Search

Objective: To assess the influence of geographies and race on the survival outcomes in patients diagnosed with cervical squamous cell carcinoma (CSCC) across three continents., Methods: This multicontinental retrospective study was conducted in 8 hospitals across Asia, Europe, and North America (NA). Clinicopathologic data of 595 patients with presumed early stages of CSCC, treated surgically, with curative intent was collected. Descriptive analysis and Cox regression models were produced., Results: A total of 595 patients, consisting of 445 (74.8 %) white, 75 (12.6 %) Blacks, and 75 (12.6 %) Asian patients were included. Geographical distribution comprised 69 % of patients from NA, 22 % from Europe, and 9 % from Asia. The median age at diagnosis was 46 years. The median overall survival (OS) and relapse-free survival (RFS) were 22.09 years and 21.19 years, respectively. Patient characteristics varied significantly across geographical regions, except for consensus tumor grade. Patients in Europe from middle-income countries with limited CC screening had a substantially higher risk of death than those in NA (HR, 1.79; 95 % CI, 1.13 to 2.79; p = 0.015). Patients from single center in Japan had higher risk of relapse than those from the four heterogeneous NA centers (sub-distribution hazard ratio, 2.19; 95 % CI, 1.22 to 3.95; p = 0.009), although OS did not differ significantly. Race remained statistically insignificant for survival outcomes across the three continents but seemed to influence survival outcomes in NA centers., Conclusion: Our study highlights impact of geographies and races on CSCC survival outcomes, emphasizing the need of considering these factors when developing targeted interventions against CSCC., Competing Interests: Declaration of competing interest The authors whose names are listed immediately below certify that they have NO affiliations with or involvement in any organization or entity with any financial interest (such as honoraria; educational grants; participation in speakers' bureaus; membership, employment, consultancies, stock ownership, or other equity interest; and expert testimony or patent-licensing arrangements), or non-financial interest (such as personal or professional relationships, affiliations, knowledge or beliefs) in the subject matter or materials discussed in this manuscript., (Copyright © 2024. Published by Elsevier Inc.)

Objective: We evaluated clinicopathologic parameters of patients with cervical squamous cell carcinoma (SCC) who were treated with initial surgical management and assessed their relation to survival outcomes. Specifically, we evaluated the relation between extent of lymphovascular invasion (LVI) and survival outcomes., Methods: All available tumor slides from patients with initially surgically treated cervical SCC were collected from 10 institutions and retrospectively analyzed. Standard clinicopathological parameters, tumor stroma, and extent of LVI were assessed (focal: <5 spaces, extensive: ≥5 spaces). PFS and OS were evaluated using Kaplan-Meier methodology. Univariable and multivariable Cox proportional hazards models were created to determine prognostic survival-related risk factors., Results: A total of 670 tumor samples were included in the analysis. Median age at diagnosis was 47 years (IQR: 38-60), 457 patients (72%) had a 2018 International Federation of Gynecology and Obstetrics (FIGO) stage I tumor, and 155 tumors (28%) were flat and/or ulcerated. There were 303 nonkeratinizing tumors (51%), 237 keratinizing tumors (40%), and 356 histologic grade 2 tumors (61%). Quantifiable LVI was present in 321 cases (51%; 23% focal and 33% extensive). On multivariable analysis for PFS, extensive and focal LVI had worse outcomes compared to negative LVI (HR: 2.38 [95% CI: 1.26-4.47] and HR: 1.54 [95% CI: 0.76-3.11], respectively; P = 0.02). The difference did not reach statistical significance for OS., Conclusion: Presence of LVI is a prognostic marker for patients with cervical SCC. Quantification (extensive vs. focal vs. negative) of LVI may be an important biomarker for oncologic outcome., Competing Interests: Declaration of Competing Interest Outside the submitted work, N.R. Abu-Rustum reports research funding paid to the institution from GRAIL. A. Iasonos reports consulting fees from Mylan. All other authors have no potential conflicts of interest to disclose., (Copyright © 2023 Elsevier Inc. All rights reserved.)

