Fax: (011) 390498211781
Localized unresectable non-rhabdo soft tissue sarcomas of the extremities in pediatric age†
Results from the Italian studies
Article first published online: 13 SEP 2005
Copyright © 2005 American Cancer Society
Volume 104, Issue 9, pages 2006–2012, 1 November 2005
How to Cite
Cecchetto, G., Alaggio, R., Dall'Igna, P., Bisogno, G., Ferrari, A., Gigante, C., Casanova, M., Sotti, G., Zanetti, I. and Carli, M. (2005), Localized unresectable non-rhabdo soft tissue sarcomas of the extremities in pediatric age. Cancer, 104: 2006–2012. doi: 10.1002/cncr.21412
Presented at the IPSO-APSA Joint Meeting, Ponte Vedra, Florida (USA), May 26, 2004.
- Issue published online: 17 OCT 2005
- Article first published online: 13 SEP 2005
- Manuscript Accepted: 3 JUN 2005
- Manuscript Revised: 19 MAY 2005
- Manuscript Received: 22 FEB 2005
- Fondazione “Città della Speranza,” Italy
- nonrhabdo soft tissue sarcoma;
- childhood sarcoma;
- tumor grade;
- extremity soft tissue sarcomas
Treatment of initially unresectable nonrhabdo soft tissue sarcomas (NRSTS) in pediatric age is debated, due to their different chemosensitivity. The authors objective was to evaluate clinical features and treatment results observed in a series of Italian patients over a 24-year period.
Fifty-two patients age 8–18 years (median 8 years) were observed (1979–2002). Primary sites were on the lower limbs in 41 and on the upper limbs in 11 cases. Clinical TNM and surgical Intergroup Rhabdomyosarcoma Staging systems were adopted. Therapeutic guidelines recommended an initial biopsy plus neoadjuvant chemotherapy. Aggressive delayed surgery of residual disease, including compartmental resections, was recommended. Radiotherapy was suggested only for patients age > 3 years when surgery was incomplete or not feasible.
The evaluation was performed dividing the patients into two categories according to their chemosensitivity. Chemosensitive (CTs) sarcomas, 21: synovial sarcoma, 11; extraosseous Ewing sarcomas, 5; primitive peripheral neuroectodermic tumors, 5. Nonchemosensitive (CTns) sarcomas, 31: fibrosarcoma, 11; malignant peripheral nerve sheet tumors, 10; liposarcoma, 2; hemangiopericitoma adult type, 2; epithelioid sarcoma, 2; and alveolar soft part sarcoma, leiomyosarcoma, clear cell sarcoma, and sarcoma NOS, each 1. Nineteen of 21 patients with CTs-NRSTS were alive without disease: the 5-year overall survival (OS) and progression-free survival (PFS) were 94.4% and 79.3%, respectively; 23 of 31 patients with CTns-NRSTS were alive without disease: 5-year OS and PFS were 75.3% and 68.3%, respectively. Response to neoadjuvant chemotherapy was complete or partial in 10 of 20 evaluable CTs and in 8 of 26 evaluable CTns tumors. The achievement of complete delayed resection was particularly important for CTns-NRSTS. Tumor size < 5 cm, distal site, and tumor grading for CTns sarcomas were often linked to a favorable outcome; no conclusive results were detected concerning age of the patients or T status of the tumor.
Multidisciplinary treatment without mutilating procedures allowed the cure of most patients with CTs and CTns-NRSTS. Relapses were cured in several cases of CTs tumors, whereas almost all patients with relapsed CTns tumors died due to the high rate of metastatic spread. Cancer 2005. © 2005 American Cancer Society.
Nonrhabdo soft tissue sarcomas (NRSTS) are tumors of mesenchymal origin, which account for approximately 4% of all pediatric malignancies. The clinical features are similar to those of rhabdomyosarcomas (RMS) and the extremities are the most common localization.
