African American patients with colorectal cancer show higher mortality than their Caucasian counterparts. Biology might play a partial role, and prior studies suggest a higher prevalence for microsatellite instability (MSI) among cancers from African Americans, albeit patients with MSI cancers have improved survival over patients with non-MSI cancers, counter to the outcome observed for African American patients. CD8+ T cell infiltration of colon cancer is postively correlated with MSI tumors, and is also related to improved outcome. Here, we utilized a 503-person, population-based colon cancer cohort comprising 45% African Americans to determine, under blinded conditions from all epidemiological data, the prevalence of MSI and associated CD8+ T cell infiltration within the cancers. Among Caucasian cancers, 14% were MSI, whereas African American cancers demonstrated 7% MSI (P = 0.009). Clinically, MSI cancers between races were similar; among microsatellite stable cancers, African American patients were younger, female, and with proximal cancers. CD8+ T cells were higher in MSI cancers (88.0 vs 30.4/hpf, P<0.0001), but was not different between races. Utilizing this population-based cohort, African American cancers show half the MSI prevalence of Caucasians without change in CD8+ T cell infiltration which may contribute towards their higher mortality from colon cancer.

Colorectal cancer is a leading cause of cancer related deaths in the U.S., with African-Americans having higher incidence and mortality rates than Caucasian-Americans. Recent studies have demonstrated that anti-tumor cytotoxic T lymphocytes provide protection to patients with colon cancer while patients deficient in these responses have significantly worse prognosis. To determine if differences in cytotoxic immunity might play a role in racial disparities in colorectal cancer 258 microsatellite-stable colon tumors were examined for infiltrating immune biomarkers via immunohistochemistry. Descriptive summary statistics were calculated using two-sample Wilcoxon rank sum tests, while linear regression models with log-transformed data were used to assess differences in race and Pearson and Spearman correlations were used to correlate different biomarkers. The association between different biomarkers was also assessed using linear regression after adjusting for covariates. No significant differences were observed in CD8+ (p = 0.83), CD57+ (p = 0.55), and IL-17-expressing (p = 0.63) cell numbers within the tumor samples tested. When infiltration of granzyme B+ cells was analyzed, however, a significant difference was observed, with African Americans having lower infiltration of cells expressing this cytotoxic marker than Caucasians (p<0.01). Analysis of infiltrating granzyme B+ cells at the invasive borders of the tumor revealed an even greater difference by race (p<0.001). Taken together, the data presented suggest differences in anti-tumor immune cytotoxicity may be a contributing factor in the racial disparities observed in colorectal cancer.

No abstract

The images for Figure 1 and Figure 2 were inadvertently swapped. Please view the correct images and legends for Figure 1 and Figure 2.

Combination therapy, with rare exception, is a requirement for successfully treating cancer patients. Advances in genetic characterization coupled with a growing understanding of tumor heterogeneity are poised to revolutionize future strategies for combination treatments. Historically, working out the best combinations has been challenging and largely through iterative cycles of clinical trial and error due to gross limitations of preclinical models to accurately predict activity. Current strategies for developing new anticancer drugs rely heavily on preclinical testing in cancer cell lines and their derived in vivo xenograft models. These models are fraught with significant limitations including highly passaged cancer cell lines grown on plastic that ignores the microenvironment and tempered representation of the complex heterogeneity of cancer. Tumor cell microenvironment has significant impact on growth kinetics, cell signaling and response to drug treatments. 3D models attempt to recapitulate elements of the microenvironment, are more biologically relevant models compared to 2D models and have gained preference among cancer researchers and drug developers. The purpose of this study is to investigate the utility of molecularly characterized patient derived lung tumor cells (PTCs) for steering decisions for effective combination treatments in the clinical setting. We have established preclinical lung cancer models using PTCs grown in a 3D culture system. Here we present a retrospective study in lung PTCs evaluating single agents and combinations of molecularly targeted as well as cytotoxic agents including erlotinib, crizotinib, etoposide, cisplatin, carboplatin, gemcitabine, paclitaxel, vinorelbine, topotecan and irinotecan. We utilized a range of methods (PCR, FISH, IF, WB) to determine the genetic and molecular features of the PTCs prior to performing drug treatments. We used high content imaging to evaluate subpopulations within PTCs, colony morphology and proliferative endpoints. Our results identified responder and non-responder populations and emphasize the need to base combination treatment strategies on subpopulation analysis of the tumor. Patient-derived tumor cell models grown in 3D culture conditions coupled with molecular characterization are essential for conducting hypothesis driven studies. This approach offers an informed starting point for subsequent in vivo studies from which data can guide personalized medicine decisions toward true translation in the clinic. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 103rd Annual Meeting of the American Association for Cancer Research; 2012 Mar 31-Apr 4; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2012;72(8 Suppl):Abstract nr 881. doi:1538-7445.AM2012-881