This Article Abstract Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Save to my personal folders Download to citation manager Rights & Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Spunt, S. L. Articles by Pappo, A. S. Search for Related Content PubMed PubMed Citation Articles by Spunt, S. L. Articles by Pappo, A. S. Social Bookmarking                            What's this?

Prognostic Factors for Children and Adolescents With Surgically Resected Nonrhabdomyosarcoma Soft Tissue Sarcoma: An Analysis of 121 Patients Treated at St Jude Children's Research Hospital

From the Departments of Hematology/Oncology, Biostatistics and Epidemiology, Surgery, Radiation Oncology, and Pathology, St. Jude Children's Research Hospital, Memphis, TN; the Medical College of Pennsylvania and Hahnemann School of Medicine, Philadelphia, PA; and the Departments of Pediatrics, Surgery, and Pathology, The University of Tennessee College of Medicine, Memphis, TN.

Address reprint requests to Sheri L. Spunt, MD, Department of Hematology/Oncology, St. Jude Children's Research Hospital, 332 N Lauderdale St, Memphis, TN 38105-2794; email sheri.spunt{at}stjude.org

	ABSTRACT

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION REFERENCES

 
PURPOSE: The rarity and heterogeneity of pediatric nonrhabdomyosarcoma soft tissue sarcoma (NRSTS) has precluded meaningful analysis of prognostic factors associated with surgically resected disease. To define a population of patients at high risk of treatment failure who might benefit from adjuvant therapies, we evaluated the relationship between various clinicopathologic factors and clinical outcome of children and adolescents with resected NRSTS over a 27-year period at our institution.

PATIENTS AND METHODS: We analyzed the records of 121 consecutive patients with NRSTS who underwent surgical resection between August 1969 and December 1996. Demographic data, tumor characteristics, treatment, and outcomes were recorded. Univariate and multivariate analyses of prognostic factors for survival, event-free survival (EFS), and local and distant recurrence were performed.

RESULTS: At a median follow-up of 9.2 years, 5-year survival and EFS rates for the entire cohort were 89% ± 3% and 77% ± 4%, respectively. In univariate models, positive surgical margins (P = .004), tumor size 5 cm (P < .001), invasiveness (P = .002), high grade (P = .028), and intra-abdominal primary tumor site (P = .055) adversely affected EFS. All of these factors except invasiveness remained prognostic of EFS and survival in multivariate models. Positive surgical margins (P = .003), intra-abdominal primary tumor site (P = .028), and the omission of radiation therapy (P = .043) predicted local recurrence, whereas tumor size 5 cm (P < .001), invasiveness (P < .001), and high grade (P = .004) predicted distant recurrence.

CONCLUSION: In this largest single-institution analysis of pediatric patients with surgically resected NRSTS, we identified clinicopathologic features predictive of poor outcome. These variables should be prospectively evaluated as risk-adapted therapies are developed.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION REFERENCES

 
NONRHABDOMYOSARCOMA soft tissue sarcoma (NRSTS) accounts for approximately 3% of all childhood malignancies.¹ These rare and heterogenous tumors include a variety of soft tissue neoplasms, the histologic, clinical, and prognostic characteristics of which differ greatly from those of rhabdomyosarcoma. The mainstay of therapy for NRSTS is surgical resection, with or without radiotherapy. With this treatment, more than 70% of patients with surgically resectable tumors are expected to achieve long-term survival.²

Despite the favorable outcome in children and adolescents with resected NRSTS, a significant proportion of patients experience disease recurrence or progression. The characteristics associated with an increased risk of treatment failure after surgical resection are largely unknown. In addition, it is uncertain whether prognostic factors in pediatric patients with NRSTS are similar to those identified in studies of adults with this disease. Because both chemotherapy and radiotherapy carry risks of short- and long-term side effects, their use would best be restricted to those at high risk of disease recurrence after surgical resection. Therefore, we retrospectively evaluated clinicopathologic features and their correlation with clinical outcome for children and adolescents with surgically resected NRSTS who were treated at our institution over a 27-year period.

