- Email: [email protected]
Failure patterns and survival in pediatric soft tissue sarcoma
IN. J. Radiation Oncology Bid. Phys., Vol. IS, pp. 37-41 Printed in the U.S.A. All rights reserved.
Copyright
0360-3016/88 $3.00 + IJO 0 1988 Pergamon Press plc
??Original Contribution
FAILURE
PATTERNS
DAVID
AND SURVIVAL
M. BRIZEL,
M.D., ’HOWARD
IN PEDIATRIC WEINSTEIN,
SOFT TISSUE
M.D.,2 MYLA
HUNT,
SARCOMA
M.S.3
‘Joint Center for Radiation Therapy, *Pediatric Oncology, Dana-Farber Cancer Institute, The Children’s Hospital (TCH), 3Department of Biostatistics, Dana-Farber Cancer Institute, Harvard Medical School, Boston, MA 02 115 We retrospectively analyzed 44 patients with localized soft tissue sarcomas who were seen and treated at the JCRT, DFCI, and TCH between 1970-1984. Patients with rhabdomyosarcoma were excluded. Primary tumors were located in the following sites: extremities 19 (43%), head and neck 9 (20%), and trunk 16 (37%). Median follow-up for survivors was 7.7 years (range 24 mo-16 years). Surgery was the initial aspect of treatment for all patients. All patients also received post-operative irradiation, 43 at presentation and one at local relapse, and 26 received adjuvant chemotherapy. Radiation was delivered to a dose of 4000 cGy (median) followed by a boost to a median dose of 5760 cGy (range 4500-7000 cGy). Actuarial 5- and lo-year disease-free survivals (DFS) were 70% and 59% while the actuarial 5- and lo-year overall survivals (OS) were both 75%. All parameters were assessed for significance by univariate analysis. OS was significantly affected by presenting stage when analyzed according to both the Intergroup Rhabdomyosarcoma Staging System (IRS) and the American Joint Committee on Cancer system (AJCC). For the IRS, OS at 10 years was 100% for Stage I, 72% for Stage II, and 54% for Stage III (p = 0.04). For the AJCC, OS at 10 years was 100% for Stage I and 65% for Stage II and III @ = 0.05). Primary site, histology, and use of adjuvant chemotherapy did not influence OS or DFS. Fourteen patients failed: 8 local, 1 distant, and 5 combined local and distant. There was no LF among the 9 pts. with primary lesions ~5 cm compared to 1 l/29 (39%) whose tumor was >5 cm @ = 0.04). Pts. with gross residual disease had a local DFS of 42%, but those with no residual or microscopic residual had a local DFS of 71% (p = 0.02). In conclusion, childhood STS has an excellent OS (75% at 10 years). Tumor size and residual tumor after surgery strongly predicted for local failure. Of interest, the pattern of failure is predominantly local in our series. This suggests that more aggressive local treatment is indicated in management of children with STS. Higher doses of irradiation as used for adult STS are probably indicated for patients with gross
residual disease. Pediatric, Soft tissue sarcoma, Radiation therapy. INTRODUCTION
non-rhabdomyosarcoma STS is sparse, however.2*‘0V” This article will review the treatment of pediatric nonrhabdomyosarcoma STS at one institution with particular attention to patterns of relapse and overall survival.
Overall, less than 1% of all cancers are sarcomas, and for children less than 15 years old, soft tissue sarcomas comprise approximately 6% of childhood malignancies. Data from the National Cancer Institute’s SEER14 program
METHODS
for the years 1973-1982 shows an annual average incidence of 360 cases of soft tissue sarcoma (STS) in chil-
AND
MATERIALS
dren under 15 years of age which corresponds to a rate of 7.9 per one million. More than half of these children had rhabdomyosarcoma, an entity which has been studied extensively. The literature dealing with pediatric
The records of all patients less than 18 years old with a diagnosis of soft tissue sarcoma who were treated at the Joint Center for Radiation Therapy, Dana Farber Can-
Presented at the 29th Annual Scientific Meeting ofthe American Society for Therapeutic Radiology and Oncology, Boston, Massachusetts, October 22, 1987. Reprint requests to: David M. Brizel, M.D., Division of Radiation Oncology, Duke University Medical Center, Durham, N.C. 27710.
Acknowledgements-The authors wish to thank Marcia Delorey, R.N., for her assistance with chart reviewing and Ms. Donna Stephenson and Ms. Patricia Taylor for preparation of the manuscript. Accepted for publication 1 February 1988.
cer Institute,
37
and The Children’s
Hospital,
Boston
from
38
I. J. Radiation Oncology 0 Biology 0 Physics Table I. Staging
Stage I
II III N.A.
