Carcinosarcoma of the ovary

Authors

  • D.-A. SILASI,

    Corresponding author
    1. Division of Gynecology Oncology, Department of Obstetrics, Gynecology, and Reproductive Sciences, Yale University School of Medicine, New Haven, Connecticut
    Search for more papers by this author
  • J.L. ILLUZZI,

    1. Division of Gynecology Oncology, Department of Obstetrics, Gynecology, and Reproductive Sciences, Yale University School of Medicine, New Haven, Connecticut
    Search for more papers by this author
  • M.G. KELLY,

    1. Division of Gynecology Oncology, Department of Obstetrics, Gynecology, and Reproductive Sciences, Yale University School of Medicine, New Haven, Connecticut
    Search for more papers by this author
  • T.J. RUTHERFORD,

    1. Division of Gynecology Oncology, Department of Obstetrics, Gynecology, and Reproductive Sciences, Yale University School of Medicine, New Haven, Connecticut
    Search for more papers by this author
  • G. MOR,

    1. Division of Gynecology Oncology, Department of Obstetrics, Gynecology, and Reproductive Sciences, Yale University School of Medicine, New Haven, Connecticut
    Search for more papers by this author
  • M. AZODI,

    1. Division of Gynecology Oncology, Department of Obstetrics, Gynecology, and Reproductive Sciences, Yale University School of Medicine, New Haven, Connecticut
    Search for more papers by this author
  • P.E. SCHWARTZ

    1. Division of Gynecology Oncology, Department of Obstetrics, Gynecology, and Reproductive Sciences, Yale University School of Medicine, New Haven, Connecticut
    Search for more papers by this author

Dan-Arin Silasi, MD, Division of Gynecologic Oncology, Department of Obstetrics, Gynecology, and Reproductive Sciences, Yale University School of Medicine, 333 Cedar Street, FMB 336, New Haven, CT 06520-8063, USA. Email: dan-arin.silasi@yale.edu

Abstract

The objective of this study was to evaluate the treatment and outcome in patients with ovarian carcinosarcoma. The Tumor Board Registry was reviewed for patients with ovarian carcinosarcoma treated at our institution from June 1993 to December 2004. The medical records were retrospectively analyzed with emphasis on cytoreduction, cytotoxic regimens, progression-free interval, and survival. Twenty-two patients were identified. All but two presented with advanced stage disease. The median survival for the entire cohort was 38 months. Median survival was 46 months for 18 optimally debulked (<1 cm) patients and 27 months for four suboptimally debulked (>1 cm) patients. Six patients were treated with optimal cytoreduction and adjuvant cisplatin (40 mg/m2× 1 day) and ifosfamide (1200 mg/m2/day × 4 days) every 28 days. Median progression-free interval in the cisplatin and ifosfamide group was 13 months, and median survival was 51 months. The combination of carboplatin (AUC 5) and taxol (175 mg/m2) every 21 days was administered to four patients as first-line chemotherapy following optimal cytoreduction. In the carboplatin and taxol group, median progression-free interval was 6 months and median survival was 38 months. The difference in survival between the cisplatin and ifosfamide group and the carboplatin and taxol group was not statistically significant (P= 0.48). In conclusion, patients with ovarian carcinosarcoma usually present with advanced stage disease. Treatment consists of optimal cytoreduction and chemotherapy. The most effective cytotoxic regimen remains to be determined. First-line cisplatin and ifosfamide or carboplatin and taxol can achieve survival rates observed in epithelial ovarian cancer.

Ovarian carcinosarcomas are exceedingly rare malignancies, and in most cases, patients present with advanced disease. Carcinosarcomas of the female genital tract, also known as (malignant) mixed müllerian (mesodermal) tumors, are mixed tumors comprising both malignant epithelial and mesenchymal (stromal) components. They most commonly are identified in the uterus.

Although the evidence for a monoclonal theory of histogenesis is better documented in uterine carcinosarcomas, recent studies propose a similar etiopathogenetic pathway for ovarian carcinosarcomas and the common epithelial cancers(1–3). The findings suggest that the epithelial component initiates tumorigenesis and the sarcomatous component represents a metaplastic phenomenon. Interestingly, the composition of metastatic lesions changes during disease progression, with the sarcomatous component becoming more pronounced(4).

