Bone Marrow Transplantation, (1997) 19, 227–231  1997 Stockton Press All rights reserved 0268–3369/97 $12.00

Do patients with metastatic and recurrent rhabdomyosarcoma benefit from high-dose therapy with hematopoietic rescue? Report of the German/Austrian Pediatric Bone Marrow Transplantation Group.

E Koscielniak1, TH Klingebiel2, C Peters3, J Hermann4, ST Burdach5, C Bender-Go¨tze6, ST Mu¨ller-Weihrich6 and J Treuner1

1Department of Pediatric Oncology, Olga Hospital, Stuttgart; 2University of Tu¨bingen, Germany; 3St Anna Kinderspital, Vienna, Austria; 4University of Jena; 5University of Du¨sseldorf; and 6University of Mu¨nchen, Germany

Summary: of patients at previously known sites; in three cases new metastatic sites were observed. Patients with pri- Patients with primary metastatic or recurrent rhabdo- mary localized tumors who had been treated with HDC myosarcoma (RMS) have a very poor prognosis. Since because of relapse did slightly better (four of nine alive -high-dose chemotherapy (HDC) ؎ TBI was thought to with NED) than patients with primary metastatic dis improve survival, many centers performed this ther- ease (five of 27 alive with NED). HDC is still of uncer- apy using different types of hematopoietic rescue (auto tain value in the therapy of poor-risk rhabdomyosar- BM or PBSC, allo BM). This is a retrospective, multi- coma and should be performed only as part of center analysis of the results of treatment in 36 patients controlled clinical trials. with primary metastatic or relapsed RMS who were Keywords: rhabdomyosarcoma; high-dose therapy; soft given HDC ؎ TBI and hematopoietic rescue between tissue sarcoma 1986 and 1994. The median age was 6 years (Ͻ1–22 years). Primary therapy was given according to either one of the Cooperative German Soft Tissue Sarcoma Studies CWS-81, -86, -91 or the European Study for Rhabdomyosarcoma (RMS) is the commonest, highly Stage IV Malignant Mesenchymal Tumors in Child- malignant soft tissue sarcoma of childhood. The prognosis hood. There were 22 alveolar RMS, 13 embryonal of patients with localized rhabdomyosarcoma has improved RMS and one undifferentiated sarcoma. The indication in the last 20 years, with an EFS rate of approximately for HDC was primary metastatic disease (27 patients) 70%.1–3 In contrast, the chance of cure in primary meta- or a relapse of a primary localized tumor (nine static or relapsed tumors is very poor. The 5 year survival patients). Thirty-two patients were in 1st or 2nd CR rates of patients with primary disseminated tumors in the when given HDC and four in VGPR. The median time Intergroup Rhabdomyosarcoma Studies IRS I, II and III from last event to HDC was 44 weeks (21–110). HDC were between 20 and 30%. In the German Soft Tissue Sar- consisted of fractionated melphalan ((4 ؋ 30– coma Study CWS-81 it was 19% and in the European Study mg/m2), VP16 40–60 mg/kg, carboplatin 3 ؋ 400– for Stage IV Malignant Mesenchymal Tumors in Childhood 45 500 mg/m2) in 26 patients, 10 of whom received (MMT Stage IV) the survival rate at 3 years was 24.7%.1–6 additional FTBI. Seven patients were treated with mel- It is interesting to note that the results of treatment have phalan alone or in combination with carboplatin. Two changed little over the last 20 years. The chance of being patients received cyclophosphamide/busulphan with cured after relapse of primary localized RMS depends on TLI (total lymphoid irradiation) and one cyclophos- the intensity of the primary chemotherapy, on the histologi- phamide with FTBI. Thirty-one patients were given cal subtype and whether radiotherapy was given. In the autologous BM or PBSC as hematopoietic rescue and German Cooperative Soft Tissue Sarcoma Studies CWS-81 five allogeneic bone marrow from HLA-identical sib- and CWS-86 only patients with embryonal rhabdomyosar- lings. Fourteen patients received GM-CSF or G-CSF coma who relapsed