Cover Letter
Title: Primary treatment and prognostic factors of carcinosarcoma of ovary, fallopian tube, and peritoneum: a Taiwanese Gynecologic Oncology Group study
Authors: Chien-Hsing Lu, I-hui Chen, Yi-Jen Chen, Kung-Liahng Wang, Jian-Tai Timothy Qiu, Hao Lin, Wu-Chou Lin, Wen-Shiung Liou, Yue-Shan Lin, Yao-Ching Hung
Dear Editors and Reviewers:
Carcinosarcoma (malignant mixed Müllerian tumor, or malignant mixed mesoderm tumor) of ovary, fallopian tube, and peritoneum is an extremely rare and highly malignant disease. The prognosis is even poorer than the more prevalent and also highly malignant high-grade serous carcinoma. Due to the rarity of this cancer, previous reports about ovarian carcinosarcoma were all case series studies and involved small case numbers. Although Carmen et al. made a
comprehensive review of ovarian carcinosarcoma from molecular and clinical perspectives, several important issues regarding prognosis and optimal treatment remain controversial or unelucidated due to the lack of relevant studies. This study was designed to delineate prognostic factors using a multivariate analysis based on a multi-center database.
After retrospectively reviewing the medical records of participating hospitals, a total of 61 cases diagnosed with the disease were included in the data analysis. Various demographic data, tumor condition, extent of operation, and chemotherapy were analyzed first with log-rank test, simple Cox regression, and multiple Cox regression tests. We found early diagnosis at stage I, omentectomy,
complete debulking operation, platinum, ifosfamide, and cisplatin/ifosfamide regimen were
all significantly correlated with or showed a trend toward improvement in survival by log-rank test. Multiple Cox regression analysis revealed that only six or more cycles of
chemotherapy may improve the prognosis of carcinosarcoma of ovary, fallopian tube, and
peritoneum.
To date, this study is the largest of its kind in terms of number of patients and also includes the 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
largest population of patients diagnosed with stage I and II carcinosarcoma. To the best of our knowledge, this is the first study to demonstrate a significant benefit of a specific number of chemotherapy cycles. The true prognostic value of various clinical variables was revealed by log-rank test as well as by multiple Cox regression tests which was used to adjust for potential confounding factors. Thus, the results of this study may lead to a better understanding of the various prognostic factors and the direction of treatment which in turn may result in better outcomes.
The paper was prepared according to the STROBE guidelines. The manuscript has never been published in any other journal, neither has it been presented in any meeting. All authors have declared no financial conflicts of interest. Institutional Review Board approvals were obtained from each participating hospital before the beginning of the study.
Funding: nil
Suggestion of reviewers:
1. Linus T. Chuang MD. Division of Gynecologic Oncology, Department of Obstetrics,
Gynecology and Reproductive Medicine, Mount Sinai School of Medicine, 1176 Fifth Avenue, Box 1173, New York, NY 10029, USA. [email protected]
2. Hung-Cheng Lai. Department of Obstetrics and Gynecology, Tri-Service General Hospital, National Defense Medical Center, 325, Section 2, Cheng-Kung Road, Neihu, Taipei 114, Taiwan. [email protected]
Best Regards,
Chien-Hsing Lu and Yao-Ching Hung May 27, 2013 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
Primary treatment and prognostic factors of carcinosarcoma of ovary, fallopian tube, and peritoneum: a Taiwanese Gynecologic Oncology Group study (Character count: 145)
Manuscript type: Original Article Authors:
Chien-Hsing Lu a,b,c , I-hui Chen d , Yi-Jen Chen e,f , Kung-Liahng Wang g,h,i , Jian-Tai Timothy Qiu j , Hao Lin k , Wu-Chou Lin l , Wen-Shiung Liou m, Yue-Shan Lin o , Yao-Ching Hung p,q*
From:
a Department of Obstetrics and Gynecology, Taichung Veterans General Hospital, Taichung,
Taiwan
b Department of Obstetrics and Gynecology, National Yang-Ming University, Taipei,Taiwan. c Institute of Biomedical Sciences, National Chung-Hsing University, Taichung, Taiwan
d Obstetrics and gynecology department, National Taiwan University Hospital, Hsin-Chu branch e Department of obstetrics and Gynecology, Taipei Veterans General Hospital.