	PATIENTS AND METHODS

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION REFERENCES

 
Between August 1969 and December 1996, 212 consecutive patients younger than 21 years with a diagnosis of NRSTS were admitted to St. Jude Children's Research Hospital. Of the 129 patients who underwent surgical resection at the time of initial diagnosis, 121 were eligible for analysis. Eight patients were excluded from the analysis because of prior malignancy (n = 4) or the absence of adequate data to confirm the diagnosis and stage of disease (n = 4). Patients with desmoid tumor were not included in the study.

Demographic data collected for this group of 121 patients included sex, race, age at the time of diagnosis, and duration of follow-up. The pathologic diagnosis was confirmed in all cases by one of the authors (J.J.J.). Histologic grade was determined using the Pediatric Oncology Group grading system,³ which takes into account patient age, histology, cellularity, necrosis, mitotic activity, and nuclear features. Tumor characteristics identified included site of primary tumor (head and neck, extremity, trunk wall, intra-abdominal), histologic pattern, size (< 5 cm or 5 cm), invasiveness into bone or neurovascular structures (yes or no), histopathologic grade (low [1,2] or high [3]), and regional lymph node status (positive or negative). All tumors arising in the abdomen, retroperitoneum, and pelvis were classified as intra-abdominal tumors.

Disease was staged according to the surgical and pathologic staging system developed by the Intergroup Rhabdomyosarcoma Study Group.⁴ Briefly, patients in clinical group I had undergone complete resection of their tumor with surgical margins and lymph nodes negative for tumor; those in clinical group II had evidence of microscopic residual disease after surgery or had undergone complete resection of lymph nodes positive for tumor. For patients with negative surgical margins (clinical group I), the minimum depth of normal tissue surrounding the tumor in the surgical specimen was further classified ( 1 cm or > 1 cm). Subsequent surgical procedures, as well as chemotherapy and radiotherapy administered after the initial surgical resection, were also recorded. Because the accuracy of staging may have improved after computed tomography scanning became available in 1977, treatment was divided into two eras: 1969 through 1977 and 1978 through 1996. Sites of disease recurrence, type of therapy after recurrence, response to therapy after recurrence, and ultimate outcome were also recorded. Local recurrence was defined as recurrence at local and/or regional sites.

Statistical Methods
Associations between categorical variables were examined using Fisher's exact test. Survival was defined as the time interval from diagnosis to death from any cause or to last follow-up evaluation. Event-free survival (EFS) was defined as the time interval from diagnosis to recurrence/progressive disease, second malignancy, or death, or to last follow-up evaluation. Postrelapse survival was defined as the time interval from disease recurrence to death or last follow-up evaluation. Survival, postrelapse survival, and EFS were estimated using the Kaplan-Meier method; SEs were calculated using the method of Peto et al.⁵ The effects of each prognostic variable on survival and EFS were examined in univariate Cox proportional hazards models. All variables significant at P .20 were entered into a multivariate model. To arrive at a parsimonious multivariate model, we removed variables if the likelihood ratio ² test indicated that they did not significantly contribute to the model. Relative risks with 95% confidence intervals were calculated. An exact log-rank test was used to examine the differences in postrelapse survival distributions by site of recurrence (local v distant/both).

Local control was defined as the time interval from diagnosis to local and/or regional recurrence of disease. Competing risks of local treatment failure included distant recurrence, second malignancy, or death before local recurrence. Distant control was defined as the time interval from diagnosis to distant recurrence. Competing risks of distant recurrence included local recurrence, second malignancy, or death before distant recurrence. Patients with simultaneous local and distant recurrence were considered to have local recurrence for the analysis of local control and distant recurrence for the analysis of distant control. The cumulative incidence of local and distant recurrence was estimated using the methods of Kalbfleisch and Prentice.⁶ Statistical tests developed by Gray⁷ were used to examine differences in the cumulative incidence of local and distant recurrence by tumor, treatment, and patient characteristics. Because of the small number of distant recurrences, risk factors for distant recurrence were examined only univariately.

	RESULTS

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION REFERENCES

 
At a median follow-up duration of 9.2 years (range, 1.6 to 28 years), 84% of the patients (n = 102) were alive. Among the survivors, more than 90% had been seen or contacted within the previous year.

