BACKGROUND:Cancer metastasis is the main contributor to breast cancer fatalities as women with the metastatic disease have poorer survival outcomes than women with localised breast cancers. There is an urgent need to develop appropriate prognostic methods to stratify patients based on the propensities of their cancers to metastasise. The insulin-like growth factor (IGF)-I: IGF binding protein (IGFBP):vitronectin complexes have been shown to stimulate changes in gene expression favouring increased breast cancer cell survival and a migratory phenotype. We therefore investigated the prognostic potential of these IGF- and extracellular matrix (ECM) interaction-induced proteins in the early identification of breast cancers with a propensity to metastasise using patient-derived tissue microarrays.METHODS:Semiquantitative immunohistochemistry analyses were performed to compare the extracellular and subcellular distribution of IGF- and ECM-induced signalling proteins among matched normal, primary cancer and metastatic cancer formalin-fixed paraffin-embedded breast tissue samples.RESULTS:The IGF- and ECM-induced signalling proteins were differentially expressed between subcellular and extracellular localisations. Vitronectin and IGFBP-5 immunoreactivity was lower while β1 integrin immunoreactivity was higher in the stroma surrounding metastatic cancer tissues, as compared to normal breast and primary cancer stromal tissues. Similarly, immunoreactive stratifin was found to be increased in the stroma of primary as well as metastatic breast tissues. Immunoreactive fibronectin and β1 integrin was found to be highly expressed at the leading edge of tumours. Based on the immunoreactivity it was apparent that the cell signalling proteins AKT1 and ERK1/2 shuffled from the nucleus to the cytoplasm with tumour progression.CONCLUSION:This is the first in-depth, compartmentalised analysis of the distribution of IGF- and ECM-induced signalling proteins in metastatic breast cancers. This study has provided insights into the changing pattern of cellular localisation and expression of IGF- and ECM-induced signalling proteins in different stages of breast cancer. The differential distribution of these biomarkers could provide important prognostic and predictive indicators that may assist the clinical management of breast disease, namely in the early identification of cancers with a propensity to metastasise, and/or recur following adjuvant therapy.

AIMS:This retrospective, population-based study investigated the patterns and risks of mortality from breast cancer, other cancers and non-cancer causes according to the age at diagnosis of breast cancer.METHODS:Mortality was assessed in all Australian women (n = 179,653) aged 30-79 years who were diagnosed with breast cancer between 1982 and 2004 and who survived a minimum of 1 year. The mean follow up was 6.3 years (range 0-23 years).RESULTS:Before December 2005, 52,934 women had died (34,459 of breast cancer, 5019 of other cancers and 13,456 of non-cancer causes). There was an inverse age-related relative risk of mortality (calculated as the standardized mortality ratio [SMR]) from breast cancer (linear trend across age P < 0.01). For breast cancer survivors the age-adjusted SMR was 0.99 for other cancers and 0.81(P < 0.01) for non-cancer causes in comparison with the general population. The SMR for other cancers and non-cancer causes was highest in the 30-39-year-old age group (2.13, P < 0.01 and 2.15, P < 0.01, respectively), and progressively decreased with increasing age, with the 70-79-year-old age group having significantly reduced SMR (0.95, P < 0.05, and 0.78, P < 0.01, respectively, compared with the age-matched general population).CONCLUSION:There was an inverse age-related relative risk of death from breast cancer, other cancers and non-cancer causes. These findings suggest that younger Australian women require long-term health surveillance and that older women with limited comorbidities require optimal treatment of their breast cancer.

Decline in cognitive function can be experienced by up to 50% of women while undergoing adjuvant chemotherapy with a subset of patients experiencing long-term-effects. Bayesian latent basis longitudinal random-effects and latent class growth models were used to assess the degree of verbal memory recovery for learning, immediate retention and delayed recall in women undergoing chemotherapy for early stage breast cancer and who were assessed before, and at 1, 6 and 18 months post chemotherapy. The latent basis model, with the initial time point fixed to 1 and the second to 0, enabled the estimation of the degree of recovery at times 3 and 4. In the single-class model, three prior distributions for the random effects were used, resulting in minimal differences in parameter estimation. These models resulted in a near complete recovery for learning, and 57-60% for immediate retention and delayed recall at 18 months post chemotherapy. A range of two-to five-class Bayesian mixture models were fitted, with classes differing for baseline and degree of recovery. In the three-class model, the 'Low' class indicated 29.0%, 11.0% and 9.6% recovery at 18 months post treatment for the three outcome measures respectively, indicating an extended time of reduced verbal memory function.

