Neo-adjuvant chemotherapy for primary high-grade extremity soft tissue sarcoma

Annals of Oncology 15: 1667– 1672, 2004 doi:10.1093/annonc/mdh431

Original article

Neo-adjuvant chemotherapy for primary high-grade extremity soft tissue sarcoma S. R. Grobmyer1, R. G. Maki2, G. D. Demetri3, M. Mazumdar4, E. Riedel4, M. F. Brennan1 & S. Singer1* Departments of 1Surgery, 2Medicine and 3Biostatistics, Memorial Sloan Kettering Cancer Center, New York, NY; 4Department of Medicine, Dana Farber Cancer Institute, Boston, MA, USA Received 12 March 2004; revised 10 June 2004; accepted 18 June 2004

Introduction Soft tissue sarcomas (STSs) are a heterogeneous group of tumors that most commonly arise in the extremities [1, 2]. Sarcoma-related mortality from extremity lesions occurs most commonly secondary to hematogenous metastasis [1]. While chemotherapy has established efficacy in reducing metastasis and prolonging survival in several specific subtypes of childhood STSs (rhabdomyosarcoma [3] and Ewing sarcoma [4]), its value in treating most other histologic types of primary sarcoma remains controversial. Many randomized trials of chemotherapy for STS conducted to date have had small sample size, risk factor imbalance between treatment and no-treatment arms, suboptimal chemotherapy and have involved a heterogeneous mix of tumor sites and grade. As a result, these trials have yielded variable results and have been difficult to interpret. Despite these inconsistent data, adjuvant or neo-adjuvant chemotherapy (NAC) is commonly used at many centers as

*Correspondence to: Dr S. Singer, The Department of Surgery, Memorial Sloan Kettering Cancer Center, 1275 York Avenue, New York, NY 10021, USA. Tel: +1-212-639-2940; Fax: +1-646-422-2300; E-mail: [email protected] q 2004 European Society for Medical Oncology

part of a treatment plan for patients with primary extremity STSs. Proponents have suggested several potential benefits including an ability to assess sarcoma response to a given chemotherapeutic regimen, earlier treatment of microscopic metastatic disease and facilitation of tumor removal [5, 6]. In spite of these theoretical benefits, the value of NAC in improving survival of patients with STS remains unproven. This study was undertaken to retrospectively determine the association of NAC with survival and recurrence in a group of patients with high risk (large, high-grade) extremity STS.

Patients and methods Prospectively maintained databases of all patients diagnosed with STS treated at either Dana Farber Cancer Institute (DFCI) (database started in 1978) or Memorial Sloan Kettering Cancer Center (MSKCC) (database _ 5 cm), highstarted in 1982) were queried to identify all primary, large (> grade, deep, extremity STS patients. A binary grading system was used as described previously (low versus high) [7]. Patients presenting with local recurrence after prior treatment were not included in this series. Patients with Ewing sarcoma or rhabdomyosarcoma were excluded from the cohort.

Downloaded from http://annonc.oxfordjournals.org/ at Kainan University on February 12, 2015

Background: The purpose of this study was to retrospectively analyze the relationship between neo-adjuvant chemotherapy (NAC) and outcome in patients with high-grade extremity sarcomas. Patients and methods: Inclusion criteria were high-grade, deep, >5 cm extremity soft tissue sarcomas. Patients diagnosed between 1990 and 2001 were treated with surgery only (n = 282) or NAC containing doxorubicin/ifosfamide/mesna (AIM) (n = 74). The stratified Cox proportional hazards model was used to test the effect of NAC on disease-specific survival and recurrence while adjusting for known prognostic factors. Results: NAC was associated with improved disease-specific survival for this cohort of patients (P = 0.02). This overall improvement appears to be driven by the benefit of NAC on disease-specific survival for patient with tumors >10 cm. The 3-year disease-specific survival for tumors >10 cm was 0.62 (95% CI: 0.53 – 0.71) for patients not receiving NAC and 0.83 (95% CI: 0.72 – 0.95) for patients receiving NAC. Conclusion: NAC with AIM was associated with a significant improvement in disease-specific survival in patients with high-grade extremity soft tissue sarcomas >10 cm. These data emphasize the need for further prospective clinical studies of neo-adjuvant or adjuvant chemotherapy for patients with large high-grade extremity sarcomas. Key words: chemotherapy, doxorubicin, ifosfamide, neo-adjuvant, sarcoma

