- Email: [email protected]
ADULT PROSTATE SARCOMA: THE M. D. ANDERSON CANCER CENTER EXPERIENCE
0022-5347/01/1662-0521/0 THE JOURNAL OF UROLOGY® Copyright © 2001 by AMERICAN UROLOGICAL ASSOCIATION, INC.®
Vol. 166, 521–525, August 2001 Printed in U.S.A.
ADULT PROSTATE SARCOMA: THE M. D. ANDERSON CANCER CENTER EXPERIENCE WADE J. SEXTON, RAYMOND E. LANCE, ADRIANA O. REYES, PETER W. T. PISTERS, SHI-MING TU AND LOUIS L. PISTERS From the Departments of Urology, Pathology, Surgical Oncology and Genitourinary Medical Oncology, The University of Texas M. D. Anderson Cancer Center, Houston, Texas
ABSTRACT
Purpose: Sarcoma of prostate origin is rare. Historically, long-term survival rates for adult patients with prostate sarcoma are poor. We analyzed the experience of 1 institution with prostate sarcoma during the last 3 decades. Materials and Methods: The records of 21 patients with prostate sarcoma were reviewed to identify symptoms at presentation, diagnostic procedures, presence and development of metastases, staging evaluation, histological subtype, grade and size of the primary tumor, and treatment sequence, including surgery, and preoperative and postoperative therapies. Several clinicopathological variables were assessed for prognostic importance. Results: Most patients presented with urinary obstruction. The diagnosis of prostate sarcoma was usually established with ultrasound guided biopsy or transurethral resection. Histological subtypes were leiomyosarcoma in 12, rhabdomyosarcoma in 4, malignant fibrous histiocytoma in 1 and unclassified sarcoma in 4 patients. At last followup, 8 patients had no evidence of disease after a median of 81.5 months (range 10 to 197). The remaining 13 patients died of sarcoma (median survival 18 months, range 3 to 94). The 1, 3 and 5-year actuarial survival rates for all 21 patients were 81%, 43% and 38%, respectively. Factors predictive of long-term survival were negative surgical margins (p ⫽ 0.0005) and absence of metastatic disease at presentation (p ⫽ 0.0004). Tumor size and grade, and the histological subtype of prostate sarcoma had no significant influence on actuarial survival. Conclusions: The long-term disease specific survival rate for adults with prostate sarcoma is poor. Early diagnosis and complete surgical resection offer patients the best chance for cure. KEY WORDS: prostate, sarcoma, neoplasms, drug therapy, surgery
Primary prostate sarcomas are rare tumors occurring from nonepithelial mesenchymal components of the prostate stroma. These tumors account for less than 0.1% of primary prostate malignancy in adults.1 Leiomyosarcoma is the most common histological type in adults,2 whereas rhabdomyosarcoma is more common in pediatric patients. Surgery has been the mainstay of treatment and usually involves cystoprostatectomy or total pelvic exenteration for more locally advanced lesions. Radical prostatectomy or more conservative surgery may be an option if the sarcoma is small and confined to the prostate2, 3 but this presentation is uncommon. Historically, the long-term survival rate for patients with prostate sarcoma has been poor.2, 4 – 6 We report the experience of 1 institution with prostate sarcoma during the last 3 decades. PATIENTS AND METHODS
From March 1972 to July 2000, 21 adult males with primary prostate sarcoma were evaluated and treated at The University of Texas M. D. Anderson Cancer Center. These patients were identified with an institutional tumor registry and from a prospective sarcoma database. Patients with carcinosarcoma, those in whom the bulk of the tumor did not occur from the prostate and those seen for a second opinion but treated elsewhere were excluded from review. At diagnosis, electron microscopy and immunohistochemi-
cal staining were performed to confirm that the tumors were of mesenchymal origin and determine the subtype of the prostate sarcoma. If the sarcoma subtype could not be determined the tumor was labeled as an unclassified prostate sarcoma. The tumor grade was determined based on the degree of differentiation and cellularity and the number of mitoses per high power field. From the medical records we extracted data on patient race, diagnostic procedures, symptoms at presentation, presence and development of metastases, staging evaluation, histological subtype, grade and size of the primary tumor, and treatment sequence, including surgery and preoperative and postoperative therapies. Followup information was obtained from medical records and telephone interviews. Size was recorded as the maximum tumor dimension on gross specimens when extirpative surgery was performed first or staging magnetic resonance imaging (MRI) and computerized tomography (CT) studies done before therapy. We determined primary tumor response to preoperative therapy in 6 patients by comparing the radiographically determined size at presentation to the tumor size at final pathological assessment. Patient tumors were retrospectively staged according to the American Joint Committee on Cancer (AJCC) staging for soft tissue sarcoma.7 The resection and margin status was reported according to the International Union Against Cancer classification, including microscopically negative, microscopically positive and grossly positive margins.8 Patient survival was determined from the date of diagnosis until death from metastatic prostate sarcoma or until the most recent patient contact. Kaplan-Meier methods were
Accepted for publication February 23, 2001. Editor’s Note: This article is the third of 5 published in this issue for which category 1 CME credits can be earned. Instructions for obtaining credits are given with the questions on pages 662 and 663. 521
522
ADULT PROSTATE SARCOMA
used to assess the effect of various clinicopathological variables on disease specific survival.9 The log-rank test was used to compare survival rates between subgroups, and p ⫽ 0.05 was considered statistically significant. RESULTS
Patients and treatment. The median age of the patient cohort was 49 years (range 17 to 77). Of the patients 71% were in the fifth or sixth decade of life. There were 19 non-Hispanic white men and 2 Hispanic men. Patient symptoms are listed in table 1. Diagnosis with prostate sarcoma via transrectal ultrasound guided biopsy was performed in 12 patients, transurethral prostatic resection in 5, transperineal biopsy in 2, suprapubic prostatectomy in 1 or CT guided biopsy in 1. Staging evaluation usually consisted of CT, MRI, chest radiograph and bone scan. There were 5 (24%) patients who presented with metastatic disease, and all 5 had lung metastasis and 2 had synchronous liver metastasis. Prostate specific antigen (PSA) data were available for 9 patients, and only 1 had increased PSA (13 ng./ml.). The PSA for 7 patients ranged from 0.6 to 2.4 ng./ml (median 1.3) and the remaining patient had a value reported as normal. Table 2 lists disease characteristics, surgeries and outcomes in the 21 patients, and table 3 lists the preoperative and postoperative treatment strategies. Surgery was the initial treatment in 8 patients, of whom 6 had high grade leiomyosarcoma, 1
TABLE 1. Symptoms at presentation No. Pts. (%) Urinary obstruction Pelvic, perineal pain Urinary frequency Urinary retention Rectal pain Constipation Hematuria Pain with ejaculation Several patients had multiple symptoms.
