Postoperative radiotherapy in the treatment of extremity soft tissue sarcomas

Copyright

03hO-3016/90 $3.00 + .olJ ‘i. 1990 Pergamon Press plc

0 Original Contribution POSTOPERATIVE RADIOTHERAPY IN THE TREATMENT OF EXTREMITY SOFT TISSUE SARCOMAS WILLIAM

J. PAO

AND MILJENKO

V. PILEPICH

Radiation Oncology Center. Mallinckrodt Institute of Radiology. Washington University School of Medicine, St. Louis. MO 63 110 From 1966 to 1983, 50 patients with extremity soft tissue sarcomas were treated with wide local excision and postoperative radiotherapy at the Mallinckrodt Institute of Radiology. The median follow-up was 70 months (range 28 to 168). Grade was the most significant factor affecting survival: all 11 patients with well differentiated tumor survived versus 6/S patients with moderate and 17/31 patients with poorly differentiated tumors (p < 0.01). In addition, lymph node involvement at diagnosis conferred a worse prognosis with only 2/5 patients alive after treatment (p < 0.05). Eleven of 50 (22%) failed locally. Factors affecting local control included gross residual tumor after operation and limited treatment volume. Among the 35 patients who did not have gross residual tumor or limited treatment volume, two patients who received less than 5000 cCy failed locally versus l/18 patients who received between 5000-6000 cGy and 2/15 patients who received more than 6000 cCy. Microscopically positive margins and a volume encompassing less than the total muscular compartment was not associated with an increase in the incidence of local failure. Eight patients developed local complication: five due to retreatment for local recurrence. Overall, 24/26 patients who are alive have had their limbs preserved with normal function. Soft tissue sarcoma. Extremity sarcoma, Limb sparing therapy, Radiation therapy

There were 30 male and 20 female patients. Their ages ranged from 18 to 9 1 with a median of 52 years; 42 were white and 8 were black. All patients’ medical records. diagnostic imaging, and therapy films were reviewed. Histo-pathology and grade were institutionally reviewed at referral or diagnosis and retained for the current analysis. Tumor size denotes the maximal diameter of tumor at presentation. All patients underwent wide local excision: none had a second reexcision for minimal (~1 cm) gross residual or microscopic margins prior to radiotherapy. All specimens were linked according to an established Washington University gross pathology section protocol. Gross residual tumor denotes tumor (usually attached to a critical structure, i.e., major vessel or nerves) that was noted to have been left in siru by the operating surgeon. Microscopic disease denotes tumor present at/or within 1 mm of the cut margin. All patients were staged in the 1988 American Joint Commission (AJC) system (I). Irradiation was begun 4 to 6 weeks postoperatively. Parallel opposed tangential fields were used in most patients with 4 or 6 MV photons. The initial treatment field was decreased to a final boost field after 4500 to 5000 cGy if the total dose exceeded 5000 cGy. Treatment volumes were retrospectively reviewed and coded as com-

INTRODUCTION In the past 2 decades, radiotherapy has established a role in the multi-disciplinary management of soft tissue sarcomas of the extremities (8,9. 11, 13, 14, 15). When combined with wide local excision, local control rates of SO90% were achieved while preserving useful function of the limbs. Most authors agree that a dose of at least 6000 to 6500 cGy is required for post-operative control of extremity sarcomas. However, the volume of treatment, and surgical and pathologic factors have been difficult to define in the management of these tumors. We present here an analysis of the results of postoperative irradiation focusing on the above factors.

METHODS

AND

MATERIALS

From I966 to 1983. 164 patients with soft tissue sarcomas were seen in consultation at the Mallinckrodt Institute of Radiology (MIR), Washington University School of Medicine. Ninety-six patients completed a course of curative radiotherapy, of whom 50 previously untreated patients received post-operative radiotherapy after wide local excision for soft tissue sarcomas of the extremities.

