Soft tissue sarcomas treated with postoperative external beam radiotherapy with and without low-dose-rate brachytherapy

Int. J. Radiation Oncology Biol. Phys., Vol. 59, No. 2, pp. 475– 480, 2004 Copyright © 2004 Elsevier Inc. Printed in the USA. All rights reserved 0360-3016/04/$–see front matter

doi:10.1016/j.ijrobp.2003.10.048

CLINICAL INVESTIGATION

Sarcoma

SOFT TISSUE SARCOMAS TREATED WITH POSTOPERATIVE EXTERNAL BEAM RADIOTHERAPY WITH AND WITHOUT LOW-DOSE-RATE BRACHYTHERAPY STEPHEN F. ANDREWS, D.O.,* PENNY R. ANDERSON, M.D.,* BURTON L. EISENBERG, M.D.,† ALEXANDRA L. HANLON, PH.D.,* AND ALAN POLLACK, M.D., PH.D.* Departments of *Radiation Oncology, Fox Chase Cancer Center, Philadelphia, PA and †Surgical Oncology, Fox Chase Cancer Center, Philadelphia, PA Purpose: Patients treated for soft tissue sarcoma with adjuvant low-dose-rate brachytherapy (BT) plus external beam radiotherapy (EBRT) were compared with those treated with adjuvant EBRT alone. The hypothesis was that higher doses from postoperative BT plus EBRT would improve local tumor control. Methods and Materials: The medical records of 130 sarcoma patients definitively treated between February 1983 and February 2001 were reviewed. Of these, 25 patients received BT followed by EBRT, and 61 were treated with EBRT alone. Overall survival, freedom from distant metastasis, and local control were calculated using Kaplan-Meier estimates. Univariate and multivariate analyses were performed. The mean postoperative radiation dose with EBRT alone was 59 Gy (range, 50 –74) and 50 Gy (range, 40 –70 Gy) when low-dose-rate BT was included. The mean implant dose was 16 Gy (range, 10 –20 Gy). Results: The 5-year Kaplan-Meier estimate for overall survival for BT plus EBRT and EBRT alone was 82% and 72% (p ⴝ 0.93), respectively. The 5-year freedom from distant metastasis and freedom from local failure rate for BT plus EBRT vs. EBRT alone was 90% vs. 78% (p ⴝ 0.15) and 90% vs. 83% (p ⴝ 0.25), respectively. In the univariate subset analysis, Stage III patients had better local control at 5 years (100% vs. 62%, p ⴝ 0.03) and a trend was noted for better local control for high-grade tumors (100% vs. 74%, p ⴝ 0.09) if treated with BT plus EBRT. No statistically significant predictors were found on multivariate analysis for local control. The median follow-up was 62 months. Conclusion: Local control at 5 years was high in both groups at 83% and 90%. On univariate analysis, Stage III patients had improved 5-year local control and a trend was found toward better local control for high-grade tumors. On multivariate analysis, no predictors were found for better local control; however, the numbers of Stage III and high-grade patients were small, which may have masked a possible benefit of BT plus EBRT in this population. © 2004 Elsevier Inc. Sarcoma, Radiotherapy, Brachytherapy, Local control.

The timing of RT before or after limb-sparing surgery has been debated and varies by institution (3). Preoperative EBRT has the advantage of a smaller field size and lower doses, which potentially provides better functional outcome (4). It is also used when tumors are bulky and/or in close proximity to structures such as bone or neurovascular bundles, and surgery would otherwise be morbid or technically difficult. Increased postoperative wound complications with preoperative therapy have been reported, including a Phase III randomized trial from the Canadian National Cancer Institute of Clinical Trials (5). A follow-up of this same study showed that these wound complications only predicted for a decreased functional decrement for up to 6 months after surgery (6).

INTRODUCTION Soft tissue sarcomas (STSs) in adults are a relatively uncommon malignancy, with an estimated incidence of 8300 new cases in 2002 (1). The primary treatment is surgery, but the extent of surgery has been reduced during the past few decades. Radical surgery, such as amputation or gross resections that remove large volumes of normal tissue at the expense of function and cosmesis, was the standard of care before the early 1980s. This changed when a prospective randomized trial from the National Cancer Institute showed that limb-sparing surgery combined with adjuvant external beam radiotherapy (EBRT) was possible without compromising survival compared with amputation alone (2). Reprint requests to: Penny R. Anderson, M.D., Department of Radiation Oncology, Fox Chase Cancer Center, 7701 Burholme Ave., Philadelphia, PA 19111. Tel: (215) 728-7476; Fax: (215) 718-2868; E-mail: [email protected] Presented at the 2002 Annual Meeting of the American Society

for Therapeutic Radiology and Oncology, New Orleans, LA, October 2002. Received Apr 15, 2003, and in revised form Oct 29, 2003. Accepted for publication Oct 31, 2003. 475