16 (76) 10 (48) 7 (33) 5 (24) 3 (14) 3 (14) 2 (10) 1 (5)
had low grade unclassified sarcoma and 1 had malignant fibrous histiocytoma. Surgical margins were positive and grossly positive in 2, respectively, of the 6 patients who underwent initial cystoprostatectomy, and they received postoperative radiotherapy to the tumor bed. The 2 patients who underwent initial radical retropubic prostatectomy had microscopically negative margins. The majority of the remaining patients (12 of 13) were treated with chemotherapy first. One patient with rhabdomyosarcoma received initial radiotherapy followed by multiagent chemotherapy. There were 6 patients who underwent delayed cystoprostatectomy after preoperative treatment, including 1 with a stable pleural metastatic lesion. We evaluated the pathological response of primary tumor to therapy in these 6 patients (table 4). All 6 surgical specimens re-
TABLE 2. Disease characteristics, surgery and outcome Pt.—Age No. yrs.
Histology
Grade
Size (cm.)
AJCC Stage
Surgery
Margin Resection
1—64
Leiomyosarcoma
High
Less than 5*
II
Cystoprostatectomy, ileal loop urinary diversion
Microscopically neg.
2—49
Leiomyosarcoma
High
2
II
3—56
Leiomyosarcoma
High
3
II
4—57 5—44
Leiomyosarcoma Leiomyosarcoma
Low High
9 6
II III
6—55
Leiomyosarcoma
High
9
III
7—41
Leiomyosarcoma
High
14
III
8—40
Leiomyosarcoma
High
21
III
9—77 10—64
Leiomyosarcoma Leiomyosarcoma
High High
Radical retropubic prostatectomy Cystoprostatectomy, ileal loop urinary diversion None Cystoprostatectomy, ileal loop urinary diversion Cystoprostatectomy, ileal loop urinary diversion Total pelvic exenteration, ileal loop urinary diversion Cystoprostatectomy, ileal loop urinary diversion None None
11—45 12—47 13—56
Leiomyosarcoma Leiomyosarcoma Unclassified
– – Microscopically neg.
14‡—40
IV IV
High High Low
3 Greater than 5* 7 18 1.5
Unclassified
High
10.5
III
15—53
Unclassified
High
13
III
16—56
Unclassified
High
3.2
IV†
17—47
Rhabdomyosarcoma
High
6.5
III
18—17
Rhabdomyosarcoma
High
7
III
None None Radical retropubic prostatectomy Cystoprostatectomy, ileal loop urinary diversion Cystoprostatectomy, ileal loop urinary diversion Cystoprostatectomy, ileal loop urinary diversion Cystoprostatectomy, Indiana continent catheterizable diversion None
19—29
Rhabdomyosarcoma Rhabdomyosarcoma
High
11
III
None
High
14
III
Malignant fibrous histiocytoma
High
11
III
Total pelvic exenteration, ileal loop urinary diversion Cystoprostatectomy, Indiana continent catheterizable diversion
20—27 21‡—54
IV IV I
Status
Survival (mos.)
Microscopically neg.
No evidence of disease, lost to followup Died of disease
41
Microscopically neg.
Died of disease
30
– Microscopically pos.
Died of disease Died of disease
15 20
Microscopically neg.
No evidence of disease No evidence of disease
12
Grossly pos.
Died of disease
15
– –
Died of disease Died of disease
8 3
Died of disease Died of disease No evidence of disease No evidence of disease No evidence of disease Died of disease
18 5 75
Microscopically neg.
Microscopically neg. Microscopically neg. Microscopically neg. Microscopically neg. – –
No evidence of disease No evidence of disease, lost to followup Died of disease
96
88
197 45 20 10 105 36
Grossly pos.
Died of disease
9
Microscopically neg.
Died of disease
94
* By physical examination. † Primary tumor decreased 40% with preoperative therapy and solitary small pleural lesion remained stable. ‡ Local recurrences treated with local salvage surgery.
523
ADULT PROSTATE SARCOMA TABLE 3. Therapy No. Pts. (No. presenting with metastasis) Surgery Surgery only Surgery, adjuvant doxorubicin/ifosfamide Surgery, adjuvant 64, 70 Gy. external beam radiotherapy Surgery, adjuvant doxorubicin/dacarbazine, 60 Gy. adjuvant external beam radiotherapy Preop. chemotherapy, surgery: Doxorubicin, ifosfamide Doxorubicin, cyclophosphamide, vinblastine Doxorubicin, cyclophosphamide, vincristine, dacarbazine Preop. chemotherapy, preop. external beam radiotherapy, surgery: Doxorubicin, dacarbazine, 50 Gy. external beam radiotherapy Doxorubicin, cyclophosphamide, vincristine, 50 Gy. external beam radiotherapy No surgery Chemotherapy only: Doxorubicin, cyclophosphamide, vincristine Doxorubicin, dacarbazine 60 Gy. external beam radiotherapy, cisplatin, ifosfamide, vincristine Multiple chemotherapy regimens for metastases
4 1 2 1 1 1 1
2 (1) 1
3 (2) 2 (1) 1 1 (1)
vealed cystic and focal necrosis that was up to 90% of primary tumor in 1 case. The 3 patients treated with preoperative chemoradiation compared to the 3 treated with chemotherapy alone had a decrease in primary tumor size. There were 5 patients who had microscopically negative and 1 who had grossly positive margins. Local recurrences developed in 2 patients that were treated with additional surgery. One patient with 11 cm. malignant fibrous histiocytoma had local recurrences resected at 33 and 74 months after initial cystoprostatectomy. Microscopic surgical margins were negative in each of the 3 operations. Metastatic disease developed in the lungs and liver 81 months after initial diagnosis, and the patient died of metastatic sarcoma at 94 months. The second patient was treated with preoperative chemotherapy after diagnosis of 10.5 cm. high grade, unclassified prostate sarcoma. Microscopic surgical margins at cystoprostatectomy were negative. Local recurrence involving the rectum developed in the patient. This recurrence was treated with abdominoperineal resection 4 years after initial surgery, and the patient remained free of disease for 11 years. A second local recurrence requiring repeat resection developed 15 years after cystoprostatectomy. The patient had no evidence of disease 197 months after original diagnosis. Patient survival. At last followup, 8 patients had no evidence of disease after a median of 81.5 months (range 10 to 197). The remaining 13 patients died of metastatic prostate sarcoma. No patients died of nonsarcoma causes or as a consequence of treatment. Median survival of these 13 patients was 18 months (range 3 to 94). Once patients were diagnosed with metastases, median survival was 13 months (range 2 to 26). The 1, 3 and 5-year actuarial survival rates for all 21 patients were 81%, 43% and 38%, respectively. For
patients with nonmetastatic disease at presentation (AJCC stages I to III) the 1, 3 and 5-year actuarial survival rates were 94%, 58% and 50%, respectively. Several of the patients in our series were treated with a combined modality approach that appeared to extend mean survival. Of the 5 patients with stage III prostate sarcoma treated with preoperative multimodality therapies 4 survived and were free of disease at a mean of 70 months (table 4). Furthermore, 7 patients in our series had stage III prostate sarcoma with tumors greater than 10 cm. in dimension. The mean survival of 4 of the 7 patients treated with multimodality therapy was 85 months compared to a mean survival of 39 months for the remaining 3. Prognostic factors. There were no differences in survival between patients grouped by histological subtype of prostate sarcoma (fig. 1). Because the AJCC stage for soft tissue sarcoma is, in part, based on tumor grade and size, we also examined survival by tumor grade and by tumors 5 cm. or less compared with those greater than 5 cm. Our analysis demonstrated no survival advantage for patients with low grade sarcoma (data not shown) or smaller prostate sarcoma (fig. 2). We determined the AJCC stage for all 21 patients and analyzed actuarial survival according to stage (table 2). Stages I and II were combined for analytical purposes because of the small number of patients in these groups. We found no significant difference in actuarial survival between patients with stage I or II prostate sarcoma and stage III (p ⫽ 0.8532). However, there was a significant difference in disease specific survival between patients presenting with metastatic prostate sarcoma (stage IV) and nonmetastatic (stage I to III) disease (fig. 3). A total of 14 patients underwent initial or delayed surgical extirpation of the prostate sarcoma and involved local structures. None of the patients in our series had clinically abnormal pelvic lymph nodes. Therefore, pelvic lymphadenectomy was not routinely performed. Microscopically negative surgical margins were achieved in 11 (79%) patients, microscopically positive margins were present in 1 and grossly positive margins were present in 2 (table 2). The 5-year actuarial survival rate for patients with negative surgical margins was 67% compared to 0% for those with positive surgical margins (fig. 4). DISCUSSION
Contemporary data on prostate sarcoma are limited. The 2 largest series in the last decade are from Cheville4 and from Russo5 et al. There are significant differences between these series and our own. Cheville et al reported 23 cases of prostate leiomyosarcoma, although with clinical information for only 14. Their review spanned 65 years and may be limited by inadequate clinical staging due to inferior imaging capabilities in earlier decades. Russo et al from the Memorial Sloan-Kettering Cancer Center reported their experience with genitourinary sarcoma. Of 43 patients 10 had high grade prostate sarcoma but accurate comparison with our patients was difficult because their patients with prostate sarcoma were analyzed together with those with other genitourinary sarcoma, including those of renal, bladder, seminal vesicle and paratesticular origin. Other series of patients with prostate sarcoma are limited by small patient number
TABLE 4. Treatment and outcome of patients who underwent preoperative combined modality therapy followed by surgery Pt. No. 7 14 15 16 17 20
Therapy Doxorubicin, dacarbazine, 50 Gy. external beam radiotherapy Doxorubicin, cyclophosphamide, vinblastine Doxorubicin, ifosfamide Doxorubicin, dacarbazine, 50 Gy. external beam radiotherapy Doxorubicin, cyclophosphamide, vincristine, 50 Gy. external beam radiotherapy Doxorubicin, cyclophosphamide, vincristine, dacarbazine
Mass Size Change (%)
Margin Resection
Decrease (46) No change No change Decrease (37) Decrease (31)
Microscopically Microscopically Microscopically Microscopically Microscopically
No change
Grossly pos.
neg. neg. neg. neg. neg.
Disease Status No evidence of disease No evidence of disease No evidence of disease Died of disease No evidence of disease Died of disease
524
ADULT PROSTATE SARCOMA
FIG. 1. Actuarial survival by prostate sarcoma subtype. Other sarcomas include unclassified subtypes in 4 patients and malignant fibrous histiocytoma in 1. No significant difference exists between patients with leiomyosarcoma and rhabdomyosarcoma (p ⫽ 0.4759), leiomyosarcoma and other sarcoma (p⫽0.0653) or rhabdomyosarcoma and other sarcoma (p ⫽ 0.5239). Values in parentheses are 5-year actuarial survival rates.
FIG. 3. Difference in actuarial survival between patients presenting with metastatic disease (stage IV) and those with all other stages combined (I to III, p ⫽ 0.0004). Values in parentheses are 5-year actuarial survival rates.
FIG. 4. Actuarial survival by surgical margins after extirpative surgery. Significant difference in survival exists between patients with microscopically negative margins (R0) compared to microscopically positive (R1) or grossly positive (R2) surgical margins (p ⫽ 0.0005). Values in parentheses are 5-year actuarial survival rates. FIG. 2. Difference in actuarial survival between patients with tumors 5 cm. or less and those with greater than 5 cm. (p ⫽ 0.7914). Values in parentheses are 5-year actuarial survival rates.
or the lack of contemporary staging.1, 6, 10, 11 Our series of patients represents a larger group treated at a single institution during a more contemporary period and is limited to prostate sarcoma. Although we did include different histological subtypes of prostate sarcoma in our review, treatment and outcome in all patients are specifically outlined. We also examined whether histological subtype affected survival. The diagnosis of prostate sarcoma is usually accomplished with transrectal ultrasound guided core biopsy, which is well tolerated, requires no anesthesia and often provides adequate tissue to render the histological diagnosis. Most prostate sarcoma causes urinary obstruction,4, 5 and transurethral prostatic resection performed because of symptoms of bladder outlet obstruction may reveal prostate sarcoma. As demonstrated in our series, the PSA value is almost always normal in patients with prostate sarcoma, which is not surprising given the nonepithelial origin of prostate sarcoma. Benign spindle cell nodules of the prostate may result from prior urological procedures, such as prostate needle biopsy or transurethral resection. These lesions could be misdiagnosed
as low grade sarcomatous neoplasms.12 Other benign proliferative prostate lesions can occur in the absence of prior genitourinary surgery, stressing the importance of careful preoperative histological characterization of atypical prostate tumors before extensive pelvic surgery.13, 14 Clinical staging is further enhanced by cross-sectional pelvic imaging. MRI has been the study of choice for extremity sarcoma.15, 16 For large prostate sarcoma we believe MRI facilitates surgical planning with multiplanar images, better delineation of soft tissue planes, and superior assessment of the relationship between tumor and adjacent viscera and neurovascular structures. The overall survival rate for adult patients with prostate sarcoma is poor. The lack of early, specific clinical symptoms result in more advanced disease at presentation. The overall 5-year actuarial survival rate for our patients was 38% compared with a range of 0% to 60% reported in other series.2, 4 – 6, 10 However, data from various series are difficult to compare because of nonuniform reporting on patient and pathological characteristics, changes with time in radiographic staging and the relatively small number of patients with prostate sarcoma. We determined whether various factors, including tumor
525
ADULT PROSTATE SARCOMA