As for adult sarcomas, the mainstay of treatment for NRSTS is immediate surgical resection alone or associated with radiotherapy (RT), with sarcomas completely resected at diagnosis usually having an excellent outcome. The therapeutic approach to initially unresectable NRSTS should be standardized due to their different chemosensitivity to the drugs usually adopted for pediatric RMS. Some NRSTS are quite chemosensitive (i.e., primitive peripheral neuroectodermic tumors [PPNET] and extraosseous Ewing sarcomas [EES]), while others are poorly chemosensitive (malignant peripheral nerve sheet tumors [MPNST], malignant fibrous histiocytoma, leiomyosarcoma, clear cell sarcomas, etc.); other histologic entities (synovial sarcoma [SS]) are poorly chemosensitive in adults but seem to be quite chemosensitive in pediatric patients.1–4
Unresectable localized NRSTS are generally larger than 5 cm, invasive, and high-grade tumors: the achievement of local control with multidisciplinary therapy is crucial to reach a cure. In particular, the role of surgically conservative procedures in extremity sites is debated.5 Only a few reports in the literature describe this particular subset of localized unresectable NRSTS of the extremities in childhood, but their number is often part of a large series of patients.2, 6 The purpose of this review is to evaluate the clinical features and the treatment results of children and adolescents with initially unresectable NRSTS of the extremities enrolled in the Italian Cooperative Studies over a 24-year period.
MATERIALS AND METHODS
Between January 1979 and December 2002, 52 eligible and evaluable patients (35 males, 17 females) affected by localized, initially unresectable NRSTS of the extremities were enrolled in Italy: 41 in the Italian Cooperative RMS studies (RMS 1979, RMS 1988, and the ongoing RMS 1996) and 11 in the Istituto Nazionale Tumori of Milan study (1975–1996). They represented 35% of 149 NRSTS of the extremities observed in the same period. Informed consent for treatment and data management was obtained from patients or parents at the beginning of therapy.
The median age at diagnosis was 8 years (range 8–18 years). In 30 cases the patients' age was 10 years or less. The primary sites were on the lower limbs in 41 cases and on the upper limbs in 11; 15 were proximally and 37 distally localized. Data were collected prospectively and, for the purposes of this study, the clinical and operative records were searched to achieve more precise information concerning clinical characteristics and therapies. Treatment results and outcome were analyzed with a follow-up, which, by May 2004, ranged from 23–267 months (median 76 months).
Patients were classified according to the TNM pretreatment staging system on the basis of the features of the tumor, the involvement of regional lymph nodes, and the presence of distant metastases. Tumors were considered T1 if they were confined to the tissue of origin and T2 if they involved contiguous structures. T1 and T2 lesions were further divided into substages ‘a’ (tumor size < 5 cm) and ‘b’ (tumor size > 5 cm). T status was detected in 51 cases: 23 were considered TI and 28 T2. The size was evaluable in 51 patients: in 39 the lesion measured more than 5 cm and in 12 less than 5 cm. On the basis of the initial surgical approach, which consisted of a biopsy or resection with macroscopic residuals, all patients were grouped according to the Intergroup Rhabdomyosarcoma Staging (IRS) system as Group III. All histologic specimens were reviewed by the national panels of pathologists. The histologic grading, if applicable, was analyzed according to the Pediatric Oncology Group (POG) evaluation system.7
The main therapeutic guidelines of the Italian studies for nonmetastatic NRSTS have not changed significantly over the years. In particular, there were no consistent differences regarding the initial surgical approach compared with that for RMS. The initial removal of the tumor was recommended only if a complete resection was feasible beyond the reactive zone but within the muscolar compartment, with a surrounding cuff of normal tissue around the mass (wide resection). Mutilating (or demolitive) procedures, such as amputations and excisions with major functional impairment, were discouraged initially. An NRSTS was considered unresectable initially when a wide, nonmutilating resection was not feasible on the basis of clinical and radiological evaluation (magnetic resonance imaging [MRI], computed tomography scan) due to the size or the site of the mass. In these cases, a surgical biopsy plus neoadjuvant chemotherapy (CT) was suggested, as for RMS. The exploration of regional lymph nodes (axillary or inguinal) was always highly recommended.
Neoadjuvant CT was administered according to various regimens in the different studies: vincristine, D-actinomycin, cyclophosfamide (VAC) in the first RMS-79 study; vincristine, dactinomycin, ifosfamide, doxorubicin (VAIA) in RMS-88; and VAIA/CEVAIE (cyclophosfamide, etoposide, vincristine, D-actinomycin, ifosfamide, doxorubicin) in the ongoing RMS-96 study. Evaluation of the response of the lesion to neoadjuvant CT was performed after 9 weeks based on the decrease of the tumor dimensions at imaging and was defined as follows, complete remission (CR): no evidence of disease at imaging; partial remission (PR): reduction to > 50% of the initial tumor volume; objective remission (OR): reduction to > 25% of the initial volume stable disease (SD): reduction to < 25% of the initial tumor volume; progression of disease (PD): increase in tumor volume or appearance of new lesions.