This study examined whether chemotherapy-induced menopause affects cognitive functioning in women with early breast cancer. The neuropsychological performance of 121 breast cancer patients (age M=9.62, SD=8.11, range=25.25-67.92) treated with chemotherapy was assessed pre-chemotherapy, as well as 1, 6, and 18 months post-chemotherapy completion. Linear mixed modeling was used to evaluate the data. Type of menopause (pre, chemotherapy-induced, and post menopause) was found to significantly interact with cognitive performance on two cognitive variables. Specifically, chemotherapy-induced menopausal women did not show any significant changes in performance on an abstract reasoning task, while the pre-menopausal and post-menopausal groups significantly improved over time. A significant interaction on a test of finger dexterity and coordination was also found, although inspection of the results indicated that this was due to a significant improvement in the pre-menopausal groups at 6 months post chemotherapy. After chemotherapy most cognitive variables showed improvements over time, although two indicators of verbal memory showed significant declines immediately after chemotherapy, with improvement by 18 months post completion. The current study found little evidence to suggest that chemotherapy-induced menopause broadly affects cognitive functioning after treatment administration. However, longer follow-up assessments are warranted to assess the long-term effects of combined chemotherapy and endocrine treatment.

Informed consent and subject protection are internationally mandated requirements for the ethical conduct of research; however, the monitoring of the day-to-day conduct of research may be insufficient for ensuring consistent compliance with required ethical ideals. Oncology nurses were surveyed about their perceptions of ethical issues relevant to cancer trials research. Utilising an investigator-developed instrument, multi-item scales assessed six ethical domains. Of 192 respondents, 95% or more held definite views in 12 of 15 items about patient understanding of cancer trials, informed consent and the welfare of participants. Approximately 95% perceived that patients consented freely and knew how to withdraw from a trial, and 81% perceived better monitoring of trial than non-trial patients. However, more than 80% of respondents perceived that at times patients had unrealistic expectations of participation, and more than 50% perceived that participants sometimes did not understand the nature and risk of cancer trials. Although the conative attributes of patients place limits on the goals of bioethics, the results of this study show first that oncology nurses have opinions about ethical constructs directly linked to the daily conduct of cancer clinical trials, and second that this link warrants further investigation in order to benchmark trial conduct against the ideals of ethical research.

This paper investigates the impact of chemotherapy on cognitive function of breast cancer patients and whether this response is homogeneous for all patients. Latent class piecewise linear trajectory (growth) models were employed to describe changes and identify subgroups in three Auditory Verbal Learning Test measures (learning, immediate retention and delayed recall) in 130 breast cancer patients taken at three time periods: before chemotherapy and 1 and 6 months post-chemotherapy. Two distinct subgroups of women exhibiting different patterns of response were identified for learning and delayed recall and three for immediate retention. The groups differed in level (intercept) at 1 month post-chemotherapy and patterns of decline and recovery. Binomial and multinomial logistic regressions on the latent classes found that age, initial National Adult Reading Test (NART)-predicted IQ, stage of cancer and the initial Functional Assessment of Cancer Therapy-Breast subscale (or subsets thereof) to be significant predictors of classes.

