Does high-dose chemotherapy have an impact on outcome of Ewing sarcoma compared to standard chemotherapy?

• Population: Ewing sarcoma
• Intervention: high-dose/myeloablative chemotherapy with autologous stem cell rescue
• Comparison: standard chemotherapy
• Outcomes: overall survival, event-free survival, toxicity

Authors: Ashika Ramamurthy, Natacha Omer, Julie Cayrol, Elizabeth Connelly, Jeremy Lewin, Vivek Bhadri, Joanna Connor, Peter Grimison, Angela Hong and the ANZSA Sarcoma Guidelines Working Party

Ewing sarcoma (ES), most frequently diagnosed in children and young adults, is a group of malignant solid tumours consisting of small blue round cells that exhibit varying degrees of neural differentiation, most often located in bone, although extraosseous primaries also
occur (1-3).  These tumours are collectively defined by the presence of a fusion involving a member of the FET family of genes (usually EWSR1) and a member of the ETS family of transcription factors, with EWSR1-FLI1 fusion identified in 85% of cases (1, 4). Small blue round cell sarcomas of bone and soft tissue that do not meet the above criteria on molecular testing have been newly grouped under three main variants – round cell sarcomas with EWSR1–non-ETS fusions, CIC-rearranged sarcomas and sarcomas with BCOR genetic alterations. These were previously collated with ES and termed the Ewing sarcoma family of tumours (ESFT)(1).

Survival rates in ES have improved significantly in the last three decades due to cooperative trials which emphasized the importance of systemic therapy in both localised and metastatic disease. Patients with localised disease that respond to multimodal therapy have a 5-year overall survival (OS) rate of over 70% (5). Adverse prognostic factors have been identified in localised disease, including poor histologic response to induction chemotherapy, large tumour volume and unfavourable tumour location (5). Outcomes in primary disseminated multifocal ES and relapsed disease remain poor with 5-year OS at or below 30% (5, 6). For these poor prognosis patients and patients with high-risk localised ES, high-dose chemotherapy (HDT) with autologous stem cell transplant (ASCT) has been evaluated as an alternative, or in addition to conventional chemotherapy. The rationale for HDT is to exploit the dose-response relationship with alkylating agents, where relatively small changes in dose may impact tumour cell kill in these highly chemo-sensitive tumours. Myelosuppression is a dose-limiting toxicity of these agents; however this can be overcome when treatment is supported by autologous stem cell rescue, enabling HDT use (7, 8). The relative benefit of HDT, however, remains uncertain and must be considered against severe haematological and non-haematological toxicities, including death, associated with treatment.

This systematic review identifies 30 studies addressing the impact of HDT with ASCT in Ewing sarcoma. These studies include a large volume of retrospective and prospective data from American and European sarcoma registries and cooperative clinical trials groups.

HDT as part of first-line treatment in high-risk or poor prognosis Ewing Sarcoma

Sixteen studies evaluated HDT as part of first-line therapy specifically in ES and provided data on survival outcomes (technical report - Table 1). This includes the published results from three different arms of two European randomised controlled clinical trials (Euro-E.W.I.N.G. 99 and EWING 2008 studies)(8-10). The observed impact of HDT differed for the different arms of these three prospective randomised trials with only one study finding evidence for the utility of HDT.

Whelan et al. examined the outcome of 240 patients with high-risk localised disease who received six vincristine/ifosfamide/doxorubicin/etoposide (VIDE) courses and one VAI consolidation course, enrolled in the R2Loc arm of the Euro-E.W.I.N.G.99 or EWING-2008 studies (8). High-risk disease was defined as patients with poor histological or radiological response to induction chemotherapy (residual viable cells >=10% or <50% radiological reduction of the soft tissue disease component in small unresected tumours) or large tumour volume at diagnosis (>=200 mL). These patients were then randomised to a further seven VAI courses or one course of high-dose BuMel chemotherapy with ASCT. A significant benefit was observed in OS (HR 0.63; 95% CI 0.41-0.95; P=0.028) and EFS (HR of 0.64; 95% CI 0.43 to 0.95; P=0.026) for patients receiving HDT.  The 3- and 8-year EFS were, respectively, 69.0% (95% CI, 60.2% to 76.6%) versus 56.7% (95% CI, 47.6% to 65.4%) and 60.7% (95%CI, 51.1% to 69.6%) versus 47.1% (95% CI, 37.7% to 56.8%). The 3- and 8-year OS were, respectively, 78.0% (95% CI, 69.6% to 84.5%) versus 72.2% (95% CI, 63.3% to 79.6%) and 64.5% (95% CI, 54.4% to 73.5%) versus 55.6% (95% CI, 45.8% to 65.1%).