An aggressive delayed surgery of the residual disease, including wide and compartmental resections, was recommended after neoadjuvant CT. Mutilating procedures were accepted when a poor response to CT was obtained. Delayed surgery was not indicated when complete remission at imaging was obtained with neoadjuvant CT.
External beam RT was indicated only for patients age > 3 years because of the high risks of late sequelae in patients age < 3 years. It was given when delayed surgery was not radical or as the only treatment when surgery was not possible (total doses: 46–52 Gy); the radiation target volume included the initial mass plus 2-cm margins. Since 1988, accelerated and hyperfractionated irradiation have been used instead of conventional fractionation RT and, lately, in the ongoing study RT was given before delayed surgery in some cases to improve the efficacy of neoadjuvant CT. Brachytherapy was adopted whenever possible for those infants and young children who had undergone incomplete, even doubtful, resections.
The Kaplan–Meier method was used to estimate prognosis according to OS and PFS rates. The OS was defined as the time from the date of diagnosis to death by any cause, and the PFS was defined as the time from the date of diagnosis to disease progression or relapse. The log rank test was used for univariate comparison of survival levels and for defining the potential value of prognostic factors.
Evaluation of the results was performed by dividing the patients into two categories according to their chemosensitivity (Table 1). There were 21 chemosensitive (CTs) sarcomas and 31 nonchemosensitive (CTns) sarcomas. Infantile fibrosarcomas were included in the group of poorly chemosensitive tumors together with adult-type fibrosarcomas due to poor results achieved in our experience with neoadjuvant chemotherapy for these histotypes.8
|CTs n = 21||CTns n = 31|
|Clear cell sarcoma||1|
The initial surgical approach was a biopsy in 46 cases (CTs: 19; CTns: 27); there was a resection with macroresidual disease in 6 (CTs: 2; CTns: 4). The response to neoadjuvant CT was evaluable in 20 of 21 CTs and in 26 of 31 CTns tumors. Complete remission or PR were achieved in 10 CTs (5 SS, 3 PPNETs, 2 EES) and also in 8 CTns tumors (4 MPNST, 2 infantile fibrosarcomas, 1 liposarcoma, 1 alveolar sarcoma). Objective remissions were achieved in 3 CTs and 4 CTns sarcomas; ST or PD were achieved in 7 CTs and 14 CTns sarcomas, respectively. In one of the patients with CTs, who showed ST after neoadjuvant CT, local CT with endoarterial infusion of VP16 and carboplatin was attempted without any result. The results of subsequent treatment is known in 49 out of 52 patients: 20 CTs and 29 CTns sarcomas (Table 2).
|CTs: 20 cases||CTns: 29 cases|
|Surgery||No.||RT||Events: 6||Surgery||No.||RT||Events: 9|
|Complete||11a||yes 2||—||Complete||13||yes 5a||—|
|no 9||2 LR||no 8a||1LR|
|Complete Mut.||1||yes 1||1Mts||Complete Mut.||6||no 6||1 PD, 1 Mts|
|1 LR + Mts|
|Incomplete||2||yes 2||1 Mts|
|No surgery||6||yes 4||1LR, 1LR + Mts|
|No surgery||10||yes 3||1 PD|
|no 2||—||no 7||1LR, 2 PD, 1Mts|
Delayed surgery was performed in 33 of 49 cases: among 14 CTs tumors the surgical procedure was conservative in 13 and demolitive in 1 (muscolar excision with functional impairment); among 19 CTns tumors the excision was conservative in 13 and demolitive in 6 (2 calf amputations, 2 thigh amputations, 1 disarticulation of the lower limb, 1 demolitive muscolar excision). A complete resection was obtained in 12 of 14 CTs and in all 19 CTns sarcomas in which a delayed surgery had been performed: 3 of 12 and 5 of 19 patients were also treated with RT. In two CTs sarcomas the excision was incomplete and RT was given.
Radiotherapy was the only delayed treatment in 7 of 49 patients (4 CTs and 3 CTns tumors). No local treatment was given in 9 of 49 patients with 2 CTs and 7 CTns tumors: in 7 because of age < 3 years and tumor still inoperable (3 PR, 4 SD after primary CT), in 1 because of CR after neoadjuvant CT, and in 1 for center decision.