f Department of Obstetrics and Gynecology, National Yang-Ming University, Taipei, Taiwan. g Department of obstetrics and Gynecology, Mackay Memorial Hospital, Taipei, Taiwan.
h Department of Nursing, Mackay Medicine, Nursing and Management College, Taipei, Taiwan. i Department of Obstetrics and Gynecology, Taipei Medical University, Taipei, Taiwan.
j Department of obstetrics and Gynecology, Chang Gung Memorial Hospital, and Chang Gung
University College of Medicine, Taoyuan, Taiwan.
k Division of GynecologicOncology, Department of Obstetrics and Gynecology, Kaohsiung Chang
Gung Memorial Hospital and Chang Gung University College of Medicine, Kaohsiung, Taiwan.
l Department of obstetrics and Gynecology, China Medical University Hospital, Taichung, Taiwan. m Department of obstetrics and Gynecology, Kaohsiung Veterans General Hospital, Kaohsiung,
Taiwan
o Department of obstetrics and Gynecology, Chi-Mei Medical Center, Tainan, Taiwan.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
p Department of obstetrics and Gynecology, China Medical University Hospital, Taiwan q College of Medicine, China Medical University, Taiwan.
Address reprint requests to:
Dr. Yao-Ching Hung
Department of Obstetrics and Gynecology, China Medical University Hospital, No. 2, Yuh-Der Road, Taichung 404, Taiwan.
Telephone: +886-4-23741258, Fax: +886-4-23503021 E-mail: [email protected]
Acknowledgements: The authors would like to thank the Biostatistics Task Force of Taichung
Veterans General Hospital for their assistance
Details of ethics approval: Institutional Review Board (IRB), Taichung Veterans General
Hospital, No. C10242. Date of approval: Jan. 12, 2011 and IRB from each participating hospital.
Short title: Ovarian carcinosarcoma (Character count: 22)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
ABSTRACT
Objectives. To determine the clinical prognostic factors in carcinosarcoma of ovary, fallopian tube, and peritoneum.
Materials and Methods. This retrospective study was undertaken by the Taiwan Gynecologic Oncology Group (TGOG). The retrieved clinical data included demographic factors, medical disease, tumor status, extent of operation, and adjuvant chemotherapy.
Results. Totally, 63 cases diagnosed with carcinosarcoma of ovary, fallopian tube, and peritoneum were identified. Only 61 patients with complete data were enrolled for further data analysis. The mean follow-up period was 2.0 ± 2.5 years. The mean overall survival was 15.4 months. By log-rank test, age, menopausal status, parity, hypertension, diabetes, tumor size, para-aortic lymph metastasis, pretreatment CA-125, preceding diagnostic operation, hysterectomy, lymphadenectomy, other operations, and paclitaxel use were not predictive of overall survival. Pelvic lymph metastasis, omentectomy, gross residual implants, and platinum were borderline significant for survival. Stage I, unilateral ovarian tumor, metastatic tumors less than 2 cm, ifosfamide, cisplatin/ifosfamide regimen, and six or more cycles of chemotherapy were
significantly predictive of survival. When significant parameters were analyzed by multiple Cox regression test, only six or more cycles of chemotherapy were predictive of better overall survival. Conclusions. Early diagnosis at stage I, omentectomy, complete debulking operation, platinum, ifosfamide, and cisplatin/ifosfamide regimen all showed a trend toward improvement of survival by log-rank test. On multiple Cox regression analysis, only six or more cycles of chemotherapy appeared to improve the prognosis of carcinosarcoma of ovary, fallopian tube, and peritoneum. (Word count: 236) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22
Introduction
Ovarian cancer is the most lethal gynecologic cancer. Despite the progress that has been made in surgical techniques and cytotoxic chemotherapeutic agents, the five-year survival rate has only increased from 36% to 44% in the past thirty years [1]. It was estimated that in 2012 there would be 22,280 new cases and 15,500 deaths due to ovarian cancers in the US. Ninety percent of ovarian cancers are epithelial cell type and comprise multiple histological types with various specific molecular changes, clinical behaviors, and treatment outcomes [2]. Carcinosarcoma (also called malignant mixed Müllerian tumor, or mixed mesodermal tumor) contains both malignant epithelial and mesenchymal elements. The more frequent epithelial component is serous, endometrioid, or undifferentiated adenocarcinoma. The sarcomatous element may be homologous tissue native to the ovary or heterologous tissue not native to the ovary. According to the Surveillance, Epidemiology and End Results (SEER) report, this is a very rare tumor and accounts for only 1.7% of all ovarian cancer patients older than 20 years of age [3]. The relative survival rate (RSR) of ovarian
carcinosarcoma was 29.8%, which was even worse than that of the more prevalent serous carcinoma (RSR 39.6 to 44.2%) [3].