The clinical and tumor characteristics of the patient population are listed in Table 1. The median age at the time of diagnosis was 11.2 years (range, birth to 20.9 years), and 10 patients were younger than 1 year at the time of diagnosis. There were 58 males (48%), and 97 patients (80%) were white. The most frequent primary tumor sites were the extremities and trunk wall. After surgical resection, 81 patients (67%) had clinical group I disease and 40 (33%) had clinical group II disease. For 61 clinical group I patients (75%), the minimum depth of normal tissue surrounding the tumor in the surgical specimen was more than 1 cm. The most common histologic diagnoses were synovial sarcoma, malignant fibrous histiocytoma, fibrosarcoma, and malignant peripheral-nerve sheath tumor. Forty-eight patients (40%) had tumors more than 5 cm in greatest dimension, 33 (27%) had invasive tumors, and 45 (37%) had high-grade tumors. Only three patients had metastatic tumor in regional lymph nodes.

We evaluated the relationship between clinical group, tumor size, histologic grade, and invasiveness and found that tumor size was significantly associated with both histologic grade (P = .022) and invasiveness (P < .001). Only 21 of 73 (29%) tumors smaller than 5 cm in diameter were high grade, whereas 24 of 48 (50%) larger tumors were high grade. Only 10 of 73 (14%) tumors smaller than 5 cm were invasive, whereas 23 of 48 (48%) larger tumors were invasive. In addition, we also found a significant association between histologic grade and invasiveness (P = .02). We found no association between clinical group and tumor size (P = .70), grade (P = .32), or invasiveness (P = .086).

Adjuvant Therapy
Fifty-one clinical group I patients (63%) were treated with surgery alone, compared with 15 clinical group II patients (38%). The remaining patients received adjuvant chemotherapy and/or radiotherapy (Table 2). Group I patients with high-grade tumors were more likely to receive adjuvant therapy than were those with low-grade tumors (52% v 27%; P = .035). Similarly, group I patients with tumors greater than 5 cm in diameter were more likely (17 of 31; 55%) than patients with smaller tumors (13 of 50; 26%; P = .017) to receive adjuvant therapy. The use of adjuvant therapy in group I patients did not correlate with tumor invasiveness (P = .99). There were no significant differences between the group II patients who did receive adjuvant therapy and those who did not in terms of histologic grade (P = .99), tumor size (P = .75), or invasiveness (P = .99).

The number of clinical group I and II patients who received adjuvant chemotherapy was similar (31% v 28%; P = .83), but clinical group II patients were significantly more likely to receive adjuvant radiotherapy (53% v 12%; P < .001). The 10 group I patients treated with radiation therapy were given a median dose of 54.9 Gy (range, 26.5 to 60.4 Gy); the median radiotherapy dose for the 21 group II patients was 59.4 Gy (range, 40 to 75 Gy).

Nineteen of 40 clinical group II patients (48%) did not receive radiotherapy immediately after surgery for the following reasons: initial treatment administered at an outside facility and referred only at the time of recurrence (n = 4), negative surgical margins (clinical group II because of positive resected lymph nodes; n = 2), tumor deemed insensitive to radiotherapy or of very low malignant potential (n = 6), chemotherapy recommended (n = 4, one of whom refused treatment), tumor recurred during postoperative recovery after initial surgery (n = 1), family refused further treatment (n = 1), or initial pathologic specimen interpreted as a benign process (n = 1).

Survival
The estimated 5-year survival rate for the entire cohort was 89.0% ± 3.3% (Fig 1). Table 3 lists the 5-year estimates of overall survival according to all of the variables explored. In univariate models, survival estimates were significantly worse for patients with large ( 5 cm), invasive, high-grade (grade 3), or intra-abdominal tumors. In a multivariate model, tumor size, histologic grade, and site of primary tumor remained significant predictors of survival, and clinical group also significantly predicted survival. Other potential prognostic factors examined in this study, including treatment era, age at diagnosis, sex, race, tumor histology, and adjuvant treatment with chemotherapy or radiotherapy, were not significantly associated with overall survival.

View larger version (11K): [in this window] [in a new window]   Fig 1. Survival (S) and EFS of children and adolescents with surgically resected NRSTS.

