- Email: [email protected]
Soft tissue sarcomas of the chest wall
J
THORAC CARDIOVASC SURG
1991;101:843-54
Soft tissue sarcomas of the chest wall Results ofsurgical resection Primary soft tissue sarcomas of the chest wall are uncommon, and data concerning treatment and results are sparse. Most studies have categorized these tumors as truncal sarcomas and inferred a poor prognosis. To assess the results of surgical treatment, we reviewed our 4O-year experience. Methods: Records of 189 patients admitted to our institution from 1948 to 1988 were reviewed. Pathologic material was available for review in the 149 cases (79%) that comprise this report. Survival was calculated by the Kaplan-Meier method, with compariso~ by log-rank analysis and significance defined as p < 0.05. Results: Ages ranged from 3 weeks to 86 years (median, 38 years); the ratio of male to female patients was 2:1. The initial complaint was mass or pain in 97% of the cases. Ninety sarcomas (60%) were high grade and 59 (40%) were low grade. Histologic types were as follows: desmoid tumor (n = 32, 21 %); liposarcoma (n = 23, 15 %); rhabdomyosarcoma (n = 18, 12 %); fibrosarcoma (n = 17, 11 %); embryonal rhabdomyosarcoma (n = 14, 9%); malignant peripheral nerve tumor (n = 13, 9%~ malignant fibrolti histiocytoma (n = 11,7%); spindle ceU sarcoma (n = 4,3%); tenosynovial sarcoma (n = 3, 3%); hemangiopericytoma (n = 3, 3% ~ alveolar soft part sarcoma (n = 3, 3% ~ and other types (n = 12, 9%). Resection was the primary treatment in 140 cases (94%). Local recurrence developed in 27%. Metastases occurred in 52 (35%) of the cases (metachronOiti in 42, synchrOOOIti in 10) and were more common in patients with high-grade disease (46/90, 51 %) than in those with low-grade disease (6/59, 10%~ Overall S-year survival was 66%. Five-year survival rate for those with high-grade sarcomas (49%) was significantly lower than that for low-grade sarcomas (90%, p < 0.0001). Tumor size and age of patient were not prognostic. Conclusions: Survival of patients with primary soft tissue sarcomas of the chest wall after resection is similar to that of patients with sarcomas of the extremities. Resection alone provides acceptable survival (90% ai5 years) for those with low-grade sarcomas, but adjuvant treatment should be co~idered for those with high-grade sarcomas.
Mark S. Gordon, MDa (by invitation), Steven I. Hajdu, MDb (by invitation), Manjit S. Bains, MD,a and Michael E. Burt, MD, PhDa (by invitation), New York, N.Y.
Emary soft tissue sarcomas of the chest wall are uncommon. Of the 5000 new cases of soft tissue sarcoma diagnosed annually in the United States, approximately 20% occur in the abdominal or chest wall.!" Thus few
From The Thoracic Service,' Department of Surgery and The Department of Pathology," Memorial Sloan-Kettering Cancer Center, New York, N.Y. Read at the Seventieth Annual Meeting of The American Association for Thoracic Surgery, Toronto, Ontario, Canada, May 7-9,1990. Address for reprints: Michael E. Burt, MD, PhD, Department of Surgery, Memorial Sloan-Kettering Cancer Center, 1275 York Ave., New York, NY 10021.
12/6/26710
centers have acquired a large experience with these tumors. With the advent of reconstructive techniques of the chest wall, extensive tumors have become technically resectable with minimal morbidity." However, data to determine the results of treatment for soft tissue sarcomas arising from the chest wall are sparse and usually involve small numbers.l The recent literature includes analyses of primary chest-wall soft tissue sarcomas in conjunction with chest-wall sarcomas of either bone or connective tissue origin.f 7 soft tissue sarcomas metastatic to the chest wall," or soft tissue sarcomas of the trunk? To further characterize primary chest-wall soft tissue sarcomas and to evaluate results of treatment, we con843
The Journal of
844
Gordon et ai.
Thoracic and Cardiovascular Surgery
Proportion Surviving
1
flti-tt4r i
i
0.8
0.6
1Hlll l ! 1thlHh
1
j
i'I liq j
1
,I
0.4 ,
0.2
o o
10
20
30
40
50
60
70
80
90
100
110
120
Time (rnos)
Fig. 1. Overall survival in 149 patients with chest-wall soft tissue sarcomas. The overall actuarial survival rates were 66% at 5 years and 56% at 10 years. Proportion Surviving
1ft;''-... ---~K-"He
-%
----
-------_._----
+---+-~.'
f..-++-"
't,,,,,
081 0.6
~+-
"t1'c
"-----+~
(p-O.0001)
L
;-j +--------
0.4 ..
0.2
.
I
--+-
- +- 1t
High Grade (n-90)
---+- low Grade (n-59)
o '----_L o 10
_
___
20
.~
~
L
30
40
50
_
60
70
Time (rnos)
80
90
100
110
120
Fig. 2. Actuarial curves of overall survival stratified by tumor grade. Patients with high-grade tumors had 5- and 10-year survival rates of 49% and 39%, respectively, compared to 5- and 10-year survival rates of 90% and 82%, respectively, among patients with low-grade tumors.
ducted a retrospective study of a 40-year experience at Memorial Sloan-Kettering Cancer Center (MSKCC). Patients and methods A computer search of the medical records at our institution from Jan. 1, 1948, to Dec. 31, 1988, yielded the charts of 189 patients with soft tissue sarcomas arising in the chest wall. This represented 5.5% of all soft tissue sarcomas (n = 3417) seen at our institution during this period. We defined primary soft tissue sarcomas of the chest wall as those arising between the clavicles and costochondral cartilages anteriorly, the glenoid fossa
and costochondral junctions laterally, and spinous processes of the scapulae and costochondral borders posteriorly. Patients with tumors of bony or cartilaginous origin (such as Ewing's sarcoma, osteosarcoma, and chondrosarcoma) were not included. All pathologic material was reviewed by one of us (S.I.H .). Patients were excluded from further study if the diagnosis of soft tissue sarcoma could not be confirmed by reevaluation of the original pathologic material. Of the 189 patients identified, pathologic material was available for review in 149 cases (79%), and these comprise the basis of this analysis. Histologic classification of sarcomas was predominantly based on sections stained with hematoxylin and eosin, as previ-
Volume 101 Number 5 May 1991
Chest-wall soft tissue sarcomas 8 4 5
Table I. Histologic types of chest-wall soft tissue sarcomas: Average age ofpatient, gradeof tumor, and distribution by sex No. by grade Histologic type
No.
32 23 18 17 14
Desmoid Liposarcoma Rhabdomyosarcoma Fibrosarcoma Embryonal rhabdomyosarcoma Malignant peripheral nerve tumor Malignant fibrous histiocytoma Spindle cell sarcoma Tenosynovial sarcoma Hemangiopericytoma Alveolar soft-part sarcoma Dermatofibrosarcoma protuberans Leiomyosarcoma Primitive neuroectodermal tumor Giant cell sarcoma Unclassified Total
13 II
4 3 3 3 2 2 I I
--..1.
149
%
Low
High
21 15 12 11 9 9 7
32 16 0 1 0 4
3
2 0 1 0 2 0 0 0
0 7 18 16 14 9 10 2 3 2
I
3 3 3
1 I I I
1
3
0 2 1 I
-l!
..1 90
59
Mean age (yr)*
No. male
33 56 44 44 17 49 44 45 37 38 20 52 62 16 52 33 41
13 20 13 12 7 10
8 3
3 3 2 1 2 0 I
--..1.
100
No. female
19 3 5 5 7 3 3
1 0 0 1 I
0 I
0
.n
49
'Overall median age. 38 years.
Table II. Chest-wall soft tissue sarcomas: Patients previously undergoing radiation therapy Previous malignancy Breast carcinoma Hodgkin's disease Breast carcinoma Lymphoma Hodgkin's disease Hodgkin's disease Breast carcinoma
Dose (cGy)
Interval between RT and sarcoma (yr)
Unknown
4 4 13 3 3
1250 3920 3588 2000 6000 4600*
8
10
Chest wall sarcoma Histologic findings
Grade
Desmoid tumor Malignant fibrous histiocytoma Fibrosarcoma Liposarcoma Malignant fibrous histiocytoma Malignant fibrous histiocytoma Malignant peripheral nerve tumor
Low High High Low High High Low
RT, Radiation therapy. 'Skin dose was 6440 cGy (cobalt 60 machine).
ously reported.!? Tissue was designated as unclassified when adequate material was available but when a definitive final classification could not be made. All sarcomas were classified as low- or high-gradeon the basisof cellular differentiation,degree of tumor cellularity and vascularity, amount of stroma and necrosis, and the number of mitoses per 10 high-power microscopicfields. IO, 11 Local recurrence was defined as the appearance of sarcoma in or adjacent to the site of previousdefinitivetherapy. Recurrent sarcoma was classified as metastatic only if it was found distinctlyremovedfrom the site of the primary tumor. Diseasefree intervalwasdefinedas time betweendefinitivetherapy and local recurrence or metastasis. The method of Kaplan-Meier was used to determine actuarialsurvival. Differences in survivalwere determined by log-rank analysis, and differences in local recurrence were determined by Fisher's exact test; significant differences were defined by p <0.05.
Results Age and sex. The patient population ranged in age from 3 weeks to 86 years. The mean and median ages at diagnosis were 41 years and 38 years, respectively. The prevalence by age was bimodal in distribution, with peaks at the third and sixth decades. Compared with ages of patients with sarcomas ofother histologic types, the mean age at diagnosis was lowest for patients with embryonal rhabdomyosarcoma (mean, 17 years) and highest for patients with liposarcoma (mean, 56 years) (Table I). One hundred patients (67%) were male and 49 (33%) were female (a ratio of 2:1). With the exception of desmoid tumors (with a ratio of female to male patients of 1.5:1), all histologic types occurred more commonly in male patients, with ratios ranging from 2:1 to 3:1. Male
The Journal of
8 4 6 Gordon et al.