1668 STS. The primary end points for the analysis were disease-specific survival (defined as time from diagnosis to death due to disease or last follow-up) and recurrence-free survival (defined as time from diagnosis to first recurrence, either local or distant). Secondary end points included time to distant recurrence (defined as time from diagnosis to development of metastatic disease), and time to local recurrence (defined as time from surgery to development of local recurrence). Three-year estimates of disease-specific survival and the recurrence end points are presented since the median follow-up time for survivors was  37 months (Table 1).

The purpose of the study was to determine the association of NAC with survival and recurrence in a group of patients with high-risk extremity

Approval for this study was obtained from the Institutional Review Boards of MSKCC and DFCI.

A stratified Cox proportional hazards model was used to compare the two treatment groups and to account for differences in known prognostic factors. These prognostic factors included size of the primary tumor (divided into two groups: 5–10 cm and >10 cm), histology (categorized into five groups: leiomyosarcoma, liposarcoma, malignant peripheral nerve sheath tumor (MPNST), synovial, and malignant fibrous histiocytoma (MFH) plus other histologies) and age (divided into two _ 60 and >60) [8]. Although grade and depth are also important groups: < prognostic factors [9], they were not included since all patients had tumors that were high grade and deep depth. It was not possible to stratify by institution since the number of people treated at DFCI was small. All analyses were done using SAS, version 8.

Table 1. Patient/disease characteristics in the surgery only and neo-adjuvant chemotherapy and surgery groups Characteristic

n

Surgery only

AIM chemotherapy + surgery

Total

MSKCC

DFCI

Total

MSKCC

DFCI

282

261

21

74

44

30

Age, years Median (range)

62 (15– 94)

62 (15–94)

54 (29–87)

50 (17–73)

48 (18–73)

55 (17–72)

< _ 60

138 (48.9%)

127 (48.7%)

11 (52.4%)

55 (74.3%)

36 (81.8%)

19 (63.3%)

>60

144 (51.1%)

134 (51.3%)

10 (47.6%)

19 (25.7%)

8 (18.2%)

11 (36.7%)

Lower

228 (80.9%)

212 (81.2%)

16 (76.2%)

58 (78.4%)

33 (75.0%)

25 (83.3%)

Upper

54 (19.2%)

49 (18.8%)

5 (23.8%)

16 (21.6%)

11 (25.0%)

5 (16.7%)

Leio

28 (10.0%)

26 (10.0%)

2 (9.5%)

6 (8.1%)

2 (4.6%)

4 (13.3%)

Lipo

49 (17.4%)

44 (16.9%)

5 (23.8%)

7 (9.5%)

2 (4.6%)

5 (16.7%)

Extremity site

Histology

MFH + other

163 (57.8%)

154 (59.0%)

9 (42.9%)

38 (51.4%)

26 (59.1%)

12 (40.0%)

MPNST

11 (3.9%)

9 (3.5%)

2 (9.5%)

5 (6.8%)

1 (2.3%)

4 (13.3%)

Synovial

31 (11.0%)

28 (10.7%)

3 (14.3%)

18 (24.3%)

13 (29.6%)

5 (16.7%)

Size of primary, cma Median (range)

10.2 (5–37)

9.5 (5.5–38)

12 (6–30)

5–10

141 (50.2%)

10 (5– 38)

129 (49.6%)

12 (57.1%)

23 (31.1%)

12.5 (6– 25) 14 (31.8%)

12 (7.9–30) 9 (30.0%)

>10

140 (49.8%)

131 (50.4%)

9 (42.9%)

51 (68.9%)

30 (68.2%)

21 (70.0%)

Micro marginsb Positive

51 (18.2%)

48 (18.5%)

3 (14.3%)