AJCC (N)
IRS (N)
11
14
1
15
22 10
15 0
1970 through 1984 were retrospectively reviewed. Fiftynine patients were initially identified. After the exclusion of those presenting with disseminated disease or rhabdomyosarcoma, 44 patients remained for analysis. Surgical excision comprised initial management in all patients. Wide excision was most often attempted, and all but one of the extremity patients, who had a lesion of the ankle, had limb-sparing procedures. All patients received post-operative irradiation, 43 at presentation and one at the time of relapse. Initial treatment volumes, which included surgical scars, received a median dose of 4000 cGy in 170-200 cGy daily fractions. Conedown fields received additional irradiation to a median total dose of 5760 cGy (range 4500-7000 cGy). All treatment was given with a 4,6, or 8 MV linear accelerator. Immobilization devices, wedges, compensators, and bolus were used where necessary to improve reproducability and dose homogeneity. Twenty-six patients received adjuvant chemotherapy. In 17 patients this consisted of either a Vincristine, Actinomycin, Cyclophosphamide (VAC) or Vincristine, Adriamycin, Cyclophosphamide (Adria-VAC) regimen. Altogether, Adriamycin was part of the chemotherapy regimen in 16 patients (median dose of 423 mg/m’). Patients treated after 1975 were staged according to American Joint Committee on Cancer (AJCC) criteria.” Those treated prior to 1975 were retrospectively restaged according to these standards where possible (Table 1). Ten patients could not be restaged: 5 could not be graded, and size could not be determined in 5. For the purposes of analysis, all patients were also restaged to conform with the requirements of the Intergroup Rhab-
Table 2. Pediatric STS histologies No. (%)
Malignant schwannoma Synovial cell Undifferentiated sarcoma Malignant fibrous histiocytoma Fibrosarcoma Leiomyosarcoma Spindle cell Angiosarcoma Other
13 (29.5)
Total
44 (100)
7(15.9) 6 (13.6) 5 (11.4) 4 (9.1) 3 (6.8) 2 (4.5)
l(2.3) 3 (6.8)
July 1988, Volume 15,Number 1
domyosarcoma Staging (IRS) system6 (Table 1). All calculations of local control, disease-free survival (DFS), and overall survival were calculated from the time of diagnosis using the Kaplan-Meier method.3 Differences were assessed for significance using Fisher’s Exact Test. RESULTS
The median age at diagnosis was 12 years (range 2 months to 18 years), and the male:female ratio was 2: 1. Median follow-up was 7.7 years with a minimum followup of 2 years. Primary lesions were distributed in the following sites: extremities19 patients (43%), trunk- 16 patients (37%), and head and neck-9 patients (20%). Histopathology varied widely with malignant schwannoma being the most common (Table 2). Disease-free and overall survival For all patients, the lo-year disease-free survival was 59%, and the IO-year overall survival was 75% (Figs. 1, 2). The two late recurrences (6 and 9 years) have both been salvaged with surgery, one with a wide excision and one with an amputation. AJCC presenting stage was a significant determinant of survival (Fig. 3). Ten-year survival for Stage I was 100% and for Stages II and III was 65% (p = 0.05). Stages II and III were considered together since there was only one Stage II patient. IRS stage also exerted a significant impact upon survival (Fig. 4) with Stages I, II, and III being lOO%, 72%, and 54% respectively (p = 0.05). Disease-free survival was not significantly influenced by AJCC stage, but it was affected by IRS stage. DFS was 34% for patients with gross residual tumor after surgery, 62% with microscopic residual, and 83% with no residual (p = 0.05). Analysis of primary site and histology revealed no significant influence upon DFS or overall survival. Likewise, the use of systemic chemotherapy exerted no significant impact upon survival. There was, however, a strong selection bias in favor of giving chemotherapy to those patients felt to have a worse prognosis. Of the 22 patients with grade III lesions, 17 received systemic treatment. A total of 26 patients received chemotherapy, and all but 9 had grade III lesions. Failure patterns Fourteen patients relapsed after treatment. Sites of first failure were as follows: local only 8, distant only 1, and local + distant 5. Pulmonary metastases were present in 5 of the 6 patients who failed in distant sites. Bone, liver, and CNS involvement were also documented in 3 patients. With respect to local failure, 77% were in-field, 8% were marginal recurrences, and 15% were indeterminant. There was a detailed univariate analysis of prognostic factors associated with local control (Table 3). Local tumor control was strongly associated with the size of
Pediatric soft tissue sarcoma 0 D. M. BRIZELet al.
39
4
$.
s.. 3.0
. 6.0. YEARS &R 30
““‘. 9.0 PROG 14
12.0
a ‘.9.0 9l *. 3.0* *. 6.0*YEARS
”
TOTAL 44
-
AJCC STAGE I STAGE _ . STAGE IIiIII
ALIVE 11 16
*.
DEAD 0 7
‘.“.’ 15.0
12.0
TOTAL 1I 23
p = 0.05
Fig. 1. Disease-free survival for all patients.
Fig. 3. Survival by AJCC stage. Stage I was significantly better than Stages II/III.
0
1.01==
-I 2 8
. 9 ?s_ , -8. lb-0 ,I .7e.-o-. -m-m1 -I .6. b--ma-e-----o .5 .4.
E
.3.
z
- *_ a_ - .-
100%
_ _ _ _ _ -0
72x
54x
.2. .l-
BL* *3.0* *. 6.0. a ‘.9.0 YEARS
ALlVE 34
”
DEAD 10
s.