EFS
Thirty-five patients experienced one or more events. First events consisted of 31 recurrences (17 local, 12 distant, two both local and distant), two second malignancies, and two deaths. The estimated 5-year EFS rate for the entire cohort was 77.1% ± 4.4% (Fig 1). For the 31 patients who experienced tumor recurrence, the median time from diagnosis to recurrence was 1.3 years (range, 2.4 months to 20.3 years). Second malignant neoplasms developed in three patients at 7.6, 9.4, and 10.4 years after diagnosis. One of these patients was alive 5.3 years after developing a malignant melanoma (no prior radiation exposure), and two have died of secondary osteosarcoma arising in the radiation field. One of the patients with secondary osteosarcoma developed the disease 9.2 years after local recurrence of malignant fibrous histiocytoma. Two patients without evidence of disease died in motor vehicle accidents.

Table 4 shows 5-year EFS estimates according to patient, tumor, and treatment factors. In univariate models, clinical group, tumor size, histologic grade, and invasiveness significantly influenced EFS. Based on a likelihood ratio ² test, the contribution of the primary tumor site variables approached significance (P = .094) and were entered into a multivariate model. In the multivariate model, clinical group, tumor size, and histologic grade were found to significantly influence EFS (P = 0.05). Other clinicopathologic features (treatment era, age at diagnosis, sex, race, tumor histology) and treatment factors (chemotherapy, radiotherapy) were examined, but none significantly affected EFS.

Treatment Failure
Seventeen patients experienced local recurrence, 12 experienced distant metastatic recurrence, and two experienced simultaneous local and distant recurrences. Competing risks of disease recurrence included second malignancy (n = 2) and death before recurrence (n = 2). For the study population, the 5-year estimate of the cumulative incidence of local failure was 12.8% ± 3.1%; that of distant failure was 11.8% ± 3.0%.

Factors Associated With Local Disease Recurrence
In univariate analyses, statistically significant differences in the cumulative incidence of local failure were observed with clinical group, site of primary tumor, and treatment with radiation (Table 5). Differences in the cumulative incidence of local recurrence by tumor size and treatment with chemotherapy approached statistical significance (P = .084 and .100, respectively). Local recurrence occurred more frequently in patients with clinical group II tumors, large tumors, or intra-abdominal disease. Conversely, the risk of local recurrence was significantly lower for patients treated with radiotherapy. In patients with positive surgical margins (clinical group II), the use of radiotherapy was associated with a significantly lower local recurrence rate (P = .001), whereas radiotherapy did not seem to influence the rate of local recurrence in patients with negative surgical margins (clinical group I; P = .37). In addition, local control rates for patients with negative surgical margins were not dependent on the extent (depth) of the surgical margin. Local recurrence occurred less frequently in patients treated with chemotherapy (5-year estimates, 9% ± 5% v 15% ± 4%). High histologic grade and invasiveness did not increase the risk of local recurrence.

Because of the presence of competing risks and the small number of local recurrences in our study population, we were unable to perform a multivariate analysis of factors associated with local recurrence. Rather, we evaluated the effects of factors observed to be significant in univariate models by adjusting for each of the other prognostic factors. There was no difference in the cumulative incidence of local recurrence among the extremity, head and neck, and trunk wall sites (P = .77), and these were combined for this analysis. Tumor size did not seem to significantly influence local control in models that were adjusted for clinical group (P = .085), site of primary tumor (P = .30), and treatment with radiotherapy (P = .075). However, clinical group, site of primary tumor (intra-abdominal v others), and treatment with radiotherapy each remained significant after the analysis was adjusted for each of the other variables (data not shown). However, these findings must be interpreted with caution. Radiotherapy was not administered uniformly, even among clinical group II patients, raising the possibility that the influence of this modality might have been either stronger or weaker in the absence of this selection bias.

Factors Associated With Distant Disease Recurrence
In univariate analyses, factors that were significantly associated with the cumulative incidence of distant disease recurrence included tumor size (P < .001), invasiveness (P < .001), and histologic grade (P = .004; Table 5). Patients with large, invasive, or high-grade tumors were at the greatest risk of distant recurrence. Because distant recurrence was uncommon, assessment of multiple covariates was not possible.