Due to the rarity of this cancer, previous investigations of ovarian carcinosarcoma were all case series studies. Although Carmen et al. conducted a comprehensive review of ovarian
carcinosarcoma from molecular and clinical aspect perspectives [4], a number of important issues regarding the prognosis and optimal treatment remain controversial or unelucidated. Thus, this study was designed to determine the prognostic factors of ovarian carcinoma by multivariate analysis using a multi-center database.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22
Materials and methods
Institutional Review Board approvals were obtained from participating medical centers in the Taiwanese Gynecologic Oncology Group (TGOG). The clinical data of patients diagnosed with carcinosarcoma of ovary, fallopian tube, and peritoneum from 1991 to 2011 were retrospectively collected from medical charts for statistical analysis.
The staging was based on the 1988 International Federation of Gynecology and Obstetrics (FIGO) staging system. Demographic data collected included age at diagnosis (younger than 55 years or 55 years and older), menopausal status, and parity (two or more or fewer than two). Medical history was assessed by reviewing medical charts and collecting data on hypertension and diabetes. Extent of tumor was evaluated and data included: stages of disease, which were
reassigned to stages I, II, III, or IV; tumor size (14 cm or more, or less than 14 cm), bilateral ovarian tumors, retroperitoneal lymph node metastasis, size of metastatic tumors (2 cm or more, or less than 2 cm), and serum level of CA-125 (200 u/ml and above, or less than 200 u/ml). The treatment policies were decided by the attending physician or by the multidisciplinary tumor board in each medical center. Extent of operations analyzed included: preceding operation leading to the diagnosis (including laparoscopy, exploratory laparotomy with biopsy, oophorectomy, etc.), hysterectomy, omentectomy, pelvic, and para-aortic lymphadenectomy, other operations (wedge resection of liver, segmental resection of small or large intestine, splenectomy, etc.), size
(macroscopic or not) and number (more than one, one, or zero) of residual tumors.
Chemotherapeutic agents analyzed included platinum (cisplatin or carboplatin), paclitaxel,
ifosfamide, platinum/paclitaxel combination, and platinum/ifosfamide combination. The number of cycles of primary adjuvant chemotherapy was also recorded. Overall survival was defined as the interval between the date of diagnosis to the date when patient died or had her last follow-up. Data were analyzed using the Statistical Package for the Social Sciences software (SPSS for Windows, version 10.0.7, SPSS Inc., Chicago, IL). Survival curves were generated using the Kaplan-Meier 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
method and differences in survival curves were calculated using the log-rank test. The cut-off values of continuous variables were determined by receiver operating characteristic (ROC) curves. Simple Cox regression tests were used first to determine the crude hazard ratio (HR) of significant variables by log-rank test. Multiple Cox regression model was then used to evaluate the adjusted HR to predict survival after controlling for significant variables identified by log-rank test. A p value < 0.05 was considered statistically significant.
1 2 3 4 5 6
Results
Initially, a total of 64 cases diagnosed with ovarian/fallopian tube/peritoneal carcinosarcoma were enrolled into the data analysis. Among them, three cases had key data missing, including stage and regimen of chemotherapy, and were thus excluded from the analysis. Finally, 61 cases were enrolled for further data processing and analysis.
In this cohort, the mean age at diagnosis was 56.6 ± 15.1 years (range, 17.6 - 89.6 years). The mean parity was 1.8 ± 1.8. (range, 0-8). Thirty-seven (61%) patients were menopausal.