Thoracic and Cardiovascular Surgery
Proportion Surviving
1~Ir;-------:--c
1 I
(p~O.02) iffi---
-~
Desmoid (n-32)
I
Low Grade (n-27) -D
oI
o
High Grade (n-gO)
_L
10
20
30
40
50
60
70
Time (rnos)
80
90
100
110
120
Fig. 3. Actuarial survival curves compare desmoid tumors (n = 32), low-grade sarcomas other than desmoid tumors (n = 27), and high-grade sarcomas (n = 90). Overall survival rates of patients with low-grade sarcomas other than desmoid tumors (5-year rate, 83%; IO-year rate, 72%) were significantly different from overall survival rates of those with high-grade sarcomas. The survival rate for patients with desmoid tumors tended to be higher than that for those with the other low-grade sarcomas, but this difference did not reach significance (p = 0.07).
Table III. Chest-wall soft tissue sarcomas (n = 149): Treatment based on extent of disease at initial examination No.
Local disease only Resection' Radiation therapy + chemotherapy Radiation therapy only Synchronous metastases Resection of primary tumor and metastases' Radiation therapy to primary tumor + chemotherapy Radiation therapy to primary tumor Chemotherapy
139 137 I I 10 3 4 1 2
'One hundred forty patients underwent resection or the primary tumor as initial therapy.
predominance was particularly notable for liposarcomas, where the ratio of male to female patients was 7:1. Clinical presentation. The initial symptom was a nonpainful mass in 109 patients (73%), a painful mass in 20 (13%), and pain alone in 17 (11%). In two patients the initial symptom was not stated, and one patient had no symptoms but had a mass seen on routine chest radiography. Only seven patients (5%) reported weight loss when seeking treatment. Of the 118 patients with documented duration of symptoms, the mean was 12 months. Seven patients had received previous radiotherapy to the chest wall for a malignant neoplasm (Table 11). Three patients had previously been treated for breast cancer,
three for Hodgkin's disease, and one for lymphoma. In one patient the radiation dose was unknown, but in the other six patients the mean dose was 3560 rad, with a range of 1250 to 6000 rad. The mean and median times from radiotherapy to the diagnosis of soft tissue sarcoma were 6.4 years and 4 years, respectively. High-grade malignant fibrous histiocytoma was the most common sarcoma, occurring in three patients (45%). Among the 32 patients with desmoid tumors, four also had Gardner's syndrome (two male and two female patients, with a mean age of 21 years). Three of these patients subsequently died of complications related to Gardner's syndrome: Adenocarcinoma of the duodenum developed in one, extensive intraabdominal desmoid tumor with small bowel necrosis and sepsis developed in one, and liver failure developed in one after multiple resections for intraabdominal desmoid tumor. Of the 13 patients with malignant peripheral nerve tumors, one patient had von Recklinghausen's disease (multiple neurofibromatosis). This patient was first seen with metastatic disease and died 3 months after the diagnosis of sarcoma. Site of primary tumor. As a group, sarcomas were not associated with any particular site of the chest wall. However, certain sites were occupied more commonly by particular sarcomas. All three tenosynovial sarcomas occurred in the interscapular area, and of the 11 sarcomas in the interscapular area, six (55%) were either tenosynovial sarcomas or fibrosarcomas. Of the 11 sarcomas in
Volume 101
Chest-wall soft tissuesarcomas 8 4 7
Number 5 May 1991
Proportion Surviving 1.-T'i¥-'r-------------------------, 0.8 0.6
(p
0.4 0.2
o " - - _ ' - - - - - _ L - _ - L -_ _--L-_--l-_--'---_---'--_----"-_--'-_---'-_ _-'_-----! o 10 20 30 40 50 60 70 80 90 100 110 120 Time (rnos) -+- Rhabdo (n-18) -8-
Embry Rhabdo (n-14)
- - Fibro (n-16) --- MFH (n-10)
Fig. 4. Actuarial survival curves for patients with the more common high-grade sarcomas, stratified according to histologic origin. Ten-year survival for patients with high-grade sarcomas originating from fibrous or connective tissue (fibrosarcoma and malignant fibrous histiocytoma) was significantly greater (p < 0.0001) than that for patients with high-grade sarcomas of muscular origin (rhabdomyosarcoma and embryonal rhabdomyosarcoma). Rhabdo, Rhabdomyosarcoma; Embry Rhabdo, embryonal rhabdomyosarcoma; Fibro, fibrosarcoma; MFH, malignant fibrous histiocytoma.
the presternal area, six (55%) were either fibrosarcomas or malignant fibrous histiocytomas. Method of diagnosis. The histologic diagnosis was made by biopsy (n = 58, incisional in 32 cases and excisional in 26) or resection (n = 84, limited in 56 cases and radical in 28). In one patient with embryonal rhabdomyosarcoma, the histologic diagnosis was made by lymph node biopsy, and in one patient who had undergone radiotherapy for Hodgkin's disease, the diagnosis was made by aspiration needle biopsy. In five patients the method of initial histologic diagnosis could not be ascertained from the medical record. Histology and grade. The most common types of tumor were desmoid tumor (21%), liposarcoma (15%), rhabdomyosarcoma (12%), and fibrosarcoma (11%) (Table I). Tumors were high grade in 90 patients (60%) and low grade in 59 (40%). As expected, all desmoid tumors (n = 32) and dermatofibrosarcoma protuberans (n = 2) were low grade, and all embryonal rhabdomyosarcomas (n = 14) were high grade. Of the 23 liposarcomas, 16 (70%) were low grade and seven (30%) were high grade. Of the remaining histologic types, most tumors (50% or more) were high grade. Therapy. Of the 149 patients with chest-wall soft tissue sarcoma, 75 (50%) underwent their initial treatment at MSKCC and 74 (50%) underwent their initial treatment at another institution. One hundred forty patients underwent resection of the primary lesion as initial treatment (Table III). Five patients were treated initially by
Table IV. Chest-wall soft tissuesarcomas: Treatment of local recurrence Therapy
No.
Surgery Surgery + radiation therapy* Surgery + radiation therapy + chemotherapy Surgery + brachytherapy Surgery + brachytherapy + radiation therapy Surgery + brachytherapy + chemotherapy Radiation therapy Radiation therapy + chemotherapy None Total
47 II 7 4 7 2 2 I
.L
82
"Radiation therapy denotes external beam.
radiation therapy and chemotherapy; four of these had distant metastases at diagnosis. Twelve patients underwent adjuvant therapy, after having undergone a complete resection. Five patients were treated with radiotherapy, six patients were treated with chemotherapy, and one patient underwent both radiotherapy and chemotherapy. After incomplete resection of the primary tumor (margins positive for disease), 17 patients were treated with radiotherapy. Of these, 11 received intraoperative radiation implants (brachytherapy). Eight patients underwent chemotherapy, combined with radiation therapy in five and without additional treatment in three. Local recurrence. Of the 140 patients who underwent
848
The Journal of Thoracic and Cardiovascular Surgery
Gordon et a/.
Table V. Chest-wall soft tissue sarcomas: Patterns of metastasis Metastases
Grade Low High Size (cm)*
<5 5-9.9 2':10 Surgical margins] Adequate Inadequate Histopathologic findings Desmoid tumor Liposarcoma Rhabdomyosarcoma Fibrosarcoma Embryonal rhabdomyosarcoma Malignant peripheral nerve tumor Malignant fibrous histiocytoma Spindle cell sarcoma Tenosynovial sarcoma Hemangiopericytoma Alveolar soft-part sarcoma Dermatofibrosarcoma protuberans Leiomyosarcoma Primitive neuroectodermal tumor Giant cell sarcoma Unclassified
Total
No.
Synchronous
Metachronous
No.
%
59 90
0 10
6 36
6 46
10 51
22 27 16
0 0 0
3 8 5
3 8 5
14 30 31
53 17
0 0
10 5
10
19 29
32 23 18 17 14
0 0 4 0 5 0 0 0
0 8 8 3 8 4 2
0 8 12 3
13 II
4 3 3 3 2
2 I I
2
I
0 0 0 0 0 0 0
5
13
4 2 2
0 35 67 18 93 31 18 50 100 67 33
2 2 2 I
3 2
0
0
0
I 0 I
I
0
50
0
0
I
I
0
50 0
'Pathologic size was only available for 65 patients (44%).
t Adequacy of margins could only be determined for the 70 patients who underwent primary operation at MSKCC.
resection as initial treatment, 82 (59%) experienced local recurrence. Sixty-three of these (77%) were initially treated at another institution and referred to MSKCC. Adequacy of resection could be determined only in the 70 patients who underwent initial resection or later resection at MSKCC. In 53 of 70 patients (76%) the margins of resection were free of tumor, and in 17 of 70 patients (24%) the resection margins were either microscopically or grossly involved with tumor. Overall local recurrence of tumors initially resected at MSKCC was 27% (19/70). Of the 53 patients with tumor-free margins of resection, 13 (24%) experienced local recurrence, compared with six (35%) of the 17 patients with involved margins (p = 0.164). Local recurrence was found in eight of the 30 patients with high-grade sarcoma who had negative margins (27%), compared with one of nine patients with high-grade sarcoma who had positive margins (II %) (p = 0.249). For low-grade sarcomas, local recurrence developed in five of 23 patients with negative margins (22%), compared with five of eight patients with positive margins (62%) (p = 0.04). Treatment of local recurrences is shown in Table IV.