Accepted

Reprint requests to: William J. Pao, Dept. of Radiation Oncology, St. Luke’s Medical Center, 2900 W. Oklahoma Ave., Milwaukee, WI 532 15. 907

for publication

26 April 1990

908

I. J. Radiation Oncology

0 Biology 0 Physics

partmental, subcompartmental, and limited. Volume was assessed as compartmental if the entire compartment was encompassed. Volume was assessed as subcompartmental if a margin of at least 5 cm was measured from the tumor bed distally and proximally; in addition, the entire muscular compartment had to be included in the transverse margin (tangential margins ofthe field). Limited treatment volume indicates that either the proximal/distal or tangential margins did not satisfy these criteria. All patients were treated with external beam only: brachytherapy was not used at MIR for the treatment of sarcomas. Eight patients received chemotherapy. All were Doxorubicin containing regimens. Seven of eight patients also received cyclophosphamide. Other agents used included vincristine, DTIC, and methotrexate. Three to nine cycles were delivered after radiotherapy was completed. Patients were seen in the follow-up clinic 4-6 weeks post therapy, then every 3 months for the first 3 years. then yearly. No patients in this group was lost to followup. Survival was calculated from the date of surgery. Statistics were performed using SPSS (Statistical Package for Social Sciences).

RESULTS

Patient charucteristiu The majority of patients referred for radiotherapy had high grade (poorly differentiated) and/or large tumors. It was the standard of care to use surgery alone in small and low grade tumors at Washington University Medical Center up to the mid 1970’s. Most of the patients in the series have either moderately or poorly differentiated tumors (Table 1). Histologically, liposarcomas and malignant fibrous histiocytomas constituted 60% of the patients (Table 2). The lower extremity, particularly the thigh, was the more frequent site of occurrence (Table 3) as consistent with other series (4, 7, 8, 1 I. 14).

Table

I. Tumor histologic

subtype grade at presentation

October 1990. Volume

19. Number

Table 2. Tumor

characteristics:

subtype

Malignant fibrous histiocytoma Liposarcoma Synovial sarcoma Rhabdomysarcoma Fibrosarcoma Pleomorphic sarcoma Extraskeletal Ewings sarcoma Leiomyosarcoma Malignant schwanoma Clear cell sarcoma Malignant mesenchymoma Unclassified Totals

Well

Well differentiated Size Less than 5 cm 5 to 10 cm More than 10 cm

Moderately differentiated

3 4

4 2

15 12

4

2

4

Table 3. Tumor

characteristics:

1 9

Total

14 6 2 1 1 1 1

19 12 6 2 1 1 1 1

1

I

1

1 3

1 4

8

31

50

4 3

1

11

site versus stage

American committee Poor

Poorly differentiated

Twenty-two patients failed either locally or with distant metastasis. All four patients who failed only locally were salvaged by re-excision and/or re-irradiation; all are alive free of disease 52 to 137 months from the time of salvage. Seven patients failed both locally and with distant metastasis. Eleven patients failed with distant disease alone. Except for one patient, all patients who failed distantly are dead with disease at the time of this report. The exceptional patient failed locally, in the inguinal node region and the lung. He had a complete response to cyclophosphamide and doxorubicin with local re-excision and reirradiation: he was alive free of disease at time of analysis at 13 months post salvage treatment. None of the 11 patients with well differentiated tumors developed distant metastases versus 3/8 (37%) patients with moderately differentiated, and I5/3 1 (48%) patients with poorly differentiated tumors (Chi-square (6); p < 0.01). In patients with moderate or poorly differentiated tumors, size was also a predictor of distant metastases: 6119 (3 1%) for tumor less than 5 cm. 8/14 (57%) for tumors 5 to 10 cm, and 4/6 (66%) in tumors greater than 10 cm. Three patients failed in the regional lymph nodes; all had local and distant failure. two had malignant fibrous histiocytomas, and one had a synovial sarcoma. Eighteen patients

versus

Moderate

size versus grade

Grade

Grade Histologic

4

Site Axilla Shoulder (deltoid) Arm Elbow Forearm wrist Buttocks (gluteus m.) Inguinal region Thigh Knee Leg Foot Total