Demographic factor did not affect survival (Supplementary Table S1). Hypertension and diabetes were noted in 12 and 7 cases, respectively, but did not influence survival (Supplementary Table S1). Ten patients had no symptoms, 11 patients had self-palpable mass, and 31 and 25 patients had abdominal pain and abdominal fullness, respectively. Four patients had abnormal vaginal bleeding, and two and three patients had nausea and vomiting, respectively. Three patients had poor appetite. The mean follow-up period was 2.0 ± 2.5 years (range, 0-10.6 years). The median overall
survival of the cohort was 1.32 years. In the cohort, only 14 (23%) cases were diagnosed at stages I (8 cases) and II (6 cases). Stage I disease could predict better overall survival by log-rank test with a p value of 0.006 (Table 1, Figure 1). Bilateral ovarian tumor was associated with poor prognosis with a p value of 0.007 (Table 1). Size of metastatic tumor less then two centimeters was predictive of better overall survival (p= 0.036). Pelvic lymph node metastasis was borderline predictive of worse overall survival (p=0.103) (Table 1). Other parameters including tumor size, para-aortic lymph node metastasis, and pre-treatment CA-125 were not predictive of overall survival (Table 1). None of the cases received neoadjuvant chemotherapy, but received primary surgical treatment by laparotomy with or without adjuvant chemotherapy instead. Four patients received preceding operation leading to the diagnosis (Table 2), including two who received laparoscopic
adnexectomy, and two who received unilateral salpingo-oophorectomy. Oophorectomy was done in all patients. Hysterectomy, lymphadenectomy, and omentectomy were done in 86%, 68%, and 82% 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
of cases, respectively. Extent of operation, either hysterectomy, pelvic lymphadenectomy, para-aortic lymphadenectomy, and other operations were not predictive of survival. Omentectomy (p=0.101), number of residual implants of one or less (p=0.137), and microscopic or no residual implants (p=0.144) were borderline predictive of survival (Table 2, Figure 2). Chemotherapy was given to all but 6 patients. The average cycle of chemotherapy was 4.2 ± 2.5 (ranged 0 - 9). Thirty-three (54%) patients received six or more cycles of primary chemotherapy. The analysis of the effects based on each chemotherapeutic agent revealed that paclitaxel was not predictive of survival. In contrast, ifosfamide and platinum were associated with significant and borderline survival (p= 0.033 and 0.053, respectively). The evaluation of the effects of combination
chemotherapy showed that platinum combined with ifosfamide was associated with significantly better survival compared with those of the other regimens (p=0.021) (Table 2, Supplementary Figure S2). However, platinum combined with paclitaxel was associated with poorer prognosis when compared with those of the other regimens (p=0.031). Six or more cycles of chemotherapy was associated with better survival compared with that of treatments with fewer cycles (p< 0.001) (Table 2, Figure 3).
The aforementioned statistically significant parameters were analyzed by simple Cox regression to calculate their crude HR for mortality (Table 3). Multiple Cox regression analyses were then performed to control for potential confounding factors and the results showed that the only
significant parameter predicting overall survival was six or more cycles of adjuvant chemotherapy. Although log-rank test initially showed that other parameters, including diagnosis at stage I, omentectomy, complete debulking operation, platinum, ifosfamide, and cisplatin/ifosfamide regimen, showed significant associations with better survival, they were not significant prognostic factors in the multiple Cox regression analysis (Table 3).
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
Discussion
Carcinosarcoma of ovary, fallopian tube, and peritoneum is regarded as a highly malignant tumor with very poor prognosis and is relatively chemoresistant [5]. Previous analyses were cohort studies from single medical centers and enrolled only a small number of cases, which limited the use of multivariate analysis to determine true independent risk factors which can predict prognosis. Reported prognostic factors may turn out not to be true independent factors if multivariate analysis is used. In this multi-center study the highest number of carcinosarcoma patients to date were analyzed. The purpose of this study was to identify true risk factors for prognosis of this aggressive cancer with a view to gaining a better understanding of the disease and improving its management.
Inthasorn et al. reported that age over 65 years was an indicator of poor prognosis using univariate analysis [6]. However, our data revealed that age 55 years and older or younger was not associated with poorer prognosis even based on log-rank test only. This finding was similar to
results reported by Rutledge et al. [7] who found age did not have an impact on survival.