Seventy-eight patients (95%) underwent resection of the local recurrence either alone (47/82,58%) or in combination with at least one other form of therapy (31/82, 38%). Metastases. Of the 149 patients with chest-wall soft tissue sarcomas, 52 (35%) experienced either synchronous or metachronous metastases (Table V). Metastases occurred in 46 out of 90 (51 %) patients with high-grade chest-wall sarcomas, as compared with six out of 59 (10%) patients with low-grade sarcomas. Metastases occurred in three out of 22 (I4%) patients with tumors less than 5 em in diameter, as compared with 13 out of 43 (30%) patients with tumors 5 cm or greater in diameter. Although adequacy of surgical margins could be determined in only 70 patients, this variable seemed unrelated to the development of metastases. Patients with desmoid tumors (n = 32) or dermatofibrosarcoma protuberans (n = 2) did not experience metastases. Of the more common histologic types (occurring in at least 10 patients), metastases were more likely to develop in patients with embryonal rhabdomyosarcoma (I3/I4, 93%) and rhabdomyosarcoma (I 21 18, 67%). Least likely to metastasize
Volume 101 Number 5
Chest-wall soft tissue sarcomas 8 4 9
May 1991
Proportion Surviving
1_c,-~~~~~~----~~-~~~~~~~~~-------,
0.8
0.6
K 0.4
0.2
-+-
No Pain (n·112)
-+-
Pain (n·37)
(p=O.042)
*
o I,--~---,--~_l.. __ ~_L~~~,--_-,--_-"-_---,--_--,-_---.J,--_----, o 10 20 30 40 50 60 70 80 90 100 110 120 Time (rnos)
Fig. 5. Actuarial survival curves stratified according to the presence or absence of pain at initial examination. Patients with pain at initial examination had a significantly decreased overall survival.
Proportion Surviving
0.6 0.4
(p=O.003)
0.2
o o
'--_L~~-'--_-'--_---'--~-"-~~~---'--~-'-~--'-~----l~---.J--,-------!
10
20
30
40
50
60
70
80
90
100
110
120
Time (rnos) -+-
Low. No Pain (n·49)
-8-
High. No Pain (n·63)
---+-
Low. Pain (n·lO) High. Pain (n·27)
Fig. 6. Actuarial survival curves stratified by grade, according to the presence or absence of pain at initial examination. Pain had a significant adverse effect on survival only among patients with high-grade sarcomas.
were fibrosarcoma and malignant fibrous histiocytoma, each of which metastasized in 18% of patients. Of the 52 patients in whom metastases developed, 28 patients (54%) experienced metastases to a single organ. The most common single site for metastases was the lung, with 20 of the 28 patients (71 %) experiencing metastases to this site. Of the remaining eight patients who experienced metastases to a single site, three (11%) experienced metastases to subcutaneous tissue in areas distinctly removed from the primary site, one (4%) experienced metastases to regional lymph nodes, and one patient each
(4%) experienced metastases to the spine, brain, liver, and retroperitoneum. Twenty-four patients (46%) experienced metastases to multiple sites, including lung, spine, bone (other than spine), pleura, brain, regional lymph nodes, retroperitoneum, and intraabdominal organs. Disease-free interval. Of the 139 patients who were initially seen with local disease, 91 (65%) experienced recurrence after initial therapy. The initial recurrence was local in 77 cases (85%), distant in 13 cases (14%), and both local and distant in one case (1 %). Disease-free
The Journal' of'
8 5 0 Gordon et al.
Thoracic and Cardiovascular Surgery
Proportion Surviving
1i&,i-;-~~~~~~~~~~~~~~--~~~~~~~_
0.8
-+-itH
-~~H''¥1'+'~+----:f,----+------jc+h
0.6 .
0.4
(p=O.89)
0.2
-+--
No Loc Recur (n'67)
-+--
Loc Recur (n'82)
O'--~-'--~-'--~~~-'--~--'---~~~~~~~~~----"-~~~~
o
10
20
30
40
50
60
70
80
90
100
110
120
Time (mas)
Fig. 7. Actuarial survival curves demonstrate the effect of local recurrence. Local recurrence had no impact on overall survival.
Proportion Surviving
~'~~~:::;::::::::J
1
-----+---
0.8
(p=O.31)
0.6
'P~l
0.4 0.2
o o
'---~L-~-'--~.L-.~..L.....~-'---~--"---~~
10
20
30
40
-+-
Low, No Rec (n·25)
-B-
High, No Rec (n·42)
50
60
70
_ _-+-~---"--~----"-~~~-
80
90
100
110
120
Time (mas) -+- Low, Rec (n'34)
High, Recur (n·48)
Fig. 8. Actuarial survival curves, stratified by grade, demonstrate the effect of local recurrence. Within each grade, local recurrence had no effect on overall survival.
intervalranged from 1 to 68 months, with a medianof 14 months. Survival. The median follow-up time in our series of 149 patients was 56 months. Overall survival was 66%at 5 years and 56%at 10 years (Fig. 1). The 5- and IOyear survivalrates for patients with low-gradesarcomas were 90%and 82%,respectively; these weresignificantly higher (p < 0.0001) than the 5- and 10-yearsurvival rates for patientswith high-gradesarcomas(49%and 39%,respectively) (Fig. 2). The 1O-year survival rate for patientswith desmoidtumors was 93%, and there was a trend suggesting that their survivalwas better than that for those with
other low-grade sarcomas (p = 0.07). The difference in survival rates between patients with the 27 other lowgrade sarcomas (5-year survival, 83%; lO-year survival, 72%) and those with high-grade sarcomas remained significant (p = 0.02) (Fig. 3). Survival was next analyzed for the more common high-gradeforms(Fig. 4). At 5 yearsthe survival rates for patients with fibrosarcoma, malignant fibrous histiocytorna, rhabdomyosarcoma, and embryonalrhabdomyosarcoma were 87%,62%,21%, and 14%, respectively. At 10 years, survival rates for patients with fibrosarcoma and malignant fibrous histiocytoma were similar, as were
Volume 101
Chest-wall soft tissue sarcomas 8 5 1
Number 5 May 1991
Proportion Surviving
1_~---------------------------,
0.8
0.6
(p
0.2
-t- Metastasis (n.52)
-+-
No Mets (n·97)
o~-~--'------'---'-------'----L-_-L_---L_--'-_---.l_---'_----.J
o
10
20
30
40
50
60
70
80
90
100
110
120
Time (mos)
Fig. 9. Actuarial curves demonstrate the effect of metastases on survival. Patients who experienced metastases had significantly decreased overall survival. Proportion Surviving
1
~F++=+=+====i;=;::=:=;=====:=l
0.8 o~
0.6 0.4 0.2 _
oL-_L--'--'-_-'-_-'----_---'---_--L-_-L_~
o
10
20
30
40
50
60
70
80
_ _ L_
90
_...L_
100
_____'_~
110
120
Time (rnos) -+- Low, No Mets (n·53) --B-
High, No Mets (n.50)
--+-
Low, Mets (n·6)
- - High, Mets (n-40)
Fig. 10. Actuarial survival curves, stratified by grade, demonstrate the effect of metastases on survival. Within each grade, metastases significantly decreased survival (low-grade, p = 0.031; high-grade, p = 0.000(1). Patients with high-grade sarcomas who did not experience metastases had overall survival similar to that of patients with low-grade sarcomas that metastasized.
rates for those with rhabdomyosarcoma and embryonal rhabdomyosarcoma. The lO-year survival rate was approximately 65% for patients with the high-grade sarcomas originating from fibrous or connective tissue, as opposed to 14% for those with the high-grade sarcomas of muscle origin (p < 0.0001). Patients whose initial symptoms included pain (n = 37) had 5- and lO-year survival rates of 48% and 37% respectively,compared with 5- and 10-year survival rates of 70% and 63%, respectively, among those patients without pain (n = 112) at initial examination (p = 0.042)
(Fig. 5). Further analysis showed this difference to be significant only for those patients with high-grade tumors; there was no survival difference between patients with low-grade tumors with and without pain at initial examination. (Fig. 6). As shown in Fig. 7, overall survival was not affected by local recurrence. When it was analyzed according to grade, local recurrence again was found to have no adverse effect on survival (Fig. 8). In the 52 patients in whom metastases developed, 5and l O-year survival rates were 27% and 19%, respec-
The Journal of .
8 5 2 Gordon et al.