Objective: The objective of this study was to determine the proportion of patients with a pre-invasive endometrial lesion who meet Mayo criteria for lymph node dissection on final pathology to determine if the use of sentinel lymph node biopsy in patients with pre-invasive lesions would be warranted., Methods: All women who underwent hysterectomy for a pre-invasive endometrial lesion (atypical hyperplasia or endometrial intra-epithelial neoplasia) between 2009 and 2019 were included for analysis. Relevant statistical tests were utilized to test the associations between patient, operative, and pathologic characteristics., Results: 141 patients met inclusion criteria. 51 patients (36%) had a final diagnosis of cancer, the majority (96%) of which were Stage IA grade 1 endometrioid carcinomas. Seven patients (5%) met Mayo criteria on final pathology (one grade 3, seven size >2 cm, one >50% myoinvasive). Three of these seven patients had lymph nodes assessed of which 0% had metastases. Six of these patients had frozen section performed, and 2 met (33%) Mayo criteria intraoperatively. Of the seven patients in the overall cohort that had lymph node sampling, six had a final diagnosis of cancer and none had positive lymph nodes. Of the 51 patients with cancer, only 10 had cancer diagnosed using frozen section, and only two met intra-operative Mayo criteria. Age > 55 was predictive of meeting Mayo criteria on final pathology (p = 0.007). No patients experienced a cancer recurrence across a median follow up of 24.3 months., Conclusions: Atypical hyperplasia and endometrial intra-epithelial neoplasia portend low risk disease and universal nodal assessment is of limited value., Competing Interests: Declaration of Competing Interest The authors declare no relevant conflicts of interest., (Copyright © 2021 Elsevier Inc. All rights reserved.)

Objective: The Silva invasion pattern-based classification system stratifies endocervical adenocarcinomas (ECAs) into 3 categories corresponding to risk of metastasis and recurrence, but has only been evaluated for HPV-associated ECAs of usual type. We examined whether the Silva system is applicable to all endocervical adenocarcinomas, especially those not associated with HPV., Methods: Complete slide sets from 341 surgical specimens of ECA were collected from 7 institutions worldwide. All specimens were associated with clinical records covering at least 5 years of follow-up. Tumors were classified as HPV-associated (HPVA) or not (NHPVA) by both morphology and detection of HPV using in situ hybridization. Recurrence and survival were analyzed by multivariate Mantel-Haenszel methods., Results: Most specimens (292; 85.6%) were HPVA, while 49 (14.3%) were NHPVA. All NHPVAs were Silva pattern C, while 76.0% of HPVAs were pattern C, 14.7% pattern A, and 9.3% pattern B. Including both HPVAs and NHPVAs, lymphovascular invasion (LVI) was detected in 0% of pattern A, 18.5% of pattern B and 62.6% of pattern C cases (p < 0.001). None of the pattern A or B cases were associated with lymph node metastases (LNM), in contrast to pattern C cases (21.8%). Among patients with Silva pattern C ECA, those with HPVA tumors had a lower recurrence rate and better survival than those with NHPVA; however, when adjusted for stage at diagnosis, the difference in recurrence and mortality was small and not statistically significant., Conclusions: Application of the Silva system is only relevant in HPVA cervical adenocarcinoma., (Copyright © 2018 Elsevier Inc. All rights reserved.)

