Hewitt, Stephen M.

Terminal duct lobular units (TDLUs) are the predominant source of future breast cancers, and lack of TDLU involution (higher TDLU counts, higher acini count per TDLU and the product of the two) is a breast cancer risk factor. Numerous breast cancer susceptibility single nucleotide polymorphisms (SNPs) have been identified, but whether they are associated with TDLU involution is unknown. In a pooled analysis of 872 women from two studies, we investigated 62 established breast cancer SNPs and relationships with TDLU involution. Poisson regression models with robust variance were used to calculate adjusted per‐allele relative risks (with the non‐breast cancer risk allele as the referent) and 95% confidence intervals between TDLU measures and each SNP. All statistical tests were two‐sided; P < 0.05 was considered statistically significant. Overall, 36 SNPs (58.1%) were related to higher TDLU counts although this was not statistically significant (p = 0.25). Six of the 62 SNPs (9.7%) were nominally associated with at least one TDLU measure: rs616488 (PEX14), rs11242675 (FOXQ1) and rs6001930 (MKL1) were associated with higher TDLU count (p = 0.047, 0.045 and 0.031, respectively); rs1353747 (PDE4D) and rs6472903 (8q21.11) were associated with higher acini count per TDLU (p = 0.007 and 0.027, respectively); and rs1353747 (PDE4D) and rs204247 (RANBP9) were associated with the product of TDLU and acini counts (p = 0.024 and 0.017, respectively). Our findings suggest breast cancer SNPs may not strongly influence TDLU involution. Agnostic genome‐wide association studies of TDLU involution may provide new insights on its biologic underpinnings and breast cancer susceptibility.

AbstractAutomated methods are needed to facilitate high‐throughput and reproducible scoring of Ki67 and other markers in breast cancer tissue microarrays (TMAs) in large‐scale studies. To address this need, we developed an automated protocol for Ki67 scoring and evaluated its performance in studies from the Breast Cancer Association Consortium. We utilized 166 TMAs containing 16,953 tumour cores representing 9,059 breast cancer cases, from 13 studies, with information on other clinical and pathological characteristics. TMAs were stained for Ki67 using standard immunohistochemical procedures, and scanned and digitized using the Ariol system. An automated algorithm was developed for the scoring of Ki67, and scores were compared to computer assisted visual (CAV) scores in a subset of 15 TMAs in a training set. We also assessed the correlation between automated Ki67 scores and other clinical and pathological characteristics. Overall, we observed good discriminatory accuracy (AUC = 85%) and good agreement (kappa = 0.64) between the automated and CAV scoring methods in the training set. The performance of the automated method varied by TMA (kappa range= 0.37–0.87) and study (kappa range = 0.39–0.69). The automated method performed better in satisfactory cores (kappa = 0.68) than suboptimal (kappa = 0.51) cores (p‐value for comparison = 0.005); and among cores with higher total nuclei counted by the machine (4,000–4,500 cells: kappa = 0.78) than those with lower counts (50–500 cells: kappa = 0.41; p‐value = 0.010). Among the 9,059 cases in this study, the correlations between automated Ki67 and clinical and pathological characteristics were found to be in the expected directions. Our findings indicate that automated scoring of Ki67 can be an efficient method to obtain good quality data across large numbers of TMAs from multicentre studies. However, robust algorithm development and rigorous pre‐ and post‐analytical quality control procedures are necessary in order to ensure satisfactory performance.