Thoracic and Cardiovascular Surgery
Proportion Surviving
1ftf---+---+--------------------_
0.8
0.6
0.4
0.2
--+-
<
--+-
5-10 em (n-27)
5
em (n-22)
-B- , 10 em (n-16)
o ---'--------J-----'---....l...-_L-----L_-'---_-'---------'_----'--_....l...---.J o 10 20 30 40 50 60 70 80 90 100 110 120 ,-I
Time (mas)
Fig. 11. Actuarial curves ofoverall survival stratified bytumor size. Nosignificant difference was noted among different sizes of primary tumors. tively, as opposedto 83%and 75%among the 97 patients without metastases (p < 0.(001) (Fig. 9). For both lowand high-grade tumors, the development of metastases had a significant adverse effect on survival (Fig. 10). Interestingly, survivalwas similar for patients with highgrade sarcomas who did not experience metastases and patients with low-grade sarcomas who experienced metastases. When analyzed according to sizeof tumor, 5-year survival rate was similar for patients with tumors smaller than 5 em, tumors between 5 and 9.9 em, and tumors at least 10em in size (Fig. 11). The 5-year survival rates for patients with tumors in these three size ranges were 82%, 74%,and 63%,respectively. At 10years the survivalrates for patients with tumors in these three size intervalswere 59%for tumors less than 5 em, 74% for tumors between 5 and 9.9 em, and 63% for tumors larger than 10 em. There was no significantdifferencein survival according to size of tumor. Finally,survival was analyzed accordingto patient age at diagnosis. At 5 years the survivalrate was66%both for patients younger than 53 years at diagnosis and for patients older than 53 years at diagnosis. At 10 years 62% of patients whose diagnoses were made at less than 53 years of age were alive,as opposed to 47%of thoseat ages greater than 53 years (p = 0.07). Discussion
Data concerning treatment and results of therapy of primary soft tissue sarcomas of the chest wall are sparse. Most published series of these relatively uncommon tumors are analyzed as truncal sarcomas and as a group
do relativelypoorly. Our bias at the start of this study was that chest-wall soft tissue sarcomas in general had an outcome of therapy comparable with that of extremity soft tissue sarcomas. Our series consisted of 149 patients with primary soft tissue chest-wall sarcomas who were admitted to our institution over a 40-year period. The predominance of male patients over female patients (2:1) and our median patient age of 38 years are similar to those in other series of chest-walland extremity soft tissue sarcomas.v 12-16 Several studies of patients with chest-wall sarcoma have recently been published.o 6, 8 All of these articles have advocated aggressive chest-wall resection, with adjuvant treatment in selected cases. Prognostic factors have includedgrade, size,status of surgical margins, and metastases.o" Analysis of data from two large series of extremity soft tissue sarcoma has additionally identified age, symptoms, local recurrence, and histopathologic nature as prognostically important factors.12, 15 The overallsurvivalrates in our seriesof 66%at 5 years and 56%at 10 years compare favorably with the 51% 5year and 34% l Oeyear survival rates reported by Greager and coworkers.' and are similarto the 5-yearsurvival rate for patients with primary chest-wallsarcomasreportedby Perry and associates," There was a significant difference in survival basedon tumor grade; patients with low-grade sarcomas had a higher survival rate than had those with high-grade sarcomas. Excluding desmoid tumors from the other low-gradetumors did not alter the significance of this difference. Interestingly, 5- and 10-year survival rates (90% and 82%, respectively) for patients with lowgrade chest-wallsarcomas wereslightlyhigher than were
Volume 101 Number 5 May 1991
similarsurvival data for patients with low-gradeextremity soft tissue sarcomas (5-year rate, 82%; 10-year rate, 74%).16 Survivalwas similarfor patients with high-grade chest-wall (5-year rate, 49%; 10-year rate, 39%) and those with extremity soft tissue sarcomas (5-year rate, 47%; 10-yearrate, 37%).16 Unlike the National Cancer Institute series of highgrade extremity soft tissue sarcoma.P we found histopathologic resultsa significant prognosticindicator.At 10 years, survival for patients with high-grade tumors of muscleoriginwassignificantly lessthan that of thosewith high-grade tumors originating in fibrous or connective tissue. Althoughonly 37 patients (25%) reported pain at initial examination, these patients had decreased survival compared with those without pain at presentation. The adverseeffectof painon survival, however, wassignificant onlyfor patients with high-grade tumors. Similar results have also been reported by Collin and colleagues'? in extremity sarcomas. In a recentlypublishedseries of chest-wall sarcomas.! status of surgical margins was found to be the most importantfactor affectingsurvival. In our series, margins of resection negativefor neoplasmsignificantly decreased the likelihood of local recurrence for patients with lowgrade tumors only. For those with high-grade tumors, surgicalmargin status had no effect on local recurrence, presumably becausepatients did not livelongenoughfor local recurrence to develop. Overall survival was not adversely affected by local recurrence. Furthermore, within each grade, local recurrence had no impact on overall survival. Metastatic diseaseappeared in 52 of our 149 patients. Metastases occurred more frequently in patients with high-gradetumors (51%) than in patientswith low-grade tumors (10%). Whether synchronous or metachronous, the development of metastases had a significant adverse effecton survival. Two large seriesof extremitysoft tissuesarcoma both identified sizeof the primary lesion as a significant prognosticfactor,I2, 15 and in one ofthese series, age at initial examination was also significant.F In our analysis, survival was not influenced by size of the primary tumor. Likewise, age was not related to survival among patients with primary chest-wall soft tissue sarcoma. Resection of low-grade chest-wall soft tissuesarcomas provides acceptable long-term survival. If necessary to achieve clear margins, the chest wall, including the parietal pleura, should be resected in continuity. Although chest-wall resection may be an extensive procedure,several reportshaveshownthat it can be performedwith low morbidity and mortality.r" Techniques for chest-wall
Chest-wall soft tissue sarcomas 8 5 3
reconstruction have been well described." Adjuvant systemic therapy does not seem indicated for patients with low-grade chest-wall sarcomas, given the 90% 5-year survival rate and the 82% 10-year survival rate after resection. However, prospective trials should be directed at decreasing the 22% local recurrence rate associated with adequate widelocal resectionalone in patients with low-grade sarcomas. Becausehigh-gradechest-wall soft tissuesarcomas are associated with a significantly decreased survival (49% and 39%at 5 and 10 years, respectively), adjuvant therapy should be considered for patients with these tumors. REFERENCES 1. Brennan MF, Shiu MH. Presentation, demographics, and prognostic factors of soft tissue sarcoma. In: Shiu MH, Brennan MF, eds. Surgical management of soft tissue sarcoma. Philadelphia: Lea & Febiger, 1989:45. 2. Rosenberg SA, Suite HD, Baker LH. Sarcomas of soft tissues. In: DeVita VT, Hellman S, Rosenberg SA, eds. Cancer: principles and practice of oncology. 2nd ed. Philadelphia: JB Lippincott Co, 1985:1244. 3. Teitelbaum SL. Twenty years' experience with soft tissue sarcomas of the chest wall at a large institution. J THORAC CARDIOVASC SURG 1972;63:585-6. 4. McCormack P, Bains MS, Beattie EJ, Martini N. New trends in skeletal reconstruction after resection of chest wall tumors. Ann Thorac Surg 1981;31:45-52. 5. Greager JA,Patel MK, BrieleHA, Walker MJ, Wood DK, Das Gupta TK. Soft tissue sarcomas of the adult thoracic wall. Cancer 1987;59:370~3. 6. Graeber GM, Synder RJ, Fleming AW, et al. Initial and long-term results in the management of primary chest wall neoplasms. Ann Thorac Surg 1982;34:664-73. 7. Shamberger RC, Grier HE, Weinstein HJ, Perez-Atayde AR, Tarbell NJ. Chest wall tumors in infancy and childhood. Cancer 1989;63:774-85. 8. Perry RP, Venzon D, Roth JA, Pass HI. Survival after surgical resection for high-grade chest wall sarcomas. Ann Thorac Surg 1990;49:363-9. 9. Suite HD, Mankin HJ, Wood WC, et al. Treatment of the patient with stage MO soft tissuse sarcoma. J Clin Oncol 1988;6:854-62. 10. Hajdu SI. Pathology of soft tissue tumors. Philadelphia: Lea & Febiger, 1979. 11. Hajdu SI. Differential diagnosis of soft tissue and bone tumors. Philadelphia: Lea & Febiger, 1986. 12. Collin CF, Friedrich C, Godbold J, Hajdu S, Brennan MF. Prognostic factors for local recurrence and survival in patients with localized extremity soft-tissue sarcoma. Semin Surg OncoI1988;4:30-7. 13. Torosian MH, Friedrich C, Godbold J, Hajdu SI, Brennan MF. Soft-tissue sarcoma: initial characteristics and prognostic factors in patients with and without metastatic disease. Semin Surg Oncol 1988;4:13-9.
The Journal of Thoracic and Cardiovascular Surgery
8 5 4 Gordon et al.
14. Shiu MH, Brennan MF. Principles of surgical management. In: Shiu MH, Brennan MF, eds. Surgical management of soft tissue sarcoma. Philadelphia: Lea & Febiger, 1989:68-76. 15. Potter DA, Kinsella T, Glatstein E, et ai. High-grade soft tissue sarcomas of the extremities. Cancer 1986;58:190205. 16. Collin C, Godbold J, Hajdu S, Brennan M. Localized extremity soft tissue sarcoma: an analysis of factors affecting survival. J Clin OncoI1987;5:60l-12.
Discussion Dr. John R. Benfield (Sacramento, Calif). This has been an important report about the management of primary chest-wall sarcomas from the viewpoint of therapeutic outcome. I believe, however, that additional emphasis needs to be placed on correct and timely diagnosis and on the role of adjuvant therapy. In my opinion, needle biopsies have little, if any, place in the diagnosis of primary chest-wall sarcomas, because they do not provide enough tissue. The primary use of frozen section is to determine whether an adequate biopsy specimen, free of necrosis, has been obtained. In these diseases the pathologist needs time to study the specimen with permanent sections electron microscopy, and special stains. Treatment planning' includes multimodality therapy often enough that it is usually best done after the best available diagnosis has been made. Adjuvant therapy may be necessary before rather than after operation. How does this translate into practicality? If the chest-wall neoplasm is small enough to be amenable to complete primary resection without mutilation, this should be done. If the tumor is large enough that a major ablation and major reconstruction
will be required, I believe an incisional biopsy should be done as a first ~tage. In some instances, such as Ewing's sarcoma, preoperative therapy may significantly reduce the tumor size, and therefore the need for multilation, and in certain childhood cancers, such as rhabdomyosarcoma, chemotherapy may be the mainstay of therapy. Faced with a possible chest-wall sarcoma there is no need to solve the problem in a single stage, unless opportunity to do so readily presents itself. My position about the diagnostic approach to chest-wall tumors" has proved surprisingly controversial. How do you at Memorial approach the diagnosis of large chest-wall tumors? Dr. Gordon. Dr. Benfield,at Memorial we would agree with your management strategy. Although soft tissue sarcomas are rare, it is important to consider this diagnosis in patients who s~k treatment for tumors of the chest wall. For small superficial tumors, excisional biopsy is a reasonable first diagnostic step. For larger tumors, incisional biopsy allows time for accurate pathologic evaluation. Additional preoperative diagnostic tests such as computed tomographic scan or magnetic resonance imaging may be warranted before proceeding with planned adequate resection. Historically, soft tissue sarcomas of the chest wall have been gro~ped with truncal sarcomas and associated with poor prognosis, Our data, however, demonstrate that survival rates for p~tients wi~h chest-wall soft tissue sarcomas, treated primarily With resection, are comparable to those for patients with soft tissue sarcomas of the extremities. We would recommend that primar~ soft tissue sarcomas of the chest wall be grouped with extremity sarcomas and not truncal sarcomas, particularly in the design of prospective studies, to evaluate adjuvant therapy. *Benfield JR. Primary chest wall tumors [Editorial]. Ann Thorac Surg
1985;39:1.