locally and in whom local irradiation after hematopoietic stem cell transfusion (HSCT). Ten was still feasible had a slightly better chance of survival patients received adjuvant IL-2. There was one toxic with a survival rate at 3 years of 62% (data not published). HDC-related death. Nine patients are alive and free of For the remaining patients the EFS rate at 2 years did not disease with a median observation time of 57 months exceed 20%.7 Since RMS is a chemosensitive tumor a (32–108). The median time from HDC to relapse was further escalation of dose intensity seemed to be the best 4 months (1–17). The tumor recurred in the majority way of improving results. Many centers in Europe and USA have performed high-dose chemotherapy with hemato- poietic stem cell transfusion (HSCT) as a late consolidation Correspondence: Dr E Koscielniak, Department of Pediatric Oncology and of therapy in patients with metastatic and relapsed solid 8 Hematology, Olga Hospital, Bismarckstr. 8, 70176 Stuttgart, Germany tumors. However, the prognostic benefit of this procedure Received 21 March 1996; accepted 9 September 1996 is still unproven.8–14 High-dose chemotherapy in rhabdomyosarcoma E Koscielniak et al 228 We present a retrospective analysis of the results of was treated according to the German Cooperative Study for treatment in patients with metastatic or relapsed rhabdo- relapsed sarcomas (CWS/CESS Relapse-Study). Treatment myosarcoma who were given HDC with HSCT as a con- regimens used included: VACA (vincristine, dactinomycin, solidation of CR or very good PR. Patients were registered cyclophosphamide and doxorubicin), VAIA (vincristine, in the German/Austrian Pediatric Bone Marrow Trans- dactinomycin, ifosfamide and doxorubicin), EVAIA plantation Register in Tu¨bingen and were treated between (additional VP16), CEVAIE (carboplatin, epiadriamycin, 1986 and 1994 in several centers in Germany and Austria vincristine, ifosfamide, dactinomycin and VP16). The rec- (Appendix 1). The endpoint for the analysis was 31 ommended chemotherapy duration for metastatic patients December 1995. in the CWS studies was 56 (CWS-81) or 46 weeks (CWS- 86). In the MMT Stage IV study it was 26 weeks. Since local therapy (irradiation or surgery on primary tumor Ϯ metastases) was also given, the complete duration of ther- Materials and methods apy before HDC was even longer than 56 weeks in sev- eral patients. The clinical characteristics of the 36 patients are given in The median interval between the last event that led to Table 1. The median age at diagnosis was 6 years (Ͻ1– eligibility for HDC and the HDC itself was 44 weeks (21– 22). Alveolar rhabdomyosarcoma was the predominant his- 110). The latest event before HDC was defined either as a tological subtype, occurring in 22 (61%) patients. Primary relapse or a diagnosis of primary metastatic RMS. Thirty- metastatic disease was the indication for HDC in 27 (75%) two patients were in first or second CR before HDC and patients. Nine patients received HDC because of relapse four in VGPR. High-dose therapy in 26 patients consisted (four systemic, four lymph nodes and one local relapse) of of fractionated melphalan (4 ϫ 30–45 mg/m2), VP16 (40– their primary localized tumors. 60 mg/kg) and carboplatin (3 ϫ 400 mg/m2), the so-called The treatment received before high-dose therapy was MEC schema, developed in Vienna and Du¨sseldorf.9,13 Ten according to the protocols of the German Cooperative Soft of 26 patients given MEC therapy received fractionated Tissue Sarcoma Studies (CWS-81, CWS-86 and CWS-91) total body irradiation (FTBI), administered to a total dose or the European Study for Stage IV Malignant Mesenchy- of 12 Gy (2 ϫ 1.5 Gy/day), parallel to melphalan appli- mal Tumors in Childhood (MMT Stage IV). One patient cation. Five patients received melphalan only (140– 180 mg/m2), two patients melphalan in combination with carboplatin