Uterine Sarcoma INTRODUCTION — Sarcomas are a rare form of uterine cancer, comprising fewer than 4 percent of uterine malignancies and occurring in 17 per million women annually. They can arise from the endometrial lining of the uterus or the myometrium. Compared to the more common types of endometrial cancer, women with uterine sarcomas have a poorer prognosis because of the aggressive nature of this tumor. EPIDEMIOLOGY — Uterine sarcomas occur primarily in women 40 to 60 years of age. The rarity of these tumors has prevented the performance of large epidemiologic studies to identify risk factors. Based upon the available United States data, there is an approximately two- to three-fold increase in risk among black compared to white women. A history of pelvic irradiation also seems to be a risk factor, noted in 5 to 10 percent of patients with this disease. Data regarding parity and time of menarche and menopause as risk factors are inconclusive. An increase in the risk of uterine sarcomas (particularly mixed mullerian tumors) appears to accompany the use of long-term adjuvant tamoxifen in women with breast cancer. CLASSIFICATION — The histologic classification of uterine sarcomas is based upon the type of cancerous cell and its presumed site of origin. In general, uterine sarcomas of the homologous type arise from native elements (eg, endometrial stromal sarcoma, leiomyosarcoma and sarcomas of nonspecific supporting tissue such as fibrous tissue, vessels, or lymphatics). In contrast, heterologous tumors contain elements with nonnative differentiation (eg, skeletal muscle, cartilage, bone). The International Society of Gynecologic Pathologists classifies uterine sarcomas according to whether the tumor is purely nonepithelial or mixed epithelial-nonepithelial A somewhat different scheme is used by the Gynecologic Oncology Group (GOG), which classifies uterine sarcomas into five basic categories: Mixed homologous mullerian sarcomas Mixed heterologous mullerian sarcomas Leiomyosarcoma Endometrial stromal tumor Other uterine sarcomas Mixed mullerian tumors (carcinosarcoma) — Malignant mixed mullerian tumors (MMMTs, now called carcinosarcomas) are the most common type of uterine sarcoma, in one series representing 43 percent (51/119) of all cases. Both carcinomatous and sarcomatous elements must be present. However, both cell types are thought to arise from a common progenitor cell, and the sarcomatous element appears to be the result of dedifferentiation. Although any combination is possible, the most common is a mixed homologous carcinosarcoma consisting of serous papillary carcinoma mixed with endometrial stromal sarcoma. Other sarcomatous components in a mixed homologous mullerian sarcoma may resemble a leiomyosarcoma, fibrosarcoma, malignant fibrous histiocytoma, or undifferentiated sarcoma. In contrast, in the mixed heterologous mullerian sarcoma (also termed mixed mesodermal sarcoma), the sarcomatous component is derived from tissue not native to the uterus, and may be a rhabdomyosarcoma, chondrosarcoma, osteosarcoma, or liposarcoma. These tumors tend to form bulky polypoid masses that often fill the uterine cavity and may extend into or through the endocervical canal. It is not uncommon to see obvious myometrial invasion. Carcinosarcomas are aggressive tumors; intraabdominal and retroperitoneal nodal metastases are frequently found. This is not surprising since the epithelial component is usually a papillary serous carcinoma, and predominates within metastatic sites. Consequently, the clinical course of patients with advanced disease mimics that of women with epithelial ovarian cancer. Leiomyosarcoma — Leiomyosarcomas account for one-third of uterine sarcomas. They appear grossly as a large (>10 cm) yellow or tan solitary mass with soft, fleshy cut surfaces exhibiting hemorrhage and necrosis. The intramural mass may bulge into the uterine cavity. Benign leiomyomas and malignant leiomyosarcomas often coexist in the same uterus. However, they are independent entities; leiomyomas probably do not degenerate into leiomyosarcomas. The three main criteria for the diagnosis of leiomyosarcoma are: Frequent mitotic figures. Significant nuclear atypia. Presence of coagulative necrosis of tumor cells Using these criteria, a low or high risk diagnosis indicates a risk of malignancy that is below or above 10 percent, respectively. The smooth-muscle tumors may be categorized into the following groups: (1) Leiomyomas: no atypia or coagulative necrosis and less than or equal to5 mitoses per 10 high power fields (HPF) (2) Leiomyomas with high mitotic activity: no atypia or coagulative necrosis and >5 mitoses per 10 HPF (3) Atypical leiomyoma with low risk: diffuse significant atypia, no coagulative necrosis, and a mitotic count