Purpose: Features assessed on magnetic resonance images (MRIs) of the knees, including cartilage volume (CV), bone marrow lesions (BMLs) and synovial effusion, are associated with pain and structural progression in subjects with knee osteoarthritis (OA). Few studies, however, have examined the association of MRI findings with total knee arthroplasty (TKA) or estimated the ability of MRI findings to improve prediction of TKA over and above routine demographic, clinical and radiographic parameters. Methods: Data from the “Progression” subcohort of the Osteoarthritis Initiative (OAI) were analyzed. Subjects included in these analyses were aged 45 to 79 years and had symptomatic radiographic knee OA (defined as pain on most days of at least one month during the past year AND a definite tibiofemoral osteophyte in the same knee) in one or both knees at baseline; had at least one annual follow-up visit; had all available knee radiographs from the baseline through their 48-month follow-up visit centrally read for Kellgren-Lawrence (KL) grade by trained readers; and had baseline knee MRIs analyzed for CV at the femoral condyle and tibial plateau (mm§), presence of BMLs in the femoral condyles, and synovial fluid (SF) volume (mm§) using fully automated quantitative methodology developed by ArthoLab Inc. (Montreal, Quebec, Canada). Fixed-flexion posterior-anterior knee radiographs and knee MRIs were obtained using standard protocols with a SynaflexerTM platform and 3.0T Siemens magnets, respectively. TKA was self-reported at annual follow-up visits through 72 months and validated with medical records. Base knee-specific multiple variable Cox proportional hazards models were constructed with the following variables from the OAI baseline visit (age, gender, race, marital status, body mass index, depressive symptoms, KOOS quality of life scores, pain on motion and effusion on physical examination and KL grade) as independent variables and time to TKA as the dependent (outcome) variable. Improvement in prediction of TKA was assessed by examining the improvement in the likelihood ratio when CV, presence and size of BMLs and logSF volume were added individually and together to the best base models. Finally, change in area under the receiver operating characteristic curve (AUC) was calculated from logistic regression models further adjusted for follow-up time. Results: Of 1390 subjects enrolled in the "Progression" subcohort, 1024 and 982 subjects with symptomatic radiographic knee OA involving the right and left knees, respectively, had complete data and were included in these analyses. There were a total of 81 (8.0%) and 83 (8.4%) TKAs in the right and left knees, respectively, among these subjects during 72 months of follow-up. In knee-specific bivariate Cox proportional hazards models, medial compartment CV (P < 0.05), size of BMLs in the medial femoral condyle (P < 0.0001) and logSF volume (P < 0.0001) were significantly associated with TKA in both knees. After addition of these MRI features individually to the best base knee-specific models, however, only size of BMLs (P = 0.02 in both knees, respectively) and logSF volume (P = 0.02 in both knees) were significantly associated with TKA. Furthermore, size of BMLs and logSF volume remained significantly associated with TKA when all 3 MRI features were added to the best base model. The improved prediction based on change in AUC, however, was minimal with an increase from 0.86 to 0.87 in both knees. Conclusions: These data suggest that some MRI findings, particularly BMLs and SF volume, minimally improve the prediction of receiving a TKA in subjects with symptomatic radiographic knee OA when added to demographic, clinical and radiographic variables.

s, 5th AOS and 9th Annual Conference of OOTR, 2013 / European Journal of Cancer 49S1 (2013) S1−S22 S15 Discussion. Polz expression can be used as the predictor for poor prognosis, which might be caused by the potential chemoradiation resistance of the cervical cancer patients. The mechanism deserves further exploration. Funding. This study was supported by Science Foundation for Young Scholars of Fudan University. The authors declared no conflicts of interest. AOSP25 EXPRESSION OF SURVIVIN, FAS, CYCLIN E, AND CASPASE-3 IN APOPTOTIC BREAST CANCER CELLS INDUCED BY ELECTROCHEMICAL THERAPY B. Zhoua, C. Weib, Z. Zhangb, T. Yangb, S. Yua, J. Yua, W. Looc *, L. Chowc, A. Yipc. aNingxia People’s Hospital, Yinchuan, China, bNingxia Medical University, Yinchuan, China, cOrganisation for Oncology and Translational Research and UNIMED Medical Institute, Hong Kong Introduction. Electrochemical therapy (EChT) in malignant tumours had been identified as a possible adjunctive treatment for cancers. However, the mechanism of action remains largely unknown. In this study, we explored further the mechanisms of EChT on breast cancer cell line in vitro by detecting cell growth, cell cycle changes, cell apoptosis, and protein expression of survivin, Fas, cyclin E, and caspase-3. Methods.Humanbreast cancer cellswere treated by EChTwith different levels of electricity at 0C (control), 3C, 5C, 10C, and 20C. Cell inhibition, cell cycle distribution, and cell apoptosis were analysed by MMT and flow cytometry analysis. The expression of survivin, Fas, cyclin E, and caspase-3 in breast cancer cells was analysed with western blot. Results.The cell inhibitory rate and necrosis were significantly increased in breast cancer cells after EChT treatment compared with control (p<0.05). The apoptosis ratio within 3C, 5C, and 10C was increased obviously in order. The proportion of G0/G1 phase was higher than that in control group (p<0.05). The expression levels of survivin, cyclin E, and caspase-3 protein in the treatment group were decreased, but the expression of Fas protein was increased with higher power of EChT treatment. Discussion. The study showed that EChT could inhibit breast cancer cell growth by regulating apoptotic pathway proteins such as survivin, Fas, and cyclin E, and arresting the cell cycle at G0/G1 phase. Considering the change in expression of the regulatory proteins and strong apoptotic effect of EChT in breast cancer cells, further investigation is warranted to explore its possible treatment strategy in breast cancer. Funding. This study was supported by Natural Science grant of Ningxia, China (NZ009164). The authors declared no conflicts of interest. AOSP26 EFFECT OF RAT HYPERPLASIA SUPPRESSOR GENE ON GLIOMA IN VIVO AND IN VITRO P. Gaoa, Y. Zoua, Z. Wangc, W. Haod, H. Guoa, B. Zhanga, W. Zhaob, B. Shena *. aThe General Hospital of Ningxia Medical University, Yinchuan, China, bNingxia Medical University, Yinchuan, China, cThe Zhengzhou No. 7 People’s Hospital, Zhengzhou, China, dThe Affiliated Hospital of Yanan University,