9 (12.2%)

5 (11.4%)

4 (13.3%)

Negative

229 (81.8%)

211 (81.5%)

18 (85.7%)

65 (87.8%)

39 (88.6%)

26 (86.7%)

25 (7–78)

38 (8–148)

38 (8– 120)

37 (11–148)

Median follow-up for survivors (range) a

37 (0.3 –152)

39 (0.3– 152)

One missing size for MSKCC patient. Two missing margins for MSKCC patients. AI, doxorubicin + ifosfamide + mesna; DFCI, Dana Faber Cancer Institute; MFH, malignant fibrous histiocytoma; MPNST, malignant peripheral nerve sheath Tumor; MSKCC, Memorial Sloan Kettering Cancer Center. b

Downloaded from http://annonc.oxfordjournals.org/ at Kainan University on February 12, 2015

Patients treated with NAC consisting of doxorubicin/ifosfamide/mesna (AIM) followed by surgery for their primary disease were identified. Chemotherapy was not administered in the setting of a protocol. Patients were treated by a heterogeneous group of physicians at two institutions who individually made treatment decisions. Doxorubicin was typically given at 75 mg/m2/cycle as divided doses i.v. bolus over 3 days and ifosfamide was given as a dose of 6–9 g/m2/cycle, in divided doses over 3 days, usually over 3 h with mesna as a urothelial protectant. Among those who received chemotherapy, the mean number of cycles of pre-operative chemotherapy given was 3 ± 0.4 (standard deviation). These patients had been diagnosed with STS of the extremity between March 1985 and July 2001. All patients with large, high-grade, deep, extremity STS who received surgery during this same time period, but did not receive any chemotherapy (neo-adjuvant or adjuvant) for their primary disease were also identified. Since only two patients were treated with NAC during the 1980s (both from DFCI, one diagnosed in 1985, the other in 1989), we restricted our cohort to all patients treated with NAC and surgery or surgery alone from 1990 to 2001. Seventy-four patients were treated with NAC and surgery (30 from DFCI and 44 from MSKCC), and 282 patients were treated with surgery alone (21 from DFCI and 261 from MSKCC). Patient records were reviewed to confirm data collected in the databases.

1669 Table 2. One- and 3-year estimates of disease-specific survival, recurrence-free survival, and freedom from distant recurrence End point

Total

% Censored

1-year (95% CI)

3-year (95% CI)

Disease-specific survival

356

71

90% (87–93%)

73% (68–78%)

Recurrence-free survival

356

52

72% (67–77%)

52% (46–57%)

Freedom from distant recurrence

356

56

76% (72–81%)

57% (51–62%)

Results There were a total of 356 patients in the study (305 from MSKCC and 51 from DFCI). The clinicopathological characteristics of the patients stratified by treatment and by institution are shown in Table 1. Overall, younger patients were treated with NAC and surgery (median 50 years, range 17 –73) as opposed to surgery-only patients (median 62 years, range 15 –94). Most tumors involved the lower extremity (80.9% in the surgery only group and 78.4% in the NAC and surgery group). Median tumor size was larger in the NAC and surgery group (12 cm, range 6–30) than in the surgery only group (10 cm, range 5–38). Patients with synovial sarcoma were more commonly treated with NAC and surgery rather than with surgery alone. Median follow-up for surviving patients in this series was 37 months in the surgery only group and 38 months in the NAC group. Overall disease-specific survival was 73% at 3 years, with a 95% confidence interval (CI) of 68–78% (Table 2; Figure 1). The unadjusted hazard ratio for the effect of NAC on diseasespecific survival was 0.75 (95% CI: 0.45 –1.2). Using the Cox proportional hazards model stratified by histology, size of primary tumor and age, NAC was associated with a significant improvement in disease-specific survival, P = 0.02 (Table 3). The hazard ratio (HR) for the effect of NAC on disease-specific survival was 0.52 (95% CI: 0.30–0.92). In order to investigate whether the effect of NAC varied among different prognostic groups, the effect of NAC was examined within strata. However since the number of patients was small in many strata when stratifying by the combination of histology, size and age,

