Objective Prognosis in women with ovarian cancer mainly depends on International Federation of Gynecology and Obstetrics stage and the ability to perform optimal cytoreductive surgery. Since ovarian cancer has a heterogeneous presentation and clinical course, predicting progression-free survival (PFS) and overall survival (OS) in the individual patient is difficult. The objective of this study was to determine predictors of PFS and OS in women with advanced stage epithelial ovarian cancer (EOC) after primary cytoreductive surgery and first-line platinum-based chemotherapy.Design Retrospective observational study.Setting Two teaching hospitals and one university hospital in the south-western part of the Netherlands.Population Women with advanced stage EOC.Methods All women who underwent primary cytoreductive surgery for advanced stage EOC followed by first-line platinumbased chemotherapy between January 1998 and October 2004 were identified. To investigate independent predictors of PFS and OS, a Cox' proportional hazard model was used. Nomograms were generated with the identified predictive parameters.Main outcome measures The primary outcome measure was OS and the secondary outcome measures were response and PFS.Results A total of 118 women entered the study protocol. Median PFS and OS were 15 and 44 months, respectively. Preoperative platelet count (P = 0.007), and residual disease <1 cm (P = 0.004) predicted PFS with a optimism corrected c-statistic of 0.63. Predictive parameters for OS were preoperative haemoglobin serum concentration (P = 0.012), preoperative platelet counts (P = 0.031) and residual disease <1 cm (P = 0.028) with a optimism corrected c-statistic of 0.67.Conclusion PFS could be predicted by postoperative residual disease and preoperative platelet counts, whereas residual disease, preoperative platelet counts and preoperative haemoglobin serum concentration were predictive for OS. The proposed nomograms need to be externally validated.

Objectives: Suboptimal debulking (>1 cm residual tumor) results in poor survival rates for patients with an advanced stage of ovarian cancer. The purpose of this study was to develop a prediction model, based on simple preoperative parameters, for patients with an advanced stage of ovarian cancer who are at risk of suboptimal cytoreduction despite maximal surgical effort. Methods: Retrospective analysis of 187 consecutive patients with a suspected clinical diagnosis of advanced-stage ovarian cancer undergoing upfront debulking between January 1998 and December 2003. Preoperative parameters were Karnofsky performance status, ascites and serum concentrations of CA 125, hemoglobin, albumin, LDH and blood platelets. The main outcome parameter was residual tumor >1 cm. Univariate and multivariate logistic regression was employed for testing possible prediction models. A clinically applicable graphic model (nomogram) for this prediction was to be developed. Results: Serum concentrations of CA 125 and blood platelets in the group with residual tumor >1 cm were higher in comparison to the optimally cytoreduced group (p < 0.0001 and <0.01, respectively). Serum albumin and hemoglobin levels were lower in the group with residual tumor (p < 0.0001 and <0.05, respectively). The frequency of preoperative ascites was higher in the group with residual tumor (p < 0.0005). The prediction model, consisting of CA 125 and albumin, for remaining with residual tumor showed an area under the receiver operating characteristics curve of 0.79. A nomogram for probability of residual tumor >1 cm based on serum levels of CA 125 and albumin was established. Conclusion: Postoperative residual tumor despite maximal surgical effort can be predicted by preoperative CA 125 and serum albumin levels. With a nomogram based on these two parameters, probability of postoperative residual tumor in each individual patient can be predicted. This proposed nomogram may be valuable in daily routine practice for counseling and to select treatment modality.

The effect of chemotherapy on the different components of uterine and ovarian carcinosarcoma is largely unknown. This report describes six patients with advanced carcinosarcoma, five of whom received 4 cycles of doxorubicin and ifosfamide (AI) directed at the sarcomatous component of the tumor. Responses in these five patients at second-look laparotomy were: one complete response, two partial responses (persistence of only the carcinomatous component), one stable disease, and one progressive disease (both components still present in both cases). Thereafter 4 cycles of a cisplatin-based regimen were scheduled. Response to the cisplatin-containing regimen was only evaluated clinically. The sixth patient (with no macroscopic disease left after initial surgery) received 6 cycles of a cisplatin-based chemotherapy from the onset and was found to be in complete response at relaparotomy. Median progression-free survival for all patients was 15 months and median survival 21 months. A literature survey showed that carcinosarcoma differs from adult soft tissue sarcomas with respect to responsiveness to chemotherapy. Cisplatin and ifosfamide are active as single agents, whereas the response to single-agent doxorubicin seems to be lower. The data suggest, however, that superior response rates and increased survival times are achieved with cisplatin/doxorubicin-based chemotherapy. The sensitivity of carcinosarcoma to cisplatin supports the recent view that carcinosarcoma of the female genital tract is possibly a high grade carcinoma with metaplastic sarcomatous elements.

No abstract

Valsalva. Subjective scores for pelvic floor muscle contraction were given on dynamic 3D and 4D using the standardized ICS terminology for assessment of pelvic muscle contraction (non-functioning, weak, normal or strong). Offline analysis of the datasets was undertaken using the software GE Kretz 4D View. Measurements were taken at the level of minimal hiatal dimensions using the inferior margin of the symphysis pubis as reference point on 2D and 3D datasets. On 2D ultrasound the diameter of the minimal anteroposterior (AP) hiatus, vertical displacement of the bladder neck and levator angle were measured. 3D datasets were used for obtaining AP and left-right diameter as well as area of the hiatus. Results: Complete datasets were available for 349 women. The proportional difference (value A rest-value A contraction/value A rest) between the above parameters was calculated. There were significant correlations (P < 0.001) in all parameters between women with no, weak, normal or strong contractions except for vertical bladder neck displacement. The proportional 2D AP difference between rest and contraction measured in the 2D volume seemed to be the strongest predictor, followed by 3D AP diameter and hiatal area. Conclusions: Proportional differences of the AP measurement between rest and contraction in 2D volumes is the easiest method for quantification of pelvic floor muscle contraction. Surprisingly, vertical bladder neck displacement did not seem to be significant between women with no, weak, normal or strong pelvic floor contractions. OC263 Are anal sphincter injuries related to elevator abnormalities?A. B. Steensma, K. J. Schweitzer, C. W. Burger, W. R. Schouten Erasmus Medical Center, NetherlandsObjectives: Extensive damage of the pelvic floor due to childbirth can lead to anatomical abnormalities such as levator defects and or anal sphincter injuries. This study was designed to investigate their incidence and prevalence in patients complaining of pelvic organ prolapse and or fecal incontinence with transperineal ultrasound. Methods: Some 164 patients were included in a prospective observational study. Transperineal ultrasound was performed in the supine position after voiding acquiring 3D/4D volumes at rest, contraction and on Valsalva using a GE Kretz Voluson 730 Expert system. For imaging of the anal sphincter complex a 3D microconvex transducer (RNA 5-9-MHz) was used. Offline analysis was performed with 4D View software with the investigator blinded to the clinical data. For quantification of the levator defects tomographic ultrasound imaging (TUI) was used. At the level of minimal hiatal dimension a set of eight slices with an interval of 2.5 mm was examined, and defects were scored from 0 to 8 for each side. The proportional difference (value rest − value contraction/value rest) for contraction was calculated for evaluation of pelvic floor function. Results: Forty-eight patients were diagnosed with an anal sphincter injury (29%), which involved the internal sphincter, external sphincter or both. A total ...