These tumors are defined by aggressive behavior, and survival is significantly reduced when compared to that in epithelial ovarian cancers.

Treatment for advanced disease consists of complete surgical staging and debulking and postoperative adjuvant chemotherapy. Because of the rare occurrence of these tumors, with fewer than 400 cases reported in the English literature(5), no consensus has emerged yet as to what constitutes the most effective chemotherapy.

We report a series of 22 patients with carcinosarcoma of the ovary, the cytotoxic treatments administered, and the outcomes after surgery and chemotherapy.

Materials and methods

The Tumor Board Registry of the Gynecologic Oncology Division was reviewed for ovarian carcinosarcomas diagnosed and treated at our institution between July 1993 and December 2004. The slides were reexamined by a gynecologic pathologist and the diagnoses reconfirmed. Medical charts were available for all patients, with good follow-up. The patients’ records were reviewed with special emphasis on surgical optimal or suboptimal debulking, cytotoxic regimens administered, and survival. Optimal debulking was defined as residual tumor of less than 1 cm in its largest diameter.

Survival was calculated from the date of diagnosis until the date of death or the date of last follow-up for the patients that were still alive with or without evidence of disease.

SAS version 9.1 software was used to compute the statistical data. Survival curves were generated by the Kaplan–Meier method. Survival between groups was compared with the log-rank test. Differences were considered statistically significant for a P value <0.05.

This study protocol (#0510000756) was approved by the Human Investigation Committee of the Yale University School of Medicine.

Results

During the period from 1993 to 2004, 22 patients with ovarian carcinosarcoma were diagnosed and treated at Yale-New Haven Hospital. The age at time of diagnosis ranged from 47 to 89 years, with a median age of 71 years.

The most common presenting complaints were bloating and increasing abdominal girth.

All patients underwent complete surgical staging and debulking. The tumors were staged according to the classification adopted by the FIGO in 1988(6,7). FIGO staging was as follows: one patient was stage IIB, one patient was stage IIC, 15 patients were stage III (two patients with stage IIIA, one patient with stage IIIB, and 12 patients with stage IIIC), and four patients were stage IV. One patient who received neoadjuvant chemotherapy was assigned stage X.

Optimal surgical debulking was achieved in 18 patients (82%), and suboptimal debulking was reported in four cases (18%).

Postoperatively, all but one patient received chemotherapy. Table 1 describes the first- and second-line chemotherapy regimens administered, as well as the total number of treatments with cytotoxic agents.

Table 1.  Clinicopathologic characteristics, treatment, and outcome in 22 patients with carcinosarcoma of the ovary
Patient No.AgeStageDebulkingFirst- and second-line chemotherapyOther chemotherapiesSurvival
  1. AWD, alive with disease; LFU, lost to follow-up; DOD, dead of disease; NED, alive with no evidence of disease; DOC, dead of other causes; POD, progression of disease; PFI, progression-free interval; VP-16, etoposide.