Abstract A genome-wide association study (GWAS) of pancreatic cancer conducted within the NCI Cohort Consortium (PanScan I and II) identified pancreatic cancer susceptibility loci on chromosomes 1q32.1/NR5A2, 5p15.33/CLPTM1L/TERT, 9q34.2/ABO, and 13q22.1. The most significant single-nucleotide polymorphism (SNP) identified on 13q22.1, rs9543325, lies in a ~600 kb gene desert; the nearest genes are KLF5, KLF12, PIBF1, DIS3, and BORA (265–586 kb away). Imputation using the 1000 Genomes and DCEG reference datasets did not improve the GWAS signal, but produced a set of highly correlated SNPs for functional follow-up. We performed eQTL analysis to test for association between the genotypes of these functional candidate variants and expression of nearby genes. Among 64 normal derived pancreatic tissue samples, DIS3 expression showed the strongest association with a 30 bp indel variant in the risk locus (P = 4.8 × 10-4), indicating risk alleles associate with reduced DIS3 expression. Mutations in DIS3 have previously been identified in acute myeloid leukemia and multiple myeloma, and its expression has been correlated with metastatic potential in colorectal cancer, suggesting DIS3 is relevant to cancer biology. Chromosome conformation capture identified a physical interaction between the indel-containing locus and a region near the DIS3 promoter. Luciferase assay for regulatory function of this indel-containing locus revealed allele-specific silencer activity for the insertion allele. Supershift electromobility shift assay (EMSA) demonstrated binding of LEF1 specifically to the insertion allele of the indel, which contains two in silico predicted LEF1 binding elements. Finally, through immunohistochemical analysis, high DIS3 protein levels associated with better survival for pancreatic cancer patients (hazard ratio = 2.87, 95% CI = 1.49–5.53, P = 0.001). Our results suggest that at least one target gene for the pancreatic cancer risk variants on chr13q22.1 may be DIS3, and that the underlying biology may be mediated by the novel indel through a long-range repressive effect on DIS3 expression. Citation Format: Jason W. Hoskins, Abdisamad Ibrahim, Mickey Emmanuel, Sarah Manmiller, Jinping Jia, Hemang Parikh, Irene Collins, Kris Ylaya, Sean F. Altekruse, Stephen M. Hewitt, Gloria M. Petersen, Laufey T. Amundadottir. Functional analysis of the chr13q22.1 pancreatic cancer risk locus suggests allele-specific effects on DIS3 expression with prognostic implications. [abstract]. In: Proceedings of the AACR Special Conference on Translation of the Cancer Genome; Feb 7-9, 2015; San Francisco, CA. Philadelphia (PA): AACR; Cancer Res 2015;75(22 Suppl 1):Abstract nr A1-09.

Abstract A genome-wide association study (GWAS) of pancreatic cancer conducted within the NCI Cohort Consortium (PanScan I and II) identified pancreatic cancer susceptibility loci on chromosomes 1q32.1/NR5A2, 5p15.33/CLPTM1L/TERT, 9q34.2/ABO, and 13q22.1. The most significant single-nucleotide polymorphism (SNP) identified on 13q22.1, rs9543325, lies in a 600 kb gene desert; the nearest genes are KLF5, KLF12, PIBF1, DIS3, and BORA (265–586 kb). Imputation using the 1000 Genomes and DCEG reference datasets did not improve the GWAS signal, but produced a set of highly correlated SNPs for functional follow-up. We performed eQTL analysis to test for association between the genotypes of these functional candidate variants and expression of nearby genes. Among 100 normal derived pancreatic tissue samples, DIS3 expression showed the strongest association with a novel 30 bp indel variant in the risk locus (P = 4.0 × 104), indicating risk alleles associate with reduced DIS3 expression. Mutations in DIS3 have previously been identified in acute myeloid leukemia and multiple myeloma, and its expression has been correlated with metastatic potential in colorectal cancer, suggesting DIS3 is relevant to cancer biology. Chromosome conformation capture identified a physical interaction between the indel-containing locus and a region near the DIS3 promoter. Luciferase assay for regulatory function of this indel-containing locus revealed allele-specific silencer activity for the insertion allele. Supershift electromobility shift assay (EMSA) demonstrated binding of LEF1 specifically to the insertion allele of the indel, which contains two in silico predicted LEF1 binding elements. Finally, through immunohistochemical analysis, high DIS3 protein levels associated with better survival for pancreatic cancer patients (hazard ratio = 2.87, 95% CI = 1.49–5.53, P = 0.001). Our results suggest that at least one target gene for the pancreatic cancer risk variants on chr13q22.1 may be DIS3, and that the underlying biology may be mediated by the novel indel through a long-range repressive effect on DIS3 expression. Citation Format: Jason W. Hoskins, Abdisamad Ibrahim, Mickey Emmanuel, Hemang Parikh, Jinping Jia, Irene Collins, Kris Ylaya, Sean F. Altekruse, Stephen M. Hewitt, Gloria M. Petersen, Laufey T. Amundadottir. Functional analysis of the chr13q22.1 pancreatic cancer risk locus suggests allele-specific effects on DIS3 expression with prognostic implications. [abstract]. In: Proceedings of the AACR Special Conference on Pancreatic Cancer: Innovations in Research and Treatment; May 18-21, 2014; New Orleans, LA. Philadelphia (PA): AACR; Cancer Res 2015;75(13 Suppl):Abstract nr B111.