Objective: A subset of uterine serous carcinoma (USC) may have better clinical behavior bringing up the possibility that there may be morphologic features, which would help in their categorization. The aim of this study is to evaluate the potential use of the MD Anderson Cancer Center 2-tier grading system for ovarian carcinoma in USC., Methods: Tumors assigned a combined score included in this analysis were 1) low-grade: tumors without marked atypia and 12 mitoses/10 high power field (HPF) and 2) high grade: tumors with severe nuclear atypia and >12 mitoses/10 HPF. Clinicopathologic parameters evaluated included patients' age, tumor size, myometrial invasion (MI), lymphovascular invasion (LVI), lymph node (LN), FIGO stage, and patient outcome., Results: 140 patients with USC were included, 30 low grade uterine serous carcinoma (LGUSC) and 110 high grade uterine serous carcinoma (HGUSC). Of all parameters only 2 (MI and stage IA) reached statistical significance. 67% of LGUSC cases showed myometrial invasion versus 93.6% HGUSC cases (p = 0.003). A higher percentage of LGUSC (63.3%) versus HGUSC (32.7%) were in stage IA (p = 0.01). However, by multivariate analysis including age, LVI, stage and tumor grade only stage was an independent prognostic factor., Conclusion: The presence of atypia and mitosis across a uterine serous carcinoma is notoriously variable in magnitude and extent, potentially making evaluation of these features difficult and subsequent grading subjective. Our findings thus show that actual prognostic utility of application of MDACC two-tier grading system to uterine serous carcinoma may not be applicable., (Copyright © 2013 Elsevier Inc. All rights reserved.)

Objectives: Uterine leiomyosarcoma (uLMS) was staged using the FIGO system for endometrial cancers. The new FIGO system takes into consideration tumor size disregarding myometrial and cervical involvement. We aimed to compare the two systems and see which more accurately predicts overall survival (OS)., Methods: 86 patients with uLMS (1984-2010) were retrospectively staged using both FIGO systems. Mean OS rates were estimated using the Kaplan-Meier method., Results: More patients had stage-I disease by the new FIGO system (42 versus 33). Five versus 18 and 27 versus 5 had old and new stage-II and III diseases respectively. Five and 4 patients with old stage II and III uLMS respectively were downstaged to stage I while 18 with old stage III were downstaged to stage II. Median follow-up was 23.5 months with a median OS of 114 (95% CI, 61-166) months. Although patients with stage I tumors had a higher mean OS rate compared to those with higher stage disease by either system, patients with old stage II-IV disease showed similar mean OS rates, with stage III-IV patients having a slightly better mean OS and a similar trend was observed with the new system. Patients with new FIGO stage III had a higher mean OS rate than those with stage II or IV disease (37.6 versus 28.1 and 34.3 months). Nonetheless, no statistical significant differences were seen in OS according to stage using either system (p=0.786 and p=0.400 respectively)., Conclusion: Neither FIGO staging system is ideal in classifying patients into four clinically significant stages., (Copyright © 2012 Elsevier Inc. All rights reserved.)

Objective: Uterine serous carcinoma (USC) constitutes 10% of uterine cancers but ~40% of deaths. Tumor size is a known prognostic factor in other solid tumors. In endometriod cancers it is one element used to identify the need for complete staging, while its significance in USC is debated. Therefore tumor size was examined as an independent prognostic factor., Methods: Clinical and pathologic variables were recorded for 236 institutional patients, and those patients in the SEER database with USC. Chi-square and Fisher exact t-tests were utilized and survival data generated via Kaplan-Meier method; multivariate analysis was performed via cox-regression., Results: The patients' mean age was 67.2 years (range 40-91). Survival ranged from 0 to 184 months (mean 42.8). We used a tumor size cut-off of 1cm and noted significant associations with myometrial invasion (p<0.0001), angiolymphatic invasion (p<0.0001), peritoneal washings (p=0.03), stage (p=0.015) and positive lymph nodes (p=0.05). Furthermore, recurrence was associated with larger tumors (p=0.03). In multivariate analysis, extra-uterine disease was the only factor associated with both recurrence and survival. Review of the SEER database noted association of larger tumors with lymph node involvement and a significant survival advantage with tumors <1cm in both univariate and multivariate analysis., Conclusions: Treatment options for USC are often predicated on the surgical stage and therefore components of the staging are vitally important. The 1cm tumor-size cut-off should be studied prospectively as a prognostic indicator of survival and recurrence in USC and considered for inclusion in USC staging., (Copyright © 2012. Published by Elsevier Inc.)