174IIIBOptimal: no visible residual diseaseCisplatin + ifosfamide × six cyclesDoxil46 months DOD/PFI 12 months
 Weekly taxolTamoxifen 
 Oral VP-16 
287IIIAOptimal: no visible residual diseaseCisplatin + ifosfamide × six cycles 49 months DOD/PFI 28 months
 Oral VP-16 
355IIICOptimal: no visible residual diseaseCisplatin + ifosfamide × nine cyclesTaxol + ifosfamide53 months DOD/PFI 0 months
 Carboplatin + taxol × six cyclesDoxil 
 Gemzar 
 Topotecan 
 Oral VP-16 
450IVOptimal: no visible residualCisplatin + ifosfamide × six cycles 52 months NED/PFI 46 months
561IIIAOptimal: no visible residual diseaseCisplatin + ifosfamide × six cyclesCarboplatin + taxol67 months DOD/PFI 14 months
 Carboplatin × three cyclesTaxol 
 Taxotere 
 Topotecan + doxil 
 Gemzar 
 Ifosfamide 
 Adriamycin + dacarbazine 
647IIICOptimal: no visible residualCisplatin + ifosfamide × two cycles 4 months DOD
775IIICSuboptimal: >1 cmCisplatin + ifosfamide × six cycles 28 months NED (LFU)
862IIICOptimal: no visible residual diseaseCarboplatin + taxol × seven cyclesCisplatin + gemzar24 months AWD/PFI 4 months
 Doxil × two cycles 
969IIICOptimal: <1 mmCarboplatin + taxol × six cycles 17 months NED/PFI 11 months
1060IVOptimal: no visible residual diseaseCarboplatin + taxol × nine cycles 18 months AWD/PFI 4 months
 Doxil + gemzar × six cycles 
1150IIICOptimal: <1 cmCarboplatin + taxol × six cyclesDoxil37 months AWD/PFI 8 months
 Cisplatin + gemzar × six cyclesTaxol 
 Oral VP-16 
 Topotecan 
1253IIICSuboptimal: >1 cmCarboplatin + taxol × four cycles PODOral VP-1622 months DOD
 Cisplatin + ifosfamide × six cycles 
1381IICOptimal no visible residualTaxol + carboplatin × six cycles 27 months NED/PFI 22 months
1467XSuboptimal: >1 cmNeoadjuvant carboplatin + taxol × six cyclesOral VP-1632 months DOD
 Cisplatin + ifosfamide × six cyclesTamoxifen 
 Cisplatin 
 Doxil 
1564IIICOptimal: no visible residual diseaseCisplatin + VP-16 + adriamycin × six cyclesTaxol125 months AWD
 Carboplatin + taxol × nine cyclesTamoxifen 
1672IIBOptimal: no visible residualCisplatin + VP-16 × six cycles 60 months NED (LFU)
 Oral VP-16 × six cycles 
1773IVOptimal: no visible residual diseaseTaxol + ifosfamide × six cycles 9 months DOD
 Oral VP-16 
1883IVOptimal: no visible residualCarboplatin + cytoxan × four cycles 7 months DOD
1974IIICOptimal: miliary 3 months DOD
2057IIICOptimal: < 0.5 cmTaxol + ifosfamide × nine cyclesTamoxifen27 months DOD
 Cisplatin + VP-16 + adriamycin × nine cyclesGemzar 
2170IIICSuboptimal: >1 cmDoxil × one cycle 2 months DOC
2289IIICOptimal: no visible residualOral VP-16 × two cycles 9 months DOD

Survival was counted from the date of diagnosis until the patient expired or the date of last follow-up visit.

The overall median survival for all patients in this report was 38 months (95% CI: 22–53 months). The median survival for patients with stage III disease was 38 months (95% CI: 22–53 months). The four patients with stage IV disease had survivals of 7, 9, 18, and 52 months.

Patients with optimal debulking had a median survival of 46 months (95% CI: 9–53 months) as opposed to 27 months in the suboptimally debulked group (95% CI: 2–32 months). However, this difference was not statistically significant (P= 0.24). Figure 1 shows the survival curves for optimally versus suboptimally cytoreduced patients.

Figure 1.

Kaplan–Meier survival curves of patients with ovarian carcinosarcoma who underwent optimal versus suboptimal cytoreductive surgery (P= 0.24).

We compared survival in the group treated with cisplatin and ifosfamide as first-line combination chemotherapy versus all other cytotoxic regimens. The doses administered every 28 days were 40 mg/m2× 1 day for cisplatin and 1200 mg/m2/day × 4 days for ifosfamide. Mesna was infused every day at doses of 120 mg/m2 before and 1200 mg/m2 during ifosfamide administration. Median survival for all patients (six optimally and one suboptimally cytoreduced) treated with cisplatin and ifosfamide as first-line chemotherapy was 53 months (95% CI: 46–67 months), and for all other patients the median survival was 27 months (95% CI: 9–38 months). Figure 2 shows the survival curves for these two groups (P= 0.16).

Figure 2.