less than 10 mitoses per 10 HPF. They are considered atypical leiomyomas if there is focal significant atypia and no coagulative necrosis regardless of the mitotic count. (4) Leiomyosarcoma: diffuse significant atypia, no coagulative necrosis, and a mitotic count >10 mitoses per 10 HPF or diffuse significant atypia with coagulative necrosis or no significant atypia but with coagulative necrosis and a mitotic count >10 mitoses per 10 HPF. Tumors of the latter type with mitotic counts <10 mitoses per 10 HPF are considered smooth-muscle tumors of low malignant potential. However, some smooth muscle tumors still fall out of this schema (eg, myxoid leiomyosarcomas) Endometrial stromal sarcomas — Endometrial stromal sarcomas arise from the endometrium and infiltrate the myometrium. There are two subtypes: The low-grade subtype is comprised of cells that resemble proliferative endometrial stromal cells that have infiltrated the myometrium, often as worm-like plugs of tumor within the vascular or lymphatic channels. Cytologic atypia is mild to moderate with less than 10 (but usually 0 to 3) mitoses per ten high power fields. They are slow growing tumors with infrequent metastasis or recurrence after therapy. High-grade stromal sarcomas demonstrate moderate to marked cellular atypia with prominent chromatin clumps and nucleoli. They have greater than or equal to10 mitoses per 10 high power fields. These tumors are poorly differentiated, without specific features. Their infiltrating pattern suggests that they may originate from endometrial stromal cells. They enlarge and metastasize quickly and are often fatal. The division of endometrial stromal tumors into low-grade and high- grade subtypes has fallen out of favor. Most high-grade stromal sarcomas are formed of anaplastic cells that show no evidence of endometrial stromal differentiation. Thus, they are better classified as undifferentiated uterine or endometrial sarcomas. The designation of endometrial stromal tumor is now limited to neoplasms that were formally referred to as low-grade stromal sarcomas. In contrast to leiomyosarcoma and carcinosarcoma, blacks are not at higher risk of endometrial stromal tumors than whites. Adenosarcoma — The rare adenosarcoma of the uterus is a mixed tumor in which a benign epithelial component is mixed with a malignant stromal (ie, sarcomatous) element. They present as polypoid masses usually arising from the fundus and generally have low malignant potential. CLINICAL MANIFESTATIONS — Vaginal bleeding is the most common presenting symptom, and occurs in most patients. The amount of bleeding ranges from spotting to menorrhagia. The bleeding may be accompanied by pain, foul smelling vaginal discharge, or symptoms of pelvic pressure. On pelvic examination, the uterus is often enlarged and, in some patients (particularly those with carcinosarcoma), part of the tumor may protrude from the uterine cavity through the cervical os. These symptoms are similar to those seen in women with leiomyomas, and most women with a rapidly enlarging uterus or uterine mass (doubling in size over a period of three to six months) do not have a sarcoma. As an example, in one study of 1332 women admitted to either of two community hospitals for hysterectomy or myomectomy of presumed uterine leiomyomas, the incidence of uterine sarcomas was extremely low (0.23 percent). Among the 341 women with a rapidly growing uterus by clinical or ultrasound examination, only one (0.27 percent) had a uterine sarcoma. Based upon these findings, an increased risk of sarcoma among women with "rapidly growing" leiomyomas could not be substantiated. However, the diagnosis of a uterine sarcoma should be considered in postmenopausal women with a pelvic mass, abnormal bleeding, and pelvic pain, where the incidence of sarcoma is higher (1 to 2 percent). DIAGNOSTIC EVALUATION — Imaging studies and clinical findings are not specific for sarcoma versus other uterine tumors. Ultrasound examination, magnetic resonance imaging, or computed tomography cannot reliably distinguish between a sarcoma and leiomyoma, endometrial cancer, lymphoma, intravenous leiomyomatosis, or adenomyosis. The diagnosis of uterine sarcomas is made from histologic examination of the entire uterus. Endometrial sampling alone can be negative in the absence of endometrial involvement. STAGING Preoperative evaluation — The preoperative evaluation of women with suspected uterine sarcoma includes a complete blood count, electrolytes, renal function, liver function tests, urinalysis, and chest x-ray. Computed tomography (CT) scans may be useful to identify occult extrauterine disease. Procedure — The staging procedure includes a total hysterectomy, usually with bilateral salpingo-oophorectomy, and exploration of the