Objectives: Since 1988, cervical gland involvement and stromal invasion defined stage IIA and stage IIB endometrial carcinoma. In 2009, FIGO changed the criteria for stage II disease to include only those with cervical stromal invasion. We wished to: 1) assess the reproducibility of pathologists to distinguish patterns of cervical spread, and 2) determine the prognostic significance of cervical involvement., Methods: Slides from 46 women with cervical involvement by endometrial adenocarcinoma were scored for 5 patterns of involvement by 6 experienced pathologists to determine reproducibility. To assess prognostic significance, 206 patients with FIGO 1988 stage II adenocarcinoma formed the study population with matched FIGO stage I controls., Results: At least 5 of the 6 pathologists agreed that the cervix was involved in the 46 cases. The reproducibility for cervical gland involvement and endocervical stromal invasion was slight (kappas of 0.15 and 0.28). The survival with any type of cervical involvement was not significantly different from that of matched stage I controls (p=0.18). The 5year recurrence-free survival rates were 84% for FIGO 1988 stage I, 73% for stage IIA, and 82% for stage IIB (FIGO 2009 stage II)., Conclusions: Pathologists reliably recognize cervical involvement by endometrial carcinoma. However, reproducibility for the determination of pattern of cervical spread by experienced pathologists is too low to be of clinical utility. Women with spread of carcinoma to the cervix do not have a significantly lower survival than matched stage I controls. Cervical spread should not be the basis for determination of stage II disease., (Copyright © 2012 Elsevier Inc. All rights reserved.)

Objectives: Endometrioid adenocarcinoma of the endometrium (EEC) is the most common histologic type of endometrial cancer, with stage being the most critical prognostic factor. Cervical involvement (CI), divided into IIA (epithelial involvement) and IIB (stromal invasion), is overall associated with decreased survival (70 vs 90%). However, the impact on prognosis of sub-stages IIA vs IIB is unclear. The purpose of this study was to investigate the prognostic significance of cervical involvement as well as its substaging in patients diagnosed with EEC., Methods: Eighty-one patients treated for stage II EEC were identified (1993-2003) in our institution. They were stratified into Group 1 (46) with available slides for review and Group 2 (35) with information obtained from the pathology report. All pathology reports, for all 81 patients, contained information on cervical glandular and stromal involvement. In Group 1, 1 to 6 slides of cervix (mean 3) were reviewed. Tumors were classified as Stage IIA or IIB according to the most recent FIGO criteria. Stromal invasion (SI) in Group 1 tumors was sub-classified in 4 subgroups based on depth of invasion; A) < or =1 mm; B) >1 mm and < or =3 mm; C) >3 mm and < or =5 mm and D) >5 mm. Other histopathologic parameters evaluated include grade, depth of myometrial invasion (MI), and lymphovascular invasion (LI). Clinical data included age, type of surgery, type of radiation, and survival. Statistical analysis was performed., Results: Patients ranged in age from 33-91 (median 64) years. In Group 1, 11 patients had stage IIA and 35 stage IIB tumors. Depth of SI ranged from 1-12 mm (mean 3.4 mm). The pathologists reviewing the slides in Group 1 agreed with the initial reported description of cervical glandular and stromal involvement. In Group 2, 15 patients had stage IIA and 20 stage IIB tumors with no further information regarding depth of SI. In Group 1, 12 EECs were Grade 1, 29 Grade 2, and 5 Grade 3. Thirty tumors had <50% MI, 15 showed >50% MI and LVI was present in 11. In Group 2, 13 tumors were Grade 1, 13 Grade 2, and 9 Grade 3. Twenty-one had <50% or no MI and 9 showed LVI. Median follow-up was 73 (range 5-210) months. Five- and 10-year survival rates were 83% and 78% for patients with stage IIA and 71% and 65% for stage IIB EECs respectively. By univariate analysis, age, MI, LVI and type of treatment affected survival but not substaging into IIA vs IIB or depth of SI. By multivariate analysis, only age (p=0.001), LVI (p=0.017), and type of treatment (p=0.022) were predictors of survival in stage II EECs., Conclusions: This study showed that the distinction between stage IIA and IIB or depth of SI does not affect survival in patients with EEC. LVI and type of hysterectomy performed were predictors of survival in stage II EECs. Our results suggest that substaging should be eliminated, women with suspect cervical SI should be offered a radical hysterectomy, and that the presence of LVI may be a useful tool in guiding recommendations about the need for adjuvant radiation therapy.