476

I. J. Radiation Oncology

● Biology ● Physics

When postoperative RT is used, it is most commonly given as either low-dose-rate brachytherapy (BT) or EBRT, or a combination of the two. The potential benefits of BT alone include a reduced treatment time, higher doses to the tumor bed, and the sparing of overlying skin. The Memorial Sloan-Kettering MSK group has published the results of a randomized trial comparing BT alone after surgery vs. no further treatment and showed statistically significant improved local control for completely resected high-grade tumors (7). A follow-up study demonstrated that patients with positive margins had a higher rate of local failure for high-grade tumors (8). This was consistent with the findings from the National Cancer Institute randomized trial and others that showed higher local failure rates in patients with positive margins treated with EBRT and conservative surgery. Thus the addition of EBRT to BT when margins are positive has been proposed by several authors (9, 10). The rationale is to use BT for control of the tumor bed and to add EBRT for wider coverage to minimize the potential for marginal failures. This was examined in a separate study from the MSK group that revealed that when EBRT was added to BT and compared with BT alone a trend was found toward better local control (p ⫽ 0.08) (10). Few have reported on the combination of BT and EBRT for STS, and, to our knowledge, BT plus EBRT and EBRT alone have not been compared. The hypothesis we investigated was whether higher radiation doses from BT plus EBRT would result in better local tumor control than with EBRT alone. Also, we attempted to identify subgroups of patients who may benefit from this combination therapy. METHODS AND MATERIALS The medical records of 130 patients definitively treated for STS between February 1983 and February 2001 were reviewed. These included tumors of the head and neck, upper and lower extremities, and trunk. Visceral sarcomas were excluded. Of these, 25 patients received adjuvant BT followed by EBRT, and 61 were treated with EBRT alone. The patient and tumor characteristics are given in Table 1. A larger percentage of tumors with positive margins were treated with EBRT alone than with BT and EBRT. However, the percentage of patients with pathologically negative margins was similar between the two groups (56% vs. 68%, p ⫽ 0.34). Margin status was defined as positive, close, negative, or unknown and was determined by the pathology report. Any negative margin documented as ⱕ2 mm was considered close. Histologically, the three most common tumors were malignant fibrous histiocytomas, liposarcoma, and leiomyosarcoma in both treatment groups. (Table 1). Most patients were initially seen and underwent surgery by an outside surgeon and were subsequently referred to the Fox Chase Cancer Center. A wide local excision was attempted as the first surgery in 23 (27%) of 86 patients. Of these 23 patients, 10 (43%) required additional surgery in an attempt to achieve clear negative margins. A single patient underwent a third wide excision. Neoadjuvant or adjuvant

Volume 59, Number 2, 2004

Table 1. Patient and tumor characteristics Characteristic Gender (n) Male Female Age (y) Median Range Size (cm) Median Range Histologic type (n) MFH/spindle cell Liposarcoma Leiomyosarcoma Synovial Other* Surgical margins (n) Negative Positive Close Unknown Grade (n) Low Intermediate High Unknown Site Head and neck Upper extremity Lower extremity Trunk

EBRT alone

BT plus EBRT

35 (56) 26 (44)

13 (52) 12 (48)

52.5 17–82

52.5 23–80

7.8 1–30

9.6 3–30

21 (34.4) 12 (19.7) 7 (11.5) 3 (4.9) 18 (29.5)

11 (44.0) 10 (40.0) 3 (12.0) 0 (0.0) 1 (4.0)

34 (55.7) 15 (24.6) 8 (13.1) 4 (6.6)

17 (68.0) 1 (4.0) 4 (16.0) 3 (12.0)

18 (29.5) 8 (13.1) 25 (41.0) 10 (16.4)

6 (24.0) 3 (12.0) 12 (48.0) 4 (16)

2 (3.3) 6 (9.8) 21 (34.4) 32 (52.2)

0 (0.0) 7 (28.0) 13 (52.0) 5 (20.0)