12.0
$
”15.0 L ’
TOTAL 44
I R S -I . _ . --
II III
*“*.3.0 Stage
6.0
*
s. 9.0*. .12.0.
YEARS
ALIVE 14 11 9
DEAD 0 4 6
”
15.0’
”
TOTAL 14 15 15 p - 0.04
Fig. 2. Overall survival for all patients.
Fig. 4. Survival by IRS Stage. I (no residual) vs. II (microscopic residual): p = 0.05. II vs. III (gross residual): p = 0.05.
40
I. J. Radiation Oncology 0 Biology 0 Physics Table 3. Prognostic factors for local failure Local failure
Radiation dose ~5000 cGy 5000-6000 cGy >6000 cGy Tumor size ~5 cm. >5 cm. Initial resection extent No/microscopic residual Gross residual
D value
419 (44%) 4/ 15 (27%) 5/18 (28%)
N.S.
O/9 1 l/29 (39%)
0.04
6/28 (21%) 7/14 (50%)
0.02
the primary and the extent of the initial resection (Table 3). There were no local failures in patients with primaries less than 5 cm compared to 11/29 (39%) in patients with tumors larger than 5 cm (p = 0.04). Patients with gross residual disease had a local control of 50% while those with microscopic or no residual had a local control of 79% (p = 0.02). Histology did not significantly affect local control, but there was a trend towards increased local failure in the patients with malignant schwannoma (p = 0.07).
Toxicity Seventy percent of the patients in this series developed mild to moderate erythema and desquamation in the irradiated volume which resolved with conservative measures. There were no severe skin reactions. Chronic subcutaneous fibrosis occurred in 28/44 (64%) of the patients. It was of moderate degree in 4 (9%), severe in 2 (5%), and mild in all of the others. Three of the patients with thoracic primaries developed symptomatic radiation pneumonitis with one requiring hospitalization. This patient had a malignant schwannoma and had received 180 cGy b.i.d. to a large portion of his lung volume. Three of the 16 patients receiving Adriamycin developed congestive heart failure which was not fatal. Median doses of Adriamycin to the groups with and without cardiotoxicity were 420 mg/m* and 375 mg/m* respectively (p = ns). Life threatening infections were seen in six patients receiving chemotherapy, and one patient died of sepsis with no evidence of tumor found at autopsy. DISCUSSION
The present series has shown an actuarial lo-year disease-free survival of 59% and lo-year survival of 75% in pediatric non-rhabdomyosarcoma STS treated with conservative surgery and radiation therapy. These results are similar to those obtained for the treatment of soft tissue sarcoma in adults.5,‘3 AJCC stage, which is based primar-
July 1988, Volume 15, Number 1
ily on tumor grade, as well as IRS stage grouping, which is based on the extent of residual disease after surgery, were both significant prognostic factors. A study from the University of Florida” (Sollacio, R.J., oral communication, April, 1987) examined 50 pediatric non-rhabdomyosarcoma STS patients treated predominantly with radical surgical resection. Five-year DFS and survival were 60% and 7 1% respectively. Grade was the single most important variable affecting survival (92% v. 59% for low grade v. high grade) as well as local and distant failure. Another review from St. Jude Children’s Hospital* examined 62 patients with non-rhabdomyosarcoma STS of whom 50 had localized presentations. Staging was performed according to IRS guidelines, and survival was significantly better in the group I and II patients than in the group III patients with gross residual (80% v 8%). Histologic grading was not performed in this study. Adjuvant chemotherapy did not alter the prognosis for the group I patients selected to receive it, but the criteria for selection were not discussed. The role of adjuvant chemotherapy in the treatment of soft tissue sarcoma remains uncertain with both positive and negative randomized trials reported1,8,9 and requires further study. This series demonstrated that tumor size and the presence of gross residual disease exerted a significant impact upon local control. Again this is similar to the data obtained in the study of adult STS. Reports from the MD Anderson Hospital4,5 have shown that in adults with soft tissue sarcoma, high grade and tumor size greater than 5 cm. are significant parameters that adversely affect local control, distant failure, and overall survival. Precise information regarding the status of surgical margins was lacking in many of the patients in those studies. Data from the Massachusetts General Hospital” can be used to address the importance of surgical margins. Patients undergoing 6000 cGy post-operatively had a local failure rate of 4% after complete resection as opposed to a local failure rate of 19% in patients with an incomplete resection. The St. Jude study* did not demonstrate any benefit to post-operative irradiation in children with soft tissue sarcoma. Only 14 patients were irradiated, however, of whom 12 had gross residual disease after surgery. Local recurrence was seen in 2/5 group II patients and in 1 I/ 14 group III patients. The median dose given was only 4200 cGy. Given such low doses to so many patients with gross residual disease, it is not surprising that no benefit was demonstrated. Local failure occurred in 13/44 (30%) patients in the present study, and distant metastases developed in 6/44 ( 14%). Although a dose response relationship was not evident, the median tumor dose was only 5760 cGy, and others have shown improved local control for doses greater than 6000 cGy.13 The failure pattern in this series
41
Pediatric soft tissue sarcoma 0 D. M. BRIZEL et al. was predominantly
local. In part this may have been influenced by the relatively low doses used for gross disease or possibly the efficacy of adjuvant chemotherapy in controlling micrometastatic disease without contributing to local control. The trend towards more local failure in patients with malignant schwannoma may also have contributed. An analysis of failure patterns in adult STS from MD Anderson7 demonstrated that malignant schwannoma was the histology associated with the highest local recurrence rate after treatment (29%) and within this histology, local failure accounted for 47% of all failures. Although childhood non-rhabdomyosarcoma STS is an uncommon entity, this study and the others available
in the literature reveal an excellent overall surviva12*” The variables of grade, tumor size, and residual disease after surgery appear to assume the same importance as they do in adult sarcoma with respect to local control and survival. The relative predominance of local failure in this series and the lower doses of irradiation given suggest a role for more aggressive local treatment in the management of children with soft tissue sarcoma. Perhaps further surgery such as re-excision to improve the status of margins followed by post-operative irradiation is indicated. When this more aggressive surgical appreach is not feasible without the loss of function, then higher doses of irradiation should be used as in adult soft tissue sarcoma.
