PMCCPMCCPMCC

Search tips
Search criteria 

Advanced

 
Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
 
Gynecol Oncol. Author manuscript; available in PMC 2013 October 1.
Published in final edited form as:
PMCID: PMC3787514
NIHMSID: NIHMS467939

Phase II trial of adjuvant pelvic radiation “sandwiched” between ifosfamide or ifosfamide plus cisplatin in women with uterine carcinosarcoma[shadowed white star],[shadowed white star][shadowed white star],[large star]

Abstract

Objective

Uterine carcinosarcoma (CS) is a rare uterine tumor with an extremely poor prognosis. In the adjuvant setting, efficacy has been shown with radiotherapy (RT), systemic chemotherapy, or both. This is the first report describing the efficacy and toxicity of adjuvant ifosfamide or ifosfamide plus cisplatin “sandwiched” with RT in patients with surgically staged and completely resected uterine carcinosarcoma.

Methods

Women with surgically staged CS with no gross residual disease were initially administered ifosfamide (1.2 g/m2/day × 5 days) with cisplatin (20 mg/m2/day × 5 days) every 3 weeks for 3 cycles followed by pelvic external beam RT and brachytherapy followed by 3 additional cycles of ifosfamide (1.0 g/m2/day) with cisplatin (20 mg/m2/day × 5 days) every 3 weeks. Similar to the GOG trial in recurrent CS (Sutton et al, 2000), the addition of cisplatin added toxicity without additional efficacy, so mid-study, the cisplatin was eliminated from the regimen. Toxicities were recorded and disease-free survival (DFS) was calculated with Kaplan-Meier statistical methods.

Results

In total, 12 patients received ifosfamide and cisplatin and 15 patients received ifosfamide alone, both ‘sandwiched’ with RT. The median follow up was 35.9 months (range 6–88). The 2 year DFS was similar in both the ifosfamide/cisplatin and ifosfamide groups (log-rank p = 0.16), so they were combined for analysis. 19 patients (70%) completed the protocol. As expected, stage 1 patients had a better 2-year DFS (18.75 ± 1.12 months; log-rank p = 0.008 when compared to stages 2, 3, 4). Also, in stages 2, 3 and 4 patients, the DFS was 15.81 ± 1.73 months. Grade 3/4 neutropenia, anemia and thrombocytopenia occurred in 18%, 4% and 4% of cycles, respectively.

Conclusions

Ifosfamide “sandwiched” with RT appears to be an efficacious regimen for surgically staged CS patients with no residual disease, even in patients with advanced stage. The addition of cisplatin to the regimen added toxicity without improving efficacy. Even with ifosfamide alone, the efficacy of this ‘sandwich’ regimen comes with a moderate but tolerable toxicity profile.

Keywords: Uterine cancer Carcinosarcoma Adjuvant therapy Radiation therapy Ifosfamide Phase II trial

Introduction

Uterine carcinosarcoma (CS) represents approximately 4% of primary uterine malignancies and 26% of uterine cancer deaths [1,2]. It is an aggressive uterine tumor with a poor prognosis and a high recurrence rate [3]. Patterns of spread are both by hematogenous and lymphatic dissemination which explains the pelvic and extrapelvic sites of failure, with most patients developing extrapelvic disease [4]. 36% of CS patients with disease clinically confined to the uterus have nodal metastasis at the time of diagnosis and the majority of these patients will die of metastatic disease within 2 years of the diagnosis [5].

Even in patients with an early stage who receive adjuvant treatment, pelvic failure is common. Chi et al. evaluated patients with stage I or II CS who were treated with surgery versus surgery followed by whole pelvic radiation therapy. There was a trend towards fewer pelvic recurrences in the patients who received radiation therapy compared to those who did not (p = 0.09) [6]. Other studies showed similar statistically significant lower rates of pelvic recurrence in patients who received radiation therapy compared to those who did not (28% versus 48% (p = 0.0002) [7] and 3% versus 55% (p < 0.0001) [8]). Although pelvic radiation has been shown to decrease the rate of pelvic recurrence, it has not been shown to improve the overall survival [6,7,914]. This finding is likely attributable to the high incidence of distant metastasis known to occur at the time of CS recurrence. In a study of CS patients who received a combination of chemotherapy and radiation therapy in an adjuvant setting, there was a longer median disease specific survival (p < 0.001) and disease free survival (p = 0.02) [15].