abdomen. Extensive local growth is a hallmark of uterine sarcomas, especially leiomyosarcoma. As is noted for epithelial tumors of the uterus, distant spread may occur by intraabdominal, lymphatic, or hematogenous spread. The staging of uterine sarcomas is the same as that described for endometrial cancer by the International Federation of Gynecologists and Obstetricians (FIGO) briefly: Stage I: disease confined to the uterine corpus Stage II: disease involving the cervix Stage III: regional spread to pelvic organs Stage IV: disease outside of the pelvis. Lymph node biopsy is recommended as part of the staging procedure; however, it is not essential unless required by a treatment protocol. In one series of 208 women with uterine leiomyosarcoma, only four of 36 who underwent lymph node sampling had positive nodes, one of whom had neither palpable nodes nor evidence of extrauterine disease. Although the presence of involved lymph nodes was of prognostic significance, lymphadenectomy was not therapeutic, since outcomes were similar for similarly staged patients who did and did not undergo lymphadenectomy. In addition, a literature review of patients with leiomyosarcoma grossly confined to the uterus/cervix showed lymph node metastases in 5 of 101 patients and ovarian metastases in 4 of 130 patients. The authors concluded lymph node dissection should be reserved for women with clinically suspicious nodes. In their experience, this occurred only in women with extrauterine disease. All uterine sarcomas have the propensity for hematogenous dissemination. This tends to occur early, and most frequently involves lung. Thus, all patients should undergo staging thoracic CT scan. TREATMENT — Extirpative surgery is the primary treatment of uterine sarcoma and may be curative for tumors confined to the uterus. Radiation therapy and chemotherapy are used as adjunctive modalities to surgery, and in patients with advanced disease. Surgery — Extrafascial abdominal hysterectomy with bilateral salpingo-oophorectomy is the only curative therapy for uterine sarcomas. There are no additional benefits from preoperative radiation, radical surgery, or optimal cytoreduction. Carcinosarcomas are more likely to be associated with intraabdominal and retroperitoneal metastases than are leiomyosarcomas. This was illustrated by a study of 62 patients with carcinosarcoma that appeared to be confined to the uterus in whom complete surgical staging revealed occult metastases (adnexae, pelvic and paraaortic nodes, omentum) in 60 percent. Thus, complete surgical staging of patients with carcinosarcoma may be useful in individualizing the use of adjuvant radiation therapy and/or chemotherapy. These issues are addressed in detail below. By comparison, studies of patients with leiomyosarcoma apparently confined to the uterus have reported a much lower risk of occult adnexal and nodal disease, generally <5 percent. Adjuvant therapy — Three types of adjuvant (postoperative) therapy have been evaluated in women with uterine sarcoma: radiation therapy (RT), chemotherapy, and combined modality therapy. Adjuvant RT — Although considerable literature exists concerning radiation therapy for early stage uterine sarcoma, much of it is limited by lack of surgical staging, outmoded radiation techniques, and limited followup. Furthermore, the majority of reports fail to distinguish between the various histologic subtypes in their analysis, and include patients with a wide variety of tumor stages. As a result, the benefit of postoperative adjuvant RT in women undergoing surgery for uterine sarcoma is unproven, and its role is controversial. No randomized trials have been reported in final form, although preliminary results are available for one European trial, and a second GOG trial focusing on carcinosarcoma is nearing the end of accrual. Randomized trials — In the only randomized trial that has completed accrual, from the European Organization for Research and Treatment (EORTC) randomly assigned 222 women with resected surgical stage I or II uterine sarcomas (103 leiomyosarcomas, 91 carcinosarcomas, 28 endometrial stromal sarcomas) to pelvic external beam RT or observation. The study closed in 2001, and a preliminary report in 2003 suggested a lower rate of local recurrence in the radiated group but no improvement in survival. The impact on pelvic control was seen only in the carcinosarcoma subtype. GOG trial 150, discussed below, focuses on this subgroup only. Uncontrolled series — The available retrospective reports provide conflicting results. Most studies that group the various histologies together note better pelvic control with adjuvant RT, and in some, a delay in the appearance of metastatic disease; however, survival has not been impacted. As an example, the