Introduction: The advent of digital slides offers new opportunities within the practice of pathology such as the use of image analysis techniques to facilitate computer aided diagnosis (CAD) solutions. Use of CAD holds promise to enable new levels of decision support and allow for additional layers of quality assurance and consistency in rendered diagnoses. However, the development and testing of prostate cancer CAD solutions requires a ground truth map of the cancer to enable the generation of receiver operator characteristic (ROC) curves. This requires a pathologist to annotate, or paint, each of the malignant glands in prostate cancer with an image editor software - a time consuming and exhaustive process.Recently, two CAD algorithms have been described: probabilistic pairwise Markov models (PPMM) and spatially-invariant vector quantization (SIVQ). Briefly, SIVQ operates as a highly sensitive and specific pattern matching algorithm, making it optimal for the identification of any epithelial morphology, whereas PPMM operates as a highly sensitive detector of malignant perturbations in glandular lumenal architecture.Methods: By recapitulating algorithmically how a pathologist reviews prostate tissue sections, we created an algorithmic cascade of PPMM and SIVQ algorithms as previously described by Doyle el al. [1] where PPMM identifies the glands with abnormal lumenal architecture, and this area is then screened by SIVQ to identify the epithelium.Results: The performance of this algorithm cascade was assessed qualitatively (with the use of heatmaps) and quantitatively (with the use of ROC curves) and demonstrates greater performance in the identification of malignant prostatic epithelium.Conclusion: This ability to semi-autonomously paint nearly all the malignant epithelium of prostate cancer has immediate applications to future prostate cancer CAD development as a validated ground truth generator. In addition, such an approach has potential applications as a pre-screening/quality assurance tool.

Abstract Background: Terminal duct lobular units (TDLU) are the predominant source of breast cancers. Lesser degrees of age-related TDLU involution have been associated with increased breast cancer risk, but factors that influence involution are largely unknown. We assessed whether circulating hormones, implicated in breast cancer risk, are associated with levels of TDLU involution using data from the Susan G. Komen Tissue Bank (KTB) at the Indiana University Simon Cancer Center (2009–2011). Methods: We evaluated three highly reproducible measures of TDLU involution, using normal breast tissue samples from the KTB (n = 390): TDLU counts, median TDLU span, and median acini counts per TDLU. RRs (for continuous measures), ORs (for categorical measures), 95% confidence intervals (95% CI), and Ptrends were calculated to assess the association between tertiles of estradiol, testosterone, sex hormone–binding globulin (SHBG), progesterone, and prolactin with TDLU measures. All models were stratified by menopausal status and adjusted for confounders. Results: Among premenopausal women, higher prolactin levels were associated with higher TDLU counts (RRT3vsT1:1.18; 95% CI: 1.07–1.31; Ptrend = 0.0005), but higher progesterone was associated with lower TDLU counts (RRT3vsT1: 0.80; 95% CI: 0.72–0.89; Ptrend &lt; 0.0001). Among postmenopausal women, higher levels of estradiol (RRT3vsT1:1.61; 95% CI: 1.32–1.97; Ptrend &lt; 0.0001) and testosterone (RRT3vsT1: 1.32; 95% CI: 1.09–1.59; Ptrend = 0.0043) were associated with higher TDLU counts. Conclusions: These data suggest that select hormones may influence breast cancer risk potentially through delaying TDLU involution. Impact: Increased understanding of the relationship between circulating markers and TDLU involution may offer new insights into breast carcinogenesis. Cancer Epidemiol Biomarkers Prev; 23(12); 2765–73. ©2014 AACR.