REFERENCES 1. Bramwell, V.H.C., Rouesse, J., Santoro, A., Buesa, J., Somers, R., Thomas, D., Sylvester, R., Pinedo, H.: European experience of adjuvant chemotherapy for soft tissue sarcoma: Preliminary report of randomized trial of cyclophosphamide, vincristine, doxorubicin, and dacarbazine. Cancer Treat. Symp. 3: 99-107, 1985. 2. Horowitz, M.E., Pratt, C.B., Webber, B.L., Hustu, H.O., Etcubanas, E., Miliauskas, J., Rao, B.N., Fleming, I.D., Kuman, A.P.M., Green, A.A.: Therapy for childhood softtissue sarcomas other than rhabdomyosarcoma: a review of 62 cases treated at a single institution. J. Clin. Oncol. 4: 559-564,
1986.
3. Kaplan, E.S., Meier, P.: Non-parametric estimation from incomplete observation. Am. Stat. Assoc. J. 53: 457-480, 1958.
4. Lindberg, R.D.: Treatment of localized soft-tissue sarcomas in adults at MD Anderson Hospital and Tumor Institute (1960-1981). Cancer Treat. Symp. 3: 59-65, 1985. 5. Lindberg, R.D., Martin, R.G., Romsdahl, M.M., Barkley, H.T.: Conservative surgery and post-operative radiotherapy in 300 adults with soft-tissue sarcomas. Cancer 47: 2391-2397,198l.
6. Maurer, H.M., Moon, T.E., Donaldson, M., Fernandez, C., Gehan, E., Hammond, D., Hays, D.M., Lawrence, W., Newton, W., Ragab, A., Raney, B., Soule, E., Sutow, W.W., Tefft, M.: The Intergroup Rhabdomyosarcoma study. Cancer 40: 20 15-2026, 1977. 7. Romsdahl, M.M., Lindberg, R.D., Martin, R.G.: Patterns
of failure after treatment Treat. Symp. 2: 25 l-258,
of soft-tissue sarcoma. Cancer 1983.
8. Rosenberg,
S.A.: Prospective randomized trials demonstrating the efficacy of adjuvant chemotherapy in adult patients with soft-tissue sarcomas. Cancer Treat. Rep. 68:
1067-1078, 1984. 9. Rosenberg, S.A., Chand, A.E., Glatstein, E.: Adjuvant che-
motherapy for treatment of extremity soft-tissue sarcomas: Review of National Cancer Institute experience. Cancer Treat. Symp. 3: 83-88,
1985.
10. Russell, W.O., Cohen, J., Edmondson, J.H., Enzinger, F., Hajdu, S.I., Heise, H., Martin, R.G., Miller, W.T., Schmitz, R.L., Suit, H.D.: Staging system for soft-tissue sarcoma. Sem. Oncol. 8: 156-159, 1981. 11. Sollaccio, R.J., Conrad, C., Mendenhall, N.P., Marcus, R.B., Million, R.R., Springfield, D.S., Enneking, W.F.: Soft-tissue sarcoma in children: analysis of prognostic factors. Znt. J. Radiat. Oncol. Biol. Phys. 12(Suppl. 1): 133, 1986. 12. Suit, H.D.: Patterns of failure after treatment of sarcoma of soft-tissue by radical surgery or by conservative surgery and radiation. Cancer Treat. Symp. 2: 241-246, 1983. 13. Suit, H.D., Mankin, H.J., Schiller, A.L., Wood, W.C., Tepper, J.E., Gebhardt, M.C.: Results of treatment of sarcoma of soft-tissue by radiation and surgery at Massachusetts General Hospital. Cancer Treat. Symp. 3: 43-47, 1985. 14. Young, J.L., Ries, L.G., Silverberg, E., Horm, J.W., Miller, R.W.: Cancer incidence, survival, and mortality for children younger than 15 years. Cancer 58: 598-602,1986.