Kaplan–Meier survival curves of patients who were treated with first-line cisplatin and ifosfamide versus other chemotherapeutic regimen (P= 0.16).

One group of six patients with metastatic ovarian carcinosarcoma received the same treatment: complete surgical staging and optimal cytoreduction followed by first-line cytotoxic therapy with cisplatin and ifosfamide. Five patients were treated with six cycles. One patient died of disease after completing two cycles. Five (83%) patients from this group were alive at 3 years, and four patients (66%) were alive at 4 years. Their survival (median survival 51 months, 95% CI: 46–67 months) did not demonstrate a significant difference (P= 0.26) when compared to all other patients (median survival 27 months, 95% CI: 9–38 months).

A second group consisted of four patients with stages III and IV ovarian carcinosarcomas that were optimally debulked and treated with carboplatin (AUC [area under the curve] 5) and taxol (175 mg/m2) every 21 days as first-line chemotherapy (six to nine cycles). Their median survival was 38 months, which was not significant (P= 0.48) when compared with the 51-month survival of the cisplatin and ifosfamide group. It should be noted, however, that all patients treated with taxol and carboplatin were alive at the time of the last follow-up visit. Since all patients but one from the cisplatin and ifosfamide group are dead of disease, we compared the median progression-free interval for optimally cytoreduced patients with stage III and IV carcinosarcoma that were treated with first-line cisplatin and ifosfamide (13 months) versus first-line carboplatin and taxol (6 months). This difference was not statistically significant (P= 0.40). The progression-free intervals for patients treated with cisplatin and ifosfamide or taxol and carboplatin as first-line chemotherapy after optimal cytoreduction are listed in Table 2.

Table 2.  Progression-free intervals for patients with advanced stage carcinosarcoma of the ovary treated with cisplatin and ifosfamide or carboplatin and taxol as first-line chemotherapy after optimal cytoreduction
Patient No.AgeStageFirst-line chemotherapyProgression-free intervalSurvival
  1. AWD, alive with disease; DOD, dead of disease; NED, alive with no evidence of disease.

174IIIBCisplatin + ifosfamide12 months46 months DOD
287IIIA28 months49 months DOD
355IIIC0 month53 months DOD
450IV46 months52 months NED
561IIIA14 months67 months DOD
647IIIC0 month4 months DOD
762IIICCarboplatin + taxol4 months24 months AWD
869IIIC11 months17 months NED
960IV4 months18 months AWD
1050IIIC8 months37 months AWD

Discussion

Ovarian carcinosarcomas are rare malignancies(8) that affect mainly the postmenopausal patient. They are aggressive cancers that are routinely widely metastatic at the time of initial presentation(9).

No uniform agreement exists about the optimal treatment of these malignancies except that achieving optimal debulking at the time of initial surgery significantly improves overall survival(10,11).

Most published data are limited to retrospective reviews, and because of the rarity of the disease, few institutions are able to accrue a sufficient number of patients for prospective studies. Table 3 summarizes data regarding outcome of patients with carcinosarcoma of the ovary in published studies.

Table 3.  Ovarian carcinosarcoma: surgical treatment, chemotherapies, and outcome in published studies
AuthorsPatients evaluatedPatients with advanced diseaseSurgical treatment (all stages)Chemotherapy (first line unless otherwise stated)Progression-free intervalSurvival (all stages)
  1. MMMT, malignant mixed müllerian tumors.

