Comparison of results of a pilot study of alternating vincristine/doxorubicin/cyclophosphamide and etoposide/ifosfamide with IRS-IV in intermediate risk rhabdomyosarcoma: A report from the Children's Oncology Group

INTRODUCTION

Over 50% of patients with newly diagnosed rhabdomyosarcoma (RMS) have intermediate risk disease, as defined by stage, site, nodal status, and histology 1,2. Table I shows the risk group assignment used for the most recent Intergroup Rhabdomyosarcoma Study (IRS) Trials. The first three predominantly North American IRS trials investigated the addition of doxorubicin to VAC chemotherapy for patients with Clinical Group III and IV RMS, without demonstrating a clear benefit for doxorubicin 3-5. The design of these studies has been criticized for the non-dose-intensive use of doxorubicin and mixing of different histologic subgroups, complicating their interpretation 6. Doxorubicin is an active agent against RMS with single agent response rates of 20%–100% in older studies and 65% in more recent phase II window studies of patient with high-risk RMS 7-9. However, the potential for cardiotoxicity and the lack of evidence of superiority of doxorubicin with VAC over VAC alone have limited its widespread use in the initial treatment of RMS. Because young children are particularly susceptible to anthracycline-induced cardiotoxicity caution regarding the use of doxorubicin is especially relevant 10,11.

In IRS-IV study, conducted from 1991 to 1997, most patients with non-metastatic RMS were randomized to receive one of three chemotherapy regimens for the first 28 weeks of treatment: vincristine and dactinomycin (VA) plus cyclophosphamide (VAC), VA plus ifosfamide (VAI), or vincristine, ifosfamide, and etoposide (VIE). All patients then received 18 additional weeks of VAC. The overall 3-year failure-free survival (FFS) and overall survival were 77% and 86%, respectively. There was no difference in outcome for patients treated with VAC alone or sequential VAI/VAC or VIE/VAC 12. Ifosfamide can cause nephrotoxicity, especially in younger children and at cumulative doses greater than 72 g/m² 13-16, and etoposide has been associated with therapy associated leukemia 17. Therefore, VAC chemotherapy remains the standard systemic therapy in North America.

Concurrent with the conduct of IRS-IV, Arndt et al. investigated a regimen of alternating cycles of vincristine, doxorubicin, and cyclophosphamide (VDC) and etoposide and ifosfamide (IE) with additional weekly vincristine. Local tumor control was delayed until week 12. Preliminary data published in 1998 included a Kaplan–Meier estimate of 3-year FFS and overall survival of 85% and 91%, respectively 18. Because of the limited study population and follow-up, it was difficult to compare to the IRS-IV results.

The purpose of this report is to update the experience with alternating VDC/IE therapy with a larger cohort of patients and longer follow-up, and to compare the outcome of patients treated on the pilot study with that of a similar group of patients treated on IRS-IV.

METHODS

The pilot study utilizing alternating cycles of VDC/IE was conducted from 1991 to 1999. Patients were defined as having intermediate risk RMS if they had stage 2 or 3 (unfavorable site) Group III disease, or any non-metastatic alveolar disease. Patients with metastatic disease were not eligible for the pilot study. A subset of “similar” patients treated on IRS-IV, conducted from 1991 to 1997, was identified to match the pilot study patients. Patients were compared for presenting features and outcome. Details of both treatment regimens and toxicity have been previously reported 12,18.

Comparisons of the outcome for patients treated on the pilot study and IRS-IV were performed in two ways. Almost half of the patients treated on the pilot study had parameningeal (PM) tumors. Therefore, the results for these patients were first compared to similar patients treated on IRS-IV, matched for PM site and meningeal extension, as these factors have been shown to be prognostically important 19. FFS was defined as the time to the first occurrence of recurrence of RMS or death as a first event, with patients not experiencing a treatment failure censored at date of last contact. Overall survival (OAS) was defined as time to death from any cause or date of last contact. Kaplan–Meier estimates of the time-to-event distributions for FFS and OAS were calculated and compared using the log-rank test.

For patients with non-PM primaries, patients from IRS-IV were matched to patients from the pilot study by primary site, histology, age, and IRS Group, and a stratified Cox model was used to assess the impact of “study” (IRS-IV or pilot) after adjusting for these prognostic factors. This analysis was also performed including the patients with PM primaries as an additional stratum. Estimates of FFS for the IRS-IV patients with non-PM primaries were calculated by weighting the IRV-IV cases so that their distribution mimicked that of the cases treated on VDC/IE.