Copyright
0360-3016/88 $3.00 + IJO 0 1988 Pergamon Press plc
??Original Contribution
FAILURE
PATTERNS
DAVID
AND SURVIVAL
M. BRIZEL,
M.D., ’HOWARD
IN PEDIATRIC WEINSTEIN,
SOFT TISSUE
M.D.,2 MYLA
HUNT,
SARCOMA
M.S.3
‘Joint Center for Radiation Therapy, *Pediatric Oncology, Dana-Farber Cancer Institute, The Children’s Hospital (TCH), 3Department of Biostatistics, Dana-Farber Cancer Institute, Harvard Medical School, Boston, MA 02 115 We retrospectively analyzed 44 patients with localized soft tissue sarcomas who were seen and treated at the JCRT, DFCI, and TCH between 1970-1984. Patients with rhabdomyosarcoma were excluded. Primary tumors were located in the following sites: extremities 19 (43%), head and neck 9 (20%), and trunk 16 (37%). Median follow-up for survivors was 7.7 years (range 24 mo-16 years). Surgery was the initial aspect of treatment for all patients. All patients also received post-operative irradiation, 43 at presentation and one at local relapse, and 26 received adjuvant chemotherapy. Radiation was delivered to a dose of 4000 cGy (median) followed by a boost to a median dose of 5760 cGy (range 4500-7000 cGy). Actuarial 5- and lo-year disease-free survivals (DFS) were 70% and 59% while the actuarial 5- and lo-year overall survivals (OS) were both 75%. All parameters were assessed for significance by univariate analysis. OS was significantly affected by presenting stage when analyzed according to both the Intergroup Rhabdomyosarcoma Staging System (IRS) and the American Joint Committee on Cancer system (AJCC). For the IRS, OS at 10 years was 100% for Stage I, 72% for Stage II, and 54% for Stage III (p = 0.04). For the AJCC, OS at 10 years was 100% for Stage I and 65% for Stage II and III @ = 0.05). Primary site, histology, and use of adjuvant chemotherapy did not influence OS or DFS. Fourteen patients failed: 8 local, 1 distant, and 5 combined local and distant. There was no LF among the 9 pts. with primary lesions ~5 cm compared to 1 l/29 (39%) whose tumor was >5 cm @ = 0.04). Pts. with gross residual disease had a local DFS of 42%, but those with no residual or microscopic residual had a local DFS of 71% (p = 0.02). In conclusion, childhood STS has an excellent OS (75% at 10 years). Tumor size and residual tumor after surgery strongly predicted for local failure. Of interest, the pattern of failure is predominantly local in our series. This suggests that more aggressive local treatment is indicated in management of children with STS. Higher doses of irradiation as used for adult STS are probably indicated for patients with gross
residual disease. Pediatric, Soft tissue sarcoma, Radiation therapy. INTRODUCTION
non-rhabdomyosarcoma STS is sparse, however.2*‘0V” This article will review the treatment of pediatric nonrhabdomyosarcoma STS at one institution with particular attention to patterns of relapse and overall survival.
Overall, less than 1% of all cancers are sarcomas, and for children less than 15 years old, soft tissue sarcomas comprise approximately 6% of childhood malignancies. Data from the National Cancer Institute’s SEER14 program
METHODS
for the years 1973-1982 shows an annual average incidence of 360 cases of soft tissue sarcoma (STS) in chil-
AND
MATERIALS
dren under 15 years of age which corresponds to a rate of 7.9 per one million. More than half of these children had rhabdomyosarcoma, an entity which has been studied extensively. The literature dealing with pediatric
The records of all patients less than 18 years old with a diagnosis of soft tissue sarcoma who were treated at the Joint Center for Radiation Therapy, Dana Farber Can-
Presented at the 29th Annual Scientific Meeting ofthe American Society for Therapeutic Radiology and Oncology, Boston, Massachusetts, October 22, 1987. Reprint requests to: David M. Brizel, M.D., Division of Radiation Oncology, Duke University Medical Center, Durham, N.C. 27710.
Acknowledgements-The authors wish to thank Marcia Delorey, R.N., for her assistance with chart reviewing and Ms. Donna Stephenson and Ms. Patricia Taylor for preparation of the manuscript. Accepted for publication 1 February 1988.
cer Institute,
37
and The Children’s
Hospital,
Boston
from
38
I. J. Radiation Oncology 0 Biology 0 Physics Table I. Staging
Stage I
II III N.A.