joint stage

I

II

III

IV

Total

1 0 2 0 0 0 0 0 6 1 1 0

0 1 0 0 0 0 2 3 0 0 0

1 3 1 1 4 1 1 1 6 2 4 3

7 ; 0 0 0 0 0 0 2 0 0 0

4 4 3 2 4 1 1 3 17 3 5 3

11

7

27

5

50

I

Postoperative

radiotherapy

in soft tissue sarcoma

Table 4. Local recurrence:

patient,

0 W. J. PAO AND

treatment

factors, and outcome

Histol

Grade

Residual

RT volume

RT dose

Other failure sites

Foot Leg

MFH MFH

Poor Poor

Micro Gross

Limit Limit

6625 5000

Lung

W W

Thigh Thigh

LIP LIP

Well Mod.

Micro Gross

Comp. Limit

4500 6400

Mult.

M

W

Thigh

FBR

Poor

Micro

Sub

6500

Mult.

53 51 40

M F M

W W W

Thigh Forearm Axilla

LIP MFH LIP

Well Poor Well

Gross Micro Micro

Comp. Comp. Comp.

5000 4500 6400

Mult. -

67

M

W

Leg

MFH

Poor

None

Sub.

5600

Lung

24 27

M F

W B

Axilla Knee

ccs SYN

Poor Poor

Micro Gross

Comp. Comp.

5600 6400

Lung Mult.

Age

Sex

Race

61 58

M M

W B

39 51

M M

53

Site

909

M. V. PILEPICH

Treatment Amp. Amp. + Cyc/ Adr Ext. + 50 Gy. None 50 Gy. f Cyc/Adr Excision Amp. Excision + 50 GY. Excision + Cyc/Adr Excision Amp.

Outcome NED 25 mo DOD 55 mo NED 82 mo DOD 1 mo Edema DODllmo Fibrosis NED 102 mo DOD 10 mo NED 60 mo NED 13mo Edema Amp. DOD 18 mo DOD 14 mo

Histology: MFH = Malignant fibrous histiocytoma; LIP = Liposarcoma: FBR = Fibrosarcoma: CCS = Clear cell sarcoma: STN = Synovial sarcoma. RT volume: Sub = Subcompartmental; Comp = Compartmental; Limit = Limited. Site of failure: Mult. = Multiple organ systems. Treatment: Amp = Amputation: Cyc/Adr = Cyclophophamide and doxorubicin (adriamycin). Outcome: All survival times calculated and reported from the time of salvage treatment.

in distant sites: eight in the lung alone, one in bone alone, and nine in multiple organ systems include skin, distant lymph nodes, and liver. failed

Loculcontrol: putient and treatment jirctors Eleven either

patients

gross

residual

failed locally tumor

(Table

or positive

4). All patients margins

had

following

or were treated to limited volume or doses below 5000 cGy. Residual gross disease after excision appears to have a role in local failure: 4/8 patients with gross residual failed locally versus 6/3 I with microscopic positive margins and l/ 10 patients with negative margins. The difference is not statistically significant (Chi-square: a < 0.09); other factors including grade, site. and stage were not significant in univariate analysis. Since this study analyzes results of therapy over a 1&year time span, a wide variety of treatment doses and volumes were used. Patients received a minimum of 4500 cGy to a maximum of 6900 cGy (mean = 6000) (Table 5). Four of the nine patients who were treated to limited volume failed locally: this is contrasted with l/l0 patients who were treated to subcompartmental and 6/3 1 who were treated to whole compartmental volumes (p < 0.09). To determine if there was a relationship between dose and local control, patients with gross residual after surgery or those who were treated to limited volume were analyzed separately. Table 6 shows the incidence of failure in three dose ranges: less than 5000 cGy, 5000 cGy to 6000 cGy. and greater than 6000 cGy, stratified by grade. Both patients who received less than 5000 cGy failed locally. The difference in local failure between patients who received 5000 to 6000 cGy and resection,