Abstract Genome-wide association studies (GWAS) of 10 different cancers have identified pleiotropic cancer predisposition loci across a region of chromosome 5p15.33 that includes the TERT and CLPTM1L genes. Of these, susceptibility alleles for pancreatic cancer have mapped to the CLPTM1L gene, thus prompting an investigation of the function of CLPTM1L in the pancreas. Immunofluorescence analysis indicated that CLPTM1L localized to the endoplasmic reticulum where it is likely embedded in the membrane, in accord with multiple predicted transmembrane domains. Overexpression of CLPTM1L enhanced growth of pancreatic cancer cells in vitro (1.3–1.5–fold; PDAY7 &lt; 0.003) and in vivo (3.46-fold; PDAY68 = 0.039), suggesting a role in tumor growth; this effect was abrogated by deletion of two hydrophilic domains. Affinity purification followed by mass spectrometry identified an interaction between CLPTM1L and non-muscle myosin II (NMM-II), a protein involved in maintaining cell shape, migration, and cytokinesis. The two proteins colocalized in the cytoplasm and, after treatment with a DNA-damaging agent, at the centrosomes. Overexpression of CLPTM1L and depletion of NMM-II induced aneuploidy, indicating that CLPTM1L may interfere with normal NMM-II function in regulating cytokinesis. Immunohistochemical analysis revealed enhanced staining of CLPTM1L in human pancreatic ductal adenocarcinoma (n = 378) as compared with normal pancreatic tissue samples (n = 17; P = 1.7 × 10−4). Our results suggest that CLPTM1L functions as a growth-promoting gene in the pancreas and that overexpression may lead to an abrogation of normal cytokinesis, indicating that it should be considered as a plausible candidate gene that could explain the effect of pancreatic cancer susceptibility alleles on chr5p15.33. Cancer Res; 74(10); 2785–95. ©2014 AACR.

Abstract Background: Terminal ductal lobular units (TDLUs), the functional units of the breast for lactation, are the structures from which most breast cancers arise. TDLU involution, reflecting atrophy and loss of TDLUs, has been related to lower subsequent breast cancer risk. The predictors of such breast tissue aging, however, are not well defined. Given that estrogens and other hormones play a central role in the etiology of postmenopausal breast cancers, we conducted a study to assess the relationship between four hormones and sex hormone-binding globulin (SHBG) with markers of TDLU involution. Methods: Subjects included healthy women who volunteered between 2009 and 2011 to participate in the Komen Tissue Bank Study. Participants completed a risk factor questionnaire, provided a blood sample, and donated breast tissue for research. Our analysis was restricted to postmenopausal women who were not currently taking menopausal hormones (n=153). A pathologist, masked to patient data, assessed tissue sections for the presence of TDLUs and, among sections containing TDLUs (n=88), measured TDLU diameters for up to ten TDLUs per section. Both presence of TDLUs and TDLU diameter are inversely related to involution. Chemiluminescent immunometric assays were used to measure serum levels of estradiol (E2), SHBG, follicle-stimulating hormone (FSH), and prolactin. Percent free E2 was calculated from measured E2 and SHBG levels. Logistic regression was used to estimate odds ratios (ORs) and 95% confidence intervals (CIs) for the presence of TDLUs (none or any) associated with tertiles of serum hormone and SHBG levels. Multivariable ordinal logistic regression was used to calculate ORs, 95% CIs, and p-trend for TDLU diameter categorized as tertiles (≥218.4, 218.4-281, and ≥281 microns) with serum hormone and SHBG levels. All models were adjusted for potential confounding factors. Results: The study population was 87.6% Caucasian with a mean age of 59.5 years, and &gt;2/3 of the women were overweight or obese. Increasing percent free E2 was significantly associated with larger TDLU diameter (OR&gt;2.53 vs. ≥1.99: 5.58, 95% CI: 1.48-21.07; p-trend=0.01), whereas increasing SHBG was associated with