AJCC (N)
IRS (N)
11
14
1
15
22 10
15 0
1970 through 1984 were retrospectively reviewed. Fiftynine patients were initially identified. After the exclusion of those presenting with disseminated disease or rhabdomyosarcoma, 44 patients remained for analysis. Surgical excision comprised initial management in all patients. Wide excision was most often attempted, and all but one of the extremity patients, who had a lesion of the ankle, had limb-sparing procedures. All patients received post-operative irradiation, 43 at presentation and one at the time of relapse. Initial treatment volumes, which included surgical scars, received a median dose of 4000 cGy in 170-200 cGy daily fractions. Conedown fields received additional irradiation to a median total dose of 5760 cGy (range 4500-7000 cGy). All treatment was given with a 4,6, or 8 MV linear accelerator. Immobilization devices, wedges, compensators, and bolus were used where necessary to improve reproducability and dose homogeneity. Twenty-six patients received adjuvant chemotherapy. In 17 patients this consisted of either a Vincristine, Actinomycin, Cyclophosphamide (VAC) or Vincristine, Adriamycin, Cyclophosphamide (Adria-VAC) regimen. Altogether, Adriamycin was part of the chemotherapy regimen in 16 patients (median dose of 423 mg/m’). Patients treated after 1975 were staged according to American Joint Committee on Cancer (AJCC) criteria.” Those treated prior to 1975 were retrospectively restaged according to these standards where possible (Table 1). Ten patients could not be restaged: 5 could not be graded, and size could not be determined in 5. For the purposes of analysis, all patients were also restaged to conform with the requirements of the Intergroup Rhab-
Table 2. Pediatric STS histologies No. (%)
Malignant schwannoma Synovial cell Undifferentiated sarcoma Malignant fibrous histiocytoma Fibrosarcoma Leiomyosarcoma Spindle cell Angiosarcoma Other
13 (29.5)
Total
44 (100)
7(15.9) 6 (13.6) 5 (11.4) 4 (9.1) 3 (6.8) 2 (4.5)
l(2.3) 3 (6.8)
July 1988, Volume 15,Number 1
domyosarcoma Staging (IRS) system6 (Table 1). All calculations of local control, disease-free survival (DFS), and overall survival were calculated from the time of diagnosis using the Kaplan-Meier method.3 Differences were assessed for significance using Fisher’s Exact Test. RESULTS
The median age at diagnosis was 12 years (range 2 months to 18 years), and the male:female ratio was 2: 1. Median follow-up was 7.7 years with a minimum followup of 2 years. Primary lesions were distributed in the following sites: extremities19 patients (43%), trunk- 16 patients (37%), and head and neck-9 patients (20%). Histopathology varied widely with malignant schwannoma being the most common (Table 2). Disease-free and overall survival For all patients, the lo-year disease-free survival was 59%, and the IO-year overall survival was 75% (Figs. 1, 2). The two late recurrences (6 and 9 years) have both been salvaged with surgery, one with a wide excision and one with an amputation. AJCC presenting stage was a significant determinant of survival (Fig. 3). Ten-year survival for Stage I was 100% and for Stages II and III was 65% (p = 0.05). Stages II and III were considered together since there was only one Stage II patient. IRS stage also exerted a significant impact upon survival (Fig. 4) with Stages I, II, and III being lOO%, 72%, and 54% respectively (p = 0.05). Disease-free survival was not significantly influenced by AJCC stage, but it was affected by IRS stage. DFS was 34% for patients with gross residual tumor after surgery, 62% with microscopic residual, and 83% with no residual (p = 0.05). Analysis of primary site and histology revealed no significant influence upon DFS or overall survival. Likewise, the use of systemic chemotherapy exerted no significant impact upon survival. There was, however, a strong selection bias in favor of giving chemotherapy to those patients felt to have a worse prognosis. Of the 22 patients with grade III lesions, 17 received systemic treatment. A total of 26 patients received chemotherapy, and all but 9 had grade III lesions. Failure patterns Fourteen patients relapsed after treatment. Sites of first failure were as follows: local only 8, distant only 1, and local + distant 5. Pulmonary metastases were present in 5 of the 6 patients who failed in distant sites. Bone, liver, and CNS involvement were also documented in 3 patients. With respect to local failure, 77% were in-field, 8% were marginal recurrences, and 15% were indeterminant. There was a detailed univariate analysis of prognostic factors associated with local control (Table 3). Local tumor control was strongly associated with the size of
Pediatric soft tissue sarcoma 0 D. M. BRIZELet al.
39
4
$.
s.. 3.0
. 6.0. YEARS &R 30
““‘. 9.0 PROG 14
12.0
a ‘.9.0 9l *. 3.0* *. 6.0*YEARS
”
TOTAL 44
-
AJCC STAGE I STAGE _ . STAGE IIiIII
ALIVE 11 16
*.
DEAD 0 7
‘.“.’ 15.0
12.0
TOTAL 1I 23
p = 0.05
Fig. 1. Disease-free survival for all patients.
Fig. 3. Survival by AJCC stage. Stage I was significantly better than Stages II/III.
0
1.01==
-I 2 8
. 9 ?s_ , -8. lb-0 ,I .7e.-o-. -m-m1 -I .6. b--ma-e-----o .5 .4.
E
.3.
z
- *_ a_ - .-
100%
_ _ _ _ _ -0
72x
54x
.2. .l-
BL* *3.0* *. 6.0. a ‘.9.0 YEARS
ALlVE 34
”
DEAD 10
s.