Rutledge et al.(17)3124 patients stages III/IV25 patients with no visible residual disease; three patients, >1 cm; two patients, 2 cm; one patient, >5 cmCisplatin 20 mg/m2× 4 days and ifosfamide 1.5 gm/m2× 4 days every 21 days × six cycles (11 patients); carboplatin AUC 6 and taxol (175 mg/m2) every 3 weeks × six cycles (16 patients)12 months in carboplatin and taxol group; not reached in cisplatin and ifosfamide group (but significantly improved)21 months (31 patients); cisplatin and ifosfamide, 81% alive at 2 years; carboplatin and taxol, 55% alive at 2 years
Mok et al.(19)TenEight patients stages III/IVEight patients with no visible residual disease; one patient, <2 cm; one patient, <1 cmCisplatin 75 mg/m2 and ifosfamide 1.2 gm/m2, two to six cycles (seven patients); platinum-based chemotherapy, all patients 46 months (ten patients)
Thigpen et al.(13)13227 patients with stages III/IV and recurrent disease (measurable disease data only)86 patients with nonmeasurable disease; 44 patients with measurable disease (measurements for residual disease not stated)Cisplatin 50 mg/m2 every 3 weeks until disease progression or unacceptable toxicity5.2 months (130 patients)11.7 months (130 patients); measurable disease responders = 19 months; nonresponders = 4.7 months; nonmeasurable disease = 15.8 months
Brown et al.(12)6552 patients stages III/IV28 patients, <2 cm; 27 patients, >2 cm; ten patients, unknownPlatinum-based regimens (37 patients); no chemotherapy (22 patients)6.4 months8.2 months (65 patients)
Harris et al.(20)4032 patients stages III/IVFour patients, no visible residual disease; nine patients, <2 cm; 27 patients, >2 cmPlatinum-based regimens (26 patients); no chemotherapy (eight patients) 8.7 months (40 patients)
Duska et al.(21)28; 26 first-line carboplatin and taxol23 patients, stages III/IVData for patients with complete response only: ten patients, <2 cm; four patients, >2 cm; two patients, unknownCarboplatin AUC 5–7.5 and taxol (175 mg/m2) × three to eight cycles (13 patients)9 months27.1 months (all stages); 25 months (stages III and IV)
Sit et al.(22)13Ten patients stages III/IVSix patients, no visible residual disease; four patients, <1 cm; one patient, 2 cm; two patients, >2 cmCarboplatin AUC 5 and taxol (175 mg/m2) every 3 weeks (first or second line) (six patients); cisplatin 50 mg/m2 and ifosfamide 5 gm/m2 (+mesna 5 gm/m2) × 1 day (first or second line) (eight patients)10 months (cisplatin and ifosfamide group, eight patients’ data only)Carboplatin and taxol group: 19 months; Cisplatin and ifosfamide group: 23 months
Sood et al.(10)47: 40 MMMT and seven sarcoma41 patients stages III/IV (MMMT and sarcoma)25 patients, <1 cm; 22 patients, >1 cmSurvival advantage for platinum-based regimens (27 patients) versus other (11 patients); comment: preoperative CA-125 < 75 U/mL associated with better survival10.5 months16 months (47 patients, MMMT and sarcoma); 5-year survival, 45% for optimal cytoreduction versus 8% for suboptimal
Le et al.(11)36; 35 staging data available26 patients stages III/IV22/35 patients with macroscopic residual disease; cisplatin/doxorubicin group 28 patients: six patients, no visible residual disease; eight patients, <2 cm; 14 patients >2 cmCisplatin 50 mg/m2 and doxorubicin 50 mg/m2 every 4 weeks up to nine cycles (28 patients)90% (19/21 patients) of recurrences occurred within 2 years after diagnosisCisplatin/doxorubicin group (28 patients): 56% survival at 2 years; 35% survival at 5 years

Brown et al. compared prospectively the clinicopathologic features and outcomes of 65 patients with carcinosarcoma of the ovary to 746 patients with serous adenocarcinoma of the ovary. They reported a median survival of 14.8 months in patients with optimally debulked FIGO stage III ovarian carcinosarcomas as opposed to 3.1 months for suboptimally or nondebulked stage III disease. Their data show a significantly lower objective response rate to platinum therapy(12). However, most authors report a higher response rate for platinum-based regimens than for nonplatinum combinations.

Thigpen et al. in a Gynecologic Oncology Group study presented data involving 136 patients with ovarian carcinosarcomas treated with cisplatin 50 mg/m2 every 3 weeks until disease progression or unacceptable toxicity occurred. They concluded that cisplatin is active as initial therapy for these tumors. The median survival in 130 patients was 11.7 months(13).