impact of pelvic RT on outcome in uterine sarcoma was studied in a subset of 109 patients with stage I or II uterine sarcoma enrolled on GOG protocol 20 (a randomized trial of adjuvant doxorubicin), in which the use of pelvic RT was optional. Of the 157 patients with two-year follow-up, 109 (87 percent carcinosarcoma) received RT. The RT group had a significantly lower rate of failure within the radiated field (23 versus 54 percent), but no differences in progression-free interval, overall survival or site of first recurrence. Other retrospective series suggest the possibility of a survival benefit in addition to improvement in local control. The largest series evaluated 103 women with stage I-IVa uterine sarcoma (42 percent leiomyosarcoma, 39 percent carcinosarcoma, 17 percent endometrial stromal tumor, 3 percent miscellaneous), 52 percent of whom received RT at the discretion of their treating physician. Irradiated patients had significantly better five-year pelvic control (76 versus 36 percent), and overall survival (73 versus 37 percent, respectively). The relationship between RT and improved pelvic control as well as survival remained significant in multivariate analysis after controlling for stage, histology, tumor grade and the presence of lymphovascular invasion. As noted previously, interpretation of the possible benefit of RT is hampered by the difficulty in comparing outcomes from series in which patients of varying stages and histologies were reported. Few series have evaluated the benefit of adjuvant RT in homogeneous populations: Endometrial stromal tumors — One series focused on 21 women with endometrial stromal tumor (eleven stage I), 15 of whom received adjuvant RT following curative intent surgery. The five year overall survival rate in this group was 72 percent and the local control rate was 93 percent. The lack of a comparison group not receiving RT hinders the interpretation of these results. However, the authors compiled a summary of published data, with the following cumulative five year survivals by treatment for endometrial stromal tumor: Surgery alone — 12 of 26 (46 percent) Surgery and RT — 24 of 39 (62 percent) Surgery and chemotherapy — 10 of 23 (43 percent) RT alone — 1 of 12 (8 percent) Carcinosarcoma — The fact that few women with carcinosarcoma relapse outside of the abdomen, and the preliminary reports of the EORTC trial discussed above suggests that adjuvant RT may be particularly beneficial for this group. Most studies that focus on carcinomasarcoma report better pelvic control with adjuvant RT. However, as with other uterine sarcomas, impact on survival is mixed; some reports suggest better survival in irradiated patients while most do not. As noted above, a phase III GOG trial (GOG 150) is currently accruing patients with completely resected carcinosarcoma to whole abdomen RT or adjuvant ifosfamide/cisplatin (see below). Eligible patients should be encouraged to enroll. Adjuvant chemotherapy — The high propensity for distant failure among women with uterine sarcomas (particularly leiomyosarcomas) suggests that adjuvant systemic therapy might provide some benefit. Unfortunately, as with adjuvant RT, the rarity of this disease has restricted the ability to conduct randomized controlled trials in this area. Despite the paucity of data supporting benefit, adjuvant chemotherapy protocols are often recommended for patients whose tumors have more than 20 mitoses per high power field, deep myometrial invasion, or extension beyond the uterus. The most commonly used agents are doxorubicin, cisplatin, and ifosfamide. Well performed trials providing clear evidence of benefit for this and other combinations are lacking. Two small randomized trials of surgery with or without postoperative single agent doxorubicin have been published, with different results. A GOG trial compared six months of adjuvant doxorubicin to no postoperative treatment in 156 women with early stage uterine sarcoma (stage I and II). There was no improvement in progression-free interval or survival with the use of doxorubicin. In contrast, benefit from adjuvant doxorubicin was suggested in a second prospective study of women with resected stage I uterine sarcoma who were randomized to surgery alone or followed by doxorubicin. Compared to surgery alone, the chemotherapy treated patients had a higher five year rates of both overall survival (63 versus 36 percent) and recurrence-free survival (RFS, 75 versus 46 percent). Neither of these trials considered the histologic subtype or the uterine sarcoma, a factor known to affect chemotherapy responsiveness. The impact of histologic type on benefit from adjuvant chemotherapy was suggested in a report in which ifosfamide (1.5 g/m2 per day for 3 days, every 28 days) plus mesna was