histology, grade, size, AJCC stage and surgical margin status, were predictive of long-term survival. There were no survival differences between patients grouped according to histological subtype (fig. 1). Although there were only 2 patients with low grade prostate sarcoma in our series, we found no survival difference between those with high versus low grade tumors (data not shown). Additionally, patients with tumors 5 cm. or smaller did not have a higher survival rate than those with tumors larger than 5 cm. (fig. 2). We also did not see any significant survival advantage for patients with AJCC stage I or II over those with stage III prostate sarcomas. These findings are in contrast to data presented in the series of Russo et al in which low tumor grade, tumor size 5 cm. or smaller and low stage were favorable predictors of long-term survival of patients with genitourinary sarcoma.5 These differences may be related to the heterogeneous anatomical sites in their report that demonstrated significant survival advantages for patients with bladder and paratesticular sarcomas compared to those with prostate and renal sarcomas. Notwithstanding the limitations of sample size, we do not believe that all genitourinary sarcoma can necessarily be characterized by a single staging system that can predict outcome. In contrast to other clinicopathological factors, surgical margin status does appear to be important for long-term survival. We achieved microscopically negative surgical margins in 79% of our patients who underwent surgery, and the actuarial survival rate was significantly higher than that for those with positive and grossly positive margins (fig. 4). This finding is supported by data from the series by Russo et al in which disease-free survival at 3 years was significantly improved in patients with negative surgical margins.5 If a patient with prostate sarcoma is believed to be at high risk for positive surgical margins preoperative therapy should be considered. Preoperative chemotherapy with subsequent cystoprostatectomy is associated with significant pathological response. In a study of 6 patients with prostate sarcoma Ahlering et al found that a combination of doxorubicin and cisplatin produced greater than 95% necrosis of primary tumors, although there was no evidence of significant radiographic response.10 We found necrosis in the pathological specimens after preoperative chemotherapy as well. However, only patients who received preoperative radiotherapy had appreciable downsizing of the primary tumor (table 4). The prognosis is poor for patients with prostate sarcoma treated with surgery alone, so multimodality therapy should be strongly considered for them. The rare occurrence of adult prostate sarcoma has limited the critical evaluation of treatment strategies and patient survival to anecdotal experiences. However, our results suggest that there may be a survival advantage for patients treated with a combined multimodality approach. Insight into multimodality treatments likely to be more effective for patients with prostate sarcoma can be gained from existing protocol based strategies for other sarcomas. For example, a current protocol for retroperitoneal and pelvic sarcomas incorporates preoperative chemoradiation followed by surgical resection and intraoperative radiotherapy.17 Innovative, combined multimodality treatment strategies may help to improve prognosis for patients with these generally poor prognosis tumors. A final revelation from our review is that patients need close long-term surveillance to detect recurrent prostate sarcoma. Late local recurrences of soft tissue sarcoma are not uncommon and are significantly associated with microscopically positive surgical margins.18, 19 Multiple metachronous local recurrences amenable to surgical resection developed in 2 patients in our series, and 1 survived and is free of disease more than 16 years after initial diagnosis. To our knowledge, our series is the first to demonstrate such a role for local salvage surgery for management of prostate sarcoma. To
detect late recurrences, we recommend that patients with no evidence of disease after treatment be followed closely with chest, abdominal and pelvic imaging. CONCLUSIONS
Prostate sarcomas are rare tumors and often extremely large at initial patient presentation. The overall survival rate remains poor regardless of initial tumor size, grade or histological subtype. Complete surgical resection, including microscopically negative margins, is almost always required to cure patients, and achieving microscopically negative surgical margins appears critical for long-term survival. Long-term surveillance is necessary to detect recurrences. Survival of patients with isolated local recurrences may be prolonged with salvage surgery. REFERENCES
1. Tetu, B., Srigley, J. R. and Bostwick, D. G.: Soft tissue tumors. In: Pathology of the Prostate. Edited by D. G. Bostwick. New York: Churchill-Livingstone, p. 117, 1990 2. Morse, M. J. and Whitmore, W. F.: Genitourinary sarcomas in adults. In: Surgical Management of Soft Tissue Sarcomas. Edited by M. H. Shiu and M. F. Brennan. Philadelphia: Lea and Febiger, pp. 189 –207, 1989 3. Quinlan, D. M., Stutzman, R. E., Peters, C. A. et al: Unilateral nerve-sparing radical prostatectomy and hemicystectomy in management of prostate sarcoma. Urology, 41: 308, 1993 4. Cheville, J. C., Dundore, P. A., Nascimento, A. G. et al: Leiomyosarcoma of the prostate: report of 23 cases. Cancer, 76: 1422, 1995 5. Russo, P., Brady, M. S., Conlon, K. et al: Adult urological sarcoma. J Urol, 147: 1032, 1992 6. Narayana, A. S., Loening, S., Weimar, G. W. et al: Sarcoma of the bladder and prostate. J Urol, 119: 72, 1978 7. American Joint Committee on Cancer: Soft tissue sarcoma. In: American Joint Committee on Cancer Staging Manual, 5th ed. Edited by I. D. Fleming, J. S. Cooper, D. E. Henson et al. Philadelphia: Lippincott-Raven, pp.149 –156, 1997 8. UICC International Union Against Cancer: Urological tumors. In: TNM Classification of Malignant Tumors, 5th ed. Edited by L. H. Sobin and Ch. Wittekind. New York: Wiley-Liss, pp. 165–194, 1997 9. Kaplan, E. L. and Meier, P.: Nonparametric estimation from incomplete observations. J Amer Stat Ass, 53: 457, 1958 10. Ahlering, T. E., Weintraub, P. and Skinner, D. G.: Management of adult sarcomas of the bladder and prostate. J Urol, 140: 1397, 1988 11. Smith, B. H. and Dehner, L. P.: Sarcoma of the prostate gland. Am J Clin Pathol, 58: 43, 1972 12. Huang, W. L., Ro, J. Y., Grignon, D. J. et al: Postoperative spindle cell nodule of the prostate and bladder. J Urol, 143: 824, 1990 13. Gaudin, P. B., Rosai, J. and Epstein, J. I.: Sarcomas and related proliferative lesions of specialized prostatic stroma: a clinicopathologic study of 22 cases. Am J Surg Pathol, 22: 148, 1998 14. Cespedes, R. D., Lynch, S. C. and Grider, D. J.: Pseudosarcomatous fibromyxoid tumor of the prostate: a case report with review of the literature. Urol Int, 57: 249, 1996 15. Chang, A. E., Matory, Y. L., Dwyer, A. J. et al: Magnetic resonance imaging versus computed tomography in the evaluation of soft tissue tumors of the extremities. Ann Surg, 205: 340, 1987 16. Hanna, S. L. and Fletcher, B. D.: MR imaging of malignant soft-tissue tumors. Magn Reson Imaging Clin North Am, 3: 629, 1995 17. Pisters, P. W. T., Patel, S., Crane, C. et al: Phase-I trial of preoperative doxorubicin-based concurrent chemoradiation and electron-beam intraoperative radiation therapy (IORT) for respectable retroperitoneal sarcomas. Proc Am Soc Clin Oncol, suppl., 19: 558a, 2000 18. Lewis, J. J., Leung, D., Casper, E. S. et al: Multifactorial analysis of long-term followup (more than 5 years) of primary extremity sarcoma. Arch Surg, 134: 190, 1999 19. Fleming, J. B., Berman, R. S., Cheng, S. C. et al: Long-term outcome of patients with American Joint Committee on Cancer Stage IIB extremity soft tissue sarcomas. J Clin Oncol, 17: 2772, 1999
Vol. 166, 521–525, August 2001 Printed in U.S.A.