Baker et al. published a prospective trial that involved 11 advanced or recurrent ovarian carcinosarcoma patients treated with cisplatin, adriamycin, and dacarbazine after surgical debulking. Their estimated 1-, 2-, and 3-year survival was 70%, 35%, and 35%, respectively(14).

Zorzou et al. treated nine patients with aggressive cytoreduction followed by anthracycline-based chemotherapy. Their median overall survival was 32.9 months with no statistical difference between early and advanced stages(15).

Crotzer et al. published a pilot study of cisplatin, ifosfamide, and mesna in the treatment of malignant mixed mesodermal tumors of the ovary. The median survival for the eight patients was 21 months(16).

Rutledge et al. evaluated retrospectively 11 patients with early and advanced disease who received cisplatin 20 mg/m2 and ifosfamide 1.5 gm/m2 for 4 days every 21 days for six cycles. The progression-free interval and overall survival for all stages were improved when compared to 16 patients treated with carboplatin (AUC 6) and paclitaxel 175 mg/m2 every 21 days for six cycles. However, this advantage disappeared when only advanced stages were considered(17).

Patsner et al. treated four patients with metastatic ovarian carcinosarcoma with mesna, doxorubicin, ifosfamide, and dacarbazine chemotherapy. Two of the four patients had optimal cytoreduction. They reported two complete responses of 34- and 46-month duration(18).

Le et al. reported on 33 patients with metastatic carcinosarcoma of the ovary. At the end of surgery, 22 patients were left with macroscopic residual disease. Adjuvant chemotherapy consisted of cisplatin and doxorubicin. Survival was 35% at 5 years. Thirteen patients had a second-look laparotomy, and the authors concluded that a second operation offers little helpful information on the management of these tumors(11).

Mok et al. presented a case series of ten patients and reviewed the use of platinum-based combination chemotherapy for ovarian carcinosarcoma at their institution. All patients underwent optimal cytoreductive surgery. Following surgery, all patients received platinum-based chemotherapy and seven patients were treated with ifosfamide (1200 mg/m2) and cisplatin (75 mg/m2). The median survival was 46 months for the entire cohort (two patients with stage IIC, seven patients with stage IIIC, and one patient with stage IV). They concluded that platinum-based combination chemotherapy after optimal cytoreductive surgery may be effective in the treatment of these tumors. They also noted that the toxicities of these regimens, especially those including ifosfamide, are significant(19).

Invariably, a diagnosis of advanced stage carcinosarcoma is associated with a dismal prognosis, and most studies record few survivors at 2 years or even 1 year. The case series presented here reports a more encouraging survival time, especially in the group of patients treated with optimal cytoreduction and adjuvant cisplatin and ifosfamide every 28 days. Administration of cisplatin and ifosfamide requires hospitalization for the duration of treatment, while the carboplatin and paclitaxel combination is infused in an outpatient setting. The latter chemotherapy regimen was administered more commonly in recent years because of the recognition of a monoclonal epithelial origin for uterine carcinosarcomas(1,4). Survivals in the two groups presented are comparable to survivals observed in ovarian epithelial malignancies. The groups of patients presented are small but homogeneous, since all cases had metastatic disease found at the initial diagnosis.

Survival curves are the only valid way to compare survival. The median survival may be misleading because, as shown in Figure 2, the patients in the noncisplatin/ifosfamide group actually survive longer in the end. However, this may be just by chance because the number of patients is small. Only two patients in the noncisplatin and ifosfamide group made this group look better overall. The curves do differ initially, and this observation merits further investigation about the possible benefits of cisplatin and ifosfamide combination regimen. However, this data set is showing that the curves actually cross in the end, which makes the statistical comparisons of survival curves less valid. The results of this study were utilized for a power analysis. While the difference in survival between the cisplatin and ifosfamide group and the carboplatin and taxol group would become statistically meaningful (power of 80%) at 279 patients in each group, the difference in progression-free interval would achieve significance with only 29 patients in each group.

For advanced stage ovarian carcinosarcoma, we recommend optimal cytoreduction and adjuvant chemotherapy. Because of the rarity of the disease, only a nationwide multi-institutional trial can provide a better answer to what constitutes the most effective cytotoxic regimen.

Ancillary