THORAC CARDIOVASC SURG
1991;101:843-54
Soft tissue sarcomas of the chest wall Results ofsurgical resection Primary soft tissue sarcomas of the chest wall are uncommon, and data concerning treatment and results are sparse. Most studies have categorized these tumors as truncal sarcomas and inferred a poor prognosis. To assess the results of surgical treatment, we reviewed our 4O-year experience. Methods: Records of 189 patients admitted to our institution from 1948 to 1988 were reviewed. Pathologic material was available for review in the 149 cases (79%) that comprise this report. Survival was calculated by the Kaplan-Meier method, with compariso~ by log-rank analysis and significance defined as p < 0.05. Results: Ages ranged from 3 weeks to 86 years (median, 38 years); the ratio of male to female patients was 2:1. The initial complaint was mass or pain in 97% of the cases. Ninety sarcomas (60%) were high grade and 59 (40%) were low grade. Histologic types were as follows: desmoid tumor (n = 32, 21 %); liposarcoma (n = 23, 15 %); rhabdomyosarcoma (n = 18, 12 %); fibrosarcoma (n = 17, 11 %); embryonal rhabdomyosarcoma (n = 14, 9%); malignant peripheral nerve tumor (n = 13, 9%~ malignant fibrolti histiocytoma (n = 11,7%); spindle ceU sarcoma (n = 4,3%); tenosynovial sarcoma (n = 3, 3%); hemangiopericytoma (n = 3, 3% ~ alveolar soft part sarcoma (n = 3, 3% ~ and other types (n = 12, 9%). Resection was the primary treatment in 140 cases (94%). Local recurrence developed in 27%. Metastases occurred in 52 (35%) of the cases (metachronOiti in 42, synchrOOOIti in 10) and were more common in patients with high-grade disease (46/90, 51 %) than in those with low-grade disease (6/59, 10%~ Overall S-year survival was 66%. Five-year survival rate for those with high-grade sarcomas (49%) was significantly lower than that for low-grade sarcomas (90%, p < 0.0001). Tumor size and age of patient were not prognostic. Conclusions: Survival of patients with primary soft tissue sarcomas of the chest wall after resection is similar to that of patients with sarcomas of the extremities. Resection alone provides acceptable survival (90% ai5 years) for those with low-grade sarcomas, but adjuvant treatment should be co~idered for those with high-grade sarcomas.
Mark S. Gordon, MDa (by invitation), Steven I. Hajdu, MDb (by invitation), Manjit S. Bains, MD,a and Michael E. Burt, MD, PhDa (by invitation), New York, N.Y.
Emary soft tissue sarcomas of the chest wall are uncommon. Of the 5000 new cases of soft tissue sarcoma diagnosed annually in the United States, approximately 20% occur in the abdominal or chest wall.!" Thus few
From The Thoracic Service,' Department of Surgery and The Department of Pathology," Memorial Sloan-Kettering Cancer Center, New York, N.Y. Read at the Seventieth Annual Meeting of The American Association for Thoracic Surgery, Toronto, Ontario, Canada, May 7-9,1990. Address for reprints: Michael E. Burt, MD, PhD, Department of Surgery, Memorial Sloan-Kettering Cancer Center, 1275 York Ave., New York, NY 10021.
12/6/26710
centers have acquired a large experience with these tumors. With the advent of reconstructive techniques of the chest wall, extensive tumors have become technically resectable with minimal morbidity." However, data to determine the results of treatment for soft tissue sarcomas arising from the chest wall are sparse and usually involve small numbers.l The recent literature includes analyses of primary chest-wall soft tissue sarcomas in conjunction with chest-wall sarcomas of either bone or connective tissue origin.f 7 soft tissue sarcomas metastatic to the chest wall," or soft tissue sarcomas of the trunk? To further characterize primary chest-wall soft tissue sarcomas and to evaluate results of treatment, we con843
The Journal of
844
Gordon et ai.
Thoracic and Cardiovascular Surgery
Proportion Surviving
1
flti-tt4r i
i
0.8
0.6
1Hlll l ! 1thlHh
1
j
i'I liq j
1
,I
0.4 ,
0.2
o o
10
20
30
40
50
60
70
80
90
100
110
120
Time (rnos)
Fig. 1. Overall survival in 149 patients with chest-wall soft tissue sarcomas. The overall actuarial survival rates were 66% at 5 years and 56% at 10 years. Proportion Surviving
1ft;''-... ---~K-"He
-%
----
-------_._----
+---+-~.'
f..-++-"
't,,,,,
081 0.6
~+-
"t1'c
"-----+~
(p-O.0001)
L
;-j +--------
0.4 ..
0.2
.
I
--+-
- +- 1t
High Grade (n-90)
---+- low Grade (n-59)
o '----_L o 10
_
___
20
.~
~
L
30
40
50
_
60
70
Time (rnos)
80
90
100
110
120
Fig. 2. Actuarial curves of overall survival stratified by tumor grade. Patients with high-grade tumors had 5- and 10-year survival rates of 49% and 39%, respectively, compared to 5- and 10-year survival rates of 90% and 82%, respectively, among patients with low-grade tumors.
ducted a retrospective study of a 40-year experience at Memorial Sloan-Kettering Cancer Center (MSKCC). Patients and methods A computer search of the medical records at our institution from Jan. 1, 1948, to Dec. 31, 1988, yielded the charts of 189 patients with soft tissue sarcomas arising in the chest wall. This represented 5.5% of all soft tissue sarcomas (n = 3417) seen at our institution during this period. We defined primary soft tissue sarcomas of the chest wall as those arising between the clavicles and costochondral cartilages anteriorly, the glenoid fossa
and costochondral junctions laterally, and spinous processes of the scapulae and costochondral borders posteriorly. Patients with tumors of bony or cartilaginous origin (such as Ewing's sarcoma, osteosarcoma, and chondrosarcoma) were not included. All pathologic material was reviewed by one of us (S.I.H .). Patients were excluded from further study if the diagnosis of soft tissue sarcoma could not be confirmed by reevaluation of the original pathologic material. Of the 189 patients identified, pathologic material was available for review in 149 cases (79%), and these comprise the basis of this analysis. Histologic classification of sarcomas was predominantly based on sections stained with hematoxylin and eosin, as previ-
Volume 101 Number 5 May 1991
Chest-wall soft tissue sarcomas 8 4 5
Table I. Histologic types of chest-wall soft tissue sarcomas: Average age ofpatient, gradeof tumor, and distribution by sex No. by grade Histologic type
No.
32 23 18 17 14
Desmoid Liposarcoma Rhabdomyosarcoma Fibrosarcoma Embryonal rhabdomyosarcoma Malignant peripheral nerve tumor Malignant fibrous histiocytoma Spindle cell sarcoma Tenosynovial sarcoma Hemangiopericytoma Alveolar soft-part sarcoma Dermatofibrosarcoma protuberans Leiomyosarcoma Primitive neuroectodermal tumor Giant cell sarcoma Unclassified Total
13 II
4 3 3 3 2 2 I I
--..1.
149
%
Low
High
21 15 12 11 9 9 7
32 16 0 1 0 4
3
2 0 1 0 2 0 0 0
0 7 18 16 14 9 10 2 3 2
I
3 3 3
1 I I I
1
3
0 2 1 I
-l!
..1 90
59
Mean age (yr)*
No. male
33 56 44 44 17 49 44 45 37 38 20 52 62 16 52 33 41
13 20 13 12 7 10
8 3
3 3 2 1 2 0 I
--..1.
100
No. female
19 3 5 5 7 3 3
1 0 0 1 I
0 I
0
.n
49
'Overall median age. 38 years.
Table II. Chest-wall soft tissue sarcomas: Patients previously undergoing radiation therapy Previous malignancy Breast carcinoma Hodgkin's disease Breast carcinoma Lymphoma Hodgkin's disease Hodgkin's disease Breast carcinoma
Dose (cGy)
Interval between RT and sarcoma (yr)
Unknown
4 4 13 3 3
1250 3920 3588 2000 6000 4600*
8
10
Chest wall sarcoma Histologic findings
Grade
Desmoid tumor Malignant fibrous histiocytoma Fibrosarcoma Liposarcoma Malignant fibrous histiocytoma Malignant fibrous histiocytoma Malignant peripheral nerve tumor
Low High High Low High High Low
RT, Radiation therapy. 'Skin dose was 6440 cGy (cobalt 60 machine).
ously reported.!? Tissue was designated as unclassified when adequate material was available but when a definitive final classification could not be made. All sarcomas were classified as low- or high-gradeon the basisof cellular differentiation,degree of tumor cellularity and vascularity, amount of stroma and necrosis, and the number of mitoses per 10 high-power microscopicfields. IO, 11 Local recurrence was defined as the appearance of sarcoma in or adjacent to the site of previousdefinitivetherapy. Recurrent sarcoma was classified as metastatic only if it was found distinctlyremovedfrom the site of the primary tumor. Diseasefree intervalwasdefinedas time betweendefinitivetherapy and local recurrence or metastasis. The method of Kaplan-Meier was used to determine actuarialsurvival. Differences in survivalwere determined by log-rank analysis, and differences in local recurrence were determined by Fisher's exact test; significant differences were defined by p <0.05.
Results Age and sex. The patient population ranged in age from 3 weeks to 86 years. The mean and median ages at diagnosis were 41 years and 38 years, respectively. The prevalence by age was bimodal in distribution, with peaks at the third and sixth decades. Compared with ages of patients with sarcomas ofother histologic types, the mean age at diagnosis was lowest for patients with embryonal rhabdomyosarcoma (mean, 17 years) and highest for patients with liposarcoma (mean, 56 years) (Table I). One hundred patients (67%) were male and 49 (33%) were female (a ratio of 2:1). With the exception of desmoid tumors (with a ratio of female to male patients of 1.5:1), all histologic types occurred more commonly in male patients, with ratios ranging from 2:1 to 3:1. Male
The Journal of
8 4 6 Gordon et al.