administered to 13 women with completely resected, moderate to high grade uterine sarcoma. For the ten patients with early stage disease, the two year progression-free survival (PFS) was 60 percent, and the two year and three year overall survival was 100 and 67 percent, respectively. Patients with mixed mullerian tumors had a longer time to progression than those with leiomyosarcoma (two year PFS 100 versus 33 percent). Multiagent regimens have also been investigated in the postoperative setting: One study evaluated the impact of adjuvant cyclophosphamide, vincristine, doxorubicin, and dacarbazine (CYVADIC) on PFS in 20 patients with stage I uterine sarcomas. All of the patients underwent surgery followed by nine monthly cycles of chemotherapy. Although there was no control group in this study, the PFS (80 and 65 percent at two and five years) did not appear to be different from historical controls treated without chemotherapy. In a study from the GOG, 65 patients with completely resected stage I (n = 50) or II (n = 15) carcinosarcoma were treated with three postoperative 21-day courses of cisplatin (20 mg/m2 daily for five days) and mesna-supported ifosfamide (1.5 g/m2 daily for five days). The two-year PFS and overall survival rates were 69 and 82 percent respectively, while overall survival at five years was 62 percent. Twelve of 23 first recurrences included pelvic sites. While these results seem better than can be achieved by surgery alone (approximately 50 percent 5-year survival for stage I uterine sarcomas,, the absence of a control group precludes any interpretation of benefit. Thus, as with adjuvant RT, the benefit of chemotherapy will likely remain unclear until randomized controlled trials are completed. Combined modality therapy — At least two nonrandomized studies suggest that postoperative combined RT and chemotherapy may provide benefit after surgical resection. In one report from Turkey, 41 women with uterine sarcoma received either pelvic RT (23 patients) or radiation plus adjuvant chemotherapy (18 patients). Three year survival rates were significantly better in the chemotherapy group (66 versus 36 percent). A second series included 38 women with clinical stage I or II mixed Mullerian tumor, all of whom underwent resection followed by RT (brachytherapy with or without external beam RT) and either four or six cycles of epirubicin (75 mg/m2) plus cisplatin (75 mg/m2). With an average follow-up of 55 months, the overall survival was 74 percent. Twenty of the 21 patients who completed the full course of adjuvant therapy were still alive, and 19 were progression-free. Although intriguing, these data do not prove the benefit of adjuvant combined modality therapy. Recommendation: adjuvant therapy — Based upon limited available data, there appears to be a role for adjuvant RT in enhancing local control following surgery; whether there is a survival benefit from adding RT remains unclear. Nevertheless, adjuvant RT is recommended for women with resected stage I or II uterine sarcomas (except for endometrial stromal sarcomas) in guidelines published by the National Comprehensive Cancer Network (NCCN). The type of RT (ie, external beam, brachytherapy) should be tailored based upon the individual clinical circumstances. There appears to be no role for chemotherapy (single agent or combination) or combined modality therapy in the adjuvant setting of resected uterine sarcoma. Neither of these approaches should be considered standard of care at present. More randomized trials are needed to determine the role of chemotherapy and/or chemoradiotherapy and to determine the optimal regimen that should be used for women with uterine sarcomas. As noted above, Gynecologic Oncology Group (GOG) protocol 150 is accruing women with completely resected surgical stage I or II carcinosarcoma to a randomized trial compared adjuvant whole abdomen RT and chemotherapy; eligible patients should be encouraged to enroll. TREATMENT FOR RECURRENT, ADVANCED OR METASTATIC DISEASE — Recurrent disease is usually established clinically by the new development of symptoms. Most relapses occur in the pelvis, followed by lung and abdomen. Bone and brain metastases are uncommon.Pulmonary relapse may be accompanied by spontaneous pneumothorax. Typically, management of metastatic uterine sarcoma conforms to treatment practice for metastatic soft tissue sarcomas of other sites. Surgical resection may be considered for isolated metastases (particuarly involving lung), radiation therapy is appropriate for symptomatic locoregional relapses, and palliative systemic chemotherapy or hormone therapy are used for all other patients Chemotherapy — There are differences in the chemotherapy responsiveness of metastatic uterine leiomyosarcomas, mixed mullerian