In patients with isolated pleural or pulmonary metastases at diagnosis (R2Pulm arm in Euro-E.W.I.N.G.99 and EWING-2008) the effect of Busulfan and Melphalan (BuMel) HDT (n = 144) was prospectively compared to standard vincristine/actinomycin/ifosfamide (VAI) consolidation chemotherapy with whole-lung irradiation (n = 143). In this group, a numerical difference in event-free survival (EFS) was observed; 43.1% versus 52.9% at 8 years with VAI plus whole-lung irradiation, and HDT respectively, although the hazard ratio was non-significant (HR 0.79, 95% CI 0.56-1.10, p=0.16) and no significant OS benefit was observed (HR 1.00, 95% CI 0.70-1.44; P=0.99)(9). Four patients died as a result of HDT-related toxicity.

In the Euro-E.W.I.N.G99 trial, 281 patients with primary disseminated multifocal ES (patients with isolated pulmonary metastases enrolled in the R2pulm arm excluded) were included in a non-randomised arm of the trial and 169 (60%) received BuMel-HDT. The reported 3-year EFS was 27% (+/-3%) and 3-year OS 34% (+/-4%) for all 281 patients (11). The incompatibility of BuMel HDT with radiation therapy to axial sites, due to lung toxicities observed with Busulfan, prompted the search for an alternative regimen. Treosulfan is a prodrug of a bifunctional alkylating agent and structurally related to busulfan and reported to have a low non-haematological toxicity (12).

In the Ewing 2008 R3 arm, Koch et al. reported on 109 patients with primary disseminated ES randomly assigned to either treosulfan-melphalan (TreoMel) HDT (n = 55) or no further treatment following consolidation chemotherapy (n = 54). Patients with pulmonary only metastases were excluded. No significant difference in EFS (HR 0.72, 95% CI 0.46-1.12, p=0.37) or OS (HR 0.96 (95% CI, 0.58-1.58, P=0.87)) was observed in the overall cohort, however a 3-year EFS benefit with HDT was seen in children aged less than 14 years old (39.3% vs 9%, p=0.016; HR 0.40 (0.19-0.87)(10). One HDT-related death was reported.

In the remaining nine retrospective and three single arm prospective studies (without a comparison arm), interpretation of OS and EFS is difficult due to wide inclusion criteria in terms of staging, inclusion of different types of small blue round cell tumours without advanced genomics testing, and differences in the backbone chemotherapy regimens used (technical report – table 1)(7, 13-23). Nine additional studies included patients treated with HDT upfront, as well as at relapse (technical report – table 3)(24-32).