Chemosensitive NRSTS: 19 out of 21 patients are alive without disease. The 5-year OS and PFS were 94.4% (confidence interval (CI): 83–100) and 79.3% (CI, 61–97), respectively (Table 3). There were 6 events occurring 18–52 months after diagnosis: 3 local relapses (LR), 2 pulmonary metastases (Mts), and 1 LR plus Mts. Two LR occurred in the group of nine patients who underwent complete excision without RT and one in a patient treated with RT only. Distant metastases were detected in one case after a complete mutilating resection and RT, and in one after incomplete resection plus RT. No events were observed in two patients in which no local treatment had been delivered. Four of the six relapsed patients were saved with multimodal therapy (one still with disease), whereas two died of disease.
|Overall survival||No.||Events||OS%||CI||P = 0.13|
|Progression free survival||No.||Events||PFS%||CI||P = 0.7|
Nonchemosensitive NRSTS: 23 out of 31 patients are alive without disease. The 5-year OS and PFS were 75.3% (CI, 59–91) and 68.3% (CI, 51–85), respectively (Table 3). There were 5 recurrences occurring 7–24 months after diagnosis: 2 LR (1 after complete excision without RT, 1 after CT alone) and 3 metastases: 2 after a complete mutilating excision (both without RT) and 1 after CT alone; 4 patients showed PD: 1 after a mutilating resection, 1 after RT, and 2 after CT only. Three out of 4 patients with PD and 3 of 5 who relapsed died of disease; 2 of 5 children treated initially with CT alone are alive without disease after amputation for LR. Among the 7 patients who underwent a mutilating surgery, 1 with CTs and 2 with CTns are alive without disease; 4 with CTns died.
Factors Influencing Outcome
Among the patients with CTns, those age < 10 years exhibited a slightly more favorable, but not significant 5-year PFS compared with that of patients aged > 10 years: 74.1% (CI, 53.4–93.6) versus 55.5% (CI, 23–88), P = 0.326; among patients with CTs, children aged < 10 years showed a more unfavorable PFS than the older ones: 77.8% (CI, 50–100) versus 81.8% (CI, 59–100), P = 0.63.
Tumors < 5 cm were associated with better but not significant 5-year PFS than tumors > 5 cm: 90.9% (CI, 73–100) vs. 66.5% (CI, 51–82), P = 0.079.
No different outcome between T1 and T2 sarcomas was observed either for CTs or for CTns. The regional lymph node involvement was evaluated in 46 cases (19 CTs, 27 CTns): metastatic invasion was detected only in 1 of 19 CTs and in 2 of 27 CTns sarcomas. All three patients are alive without disease.
The outcome was only partially influenced by the response to neoadjuvant CT: among 46 evaluable cases, 9 of 10 CTs and 8 of 8 CTns sarcomas with CR or PR after neoadjuvant CT were cured. Conversely, 8 of 10 CTs and 12 of 18 CTns tumors with OR or ST, respectively, were cured. The outcome of patients was evaluated by type of regimen adopted in the subsequent studies: no statistically significant differences of OS and PFS were observed.
The achievement of complete surgery was particularly important in the outcome of 29 CTns sarcomas: in fact 16 of 19 (84%) with complete excision are alive without disease, whereas only 6 of 10 with incomplete or no surgery are alive; among CTs sarcomas the completeness of excision did not improve the outcome: 10 of 12 (83%) with complete and 7 of 8 (87%) with incomplete excision are alive without disease.
Among distally localized sarcomas, 14 of 15 CTs (93%) and 18 of 22 CTns (81%) achieved a cure, while only 4 of 6 CTs (66%) and 4 of 9 CTns (44%) of those proximally localized were cured.
No substantial differences were detected in the outcome of patients concerning the type of external beam RT: in 9 of 16 patients in whom accelerated hyperfractionated RT was given, 6 did not show any event, 1 with LR, 1 with metastasis, and 1 with LR plus metastases. In 7 of 16 patients with conventional RT, 5 did not show any event, 1 had PD, and 1 had metastasis. No data are available regarding the efficacy of preoperative RT because only 1 out of 3 evaluable patients enrolled since 1996 received this treatment.