Thoracic and Cardiovascular Surgery
Proportion Surviving
1~Ir;-------:--c
1 I
(p~O.02) iffi---
-~
Desmoid (n-32)
I
Low Grade (n-27) -D
oI
o
High Grade (n-gO)
_L
10
20
30
40
50
60
70
Time (rnos)
80
90
100
110
120
Fig. 3. Actuarial survival curves compare desmoid tumors (n = 32), low-grade sarcomas other than desmoid tumors (n = 27), and high-grade sarcomas (n = 90). Overall survival rates of patients with low-grade sarcomas other than desmoid tumors (5-year rate, 83%; IO-year rate, 72%) were significantly different from overall survival rates of those with high-grade sarcomas. The survival rate for patients with desmoid tumors tended to be higher than that for those with the other low-grade sarcomas, but this difference did not reach significance (p = 0.07).
Table III. Chest-wall soft tissue sarcomas (n = 149): Treatment based on extent of disease at initial examination No.
Local disease only Resection' Radiation therapy + chemotherapy Radiation therapy only Synchronous metastases Resection of primary tumor and metastases' Radiation therapy to primary tumor + chemotherapy Radiation therapy to primary tumor Chemotherapy
139 137 I I 10 3 4 1 2
'One hundred forty patients underwent resection or the primary tumor as initial therapy.
predominance was particularly notable for liposarcomas, where the ratio of male to female patients was 7:1. Clinical presentation. The initial symptom was a nonpainful mass in 109 patients (73%), a painful mass in 20 (13%), and pain alone in 17 (11%). In two patients the initial symptom was not stated, and one patient had no symptoms but had a mass seen on routine chest radiography. Only seven patients (5%) reported weight loss when seeking treatment. Of the 118 patients with documented duration of symptoms, the mean was 12 months. Seven patients had received previous radiotherapy to the chest wall for a malignant neoplasm (Table 11). Three patients had previously been treated for breast cancer,
three for Hodgkin's disease, and one for lymphoma. In one patient the radiation dose was unknown, but in the other six patients the mean dose was 3560 rad, with a range of 1250 to 6000 rad. The mean and median times from radiotherapy to the diagnosis of soft tissue sarcoma were 6.4 years and 4 years, respectively. High-grade malignant fibrous histiocytoma was the most common sarcoma, occurring in three patients (45%). Among the 32 patients with desmoid tumors, four also had Gardner's syndrome (two male and two female patients, with a mean age of 21 years). Three of these patients subsequently died of complications related to Gardner's syndrome: Adenocarcinoma of the duodenum developed in one, extensive intraabdominal desmoid tumor with small bowel necrosis and sepsis developed in one, and liver failure developed in one after multiple resections for intraabdominal desmoid tumor. Of the 13 patients with malignant peripheral nerve tumors, one patient had von Recklinghausen's disease (multiple neurofibromatosis). This patient was first seen with metastatic disease and died 3 months after the diagnosis of sarcoma. Site of primary tumor. As a group, sarcomas were not associated with any particular site of the chest wall. However, certain sites were occupied more commonly by particular sarcomas. All three tenosynovial sarcomas occurred in the interscapular area, and of the 11 sarcomas in the interscapular area, six (55%) were either tenosynovial sarcomas or fibrosarcomas. Of the 11 sarcomas in
Volume 101
Chest-wall soft tissuesarcomas 8 4 7
Number 5 May 1991
Proportion Surviving 1.-T'i¥-'r-------------------------, 0.8 0.6
(p
0.4 0.2
o " - - _ ' - - - - - _ L - _ - L -_ _--L-_--l-_--'---_---'--_----"-_--'-_---'-_ _-'_-----! o 10 20 30 40 50 60 70 80 90 100 110 120 Time (rnos) -+- Rhabdo (n-18) -8-
Embry Rhabdo (n-14)
- - Fibro (n-16) --- MFH (n-10)
Fig. 4. Actuarial survival curves for patients with the more common high-grade sarcomas, stratified according to histologic origin. Ten-year survival for patients with high-grade sarcomas originating from fibrous or connective tissue (fibrosarcoma and malignant fibrous histiocytoma) was significantly greater (p < 0.0001) than that for patients with high-grade sarcomas of muscular origin (rhabdomyosarcoma and embryonal rhabdomyosarcoma). Rhabdo, Rhabdomyosarcoma; Embry Rhabdo, embryonal rhabdomyosarcoma; Fibro, fibrosarcoma; MFH, malignant fibrous histiocytoma.
the presternal area, six (55%) were either fibrosarcomas or malignant fibrous histiocytomas. Method of diagnosis. The histologic diagnosis was made by biopsy (n = 58, incisional in 32 cases and excisional in 26) or resection (n = 84, limited in 56 cases and radical in 28). In one patient with embryonal rhabdomyosarcoma, the histologic diagnosis was made by lymph node biopsy, and in one patient who had undergone radiotherapy for Hodgkin's disease, the diagnosis was made by aspiration needle biopsy. In five patients the method of initial histologic diagnosis could not be ascertained from the medical record. Histology and grade. The most common types of tumor were desmoid tumor (21%), liposarcoma (15%), rhabdomyosarcoma (12%), and fibrosarcoma (11%) (Table I). Tumors were high grade in 90 patients (60%) and low grade in 59 (40%). As expected, all desmoid tumors (n = 32) and dermatofibrosarcoma protuberans (n = 2) were low grade, and all embryonal rhabdomyosarcomas (n = 14) were high grade. Of the 23 liposarcomas, 16 (70%) were low grade and seven (30%) were high grade. Of the remaining histologic types, most tumors (50% or more) were high grade. Therapy. Of the 149 patients with chest-wall soft tissue sarcoma, 75 (50%) underwent their initial treatment at MSKCC and 74 (50%) underwent their initial treatment at another institution. One hundred forty patients underwent resection of the primary lesion as initial treatment (Table III). Five patients were treated initially by
Table IV. Chest-wall soft tissuesarcomas: Treatment of local recurrence Therapy
No.
Surgery Surgery + radiation therapy* Surgery + radiation therapy + chemotherapy Surgery + brachytherapy Surgery + brachytherapy + radiation therapy Surgery + brachytherapy + chemotherapy Radiation therapy Radiation therapy + chemotherapy None Total
47 II 7 4 7 2 2 I
.L
82
"Radiation therapy denotes external beam.
radiation therapy and chemotherapy; four of these had distant metastases at diagnosis. Twelve patients underwent adjuvant therapy, after having undergone a complete resection. Five patients were treated with radiotherapy, six patients were treated with chemotherapy, and one patient underwent both radiotherapy and chemotherapy. After incomplete resection of the primary tumor (margins positive for disease), 17 patients were treated with radiotherapy. Of these, 11 received intraoperative radiation implants (brachytherapy). Eight patients underwent chemotherapy, combined with radiation therapy in five and without additional treatment in three. Local recurrence. Of the 140 patients who underwent
848
The Journal of Thoracic and Cardiovascular Surgery
Gordon et a/.
Table V. Chest-wall soft tissue sarcomas: Patterns of metastasis Metastases
Grade Low High Size (cm)*
<5 5-9.9 2':10 Surgical margins] Adequate Inadequate Histopathologic findings Desmoid tumor Liposarcoma Rhabdomyosarcoma Fibrosarcoma Embryonal rhabdomyosarcoma Malignant peripheral nerve tumor Malignant fibrous histiocytoma Spindle cell sarcoma Tenosynovial sarcoma Hemangiopericytoma Alveolar soft-part sarcoma Dermatofibrosarcoma protuberans Leiomyosarcoma Primitive neuroectodermal tumor Giant cell sarcoma Unclassified
Total
No.
Synchronous
Metachronous
No.
%
59 90
0 10
6 36
6 46
10 51
22 27 16
0 0 0
3 8 5
3 8 5
14 30 31
53 17
0 0
10 5
10
19 29
32 23 18 17 14
0 0 4 0 5 0 0 0
0 8 8 3 8 4 2
0 8 12 3
13 II
4 3 3 3 2
2 I I
2
I
0 0 0 0 0 0 0
5
13
4 2 2
0 35 67 18 93 31 18 50 100 67 33
2 2 2 I
3 2
0
0
0
I 0 I
I
0
50
0
0
I
I
0
50 0
'Pathologic size was only available for 65 patients (44%).
t Adequacy of margins could only be determined for the 70 patients who underwent primary operation at MSKCC.
resection as initial treatment, 82 (59%) experienced local recurrence. Sixty-three of these (77%) were initially treated at another institution and referred to MSKCC. Adequacy of resection could be determined only in the 70 patients who underwent initial resection or later resection at MSKCC. In 53 of 70 patients (76%) the margins of resection were free of tumor, and in 17 of 70 patients (24%) the resection margins were either microscopically or grossly involved with tumor. Overall local recurrence of tumors initially resected at MSKCC was 27% (19/70). Of the 53 patients with tumor-free margins of resection, 13 (24%) experienced local recurrence, compared with six (35%) of the 17 patients with involved margins (p = 0.164). Local recurrence was found in eight of the 30 patients with high-grade sarcoma who had negative margins (27%), compared with one of nine patients with high-grade sarcoma who had positive margins (II %) (p = 0.249). For low-grade sarcomas, local recurrence developed in five of 23 patients with negative margins (22%), compared with five of eight patients with positive margins (62%) (p = 0.04). Treatment of local recurrences is shown in Table IV.
