4. DISCUSSION
The combined analysis of the results of these two cooperative studies suggests that different strategies of chemotherapy may have an impact in the outcome of a subgroup of patients with ARMS and regional nodal involvement.
The prognosis of patients with ARMS N1 has been reported to be poor in the historical series of European co-operative studies. In the CWS-86 study, 3-year EFS was 25% and OS 29% [11]. In the SIOP MMT84 study, 5-year EFS was 31% comparable to that of stage IV disease [16].
The experience of similar patients in North American cooperative groups has been better than that of the European studies. Of 125 patients with localized RMS and nodal involvement enrolled in the IRS-IV study, Rodeberg et al [17] reported 43% five-year failure-free survival of patients with ARMS and nodal involvement. The overall outcome of patients with alveolar histology and N1 disease reported in that study was similar to that of patients with metastatic disease at a single site.
We recently reported the experience of EpSSG RMS2005 [12] in patients with ARMS N1 and showed results similar to the IRS-IV study. The reasons for this apparent improvement in the outcome of these patients, compared to those treated within previous European studies, could be due in part to better risk stratification, more adequate treatment with intensified chemotherapy, systematic local treatment and/or improvement in supportive care.
Failures among patients reported in the present study were predominantly with metastatic disease and was similar in both cohorts. Their tumors frequently presented with advanced IRS Group and unfavorable location, characteristics related to an increased risk of distant metastatic disease [18-20]. Local recurrences represented one-third of all relapses in both protocols and local treatment in RMS2005 and in ARST0531 was identical. However, Casey et al [21] reported more loco-regional relapses in patients enrolled in ARST0531 compared with the previous COG intermediate-risk RMS study, D9803. The reasons for this increase in local relapses were not clear but may have been related to changes in cyclophosphamide dosing, changes in RT administration and fewer surgical procedures.
The most important difference between EpSSG and COG protocols was the type of chemotherapy. Patients received three different regimens based on the use of the standard combination (IVA in EpSSG and VAC in COG) to which was added doxorubicin and 6 months of maintenance chemotherapy with vinorelbine and cyclophosphamide in the European study or irinotecan in the North American study. In RMS2005, doses of alkylating agents included 54 g/m2 of ifosfamide plus 4.5 g/m2 of cyclophosphamide. In contrast, COG patients received 16.8 g/m2 (VAC) or 8.4 g/m2of cyclophosphamide (VAC/VI). Moreover, patients in EpSSG received doxorubicin at a total dose of 240 mg/m2. One-half of the patients in COG received irinotecan. Total duration of chemotherapy was 42 weeks in COG protocol vs 50 weeks in EpSSG.
There is no evident advantage of adding doxorubicin or irinotecan to the standard chemotherapy in these patients. Hawkins et al [13] have previously reported a lack of improvement with the addition of irinotecan to VAC in the same of group of patients as the present analysis. The European experience demonstrated that the addition of doxorubicin to IVA in patients with high-risk disease also failed to improve outcome when compared to standard chemotherapy [22]. Moreover, previous experiences in Europe demonstrated that the addition of carboplatin, etoposide, and epirubicin did not improve the outcome for patients with high-risk disease enrolled in the SIOP MMT-95 study [8]. In RMS2005, two toxic deaths and one secondary neoplasm occurred, so that when designing future studies, we consequently should search for a balance between the total burden of chemotherapy and the risk of toxicity.
The impact of FOXO1 fusion status in the outcome of patients in both cohorts showed important differences. In the European study, fusion-positive tumors had significantly worse EFS as well as a trend toward inferior OS, while the EFS/OS of patients enrolled in the COG study was the same whether they had fusion-positive or fusion-negative tumors. These contradictory results are intriguing. Numerous studies have reported the impact of fusion status on the outcome of patients with ARMS [23-25]. However, this impact could be different in patients with other adverse prognostic factors. For example, in patients classified as low or intermediate risk in COG studies, the presence ofFOXO1 fusion has a strong impact in prognosis [25,26] while in patients of high risk with metastatic disease, the clinical prognostic factors have a stronger impact than the fusion status [27]. However, these observations do not explain the differences we observed between our two homogeneous cohorts. In the European study the EFS of patients with fusion-negative tumors was 73%, similar to that of patients considered as high risk in the same protocol [22].
We can hypothesize that those patients with fusion-negative tumors benefit from a more intense chemotherapy. Patients in ARST0531 received a lower cumulative dose of cyclophosphamide than in previous COG RMS studies, with a possible negative impact on outcome [13]. In contrast, patients with ARMS N1 in RMS2005 could have had a benefit from the addition of maintenance chemotherapy, similar to the results of high-risk patients [28].
However, our study presents some limitations, such as the relatively small number of patients with fusion-negative tumors in both cohorts, and the higher proportion of patients with fusion status unknown in the EpSSG cohort, which could reduce the precision of the EFS estimations.
In conclusion, there are some lessons learned from the present combined analysis that should be further explored in the upcoming studies of both cooperative groups: First, very different treatment strategies in two concurrent clinical trials generated similar outcomes for patients with ARMS and regional lymph node involvement. Second, among patients withFOXO1 fusion-negative ARMS and regional lymph node involvement, the cumulative dose of alkylators and/or maintenance chemotherapy may influence outcome and, third, a need persists for innovative therapeutic strategies for those patients with fusion-positive tumors.