Survival After Recurrence
Of 31 patients who experienced tumor recurrence, 16 have died; the 5-year estimate of survival after recurrence was 57.9% ± 9.7%. The median follow-up duration for the 15 patients who survived after recurrence was 8.2 years (range, 1.8 to 19.7 years). Postrelapse survival distributions were significantly different by site of relapse (local v distant/both; P = .007). Five-year postrelapse survival estimates were 76.5% ± 11.7% for patients with local recurrence only and 35.7% ± 11.7% for patients with distant (or both local and distant) disease recurrence.

Of 17 patients with isolated local recurrences, five have died. One patient died of local tumor progression, three of distant metastatic disease, and one of a secondary malignancy. Although a statistical analysis was not possible, high histologic grade and invasiveness at initial diagnosis seemed to predict poor survival after recurrence for these patients. Four of six patients with high-grade tumors, three of five with invasive tumors, and all three with tumors that were both high grade and invasive have died of their recurrent disease. In contrast, nine of 10 patients with low-grade tumors and 11 of 12 patients with noninvasive tumors were alive 1.8 to 19.7 years (median, 5.8 years) after recurrence. The only patient who died after local recurrence of a low-grade tumor succumbed to a radiation-induced osteosarcoma.

Local recurrence was treated with various combinations of surgery, chemotherapy, and radiotherapy. However, six of 12 survivors who have no evidence of residual disease were treated with surgical resection alone. Neither the extent of surgical resection nor the type of adjuvant therapy, if any, seemed to influence outcome after local recurrence.

Of the 14 patients with distant or simultaneous local and distant recurrence, 11 have died. Two of these patients died from treatment complications; the remaining nine died from metastatic tumor. Again, patients with high-grade tumors seemed to fare more poorly than those with low-grade tumors. Nine of 10 patients with high-grade tumors died, whereas two of four patients with low-grade tumors died. Tumor invasiveness and size did not seem to influence survival after recurrence. Although these results suggest the possibility of poorer outcome for patients with high-grade tumors, there were too few distant recurrences to confirm a statistically significant relationship between histologic grade and survival.

The lung was the most common site of distant recurrence (nine of 14 cases). Other sites included bone (n = 2), soft tissues (n = 2), bone marrow (n = 1), liver (n = 1), abdominal cavity (n = 1), omentum (n = 1), and CSF (n = 1). All three patients who survived after distant recurrence had isolated lung metastases. Two underwent complete surgical resection of the metastases, and one underwent partial surgical resection followed by chemotherapy (vincristine, dactinomycin, cyclophosphamide, and doxorubicin) and 12-Gy whole-lung irradiation.

	DISCUSSION

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION REFERENCES

 
In this largest single-institution retrospective analysis of children with surgically resected NRSTS, we found that clinical outcome is excellent, with more than 80% surviving longer than 5 years. In addition, we identified clinically useful prognostic factors that are strongly associated with survival and local and distant disease control.

In our series of 121 patients, factors that predicted significantly poorer EFS and overall survival included large tumor size ( 5 cm), high histologic grade, intra-abdominal primary tumor, and microscopic residual disease after initial resection. Our findings expand previous observations reported by one of the authors (B.N.R.) that high histologic grade was associated with a less favorable survival regardless of extent of disease at presentation.⁸ Our findings are also in accord with two other small series of pediatric patients with NRSTS in which tumor size 5 cm and/or high histologic grade adversely influenced survival.^2,9 Furthermore, our findings are remarkably similar to those reported in two large adult trials in which high histologic grade, tumor size 5 cm, and microscopically positive margins negatively influenced survival.^10,11 These observations suggest that pretreatment characteristics can aid in risk stratification of patients with surgically resected NRSTS at the time of initial diagnosis, and that pediatric and adult NRSTS have similar clinical behavior.

In our analysis, we found that the risk factors associated with local recurrence differ from those associated with distant recurrence. Factors associated with local recurrence included postoperative microscopic residual disease (clinical group II), intra-abdominal primary tumor, and the omission of adjuvant radiotherapy. In our patients with negative surgical margins (clinical group I), the rate of local recurrence was similar for patients whose surgical margin was less than or greater than 1 cm. This observation agrees with that of Sadoski et al¹² and suggests that negative surgical margins improve local control rates even when the margin is very thin. Thus, although negative surgical margins should be the goal of resection, aggressive attempts to remove a large margin of normal tissue around the tumor should be avoided. These recommendations should be confirmed in larger, prospective trials to define better the optimal extent of surgical resection and its impact on clinical outcome.