INTRODUCTION:HER2 gene amplification or overexpression occurs in 15% to 25% of breast cancers and has implications for treatment and prognosis. The most commonly used methods for HER2 testing are fluorescence in situ hybridization (FISH) and immunohistochemistry. FISH is considered to be the reference standard and more accurately predicts response to trastuzumab, but is technically demanding, expensive, and requires specialized equipment. In situ hybridization is required to be eligible for adjuvant treatment with trastuzumab in Australia. Bright-field in situ hybridization is an alternative to FISH and uses a combination of in situ methodology and a peroxidase-mediated chromogenic substrate such as diaminobenzidine [chromogenic in situ hybridization (CISH)] or multimer technology coupled with enzyme metallography [silver-enhanced in situ hybridization (SISH)] to create a marker visible under bright-field microscopy. CISH was introduced into diagnostic testing in Australia in October 2006. SISH methodology is a more recent introduction into the testing repertoire. An evaluation of CISH and SISH performance to assess patient outcome were performed using tissue microarrays. MATERIALS AND METHODS:Tissue microarrays were constructed in duplicate using material from 593 patients with invasive breast carcinoma and assessed using CISH and SISH. Gene amplification was assessed using the American Society of Clinical Oncology/College of American Pathologists guideline and Australian HER2 Advisory Board criteria (single probe: diploid, 1 to 2.5 copies/nucleus; polysomy >2.5 to 4 copies/nucleus; equivocal, >4 to 6 copies/nucleus; low-level amplification, >6 to 10 copies/nucleus and high-level amplification >10 copies/nucleus; dual probe HER2/CHR17 ratio: nonamplified <1.8, equivocal 1.8 to 2.2, amplified >2.2). RESULTS:Results were informative for 337 tissue cores comprising 230 patient samples. Concordance rates were 96% for HER2 single probe CISH and SISH and 95.5% for single probe CISH and dual probe HER2/CHR17 SISH. Both bright-field methods correlated with immunohistochemistry results and with breast cancer-specific survival. CONCLUSIONS:HER2 SISH testing combines the advantages of automation and bright-field microscopy to facilitate workflow within the laboratory, improves turnaround time, and correlates with patient outcome.

Abstract The objective of this study is to identify whether decline in cognitive functioning after chemotherapy in women with breast cancer is associated with health/disease, treatment, and psychological variables. Neuropsychological performance, health/disease, and treatment-related information of 136 women with breast cancer (age M = 49.38; SD = 7.92; range = 25.25–67.92) was assessed pre-chemotherapy and 1-month post-chemotherapy. The Reliable Change Index corrected for practice (RCIp) identified women whose performance significantly declined, while Pearson correlations assessed the relationship between cognitive change and predictor variables. A total of 16.9% of women showed significant decline post-chemotherapy, with affected domains including verbal learning and memory, abstract reasoning, and motor coordination. Decline in hemoglobin levels and increased anxiety over the course of chemotherapy was found to significantly predict impairment in multiple cognitive measures. Change in specific cognitive measures was significantly associated with baseline fatigue, depression, and functional well-being (r = 0.23 to 0.33; p = .01 to < .001). Although the effects are small, there is evidence that psychological and health factors may increase vulnerability to cognitive dysfunction after chemotherapy for breast cancer. Significant associations reported in this study may be useful in the identification and treatment of at-risk individuals. (JINS, 2009, 15, 951–962.)

The main evidences of epidemiology, diagnostic imaging, pathology, surgery, radiotherapy, chemotherapy and follow-up are reviewed to optimize the routine treatment of rectal cancer according to a multidisciplinary approach. This paper reports on the knowledge shared between different specialists involved in the design and management of the multidisciplinary ESTRO Teaching Course on Rectal Cancer. The scenario of ongoing research is also addressed.In this time of changing treatments, it clearly appears that a common standard for large heterogeneous patient groups have to be substituted by more individualised therapies based on clinical-pathological features and very soon on molecular and genetic markers. Only trained multidisciplinary teams can face this new challenge and tailor the treatments according to the best scientific evidence for each patient.