Many chemotherapeutic agents that have demonstrated therapeutic efficacy in uterine CS include adriamycin, cisplatin, ifosfamide and paclitaxel [1620]. The use of chemotherapy in the adjuvant setting has been explored as a means of attempting to impact subclinical extrauterine metastasis. The Gynecologic Oncology Group (GOG) evaluated ifosfamide and cisplatin in the adjuvant setting in patients with completely resected Stage I or II CS. At a minimum of 2 years of follow-up, 69% were progression free and 82% remained alive [21]. However, 12 out of 23 patients recurred in the pelvis. The authors concluded that this regimen, in the 4-day schedule, is tolerable and the impact on progression free and overall survival was difficult to evaluate in the absence of simultaneous controls but were better than historical controls. The high incidence of pelvic failures suggested that a combined approach with chemotherapy and radiotherapy should be evaluated. The results from this trial led to a randomized trial of ifosfamide versus ifosfamide with cisplatin in patients with advanced, persistent, recurrent CS with measurable disease after primary surgery [20]. This regimen was associated with significant toxicity necessitating a 20% dose reduction of ifosfamide with the same dose of cisplatin during the course of the trial. Results from this study identified a statistically significant progression free interval (p = 0.02) in the ifosfamide/cisplatin group (6 months) compared to the ifosfamide alone group (4 months). However, there was no significant difference in overall survival (p = 0.07) as well as considerable additional toxicity. In a separate GOG CS trial, chemotherapy was compared to whole abdominal irradiation (GOG 150) [22]. After adjusting for stage and age the estimated death rate was 29% lower in the chemotherapy group than the WAI group. This difference was seen despite the administration of only 3 cycles of chemotherapy. The WAI group had more abdominal recurrences and the chemotherapy group had more vaginal recurrences, but these were not statistically significant.

“Sandwich” sequencing allows for treatment of systemic disease with chemotherapy while controlling micrometastatic disease in the pelvis with radiation therapy. In addition, this treatment allows for the maximum therapeutic dosing for both chemotherapy and radiation therapy while limiting the overall toxicity. Our group has had experience with combining sequential radiation therapy sandwiched with paclitaxel and carboplatin chemotherapy in uterine papillary serous carcinoma (UPSC) [23,24]. Patients with stages I–IV UPSC, without evidence of gross residual disease received adjuvant paclitaxel/platinum for three cycles, followed by external beam radiotherapy (EBRT) and brachytherapy and then three more cycles of paclitaxel/platinum. In this trial, the 3-year % survival probability for stage I/II patients was 84% and stage III/IV patients was 50%. The vast majority of the patients completed the prescribed protocol and most of the toxicities were self-limiting. Grade 3 hematologic toxicities occurred in 14% of cycles and grade 4 hematologic toxicities occurred in 13% of cycles. This regimen was considered highly efficacious and well tolerated in patients with completely resected stage I–IV UPSC. The objective of this study was to evaluate the toxicity and efficacy of pelvic radiation “sandwiched” between cycles of ifosfamide alone or ifosfamide and cisplatin in patients with CS, including defining patterns of recurrence.

Patients and methods

After IRB approval, eligible patients with surgically staged I–IV CS without evidence of gross residual disease after primary surgery were recruited from 1999 to 2009 to this registered phase II trial (clinicaltrials.gov identifier: NCT00231842). Eligible patients underwent surgical staging comprised of total abdominal hysterectomy, bilateral salpingo-oophorectomy, peritoneal cytology, lymph node sampling, and omental biopsy, when clinically indicated. All eligible patients were ≥ 18 years of age, had an ECOG performance status of 0 or 1, adequate hematologic function (hematocrit ≥ 30%, WBC ≥ 300/mm3, platelet count ≥ 100,000/mm3), BUN ≤ 25 mg%, creatinine ≤ 2 mg%, total bilirubin ≤ 1.5 mg/dl, aminotransferases ≤ 2.5 times the institutional upper limit of normal. Patients with concurrent medical conditions limiting their life expectancy to ≤ 3 months or those who received prior chemotherapy and/or RT for pelvic malignancy were excluded.