tumors, and carcinosarcomas. Leiomyosarcoma — In comparison to its unproven role in the adjuvant setting, single agent doxorubicin is an effective drug for advanced leiomyosarcoma arising in the uterus but not for mixed mullerian tumors. The addition of dacarbazine or cyclophosphamide with or without mitomycin and cisplatin does not add any survival benefit, but does appear to produce higher response rates. Ifosfamide has modest activity; combinations with doxorubicin increase the objective response rate but add substantial toxicity. More recent data suggest a high level of antitumor activity against uterine leiomyosarcomas with gemcitabine plus docetaxel. Patients with unresectable uterine (n = 29) or other primary site (n = 5) leiomyosarcomas received gemcitabine (900 mg/m2 on days 1 and 8) with docetaxel (100 mg/m2 on day 8) and granulocyte colony stimulating factor (G-CSF) on days 9 to 15, repeated every 21 days. The overall response rate was 53 percent, including three complete responders; 50 percent of those previously treated with doxorubicin with or without ifosfamide responded. Despite the use of G-CSF, grade 3 or 4 neutropenia and febrile neutropenia developed in 21 and 6 percent, respectively. Median survival was 17.9 months, which is longer than any reported to date in a phase II trial Antitumor activity for temozolomide was suggested in a small retrospective report of19 patients with uterine leiomyocarcoma, the majority of whom had received two prior chemotherapy regimens. One patient had a prolonged partial response and one a near complete response after 13 months of therapy which persisted for over 10 months following treatment discontinuation. Further experience with this agent is needed. Carcinosarcoma — Cisplatin and ifosfamide appear to have greater activity against mixed (heterologous) mesodermal tumors than does doxorubicin. However, the response rate with monotherapy is still quite low (8 percent complete and 11 percent partial response with cisplatin alone in one study). Response rates may be higher with combinations of cisplatin with doxorubicin and/or ifosfamide, although the available data are conflicting: In an EORTC study, 48 women (32 with measurable persistent or relapsed disease) received doxorubicin (45 mg/m2 on day 1) followed 30 minutes later by cisplatin (50 mg/m2 over 3 hours) then ifosfamide (5 g/m2 in 4 L normal saline over 24 hours with IV mesna 1.25 g/m2 administered prior to, 2.5 g/m2 administered concurrently, and 1.25 g/m2 over 12 hours following ifosfamide), with repeated courses every 21 days x 6. Complete and partial responses were noted in 11 and 7, respectively (overall response rate 56 percent), and the median survival of the entire cohort was 26 months. Treatment-related toxicity was prominent, with severe leukopenia and/or thrombocytopenia requiring dose reduction in 60 percent, and renal dysfunction which was fatal in one patient. The combination of cisplatin, doxorubicin, and dacarbazine was administered to 13 patients with mixed mesodermal sarcomas of the uterus following surgical debulking. Six of 13 (46 percent) remained without evidence of progression for periods of up to 36 months, and the estimated one- and three-year survivals were 68 and 51 percent, respectively. Although uncontrolled, these survival rates are much higher than would be anticipated for patients with unresectable disease. Less favorable results were noted with combined cisplatin (75 mg/m2 on day 1), ifosfamide (1.2 g/m2 over 24 hours, days 1 through 5) and mesna (240 mg/m2 IV prior to and 4 and 8 hours following the start of each dose of ifosfamide) administered in 28 day cycles to 16 women with primary or recurrent mixed mesodermal tumors. There were two partial responses among six patients with measurable disease, only one of whom was still alive at 7.5 months. One patient without measurable disease remained alive 6.5 years after treatment began. Toxicity was prominent, with grade 3 or 4 neutropenia in 36 percent, and grade 1 gastrointestinal toxicity in all patients. Similarly, the benefit of combined cisplatin/ifosfamide was called into question in a randomized trial in which 194 women with advanced, persistent, or recurrent carcinosarcoma were randomly assigned to ifosfamide (1.5 g/m2 daily for four days) with or without cisplatin (20 mg/m2 daily for four days). As expected, toxicity was more severe with combined therapy, and there were 6 treament-related deaths compared to none with ifosfamide alone. The modestly higher objective response rate with combined therapy (54 versus 36 percent) and slight prolongation in progression-free survival (6 versus 4 months, p = 0.02) did not translate into a significant improvement in overall survival. The activity of paclitaxel against