Downloaded from http://annonc.oxfordjournals.org/ at Kainan University on February 12, 2015

Figure 1. Disease-specific survival for all patients.

the effect of NAC was examined by stratifying by each of the three prognostic variables individually. The effect of NAC on disease-specific survival was not significant when stratifying by histology or age (Table 3). However, the effect of NAC did vary significantly when stratification was done by tumor size (Table 3). The benefit of NAC was seen in patients with tumors >10 cm, but not in patients with tumors 5–10 cm. The diseasespecific mortality HR for NAC in patients with tumors 5–10 cm was 1.4 (95% CI: 0.57 –3.3). Although it appears that NAC has an adverse effect on disease-specific survival in patients with tumors 5–10 cm, it is important to note that the number of patients in the NAC-treated group was small (n = 23) and the 95% CI is very wide. The disease-specific mortality HR for NAC in patients with tumors >10 cm was 0.45 (95% CI: 0.25–0.83). The 3-year disease-specific survival for tumors >10 cm was 62% (95% CI: 53 –71%) for patients not receiving NAC and 83% (95% CI: 72–95%) for patients receiving NAC. Survival curves demonstrating the effect of NAC on larger tumors are shown in Figure 2 (curves 2 and 4). These results indicate that the overall benefit of NAC on disease-specific survival may be driven by the effect of NAC on those patients with tumors >10 cm. The 3-year estimate of recurrence-free survival was 52% (95% CI, 46 –57%) (Table 2). The unadjusted hazard ratio for the effect of NAC on recurrence-free survival was 0.84 (95% CI: 0.58 –1.2). Using the Cox proportional hazards model stratified by histology, size of primary tumor and age, NAC showed a benefit on recurrence-free survival (HR: 0.76; 95% CI: 0.51 –1.1); however, this result was not significant, P = 0.19 (Table 3). Again, the effect of NAC was examined within strata. The results of the association of NAC on recurrence-free survival in the stratified analyses in each case did not reach statistical significance at the 0.05 level (Table 3). However, there was a trend towards benefit of NAC in those patients with tumors >10 cm (Table 3), similar to results found for disease-specific survival. The disease-specific mortality HR for NAC in patients with tumors >10 cm was 0.62 (95% CI: 0.39– 0.98); the disease-specific mortality HR for NAC in patients with tumors 5–10 cm was 1.2 (95% CI: 0.64– 2.3). Overall probability of local recurrence-free survival was 83% at 3 years (95% CI: 0.78 –0.88). Since the number of local recurrences is very small in the NAC group (nine local recurrences), further analysis was not attempted. The 3-year estimate for freedom from distant recurrence was 57% (95% CI: 51–62%) (Table 2). The results for freedom from distant recurrence were similar to the results for recurrence-free survival, as would be expected given that the majority of recurrences occurred distantly (Table 3). Although an overall

1670 Table 3. Stratified analyses to test for effect of neo-adjuvant chemotherapy (NAC) End point

Models to test for effect of NAC

HR for NAC and surgery versus surgery alone (95% CI)

P

Disease-specific survival

Stratified by size, histology and age

0.52 (0.30–0.92)

0.02

Stratified by histology alone

0.69 (0.41–1.1)

0.15

Stratified by age alone

0.85 (0.51–1.4)

0.52

Stratified by size alone

0.60 (0.36–0.99)

0.05

5 –10 cm tumors

1.4 (0.57–3.3)

>10 cm tumors Recurrence-free survival

0.45 (0.25–0.83)

Stratified by size, histology and age

0.76 (0.51–1.1)

0.19

Stratified by histology alone

0.82 (0.56–1.2)

0.30

Stratified by age alone

0.88 (0.60–1.3)

0.51

Stratified by size alone

0.76 (0.52–1.1)

0.15

1.2 (0.64–2.3)

>10 cm tumors Freedom from distant recurrence

0.62 (0.39–0.98)

Stratified by size, histology and age

0.79 (0.52–1.2)

0.28

Stratified by histology alone

0.89 (0.61–1.3)

0.57

Stratified by age alone

0.97 (0.66–1.4)

0.87

Stratified by size alone

0.83 (0.56–1.2)

0.35

5 –10 cm tumors >10 cm tumors

1.4 (0.74–2.7) 0.67 (0.41–1.1)

HR, hazard ratio.