Pre-treatment evaluation and follow up

At screening, all patients had protocol-required lab testing, including tumor markers, EKG, chest X-ray, and CT scan of the chest, abdomen, and pelvis. In addition to a physical examination, complete blood counts, serum electrolytes, and markers were performed prior to each cycle of chemotherapy, after completion of chemotherapy, every 3 months for 24 months, and then every 6 months thereafter. Imaging was repeated after treatment and during follow-up as clinically indicated. Adverse events were monitored for each cycle during therapy and during follow-up and graded using the National Cancer Institute Common Toxicity Criteria (CTC) version 3.0 as this trial was initiated in 1999, prior to instituting NCI CTC v4.0.

Treatment plan

Chemotherapy

At initiation of this protocol, cisplatin at 20 mg/m2/day with ifosfamide 1.2 g/m2/day for 5 days was administered. Similar to what was found in the GOG trial in recurrent CS [20], in the initial phase of this trial’s two-stage design, the addition of cisplatin added toxicity without additional efficacy, so cisplatin was dropped in the remainder of the recruited patients [25]. The ifosfamide and RT dosing remained the same. Ifosfamide 1.2 g/m2/day was administered parenterally for 5 days every 21 days for 3 cycles prior to RT and 1.0 g/m2/day for 5 days every 21 days for 3 cycles after RT. EBRT generally lasted for 5 weeks, depending on the treatment fields and boosts, when appropriate. This was followed by brachytherapy in all patients. To avoid chemotherapy delay, the fourth cycle of chemotherapy was initiated the same week as the last brachytherapy insertion. Standard premedications for nausea were given. Mesna 400 mg IV bolus followed by 1200 mg IV divided into 3 L/day × 5 days for cycles 1 through 3, and 333 mg IV bolus followed by 1000 mg IV divided into 3 L/day × 5 days for cycles 4 through 6 was administered sequentially with ifosfamide.

Prior to each subsequent cycle of chemotherapy, patients were required to have recovered to an ANC ≥ 1500/mm3 or WBC ≥ 3000/mm3, platelets ≥ 100,000/mm3 and renal and hepatic parameters the same as for screening. Treatment modifications for hematologic toxicities included cycle delay until recovery with subsequent dose reduction, in addition to G-CSF and/or erythropoietin, when clinically appropriate. All toxicities, dose delays, and dose reductions were recorded.

Radiation

EBRT was delivered after the 3rd cycle of chemotherapy. The total dose of EBRT was 45 Gy over 5 weeks. Patients were treated once a day, 5 days a week, with a daily fraction of 1.8 Gy. Four-field technique (AP-PA opposed and lateral opposed fields) was used with a megavoltage beam of ≥ 6 MV. Extended field radiation to the para-aortic nodes was administered in the case of two or more positive pelvic nodes or documented para-aortic lymph node metastasis.

High dose radiation (HDR) brachytherapy was prescribed to the proximal 2/3 of the vagina using the nucleotron microselectron remote after loading technique. Three fractions of 5 Gy each to 5 mm depth from the vaginal cylinder surface was delivered. HDR was given once a week for 3 weeks. The vaginal surface dose was calculated at the vaginal surface lateral to the midpoint of the surface of the cylinder.

Statistical considerations

Response was evaluated using an endpoint of 2-year disease-free survival. Disease free survival (DFS) was calculated from the date of study entry to the date of recurrence of disease. Disease free survival curves were constructed using the Kaplan–Meier (KM) method. For added power, stages I and II, or early stage patients, were combined as well as stages III and IV patients with completely resected metastatic disease. Site(s) and date of relapse were recorded and defined as pelvic or distant. Pelvic sites were specified as vaginal or other, and distant sites were specified as to their anatomic location. Relapse was confirmed by histologic or cytologic evaluation when possible. Frequencies for toxicity and adverse events were also recorded and tabulated.

Results

A total of 30 patients were enrolled with three patients withdrawing from the protocol prior to proceeding with therapy (see Fig. 1). Twenty-seven subjects received the first three cycles of chemotherapy and RT and were included in the analysis. The initial 12 subjects received cisplatin and ifosfamide until the protocol was amended mid-trial due to the published mature results of the GOG trial and observed significant grades 3 and 4 toxicities with this combination [20]. The remaining 15 subjects received ifosfamide alone in a “sandwich” regimen. Of the 27 subjects, one was not a prescribed brachytherapy. Patient demographics and stage are listed in Table 1. The mean age was 65 years (range 41–80 years). The mean BMI was 32.6 (range 21–50).