and one patient received MEC with BCNU. Two of five patients who were given allogeneic bone mar- Table 1 Patients characteristics and outcome (n ϭ 36) row from related donors received cyclophosphamide with busulphan and total lymphoid irradiation (TLI). One patient Median age (at diagnosis) years 6(Ͻ1–22) received cyclophosphamide with FTBI. The hematopoietic Histology rescue consisted of autologous BM or peripheral hemato- RMS alveolar 22 RMS embryonal 13 poietic stem cells in 31 patients (86%), 14 of whom were Undifferentiated sarcoma 1 given G-CSF or GM-CSF support. In five patients allo- Indication for HDC geneic BM from related HLA-identical donors was trans- Primary metastatic 27 fused. Ten of 33 patients who were given autologous HSCT Relapsed 9 received immunologic modulation for 2–3 months with Tumor status at HDC 1CR 27 recombinant DNA-derived interleukin 2 (Proleukin; Euro- 2CR 5 Cetus, Frankfurt, Germany). Three cycles were adminis- VGPR 4 tered at increasing doses from 6 (day 1), 9 (day 2), to 12 Therapy before HDC (days 3 to 5) ϫ 106 IU/kg with a 2–3 week interval CWS-81, CWS-86, CWS-91 23 MMT Stage IV 12 between cycles. CWS/CESS relapse protocol 1 HDC MEC ϩ TBI 26 Melphalan 5 Results Melphalan ϩ Carboplatin (ME) 2 MEC ϩ BCNU 1 CY-BU ϩ TLI 2 Nine patients are alive with no evidence of disease (NED) Interval last event to graft (weeks) with a median follow-up of 57 months (32–108) after diag- Median 44 nosis and 27 months (20–100) after HDC and one is alive Range 21–110 Graft source in 2nd CR after additional local treatment of relapse occur- Autologous ring 8 months post-HDC. There was one toxic death as a BM 26 result of sepsis. There were no statistically significant dif- PBSC 5 ferences in the distribution of survivors among different Allogeneic BM 5 subgroups (histological and therapeutic) as shown in Figure Outcome Alive 10 1. However, some trends did emerge. Patients with primary Therapy-related death 1 localized tumors who had been treated with HDC because Disease-related death 25 of relapse did slightly better (four of nine alive with NED) than patients with primary metastatic disease (five of 27 High-dose chemotherapy in rhabdomyosarcoma E Koscielniak et al 229 given HDC with HSCT as late consolidation of therapy. 100 1 The patients were registered in the German/Austrian Pedi- 5 4 2 9 4 8 7 atric Bone Marrow Transplantation Register. They formed 80 4 a relatively homogenous group with respect to the diagnos- tic and therapeutic concept prior to HDC since they were 60 all treated either according to the German Cooperative Soft Tissue Sarcoma Studies or the European Study for Meta- 9 40 27 22 18 188 19 static Mesenchymal Tumors. The HDC consisted, in the Percentage 9 5 majority of patients, of fractionated melphalan, carboplatin 20 and VP16. This combination was developed for consoli- dation of therapy in children with solid tumors in St Anna 0 9 Total Stage IV Relapse Alv. Embr. TBI No TBI IL2 No IL2 Children’s Hospital in Vienna and was shown to be effec- tive in patients with metastatic Ewing’s sarcoma.13 There are no data about the effectiveness of single agents such as Failure carboplatin and VP16 in rhabdomyosarcoma patients. No evidence of disease There is, however, evidence that VP16 combined with car- boplatin is effective in relapsed solid tumors in childhood.17 Figure 1 Distribution of survivors with no evidence of disease (NED) among different clinical subgroups. Melphalan has been proven to have high activity against rhabdomyosarcoma in xenograft models and in vivo.18 However, there is currently no evidence that high-dose melphalan alone improves the prognosis of children with alive with NED). Of the four patients who are alive after metastatic RMS.11 Many centers cooperating in the receiving HDC after relapse, three had lymph nodes recur- German/Austrian Pediatric Bone Marrow Transplantation rence