Abbreviations: EBRT ⫽ external beam radiotherapy; BT ⫽ brachytherapy; MFH ⫽ malignant fibrous histiocytoma. Numbers in parentheses are percentages. * Other histologic subtypes included angiosarcoma, clear cell, giant cell tumor of small parts, fibromyxoid, epitheloid, malignant schwanomma, alamtinoma, and hemangiopericytoma.

chemotherapy was used infrequently. Eight percent (2 of 25) and 5% (3 of 61) of patients who received BT plus EBRT and EBRT alone were treated with chemotherapy, respectively. The decision to use BT was made before surgery on the basis of the perceived risk of having microscopically positive margins. This assessment was made by the surgeon and radiation oncologist and took into account factors such as size of the gross tumor, grade, and proximity to nonresectable structures. In all cases, catheters were placed 1 cm apart to cover the tumor bed adequately, plus a margin of 1 to 2 cm when possible. The number of catheters was dependent on size of the tumor bed and ranged from 4 to 24. The implant was loaded within 5 days in all cases and by 2 days in most cases. 192Ir was used as the isotope for all implants. The mean implant dose and dose rate was 16 Gy (range, 10 –20 Gy) and 48 cGy/hr (range, 21– 80), respectively. EBRT was administered by linear accelerators in 81 (94%) of 86 cases. Five individuals were treated with 60 Co. Since 1994, approximately 90% of patients receiving EBRT underwent CT simulation as part of their treatment plan-

Postoperative RT for soft tissue sarcoma

● S. F. ANDREWS et al.

477

Fig. 1. Five-year Kaplan-Meier estimates of overall survival between external beam radiotherapy (EBRT) alone and brachytherapy (BT) plus EBRT.

ning. Radiation energy and beam arrangements were selected by the attending radiation oncologist and were driven by tumor location and patient anatomy. The initiation of EBRT after BT occurred on average 32 days (range, 17–54) after implant. The mean radiation dose when EBRT was used alone was 59 Gy (range, 50 –74 Gy) and 50 Gy (range, 40 –70 Gy) when low-dose-rate BT was included. Overall survival, freedom from distant metastasis, and local control were calculated using Kaplan-Meier estimates. Multivariate analysis (MVA) for predictors of overall survival, distant metastasis, and local failure included tumor size (⬍8 cm vs. ⱖ8 cm), grade (low vs. intermediate vs. high), age (ⱕ50 years vs. ⬎50 years), histologic type (malignant fibrous histiocytoma/spindle cell vs. liposarcoma vs. leiomyosarcoma vs. synovial vs. other), stage (American Joint Committee on Cancer 1997), margin status (positive/close/unknown vs. negative) and site (head and neck vs. upper extremity vs. lower extremity and trunk). The median follow-up was 63 and 62 months for EBRT and BT plus EBRT, respectively. RESULTS The 5-year Kaplan-Meier estimates for overall survival and distant metastasis were not statistically different at 82% vs. 73% (p ⫽ 0.93) and 90% vs. 78% (p ⫽ 0.15) for BT plus EBRT compared with EBRT alone, respectively (Figs. 1 and 2). No statistically significant difference was found in the 5-year local control rate between the two treatment

groups at 83% for EBRT patients and 90% for BT plus EBRT patients (p ⫽ 0.25; Fig. 3). On univariate analysis of local control, tumor size, stage, grade, and margin status were analyzed for the effect of the type of treatment (Table 2). Significantly better local control was found at 5 years in Stage III patients (100% vs. 62%, p ⫽ 0.03) between the two treatment groups. No local control benefit occurred when Stage I and II patients were combined between EBRT alone and BT plus EBRT (85% and 80%, respectively; p ⫽ 0.79). A trend was noted toward better local control for high-grade tumors when BT was added to EBRT (100% vs. 74%, p ⫽ 0.09). On MVA, no predictors of improved local control were found. With respect to overall survival, only stage was significant on MVA (p ⫽ 0.0034). When reviewing toxicities, we recorded events during treatment and within the first several months of followup. RT was well tolerated in both treatment groups. Eight percent (5 of 61) of patients required oral or intravenous antibiotics for cellulitis or an infected wound either during treatment or within 3 months of completing EBRT alone. In two of these cases, an open wound associated with the infection was present. One patient healed within 1 month of treatment, and the other was eventually diagnosed with a locally recurrent tumor at the site of the nonhealing wound. None of 25 patients who received combination therapy developed an infection requiring treatment with antibiotics. Two cases (8%) of a noninfected open wound that healed within 1 month of completing BT and EBRT occurred.