Some of these older studies have reported improvements in outcomes in patients receiving HDT versus conventional chemotherapy.  Drabko et al. retrospectively evaluated 102 paediatric and adolescent patients (age 1-19 years) with ES and primitive neuroectodermal tumours (PNET). High risk disease was defined as metastatic disease or localised with unresectable or poor histological response to induction chemotherapy. They found a significant improvement in 2-year OS and relapse-free survival (RFS) in 35 high-risk patients who underwent HDT and radiotherapy versus 21 patients who received conventional chemotherapy (OS  71% vs. 31%, p<0.01 and RFS 66% vs. 27% respectively, p < 0.01)(14). Of those treated with HDT, 69% were less than 14 years old. Serlo et al. retrospectively analysed an aggregated group of 76 patients receiving various unspecified HDT or conventional chemotherapy consolidation regimens (32). In 15 patients with metastatic disease, where 9 received HDT, a significant increase in 5-year disease-specific survival post HDT (74% vs. 0%, p<0.001) was observed.  In patients with localised disease, no significant improvement in 5-year disease-specific survival was seen with HDT (67% with HDT vs. 71%, p=0.662), however a non-significant improvement in disease-free survival (DFS) was observed (67% with vs. 59%, p=0.817). Other studies were not able to show a significant difference in OS or EFS for patients with primary metastatic or relapsed ES (22, 24), or localised disease with poor radiological or histological response treated with HDT (15). Gaspar et al. and Jahnukainen et al. report lower 5-year and 10-year OS rates for patients who underwent HDT compared to conventional chemotherapy, however reliable conclusion regarding HDT cannot be drawn from these studies, given allocated treatments were based on patients’ risk group (16, 28). In the Gaspar et al. study, the 5-year OS of 48 high-risk patients (poor histological response or unresectable primary) who received HDT was 48% (95% CI 35-68%) versus 54% (95% CI 40-68%) and 70% (95% CI 61-77%) for intermediate and standard-risk patients receiving conventional chemotherapy respectively (16). Jahnukainen et al. showed a 10-year OS rate of 73% (SD 16%) in 8 high-risk patients treated with HDT in first complete remission versus 90% (SD 9%) in 16 standard-risk patients treated with conventional chemotherapy (28).

Other non-comparative studies are small, enrolling between 18 and 57 participants who underwent HDT (see technical report, Table 1&3). In summary, most of these studies are retrospective and heterogeneous in nature and the evidence is weak towards the benefit of using HDT as part of first-line therapy.

Univariate and multivariate analyses in first line studies
The impact of three notable variables on outcomes after HDT were examined across multiple studies: initial stage at diagnosis, clinical remission pre-HDT and histological response to induction chemotherapy.  Diaz et al. reported the presence of metastases at diagnosis to be the most significant negative prognosticator (13). Incomplete response at time of HDT was also associated with lower EFS or OS in newly diagnosed ES patients in multiple studies (11, 13, 14, 23, 30).  Primary multifocal disseminated ES patients in partial remission (PR) pre-HDT achieved a lower 3-year EFS rate post-HDT versus patients in complete remission (CR), and patients with stable (SD) or progressive disease (PD) at the time of HDT achieved poorer results versus patients in partial remission (3-year EFS 24% ± 7% in SD/PD vs. 32% ± 5% in PR vs. 57% ± 10% in CR, p=0.01)(11).

The impact of histological response to induction chemotherapy in primary localised ES on outcome post-HDT has been reported inconsistently. Laurence et al. reported a significant survival benefit from HDT in ESFT patients who had a good histological response (17). Gaspar et al suggested patients with poor histological response seemed to benefit from HDT, when EFS was compared with historical control that did not receive HDT (EFS 45% (95% CI, 30-60) and 20% (95% CI, 7-43) for EW93 (HDT therapy) and EW88 study patients, respectively)(16). In Whelan et al., patients with an intermediate poor response to initial therapy benefited more from HDT than those with a very poor response (10% to 29% vs. 30% viable cells; P=0.06)(8).

Additionally, in univariate analysis, younger patients might benefit more from HDT than older patients. In Whelan et al., patients younger than 25 years old treated with HDT had a better outcome (p=0.12)(8). Similarly, patients with primary disseminated multifocal ES younger than 14 years old had a better outcome following TreoMel HDT (HR 0.40, 95% CI 0.19-0.87)(10). Pawlowska et al. identified older age (>18) to confer a higher risk of relapse regardless of remission status at the time of HDT (30).