and one distant metastases. Similarly, those with Group have therefore adapted (after 1990) the MEC schema embryonal RMS had a better chance of disease-free sur- as HDC in patients with poor prognosis rhabdomyo- vival than patients with alveolar RMS, but the difference sarcoma. was not significant (four of 13 vs four of 22). There was Nine of 36 patients analyzed in this study are alive with no trend for a better survival in patients treated after HDC NED with a median follow-up of 57 months after diagnosis. with IL-2. Similarly, there was no difference in prognosis There was a trend for better survival in patients who were between patients who received FTBI vs no FTBI. None of treated with HDC because of relapse of localized RMS in the five patients who received allogeneic BM are alive. The comparison to patients with primary metastatic disease. median interval between the last event that led to eligibility However, the treatment results are no better when com- for HDC and the HDC itself was similar for the survivors pared to patients with metastatic or relapsed RMS regis- as compared with the entire group, ie 43 weeks (21–110) tered in the German Soft Tissue Sarcoma Studies who were vs 44 (21–110). Relapse after HDC occurred early with a not given HDC.5,6 median interval of 4 months (1–17). Interestingly, most There are many questions concerning dose escalation and relapses (23) were at previously known sites with only three its relevance for cure. It has been shown in the case of acute relapsing in new metastatic sites. The estimated EFS rate myelogenous leukemia that no consistent gain in survival at 2 years after HDC is 36 Ϯ 7%, at 2 years after diagnosis although some improvement in response rate was seen with it is 55 Ϯ 8%. a dose of daunorubicin greater than 30 mg/m2.19 No one knows what is more important for primary metastatic RMS: short high-dose intensive therapy or less intensive therapy over a longer duration. Perhaps these two factors should be Discussion considered separately depending on the phase of therapy. In the first phase, until a complete remission is achieved, Despite the major improvement in prognosis for patients defined as no clinical signs of disease, more intensive ther- with primary localized rhabdomyosarcoma during the last apy is probably warranted. However, by what means should 20 years, no significant progress has been made in the ongoing microscopic disease be eradicated? Intensive short treatment of patients with primary metastatic or relapsed consolidation or low dose but longer therapy? This question disease.1–15 Rhabdomyosarcoma is a chemosensitive has not been answered for poor prognosis RMS. Based on tumor and the role of chemotherapy, both adjuvant and the CWS studies treatment results, one can see that the preoperative, is well established. The intensification of dose–response relationship does not always correlate with chemotherapy by escalating the dose and combining more better survival. In the first German Soft Tissue Sarcoma drugs has been shown to improve results, especially in Study CWS-81, relative intensive first-line chemotherapy locally advanced, primary non-resectable rhabdomyosar- with cyclophosphamide, doxorubicin, dactinomycin and coma.1,3,16 The dose–response relationship which has been vincristine was recommended for all patients. The treatment seen in many solid tumors was the theoretical basis for results were very good3 compared to other published ser- dose escalation beyond the ability of the bone marrow to ies.1,2,4,16 In the CWS-86 Study there was an intensification reconstitute spontaneously within an acceptable period. of first-line chemotherapy by replacing cyclophosphamide We have performed a retrospective analysis of patients by ifosfamide. An improvement in response rate without with primary metastatic and relapsed RMS who have been any change in the overall results