~6000 cGy were not significant. size did not show any significant

Similar stratification difference.

by

Survival Our overall results are similar to those of other larger series ( 13- 15). At the time of analysis, 26/50 (52%) of the patients were alive continuously without recurrence. Five patients are alive free of disease after salvage treatment for recurrent disease. Three are alive with disease. Two patients are dead due to intercurrent disease: one of stroke at 20 months and the other due to a myocardial infarction at 54 months. Fourteen patients died due to recurrent soft tissue sarcoma. The median follow-up in this series of patients is 70 months (range 28 to 168 months). Grade was the major determinant of survival. Figure 1 illustrates the actuarial survival by grade. Patients with well differentiated tumors had significantly better survival than their counterparts with moderate or poorly differentiated histology. Survival rates between moderate or poorly differentiated tumor patients were not significantly

Table 5. Soft tissue sarcomas of the extremities radiotherapy dose Radiotherapy

dose range (cGy)

4500-4999 5000-5499 5500-5999 6000-6499 6500-6800

postoperative

No. failed/no. 212 219 2/g 3120 2/l 1

treated

910

I. J. Radiation Oncology 0 Biology 0 Physics

October 1990, Volume

Table 6. Local control: grade versus radiotherapeutic dose (gross residual disease and limited volume excluded)

Grade I :! 7

5000-6000

O/I

O/2

.80000 .70000 -

>6000 cGy

414 414 9110

O/I

Tdial

cGy

l7/18

Total

314 211 819

719 616 17/X)

13/15

30135

E .60000 z &

I

\ 1’

.90000 -

No. with local control/ no. treated

RT dose <5000 cGy

1.00000

19, Number 4 I

\! \.\\ \:+.L.I.

\ ‘\ ‘~Llll‘\ \,

s\ ‘$+

.50000 -

(Moderate)

b

I

CL.40000 -

(well)

‘I

.30000 .20000 .10000 I 5

.ooooo

different. Figure 2 illustrates the disease-free survival by grade. Note that patients with well differentiated tumors tend to have late failures, usually local, which were salvaged. Other patient or treatment parameters including gender, age, site (proximal versus distal), size, and residual disease (gross vs. microscopic) were not significant as a predictor of survival in our series. Stage was also a significant predictor of survival, being closely associated with grade. Patients with State IV disease (i.e.. positive regional lymph-nodes) had significantly worse survival when factored out of grade (Fig. 3).

Eight patients developed complications of therapy. Five of the patients developed complication secondary to retreatment of local recurrence (Table 4). Three patients developed complications of therapy from their initial and only course of therapy: two had edema of the extremities after 5000 and 5400 cCy postoperatively. and one patient developed superficial necrosis of soft tissue after 6600 cGy. None of the above three patients required hospitalization and all had useful function of their limbs. Five patients required amputation: four for recurrent disease. and one for complications secondary to retreatment. Twenty-four

0

Years Fig. 2. NED survival by grade. Many well differentiated tumors failed locally and were ultimately salvaged (see Fig. I).

of the 36 patients (92%) alive without disease at the time of analysis had satisfactory function of their limbs: 45/50 (90%) patients in this study had their limbs preserved. DISCUSSION This relatively unselected population of patients with soft tissue sarcomas parallels the findings of others. Grade is the most significant prognostic factor for survival (7. 8, 14). Histologically proven nodal involvement was also shown to be a predictor of survival. This has also been shown to affect survival at Memorial Cancer Center adversely (4). The incidence of nodal involvement at presentation in our series (IO%) is slightly higher than in the series reported by Lindberg (3.9%) (8) or by Suit A utl. ( 1.3%) ( 14). The size of the tumor at presentation is also a prognostic factor for survival as demonstrated by series from M. D. Anderson Hospital (4) the National Cancer Institute (NCI) (11) and Massachusetts General Hospital ( 14). Although our data show a trend in increasing distant

\ <> ’

1.00000

.90000

PC

.80000

.90000 -

>

.80000 ,70000 _

.60000

‘i~l.J.__.__

1..