HDT for relapsed disease in Ewing Sarcoma

Fourteen studies evaluated patients with relapsed or refractory Ewing sarcoma (24-37), including nine studying outcome post HDT both upfront and at relapse (Table 2 & 3). The four retrospective cohort studies reporting specifically the outcome of patients with relapsed/refractory ES after HDT reported a significant increase in outcomes with melphalan-based HDT versus conventional chemotherapy (34-37). In Ferrari et al., 5-year post-relapse survival was 50% (95%, CI 28-72%) in 20 patients who responded to high-dose ifosfamide and proceeded to receive BuMel HDT; 12% (95% CI 2-25%) in 30 patients treated with high-dose ifosfamide chemotherapy alone due to prior HDT or inadequate response to ifosfamide; and 5% (95% CI 4-14%) for other conventional treatment regimens (n = 30)(34). Rasper et al. specifically reported the outcome of relapsed patients who achieved CR or PR following 4 to 6 cycles of conventional chemotherapy, some then receiving HDT (35). The 2-year and 5-year EFS were 44% (SD 9%) and 22% (SD 9%, p=0.93) and the 2-year and 5-year OS were 59% (SD 9%) and 41% (SD 10%, p=0.13) respectively for the 7 patients who received BuMel or TreoMel HDT consolidation in CR or PR. For the 34 patients in CR or PR after conventional chemotherapy without HDT consolidation, the 2-year and 5-year EFS were 31% (SD 8%) and 18% (SD 8%) and the 2-year and 5-year OS were 45% (SD 9%) and 25% (SD 8%)(35). A small retrospective cohort study by Shankar et al. described a median survival of 49 months for the 7 patients who received HDT, including 4 in CR and 2 in PR at the time of HDT (response rate (RR) 85%), compared to 14 months for 28 patients continuing a conventional regimen, including 3 in CR and 11 in PR (RR 58%)(36). Windsor et al. reported a 2-year and 5-year post-relapse survival of 67.9% (SD 5.9%) and 52.7% (SD 6.5%) respectively in 64 patients who underwent HDT, compared to 20.5% (SD 20.5%) and 2% (SD 1.5%) respectively in 98 patients continuing conventional chemotherapy (37). Of the 64 patients receiving HDT, response to induction chemotherapy prior to HDT was available for 55; responses included CR, PR, SD, PD in 21 (38%), 28 (51%), 4 (7%), and 2 (3%) patients were respectively. The remission status of patients not receiving HDT was not reported. The remaining studies are non-controlled and with heterogeneous inclusion criteria, with results as shown in table 3 (technical report). In the largest study (n=109), 27 patients with localized disease at diagnosis received HDT after recurrence and had a 5-year PFS of 14% (95% CI 3–30%)(27).

Univariate and multivariate analysis in relapsed/refractory studies

Gardner et al. reported poorer survival in patients with refractory disease immediately prior to HDT (27). In the Rasper and Pawloswki studies, multivariate analyses both found early relapse (<2 years from diagnosis) was an unfavourable prognostic factor (HR 4.76 (EFS, p<0.01) and 3.70 (OS, p<0.01)(35) and similarly, HR 0.36 (95% CI, 0.14% to 0.92%; P = .032) for patients with late relapse, who received HDT at relapse (30). Finally, in the Rasper study, patients in CR/PR had more risk of event when not receiving HDT, with HR of 2.90 (EFS, P < 0.01) and 2.61 (OS, P=0.01).

Toxicities, treatment related mortality and secondary malignancies

Twenty-six studies reported on treatment related deaths, six studies reported on secondary malignancies (8-10, 14, 17, 35), and five studies provided a comparison of toxicities in each treatment group (8-10, 15, 22). Four of the 30 studies provided no specific information on toxicity (7, 32, 36, 37).

High rates of severe toxicities were observed with HDT. In the prospective randomised studies by Whelan et al., and Koch et al, grade 4 haematological and grade 3 and above non-haematological toxicities were classified as severe toxicities (8, 10). In Whelan et al. severe haematological, gastro-intestinal or respiratory tract toxicities, and fever or infection were observed in 98.8.%, 71.2% and 32.5%, respectively, of patients who received BuMel chemotherapy (8). In Koch et al. study, severe haematological, gut, infection and deterioration in general condition toxicities were recorded in 98%, 48%, 39% and 26%, respectively, of patients who received TreoMel-HDT (10).