6619 Background: Survivorship research after a breast cancer (BC) diagnosis has identified psychosocial, functional and comorbidity outcomes that adversely influence daily life. Little is known, however, about non-breast cancer (NBC) mortality after a BC diagnosis. Methods: This retrospective cohort study included Australian women aged 30–79 years diagnosed with BC between January 1982 and December 2004, with mortality examined from January 1993 to December 2005. The age-standardized mortality ratio (SMR) was used to compare NBC causes of death in the BC cohort and the age-matched Australian female population between January 1993 and December 2005. Results: The median follow-up time for the cohort of 159,550 Australian women diagnosed with BC was 7.04 years (range 0–24.7 years). During the follow-up period, 30,819 (19.3%) died of BC, 5,403 (3.4%) died of other cancers (OC), 18,349 (11.5%) died of non-cancer (NC) causes, while 104,979 (65.8%) were still alive at the end of the follow-up period. Excluding the first year after diagnosis, women with a prior BC diagnosis had a higher risk of mortality (SMR = 328.6.95% CI = 324.8–332.3, p < 0.001) than the age matched female population. NBC mortality risk was elevated for OC (SMR = 103.7, CI = 100.1–107.5, p = 0.022) but reduced for NC causes (SMR = 94.9, CI = 92.3–97.5, p < 0.001). For the age groupings 30–49 years (n = 46,736), 50–69 years (n = 82,967), and 70–79 years (n = 29,847), SMR's for NC causes were 168.2 (CI = 144.1–195.1, p < 0.001), 100.9 (CI = 95.9–106.1, p = 0.355), and 90.5 (CI = 87.5–93.5, p < 0.001) respectively. Following the first year of diagnosis, the SMR for OC was non-significantly elevated with time after diagnosis. In comparison, the SMR for NC causes reduced with time after diagnosis; SMR = 96.4 (CI = 92.1–100.8, p = 0.055) at 2–5 years after diagnosis, 98.2 (CI = 93.7–102.8, p = 0.222) at 6–10 years, 92.8 (CI = 87.2–98.7, p = 0.008) at 11–15 years and 82.2 (CI = 74.8–90.2, p < 0.001) at more than 15 years. Conclusions: BC survivors have a higher risk of mortality from OC compared to the aged-matched Australian female population, but a reduced risk from NC causes. Although NC mortality is lower in older women and as time from BC diagnosis increases, women less than 50 years have a higher risk of NC mortality. No significant financial relationships to disclose.

e11513 Background: This study examines breast cancer patients’ individual changes in health related quality of life (HRQOL) measures over time and the relationships between these measures within the framework of Structural equation modeling (SEM). Methods: One hundred and fifty five Australian women with breast cancer were asked to complete the Functional Assessment of Cancer Therapy Breast, Version 4 (FACT-B) pre- and post-chemotherapy as measures of HRQOL. In SEM, growth curve modelling was required for the estimation of individual differences in developmental changes of HRQOL scales over time. A growth mixture model was also used for identifying the cross-group differences in the pattern of changes in a given population over time. Results: Reliability and validity of the FACT-G including the fatigue subscale and breast cancer subscale for evaluating Australian breast cancer patients was determined to be sufficient for research purposes. After adjustment for covariates, two subgroups of women were identified with respect of patterns of response, representing largely increasing and decreasing trends in HRQOL respectively. In the former group, on average 85% of breast cancer women reported changes over time with respect to individual HRQOL measures and 78% reported such changes with respect to combined HRQOL Scores. A small group of women reported changes with respect to HRQOL measures in the decreasing group. The greatest contribution to variation in HRQOL measures was in the breast cancer subscale for both individual and combined scores. Subsequent factor analysis of the breast cancer subscale revealed that with one exception, all items loaded onto three separate factors described as self-esteem, physical and psychological. The latter two improved over time, but self- esteem was largely unchanged. Conclusions: The study presents a technique that allows us to identify potential subgroups of study subjects that experience different trajectories of responses. The results of this study provide greater insight into HRQOL changes over time and can help to plan interventions in a better way, ultimately contributing to improvement in HRQOL of women with breast cancer. No significant financial relationships to disclose.

BACKGROUNDClinical trials frequently report acute myeloid leukemia (AML) as a complication of adjuvant chemotherapy for breast cancer (BC).PATIENTS AND METHODSThis retrospective population-based study investigated AML risk after a prior BC diagnosis and compared the results with women after a prior diagnosis of hematological malignancies (HM), other cancers combined (OCC), and the age-matched Australian female population.RESULTSWomen with a prior BC diagnosis had 2.56 times the risk of developing AML compared with the Australian female population (P<0.001). AML risk was also elevated after prior HM and OCC diagnoses (4.73, P<0.001, and 1.70, P<0.001, respectively). Although the incidence of AML rose sharply with age in all cohorts, the age-specific relative risk was highest in the 30- to 49-age-group and decreased with increasing age. AML risk increased with the duration of follow-up but there was no change of risk during the 23 years of this study.CONCLUSIONAML risk was elevated after a prior diagnosis of BC but there was no evidence of an increasing risk of AML after a BC diagnosis or, in any of the other cancer cohorts, during this era of expansion of the evidence base for more intensive treatments.