Figure 2. Disease-specific survival stratified by tumor size and chemotherapy treatment status.

benefit in freedom from distant recurrence was not seen, a benefit of NAC was seen in those patients with tumors >10 cm (HR: 0.67; 95% CI: 0.39–0.98).

Discussion The value of chemotherapy in the management of adult primary STS has been controversial. Doxorubicin has been the most widely utilized drug in the treatment of STS [1, 10, 11]. A meta-analysis [12] found that adjuvant doxorubicin-based

chemotherapy for STS was associated with a significant (10%) improvement in recurrence-free survival at 10 years, but was not associated with an improvement in overall survival. Subgroup analysis of extremity-only STS found a 7% improvement in overall survival at 10 years, a statistically significant result. As noted previously [13], the meta-analysis did not analyze pathology samples centrally, and it is clear that at least a small percentage of these tumors were not sarcomas. In addition, the extremity sarcoma result represents an unplanned subset analysis and as a result should be hypothesis-generating rather than conclusive. More recently there has been interest in the combination of doxorubicin and ifosfamide in the management of STS. The combination of doxorubicin, ifosfamide and mesna (AIM) has been reported to be more effective than doxorubicin alone in the treatment of advanced STS [14]. When ifosfamide has been used in the neo-adjuvant setting for STS, response rates ranging from 29 to 62.5% [6, 15] have been reported. Recently published prospective randomized trials of adjuvant and/or neo-adjuvant chemotherapy with anthracycline/ ifosfamide combinations for STS have yielded inconsistent results [15–17], and these studies have been criticized for their small sample size and/or heterogeneity of tumor types and sizes [2, 18, 19]. Two trials showed no survival advantage in those patients treated with doxorubicin/ifosfamide-based adjuvant therapy [15, 16]. The most encouraging trial to date has been that of Frustaci et al. [17]. This trial demonstrated a significant survival benefit at 4 years in those patients treated with adjuvant epirubicin, ifosfamide and mesna. The study

Downloaded from http://annonc.oxfordjournals.org/ at Kainan University on February 12, 2015

5 –10 cm tumors

1671 especially important given that patients with synovial sarcoma, tumors >10 cm and younger patients were more commonly treated with NAC in the present series indicating potential bias in selection of patients for chemotherapy. However, there could be imbalances in variables that have not been accounted for in the present model. It must be noted that performance status was not compared in the two groups in this series. It is possible that performance status was higher in those selected for NAC and could account for some of the observed benefit of NAC. While generally well tolerated, AIM chemotherapy may be associated with severe neutropenia, thrombocytopenia and/or anemia in some cases, and in some cases these side-effects of chemotherapy may be fatal [15–17]. The decision to employ NAC containing AIM should be made with these potential toxic effects in mind. In conclusion, this study suggests an association between improved disease-specific survival (a 21% survival benefit at 3 years) and the use of NAC in patients with high-grade extremity sarcomas >10 cm. However, these data represent a retrospective analysis and can by no means replace a prospective randomized study of chemotherapy. We suggest that, in order to obtain clearer data, consideration should be given to nationwide adjuvant or neo-adjuvant chemotherapy studies restricted to a single well-defined histological subtype, such as leiomyosarcoma or synovial sarcoma. A design to consider that may be more palatable than a study in which there is a no-treatment arm would be a trial designed to identify early biochemical and molecular markers of response to therapy and treatment benefit for individual patients. For the group of patients unresponsive to AIM chemotherapy, there are woefully inadequate options for treatment, underscoring the need to screen and identify new therapeutic targets to enhance response rates and improve survival in patients with high-grade extremity sarcoma. With the development of such agents we hope to find more significant differences in improvement in response rate and survival in the neo-adjuvant setting, which may render moot the argument of whether to give chemotherapy or not.