Although radiotherapy was administered inconsistently in our series, local control was significantly improved in patients with microscopic residual disease who received this postoperative treatment modality. Unfortunately, the addition of radiotherapy did not ultimately impact on EFS or overall survival, a finding consistent with that reported by Yang et al and other investigators.^11,13 This observation suggests that micrometastatic disease present at diagnosis may ultimately lead to unfavorable outcome, regardless of the type of local therapy administered. Other tumor characteristics (size, grade, invasiveness) and the use of adjuvant chemotherapy did not influence the risk of local recurrence.

Factors associated with distant recurrence in our series included tumor size 5 cm, high histologic grade, and invasiveness. These observations agree with those of several studies of adults.^10,11 In our series, the risk of distant recurrence for patients with and without postoperative microscopic residual disease was similar, suggesting that the extent of surgical resection at initial presentation may not influence the risk of developing metastatic disease. A similar conclusion was reached by Heslin et al,¹⁴ who found that adults with positive microscopic margins after limb salvage surgery had the same risk of distant metastases as those with negative margins after amputation. The administration of adjuvant chemotherapy also did not alter the risk of distant disease recurrence. This finding is in accord with several trials in adults^15,16 and with those of a recently completed Pediatric Oncology Group study in which adjuvant chemotherapy with vincristine, dactinomycin, cyclophosphamide, and doxorubicin failed to improve outcome.¹⁷

Survival after isolated local recurrence was significantly better than that after distant or combined local and distant recurrence, and death after local recurrence was caused by the subsequent development of metastatic disease in most cases. Although the number of patients is small, death after local recurrence seemed to be associated with risk factors present at the time of initial diagnosis that are known to predispose to the development of distant recurrence (large tumor size, high histologic grade, or invasive disease). Thus, pretreatment factors may ultimately be responsible for tumor-related mortality, regardless of the success of local control measures. These findings are similar to those of Lewis et al,¹⁸ who found that tumor-related mortality after local recurrence was predicted by high histologic grade or tumor size 5 cm at the time of initial diagnosis. Survival after distant recurrence was rare in our experience.

Based on our findings, adjuvant therapy is not recommended for all pediatric patients with resected NRSTS. At particular anatomic sites, adjuvant radiotherapy may be useful in conjunction with limited surgery to avoid disfigurement and improve functional outcome. Although this treatment modality does not impact ultimate survival, it may be used to achieve local control in patients at high risk for local recurrence, such as those with postoperative microscopic residual disease or unfavorable primary tumor sites. Studies in adults suggest that the benefit of radiotherapy is limited to those with high-grade tumors.¹⁹ Similarly, we conclude that only children with high-grade tumors should be candidates for adjuvant radiotherapy. Omission of radiotherapy for patients with negative margins (clinical group I) after resection of high-grade tumors may be reasonable, because the risks of radiotherapy in a growing child may outweigh the limited local control benefit. The use of adjuvant radiotherapy for low-grade NRSTS should be limited to tumors that are actively growing and unresectable and that pose a risk of significant morbidity or mortality.

There is little evidence from our series or from other pediatric¹⁷ and adult^15,16,20 studies that adjuvant chemotherapy significantly improves survival in NRSTS. Thus, we do not recommend adjuvant chemotherapy for pediatric patients with resectable tumors. If effective adjuvant chemotherapy for NRSTS is identified in the future, this therapy might be appropriately extended to children with resectedtumors who are at high risk for distant recurrence on the basis of features such as high histologic grade or large tumor size.

In summary, we have identified clinicopathologic features associated with clinical outcome and with local and distant recurrence in children with resected NRSTS. These results require confirmation by larger, prospective multi-institutional trials. These findings could then be used to develop risk-directed adjuvant therapies for children at high risk of treatment failure.

	ACKNOWLEDGMENTS

 
Supported in part by Cancer Center grant no. CA 23099 and Cancer Center Support CORE grant no. P30 CA 21765 from the National Cancer Institute, Bethesda, MD, and by the American Lebanese Syrian Associated Charities, Memphis, TN.