J

I I

II I

I I

li (Stage

I)

01

.70000 n.

I 15

(well)

1.00000

E

I 10

(poor) E .60000 -

; :::. \\ “‘>\% ~;~,+&j+!.(StW 11) :u~_1_________.____l___________~ (stage 111) ‘:

; .50000 -

:,

.40000 -

:

a

P < .05

I

(StageI”) >

.30000 .20000 -

.ooooo1 0

I

I

I

5

10

15

Years Fig. 1. Actuarial survival by grade. Note that the survival of patients with well differentiated tumors is significantly better than those with moderately or poorly differentiated lesions. (Lee Desu Statistic (5))

.10000 .ooooo 0

I 10

I 5

I 15

Years Fig. 3. Actuarial survival by stage. The patients who had regional lymph node involvement (Stage IVA) had significantly worse survival than those who did not. (Lee Desu Statistic (5).)

911

Postoperative radiotherapy in soft tissue sarcoma 0 W. J. PAO AND M. V. PILEPICH

metastases with size, our small numbers do not allow enough statistical power for the data to be significantly different. Local control in our series was acceptable (78%). Our data suggest that gross residual disease has a detrimental effect on local control. On the other hand, true microscopic margins may not be a significant factor in local recurrence. This was the finding in the multivariate analysis of extremity soft tissue sarcomas in the report from the NC1 (I 1) and also from the univariate analysis from MGH ( 14). Our data show a trend towards a higher recurrence rate toward patients who do have microscopic residual disease, but similar to the above studies, this was not statistically significant. With gross residual disease, a re-excision should be attempted if at all possible. In addition, afterloading brachytherapy catheters can be placed at the time of second excision or if there is disease attached to bone, major vessels or nerves. This approach has been selectively and successfully utilized at Memorial Sloan-Kettering Cancer Center (2). In a randomized trial, they were able to demonstrate a reduction of local failure from 18% to 5% or less with the addition of brachytherapy. The majority of the larger modern series reported their results with postoperative doses of 6000 cGy or more to the tumor bed. Our limited data, when corrected for volume, show that local control can be achieved with a dose range of 5000 to 6000 cGy when margins of resection are involved only microscopically. Only 1 of 18 patients treated in the above dose range failed locally. Our local control with doses of greater than 6000 cGy was as fa-

vorable. In the series from the University of California, San Francisco, Leibel et al. (7) reported only 3 patients out of 29 who failed locally after local or en-bloc resection treated to 5000 to 7000 cGy. Most of the patients from the UCSF series received 6000 cGy or greater. A “wide local excision” can vary with each institution and surgeon, and may be 1 to 10 cm of margin of normal tissue depending on the site of disease. Pathologic sectioning protocols and interpretation also vary widely. Several studies have shown concordant rates between pathologists to be 60 to 70% (3, 12). Due to the above factors, a negative or microscopic margin can represent different levels of tumor burden left behind following surgery. We do not recommend a dose of lower than 6000 cGy to most operative sarcoma sites after conservative or en-bloc local excision. We believe that the above data show that a field reduction after 5000 cGy can be safely accomplished. Volume may be a more important factor than the current dose range that is used to treat extremity soft tissue sarcomas. Our data suggest that the entire muscular compartment should be included tangentially and that the entire muscular compartment need not be treated longitudinally if the proximal or distal margins are at least 5 cm within the muscular compartment. This is especially true today with good preoperative imaging by CT and MRI (10, 16). Nevertheless, successful therapy is seldom accomplished without multimodality input in the original treatment plan. Attention to details such as intraoperative placement blocking, to limb

of clips at the margins and

careful

preservation

treatment in extremity

of resection, planning, soft tissue

customized all contribute sarcomas.

