No abstract

Introduction. Several studies have applied gene expression profiling to inflammatory breast cancer (IBC). Most of these studies were underpowered. Here, we present an integrated analysis of 3 distinct gene expression data sets of IBC and non-IBC (nIBC) samples to further uncover the IBC-specific molecular biology with enhanced statistical power. Materials & Methods. Three Affymetrix gene expression data sets were combined, resulting in a series of 137 IBC and 252 nIBC samples. IBC was diagnosed clinically. Each sample was classified according to several published gene signatures. Transcriptional heterogeneity was investigated using hierarchical clustering, coupled with silhouette score analysis. IBC-specific, molecular subtype-independent differences in gene expression were identified using linear regression modeling. Differentially expressed genes were translated into pathways using Ingenuity Pathway Analysis. Cox regression analysis was used to identify variables influencing distant metastasis-free survival (DMFS) in IBC. Finally, we focussed on the molecular aspects of pathological response to neodjuvant chemotherapy in patients with IBC. Results. In our series of IBC samples, 4 robust sample clusters were identified. These sample clusters were mainly associated with the different molecular subtypes (P<0.0001), all of which were identified in IBC with a similar prevalence in nIBC, except for the Luminal A subtype (9% vs. 40%; P<0.0001) and the ErbB2+ subtype (23% vs. 8%; P=0.0002). A total of 632 genes were differentially expressed. Analysis of this gene list identified an IBC-repressed network centered on TGFβ. Activated TGFβ-profiles and SMAD-profiles in the nIBC samples corroborated these findings. Consistent with published poor prognosis signatures, current survival analysis indicated that the molecular subtypes are significantly associated with prognosis in IBC. Surprisingly, in IBC, the Luminal A samples exhibited the shortest DMFS-interval (HR=4.02; P<0.05). Comparison of responders and non-responders to neoadjuvant chemotherapy suggests a prominent role for inflammation/immunity-related processes in determining the efficacy of neoadjuvant chemotherapy in IBC. Conclusions. IBC, like nIBC, is transcriptionally heterogeneous as exemplified by the identification of 4 robust sample clusters in the present series. This observation is further corroborated by the identification of all known molecular subtypes in IBC, albeit with a different distribution pattern characterized by a low frequency of Luminal A samples. Nevertheless, this phenotype is clinically relevant, as demonstrated by the poor prognosis profile. Our observations can be explained by the IBC-specific repression of TGFβ, which is a key molecule of epithelial-to-mesenchymal transition and is also known to prevent ER-expressing cells from proliferating. Finally, as in nIBC, inflammation- and immunity-related processes are important aspects of response to neoadjuvant chemotherapy in IBC. Citation Information: Cancer Res 2011;71(24 Suppl):Abstract nr PD03-01.

Background-The authors compared longitudinal patient-reported outcomes and physicianrated cosmesis with conventionally fractionated whole-breast irradiation (CF-WBI) versus hypofractionated whole-breast irradiation (HF-WBI) within the context of a randomized trial.