Tumor grading was analyzed according to the POG evaluation system in 5 of 11 patients with synovial sarcoma and in 30 of 31 with CTns NRSTS: all patients with synovial sarcoma (1 low- and 4 high-grade sarcomas) showed a PR to neoadjuvant CT; they are alive and well. Among patients with CTns NRSTS, 9 of 12 with low-grade sarcomas were evaluable for response to CT: 2 had PR, 7 had SD. Among 18 with high-grade sarcomas, 16 were evaluable for response to CT: 4 showed PR, 4 OR, and 8 SD. The outcome was favorable for 10 of 12 with low-grade tumors but for only 8 of 18 with high-grade tumors.
Nonrhabdo soft tissue sarcomas are a heterogeneous group of sarcomas that share some clinical and biological characteristics but vary in histological patterns and response to neoadjuvant chemotherapy. Chemosensitivity in particular has been a matter of discussion for some histotypes that present a different behavior in childhood compared to that observed in adults. Peripheral primitive neuroectodermal tumors and extraosseous Ewing sarcomas are a family of high-grade tumors generally characterized by a good response to CT. For this reason, in several reports they are not considered in the group of NRSTS, which mainly include ‘adult-type’ sarcomas with uncertain chemosensitivity. Also, the chemosensitivity of synovial sarcomas has been debated in the past and the behavior of this tumor is probably different in children than in adults: in children it is considered a ‘quite’ chemoresponsive tumor. A recent pediatric report shows that the response rate to neoadjuvant CT was quite high (60%).3
The analysis of our subset of patients with localized unresectable NRSTS of the extremities was performed, distinguishing NRSTS in two categories according to their chemosensitivity. Most of our cases were localized in the lower limbs and presented with deep, invasive primary tumors, greater than 5 cm. In general, the patients were treated with neoadjuvant CT after the initial tumor biopsy: the response was satisfying for CTs sarcomas, as expected, but it was also quite good in patients with CTns. The response of some with MPNST was remarkable, as previously reported.9
Our analysis suggests that approximately 79% and 68% of patients with, respectively, CTs and CTns localized unresectable NRSTS of the extremities can be expected to achieve and maintain complete remission with a multidisciplinary treatment, adopting mutilating procedures only in a small number of cases. The difference between the OS of the two groups (94% and 75%, respectively) is probably due to the fact that the salvage therapy was more effective for relapsed CTs tumors. Conversely, almost all relapsed patients with CTns sarcomas died: several of these relapses were metastatic.
In our experience the rate of distant relapses was slightly greater than local failures. This is in agreement with experiences reported in adults.10, 11 In a series of pediatric patients, Spunt4, 6, 20 observed that local failures, initially frequent, became less common in the last decade and this was probably related to the improvement of RT. The unexpected good response of CTs sarcomas and of some CTns sarcomas to neoadjuvant CT underline the efficacy of the regimens adopted; the response was associated with a good prognosis in most cases. The relationship between response to primary CT and outcome was confirmed by other studies.12 Moreover, the high number of distant relapses in our series, especially in CTns sarcomas, shows the necessity of finding new drugs and new CT regimens, as in those adopted in adult patients. A possible option could be intraarterial CT, widely adopted in adults, which is useful in making tumors resectable:13–16 our experience in one patient was negative, confirming the uncertain results obtained until now with this treatment in childhood.17
A crucial point of treatment of STS concerns local therapy. It is well known that an initial complete excision represents the cornerstone of treatment for patients with both CTs and CTns sarcomas: when obtained, the likelihood of survival is excellent.6, 17, 18 The survival rate decreases, however, when a complete conservative excision is not feasible at diagnosis. Some authors suggest that patients who have an initial resection with macroscopic residuals have a better prognosis than patients in whom a biopsy was performed.6 From our experience we cannot draw conclusions, because only a few patients underwent initial resection with macroscopical residual disease. An important issue regards the concept of resectability, which depends on the technical expertise of the treating surgeon: in our series, as well as in other reports,6 some tumors were considered initially unresectable; however, after an ineffective neoadjuvant CT they were resected without mutilations. It is possible that the tumor became more compact and well bounded after the treatment, making the resection feasible.
In our experience, the second operation represents a very important step in the overall treatment, and every effort should be made to choose the best timing and the adequate strategy. A wide excision, providing an adequate margin of normal tissue, is appropriate in most cases, but compartmental resections may be necessary for some lesions. Reconstructive measures (nerve repair, replacement of vessels, free-flap plasty, and even bone reconstruction) must be included in the surgical strategy. Fine surgical resections are especially necessary in children at ages up to 3 years, as external beam RT is discouraged. For these patients brachytherapy is often adopted, considering the lower risks of late effects.