It is reasonable to assume that early stage disease may warrant a better prognosis; however, to the best of our knowledge, no studies to date have been able to demonstrate stage as a prognostic factor using multivariate analysis. The statistical power in previous case series was low due to the rarity of early stage cases. In our study, simple Cox regression analysis revealed a crude HR of 6.93 (p=0.031) for death comparing stage I to stages II, III, and IV (Table 1, Figure 1). The significance disappeared when multivariate analysis was applied even though the adjusted HR was 2.64
(p=0.379). The disappearance of significance may be due to the small case number, even though this study included one of the largest samples of patients with stage I and II in the literature (14 cases). Hellström AC et al. found patients with stage I or II had better survival than those with stage III or IV [8], although the p value was less than 0.07, which is usually regarded as not statistically significant. Moreover, 75% of the early stage patients in the aforementioned study were in stage II. Thus, stage II patients appeared to have a relatively good prognosis, which was inconsistent with 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
the findings of the present study. In Figure 1, it can be seen that patients with stage II disease had quite a similar Kaplan-Meier survival curve to those with more advanced stage disease. A poor prognosis in stage II patients was also demonstrated in a study by Brown et al. which showed that 2-year survival was zero [5]. However, stage as a prognostic factor was not examined in the study. Studies by Sood et al. and Rutledge et al. demonstrated that stage was a better prognostic factor using univariate analysis, but not using multivariate analysis [7; 9; 10]. Stage was not predictive of survival in other studies [8; 11].
Optimal cytoreduction surgery may be the best strategy for improving patient survival in epithelial ovarian cancer and the same appears to be true for carcinosarcoma of ovary [12]. Our data also support the finding that cytoreduction to no gross residual tumors and no more than one residual tumor were associated with better survival (p=0.144, and 0.137, respectively). Previous reports with fewer than 30 cases consistently showed either optimal debulking to less than one or two centimeters of residual tumor as a significant or beneficial factor for survival [5; 10; 11; 12]. However, the survival benefit of optimal debulking surgery as an independent prognostic factor by multivariate analysis has only been shown in one study [10]. However, the study included 15% of ovarian sarcoma. In one of the largest case series (N=50), Rauh-Hain et al. reported median overall survival (OS) was 47 months for women with only microscopic residual tumor, 18 months in patients with optimal but macroscopic residual tumor, and 8 months in patients with suboptimal cytoreduction (larger than 1 cm residual tumor)(p=0.02) [12]. However, the survival benefit was found only in univariate analysis, and the same conditions existed in other series [5; 11; 12; 13]. Although survival benefit was found in those studies with a relatively higher number of cases, there was no statistical significance using multivariate analysis, except in the abovementioned study by Sood et al. In summary, there still were no strong evidences whether optimal debulking is an independent prognostic factors or not, but there exists a trend toward better survival with optimal debulking surgery. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
In earlier trials, it was found that doxorubicin was not active in adjuvant chemotherapy [14]. Later studies showed that platinum, ifosfamide alone or combination therapy, as well as
carboplatin/paclitaxel regimen improved survival [13; 15; 16; 17; 18]. As ovarian carcinosarcoma is known to be highly malignant, and single agent cisplatin (progression free survival of 5.2 months and overall survival of only 11.7 months) has been shown to have poor efficacy [18], in addition to clinical experience gained treating other type of epithelial ovarian cancer, it is reasonable to
postulate that platinum combination with paclitaxel may show greater benefit than other modalities in the management of this disease. Although this is the most reported adjuvant chemotherapy regimen, the number of reports was still limited and they are retrospective studies. Duska reported a response rate of 72% and a median overall survival time of 27 months [19]. Rauh-Hain et al. reported a response rate of 62% and a median overall survival time of 24 months. An even more important issue is which regimen should be the first choice for adjuvant chemotherapy after debulking operation. Anita et al. were the first to compare cisplatin/paclitaxel with
platinum/ifosfamide [16]. They found a median overall survival of 19 (N=6) and 23 months (N=8) in each group, respectively. Another report investigated a total of 22 cases, and found no significant difference in survival [20]. However, a report by Rutedge et al. showed improved median overall survival (p=0.03) and progression-free survival (p=0.005) comparing cisplatin/ifosfamide to platinum/paclitaxel when the analysis included all cases [7]. When the analyses were limited to advanced cases, there was no significant survival benefit (p=0.13). There was also no significance found when analyzed using multiple Cox regression, which was similar to the finding in this study. Platinum combined with ifosfamide was predictive of better survival compared to that of other regimens with a crude HR of 0.33 for death (p=0.045). However, when this variable was analyzed using multiple Cox regression analysis, the statistical significance disappeared (adjusted HR=0.35, p=0.153). Based on the current published data including the results presented herein, no firm conclusion about whether cisplatin/ifosfamide is better than carboplatin/paclitaxel can be made. To 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
date, no multivariate analysis has proved the superiority of cisplatin/ifosfamide with regard to survival after controlling for other potential confounding factors. Moreover, although there exists a trend toward better survival with the ifosfamide-containing regimen in all studies comparing these two regimens, it is associated with more adverse effects [7; 15; 16; 20]. Thus, the choice of
postoperative adjuvant chemotherapy still requires comprehensive clinical judgments.