smaller TDLU diameter (OR&gt;73nmol/L vs. ≥38 nmol/L: 0.19, 95% CI: 0.05-0.74; p-trend=0.02). Total E2 (p-trend=0.49), FSH (p-trend=0.85), and prolactin (p-trend=0.58) were not associated with TDLU diameter. We did not find significant relationships between the presence of TDLUs and any of the serum levels. Our observed relationships were not modified by age or other breast cancer risk factors, including BMI, years since menopause, and history of bilateral oophorectomy. Conclusion: Based on limited data, these findings suggest that sex hormones may play a role in TDLU involution. Thus, TDLU involution may represent one mechanism by which hormone levels influence breast cancer risk. TDLU measurements may therefore be helpful in assessing breast cancer risk as a marker of breast tissue aging. Analysis of the relationships between hormones and TDLU involution among premenopausal women may provide additional insights into early events in breast carcinogenesis. Citation Format: Zeina G. Khodr, Mark E. Sherman, Ruth M. Pfeiffer, Gretchen L. Gierach, Louise A. Brinton, Roni T. Falk, Deesha Patel, Laura Linville, Susan E. Clare, Daniel Visscher, Carolyn Mies, Stephen Hewitt, Anna Maria Storniolo, Jonine D. Figueroa. Endogenous hormone levels and lobular involution of the breast among healthy, postmenopausal women in the Komen Tissue Bank Study. [abstract]. In: Proceedings of the Eleventh Annual AACR International Conference on Frontiers in Cancer Prevention Research; 2012 Oct 16-19; Anaheim, CA. Philadelphia (PA): AACR; Cancer Prev Res 2012;5(11 Suppl):Abstract nr B89.

Abstract Background: Terminal duct lobular units (TDLUs) are the anatomical source of most breast cancers. TDLU involution, manifested as atrophy and loss of TDLUs, is associated with reduced breast cancer risk. We hypothesize that breast cancer susceptibility loci are associated with TDLU involution. Accordingly, we evaluated relationships between 25 established breast cancer susceptibility markers and their association with TDLU involution in the Susan G. Komen for the Cure® Tissue Bank (KTB). Methods: Subjects were women ages 18-84 (N=680), who had no personal history of any cancer, were not currently taking menopausal hormones, completed a risk factor questionnaire, provided an Oragene® saliva sample, and donated a 10-guage needle tissue core from the upper outer breast quadrant for research. We genotyped 25 known single nucleotide polymorphisms (SNPs) associated with breast cancer risk using TaqMan® assays. Pathologists masked to risk data classified three features inversely related to TDLU involution using one histological section of breast tissue/subject: number of TDLUs per section (TDLU density), TDLU diameters and number of acini per TDLU. SNPS were compared to metrics of TDLU involution categorized in tertiles: TDLU density (0-1, 2-8, 9+), mean TDLU diameter (53-291, 292-415, 416-1143 microns), and mean number of acini per TDLU (2-10, 11-20, 21+). We estimated age-adjusted per-allele odds ratios (ORs) and 95% confidence intervals (CIs) using ordinal logistic regression models with TDLU measures as the outcome and SNPs as the explanatory variables. Associations with TDLU density were based on up to 680 women with genotype data, while associations with diameter and acini measures were based on up to 469 women with observed TDLUs. Results: The rs3803662 (TOX3/TNRC9) SNP risk allele showed an association with increased TDLU density, larger TDLUs, and more acini in TDLUs (TDLU density per-allele OR=1.29 95%CI=1.04-1.60; TDLU diameter OR=1.32 95%CI=1.00-1.74; and TDLU acini OR=1.34 95%CI=1.02-1.77), consistent with previous observations of significant associations seen for this SNP with mammographic density (a well-established risk factor for breast cancer), breast cancer risk, and survival of breast cancer. In addition, we observed associations for four other SNPs with TDLU density (rs311499, rs10483813, rs2981582, rs3817198), for one other SNP with mean TDLU diameter (rs11249433), and for three other SNPs with mean TDLU acini number (rs704010, rs10941679, rs11249433). Conclusion: These data suggest that breast cancer susceptibility SNPs may be associated with TDLU involution levels, and could represent a mechanism by which they influence breast cancer risk. These findings warrant replication in other studies and provide support for the concept that levels of TDLU involution represent an intermediate marker of breast cancer risk. Citation Format: Jonine D. Figueroa, Gretchen L. Gierach, Ruth Pfeiffer, Stephen Chanock, Louise Brinton, Deesha A. Patel, Daniel Visscher, Carolyn Mies, Stephen Hewitt, Susan Clare, Anna Maria Storniolo, Mark Sherman. Associations between breast cancer susceptibility markers and terminal duct lobular unit involution in normal breast tissues among women from the Susan G. Komen for the Cure® Tissue Bank. [abstract]. In: Proceedings of the AACR Special Conference on Post-GWAS Horizons in Molecular Epidemiology: Digging Deeper into the Environment; 2012 Nov 11-14; Hollywood, FL. Philadelphia (PA): AACR; Cancer Epidemiol Biomarkers Prev 2012;21(11 Suppl):Abstract nr 58.

Abstract Little is known about the effects of many recognized breast cancer risk factors on the normal breast. We hypothesize that risk factors affect the number and morphometry of terminal ductal lobular units (TDLUs), the microanatomical structure from which most breast cancers arise. To assess this hypothesis, we evaluated relationships between breast cancer risk factors and TDLU number, TDLU diameter, and acini per TDLU in breast tissue donated by healthy volunteers. Subjects included 786 healthy women ages 18-84 years without a personal history of any cancer, who provided a risk factor questionnaire and donated a 10-guage needle core from the upper outer breast quadrant. A pathologist, masked to subject data, assessed TDLU number per section, TDLU diameter and acini per TDLU using digitized images of hematoxylin and eosin stained sections of tissue. Seventy-two randomly selected images were examined by two additional pathologists. We generated ordered categorical variables for the number of TDLUs per section (0-1, 2-8, 9+), median TDLU diameter (53-262, 263-377, 378-1375 microns) and median number of acini per TDLU (2-10, 11-20, 21+). We computed Spearman correlations between measures, and intraobserver and interobserver agreement. Associations between breast cancer risk factors and TDLU measures were assessed by estimating odds ratios (OR) and 95% confidence intervals (CI) using ordinal logistic regression models, adjusted for age. TDLU diameter and acini count were highly correlated (rho =0.71); TDLU number per section was less strongly related to the diameter (rho= 0.18) and acini measures (rho=0.20). Inter and intraobserver agreement for TDLU measures were highly correlated (rho range 0.68-0.98). All TDLU measurements were significantly lower among women age 50 years and older compared with younger women. Compared with Caucasians, African American women had more TDLUs per section (ORadj=2.03, 95% CI= 1.06-3.89). Greater numbers of TDLUs were also related to having a family history of two or more relatives with breast cancer compared with none (ORadj=2.01, 95% CI=0.91-4.46); and for parous compared with nulliparous women (ORadj=1.61, 95% CI=1.16-2.25). In contrast, parity was associated with reduced median number of acini per TDLU (ORadj=0.60, 95% CI = 0.40-0.90) and median diameter (ORadj=0.64, 95% CI = 0.42-0.96) compared with nulliparity. Our data suggest that TDLUs are more frequent among women with a positive family history of breast cancer or African Americans. Parity, an established protective factor for breast cancer risk, was related to increased numbers of TDLUs, but fewer acini per TDLU, suggesting that TDLUs may increase with pregnancy and then involute. Our data, along with published results linking TDLU involution with a reduction in breast cancer risk, suggest TDLU morphometry may reflect early biologic changes marking breast cancer susceptibility. 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 4465. doi:1538-7445.AM2012-4465