In general, our data show that the association of delayed complete surgery and RT was the best choice: all the 9 of 49 patients with CTs and CTns sarcomas who were treated with a complete delayed resection plus RT maintained local control. Brachytherapy was very helpful in cases with uncertain complete excision.
The achievement of a complete surgical control was particularly important for CTns sarcomas; several CTs tumors were cured without having obtained a complete resection. External beam RT, adopted as the only local treatment, could not avoid relapses in 3 of 7 cases. Most of the failures in our series were registered in patients with CTns sarcomas who did not receive any local therapy: in particular, they were observed in young patients, either because of the inoperability of the tumor without demolitive procedures or the impossibility to deliver RT. In this subset of patients a more aggressive surgery would have improved the survival rate.
Mutilating resections were adopted in a few cases, mainly in patients with CTns sarcomas after a poor response to other treatments: however, they led to a cure in fewer than 50% of cases. Maintenance of normal function is an important consideration in childhood but a mutilating surgery is occasionally indicated for NRSTS of the extremities.5, 19 It may represent the only option, especially in young patients, in whom external beam RT may cause serious late sequelae.
Among clinical prognostic factors, the tumor site plays an important role in the surgical approach and in the feasibility of a complete resection. In our series, patients with distally localized tumors had a better outcome than patients with proximally localized tumors.
Younger age was associated with a better clinical outcome in another series.6, 20 This finding was confirmed in our data only for patients with CTns sarcomas, while among CTs patients those age >10 years had a slightly better outcome. Tumor size was the most important, but not statistically significant, prognostic factor: among CTs and CTns tumors, patients with a <5 cm tumor mass had a better PFS. The role of regional node dissection in NRSTS of the extremities does not seem to have the same relevance as reported in RMS:14–21 our results support the finding that involvement of the lymph nodes is rare.
Different opinions are expressed in the literature concerning the association between histologic grade of NRSTS and survival:22–24 our data, mainly concerning CTns, show that patients with low-grade tumors had a better prognosis than those with high-grade tumors; however, these data should be interpreted with caution, given the small number of patients in our series.
In conclusion, our data seem to demonstrate that patients with all kinds of unresectable localized NRSTS of the extremities benefited from neoadjuvant chemotherapy and that the multidisciplinary treatment without mutilating procedures allowed a cure for most patients. The achievement of local control with delayed complete surgery and RT was particularly important for CTns sarcomas. Relapses were cured only in most cases of CTs sarcomas, whereas almost all patients with relapsed CTns died due to a high rate of nonchemosensitive metastatic spread. The size and the site of the tumor mass, and the histologic grading for CTns sarcomas, were important factors in the outcome.
These findings, which reflect a small number of patients observed in an extended study period, require confirmation by larger randomized trials to achieve better knowledge of the clinical pathologic features of these tumors. Further clinical studies on new and more effective drugs, especially for CTns sarcomas, might prevent relapses after local therapy.
The authors thank Dr. Bhaskar Rao from St. Jude Children's Research Hospital, Memphis, TN, for suggestions in preparing the article and for reviewing the text.
- 9Malignant peripheral nerve sheath tumors in childhood. A combined experience of the Italian and German Co-operative study Groups. Proceedings of the XXXIII SIOP Meeting. Med Pediatr Oncol. 2001; 37: 184., , , et al.
- 12Impact of neoadjuvant therapy on local control of soft tissue sarcoma. [Abstract 2215] Proc Am Soc Clin Oncol. 2000; 19: 558a., , , et al.
- 13Neoadjuvant chemotherapy and radiotherapy for large extremity soft-tissue sarcomas. Int J Radiat Oncol Biol Phys. 2003; 6: 915–916., , .
- 14Soft tissue sarcomas. In: PinkertonR, PlowmanPN, PietersR, editors. Paediatric oncology, 3rd ed. London: Arnold, 2004: 339–370., , , , .
- 17Localized soft tissue sarcomas of extremities: the experience of the Italian studies. Proceedings of the XXXIV IPSO Meeting. Med Pediatr Oncol. 2002; 39: 220., , , et al.
- 22Prognostic factors for the outcome of chemotherapy in advanced soft tissue sarcoma. An analysis of 2,185 patients treated with antracycline containing first line regimen—A European Organization for Research and Treatment of Cancer Soft Tissue and Bone Sarcoma Group Study. J Clin Oncol. 1999; 17: 150–157., , , et al.