e11515 Background: Racial disparities in breast cancer outcomes are attributed to differences in baseline tumor characteristics, stage, and socioeconomic factors. However, little is known about racial differences in treatment-related toxicities. We hypothesized that racial and ethnic differences result in differential tolerance to chemotherapy and possibly compromise to the dose intensity of adjuvant/neoadjuvant chemotherapy. Methods: Data was collected from 4 international collaborating centers (University of Miami, JBCRG (Japan Breast Cancer Research Group), University of Hong Kong, and Tom Baker Cancer Center) at which patients of different ethnic background have been treated for non metastatic breast cancer with same adjuvant or neoadjuvant chemotherapy of FEC 100: fluorouracil 500 mg/m 2 , epirubicin 100 mg/m 2 , and cyclophosphamide 100 mg/m 2 ). Racial/ethnic differences in toxicities were assessed by first episode of grade 2 or higher toxicity. Analysis of data was performed at the University of Miami. Results: Treatment-related toxicities are compared according to four race/ethnicity groups (120 Caucasian from USA and Canada (C), 16 African American (AA) from USA and Canada, 141 Japanese Asian (JA), and 23 Asian from Hong Kong (HKA)) (Table). JA and HKA had a significant higher rate of grade 3 or higher toxicity compared with C or AA women; 65%, 61%, 28%, and 31% respectively. However, there were no significant differences in chemotherapy dose intensity or density across the 4 race/ethnicity groups. Conclusions: This unique study noted racial differences in acute toxicity in women with breast cancer who were treated with FEC 100 chemotherapy. However, there are several limitations including the retrospective nature of our study, differences in practice across four countries, and different number of patients available for comparison. This study is ongoing and further statistical analyses are planned when a larger sample size is reached. [Table: see text] No significant financial relationships to disclose.

Locally advanced breast cancer (LABC) is a major problem, especially in developing countries. The standard treatment for LABC is neoadjuvant chemotherapy, with or without anti-Her2 therapy, followed by surgery, radiotherapy, and adjuvant systemic treatment if appropriate. However, there are few data in the literature addressing alternatives when neoadjuvant chemotherapy fails to reduce the tumour for surgery.We conducted a retrospective study including all patients who had non-metastatic LABC treated with neoadjuvant chemotherapy and who were not eligible for surgical resection; these patients were submitted to salvage radiotherapy (RTX) between January 2000 and December 2012 at the Brazilian National Cancer Institute.Fifty-seven patients were included, with a median age of 51 (23–72) years. The most frequent clinical stages were IIIA and IIIB, corresponding to 19.3% and 70.2%, respectively; mean tumour size was 8.74 (3–18) cm, and 44 patients (77.2%) had nodal involvement. Chemotherapeutic regimens containing anthracyclines were prescribed to 98.2% of the patients. Fifteen patients (26.3%) received taxanes and anthracyclines. Radiation dose was 50 Gy divided into 25 fractions; 43 patients (75.4%) had their tumours downsized by RTX and underwent mastectomy. Overall survival (OS) was 38 (23–52) months. Patients who were submitted to surgery had an OS of 49 (28–70) months and those who were not eligible for mastectomy after radiotherapy had an OS of 18 (9–27) months.This retrospective study confirms that RTX is an effective treatment to downsize LABC tumours with low or no response to chemotherapy, thereby enabling surgical resection which may improve overall patient outcome.