was observed.20 Similarly High-dose chemotherapy in rhabdomyosarcoma E Koscielniak et al 230 the prospective German–Italian study comparing ifosfam- 6 Carli M, Pinkerton R, Oberlin O et al. Risk group analysis ide 6 g/m2 vs 9 g/m2 in patients with primary localized non- in metastatic soft tissue sarcomas (STS) children. European resectable RMS showed a better response for the 9 g/m2 Intergroup Study MMT 89. Proc ASCO 1995; 14: 449. arm but the overall survival did not differ.21,22 One can con- 7 Klingebiel Th, Bode U, Ju¨rgens H et al. Treatment of relapse clude that the dose-response-survival correlation proved in soft tissue and Ewing’s sarcoma patients. A phase II trial 19 (CESS/CWS REZ 91). Med Pediatr Oncol 1993; 21: 573 true only for some degree of dose escalation. To our Abstr. knowledge, there are no published data on survival of larger 8 Seeger RC, Reynolds CP. Treatment of high-risk solid tumors series of patients with poor prognosis RMS after HDC. of childhood with intensive therapy and autologous bone mar- Most published series include different soft tissue sarcomas row transplantation. Pediatr Clin North Am 1991; 38: 393– or even other solid tumors with different biological 424. behavior.8,11 The 1995 analysis of the European Bone Mar- 9 Emminger W, Emminger-Schmidmeier W, Hawliczek R et al. row Transplantation (EBMT) Solid Tumor Registry showed High-dose melphalan, etoposide Ϯ carboplatin (MEC) com- a 5-year survival rate of 18% for 197 patients with rhabdo- bined with 12-gray fractionated total-body irradiation in chil- myosarcoma who received high-dose therapy with HSCT dren with generalized solid tumors. Pediatr Hematol Oncol and 28% for those who were grafted in first CR (74 1991; 8: 13–22. patients). 10 Shuster JJ, Cantor AB, McWilliams N et al. The prognostic The survival of children who received IL-2 after HDC significance of autologous bone marrow transplant in advanced neuroblastoma. J Clin Oncol 1991; 9: 1045–1049. was no better. Administration of rIL-2 during the first 11 Pinkerton CR, Groot-Loonen J, Barret A et al. Rapid VAC, months after HDC was thought to be effective because of high dose melphalan regimen. A novel chemotherapy a relative excess of circulating NK cells with potential approach in childhood soft tissue sarcomas. Br J Cancer 1991; lymphokine-activated killer (LAK) activity after rIL-2 64: 381–385. stimulation.23,24 To date there have been no published data 12 Pinkerton CR. Massive chemotherapy followed by bone mar- showing that the therapeutic concept of immunomodul- row graft in pediatric oncology: arguments against. Bull Can- ation with rIL-2 or other cytokines after HDC is effective cer Paris 1995; 82: 42–45. in improving the prognosis of children with solid tumors. 13 Burdach St, Ju¨rgens H, Peters C et al. Myeloablative radio- The role of total body irradiation in the eradication of chemotherapy and hematopoietic stem cell rescue in poor residual rhabdomyosarcoma cells is unclear. Some centers prognosis Ewing sarcoma. J Clin Oncol 1993; 11: 1482–1488. used it particularly in patients with multiple bone and/or 14 Ladenstein R, Hartmann O, Pinkerton CR. The role of mega- bone marrow metastases in whom effective local therapy with autologous bone marrow rescue in solid tumours of childhood. Ann Oncol 1993; 4 (Suppl. 1): 45–58. irradiation to all known tumor sites is not feasible. In 15 Treuner J, Flamant F, Carli M. Results of treatment of rhabdo- patients with metastatic Ewing’s sarcoma TBI has been 13 myosarcoma in the European studies. In: Maurer HM, Ruy- shown to produce some prognostic benefit. In five mann FB, Pochedly C (eds). Rhabdomyosarcoma and Related patients, allogeneic BMT had been performed (before Tumors in Children and Adolescents. CRC Press: Boca Raton, 1990) because of speculation that the immune function 1991, pp 227–241. of allogeneic BMT may be used as a therapeutic tool