The optimal cycles of adjuvant chemotherapy in cancer treatment should be based on well-designed clinical trials. However, data for epithelial ovarian cancer did not differentiate among various histological subtypes, and thus it is not known how many cycles of chemotherapy can achieve optimal efficacy in carcinosarcoma. The present study is the first to demonstrate that six or more cycles of chemotherapy may improve overall survival in this lethal cancer even after
adjusting for confounders using multiple Cox regression test (HR = 0.15, p< 0.001) (Table 3, Figure 2). Thus, six or more cycles of chemotherapy should be given to all newly diagnosed cases of carcinosarcoma. However, the optimal number of cycles needed to achieve best survival benefit while minimizing the toxicity remains unknown.
In summary, although this study included the largest number of carcinosarcoma patients compared with previous reports, there was still an insufficient number of cases to determine the role of various clinical factors in survival using multivariate analysis. This problem can only be overcome by enrolling more patients in future studies which can be achieved by close cooperation among medical centers. Multiple Cox regression analysis revealed that six or more cycles of chemotherapy in the present study were associated with better survival. Other factors, such as bilateral ovarian tumors and metastatic tumor larger than 2 cm were associated with poor survival by log-rank test, but not using multiple Cox regression test. Diagnosis at stage I, optimal debulking to no gross residual tumor, omentectomy, and combination chemotherapy with careful toxicity monitoring and management all showed a trend toward better survival. Thus, improved survival rates may be achieved by applying the abovementioned treatment modality for all newly diagnosed 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
carcinosarcoma of ovary/tube/peritoneum. 1
Disclosure of Interests: None.
References
[1] Siegel R, Naishadham D, Jemal A. Cancer Statistics, 2012. CA Cancer J Clin 2012;62:10-29. [2] Kurman RJ, S. IeM. Molecular pathogenesis and extraovarian origin of epithelial ovarian
cancer--shifting the paradigm. Hum Pathol 2011;42:918-931.
[3] Kosary CL (2007) Chapter 16: Cancer of the Ovary. In: Ries LAG, John L, Young J, Keel GE, Eisner MP, Lin YD, et al., editors. SEER Survival Monograph: Cancer Survival among Adults: US SEER Program, 1988–2001, Patient and Tumor Characteristics. Bethesda, MD: National Cancer Institute. pp. 133–144.
[4] del Carmen MG, Birrer M, Schorge JO. Carcinosarcoma of the ovary: a review of the literature. Gynecol Oncol 2011;125:271-277.
[5] Brown E, Stewart M, Rye T, Al-Nafussi A, Williams AR, Bradburn M, et al. Carcinosarcoma of the ovary: 19 years of prospective data from a single center. Cancer 2004;100:2148-2153. [6] Inthasorn P, Beale P, Dalrymple C, Carter J. Malignant mixed mullerian tumour of the ovary: prognostic factor and response of adjuvant platinum-based chemotherapy. Aust N Z J Obstet Gynaecol 2003;43:61-64.