Among women with non-inflammatory triple-negative and triple-negative inflammatory breast cancer (IBC), the 5-yr actuarial rates of local failure after radiation are 11%–35% and 45%, respectively, in part influenced by the contribution of radioresistant cancer stem cells to these cancers. Herein we explored the radiosensitization of breast cancer stem-like cells in vitro and examined the influence on local control after post-mastectomy radiation (PMRT) among IBC patients taking statins. SUM149, SUM159 and MCF-7 cells were cultured in standard monolayer cultures and stem cell enriching anchorage independent clonogenic cultures with simvastatin and treated with increasing concentrations of radiation. Survival curves were generated and t-test was used to compare surviving fraction (SF) of groups. p < 0.05 was considered significant. Clinical information was extracted from the charts of 534 Stage III IBC patients treated with neoadjuvant systemic therapy, mastectomy and PMRT from 1970–2011. Clinical characteristics between statin and no statin cohorts were compared using the chi-square statistic. Local regional recurrence-free survival (LRFS) was determined using the Kaplan Meier method and compared with log rank. Simvastatin radiosensitized all cell lines in both types of clonogenic culture assays. The triple-negative IBC cell line SUM149 had the greatest response to combined treatment regardless of the radiation dose used in monolayer cultures (SF2: 0.417 vs 0.319, SF4: 0.136 vs 0.075, SF6: 0.026 vs 0.018, in control vs treated respectively, all p < 0.0001) and in anchorage independent cultures (SF2: 0.880 vs 0.762, SF4: 0.863 vs 0.492, SF6: 0.920 vs 0.492, in control vs treated respectively, all p < 0.0001). The triple-negative non-IBC cell line SUM159 was more sensitive to combined treatment in monolayer cultures (p < 0.001) than in stem cell enriching cultures. In a retrospective clinical study of 534 IBC patients, 84 patients used a statin. Median follow-up time for the entire cohort was 2.6 yrs. As expected, statin users were older (80% vs 44% over 50, p < 0.001) and were more often obese than non-statin users (59% vs 42%, p < 0.001). Receptor status was balanced between these groups. Statin users were more often pathologically node-negative (p < 0.01) and had fewer grade III tumors (p < 0.04). On univariate analysis actuarial 3-yr LRFS was higher among statin users, 89% vs 75%, p = 0.009. On multivariate analyisis TNBC (HR = 0.369, p < 0.0001), lymphatic invasion (HR = 0.430, p = 0.0007), neoadjuvant pathological response (HR = 0.242, p = 0.006) and statin use (HR = 0.422, p = 0.03) were independently associated with higher LRFS. Patients with IBC and triple negative non-IBC breast cancer have the highest rates of local failure and no available known radiosensitizers. Here we report significant improvement in local control after PMRT among statin users with IBC and significant radiosensitization across triple-negative and IBC cell lines of multiple subtypes using simvastatin. Clinical value in patients without hypercholesterolemia remains to be established. These encouraging data suggest simvastatin may be an appropriate radiosensitizing agent for clinical trials. Citation Information: Cancer Res 2012;72(24 Suppl):Abstract nr PD03-06.

Background: Very little is known about the survival of patients with inflammatory breast cancer (IBC) and distant metastasis. Furthermore, the American Joint Committee on Cancer classification of breast cancer does not recognize metastatic IBC as a distinct entity within stage IV. We hypothesized that the survival of patients with IBC and distant metastasis is worse than the survival of patients with stage-matched non-IBC. Patients and Methods: We retrospectively reviewed 5314 consecutive patients with stage III or IV breast cancer (IBC or non-IBC) who were treated at our institution between 1986 and 2012. A total of 1079 patients presented with IBC (stage III: 861; stage IV: 218) and 4235 non-IBC (stage III: 2781; stage IV: 1454). We compared the time to distant metastasis from initial diagnosis, distant metastasis–free survival (DMFS), and overall survival (OS) in stage-matched patients with IBC or non-IBC. Results: The median follow-up periods were 3.3 years for patients with stage III disease (range, 0-32.2 years) and 1.8 years for patients with stage IV disease (range, 0-19.9 years). The total number of recorded events (metastasis/death) was 1657 for stage III, while the numbers of deaths for stage III and IV were 1337 and 973, respectively. In patients with stage III, the time to distant metastasis was shorter in IBC than in non-IBC (median 1.3 vs. 1.7 years, P < .001). DMFS and OS were shorter in patients with stage III IBC than in those with stage III non-IBC (2.5 vs. 6.9 years, P < .001; and 4.7 vs. 8.9 years, P < .001; respectively). However, there was no significant difference in OS after development of distant metastasis between stage III IBC and non-IBC (median for both 1.3 years, P = .83). In multivariate analysis, the diagnosis of IBC remained significantly associated with mortality after adjusting for potential confounders. De novo stage IV IBC presented more frequently with multiple sites of metastasis than de novo stage IV non-IBC (P = .02). In patients with de novo stage IV disease, OS was shorter in IBC than in non-IBC (2.3 vs. 3.4 years, P = .004). In the multicovariate Cox model, while ethnicity, tumor grade, hormone receptor status and HER2 status, site of metastasis, number of sites of metastasis, and definitive breast surgery by 1 year were all significant factors in OS for stage IV breast cancer, the diagnosis of IBC conferred a hazard ratio of 1.33 (95% confidence interval: 1.05 - 1.69) in multivariate analysis. Conclusion: Our findings suggest that IBC patients with metastasis at diagnosis have worse outcomes than stage-matched non-IBC patients. IBC patients presenting with de novo stage IV disease should be considered as a separate subcategory of stage IV in the tumor-node-metastasis classification because their clinical course and prognosis are different from those of patients with stage IV non-IBC. Citation Information: Cancer Res 2013;73(24 Suppl): Abstract nr P6-12-02.