12.0
$
”15.0 L ’
TOTAL 44
I R S -I . _ . --
II III
*“*.3.0 Stage
6.0
*
s. 9.0*. .12.0.
YEARS
ALIVE 14 11 9
DEAD 0 4 6
”
15.0’
”
TOTAL 14 15 15 p - 0.04
Fig. 2. Overall survival for all patients.
Fig. 4. Survival by IRS Stage. I (no residual) vs. II (microscopic residual): p = 0.05. II vs. III (gross residual): p = 0.05.
40
I. J. Radiation Oncology 0 Biology 0 Physics Table 3. Prognostic factors for local failure Local failure
Radiation dose ~5000 cGy 5000-6000 cGy >6000 cGy Tumor size ~5 cm. >5 cm. Initial resection extent No/microscopic residual Gross residual
D value
419 (44%) 4/ 15 (27%) 5/18 (28%)
N.S.
O/9 1 l/29 (39%)
0.04
6/28 (21%) 7/14 (50%)
0.02
the primary and the extent of the initial resection (Table 3). There were no local failures in patients with primaries less than 5 cm compared to 11/29 (39%) in patients with tumors larger than 5 cm (p = 0.04). Patients with gross residual disease had a local control of 50% while those with microscopic or no residual had a local control of 79% (p = 0.02). Histology did not significantly affect local control, but there was a trend towards increased local failure in the patients with malignant schwannoma (p = 0.07).
Toxicity Seventy percent of the patients in this series developed mild to moderate erythema and desquamation in the irradiated volume which resolved with conservative measures. There were no severe skin reactions. Chronic subcutaneous fibrosis occurred in 28/44 (64%) of the patients. It was of moderate degree in 4 (9%), severe in 2 (5%), and mild in all of the others. Three of the patients with thoracic primaries developed symptomatic radiation pneumonitis with one requiring hospitalization. This patient had a malignant schwannoma and had received 180 cGy b.i.d. to a large portion of his lung volume. Three of the 16 patients receiving Adriamycin developed congestive heart failure which was not fatal. Median doses of Adriamycin to the groups with and without cardiotoxicity were 420 mg/m* and 375 mg/m* respectively (p = ns). Life threatening infections were seen in six patients receiving chemotherapy, and one patient died of sepsis with no evidence of tumor found at autopsy. DISCUSSION
The present series has shown an actuarial lo-year disease-free survival of 59% and lo-year survival of 75% in pediatric non-rhabdomyosarcoma STS treated with conservative surgery and radiation therapy. These results are similar to those obtained for the treatment of soft tissue sarcoma in adults.5,‘3 AJCC stage, which is based primar-
July 1988, Volume 15, Number 1
ily on tumor grade, as well as IRS stage grouping, which is based on the extent of residual disease after surgery, were both significant prognostic factors. A study from the University of Florida” (Sollacio, R.J., oral communication, April, 1987) examined 50 pediatric non-rhabdomyosarcoma STS patients treated predominantly with radical surgical resection. Five-year DFS and survival were 60% and 7 1% respectively. Grade was the single most important variable affecting survival (92% v. 59% for low grade v. high grade) as well as local and distant failure. Another review from St. Jude Children’s Hospital* examined 62 patients with non-rhabdomyosarcoma STS of whom 50 had localized presentations. Staging was performed according to IRS guidelines, and survival was significantly better in the group I and II patients than in the group III patients with gross residual (80% v 8%). Histologic grading was not performed in this study. Adjuvant chemotherapy did not alter the prognosis for the group I patients selected to receive it, but the criteria for selection were not discussed. The role of adjuvant chemotherapy in the treatment of soft tissue sarcoma remains uncertain with both positive and negative randomized trials reported1,8,9 and requires further study. This series demonstrated that tumor size and the presence of gross residual disease exerted a significant impact upon local control. Again this is similar to the data obtained in the study of adult STS. Reports from the MD Anderson Hospital4,5 have shown that in adults with soft tissue sarcoma, high grade and tumor size greater than 5 cm. are significant parameters that adversely affect local control, distant failure, and overall survival. Precise information regarding the status of surgical margins was lacking in many of the patients in those studies. Data from the Massachusetts General Hospital” can be used to address the importance of surgical margins. Patients undergoing 6000 cGy post-operatively had a local failure rate of 4% after complete resection as opposed to a local failure rate of 19% in patients with an incomplete resection. The St. Jude study* did not demonstrate any benefit to post-operative irradiation in children with soft tissue sarcoma. Only 14 patients were irradiated, however, of whom 12 had gross residual disease after surgery. Local recurrence was seen in 2/5 group II patients and in 1 I/ 14 group III patients. The median dose given was only 4200 cGy. Given such low doses to so many patients with gross residual disease, it is not surprising that no benefit was demonstrated. Local failure occurred in 13/44 (30%) patients in the present study, and distant metastases developed in 6/44 ( 14%). Although a dose response relationship was not evident, the median tumor dose was only 5760 cGy, and others have shown improved local control for doses greater than 6000 cGy.13 The failure pattern in this series
41
Pediatric soft tissue sarcoma 0 D. M. BRIZEL et al. was predominantly
local. In part this may have been influenced by the relatively low doses used for gross disease or possibly the efficacy of adjuvant chemotherapy in controlling micrometastatic disease without contributing to local control. The trend towards more local failure in patients with malignant schwannoma may also have contributed. An analysis of failure patterns in adult STS from MD Anderson7 demonstrated that malignant schwannoma was the histology associated with the highest local recurrence rate after treatment (29%) and within this histology, local failure accounted for 47% of all failures. Although childhood non-rhabdomyosarcoma STS is an uncommon entity, this study and the others available
in the literature reveal an excellent overall surviva12*” The variables of grade, tumor size, and residual disease after surgery appear to assume the same importance as they do in adult sarcoma with respect to local control and survival. The relative predominance of local failure in this series and the lower doses of irradiation given suggest a role for more aggressive local treatment in the management of children with soft tissue sarcoma. Perhaps further surgery such as re-excision to improve the status of margins followed by post-operative irradiation is indicated. When this more aggressive surgical appreach is not feasible without the loss of function, then higher doses of irradiation should be used as in adult soft tissue sarcoma.