ADULT PROSTATE SARCOMA: THE M. D. ANDERSON CANCER CENTER EXPERIENCE WADE J. SEXTON, RAYMOND E. LANCE, ADRIANA O. REYES, PETER W. T. PISTERS, SHI-MING TU AND LOUIS L. PISTERS From the Departments of Urology, Pathology, Surgical Oncology and Genitourinary Medical Oncology, The University of Texas M. D. Anderson Cancer Center, Houston, Texas
ABSTRACT
Purpose: Sarcoma of prostate origin is rare. Historically, long-term survival rates for adult patients with prostate sarcoma are poor. We analyzed the experience of 1 institution with prostate sarcoma during the last 3 decades. Materials and Methods: The records of 21 patients with prostate sarcoma were reviewed to identify symptoms at presentation, diagnostic procedures, presence and development of metastases, staging evaluation, histological subtype, grade and size of the primary tumor, and treatment sequence, including surgery, and preoperative and postoperative therapies. Several clinicopathological variables were assessed for prognostic importance. Results: Most patients presented with urinary obstruction. The diagnosis of prostate sarcoma was usually established with ultrasound guided biopsy or transurethral resection. Histological subtypes were leiomyosarcoma in 12, rhabdomyosarcoma in 4, malignant fibrous histiocytoma in 1 and unclassified sarcoma in 4 patients. At last followup, 8 patients had no evidence of disease after a median of 81.5 months (range 10 to 197). The remaining 13 patients died of sarcoma (median survival 18 months, range 3 to 94). The 1, 3 and 5-year actuarial survival rates for all 21 patients were 81%, 43% and 38%, respectively. Factors predictive of long-term survival were negative surgical margins (p ⫽ 0.0005) and absence of metastatic disease at presentation (p ⫽ 0.0004). Tumor size and grade, and the histological subtype of prostate sarcoma had no significant influence on actuarial survival. Conclusions: The long-term disease specific survival rate for adults with prostate sarcoma is poor. Early diagnosis and complete surgical resection offer patients the best chance for cure. KEY WORDS: prostate, sarcoma, neoplasms, drug therapy, surgery
Primary prostate sarcomas are rare tumors occurring from nonepithelial mesenchymal components of the prostate stroma. These tumors account for less than 0.1% of primary prostate malignancy in adults.1 Leiomyosarcoma is the most common histological type in adults,2 whereas rhabdomyosarcoma is more common in pediatric patients. Surgery has been the mainstay of treatment and usually involves cystoprostatectomy or total pelvic exenteration for more locally advanced lesions. Radical prostatectomy or more conservative surgery may be an option if the sarcoma is small and confined to the prostate2, 3 but this presentation is uncommon. Historically, the long-term survival rate for patients with prostate sarcoma has been poor.2, 4 – 6 We report the experience of 1 institution with prostate sarcoma during the last 3 decades. PATIENTS AND METHODS
From March 1972 to July 2000, 21 adult males with primary prostate sarcoma were evaluated and treated at The University of Texas M. D. Anderson Cancer Center. These patients were identified with an institutional tumor registry and from a prospective sarcoma database. Patients with carcinosarcoma, those in whom the bulk of the tumor did not occur from the prostate and those seen for a second opinion but treated elsewhere were excluded from review. At diagnosis, electron microscopy and immunohistochemi-
cal staining were performed to confirm that the tumors were of mesenchymal origin and determine the subtype of the prostate sarcoma. If the sarcoma subtype could not be determined the tumor was labeled as an unclassified prostate sarcoma. The tumor grade was determined based on the degree of differentiation and cellularity and the number of mitoses per high power field. From the medical records we extracted data on patient race, diagnostic procedures, symptoms at presentation, presence and development of metastases, staging evaluation, histological subtype, grade and size of the primary tumor, and treatment sequence, including surgery and preoperative and postoperative therapies. Followup information was obtained from medical records and telephone interviews. Size was recorded as the maximum tumor dimension on gross specimens when extirpative surgery was performed first or staging magnetic resonance imaging (MRI) and computerized tomography (CT) studies done before therapy. We determined primary tumor response to preoperative therapy in 6 patients by comparing the radiographically determined size at presentation to the tumor size at final pathological assessment. Patient tumors were retrospectively staged according to the American Joint Committee on Cancer (AJCC) staging for soft tissue sarcoma.7 The resection and margin status was reported according to the International Union Against Cancer classification, including microscopically negative, microscopically positive and grossly positive margins.8 Patient survival was determined from the date of diagnosis until death from metastatic prostate sarcoma or until the most recent patient contact. Kaplan-Meier methods were
Accepted for publication February 23, 2001. Editor’s Note: This article is the third of 5 published in this issue for which category 1 CME credits can be earned. Instructions for obtaining credits are given with the questions on pages 662 and 663. 521
522
ADULT PROSTATE SARCOMA
used to assess the effect of various clinicopathological variables on disease specific survival.9 The log-rank test was used to compare survival rates between subgroups, and p ⫽ 0.05 was considered statistically significant. RESULTS
Patients and treatment. The median age of the patient cohort was 49 years (range 17 to 77). Of the patients 71% were in the fifth or sixth decade of life. There were 19 non-Hispanic white men and 2 Hispanic men. Patient symptoms are listed in table 1. Diagnosis with prostate sarcoma via transrectal ultrasound guided biopsy was performed in 12 patients, transurethral prostatic resection in 5, transperineal biopsy in 2, suprapubic prostatectomy in 1 or CT guided biopsy in 1. Staging evaluation usually consisted of CT, MRI, chest radiograph and bone scan. There were 5 (24%) patients who presented with metastatic disease, and all 5 had lung metastasis and 2 had synchronous liver metastasis. Prostate specific antigen (PSA) data were available for 9 patients, and only 1 had increased PSA (13 ng./ml.). The PSA for 7 patients ranged from 0.6 to 2.4 ng./ml (median 1.3) and the remaining patient had a value reported as normal. Table 2 lists disease characteristics, surgeries and outcomes in the 21 patients, and table 3 lists the preoperative and postoperative treatment strategies. Surgery was the initial treatment in 8 patients, of whom 6 had high grade leiomyosarcoma, 1
TABLE 1. Symptoms at presentation No. Pts. (%) Urinary obstruction Pelvic, perineal pain Urinary frequency Urinary retention Rectal pain Constipation Hematuria Pain with ejaculation Several patients had multiple symptoms.