Conflict of interest R.G.M. has received research funding from PharmaMar, Pfizer and Novartis. G.D.D. has received research funding from PharmaMar, Pharmacia and Novartis; acted as a consultant for Johnson and Johnson, Pharmacia, Novartis and PharmaMar; and has received honoraria from Johnson and Johnson, Pharmacia, Novartis and PharmaMar.

Acknowledgements R.G.M., E.R., M.F.B. and S.S. receive funding for part of their work from NIH Program Project Grant P01CA47179. S.R.G. is recipient of the Kristen Ann Carr Sarcoma Fellowship. The funding source had no role in the design, collection, analysis or interpretation of data. Portions of this manuscript were presented at the American Society of Clinical Oncology meeting in Spring 2003 in Chicago, IL.

Downloaded from http://annonc.oxfordjournals.org/ at Kainan University on February 12, 2015

had been closed for improvement in disease-free survival and overall survival, but over time has shown loss of significance of the benefit of chemotherapy in terms of overall survival [20]. In addition, the control and treatment groups in this study were not balanced with regard to histologic subtype. It is noteworthy that the incidence of distant failures was equal in the chemotherapy and non-chemotherapy groups at 4 years, consistent with the overall survival data from the recent study update [17]. There have been two previously published retrospective series of patients receiving NAC containing doxorubicin and ifosfamide [5, 21]. Both series reported an improvement in disease-specific survival among those treated with NAC compared to historical controls. In this study, we retrospectively analyzed the association of NAC with disease-specific survival in patients with high-grade extremity sarcomas from prospectively acquired databases of patients from two separate institutions. NAC was associated with an overall improvement in disease-specific survival for the complete cohort of patients and this improvement appears to be driven by the benefit of NAC in patients with extremity sarcomas >10 cm. In this high-risk group, there was a 21% improvement in disease-specific survival at 3 years. Conversely, no association was seen between improved disease-specific survival in patients with extremity sarcomas between 5 and 10 cm. There was also a trend towards an improvement in recurrence-free survival in patients with tumors >10 cm treated with NAC. The effect of NAC in these patients has been previously analyzed and presented in abstract form using a case-matched technique [22]. In the previous analysis, we found a significant improvement in 2-year disease-specific survival in those treated with NAC but no improvement in long-term diseasespecific survival. Limitations in the design of the previous analysis (a variable matching ratio for each strata and the lack of randomness associated with small group sizes) prompted us to finally perform the stratified analysis presented here. This difference in analytic techniques accounts for the similar but slightly different results and conclusions in the present report. Other changes that should be noted in this analysis include: tightening the study period to 1990 –2001 from 1978 –2001, and removing margin status from the analysis. Most disease-specific mortality in patients with high-risk extremity sarcomas is related to the development of metastatic disease. Therefore, the finding that NAC was not associated with an improvement in distant recurrence-free survival must be noted as it is seemingly inconsistent with the noted improvement in overall survival. It remains possible that NAC does not delay the appearance of metastatic disease but does slow the progression of metastatic disease thereby improving disease-specific survival. It should be underscored that this study design has distinct limitations. This is a retrospective cohort study, not a randomized trial. Chemotherapy in this study was not given in a protocol fashion, but rather according to individual physician judgment. We adjusted for several known prognostic variables for STS (age, tumor size and tumor histology). This was

1672

References

Downloaded from http://annonc.oxfordjournals.org/ at Kainan University on February 12, 2015