We thank Pamela Hays (Tumor Registry), Flo Witte and Sharon Naron (Scientific Editing), the Solid Tumor data managers for assistance with data collection, and Sherry Stephens for secretarial support.

	REFERENCES

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION REFERENCES

 
1. Grovas A, Fremgen A, Rauck A, et al: The National Cancer Data Base report on patterns of childhood cancers in the United States. Cancer80:2321-2332, 1997[Medline]

2. Marcus KC, Grier HE, Shamberger RC, et al: Childhood soft tissue sarcoma: A 20-year experience. J Pediatr131:603-607, 1997[Medline]

3. Parham DM, Webber BL, Jenkins JJ, et al: Nonrhabdomyosarcomatous soft tissue sarcomas of childhood: Formulation of a simplified system for grading. Mod Pathol8:705-710, 1995[Medline]

4. Maurer HM, Beltangady M, Gehan EA, et al: The Intergroup Rhabdomyosarcoma Study: I. A final report. Cancer61:209-220, 1988[Medline]

5. Peto R, Pike MC, Armitage P: Design and analysis of randomized clinical trials requiring prolonged observation of each patient: II. Analysis and examples. Br J Cancer35:1-39, 1977[Medline]

6. Kalbfleisch JD, Prentice RL: The Statistical Analysis of Failure Time Data. New York, NY, Wiley, 1980

7. Gray RJ: A class of K-sample tests for comparing the cumulative incidence of a competing risk. Ann Stat16:1141-1154, 1988

8. Rao BN: Nonrhabdomyosarcoma in children: Prognostic factors influencing survival. Semin Surg Oncol9:524-531, 1993[Medline]

9. McCoy DM, Levine EA, Ferrer K, et al: Pediatric soft tissue sarcomas of nonmyogenic origin. J Surg Oncol53:149-153, 1993[Medline]

10. Pisters PW, Leung DH, Woodruff J, et al: Analysis of prognostic factors in 1,041 patients with localized soft tissue sarcomas of the extremities. J Clin Oncol14:1679-1689, 1996[Abstract/Free Full Text]

11. Coindre JM, Terrier P, Bui NB, et al: Prognostic factors in adult patients with locally controlled soft tissue sarcoma: A study of 546 patients from the French Federation of Cancer Centers Sarcoma Group. J Clin Oncol14:869-877, 1996[Abstract/Free Full Text]

12. Sadoski C, Suit HD, Rosenberg A, et al: Preoperative radiation, surgical margins, and local control of extremity sarcomas of soft tissues. J Surg Oncol52:223-230, 1993[Medline]

13. Yang JC, Chang AE, Baker AR, et al: Randomized prospective study of the benefit of adjuvant radiation therapy in the treatment of soft tissue sarcomas of the extremity. J Clin Oncol16:197-203, 1998[Abstract/Free Full Text]

14. Heslin MJ, Woodruff J, Brennan MF: Prognostic significance of a positive microscopic margin in high-risk extremity soft tissue sarcoma: Implications for management. J Clin Oncol14:473-478, 1996[Abstract/Free Full Text]

15. Antman K, Amato D, Pilepich M, et al: A randomized intergroup trial of adjuvant doxorubicin (DOX) for soft tissue sarcomas (STS): Lack of apparent difference between treatment groups. Proc Am Soc Clin Oncol6:134, 1987 (abstr)

16. Alvegard TA, Sigurdsson H, Mouridsen H, et al: Adjuvant chemotherapy with doxorubicin in high-grade soft tissue sarcoma: A randomized trial of the Scandinavian Sarcoma Group. J Clin Oncol7:1504-1513, 1989[Abstract]

17. Pratt CB, Pappo AS, Gieser P, et al: Role of adjuvant chemotherapy in the treatment of surgically resected pediatric nonrhabdomyosarcomatous soft tissue sarcomas: A Pediatric Oncology Group study. J Clin Oncol17:1219-1226, 1999[Abstract/Free Full Text]

18. Lewis JJ, Leung D, Heslin M, et al: Association of local recurrence with subsequent survival in extremity soft tissue sarcoma. J Clin Oncol15:646-652, 1997[Abstract/Free Full Text]

19. Harrison LB, Franzese F, Gaynor JJ, et al: Long-term results of a prospective randomized trial of adjuvant brachytherapy in the management of completely resected soft tissue sarcomas of the extremity and superficial trunk. Int J Radiat Oncol Biol Phys27:259-265, 1993[Medline]

20. Tierney JF, Stewart LA, Parmar MK, et al: Adjuvant chemotherapy for localised resectable soft-tissue sarcoma of adults: Meta-analysis of individual data. Lancet350:1647-1654, 1997[Medline]

Submitted April 29, 1999; accepted July 22, 1999.