Fig. 1
Treatment numbers on protocol. Toxicity analysis performed on the 27 patients who received at least one dose of chemotherapy. Survival analysis was performed in all patients who received the first 3 cycles of chemotherapy followed by RT.
Table 1
Patient characteristics.

Toxicity

Out of the 27 subjects who received chemotherapy and RT, 19 (70%) completed all protocol-prescribed treatment. Four were discontinued from further chemotherapy after RT after being dose reduced to the lowest level due to grade 3/4 hematologic toxicities. Two declined further chemotherapy after RT. Two subjects progressed during therapy. A total of 138 out of planned 162 cycles (85%) were administered. Table 2 is a summary of the hematologic and non-hematologic toxicities from the 27 evaluable subjects. Grade 3/4 neutropenia, anemia and thrombocytopenia occurred in 18%, 4% and 4% of cycles, respectively. Of note, 29/39 (74.4%) of grade 3 and 4 hematologic toxicities occurred prior to RT. Most of these toxicities occurred in the subjects who received cisplatin and ifosfamide. Table 3 summarizes the number of dose reductions and cycle delays. 7.2% of cycles required a dose reduction and 17.4% of cycles were delayed. Three out of the 24 cycles that were delayed were not due to treatment-related toxicities.

Table 2
Summary of grades 3 and 4 hematologic and non-hematologic toxicities (n = 138 total cycles).
Table 3
Number of dose reductions and delays by cycle (n = 138 total cycles).

Survival

In the 27 evaluable subjects, the median follow up time was 34.2 months (range 6.4–88 months). For 2-year DFS the median follow-up time was 19.6 months (range 6.4–24 months). Two subjects were lost to follow up after at least 14 months from initiation of treatment. The 2-year DFS was similar in both the cisplatin/ifosfamide and the ifosfamide groups (p = 0.16) so both groups were combined for analysis. During the study and post-treatment follow-up period, 10/27 (37%) patients recurred, 3/17 (17.6%) of stage I/II and 7/10 (70%) of stage III/IV patients. There were no preferential patterns of failure. Of the 10 subjects that recurred, 6 had a distant recurrence (4 in the lung, and 2 in the abdomen), and 4 recurred in the pelvis (2 were nodal recurrences). Fig. 2 shows the survival curve stratified by early stage (stage I/II) and advanced stage (stage III/IV) patients. The 2-year DFS time for early stage patients was 18.07±1.21 (mean ± standard error) months and for advanced stage patients was 15.72±1.79 months. The survival probability at 2 years was 80.8% in stage I/II and 30.3% in stage III/IV patients.

Fig. 2
Kaplan-Meier survival curve of early stage (stage I and II) and late stage (stage III and IV) uterine CS patients treated with radiation “sandwiched” between three cycles of cisplatin/ifosfamide or ifosfamide chemotherapy before and after ...

Discussion

There still is no clear consensus on the best adjuvant therapy for patients with uterine CS, where survival is exceedingly poor. While combined chemotherapy regimens which include cisplatin, ifosfamide, and paclitaxel have been found to be efficacious in a treatment setting for uterine CS [1720], there are few prospective studies where therapy has been used in an adjuvant setting. This study supports that combination chemotherapy and radiotherapy in a ‘sandwich’ fashion has efficacy in the adjuvant treatment of uterine CS. Toxicity with single agent ifosfamide was moderate and mostly self-limiting. However, due to the need for inpatient administration because of the concomitant parenteral Mesna as well as the overall toxicity profile of this combined therapy, other chemotherapeutic agents, such as paclitaxel and carboplatin “sandwich” therapies should be explored. Our group has been using paclitaxel and carboplatin “sandwiched” with RT in a similar rare uterine cancer, uterine papillary serous carcinoma [23,24]. This sequence was shown to be both efficacious and tolerable in UPSC.

Combined paclitaxel and ifosfamide in recurrent or persistent uterine CS was evaluated in GOG 161 which showed improved overall survival, 14 months versus 8 months with ifosfamide alone [26]. The response rate was 45% compared to 29% with ifosfamide alone but the toxicity was worse when combined with paclitaxel. In our study, we found the same additional toxicity when adding cisplatin to ifosfamide.