1. Brennan MF, Alektiar KM, Maki RG. Soft Tissue Sarcoma. In DeVita VT, Hellman S, Rosenberg SA (eds): Cancer: Principles and Practice of Oncology, 6th edition. Philadelphia, PA: Lippincott, Williams, and Wilkins 2001; 1841–1890. 2. Singer S, Demetri GD, Baldini EH, Fletcher CD. Management of softtissue sarcomas: an overview and update. Lancet Oncol 2000; 1: 75–85. 3. Crist W, Gehan EA, Ragab AH et al. The Third Intergroup Rhabdomyosarcoma Study. J Clin Oncol 1995; 13: 610–630. 4. Raney RB, Asmar L, Newton WA Jr et al. Ewing’s sarcoma of soft tissues in childhood: a report from the Intergroup Rhabdomyosarcoma Study, 1972 to 1991. J Clin Oncol 1997; 15: 574–582. 5. Eilber FC, Rosen G, Eckardt J et al. Treatment-induced pathologic necrosis: a predictor of local recurrence and survival in patients receiving neoadjuvant therapy for high-grade extremity soft tissue sarcomas. J Clin Oncol 2001; 19: 3203–3209. 6. Meric F, Hess KR, Varma DG et al. Radiographic response to neoadjuvant chemotherapy is a predictor of local control and survival in soft tissue sarcomas. Cancer 2002; 95: 1120–1126. 7. Hajdu S. Pathology of Soft Tissue Tumors. Philadelphia, PA: Lea and Febiger 1979. 8. Pisters PW, Harrison LB, Leung DH et al. Long-term results of a prospective randomized trial of adjuvant brachytherapy in soft tissue sarcoma. J Clin Oncol 1996; 14: 859–868. 9. Grobmyer SR, Brennan MF. Predictive variables detailing the recurrence rate of soft tissue sarcomas. Curr Opin Oncol 2003; 15: 319–326. 10. Greenall MJ, Magill GB, DeCosse JJ, Brennan MF. Chemotherapy for soft tissue sarcoma. Surg Gynecol Obstet 1986; 162: 193–198. 11. Demetri GD, Elias AD. Results of single-agent and combination chemotherapy for advanced soft tissue sarcomas. Implications for decision making in the clinic. Hematol Oncol Clin North Am 1995; 9: 765 –785.

12. Sarcoma Meta-analysis Collaboration. Adjuvant chemotherapy for localised resectable soft-tissue sarcoma of adults: meta-analysis of individual data. Lancet 1997; 350: 1647–1654. 13. Verweij J, Seynaeve C. The reason for confining the use of adjuvant chemotherapy in soft tissue sarcoma to the investigational setting. Semin Radiat Oncol 1999; 9: 352 –359. 14. Edmonson JH, Ryan LM, Blum RH et al. Randomized comparison of doxorubicin alone versus ifosfamide plus doxorubicin or mitomycin, doxorubicin, and cisplatin against advanced soft tissue sarcomas. J Clin Oncol 1993; 11: 1269–1275. 15. Gortzak E, Azzarelli A, Buesa J et al. A randomised phase II study on neo-adjuvant chemotherapy for ‘high-risk’ adult soft-tissue sarcoma. Eur J Cancer 2001; 37: 1096–1103. 16. Brodowicz T, Schwameis E, Widder J et al. Intensified adjuvant IFADIC chemotherapy for adult soft tissue sarcoma: a prospective randomized feasibility trial. Sarcoma 2000; 4: 151–160. 17. Frustaci S, Gherlinzoni F, De Paoli A et al. Adjuvant chemotherapy for adult soft tissue sarcomas of the extremities and girdles: results of the Italian randomized cooperative trial. J Clin Oncol 2001; 19: 1238–1247. 18. Judson I. Adjuvant chemotherapy of soft-tissue sarcoma—current status. Sarcoma 2000; 4: 149–150. 19. Bramwell VH. Adjuvant chemotherapy for adult soft tissue sarcoma: is there a standard of care. J Clin Oncol 2001; 19: 1235–1237. 20. Frustaci S, De Paoli A, Bidoli E et al. Ifosfamide in the adjuvant therapy of soft tissue sarcomas. Oncology 2003; 65: 80–84. 21. DeLaney TF, Spiro IJ, Suit HD et al. Neoadjuvant chemotherapy and radiotherapy for large extremity soft-tissue sarcomas. Int J Radiat Oncol Biol Phys 2003; 56: 1117–1127. 22. Grobmyer SR, Maki R, Demetri GD et al. Neo-adjuvant chemotherapy for primary high grade extremity soft tissue sarcomas: A matched case control study. Proc Am Soc Clin Oncol 2003; 22: 815.