16 (76) 10 (48) 7 (33) 5 (24) 3 (14) 3 (14) 2 (10) 1 (5)
had low grade unclassified sarcoma and 1 had malignant fibrous histiocytoma. Surgical margins were positive and grossly positive in 2, respectively, of the 6 patients who underwent initial cystoprostatectomy, and they received postoperative radiotherapy to the tumor bed. The 2 patients who underwent initial radical retropubic prostatectomy had microscopically negative margins. The majority of the remaining patients (12 of 13) were treated with chemotherapy first. One patient with rhabdomyosarcoma received initial radiotherapy followed by multiagent chemotherapy. There were 6 patients who underwent delayed cystoprostatectomy after preoperative treatment, including 1 with a stable pleural metastatic lesion. We evaluated the pathological response of primary tumor to therapy in these 6 patients (table 4). All 6 surgical specimens re-
TABLE 2. Disease characteristics, surgery and outcome Pt.—Age No. yrs.
Histology
Grade
Size (cm.)
AJCC Stage
Surgery
Margin Resection
1—64
Leiomyosarcoma
High
Less than 5*
II
Cystoprostatectomy, ileal loop urinary diversion
Microscopically neg.
2—49
Leiomyosarcoma
High
2
II
3—56
Leiomyosarcoma
High
3
II
4—57 5—44
Leiomyosarcoma Leiomyosarcoma
Low High
9 6
II III
6—55
Leiomyosarcoma
High
9
III
7—41
Leiomyosarcoma
High
14
III
8—40
Leiomyosarcoma
High
21
III
9—77 10—64
Leiomyosarcoma Leiomyosarcoma
High High
Radical retropubic prostatectomy Cystoprostatectomy, ileal loop urinary diversion None Cystoprostatectomy, ileal loop urinary diversion Cystoprostatectomy, ileal loop urinary diversion Total pelvic exenteration, ileal loop urinary diversion Cystoprostatectomy, ileal loop urinary diversion None None
11—45 12—47 13—56
Leiomyosarcoma Leiomyosarcoma Unclassified
– – Microscopically neg.
14‡—40
IV IV
High High Low
3 Greater than 5* 7 18 1.5
Unclassified
High
10.5
III
15—53
Unclassified
High
13
III
16—56
Unclassified
High
3.2
IV†
17—47
Rhabdomyosarcoma
High
6.5
III
18—17
Rhabdomyosarcoma
High
7
III
None None Radical retropubic prostatectomy Cystoprostatectomy, ileal loop urinary diversion Cystoprostatectomy, ileal loop urinary diversion Cystoprostatectomy, ileal loop urinary diversion Cystoprostatectomy, Indiana continent catheterizable diversion None
19—29
Rhabdomyosarcoma Rhabdomyosarcoma
High
11
III
None
High
14
III
Malignant fibrous histiocytoma
High
11
III
Total pelvic exenteration, ileal loop urinary diversion Cystoprostatectomy, Indiana continent catheterizable diversion
20—27 21‡—54
IV IV I
Status
Survival (mos.)
Microscopically neg.
No evidence of disease, lost to followup Died of disease
41
Microscopically neg.
Died of disease
30
– Microscopically pos.
Died of disease Died of disease
15 20
Microscopically neg.
No evidence of disease No evidence of disease
12
Grossly pos.
Died of disease
15
– –
Died of disease Died of disease
8 3
Died of disease Died of disease No evidence of disease No evidence of disease No evidence of disease Died of disease
18 5 75
Microscopically neg.
Microscopically neg. Microscopically neg. Microscopically neg. Microscopically neg. – –
No evidence of disease No evidence of disease, lost to followup Died of disease
96
88
197 45 20 10 105 36
Grossly pos.
Died of disease
9
Microscopically neg.
Died of disease
94
* By physical examination. † Primary tumor decreased 40% with preoperative therapy and solitary small pleural lesion remained stable. ‡ Local recurrences treated with local salvage surgery.
523
ADULT PROSTATE SARCOMA TABLE 3. Therapy No. Pts. (No. presenting with metastasis) Surgery Surgery only Surgery, adjuvant doxorubicin/ifosfamide Surgery, adjuvant 64, 70 Gy. external beam radiotherapy Surgery, adjuvant doxorubicin/dacarbazine, 60 Gy. adjuvant external beam radiotherapy Preop. chemotherapy, surgery: Doxorubicin, ifosfamide Doxorubicin, cyclophosphamide, vinblastine Doxorubicin, cyclophosphamide, vincristine, dacarbazine Preop. chemotherapy, preop. external beam radiotherapy, surgery: Doxorubicin, dacarbazine, 50 Gy. external beam radiotherapy Doxorubicin, cyclophosphamide, vincristine, 50 Gy. external beam radiotherapy No surgery Chemotherapy only: Doxorubicin, cyclophosphamide, vincristine Doxorubicin, dacarbazine 60 Gy. external beam radiotherapy, cisplatin, ifosfamide, vincristine Multiple chemotherapy regimens for metastases
4 1 2 1 1 1 1
2 (1) 1
3 (2) 2 (1) 1 1 (1)
vealed cystic and focal necrosis that was up to 90% of primary tumor in 1 case. The 3 patients treated with preoperative chemoradiation compared to the 3 treated with chemotherapy alone had a decrease in primary tumor size. There were 5 patients who had microscopically negative and 1 who had grossly positive margins. Local recurrences developed in 2 patients that were treated with additional surgery. One patient with 11 cm. malignant fibrous histiocytoma had local recurrences resected at 33 and 74 months after initial cystoprostatectomy. Microscopic surgical margins were negative in each of the 3 operations. Metastatic disease developed in the lungs and liver 81 months after initial diagnosis, and the patient died of metastatic sarcoma at 94 months. The second patient was treated with preoperative chemotherapy after diagnosis of 10.5 cm. high grade, unclassified prostate sarcoma. Microscopic surgical margins at cystoprostatectomy were negative. Local recurrence involving the rectum developed in the patient. This recurrence was treated with abdominoperineal resection 4 years after initial surgery, and the patient remained free of disease for 11 years. A second local recurrence requiring repeat resection developed 15 years after cystoprostatectomy. The patient had no evidence of disease 197 months after original diagnosis. Patient survival. At last followup, 8 patients had no evidence of disease after a median of 81.5 months (range 10 to 197). The remaining 13 patients died of metastatic prostate sarcoma. No patients died of nonsarcoma causes or as a consequence of treatment. Median survival of these 13 patients was 18 months (range 3 to 94). Once patients were diagnosed with metastases, median survival was 13 months (range 2 to 26). The 1, 3 and 5-year actuarial survival rates for all 21 patients were 81%, 43% and 38%, respectively. For
patients with nonmetastatic disease at presentation (AJCC stages I to III) the 1, 3 and 5-year actuarial survival rates were 94%, 58% and 50%, respectively. Several of the patients in our series were treated with a combined modality approach that appeared to extend mean survival. Of the 5 patients with stage III prostate sarcoma treated with preoperative multimodality therapies 4 survived and were free of disease at a mean of 70 months (table 4). Furthermore, 7 patients in our series had stage III prostate sarcoma with tumors greater than 10 cm. in dimension. The mean survival of 4 of the 7 patients treated with multimodality therapy was 85 months compared to a mean survival of 39 months for the remaining 3. Prognostic factors. There were no differences in survival between patients grouped by histological subtype of prostate sarcoma (fig. 1). Because the AJCC stage for soft tissue sarcoma is, in part, based on tumor grade and size, we also examined survival by tumor grade and by tumors 5 cm. or less compared with those greater than 5 cm. Our analysis demonstrated no survival advantage for patients with