Alkushi A.,King Fahad National Guard Hospital |
Kobel M.,University of Calgary |
Kobel M.,Genetic Pathology Evaluation Center |
Kalloger S.E.,University of British Columbia |
Gilks C.B.,University of British Columbia
International Journal of Gynecological Pathology | Year: 2010
High-grade endometrial carcinomas are a heterogeneous group of tumors and include grade 3 endometrioid (EC-3), serous (SC), and clear cell carcinomas (CCC). There are conflicting data about the prognosis of these subtypes of high-grade endometrial carcinoma; this may be a result of lack of reproducibility in classifying tumor cell type. The purpose of this study was to examine differences in immunophenotype and outcome in a series of high-grade endometrial carcinomas, focusing on the comparison of EC-3 versus SC. We selected 180 endometrial carcinomas of SC, EC, or CCC type. No mixed carcinomas were included in the study. We chose the following immunohistochemical markers, estrogen receptor (ER), insulin-like growth factor 2 mRNA-binding protein 3 (IMP3), p16, p53, progesterone receptor (PR), and phosphatase and tensin homolog (PTEN) as being significantly differentially expressed in endometrial carcinoma subtypes. The tumors were stratified into 4 groups on the basis of their cell type and grade: EC grade 1 or 2, EC-3, SC, and CCC. Univariate survival analysis revealed significant differences in outcome between the 4 groups (P<0.0001), with significantly longer disease-specific survival for grade 1 or 2 EC versus EC-3 (P=0.0001), and EC-3 versus SC (P=0.0003). p16, PTEN, and IMP3 expression was observed more frequently in SC compared with EC-3 (P<0.0001, P=0.021, and P=0.031, respectively). These 3 markers showed the highest sensitivity and specificity in distinguishing between EC-3 and SC, with receiver operating characteristics area under the curve of 0.85, 0.69, and 0.71, respectively. ER and p53 approached but did not reach significance for differential expression in EC-3 versus SC (P=0.055 and P=0.068, respectively). A combination of p16 and PTEN predicts EC-3 versus SC with a sensitivity of 90.0% and specificity of 96.8%. p16 and PTEN can aid in distinguishing between EC-3 and SC of the endometrium, and are superior to ER, PR, and p53 for this purpose. EC-3 carcinomas have a significantly better prognosis than SC carcinomas of the endometrium. © 2010 International Society of Gynecological Pathologists.
Ota T.,Mayo Medical School |
Clayton A.C.,Mayo Medical School |
Minot D.M.,Mayo Medical School |
Shridhar V.,Mayo Medical School |
And 3 more authors.
Modern Pathology | Year: 2011
Minichromosome maintenance protein 7 (MCM7) is involved in replicative licensing and synthesis of DNA. It was previously identified as an overexpressed gene in high-grade serous carcinomas compared with serous borderline tumors of the ovary in cDNA microarray studies. In this study, we sought to validate MCM7 expression in 342 ovarian tumors on tissue microarrays. MCM7 expression was quantified as the MCM7 labeling index, and it was independently generated by two methods: a score provided by manual review of each sample by a pathologist observer and by an automated cellular imaging system. Analyses of MCM7 scores indicated a high degree of concordance and distribution between the observer- and machine-generated MCM7 labeling indexes. MCM7 expression was significantly higher in high-grade serous carcinomas than in serous borderline tumors or other histological subtypes of ovarian cancer. For both observer- and machine-derived scores, univariate analyses indicated the significant association of a high MCM7 labeling index with better progression-free survival in high-grade serous carcinomas. These results suggest the clinical importance of MCM7 expression in high-grade serous carcinomas of the ovary and the need for further evaluation of MCM7 as a potential prognostic factor in ovarian cancer. © 2011 USCAP, Inc.
Yerushalmi R.,BC Cancer Agency |
Gelmon K.A.,BC Cancer Agency |
Leung S.,Genetic Pathology Evaluation Center |
Gao D.,Genetic Pathology Evaluation Center |
And 4 more authors.
Breast Cancer Research and Treatment | Year: 2012
Insulin-like growth factor-1 receptor (IGF-1R) is expressed in normal and malignant breast tissue and has been implicated in cell survival and resistance to cytotoxic therapies. We sought to assess the prognostic impact of IGF-1R expression among patients with early breast cancer and among breast cancer subtypes. Patients with stages I-III breast cancer with archival tumor tissue were included. Paraffin tissue blocks were used to construct a tissue microarray that was stained for ER, PR, Ki-67, HER2, EGFR, and cytokeratins 5/6 to classify the breast subgroups and for expression of IGF-1R, p27, and Bcl2 by immunohistochemistry. Kaplan-Meier plots were created by subtypes. Associations between IGF-1R and prognostic variables were examined in multivariate analysis. Among 2,871 eligible women the prognostic cut point for IGF-1R expression for breast-cancer-specific survival (BCSS) was Allred score <7 versus ≥7. IGF-1R was ≥7 in 52% (LuminalA), 57.5% (LuminalB), 44.8% (LuminalHER2), 9.7% HER2-enriched, and 22.5% (Basal-like), P = 1.3 × 10 -52. IGF-1R+ was associated with age ≥50, lower histopathology grade, ER+, HER2 negativity (-), high p27 and high Bcl2 score. IGF-1R ≥7 was associated with better BCSS among LuminalB patients, hazard ratio = 0.64 (0.49-0.84); P = 1.2 × 10 -3, and worse outcome in the HER2-enriched subtype, hazard ratio = 2.37 (1.21-4.64); P = 0.012. IGF-1R correlates with good prognostic markers among patients with early breast cancer and is differentially expressed with variable prognostic impact among breast cancer subtypes. Results may have relevance to the development of therapeutics targeting IGF-1R. © 2011 Springer Science+Business Media, LLC.