7052 Background: AML is a widely reported complication of chemotherapy for HM and BC but few population based studies have investigated the nature and magnitude of AML risk after a diagnosis of cancer. Methods: This retrospective study examined 3 cohorts of female patients 30 years and over reported to state and territory cancer registries in Australia between 1982 and 2002. The cohorts comprised 40,030 women with HM, 162,142 with BC and 379,347 with OC. Results: The AML incidence rate/100,000 (IR) was 4.1 for all Australian females. The rate was much higher in the cohort with HM (IR 36.2; CI 25.3 –51.7), and slightly higher among those with BC (IR 11.7; CI 9.4 –14.9) and OC (IR 7.7; CI 6.4 –9.2). Standardized incidence ratio (SIR) estimates confirmed that women in the 3 cohorts had higher risks of subsequent AML than the general female population. These ranged from 5.5 times higher for those with HM (p <0.001), 2.7 for BC (p <0.001) and 1.7 for OC (p <0.001). Although the IR for AML rose sharply with age, the SIR estimates for HM and BC, but not OC, were highest in the young age group and decreased with increasing age (30–49 years HM 21.6, BC 5.1; 50–69 years HM 10.4, BC 2.9; 70 years and older HM 1.8, BC 1.9; all p values < 0.001). There was no evidence that time to diagnosis curves for AML were different for the 3 cohorts (53% within 5 years for HM, 60% for BC and 59% for OC). Conclusion: AML risk was higher for female cancer survivors than for the non-cancer population and was a continuing risk for long term survivors. Although the absolute risk increased with age, the relative risk was highest in young survivors of HM and BC, and decreased with age. There was no evidence of a disproportionate increase of AML risk for cancer survivors during this era of expansion of the evidence base for more extensive chemotherapy, although there was a significant increase in the risk of AML for the Australian female population. No significant financial relationships to disclose.

General Sessions [#11-82] S5–S23 Poster Discussion Sessions [#101-511] S24–S40 Poster Session I [#1001-1119] S41–S83 Poster Session II [#2001-2121] S84–S125 Poster Session III [#3001-3113] S126–S165 Poster Session IV [#4001-4117] S166–S206 Poster Session V [#5001-5119 (excl. 5015)] S207–S246 Poster Session VI [#6001-6119 plus 5015] S247–S287 Author index for abstracts S288–S302

Aim: HER-2/neu amplification occurs in 15-25% of breast carcinomas. This oncogene, also referred to as c-erbB-2, encodes a transmembrane tyrosine kinase receptor belonging to the epidermal growth factor receptor family. HER-2 over-expression is reported to be associated with a poor prognosis in breast carcinoma patients and in some studies is associated with a poorer response to anti-oestrogen therapy. These patients are less likely to benefit from CMF (cyclophosphamide, methotrexate, fluorouracil)-based chemotherapy compared with anthracycline-based chemotherapy. The aim of this study was to evaluate breast carcinomas to determine hormone receptor status and if there is a difference in breast cancer specific survival for HER-2 positive patients.Methods: A total of 591 breast carcinomas were evaluated using immunohistochemistry (IHC) for oestrogen receptor (ERp), progesterone receptor (PRp) and three different HER2 antibodies (CB11, A0485 and TAB250). Percentage of tumour cells and intensity of staining for ERp were evaluated using a semiquantitative method.Results: Of the 591 tumours, 91 (15.4%) showed 3+ membrane staining for HER-2 with one or more antibodies. Of these 91 tumours, 41 (45.1%) were ERp+/ PRp+, seven (7.7%) were ERp+/PR-, six (6.6%) were ERp-/PRp+ and 37 (40.7%) were ERp-/PR-. Of HER-2 positive tumours, 5.5% showed > 80% 3+ staining for ERp compared with 31.8% of 0-2+ HER-2 tumours; 24.2% of HER-2-positive tumours showed 60% or more cells with 2+ or 3+ staining for ERp. Treatment data were available for 209 patients and no difference was observed in breast cancer specific survival (BCSS) with HER-2 status and tamoxifen.Conclusion: Oestrogen receptor status cannot be used to select tumours for evaluation of HER-2 status, and oestrogen and progesterone receptor positivity does not preclude a positive HER-2 status. There is a higher proportion of ERp negative tumours associated with HER-2 positivity, however, more than 20% of HER-2 positive tumours show moderate or strong staining for ERp. HER-2 positive patients in this study did not show an adverse BCSS with tamoxifen treatment unlike some previous studies.