low grade sarcoma (data not shown) or smaller prostate sarcoma (fig. 2). We determined the AJCC stage for all 21 patients and analyzed actuarial survival according to stage (table 2). Stages I and II were combined for analytical purposes because of the small number of patients in these groups. We found no significant difference in actuarial survival between patients with stage I or II prostate sarcoma and stage III (p ⫽ 0.8532). However, there was a significant difference in disease specific survival between patients presenting with metastatic prostate sarcoma (stage IV) and nonmetastatic (stage I to III) disease (fig. 3). A total of 14 patients underwent initial or delayed surgical extirpation of the prostate sarcoma and involved local structures. None of the patients in our series had clinically abnormal pelvic lymph nodes. Therefore, pelvic lymphadenectomy was not routinely performed. Microscopically negative surgical margins were achieved in 11 (79%) patients, microscopically positive margins were present in 1 and grossly positive margins were present in 2 (table 2). The 5-year actuarial survival rate for patients with negative surgical margins was 67% compared to 0% for those with positive surgical margins (fig. 4). DISCUSSION
Contemporary data on prostate sarcoma are limited. The 2 largest series in the last decade are from Cheville4 and from Russo5 et al. There are significant differences between these series and our own. Cheville et al reported 23 cases of prostate leiomyosarcoma, although with clinical information for only 14. Their review spanned 65 years and may be limited by inadequate clinical staging due to inferior imaging capabilities in earlier decades. Russo et al from the Memorial Sloan-Kettering Cancer Center reported their experience with genitourinary sarcoma. Of 43 patients 10 had high grade prostate sarcoma but accurate comparison with our patients was difficult because their patients with prostate sarcoma were analyzed together with those with other genitourinary sarcoma, including those of renal, bladder, seminal vesicle and paratesticular origin. Other series of patients with prostate sarcoma are limited by small patient number
TABLE 4. Treatment and outcome of patients who underwent preoperative combined modality therapy followed by surgery Pt. No. 7 14 15 16 17 20
Therapy Doxorubicin, dacarbazine, 50 Gy. external beam radiotherapy Doxorubicin, cyclophosphamide, vinblastine Doxorubicin, ifosfamide Doxorubicin, dacarbazine, 50 Gy. external beam radiotherapy Doxorubicin, cyclophosphamide, vincristine, 50 Gy. external beam radiotherapy Doxorubicin, cyclophosphamide, vincristine, dacarbazine
Mass Size Change (%)
Margin Resection
Decrease (46) No change No change Decrease (37) Decrease (31)
Microscopically Microscopically Microscopically Microscopically Microscopically
No change
Grossly pos.
neg. neg. neg. neg. neg.
Disease Status No evidence of disease No evidence of disease No evidence of disease Died of disease No evidence of disease Died of disease
524
ADULT PROSTATE SARCOMA
FIG. 1. Actuarial survival by prostate sarcoma subtype. Other sarcomas include unclassified subtypes in 4 patients and malignant fibrous histiocytoma in 1. No significant difference exists between patients with leiomyosarcoma and rhabdomyosarcoma (p ⫽ 0.4759), leiomyosarcoma and other sarcoma (p⫽0.0653) or rhabdomyosarcoma and other sarcoma (p ⫽ 0.5239). Values in parentheses are 5-year actuarial survival rates.
FIG. 3. Difference in actuarial survival between patients presenting with metastatic disease (stage IV) and those with all other stages combined (I to III, p ⫽ 0.0004). Values in parentheses are 5-year actuarial survival rates.
FIG. 4. Actuarial survival by surgical margins after extirpative surgery. Significant difference in survival exists between patients with microscopically negative margins (R0) compared to microscopically positive (R1) or grossly positive (R2) surgical margins (p ⫽ 0.0005). Values in parentheses are 5-year actuarial survival rates. FIG. 2. Difference in actuarial survival between patients with tumors 5 cm. or less and those with greater than 5 cm. (p ⫽ 0.7914). Values in parentheses are 5-year actuarial survival rates.
or the lack of contemporary staging.1, 6, 10, 11 Our series of patients represents a larger group treated at a single institution during a more contemporary period and is limited to prostate sarcoma. Although we did include different histological subtypes of prostate sarcoma in our review, treatment and outcome in all patients are specifically outlined. We also examined whether histological subtype affected survival. The diagnosis of prostate sarcoma is usually accomplished with transrectal ultrasound guided core biopsy, which is well tolerated, requires no anesthesia and often provides adequate tissue to render the histological diagnosis. Most prostate sarcoma causes urinary obstruction,4, 5 and transurethral prostatic resection performed because of symptoms of bladder outlet obstruction may reveal prostate sarcoma. As demonstrated in our series, the PSA value is almost always normal in patients with prostate sarcoma, which is not surprising given the nonepithelial origin of prostate sarcoma. Benign spindle cell nodules of the prostate may result from prior urological procedures, such as prostate needle biopsy or transurethral resection. These lesions could be misdiagnosed
as low grade sarcomatous neoplasms.12 Other benign proliferative prostate lesions can occur in the absence of prior genitourinary surgery, stressing the importance of careful preoperative histological characterization of atypical prostate tumors before extensive pelvic surgery.13, 14 Clinical staging is further enhanced by cross-sectional pelvic imaging. MRI has been the study of choice for extremity sarcoma.15, 16 For large prostate sarcoma we believe MRI facilitates surgical planning with multiplanar images, better delineation of soft tissue planes, and superior assessment of the relationship between tumor and adjacent viscera and neurovascular structures. The overall survival rate for adult patients with prostate sarcoma is poor. The lack of early, specific clinical symptoms result in more advanced disease at presentation. The overall 5-year actuarial survival rate for our patients was 38% compared with a range of 0% to 60% reported in other series.2, 4 – 6, 10 However, data from various series are difficult to compare because of nonuniform reporting on patient and pathological characteristics, changes with time in radiographic staging and the relatively small number of patients with prostate sarcoma. We determined whether various factors, including tumor
525
ADULT PROSTATE SARCOMA