478

I. J. Radiation Oncology

● Biology ● Physics

Volume 59, Number 2, 2004

Fig. 2. Five-year Kaplan-Meier estimates of freedom from distant metastasis between external beam radiotherapy (EBRT) alone and brachytherapy (BT) plus EBRT.

DISCUSSION The overall 5-year local control rates of 83% for EBRT alone and 90% for BT plus EBRT are comparable to those

of other large reported series in the literature. The Massachusetts General Hospital group (11) reported on 258 patients between 1971 and 1985 who were treated with either pre- or postoperative RT and had a 5-year local control rate

Fig. 3. Five-year Kaplan-Meier estimates of local control between external beam radiotherapy (EBRT) alone and brachytherapy (BT) plus EBRT.

Postoperative RT for soft tissue sarcoma

Table 2. Univariate analysis: 5-year local control Outcome Overall Margin status Positive/close/unknown Negative High grade Size (cm) ⬍8 ⱖ8 Stage III I–II

EBRT alone (%)

BT plus EBRT (%)

p

83 (10/61)

90 (2/25)

0.25

69 (6/23) 90 (4/34) 74 (5/25)

100 (0/5) 94 (1/17) 100 (0/12)

0.25 0.63 0.09

82 (6/33) 85 (4/26)

100 (0/12) 81 (2/13)

0.12 0.81

62 (5/15) 85 (4/36)

100 (0/10) 81 (2/11)

0.03 0.79

Abbreviations as in Table 1.

of 86% and 91%, respectively. Barkley et al. (12) from M.D. Anderson had 80% local control for those patients treated with EBRT alone postoperatively. Alekitar et al. (13) at MSK reported a 5-year local control rate of 84% for 204 patients with high-grade tumors of ⬍5 cm who were treated adjuvantly with EBRT, EBRT plus BT, or BT alone. The group at MSK has demonstrated a benefit in resected STS with the addition of BT alone compared with observation for high-grade tumors. In our series, when BT was added to EBRT, a trend was noted toward improved local control for high-grade tumors compared with EBRT alone, but this did not hold up on MVA, most likely because of the small patient numbers. With respect to margin status, we did not show a statistically significant local control benefit with the addition of BT to EBRT, most likely, again, because of the small patient numbers. No predictors of local control were found on MVA. Another factor that may have biased against a local control benefit was that a greater percentage of patients undergoing combination therapy had lesions in the upper extremity compared with patients treated with EBRT alone (31% vs. 9%). Upper extremity tumors have been shown to have significantly worse local control than lower extremity tumors in patients treated with RT (8). The sequencing of chemotherapy with EBRT also remains controversial. Only 2 of 13 randomized trials using first-generation chemotherapy regimens showed an overall survival benefit in the adjuvant setting (14, 15). However, a

● S. F. ANDREWS et al.

479

meta-analysis of first-generation adjuvant chemotherapy agents did show improved local control, distant metastasisfree and disease-free survival (p ⫽ 0.016, p ⫽ 0.0003, p ⫽ 0.0001), but not overall survival (p ⫽ 0.12), with a median follow-up of 9.4 years. A recently published prospective randomized trial using ifosfamide for high-risk sarcomas showed an overall survival and disease-free survival benefit to adjuvant chemotherapy but only a trend toward better local control at 4 years (p ⫽ 0.09) (16). In this study, as in others, despite good local control, distant metastases occurred in close to one-half of patients. Thus, neoadjuvant chemotherapy and RT is now being investigated to try to improve systemic and local control of disease. Massachusetts General Hospital completed a Phase II trial of adult patients with deep extremity STSs ⱖ8 cm treated with neoadjuvant chemotherapy and EBRT followed by resection and then by additional chemotherapy and RT. Patients completing this regimen were compared with historical controls and demonstrated an actuarial 5-year local control, disease-free survival, and overall survival rate of 100%, 84%, and 93%, respectively, compared with 97%, 45%, and 60% in the historical controls (17). This strategy is being further studied by the Radiation Therapy Oncology Group (RTOG) in RTOG 9514, a Phase II trial, which treated a similar subset of high-risk patients with large high-grade tumors with preoperative EBRT and MAID (Mesna, Adriamycin, Ifosfamide, Dacarbazine) chemotherapy. Thus, at present, our RT approach has been more preoperative than postoperative and, as a consequence, the combination of BT and EBRT is not being used that frequently. When possible, neoadjuvant chemotherapy is sequenced with preoperative EBRT. However, when tumor resection has been performed elsewhere and re-excision with postoperative RT is planned, our data are the first to suggest that the addition of BT to EBRT may improve local tumor control in high-grade and/or locally advanced STSs compared with EBRT alone. As with most single-institution retrospective studies, caution should be used when interpreting the results of this study. Factors such as stage migration, selection bias, and imbalances in patient characteristics such as size, histologic type, and site are the limitations of this analysis. Studies with larger cases of BT and EBRT are needed to confirm these results.