Objective: We undertook an extensive molecular characterization of the epidermal growth factor receptor (EGFR) gene in vulvar squamous cell carcinomas to investigate EGFR mutation and/or genomic amplification and its association with EGFR protein expression, high-risk human papillomavirus (HPV) status and clinical outcome., Methods: A cohort of 51 vulvar cancer patients distributed across all FIGO stages was selected for immunohistochemistry (IHC) and fluorescence in situ hybridization. EGFR expression and gene amplification were correlated with high-risk HPV status, EGFR mutational status and clinical prognostic variables. Fisher's exact tests, Kaplan-Meier survival estimates and a Cox proportional-hazards model were utilized., Results: EGFR gene amplification and chromosome 7 high polysomy were observed in 12% and 6% of cases, respectively. IHC of malignant tissue with 3+ staining demonstrated 100% sensitivity and 79% specificity to detect EGFR gene amplification, yielding a 39% positive predictive value. Decreased survival (p<0.025) was observed in patients with gene amplification, and was associated with a more statistically robust 3.3 hazard ratio (p<0.005) in the Cox proportional-hazards model that controlled for age at diagnosis, stage and lymph node metastasis. Univariate analysis confirmed that EGFR gene amplification was associated with the absence of high-risk HPV (p<0.001). Common activating EGFR gene mutations were not identified., Conclusion: A subset of patients with vulvar squamous cell carcinoma was identified with EGFR gene amplification that was HPV-independent and associated with poor prognosis. Given the association of EGFR amplification with response to targeted therapy in other tumor types, these patients may be candidates for therapeutic strategies that target the EGFR pathway.

Objective: The aim of this study is to investigate whether the presence of endometriosis is a prognostic factor in patients diagnosed with clear cell carcinoma (CCC) of the ovary., Methods: Retrospective chart review was performed to all patients diagnosed with CCC and endometriosis between 1975 and 2002. All pathology reports were reviewed and slides were reviewed when available. Cox regression analysis and Kaplan-Meier test were used to calculate survival prognostic factors. The level of significance was set at 0.05., Results: Eighty-four patients with CCC were identified with a 49% rate of coexisting endometriosis. Patients with tumors arising in endometriosis (n=15), with endometriosis found elsewhere in the specimen (n=26), and those without endometriosis (n=43) were analyzed comparatively. Patients with CCCs arising in endometriosis were 10 years younger (95% C.I. 0.6-18 years) than those with CCC not arising in endometriosis (P<0.05). Patients with endometriosis anywhere in the surgical specimen presented at early stage 66% of the times versus 42% for patients without endometriosis (P<0.05). Median overall survival (OS) for patients with endometriosis was 196 months (95% C.I. 28-363) versus 34 months (95% C.I. 13-55) for patients without endometriosis (P=0.01). Advanced tumor stage at diagnosis (HR 13, 95% C.I. 5-29, P=0.001) and absence of endometriosis (HR 2, 95% C.I. 1-3.9, P=0.03) were the only significant prognostic factors associated with poor survival. Disease recurrence or death among optimally and completely cytoreduced patients was 31% and 59% for those with and without endometriosis respectively (P>0.05)., Conclusions: Our study suggests that the presence of endometriosis in patients with CCC of the ovary is associated with progression free and OS advantages with no difference in initial resectability.