carcinosarcomas is uncertain: In one report of 53 women receiving single agent paclitaxel (135 to 170 mg/m2 every three weeks), four complete and four partial responses were noted. On the other hand, the response rate was significantly lower (8 percent) in a GOG trial of 53 women receiving similar doses of paclitaxel for recurrent or persistent uterine leiomyosarcoma. A small study of six patients documented four complete responses and a median progression-free survival of 18 months using 21 day cycles of paclitaxel (175 mg/m2 over 3 hours) and carboplatin (dosed at an area under the concentration x time curve [AUC] = 6). Endometrial stromal tumors — Systemic therapy has not been well studied in women with endometrial stromal sarcoma. Ifosfamide alone has been associated with objective clinical responses in seven of 21women in one report (three complete), but the median progression-free interval was only 3 months Summary and recommendation — Symptomatic patients with advanced unresectable uterine sarcoma should be offered palliative chemotherapy. As with soft tissue sarcomas at other sites, single agent doxorubicin (at least 60 mg/m2 per cycle) is a reasonable option for women with advanced uterine sarcoma who are eligible for chemotherapy bullet The combination of gemcitabine plus docetaxel has been associated with high objective response rate in patients with leiomyosarcoma. However, randomized trials are needed before it can be concluded that higher response rates lead to important differences in priogression-free or overall survival. For patients with carcinosarcoma, single agent ifosfamide, or cisplatin plus ifosfamide with or without doxorubicin also represent appropriate options. However, combination therapy is associated with greater toxicity than monotherapy, and patients should be carefully selected for this approach based upon age, performance status, and comorbidities. Hormone therapy — Estrogen, progesterone, and other hormone receptors are present in LMSs and ESTs; however, their presence does not reliably predict hormone responsiveness. In one report, for example, 48 percent of 60 uterine sarcomas were estrogen receptor positive and 30 percent were progesterone receptor positive. Only one of 28 patients with residual or recurrent disease following surgery had an objective response to hormone therapy. Some authors have recommended that progestin therapy (eg, megestrol acetate) should be considered for adjuvant therapy and for the treatment of recurrent ESTs. Other hormone therapies described as successful in case reports include use of aromatase inhibitors and gonadotropin-releasing hormone analogs Estrogen replacement therapy after oophorectomy or at menopause is not recommended because it may increase recurrences in patients with hormone responsive tumors, especially those with ESTs PROGNOSIS — In general, uterine sarcomas carry a poorer prognosis compared to other gynecologic malignancies. The five year survival rate for stage I ovarian, endometrial, cervical, vaginal, or vulvar cancer is approximately 90 percent. In contrast, stage I uterine sarcomas have a five year survival rate of only 50 percent, falling to 20 to 30 percent when there is spread beyond the uterus (stage III or IV). The prognosis for all histologic types is similar after correcting for stage of disease. Approximately 50 percent of women with stage I sarcoma and 80 to 90 percent with stage II, III, and IV disease will develop either local or distant recurrence. However, tumor stage, grade, and mitotic count are consistently observed to correlate with prognosis. Carcinosarcomas, high-grade leiomyosarcomas, and undifferentiated sarcomas behave aggressively, while low grade leiomyosarcomas, endometrial stromal tumors, and adenosarcomas have a slower pattern of growth. For carcinosarcomas, one Gynecologic Oncology Group (GOG) series of factors that adversely influenced progression-free interval in early stage cases were tumor size and depth of uterine invasion, lymphatic or vascular involvement, histologic grade and cell type, adnexal extension, lymph node metastases, and peritoneal cytologic finding. For leiomyosarcomas, the GOG study found that the mitotic index was the only factor significantly related to the progression-free interval. If the leiomyosarcoma arose in a benign leiomyoma, the prognosis was better. However, others have also noted an association between increasing age and lower survival, particularly for uterine leiomyosarcomas. A study of 209 women with uterine sarcomas suggested that the strongest predictor of longer survival was premenopausal status at time of diagnosis. A series of 208 women with leiomyosarcomas from the Mayo Clinic also noted age over 51 years was a poor prognostic factor, as was size greater than 5 cm