histology, grade, size, AJCC stage and surgical margin status, were predictive of long-term survival. There were no survival differences between patients grouped according to histological subtype (fig. 1). Although there were only 2 patients with low grade prostate sarcoma in our series, we found no survival difference between those with high versus low grade tumors (data not shown). Additionally, patients with tumors 5 cm. or smaller did not have a higher survival rate than those with tumors larger than 5 cm. (fig. 2). We also did not see any significant survival advantage for patients with AJCC stage I or II over those with stage III prostate sarcomas. These findings are in contrast to data presented in the series of Russo et al in which low tumor grade, tumor size 5 cm. or smaller and low stage were favorable predictors of long-term survival of patients with genitourinary sarcoma.5 These differences may be related to the heterogeneous anatomical sites in their report that demonstrated significant survival advantages for patients with bladder and paratesticular sarcomas compared to those with prostate and renal sarcomas. Notwithstanding the limitations of sample size, we do not believe that all genitourinary sarcoma can necessarily be characterized by a single staging system that can predict outcome. In contrast to other clinicopathological factors, surgical margin status does appear to be important for long-term survival. We achieved microscopically negative surgical margins in 79% of our patients who underwent surgery, and the actuarial survival rate was significantly higher than that for those with positive and grossly positive margins (fig. 4). This finding is supported by data from the series by Russo et al in which disease-free survival at 3 years was significantly improved in patients with negative surgical margins.5 If a patient with prostate sarcoma is believed to be at high risk for positive surgical margins preoperative therapy should be considered. Preoperative chemotherapy with subsequent cystoprostatectomy is associated with significant pathological response. In a study of 6 patients with prostate sarcoma Ahlering et al found that a combination of doxorubicin and cisplatin produced greater than 95% necrosis of primary tumors, although there was no evidence of significant radiographic response.10 We found necrosis in the pathological specimens after preoperative chemotherapy as well. However, only patients who received preoperative radiotherapy had appreciable downsizing of the primary tumor (table 4). The prognosis is poor for patients with prostate sarcoma treated with surgery alone, so multimodality therapy should be strongly considered for them. The rare occurrence of adult prostate sarcoma has limited the critical evaluation of treatment strategies and patient survival to anecdotal experiences. However, our results suggest that there may be a survival advantage for patients treated with a combined multimodality approach. Insight into multimodality treatments likely to be more effective for patients with prostate sarcoma can be gained from existing protocol based strategies for other sarcomas. For example, a current protocol for retroperitoneal and pelvic sarcomas incorporates preoperative chemoradiation followed by surgical resection and intraoperative radiotherapy.17 Innovative, combined multimodality treatment strategies may help to improve prognosis for patients with these generally poor prognosis tumors. A final revelation from our review is that patients need close long-term surveillance to detect recurrent prostate sarcoma. Late local recurrences of soft tissue sarcoma are not uncommon and are significantly associated with microscopically positive surgical margins.18, 19 Multiple metachronous local recurrences amenable to surgical resection developed in 2 patients in our series, and 1 survived and is free of disease more than 16 years after initial diagnosis. To our knowledge, our series is the first to demonstrate such a role for local salvage surgery for management of prostate sarcoma. To
detect late recurrences, we recommend that patients with no evidence of disease after treatment be followed closely with chest, abdominal and pelvic imaging. CONCLUSIONS
Prostate sarcomas are rare tumors and often extremely large at initial patient presentation. The overall survival rate remains poor regardless of initial tumor size, grade or histological subtype. Complete surgical resection, including microscopically negative margins, is almost always required to cure patients, and achieving microscopically negative surgical margins appears critical for long-term survival. Long-term surveillance is necessary to detect recurrences. Survival of patients with isolated local recurrences may be prolonged with salvage surgery. REFERENCES
1. Tetu, B., Srigley, J. R. and Bostwick, D. G.: Soft tissue tumors. In: Pathology of the Prostate. Edited by D. G. Bostwick. New York: Churchill-Livingstone, p. 117, 1990 2. Morse, M. J. and Whitmore, W. F.: Genitourinary sarcomas in adults. In: Surgical Management of Soft Tissue Sarcomas. Edited by M. H. Shiu and M. F. Brennan. Philadelphia: Lea and Febiger, pp. 189 –207, 1989 3. Quinlan, D. M., Stutzman, R. E., Peters, C. A. et al: Unilateral nerve-sparing radical prostatectomy and hemicystectomy in management of prostate sarcoma. Urology, 41: 308, 1993 4. Cheville, J. C., Dundore, P. A., Nascimento, A. G. et al: Leiomyosarcoma of the prostate: report of 23 cases. Cancer, 76: 1422, 1995 5. Russo, P., Brady, M. S., Conlon, K. et al: Adult urological sarcoma. J Urol, 147: 1032, 1992 6. Narayana, A. S., Loening, S., Weimar, G. W. et al: Sarcoma of the bladder and prostate. J Urol, 119: 72, 1978 7. American Joint Committee on Cancer: Soft tissue sarcoma. In: American Joint Committee on Cancer Staging Manual, 5th ed. Edited by I. D. Fleming, J. S. Cooper, D. E. Henson et al. Philadelphia: Lippincott-Raven, pp.149 –156, 1997 8. UICC International Union Against Cancer: Urological tumors. In: TNM Classification of Malignant Tumors, 5th ed. Edited by L. H. Sobin and Ch. Wittekind. New York: Wiley-Liss, pp. 165–194, 1997 9. Kaplan, E. L. and Meier, P.: Nonparametric estimation from incomplete observations. J Amer Stat Ass, 53: 457, 1958 10. Ahlering, T. E., Weintraub, P. and Skinner, D. G.: Management of adult sarcomas of the bladder and prostate. J Urol, 140: 1397, 1988 11. Smith, B. H. and Dehner, L. P.: Sarcoma of the prostate gland. Am J Clin Pathol, 58: 43, 1972 12. Huang, W. L., Ro, J. Y., Grignon, D. J. et al: Postoperative spindle cell nodule of the prostate and bladder. J Urol, 143: 824, 1990 13. Gaudin, P. B., Rosai, J. and Epstein, J. I.: Sarcomas and related proliferative lesions of specialized prostatic stroma: a clinicopathologic study of 22 cases. Am J Surg Pathol, 22: 148, 1998 14. Cespedes, R. D., Lynch, S. C. and Grider, D. J.: Pseudosarcomatous fibromyxoid tumor of the prostate: a case report with review of the literature. Urol Int, 57: 249, 1996 15. Chang, A. E., Matory, Y. L., Dwyer, A. J. et al: Magnetic resonance imaging versus computed tomography in the evaluation of soft tissue tumors of the extremities. Ann Surg, 205: 340, 1987 16. Hanna, S. L. and Fletcher, B. D.: MR imaging of malignant soft-tissue tumors. Magn Reson Imaging Clin North Am, 3: 629, 1995 17. Pisters, P. W. T., Patel, S., Crane, C. et al: Phase-I trial of preoperative doxorubicin-based concurrent chemoradiation and electron-beam intraoperative radiation therapy (IORT) for respectable retroperitoneal sarcomas. Proc Am Soc Clin Oncol, suppl., 19: 558a, 2000 18. Lewis, J. J., Leung, D., Casper, E. S. et al: Multifactorial analysis of long-term followup (more than 5 years) of primary extremity sarcoma. Arch Surg, 134: 190, 1999 19. Fleming, J. B., Berman, R. S., Cheng, S. C. et al: Long-term outcome of patients with American Joint Committee on Cancer Stage IIB extremity soft tissue sarcomas. J Clin Oncol, 17: 2772, 1999