REFERENCES 1. Cancer facts and figures. American Cancer Society; Atlanta, GA; 2002. p. 1– 48. 2. Rosenberg S, Tepper J, Glatstein E, et al. The treatment of soft-tissue sarcomas of the extremities, prospective randomized evaluations of (1) limb-sparing surgery plus radiation therapy compared with amputation and (2) the role of adjuvant chemotherapy. Ann Surg 1982;196:305–315. 3. Pollack A, Zagars G, Goswitz MS, et al. Preoperative vs. postoperative radiotherapy in the treatment of soft tissue sarcomas: A matter of presentation. Int J Radiat Oncol Biol Phys 1998;42:563–572. 4. Karasek K, Constine LS, Rosier R. Sarcoma therapy: Func-

tional outcome and relationship to treatment parameters. Int J Radiat Oncol Biol Phys 1992;24:651–656. 5. O’Sullivan B, Davis AM, Turcotte R, et al. Preoperative versus postoperative radiotherapy in soft-tissue sarcoma of the limbs: A randomized trial. Lancet 2002;359:2235–2241. 6. Davis AM, O’Sullivan B, Bell RS, et al. Function and health status outcomes in a randomized trial comparing preoperative and postoperative radiotherapy in extremity soft tissue sarcoma. J Clin Oncol 2002;20:4472–4477. 7. Pisters P, Harrison L, Leung D, et al. Long-term results of prospective randomized trial of adjuvant brachytherapy in soft tissue sarcomas. J Clin Oncol 1996;14:859–868.

480

I. J. Radiation Oncology

● Biology ● Physics

8. Alektiar K, Leung D, Zelefsky M, et al. Adjuvant brachytherapy for primary high-grade soft tissue sarcoma of the extremity. Ann Surg Oncol 2002;9:48–56. 9. Schray M, Gunderson M, Sim F, et al. Soft tissue sarcoma, integration of brachytherapy, resection and external irradiation. Cancer 1990;66:451–456. 10. Alekhteyar K, Leung D, Brennan M, et al. The effect of combined external beam radiotherapy and brachytherapy on local control and wound complications in patients with high-grade soft tissue sarcomas of the extremity with positive microscopic margins. Int J Radiat Oncol Biol Phys 1996;36:321–324. 11. Suit H, Mankin H, Wood W, et al. Treatment of the patient with stage M0 soft tissue sarcoma. J Clin Oncol 1988;6:854–862. 12. Barkley HT, Jr, Martin RG, Romsdahl MM, et al. Treatment of soft tissue sarcomas by preoperative irradiation and conservative surgical resection. Int J Radiat Oncol Biol Phys 1988;14:693–699.

Volume 59, Number 2, 2004

13. Alektiar KM, Leung D, Zelefsky MJ, et al. Adjuvant radiation for stage II-B soft tissue sarcoma of the extremity. J Clin Oncol 2002;20:1643–1650. 14. Gherlinzoni F, Pignatti G, Fontana M, et al. Soft tissue sarcomas: The experience at the Istituto Ortopedico Rizzoli. Chir Organi Mov 1990;75:150–154. 15. Ravaud A, Bui NB, Coindre JM, et al. Adjuvant chemotherapy with Cyvadic in high risk soft tissue sarcoma: A randomized prospective trial. Adjuv Ther Cancer 1990;6: 556–566. 16. Frustaci S, Gherlinzoni F, De Paoli A, et al. Adjuvant chemotherapy for adult soft tissue sarcomas of the extremities and girdles: Results of the Italian randomized cooperative trial. J Clin Oncol 2001;19:1238–1247. 17. Spiro IJ, Gebhrdt M, Springfield D, et al. Neoadjuvant chemotherapy and radiotherapy for large soft tissue sarcomas [Abstract]. Proc Am Soc Clin Oncol 1996;15:524.