[7] Rutledge TL, Gold MA, McMeekin DS, Huh WK, Powell MA, Lewin SN, et al. Carcinosarcoma of the ovary-a case series. Gynecol Oncol 2006;100:128-132.
[8] Hellström AC, Tegerstedt G, Silfverswärd C, Pettersson F. Malignant mixed müllerian tumors of the ovary: histopathologic and clinical review of 36 cases. Int J Gynecol Cancer
1999;9:312-316.
[9] Chang J, Sharpe JC, A'Hern RP, Fisher C, Blake P, Shepherd J, et al. Carcinosarcoma of the ovary: incidence, prognosis, treatment and survival of patients. Ann Oncol 1995;6:755-758. [10] Sood AK, Sorosky JI, Gelder MS, Buller RE, Anderson B, Wilkinson EJ, et al. Primary
ovarian sarcoma: analysis of prognostic variables and the role of surgical cytoreduction. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
Cancer 1998;82:1731-1737.
[11] Harris MA, Delap LM, Sengupta PS, Wilkinson PM, Welch RS, Swindell R, et al. Carcinosarcoma of the ovary. Br J Cancer 2003;88:654-657.
[12] Rauh-Hain JA, Growdon WB, Rodriguez N, Goodman AK, Boruta DM 2nd, Schorge JO, et al. Carcinosarcoma of the ovary: a case control study. Gynecol Oncol 2011;121:477-481.
[13] Leiser AL, Chi DS, Ishill NM, Tew WP. Carcinosarcoma of the ovary treated with platinum and taxane: the memorial Sloan-Kettering Cancer Center experience. Gynecol Oncol
2007;105:657-661.
[14] Morrow CP, Bundy BN, Hoffman J, Sutton G, Homesley H. Adriamycin chemotherapy for malignant mixed mesodermal tumor of the ovary. A Gynecologic Oncology Group Study. Am J Clin Oncol 1986;9:24-26.
[15] Crotzer DR, Wolf JK, Gano JB, Gershenson DM, Levenback C. A pilot study of cisplatin, ifosfamide and mesna in the treatment of malignant mixed mesodermal tumors of the ovary. Gynecol Oncol 207;105:399-403.
[16] Sit AS, Price FV, Kelley JL, Comerci JT, Kunschner AJ, Kanbour-Shakir A, et al. Chemotherapy for malignant mixed Müllerian tumors of the ovary. Gynecol Oncol 2000;79:196-200.
[17] Sutton GP, Blessing JA, Homesley HD, Malfetano JH. A phase II trial of ifosfamide and mesna in patients with advanced or recurrent mixed mesodermal tumors of the ovary
previously treated with platinum-based chemotherapy: a Gynecologic Oncology Group study. Gynecol Oncol 1994;53:24-26.
[18] Tate Thigpen J, Blessing JA, DeGeest K, Look KY, Homesley HD; Gynecologic Oncology Group. Cisplatin as initial chemotherapy in ovarian carcinosarcomas: a Gynecologic Oncology Group study. Gynecol Oncol 2004;93:336-339.
[19] Duska LR, Garrett A, Eltabbakh GH, Oliva E, Penson R, Fuller AF. Paclitaxel and platinum 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
chemotherapy for malignant mixed müllerian tumors of the ovary. Gynecol Oncol 2002;85:459-463.
[20] Silasi DA, Illuzzi JL, Kelly MG, Rutherford TJ, Mor G, Azodi M, et al. Carcinosarcoma of the ovary. Int J Gynecol Cancer 2008;18:22-29.