In the five studies where a comparison of toxicities in each treatment group was provided, all concluded additional severe acute toxicities occurred in the HDT arm. Koch et al. documented more infections, gastro-intestinal (GI) and renal toxicities in the TreoMel-HDT group (P < 0.05)(10). It was also noted that severe toxicity (grade 4 hematologic toxicities and grade 3+ non-hematologic toxicities as per the National Cancer Institute Common Terminology Criteria for Adverse Events (38)) were more likely in this group. Dirksen et al. reported severe acute toxicities, in particular, GI, liver and haematological toxicities, infections, as well as other toxicities such as pain, fatigue, fever, chills and malaise, were more common in patients receiving BuMel (P < 0.05)(9). Comparable toxicities were observed in the study by Whelan et al. (P < 0.05) and Paulussen et al (8, 22). Ferrari et al. noted that haematological toxicities were higher in the HDT group, reporting 3% of patients to be thrombocytopenic following VAC-IVA-IE versus 95% following HDT (15). Although data of late effects was not collected, Dirksen et al. discuss infertility, in male and female patients, is an additional side effect anticipated with HDT that is not expected in female patients after standard therapy and WLI (9).

Treatment related deaths, with HDT, were recorded in 25 of 26 studies where treatment related toxicities were commented upon. Gardner et al. found day 100 and 1-year post HDT treatment-related mortality (TRM) to be 4% and 9% respectively in ES patients treated with various myeloablative chemotherapy regimens (with or without total body irradiation) (27).

Secondary malignancies were discussed in six studies. Drabko et al. reported three secondary malignancies, all in the group without HDT (14). Rasper et al. reported four deaths from secondary malignancies (acute myeloid leukemia and melanoma), without precision on allocated treatment(35). Laurence et al. did not find any secondary malignancies in the surviving patients at a median follow-up of 7.1 years (17). Three other studies showed the number of secondary malignancies to be similar across the HDT and conventional chemotherapy arms (8-10).

Discussion

The literature, which includes prospective international multicentre randomised studies, demonstrates inconsistent results for survival benefit with HDT.  As first line therapy of ES patients, melphalan-based myeloablative chemotherapy with ASCT may convey an OS and EFS benefit to ES patients with high-risk localised disease. Survival benefits are not confirmed for patients with metastatic disease at initial diagnosis. This benefit for high-risk localised disease, however, was demonstrated in European studies where the backbone of first-line chemotherapy was not interval compressed chemotherapy with vincristine, doxorubicin, cyclophosphamide (VDC) alternating with ifosfamide/etoposide (IE) every 2 weeks. This two-weekly dose-intensive regimen is now proven to be empirically superior and is the recommended first line chemotherapy regimen in ES (39). Therefore, for localised ES patients treated with the now standard, 2-weekly VDC/IE regimen, the role of HDT is unknown. Clinical trials testing HDT consolidation with this regimen are needed to determine if it remains beneficial, and thus HDT for high-risk localised ES can not be recommended at this time. For treatment of relapsed/refractory ES, HDT with ASCT remains a potential option. The evidence appears greatest for patients who achieve complete or partial remission before HDT. In these patients, there seems to be a survival benefit over conventional chemotherapy only (23-26). Patients and/or their guardians should be guided judiciously through the potential but uncertain benefits of HDT, where toxicities are often severe in grade and include a higher risk of treatment related death. Important factors such as child and carer psychological distress, changes in functional status and quality of life following HDT are not considered in any of the included studies.

Evidence summary	Level	References
Melphalan-based high-dose chemotherapy may be associated with improved survival in primary localised high-risk Ewing sarcoma treated with a VIDE chemotherapy backbone*, but not in primary metastatic disease.	II	(8-10, 20)
Melphalan-based high-dose chemotherapy may be associated with improved survival in selected patients with refractory/relapsed Ewing sarcoma.	III	(34-37)
Treatment response to initial therapy prior to high-dose chemotherapy is an important determinant of survival outcomes in patients undergoing high dose chemotherapy with autologous stem cell transplant (CR > PR > SD or PD).	III, IV	(11, 14, 23, 30)
Severe acute toxicities are more common following high-dose chemotherapy compared to conventional chemotherapy.	II	(8-10)
Evidence-based recommendation	Grade
Selected patients with relapsed Ewing sarcoma could be considered for high dose chemotherapy with autologous stem cell transplant. These patients should be managed in specialised sarcoma centres.	B

*Former first line European regimen now replaced by interval compressed chemotherapy with vincristine, doxorubicin, cyclophosphamide alternating with ifosfamide/etoposide 2 weekly (39)

Administration Report

Technical Report

References

► Topic 3: Paediatric/AYA Sarcoma