Data from our prospective trial support the benefit of combined chemotherapy and RT for completely resected CS in the adjuvant setting. The 2-year DFS was 18.07 ± 1.21 months for stage I/II patients and 15.72±1.79 months for stage III/IV patients. The most common toxicities were mostly self-limiting grades 3 and 4 hematologic toxicities, the majority of which occurred prior to RT. RT did not appear to add more post-RT hematologic toxicity, despite the added marrow exposure. Similar to historical data, early stage (stage I/II) patients had better progression free survival compared to advanced stage (stage III/IV) patients.

Although our study was a prospective trial, there are limitations. There is potential for bias that is attributed to any single institution trial. Due to the relative paucity of CS, our sample size is small, but sizable considering a single institution study. Our favorable findings warrant a larger study with a “sandwich” regimen with chemotherapy that is easier to use with a better toxicity profile.

Footnotes

[shadowed white star]Funding: This project was funded, in part, by the Albert Einstein Cancer Center NIH/NCIP30CA013330.

[shadowed white star][shadowed white star]Clinicaltrial.gov identifier: NCT00231842.

[large star]A portion of this data was presented at the Society of Gynecologic Oncologists (SGO) 2009 Annual Meeting in San Antonio, TX.

Conflict of interest statement

The authors declare that there are no conflicts of interest

References

1. Sherman ME, Devesa SS. Analysis of racial differences in incidence, survival, and mortality for malignant tumors of the uterine corpus. Cancer. 2003 Jul 1;98(1):176–86. [PubMed]
2. Nordal RR, Thoresen SO. Uterine sarcomas in Norway 1956–1992: incidence, survival and mortality. Eur J Cancer. 1997 May;33(6):907–11. [PubMed]
3. Major FJ, Blessing JA, Silverberg SG, Morrow CP, Creasman WT, Currie JL, et al. Prognostic factors in early-stage uterine sarcoma. A Gynecologic Oncology Group study. Cancer. 1993 Feb 15;71(4 Suppl):1702–9. [PubMed]
4. Rose PG, Piver MS, Tsukada Y, Lau T. Patterns of metastasis in uterine sarcoma. An autopsy study. Cancer. 1989 Mar 1;63(5):935–8. [PubMed]
5. Chen SS. Propensity of retroperitoneal lymph node metastasis in patients with stage I sarcoma of the uterus. Gynecol Oncol. 1989 Feb;32(2):215–7. [PubMed]
6. Chi DS, Mychalczak B, Saigo PE, Rescigno J, Brown CL. The role of whole-pelvic irradiation in the treatment of early-stage uterine carcinosarcoma. Gynecol Oncol. 1997 Jun;65(3):493–8. [PubMed]
7. Callister M, Ramondetta LM, Jhingran A, Burke TW, Eifel PJ. Malignant mixed Mullerian tumors of the uterus: analysis of patterns of failure, prognostic factors, and treatment outcome. Int J Radiat Oncol Biol Phys. 2004 Mar 1;58(3):786–96. [PubMed]
8. Gerszten K, Faul C, Kounelis S, Huang Q, Kelley J, Jones MW. The impact of adjuvant radiotherapy on carcinosarcoma of the uterus. Gynecol Oncol. 1998 Jan;68(1):8–13. [PubMed]
9. Le T. Adjuvant pelvic radiotherapy for uterine carcinosarcoma in a high risk population. Eur J Surg Oncol. 2001 Apr;27(3):282–5. [PubMed]
10. Hornback NB, Omura G, Major FJ. Observations on the use of adjuvant radiation therapy in patients with stage I and II uterine sarcoma. Int J Radiat Oncol Biol Phys. 1986 Dec;12(12):2127–30. [PubMed]
11. Nielsen SN, Podratz KC, Scheithauer BW, O’Brien PC. Clinicopathologic analysis of uterine malignant mixed mullerian tumors. Gynecol Oncol. 1989 Sep;34(3):372–8. [PubMed]
12. Perez CA, Askin F, Baglan RJ, Kao MS, Kraus FT, Perez BM, et al. Effects of irradiation on mixed mullerian tumors of the uterus. Cancer. 1979 Apr;43(4):1274–84. [PubMed]