No abstract

Background Inflammatory breast cancer (IBC) is the most aggressive type of breast cancer. HMG-CoA reductase inhibitors (statins) are cholesterol reducing agents with pleiotropic effects, including antitumorigenic and anti-inflammatory properties. We hypothesized that statins reduce the metastatic potential in primary IBC. Methods We retrospectively reviewed 724 patients diagnosed with and treated for primary IBC at The University of Texas MD Anderson Cancer Center between Jan. 12, 1995 and Jan. 27, 2011. Patients with records indicating statin use at the time of IBC diagnosis on the electronic medical record were compared with those without. We further compared outcomes stratified by statin type (hydrophilic [H] versus lipophilic [L]). We used the Kaplan-Meier method to estimate the median disease-free survival (DFS) after surgery, overall survival (OS), and disease specific survival (DSS), followed by Cox proportional hazards regression model to test statistical significance of several potential prognostic factors. Results For primary IBC patients who had information on their statin use status at IBC diagnosis, the median DFS time were 4.88 years, 2.47 years and 1.76 years (P= 0.04); the median OS time 5.05 years, 3.79 years and 4.32 years (P= 0.35); and the median DSS time 5.10 years, 3.79 years and 4.52 years (P= 0.37), for patients who took “ H”, “L” and no statin, respectively. In multivariable Cox model stratified by radiation therapy, ER/PR status and HER2 status, statin “H” use was associated with significantly improved DFS compared to no statin use (HR=0.49; 95% CI: 0.28–0.84; p<0.01), adjusted for lymphatic/vascular invasion. Although there is a trend that patients who used statin “H” had a longer time to death compared to patients who did not take statin, it did not reach statistical significance for OS (HR=0.80; 95% CI: 0.43–1.49; p=0.49) and DSS (HR=0.85; 95% CI: 0.46–1.57, p=0.59) after adjustment for lymphatic/vascular invasion, nuclear grade and surgery status within one year. Conclusions Hydrophilic statin use was associated with improved DFS. There was a trend for reduced HR in OS and DSS among primary IBC patient who used hydrophilic statins. A prospective randomized study to evaluate the potential survival benefits of statins in primary IBC population is warranted. Citation Information: Cancer Res 2012;72(24 Suppl):Abstract nr PD03-08.

somatic BRAF-V600 mutations, which have been associated with the uncontrolled proliferation, radioresistance, and immune evasion of melanoma cells. The objective of this study was to investigate the prognostic value of BRAF-V600 mutational status in the progression to melanoma BM. Materials/Methods: We merged radiation oncology, neurosurgical, and cancer registry databases with molecular pathology records from a single institution to identify a retrospective cohort of 225 melanoma patients whose BRAF mutational status (BRAF-WT or BRAF-V600) had been classified. Demographic, clinicopathologic, clinical management, and outcomes data were collected from the cancer registry and electronic medical records. The primary outcome measures were the presence of BM and BM-free survival, which was defined as the time from the initial melanoma diagnosis to radiographic diagnosis of BM. A secondary outcome was survival from BM diagnosis. Results: Eighty-three of the 225 patients (36.9%) were found to have a BRAF-V600 mutation. At primary melanoma diagnosis, BRAF-V600 mutational status was associated with a younger age (P .001), female sex (P Z .0037), a more advanced AJCC pathological stage (P Z .030), lymph node involvement (P Z .047), and a family history of cancer (P Z .044). Compared to BRAF-WT patients and other sites of metastasis, BRAF-V600 patients had a higher frequency of BM (36.1% vs 24.7%), but this result did not reach statistical significance (P Z .070). American Joint Committee on Cancer pathological stage at primary diagnosis was the only factor found to be associated with BM (P Z .0097). BRAF-WT and BRAF-V600 patients had median BM-free survivals of 2.49 years versus 2.80 years, respectively. After adjusting for AJCC pathological stage, BRAF-V600 mutational status was found to be an independent predictive factor for BM-free survival (HR Z 0.57; P Z .037). BRAF-WT and BRAF-V600 patients had median survivals from BM diagnosis of 8.7 months versus 12.7 months, respectively (P Z .58). Conclusion: In this single center retrospective cohort, BRAF-V600 mutational status was associated with a higher frequency of melanoma BM. This study provides evidence on the prognostic importance of BRAF-V600 mutational status in the progression to melanoma BM.