in 16 Flamant F, Hill C. The improvement in survival associated producing an anti-tumor effect. At present no benefit from with combined chemotherapy in childhood rhabdomyosar- allogeneic vs autologous BM has been shown in patients coma. Cancer 1984; 53: 2417–2421. with solid tumors.25 17 Calaminus G, Ju¨rgens H, Schwamborn D et al. Experience In summary, it can be said that the HDC in poor prog- with the combination carboplatin/Vp16 in recurrent childhood nosis rhabdomyosarcoma cannot be regarded as an estab- tumors. Klin Paediatr 1989; 13: 311–315. lished therapy method leading to a better prognosis. In 18 Horowitz ME, Etcubanas E, Christensen ML et al. Phase II testing of melphalan in children with newly diagnosed rhabdo- order to answer definitively, whether HDC has a place in myosarcoma: a model for anticancer drug development. J Clin the treatment of rhabdomyosarcoma, it should be performed Oncol 1988; 6: 308–314. only within prospective controlled trials. 19 Berman E. Chemotherapy in acute myelogenous leukemia: high-dose, higher expectations? J Clin Oncol 1995; 13: 1–4. 20 Treuner J, Koscielniak E, Keim M. Comparison of the rates References of response to ifosfamide and cyclophosphamide in primary unresectable rhabdomyosarcomas. Cancer Chemother Pharm- 1 Maurer HM, Beltangady M, Gehan EA et al. The Intergroup acol 1989; 24 (Suppl. 1): 48–50. Rhabdomyosarcoma Study-I. A final report. Cancer 1988; 61: 21 Koscielniak E, Carli M, Andrello L, Treuner J. Two different 209–220. regimens with ifosfamide as preoperative chemotherapy in 2 Crist WM, Gehan EA, Ragab AH et al. The third Intergroup children with unresectable rhabdomyosarcoma. A report Rhabdomyosarcoma Study. J Clin Oncol 1995; 13: 610–630. from the German Soft Tissue Sarcoma Study CWS-86 and 3 Koscielniak E, Ju¨rgens H, Winkler K et al. Treatment of soft the Italian Rhabdomaosarcoma Study ICG-88. In: Banzet P, tissue sarcoma in childhood and adolescence. Cancer 1992; Holland JF, Khayat D, Weil M (eds). Proceedings of the 70: 2557–2567. Third International Congress on Neo-Adjuvant Chemo- 4 Maurer HM, EA Gehan, Beltangady M et al. The Intergroup therapy. Springer Verlag: Paris, Heidelberg, New York, Rhabdomyosarcoma Study-II. Cancer 1993; 71: 1904–1922. 1991, pp 331–334. 5 Koscielniak E, Rodary C, Flamant F et al. Metastatic rhabdo- 22 Carli M, Treuner J, Koscielniak E et al. Ifosfamide (ifo) more myosarcoma and histologically similar tumors in childhood: a is better? 6 g vs 10 g/m2 in VAIA may influence the tumor retrospective European multi-center analysis. Med Pediatr response rate in childhood rhabdomyosarcoma (RMS)? The Oncol 1992; 20: 209–214. experience of the German (CWS 86) and the Italian (ICS- High-dose chemotherapy in rhabdomyosarcoma E Koscielniak et al 231 RMS 88) cooperative studies. Proc ASCO 1991; 11: 319 (Abstr). Appendix: Participating centers: 23 Favrot MC, Michon J, Floret D et al. Interleukin 2 immun- otherapy in children with neuroblastoma after high dose Olga Hospital, Stuttgart chemotherapy and autologous bone marrow transplantation. University Childrens Hospital, Tu¨bingen Pediatr Hematol Oncol 1990; 7: 275–284. University Childrens Hospital, Munich 24 Laws HJ, Diloo D, Hanenberg H et al. Il2 therapy post auto- University Childrens Hospital, Jena logous stem cell transplantation stimulates the production of University Childrens Hospital, Duesseldorf tumor cytotoxic cytokines in children and adolescents with University Childrens Hospital, Graz solid tumors. Klin Pa¨diatr 1993; 205: 257–263. University Childrens Hospital, Berlin 25 Ladenstein R, Lasset C, Hartmann O et al. Comparison of auto versus allografting as consolidation of primary treatments University Childrens Hospital, Hannover in advanced neuroblastoma over one year of age at diagnosis. University Childrens Hospital, Hamburg Bone Marrow Transplant 1994; 14: 37–46. St Anna Kinderspital, Vienna