Objective: Uterine mixed malignant mullerian tumors (MMMT) have traditionally been excluded from clinical trials of endometrial cancer because of a belief that they are more correctly included in the sarcoma category. Recently, investigators have suggested that uterine MMMTs are actually dedifferentiated epithelial tumors and should be treated as such. The current study was undertaken to compare outcomes, stage for stage, of uterine MMMT with poor prognosis endometrial adenocarcinomas., Methods: Cases of MMMT from 1996 to 2004 were identified from the Tumor Registry after IRB consent was obtained. Retrospective chart review was performed. Cases were matched by age, stage, performance status, and surgical procedure to controls consisting of grade 3 endometrioid, papillary serous, and clear cell endometrial carcinomas from the same time period. Overall survival was compared using the Log-Rank test., Results: 68 patients with MMMT were identified. 23 were excluded due to incomplete records. Patients with MMMT ranged in age from 51 to 95 years (mean 75.3 years). Approximately half of the patients (53%) had stage III or IV disease. Of the controls, 31 (69%) had grade 3 endometrioid, 11 (24%) papillary serous, and 3 (7%) clear cell carcinoma. Median overall survival for all patients with MMMT was significantly shorter than for controls, 18 months (range 0.5-72) versus 36 months (range 0.5-123) (P = 0.02). Patients with early stage disease (stage I or II) had shorter median survival than controls, 26 months (range 3-7) vs. 95 months (range 4-123) (P = 0.003). There was no difference in median survival when comparing advanced disease (stage III or IV) to matched controls, 15 months (range 0.5-70) vs. 19 months (range 0.5-100) (P = NS)., Conclusions: Patients with uterine MMMT have a poorer prognosis than those patients with high grade epithelial tumors, especially for those with early stage disease. Given the discrepancy in survival, these patients should not be included in studies of endometrial carcinoma. Further investigations are necessary to identify factors to improve survival of these patients.

Background: Extragastrointestinal stromal tumors (eGISTs) are rare mesenchymal-derived tumors arising outside of the GI tract. eGISTs are often histologically confused with leiomyosarcoma. Distinction between eGIST and leiomyosarcoma is critical because of the unique responsiveness of eGISTs to the molecularly targeted agent imatinib., Case: A woman presented with a history of tubal spindle cell tumor that was initially diagnosed and treated as a leiomyosarcoma. Because of minimal response to sarcoma directed chemotherapy, the possibility that the tumor was in fact an eGIST was investigated and supported by immunohistochemical and mutational analyses of the c-Kit receptor tyrosine kinase. The patient currently has stable disease control on imatinib for the last 18 months., Conclusions: The possibility of eGIST should be considered in the differential diagnosis of tumors with a spindle cell morphology in the gynecologic tract especially when involving the ovary, fallopian tube, or uterine serosa.

Objective: BAG-1 has anti-apoptotic actions and is known to bind BCL-2 and steroid receptors. High levels of BAG-1 have been implicated as a prognostic indicator in breast cancer. Whether this observation can be generalized to endometrial cancer remains unknown., Methods: IRB permission was obtained for use of human discarded tissue. Immunohistochemical analyses were performed on: proliferative endometrium (PEM, 6), secretory endometrium (SEM, 28), "low-grade" neoplastic lesions (complex atypical hyperplasia and grade 1 endometrial adenocarcinomas) (19), and "high-grade" cancers (grade 2 and 3 endometrial adenocarcinomas) (13). The level of total BAG-1 and its isoforms was evaluated by Western blot in lysates from Ishikawa cells (grade 1), MFE 296 cells (grade 2), and SK-UT(2) cells (grade 3)., Results: The proportion of "high-grade" cancers with positive cytoplasmic staining for BAG-1 was higher than that of secretory endometrium (P = 0.006). Additionally, the proportion of specimens with positive staining for nuclear BAG-1 expression was significantly higher among high-grade carcinoma specimens compared to secretory specimens (P = 0.009). A high proportion (91%) of all specimens were positive for BCL-2, limiting the ability to subcategorize the other variables analyzed. There was no relationship between positive nuclear BAG-1 expression and either estrogen receptor (ER) or progesterone receptor (PR) expression. BAG-1 was expressed in the three cell lines evaluated and total BAG-1 level was not different among the different cell lines., Conclusion: BAG-1 is expressed in the endometrium. High-grade cancers stain more frequently than secretory endometrium for both cytoplasmic and nuclear BAG-1 expression, perhaps indicating an association between expression of BAG-1 and prognosis.