RESULTS

DISCUSSION

The survival of patients with RMS has improved significantly over the past 3 decades, from approximately 25% in 1970 to 75% on IRS-IV for patients with non-metastatic disease. Despite randomized investigation in Clinical Group III and IV patients on IRS-I, -II, and -III, the role of doxorubicin remains controversial in the treatment of RMS. Randomized clinical trials of IRSG have failed to show a clear benefit to the addition of doxorubicin to other agents for treatment of RMS 4,5. However, the results from IRS-III suggested that two patient subsets potentially benefited from the addition of doxorubicin. First, patients with Clinical Group II favorable histology tumors (excluding orbit, head, and paratesticular sites), were randomized between VA and radiation therapy with or without doxorubicin. There was suggestive but not conclusive statistical evidence that the addition of doxorubicin to the standard VA regimen improved outcome 5. Second, Clinical Group I and II unfavorable histology tumors received pulsed VDC, VAC plus cisplatin, and radiation therapy and showed significant improvement in progression-free survival and survival compared to a comparable group treated with VA or VAC on IRS-II 5. However, the regimens were complex, and also contained cisplatin and cyclophosphamide, making the contribution of doxorubicin to the improvement in outcome difficult to assess.

Similarly, the role of ifosfamide remains unproven in the treatment of RMS. On IRS-IV, there was no difference in outcome for patients with non-metastatic disease when they were treated with VAC, VIE, or VAI 12. However, IRS-IV was designed as a substitution study during the initial 28 weeks of therapy, with all patients receiving VAC for the final 18 weeks of treatment. It is possible that sequential design, rather than alternating courses of therapy, contributed to the equivalent outcome of the three arms of IRS-IV.

The definition and recognition of risk groups among patients with RMS has also evolved, making comparisons between studies difficult. In this report, we compared the outcomes of patients treated on a pilot study of alternating therapy with VDC and IE and a similar subset of patients treated on IRS-IV with VAC, VIE/VAC, or VAI/VAC, using a stratified Cox model. Our results suggest that when all subsets of patients are combined (PM and non-PM), the relative risk of failure for patients treated on the doxorubicin-containing VDC/IE regimen of alternating therapy was lower than for similar patients treated on IRS-IV (relative risk of failure 0.5, P-value = 0.06). The outcome for our patients treated with alternating therapy is among the best reported in the literature.

This analysis has several limitations. First of all, the VDC/IE study had only 46 patients compared to 342 patients on IRS-IV. Second, there were more stage 2 patients on the VDC/IE study than on IRS-IV, although we attempted to adjust for this using a stratified Cox proportional hazards model. Third, this was not a randomized study. Fourth, it is not possible to determine whether the good outcome observed on the VDC/IE study was related to use of alternating combinations of non-cross-resistant agents as described by Norton and Simon 20, or the use of doxorubicin, or both. Fifth, methods of local control may have differed between the two studies and this was not analyzed. Lastly, the VDC/IE study was a limited institution study, and limited institution pilot studies often have better results when compared to multi-institutional studies. There may also have been other unmatched unknown risk factors which differed between the two studies.

Because of these limitations, this analysis neither proves nor disproves that the VDC/IE regimen used in this pilot study is superior to the regimen described in IRS-IV. Of note, doxorubicin and etoposide/ifosfamide have also been recently used in the Children's Oncology Group study for first relapse of RMS. Moreover, Womer et al. 21 have also noted excellent preliminary results in patients with newly diagnosed RMS with alternating therapy and interval compression as a means of dose intensification. Given the excellent outcome of the group of patients treated with alternating therapy on our and the Womer et al. pilot study, the Soft Tissue Sarcoma Committee of the Children's Oncology Group (COG) is exploring this strategy along with use of interval compression of chemotherapy cycles and addition of irinotecan as a radiation sensitizer in newly diagnosed patients with high-risk RMS.

If the outcome of high-risk patients is improved with this strategy, consideration will be given to incorporating it into future therapy for intermediate risk group patients. In the meantime however, COG has elected to evaluate newer topoisomerase I agents in frontline intermediate risk trials, since IRS has already committed many patients to doxorubicin randomized studies and IRS-IV to randomized evaluation of IE, and topoisomerase inhibitors have shown promise both in preclinical models and in up-front window studies of high-risk patients 22,23.