherapy if they refused to be randomized. A second randomization, done at the end of chemotherapy consolidation or autologous BMT, assigned patients to receive or not to receive 6 months of 13-cis-retinoic acid (isotretinoin [Accutane]). In vitro and some in vivo studies have reported this differentiating agent to have efficacy against neuroblastoma.
The outcome of this study, still blinded for analysis, will help determine the future role of myeloablative therapy, as well as the potential utility of differentiating agents, in high-risk neuroblastoma. Current and future studies will examine repetitive stem-cell transplants, more intensive induction therapies, and the use of alternative stem-cell sources.
Source and Purity of Stem Cells for Ablative Therapy
The source and purity of the stem cells used for ablative therapy in neuroblastoma have been the focus of concern. Immunocytology techniques have shown that as many as 70% of patients have bone marrow tumor at diagnosis and 50% have circulating tumor cells in peripheral blood. Even after several cycles of induction chemotherapy, 25% of patients still have some detectable tumor by bone marrow immunocytology and 7% have circulating tumor cells.[22]
Such cells in bone marrow have the potential for tumorigenicity. This has been inferred from studies showing the development of tumor cell lines from harvested bone marrow and by reports of miliary lung metastases following autologous BMT,[35] and has been demonstrated more definitively by the gene-marking studies of Rill and co-workers.[36] In these gene-marking studies, autologous unpurged but histologically tumor-free marrow was transfected with the neomycin resistance gene and then reinfused into patients, some of whom subsequently relapsed with neuroblastoma tumors expressing the marker.
More recently, testing of peripheral blood stem-cell collections and bone marrow with PCR has confirmed the problem of contamination with tumor cells, even in CD34-selected preparations. The significance of such low-level contamination is still undetermined but will be examined in future studies. Such contamination suggests the need for effective methods to purge both peripheral blood stem cells and bone marrow prior to reinfusion.[37]
Novel Therapeutic Approaches for Metastatic Disease
Dose-intensive therapy, with or without stem-cell support, is rapidly approaching the limits of toxicity, both in terms of acute and late effects. Novel approaches that are more tumor-specific and less toxic are required to make further progress in metastatic neuroblastoma. Approaches under investigation that show some promise in the laboratory and in preliminary phase I and II studies include immunologic modulators, including tumor-specific antibody, antibody given with gran_ulocyte colony-stimulating factor (GM-CSF, filgrastim [Neupogen]) or interleukin-2 (IL-2 [Leukine]), fusion proteins of antibody with cytokines, and tumor vaccines; tumor-targeted therapy utilizing radiolabeled antibody or MIBG; and tumor-differentiating agents.
Immunologic Modulators
Phase I and II studies of anti-GD2 antibodies have shown some modest antitumor efficacy in patients with relapsed neuroblastoma.[38,39] Both the murine monoclonal 3F8 and 14G2a (murine), the latter tested with IL-2 and with GM-CSF, have shown some tumor responses, as well as in vitro stimulation of antibody-dependent cellular cytotoxicity (ADCC). More recently, Yu and co-workers have completed a phase II trial of the human chimeric anti-GD2 antibody, CH14.18 with GM-CSF, which again shows some responses, most frequently in bone marrow tumor.[40]
Targeted Therapy
Targeted therapy using antibody or MIBG for the delivery of radiation in the form of iodine-131 has also been tested in clinical trials. Cheung and co-workers have documented responses to iodine-131-3F8 among patients with refractory neuroblastoma,[41] and are currently conducting a study in which newly diagnosed patients are receiving iodine-131-3F8 in ablative doses followed by bone marrow rescue, with further treatment with cold antibody post-transplant.
Iodine-131-MIBG has been widely tested in Europe for refractory neuroblastoma and, more recently, for initial therapy in patients with regional disease. We have conducted a phase I dose escalation trial of iodine-131-MIBG at the University of California (San Francisco) and have determined the maximal non-marrow-ablative dose and the maximal practical ablative dose with stem-cell rescue.[42] In 30 patients, we observed a 37% response rate and no significant toxicity other than hematologic effects. Future studies will utilize combinations of ablative chemotherapy with iodine-131-MIBG.
Differentiating Agents
Differentiating agents are another approach for circumventing the undesirable effects of cytotoxic agents. For many years, laboratory studies have shown the capacity for either spontaneous or induced differentiation and growth arrest of neuroblastoma cell lines in culture.
Currently, various derivatives of retinoic acid are the differentiating compounds in clinical testing. One such derivative, 13-cis-retinoic acid, has been reported to cause remissions in refractory patients, and a recently completed phase II trial showed a 10% response rate in relapsed neuroblastoma.[43-45] The recently completed randomized CCG-3891 study randomized all patients at the end of consolidation therapy to receive 13-cis-retinoic acid or no further therapy; results of this trial are pending.
Other retinoids currently in phase I investigation include all-trans-retinoic acid (ATRA [Vesanoid]) combined with interferon-alfa (Intron A, Roferon-A), 9-cis-retinoic acid, and fenretinide, another analog that causes growth arrest and apoptosis in vitro, even in cell lines shown to be resistant to trans-retinoic acid.
Conclusions
Substantial advances are occurring in the elucidation of the molecular pathogenesis of neuroblastoma, as well as in the definition of clinical prognostic groups. In addition, modest but significant improvements in the treatment of metastatic disease have been achieved by increasing dose intensity with the use of hematopoietic support. However, in order to overcome chemotherapy resistance, new therapeutic approaches are required using tumor-targeting, differentiating, or apoptotic agents; stimulation of host immune response; or genetic manipulation.