13. Salazar OM, Bonfiglio TA, Patten SF, Keller BE, Feldstein ML, Dunne ME, et al. Uterine sarcomas: analysis of failures with special emphasis on the use of adjuvant radiation therapy. Cancer. 1978 Sep;42(3):1161–70. [PubMed]
14. Spanos WJ, Jr, Peters LJ, Oswald MJ. Patterns of recurrence in malignant mixed mullerian tumor of the uterus. Cancer. 1986 Jan 1;57(1):155–9. [PubMed]
15. Gonzalez Bosquet J, Terstriep SA, Cliby WA, Brown-Jones M, Kaur JS, Podratz KC, et al. The impact of multi-modal therapy on survival for uterine carcinosarcomas. Gynecol Oncol. 2010 Mar;116(3):419–23. [PubMed]
16. Omura GA, Major FJ, Blessing JA, Sedlacek TV, Thigpen JT, Creasman WT, et al. A randomized study of adriamycin with and without dimethyl triazenoimidazole carboxamide in advanced uterine sarcomas. Cancer. 1983 Aug 15;52(4):626–32. [PubMed]
17. Thigpen JT, Blessing JA, Orr JW, Jr, DiSaia PJ. Phase II trial of cisplatin in the treatment of patients with advanced or recurrent mixed mesodermal sarcomas of the uterus: a Gynecologic Oncology Group Study. Cancer Treat Rep. 1986 Feb;70(2):271–4. [PubMed]
18. Sutton GP, Blessing JA, Rosenshein N, Photopulos G, DiSaia PJ. Phase II trial of ifosfamide and mesna in mixed mesodermal tumors of the uterus (a Gynecologic Oncology Group study) Am J Obstet Gynecol. 1989 Aug;161(2):309–12. [PubMed]
19. Curtin JP, Blessing JA, Soper JT, DeGeest K. Paclitaxel in the treatment of carcinosarcoma of the uterus: a gynecologic oncology group study. Gynecol Oncol. 2001 Nov;83(2):268–70. [PubMed]
20. Sutton G, Brunetto VL, Kilgore L, Soper JT, McGehee R, Olt G, et al. A phase III trial of ifosfamide with or without cisplatin in carcinosarcoma of the uterus: a Gynecologic Oncology Group Study. Gynecol Oncol. 2000 Nov;79(2):147–53. [PubMed]
21. Sutton G, Kauderer J, Carson LF, Lentz SS, Whitney CW, Gallion H. Adjuvant ifosfamide and cisplatin in patients with completely resected stage I or II carcinosarcomas (mixed mesodermal tumors) of the uterus: a Gynecologic Oncology Group study. Gynecol Oncol. 2005 Mar;96(3):630–4. [PubMed]
22. Wolfson AH, Brady MF, Rocereto T, Mannel RS, Lee YC, Futoran RJ, et al. A gynecologic oncology group randomized phase III trial of whole abdominal irradiation (WAI) vs. cisplatin-ifosfamide and mesna (CIM) as post-surgical therapy in stage I–IV carcinosarcoma (CS) of the uterus. Gynecol Oncol. 2007 Nov;107(2):177–85. [PMC free article] [PubMed]
23. Fields AL, Einstein MH, Novetsky AP, Gebb J, Goldberg GL. Pilot phase II trial of radiation “sandwiched” between combination paclitaxel/platinum chemotherapy in patients with uterine papillary serous carcinoma (UPSC) Gynecol Oncol. 2008 Jan;108(1):201–6. [PubMed]
24. Einstein MH, et al. Phase II trial of adjuvant pelvic radiation “sandwiched: between combination paclitaxel and carboplatin in women with uterine papillary serous carcinoma. Gynecol Oncol. in press. [PMC free article] [PubMed]
25. Einstein MH, et al. Phase II efficacy and toxicity of sequential ifosfamide or ifosfamide plus cisplatin “sandwiched” with radiation therapy as adjuvant therapy for patient s with surgically-staged carcinoma; Poster presentation to Society of Gynecologic Oncologists Annual Meeting; San Antonio, TX. 2009.
26. Homesley HD, Filiaci V, Markman M, Bitterman P, Eaton L, Kilgore LC, et al. Phase III trial of ifosfamide with or without paclitaxel in advanced uterine carcinosarcoma: a Gynecologic Oncology Group Study. J Clin Oncol. 2007 Feb 10;25(5):526–31. [PubMed]