REFERENCES 1. American Joint Committee on Cancer. Manual for staging of cancer, 3rd edition. Philadelphia: J. B. Lippincott Co.: 1988:127-129. 2. Brennan, M. F.: Hilaris, B.; Shui. M. H.; Lane, J.: McGill, G.; Friedrich, C.; Hadju. S. I. Local recurrence in soft tissue sarcoma. A randomized trial of brachytherapy. Arch. Surg. 122:1289-1293; 1987. 3. Coindre, J. M.; Trojani, M.; Contesso. G.: David, M.; Rouesse J.; Bui, N. B.; Bodaert, A.: Mascarel, I. D.: Mascarel, A.: Goussot. J. Reproducibility of a Histopathologic Grading system for adult soft tissue sarcomas. Cancer 58:306-309: 1986. 4. Collin, C. F.: Friedrich. C.: Godbold. J.; Hadju, S.: Brennan, M. F. Prognostic factors for local recurrence and survival in patients with localized extremity soft tissue sarcoma. Sem. Surg. Oncol. 4:30-37: 1988. 5. Lee, E.; Desu, M. A computer program for comparing K samples with right-censored data. Comp. Prog. Biomed. 2: 315-321: 1972. 6. Fisher, R. A. The use of multiple measurements in taxanomic problems. Ann. Eugenias 7: 179-l 88; 1936. 7. Leibel, S. A.; Tranbaugh. R. F.; Wara, W. M.: Beckstead. J. H.; Bovill, E. G.: Phillips, T. L. Soft tissue sarcomas of the extremities: survival and patterns of failure with conservative surgery and postoperative irradiation compared to surgery alone. Cancer 50:1076-1083: 1982. 8. Lindberg, R. Treatment of localized soft tissue sarcomas in adults at M. D. Anderson Hospital and Tumor Institute (1960-1980). Cancer Treat. Symp. 3:59-65; 1985.

9. Lindberg, R. D.: Martin, R. G.; Rhomsdahl, M. M.: Barkley, H. T. Conservative surgery and postoperative radiotherapy in 300 adults with soft-tissue sarcomas. Cancer 47:23912397; 1981. 10. Neifeld, J. P.: Walsh, J. W.; Lawrence. W. Computed Tomography in the management of soft tissue tumors. Surg. Obst. Gyn. 155:535-540; 1982. 11. Potter, D.: Kinsella, T.: Glatstein. E.; Wesley, T.: White, D. E.: Siepp, C. A.: Chang, A. E.; Lack, E.; Costa, J.; Rosenberg. S. A. High grade soft tissue sarcomas of the extremities. Cancer 58: 190-205; 1986. 12. Presant, C. A.: Russell, W. 0.; Alexander, R. W.; Fu, Y. S. Soft tissue and bone histopathology peer review: the frequency of disagreement in diagnosis and the need for second pathology opinions. The Southeastern Cancer Study Group experience. J. Clin. Oncol. 4:1658-1661: 1986. 13. Rosenberg. S. A.: Tepper. J.; Glatstein. E.; Costa, J.: Baker, A.: Brennan, M. F.; DeMoss, E. V.; Siepp, C.: Sindelar, W. F.: Sugarbaker, P.: Wesley, R. The treatment of soft tissue sarcomas ofthe extremities. Ann. Surg. 196:305-3 14; 1982. 14. Suit, H. D.; Mankin, H. J.; Wood, W. C.; Gebhardt, M. C.; Harmon, D. C.; Rosenberg, S. A.; Tepper. J. E.: Rosenthal, D. Treatment of the patient with stage MO soft tissue sarcoma. J. Clin. Oncol. 6:854-862; 1988. 15. Suit. H. D.; Russell, W. D.; Martin, R. D. Sarcomas of soft tissue: clinical and histologic parameters and response to treatment. Cancer 35:1478-1483: 1975. 16. Tatty, W. G.: Murphy, W. A.: Lee. J. K. T. Soft tissue tumors: MR imaging. Radiology 160: 135- 14 1: 1986.