CiteULike    Complore    Connotea    Del.icio.us    Digg    Facebook    Reddit    Technorati    Twitter    What's this?

This article has been cited by other articles:

				F. Chotel, A. Unnithan, C. R. Chandrasekar, R. Parot, L. Jeys, and R. J. Grimer Variability in the presentation of synovial sarcoma in children: A PLEA FOR GREATER AWARENESS J Bone Joint Surg Br, August 1, 2008; 90-B(8): 1090 - 1096. [Abstract] [Full Text] [PDF]

				S. L. Spunt, S. X. Skapek, and C. M. Coffin Pediatric Nonrhabdomyosarcoma Soft Tissue Sarcomas Oncologist, June 1, 2008; 13(6): 668 - 678. [Abstract] [Full Text] [PDF]

				S. L. Spunt and A. S. Pappo Childhood Nonrhabdomyosarcoma Soft Tissue Sarcomas Are Not Adult-Type Tumors J. Clin. Oncol., April 20, 2006; 24(12): 1958 - 1959. [Full Text] [PDF]

				A. S. Pappo, M. Devidas, J. Jenkins, B. Rao, R. Marcus, P. Thomas, M. Gebhardt, C. Pratt, and H. E. Grier Phase II Trial of Neoadjuvant Vincristine, Ifosfamide, and Doxorubicin With Granulocyte Colony-Stimulating Factor Support in Children and Adolescents With Advanced-Stage Nonrhabdomyosarcomatous Soft Tissue Sarcomas: A Pediatric Oncology Group Study J. Clin. Oncol., June 20, 2005; 23(18): 4031 - 4038. [Abstract] [Full Text] [PDF]

				A. Ferrari, M. Casanova, P. Collini, C. Meazza, R. Luksch, M. Massimino, G. Cefalo, M. Terenziani, F. Spreafico, S. Catania, et al. Adult-Type Soft Tissue Sarcomas in Pediatric-Age Patients: Experience at the Istituto Nazionale Tumori in Milan J. Clin. Oncol., June 20, 2005; 23(18): 4021 - 4030. [Abstract] [Full Text] [PDF]

				L. H. Baker Medical and Pediatric Oncology, Not Adult and Pediatric Oncology J. Clin. Oncol., June 20, 2005; 23(18): 4003 - 4005. [Full Text] [PDF]

				M. F. Okcu, M. Munsell, J. Treuner, A. Mattke, A. Pappo, A. Cain, A. Ferrari, M. Casanova, A. Ozkan, and B. Raney Synovial Sarcoma of Childhood and Adolescence: A Multicenter, Multivariate Analysis of Outcome J. Clin. Oncol., April 15, 2003; 21(8): 1602 - 1611. [Abstract] [Full Text] [PDF]

				S. L. Spunt, D. A. Hill, A. M. Motosue, C. A. Billups, A. M. Cain, B. N. Rao, C. B. Pratt, T. E. Merchant, and A. S. Pappo Clinical Features and Outcome of Initially Unresected Nonmetastatic Pediatric Nonrhabdomyosarcoma Soft Tissue Sarcoma J. Clin. Oncol., August 1, 2002; 20(15): 3225 - 3235. [Abstract] [Full Text] [PDF]

				H. E. Grier Soft Tissue Sarcoma: Apples, Oranges, and Passion Fruit J. Clin. Oncol., December 1, 1999; 17(12): 3695 - 3696. [Full Text] [PDF]

This Article Abstract Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Save to my personal folders Download to citation manager Rights & Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Spunt, S. L. Articles by Pappo, A. S. Search for Related Content PubMed PubMed Citation Articles by Spunt, S. L. Articles by Pappo, A. S. Social Bookmarking                            What's this?