Objectives: The role of MDR1 in clinical paclitaxel resistance remains poorly characterized. This study sought to identify the incidence and significance of P-glycoprotein (P-gp) over-expression on survival, tumor response to paclitaxel and the effect of prior cytotoxic exposure on P-gp expression in patients with paired primary and recurrent ovarian cancer samples., Methods: Retrospective survival analysis. P-gp expression was evaluated immunohistochemically with antibodies c494 and c219., Results: Thirty-two patients were identified from the tumor registry. Median interval between primary and secondary surgery was 17.9 (5.7-40.9) months. Only five primary tumors (16%) demonstrated +++ staining for P-gp. First-line treatment contained paclitaxel in 17 patients (53%) and 26 patients (81%) had been exposed to P-gp exportable chemotherapy before second surgery. Only seven of the recurrent tumors (22%) were +++. Only one of seven (14% (95% CI 0-46%)) recurrent tumors with ++ or +++ staining responded to subsequent paclitaxel, while 8 of 10 (80% (CI 46-100%)) recurrent tumors with 0/+ staining responded (P = 0.025). In multivariate analysis of outcome following second surgery, response to paclitaxel (P = 0.004) and P-gp over-expression (P < 0.001) were significant predictors of survival., Conclusions: De novo strong P-gp over-expression is uncommon, appears to change little over time or with prior exposure to chemotherapy. However, P-gp over-expression is a significant prognostic factor, and at the time of disease, relapse is inversely correlated with tumor response to paclitaxel.

Objective: Malignant mixed müllerian tumor (MMMT) of the ovary is a rare tumor with a dismal prognosis. The most effective therapy is unknown. The current study was undertaken to characterize a group of patients treated as if they had aggressive epithelial ovarian tumors, with cytoreductive surgery and combination paclitaxel/platinum chemotherapy., Methods: Retrospective analysis of data obtained from tumor registry and hospital records of cases of malignant mixed müllerian tumor between January 1, 1992 and January 1, 2000 treated at the Massachusetts General Hospital, Brigham and Women's Hospital, and University of Vermont was performed. Only patients treated with combination paclitaxel and platinum therapy were included in the analysis. Data were collected regarding cytoreduction, response to chemotherapy, disease-free interval, and survival., Results: Fifty-five patients were identified with MMMT. Twenty-eight patients with a clearly ovarian primary had received treatment with combination paclitaxel and platinum. Paclitaxel and carboplatin was given as second-line therapy in 2 patients who had chemoresponsive but incurable disease; the remaining patients were treated with paclitaxel and platinum therapy as first-line therapy. These 28 patients had a median (range) age of 66 (46-84 years) and stage was I in 2 patients, II in 3, III in 18, and IV in 5. Treatment was generally well tolerated. Sixteen patients of 26 treated with paclitaxel and platinum as first-line therapy achieved a complete clinical response (55%) and 6 patients achieved partial response for a total response rate of 72%. Optimal cytoreduction was associated with increased time to recurrence (P = 0.001) but not with survival. Overall median survival for the 28 patients is 27.1 months., Conclusion: Although treatment fails many patients, a minority of patients with MMMT in this highly selected population do unexpectedly well. An aggressive approach with surgery and combination paclitaxel-platinum chemotherapy appears to offer very effective therapy.