1 2 3 4
Table and figure legends
Table 1
The extent of disease predicting overall survival by log-rank test Table 2
The surgical and chemotherapeutic strategies influencing overall survival by log-rank test Table 3
Simple and multiple Cox regression analysis of significant clinical factors that were identified by log-rank tests and that were predictive
Figure 1
Kaplan-Meier survival curve of stages Figure 2
Kaplan-Meier survival curve of platinum/ifosfamide Figure 3
Kaplan-Meier survival curve of number of cycles of chemotherapy 1 2 3 4 5 6 7 8 9 10 11 12 13 14
Table 1
Parameter No. Death (%) P
Stage 1 8 1(12.5) 0.006 2 6 3 (50.0) 3 38 26 (68.4) 4 9 8 (88.9) Tumor size (cm) <14 29 19 (65.5) 0.605 >14 27 16 (59.3)
Bilateral ovarian tumors N 27 13 (48.1) 0.007
Y 26 21 (80.8)
Pelvic lymph node metastasis N 53 31 (58.5) 0.103
Y 7 6 (85.7)
Para-aortic lymph node metastasis N 56 34 (60.7) 0.769
Y 4 3 (75.0)
Size of metastatic tumors (cm) <2 29 14 (48.3) 0.036
>2 20 15 (75.0)
Pre-treatment CA-125 (U/ml) <200 26 14 (53.9) 0.734
>200 27 18 (66.7)
1
Table 2
Parameter No. Death (%) P
Preceding diagnostic operations N 57 37 (64.9) 0.217
Y 4 1 (25.0) Hysterectomy N 8 5 (62.5) 0.471 Y 51 33 (64.7) Omentectomy N 11 8 (72.7) 0.101 Y 49 29 (59.2) Pelvic lymphadenectomy N 26 17 (65.4) 0.642 Y 34 20 (58.8) Para-aortic lymphadenectomy N 34 22 (64.7) 0.395 Y 26 15 (57.7) Other operations N 48 30 (62.5) 0.587 Y 12 7 (58.3)
Gross residual implants N 23 11 (47.8) 0.144
Y 30 22 (73.3)
No. of residual implants 0 or 1 23 10 (43.5) 0.137
> 1 29 22 (75.9) Platinum N 9 7 (77.8) 0.051 Y 52 31 (59.6) Paclitaxel N 24 15 (62.5) 0.425 Y 37 23 (62.2) Ifosfamide N 48 32 (66.7) 0.033 Y 13 6 (46.2) Platinum + paclitaxel N 10 4 (40.0) 0.031 Y 34 21 (61.8) Platinum + ifosfamide N 35 22 (62.9) 0.021 Y 9 3 (33.3) Cycles of chemotherapy <6 29 22 (75.9) <0.001 >6 32 16 (50.0) 1 2
Table 3
Parameters Simple Cox regression Multiple Cox regression
Crude HR (95% C.I.) p Adjusted HR (95% C.I.) p
Stage Stage 1 1.00 1.00
Stage 2, 3, or 4 6.94 (1.22-65.37) 0.031 2.64 (0.30-22.80) 0.379
Bilaterality N 1.00 1.00
Y 2.52 (1.26-5.05) 0.009 3.12 (0.78-12.51) 0.108
Size of metastatic tumor (cm) < 2 1.00 1.00
> 2 2.14 (1.03-4.45) 0.041 0.80 (0.23-2.84) 0.732
Platinum + ifosfamide N 1.00 1.00
Y 0.33 (0.11-0.97) 0.045 0.35 (0.08-1.48) 0.153
Cycles of chemotherapy < 6 1.00 1.00
> 6 0.21 (0.11-0.42) <0.001 0.15 (0.05-0.44) <0.001
*HR: hazard ratio; C.I.: confidence interval. 1
Figure 1 p=0.006 1 2 3 4 5 6 7 8 9 10 11
Figure 2 p=0.021 1 2 3 4 5 6 7 8 9 10 11
Figure 3
P<0.001 1
Supplementary Table S1
Demographic factors and medical factors predicting overall survival by Kaplan-Meier method Parameter N Death (%) P Age < 55 24 16 (66.7) 0.725 >55 37 22 (59.7) Menopause N/A 6 4 (66.7) 0.728 N 18 12 (66.7) Y 37 22 (59.7) Parity <2 26 14 (53.8) 0.619 >2 34 23 (67.6) Hypertension N 49 29 (59.7) 0.169 Y 12 9 (75.0) Diabetes N 54 32 (59.3) 0.282 Y 7 6 (85.7)
* N/A: not available. 1
2 3
Supplementary Figure S1
Kaplan-Meier survival curve of residual implants
p=0.144 1
2
3 4
Highlight
> On multiple Cox regression analysis, only six or more cycles of chemotherapy appeared to improve prognosis.
> Stage I, omentectomy, complete debulking operation, platinum, ifosfamide, and cisplatin/ifosfamide trended toward better survival.
1 2 3 4 5