Al-Agha O.M.,Genetic Pathology Evaluation Center |
Al-Agha O.M.,Applied Genomics |
Huwait H.F.,Vancouver General Hospital |
Chow C.,Genetic Pathology Evaluation Center |
And 9 more authors.
American Journal of Surgical Pathology | Year: 2011
Sex cord-stromal tumors (SCSTs) of the ovary are relatively uncommon tumors. Diagnosis of SCST rests primarily on the histomorphology of these tumors, and tumors with an atypical or unconventional appearance can pose diagnostic challenges. Previously, we had identified FOXL2 (402C→G) mutation as being characteristic of adult granulosa cell tumors (aGCTs). However, molecular screening for this mutation is not always possible and adds time and cost to the diagnostic process. In this study, we investigated the potential diagnostic use of immunostaining for FOXL2 on formalin-fixed paraffin-embedded tissue sections. Using a commercially available polyclonal antiserum against FOXL2 protein, immunoexpression of FOXL2 was tested in 501 ovarian tumor samples, including 119 SCSTs, using whole tissue sections and tissue microarrays. Staining was correlated with FOXL2 mutation status. In addition, we compared FOXL2 immunoexpression with that of α-inhibin and calretinin, the 2 traditional immunomarkers of SCST, in a subset of 89 SCSTs. FOXL2 immunostaining was present in 95 of 119 (80%) SCSTs, including >95% of aGCTs, juvenile granulosa cell tumors, fibromas, and sclerosing stromal tumors. Only 50% of Sertoli-Leydig cell tumors (N=40) expressed FOXL2. One of 11 steroid cell tumors and 3 of 3 female adnexal tumors of probable Wolffian origin showed FOXL2 immunoreactivity, whereas all other non-SCSTs tested (N=368) were negative for FOXL2 expression. Thus, the sensitivity and specificity of FOXL2 immunoreactivity for SCST are 80% and 99%, respectively. The FOXL2 (402C→G) mutation was confirmed to be both a sensitive and relatively specific indicator of aGCT. Forty-five of 119 SCSTs were mutation positive. These cases were 39 of 42 (93%) aGCTs, 3 of 40 Sertoli-Leydig cell tumors, 2 of 5 thecomas, and 1 of 4 (25%) SCSTs of unclassified type. SCSTs harboring a FOXL2 mutation consistently immunoexpressed FOXL2 (44 of 45, 98%), but FOXL2 immunostaining was also seen in many SCSTs that lacked a mutation (49 of 73, 67%). FOXL2 immunostaining showed higher sensitivity for the diagnosis of SCST, compared with α-inhibin and calretinin, and FOXL2 staining was typically more intense in positive cases compared with either α-inhibin or calretinin. In the SCSTs that were negative for FOXL2 expression, α-inhibin and/or calretinin immunostaining yielded positive results. In conclusion, FOXL2 is a relatively sensitive and highly specific marker for SCST. FOXL2 staining is present in almost all SCSTs with a FOXL2 mutation, and also in a majority of SCSTs without a mutation. FOXL2, together with α-inhibin and calretinin, forms an immunomarker panel that will result in positive staining with 1 or more markers in essentially all cases of SCST. © 2011 Lippincott Williams & Wilkins.
Ali A.M.G.,University of Cambridge |
Dawson S.-J.,University of Cambridge |
Dawson S.-J.,Cancer Research UK Research Institute |
Blows F.M.,University of Cambridge |
And 8 more authors.
British Journal of Cancer | Year: 2011
Background:Tissue micro-arrays (TMAs) are increasingly used to generate data of the molecular phenotype of tumours in clinical epidemiology studies, such as studies of disease prognosis. However, TMA data are particularly prone to missingness. A variety of methods to deal with missing data are available. However, the validity of the various approaches is dependent on the structure of the missing data and there are few empirical studies dealing with missing data from molecular pathology. The purpose of this study was to investigate the results of four commonly used approaches to handling missing data from a large, multi-centre study of the molecular pathological determinants of prognosis in breast cancer.Patients and Methods:We pooled data from over 11 000 cases of invasive breast cancer from five studies that collected information on seven prognostic indicators together with survival time data. We compared the results of a multi-variate Cox regression using four approaches to handling missing data-complete case analysis (CCA), mean substitution (MS) and multiple imputation without inclusion of the outcome (MI) and multiple imputation with inclusion of the outcome (MI). We also performed an analysis in which missing data were simulated under different assumptions and the results of the four methods were compared.Results:Over half the cases had missing data on at least one of the seven variables and 11 percent had missing data on 4 or more. The multi-variate hazard ratio estimates based on multiple imputation models were very similar to those derived after using MS, with similar standard errors. Hazard ratio estimates based on the CCA were only slightly different, but the estimates were less precise as the standard errors were large. However, in data simulated to be missing completely at random (MCAR) or missing at random (MAR), estimates for MI were least biased and most accurate, whereas estimates for CCA were most biased and least accurate.Conclusion:In this study, empirical results from analyses using CCA, MS, MI and MI were similar, although results from CCA were less precise. The results from simulations suggest that in general MI is likely to be the best. Given the ease of implementing MI in standard statistical software, the results of MI and CCA should be compared in any multi-variate analysis where missing data are a problem. © 2011 Cancer Research UK. All rights reserved.