REFERENCES 1. Bramwell, V.H.C., Rouesse, J., Santoro, A., Buesa, J., Somers, R., Thomas, D., Sylvester, R., Pinedo, H.: European experience of adjuvant chemotherapy for soft tissue sarcoma: Preliminary report of randomized trial of cyclophosphamide, vincristine, doxorubicin, and dacarbazine. Cancer Treat. Symp. 3: 99-107, 1985. 2. Horowitz, M.E., Pratt, C.B., Webber, B.L., Hustu, H.O., Etcubanas, E., Miliauskas, J., Rao, B.N., Fleming, I.D., Kuman, A.P.M., Green, A.A.: Therapy for childhood softtissue sarcomas other than rhabdomyosarcoma: a review of 62 cases treated at a single institution. J. Clin. Oncol. 4: 559-564,
1986.
3. Kaplan, E.S., Meier, P.: Non-parametric estimation from incomplete observation. Am. Stat. Assoc. J. 53: 457-480, 1958.
4. Lindberg, R.D.: Treatment of localized soft-tissue sarcomas in adults at MD Anderson Hospital and Tumor Institute (1960-1981). Cancer Treat. Symp. 3: 59-65, 1985. 5. Lindberg, R.D., Martin, R.G., Romsdahl, M.M., Barkley, H.T.: Conservative surgery and post-operative radiotherapy in 300 adults with soft-tissue sarcomas. Cancer 47: 2391-2397,198l.
6. Maurer, H.M., Moon, T.E., Donaldson, M., Fernandez, C., Gehan, E., Hammond, D., Hays, D.M., Lawrence, W., Newton, W., Ragab, A., Raney, B., Soule, E., Sutow, W.W., Tefft, M.: The Intergroup Rhabdomyosarcoma study. Cancer 40: 20 15-2026, 1977. 7. Romsdahl, M.M., Lindberg, R.D., Martin, R.G.: Patterns
of failure after treatment Treat. Symp. 2: 25 l-258,
of soft-tissue sarcoma. Cancer 1983.
8. Rosenberg,
S.A.: Prospective randomized trials demonstrating the efficacy of adjuvant chemotherapy in adult patients with soft-tissue sarcomas. Cancer Treat. Rep. 68:
1067-1078, 1984. 9. Rosenberg, S.A., Chand, A.E., Glatstein, E.: Adjuvant che-
motherapy for treatment of extremity soft-tissue sarcomas: Review of National Cancer Institute experience. Cancer Treat. Symp. 3: 83-88,
1985.
10. Russell, W.O., Cohen, J., Edmondson, J.H., Enzinger, F., Hajdu, S.I., Heise, H., Martin, R.G., Miller, W.T., Schmitz, R.L., Suit, H.D.: Staging system for soft-tissue sarcoma. Sem. Oncol. 8: 156-159, 1981. 11. Sollaccio, R.J., Conrad, C., Mendenhall, N.P., Marcus, R.B., Million, R.R., Springfield, D.S., Enneking, W.F.: Soft-tissue sarcoma in children: analysis of prognostic factors. Znt. J. Radiat. Oncol. Biol. Phys. 12(Suppl. 1): 133, 1986. 12. Suit, H.D.: Patterns of failure after treatment of sarcoma of soft-tissue by radical surgery or by conservative surgery and radiation. Cancer Treat. Symp. 2: 241-246, 1983. 13. Suit, H.D., Mankin, H.J., Schiller, A.L., Wood, W.C., Tepper, J.E., Gebhardt, M.C.: Results of treatment of sarcoma of soft-tissue by radiation and surgery at Massachusetts General Hospital. Cancer Treat. Symp. 3: 43-47, 1985. 14. Young, J.L., Ries, L.G., Silverberg, E., Horm, J.W., Miller, R.W.: Cancer incidence, survival, and mortality for children younger than 15 years. Cancer 58: 598-602,1986.