Seventy-eight patients (95%) underwent resection of the local recurrence either alone (47/82,58%) or in combination with at least one other form of therapy (31/82, 38%). Metastases. Of the 149 patients with chest-wall soft tissue sarcomas, 52 (35%) experienced either synchronous or metachronous metastases (Table V). Metastases occurred in 46 out of 90 (51 %) patients with high-grade chest-wall sarcomas, as compared with six out of 59 (10%) patients with low-grade sarcomas. Metastases occurred in three out of 22 (I4%) patients with tumors less than 5 em in diameter, as compared with 13 out of 43 (30%) patients with tumors 5 cm or greater in diameter. Although adequacy of surgical margins could be determined in only 70 patients, this variable seemed unrelated to the development of metastases. Patients with desmoid tumors (n = 32) or dermatofibrosarcoma protuberans (n = 2) did not experience metastases. Of the more common histologic types (occurring in at least 10 patients), metastases were more likely to develop in patients with embryonal rhabdomyosarcoma (I3/I4, 93%) and rhabdomyosarcoma (I 21 18, 67%). Least likely to metastasize
Volume 101 Number 5
Chest-wall soft tissue sarcomas 8 4 9
May 1991
Proportion Surviving
1_c,-~~~~~~----~~-~~~~~~~~~-------,
0.8
0.6
K 0.4
0.2
-+-
No Pain (n·112)
-+-
Pain (n·37)
(p=O.042)
*
o I,--~---,--~_l.. __ ~_L~~~,--_-,--_-"-_---,--_--,-_---.J,--_----, o 10 20 30 40 50 60 70 80 90 100 110 120 Time (rnos)
Fig. 5. Actuarial survival curves stratified according to the presence or absence of pain at initial examination. Patients with pain at initial examination had a significantly decreased overall survival.
Proportion Surviving
0.6 0.4
(p=O.003)
0.2
o o
'--_L~~-'--_-'--_---'--~-"-~~~---'--~-'-~--'-~----l~---.J--,-------!
10
20
30
40
50
60
70
80
90
100
110
120
Time (rnos) -+-
Low. No Pain (n·49)
-8-
High. No Pain (n·63)
---+-
Low. Pain (n·lO) High. Pain (n·27)
Fig. 6. Actuarial survival curves stratified by grade, according to the presence or absence of pain at initial examination. Pain had a significant adverse effect on survival only among patients with high-grade sarcomas.
were fibrosarcoma and malignant fibrous histiocytoma, each of which metastasized in 18% of patients. Of the 52 patients in whom metastases developed, 28 patients (54%) experienced metastases to a single organ. The most common single site for metastases was the lung, with 20 of the 28 patients (71 %) experiencing metastases to this site. Of the remaining eight patients who experienced metastases to a single site, three (11%) experienced metastases to subcutaneous tissue in areas distinctly removed from the primary site, one (4%) experienced metastases to regional lymph nodes, and one patient each
(4%) experienced metastases to the spine, brain, liver, and retroperitoneum. Twenty-four patients (46%) experienced metastases to multiple sites, including lung, spine, bone (other than spine), pleura, brain, regional lymph nodes, retroperitoneum, and intraabdominal organs. Disease-free interval. Of the 139 patients who were initially seen with local disease, 91 (65%) experienced recurrence after initial therapy. The initial recurrence was local in 77 cases (85%), distant in 13 cases (14%), and both local and distant in one case (1 %). Disease-free
The Journal' of'
8 5 0 Gordon et al.
Thoracic and Cardiovascular Surgery
Proportion Surviving
1i&,i-;-~~~~~~~~~~~~~~--~~~~~~~_
0.8
-+-itH
-~~H''¥1'+'~+----:f,----+------jc+h
0.6 .
0.4
(p=O.89)
0.2
-+--
No Loc Recur (n'67)
-+--
Loc Recur (n'82)
O'--~-'--~-'--~~~-'--~--'---~~~~~~~~~----"-~~~~
o
10
20
30
40
50
60
70
80
90
100
110
120
Time (mas)
Fig. 7. Actuarial survival curves demonstrate the effect of local recurrence. Local recurrence had no impact on overall survival.
Proportion Surviving
~'~~~:::;::::::::J
1
-----+---
0.8
(p=O.31)
0.6
'P~l
0.4 0.2
o o
'---~L-~-'--~.L-.~..L.....~-'---~--"---~~
10
20
30
40
-+-
Low, No Rec (n·25)
-B-
High, No Rec (n·42)
50
60
70
_ _-+-~---"--~----"-~~~-
80
90
100
110
120
Time (mas) -+- Low, Rec (n'34)
High, Recur (n·48)
Fig. 8. Actuarial survival curves, stratified by grade, demonstrate the effect of local recurrence. Within each grade, local recurrence had no effect on overall survival.
intervalranged from 1 to 68 months, with a medianof 14 months. Survival. The median follow-up time in our series of 149 patients was 56 months. Overall survival was 66%at 5 years and 56%at 10 years (Fig. 1). The 5- and IOyear survivalrates for patients with low-gradesarcomas were 90%and 82%,respectively; these weresignificantly higher (p < 0.0001) than the 5- and 10-yearsurvival rates for patientswith high-gradesarcomas(49%and 39%,respectively) (Fig. 2). The 1O-year survival rate for patientswith desmoidtumors was 93%, and there was a trend suggesting that their survivalwas better than that for those with
other low-grade sarcomas (p = 0.07). The difference in survival rates between patients with the 27 other lowgrade sarcomas (5-year survival, 83%; lO-year survival, 72%) and those with high-grade sarcomas remained significant (p = 0.02) (Fig. 3). Survival was next analyzed for the more common high-gradeforms(Fig. 4). At 5 yearsthe survival rates for patients with fibrosarcoma, malignant fibrous histiocytorna, rhabdomyosarcoma, and embryonalrhabdomyosarcoma were 87%,62%,21%, and 14%, respectively. At 10 years, survival rates for patients with fibrosarcoma and malignant fibrous histiocytoma were similar, as were
Volume 101
Chest-wall soft tissue sarcomas 8 5 1
Number 5 May 1991
Proportion Surviving
1_~---------------------------,
0.8
0.6
(p
0.2
-t- Metastasis (n.52)
-+-
No Mets (n·97)
o~-~--'------'---'-------'----L-_-L_---L_--'-_---.l_---'_----.J
o
10
20
30
40
50
60
70
80
90
100
110
120
Time (mos)
Fig. 9. Actuarial curves demonstrate the effect of metastases on survival. Patients who experienced metastases had significantly decreased overall survival. Proportion Surviving
1
~F++=+=+====i;=;::=:=;=====:=l
0.8 o~
0.6 0.4 0.2 _
oL-_L--'--'-_-'-_-'----_---'---_--L-_-L_~
o
10
20
30
40
50
60
70
80
_ _ L_
90
_...L_
100
_____'_~
110
120
Time (rnos) -+- Low, No Mets (n·53) --B-
High, No Mets (n.50)
--+-
Low, Mets (n·6)
- - High, Mets (n-40)
Fig. 10. Actuarial survival curves, stratified by grade, demonstrate the effect of metastases on survival. Within each grade, metastases significantly decreased survival (low-grade, p = 0.031; high-grade, p = 0.000(1). Patients with high-grade sarcomas who did not experience metastases had overall survival similar to that of patients with low-grade sarcomas that metastasized.
rates for those with rhabdomyosarcoma and embryonal rhabdomyosarcoma. The lO-year survival rate was approximately 65% for patients with the high-grade sarcomas originating from fibrous or connective tissue, as opposed to 14% for those with the high-grade sarcomas of muscle origin (p < 0.0001). Patients whose initial symptoms included pain (n = 37) had 5- and lO-year survival rates of 48% and 37% respectively,compared with 5- and 10-year survival rates of 70% and 63%, respectively, among those patients without pain (n = 112) at initial examination (p = 0.042)
(Fig. 5). Further analysis showed this difference to be significant only for those patients with high-grade tumors; there was no survival difference between patients with low-grade tumors with and without pain at initial examination. (Fig. 6). As shown in Fig. 7, overall survival was not affected by local recurrence. When it was analyzed according to grade, local recurrence again was found to have no adverse effect on survival (Fig. 8). In the 52 patients in whom metastases developed, 5and l O-year survival rates were 27% and 19%, respec-
The Journal of .
8 5 2 Gordon et al.
Thoracic and Cardiovascular Surgery
Proportion Surviving
1ftf---+---+--------------------_
0.8
0.6
0.4
0.2
--+-
<
--+-
5-10 em (n-27)
5
em (n-22)
-B- , 10 em (n-16)
o ---'--------J-----'---....l...-_L-----L_-'---_-'---------'_----'--_....l...---.J o 10 20 30 40 50 60 70 80 90 100 110 120 ,-I
Time (mas)
Fig. 11. Actuarial curves ofoverall survival stratified bytumor size. Nosignificant difference was noted among different sizes of primary tumors. tively, as opposedto 83%and 75%among the 97 patients without metastases (p < 0.(001) (Fig. 9). For both lowand high-grade tumors, the development of metastases had a significant adverse effect on survival (Fig. 10). Interestingly, survivalwas similar for patients with highgrade sarcomas who did not experience metastases and patients with low-grade sarcomas who experienced metastases. When analyzed according to sizeof tumor, 5-year survival rate was similar for patients with tumors smaller than 5 em, tumors between 5 and 9.9 em, and tumors at least 10em in size (Fig. 11). The 5-year survival rates for patients with tumors in these three size ranges were 82%, 74%,and 63%,respectively. At 10years the survivalrates for patients with tumors in these three size intervalswere 59%for tumors less than 5 em, 74% for tumors between 5 and 9.9 em, and 63% for tumors larger than 10 em. There was no significantdifferencein survival according to size of tumor. Finally,survival was analyzed accordingto patient age at diagnosis. At 5 years the survivalrate was66%both for patients younger than 53 years at diagnosis and for patients older than 53 years at diagnosis. At 10 years 62% of patients whose diagnoses were made at less than 53 years of age were alive,as opposed to 47%of thoseat ages greater than 53 years (p = 0.07). Discussion
Data concerning treatment and results of therapy of primary soft tissue sarcomas of the chest wall are sparse. Most published series of these relatively uncommon tumors are analyzed as truncal sarcomas and as a group
do relativelypoorly. Our bias at the start of this study was that chest-wall soft tissue sarcomas in general had an outcome of therapy comparable with that of extremity soft tissue sarcomas. Our series consisted of 149 patients with primary soft tissue chest-wall sarcomas who were admitted to our institution over a 40-year period. The predominance of male patients over female patients (2:1) and our median patient age of 38 years are similar to those in other series of chest-walland extremity soft tissue sarcomas.v 12-16 Several studies of patients with chest-wall sarcoma have recently been published.o 6, 8 All of these articles have advocated aggressive chest-wall resection, with adjuvant treatment in selected cases. Prognostic factors have includedgrade, size,status of surgical margins, and metastases.o" Analysis of data from two large series of extremity soft tissue sarcoma has additionally identified age, symptoms, local recurrence, and histopathologic nature as prognostically important factors.12, 15 The overallsurvivalrates in our seriesof 66%at 5 years and 56%at 10 years compare favorably with the 51% 5year and 34% l Oeyear survival rates reported by Greager and coworkers.' and are similarto the 5-yearsurvival rate for patients with primary chest-wallsarcomasreportedby Perry and associates," There was a significant difference in survival basedon tumor grade; patients with low-grade sarcomas had a higher survival rate than had those with high-grade sarcomas. Excluding desmoid tumors from the other low-gradetumors did not alter the significance of this difference. Interestingly, 5- and 10-year survival rates (90% and 82%, respectively) for patients with lowgrade chest-wallsarcomas wereslightlyhigher than were
Volume 101 Number 5 May 1991
similarsurvival data for patients with low-gradeextremity soft tissue sarcomas (5-year rate, 82%; 10-year rate, 74%).16 Survivalwas similarfor patients with high-grade chest-wall (5-year rate, 49%; 10-year rate, 39%) and those with extremity soft tissue sarcomas (5-year rate, 47%; 10-yearrate, 37%).16 Unlike the National Cancer Institute series of highgrade extremity soft tissue sarcoma.P we found histopathologic resultsa significant prognosticindicator.At 10 years, survival for patients with high-grade tumors of muscleoriginwassignificantly lessthan that of thosewith high-grade tumors originating in fibrous or connective tissue. Althoughonly 37 patients (25%) reported pain at initial examination, these patients had decreased survival compared with those without pain at presentation. The adverseeffectof painon survival, however, wassignificant onlyfor patients with high-grade tumors. Similar results have also been reported by Collin and colleagues'? in extremity sarcomas. In a recentlypublishedseries of chest-wall sarcomas.! status of surgical margins was found to be the most importantfactor affectingsurvival. In our series, margins of resection negativefor neoplasmsignificantly decreased the likelihood of local recurrence for patients with lowgrade tumors only. For those with high-grade tumors, surgicalmargin status had no effect on local recurrence, presumably becausepatients did not livelongenoughfor local recurrence to develop. Overall survival was not adversely affected by local recurrence. Furthermore, within each grade, local recurrence had no impact on overall survival. Metastatic diseaseappeared in 52 of our 149 patients. Metastases occurred more frequently in patients with high-gradetumors (51%) than in patientswith low-grade tumors (10%). Whether synchronous or metachronous, the development of metastases had a significant adverse effecton survival. Two large seriesof extremitysoft tissuesarcoma both identified sizeof the primary lesion as a significant prognosticfactor,I2, 15 and in one ofthese series, age at initial examination was also significant.F In our analysis, survival was not influenced by size of the primary tumor. Likewise, age was not related to survival among patients with primary chest-wall soft tissue sarcoma. Resection of low-grade chest-wall soft tissuesarcomas provides acceptable long-term survival. If necessary to achieve clear margins, the chest wall, including the parietal pleura, should be resected in continuity. Although chest-wall resection may be an extensive procedure,several reportshaveshownthat it can be performedwith low morbidity and mortality.r" Techniques for chest-wall
Chest-wall soft tissue sarcomas 8 5 3
reconstruction have been well described." Adjuvant systemic therapy does not seem indicated for patients with low-grade chest-wall sarcomas, given the 90% 5-year survival rate and the 82% 10-year survival rate after resection. However, prospective trials should be directed at decreasing the 22% local recurrence rate associated with adequate widelocal resectionalone in patients with low-grade sarcomas. Becausehigh-gradechest-wall soft tissuesarcomas are associated with a significantly decreased survival (49% and 39%at 5 and 10 years, respectively), adjuvant therapy should be considered for patients with these tumors. REFERENCES 1. Brennan MF, Shiu MH. Presentation, demographics, and prognostic factors of soft tissue sarcoma. In: Shiu MH, Brennan MF, eds. Surgical management of soft tissue sarcoma. Philadelphia: Lea & Febiger, 1989:45. 2. Rosenberg SA, Suite HD, Baker LH. Sarcomas of soft tissues. In: DeVita VT, Hellman S, Rosenberg SA, eds. Cancer: principles and practice of oncology. 2nd ed. Philadelphia: JB Lippincott Co, 1985:1244. 3. Teitelbaum SL. Twenty years' experience with soft tissue sarcomas of the chest wall at a large institution. J THORAC CARDIOVASC SURG 1972;63:585-6. 4. McCormack P, Bains MS, Beattie EJ, Martini N. New trends in skeletal reconstruction after resection of chest wall tumors. Ann Thorac Surg 1981;31:45-52. 5. Greager JA,Patel MK, BrieleHA, Walker MJ, Wood DK, Das Gupta TK. Soft tissue sarcomas of the adult thoracic wall. Cancer 1987;59:370~3. 6. Graeber GM, Synder RJ, Fleming AW, et al. Initial and long-term results in the management of primary chest wall neoplasms. Ann Thorac Surg 1982;34:664-73. 7. Shamberger RC, Grier HE, Weinstein HJ, Perez-Atayde AR, Tarbell NJ. Chest wall tumors in infancy and childhood. Cancer 1989;63:774-85. 8. Perry RP, Venzon D, Roth JA, Pass HI. Survival after surgical resection for high-grade chest wall sarcomas. Ann Thorac Surg 1990;49:363-9. 9. Suite HD, Mankin HJ, Wood WC, et al. Treatment of the patient with stage MO soft tissuse sarcoma. J Clin Oncol 1988;6:854-62. 10. Hajdu SI. Pathology of soft tissue tumors. Philadelphia: Lea & Febiger, 1979. 11. Hajdu SI. Differential diagnosis of soft tissue and bone tumors. Philadelphia: Lea & Febiger, 1986. 12. Collin CF, Friedrich C, Godbold J, Hajdu S, Brennan MF. Prognostic factors for local recurrence and survival in patients with localized extremity soft-tissue sarcoma. Semin Surg OncoI1988;4:30-7. 13. Torosian MH, Friedrich C, Godbold J, Hajdu SI, Brennan MF. Soft-tissue sarcoma: initial characteristics and prognostic factors in patients with and without metastatic disease. Semin Surg Oncol 1988;4:13-9.
The Journal of Thoracic and Cardiovascular Surgery
8 5 4 Gordon et al.
14. Shiu MH, Brennan MF. Principles of surgical management. In: Shiu MH, Brennan MF, eds. Surgical management of soft tissue sarcoma. Philadelphia: Lea & Febiger, 1989:68-76. 15. Potter DA, Kinsella T, Glatstein E, et ai. High-grade soft tissue sarcomas of the extremities. Cancer 1986;58:190205. 16. Collin C, Godbold J, Hajdu S, Brennan M. Localized extremity soft tissue sarcoma: an analysis of factors affecting survival. J Clin OncoI1987;5:60l-12.
Discussion Dr. John R. Benfield (Sacramento, Calif). This has been an important report about the management of primary chest-wall sarcomas from the viewpoint of therapeutic outcome. I believe, however, that additional emphasis needs to be placed on correct and timely diagnosis and on the role of adjuvant therapy. In my opinion, needle biopsies have little, if any, place in the diagnosis of primary chest-wall sarcomas, because they do not provide enough tissue. The primary use of frozen section is to determine whether an adequate biopsy specimen, free of necrosis, has been obtained. In these diseases the pathologist needs time to study the specimen with permanent sections electron microscopy, and special stains. Treatment planning' includes multimodality therapy often enough that it is usually best done after the best available diagnosis has been made. Adjuvant therapy may be necessary before rather than after operation. How does this translate into practicality? If the chest-wall neoplasm is small enough to be amenable to complete primary resection without mutilation, this should be done. If the tumor is large enough that a major ablation and major reconstruction
will be required, I believe an incisional biopsy should be done as a first ~tage. In some instances, such as Ewing's sarcoma, preoperative therapy may significantly reduce the tumor size, and therefore the need for multilation, and in certain childhood cancers, such as rhabdomyosarcoma, chemotherapy may be the mainstay of therapy. Faced with a possible chest-wall sarcoma there is no need to solve the problem in a single stage, unless opportunity to do so readily presents itself. My position about the diagnostic approach to chest-wall tumors" has proved surprisingly controversial. How do you at Memorial approach the diagnosis of large chest-wall tumors? Dr. Gordon. Dr. Benfield,at Memorial we would agree with your management strategy. Although soft tissue sarcomas are rare, it is important to consider this diagnosis in patients who s~k treatment for tumors of the chest wall. For small superficial tumors, excisional biopsy is a reasonable first diagnostic step. For larger tumors, incisional biopsy allows time for accurate pathologic evaluation. Additional preoperative diagnostic tests such as computed tomographic scan or magnetic resonance imaging may be warranted before proceeding with planned adequate resection. Historically, soft tissue sarcomas of the chest wall have been gro~ped with truncal sarcomas and associated with poor prognosis, Our data, however, demonstrate that survival rates for p~tients wi~h chest-wall soft tissue sarcomas, treated primarily With resection, are comparable to those for patients with soft tissue sarcomas of the extremities. We would recommend that primar~ soft tissue sarcomas of the chest wall be grouped with extremity sarcomas and not truncal sarcomas, particularly in the design of prospective studies, to evaluate adjuvant therapy. *Benfield JR. Primary chest wall tumors [Editorial]. Ann Thorac Surg
1985;39:1.