Anatomic Abstracts
| Anatomic Abstracts |
|
|
|
November 2012 Editors:
Classifying papillary lesions of the breast on core biopsy is challenging. These lesions traditionally were surgically excised, but today, conservative management of benign lesions is being advocated. Therefore, accurately classifying papillary lesions on core biopsy is all the more imperative. The extent to which subspecialty training in breast pathology might mitigate such difficulties in diagnosis has not yet been reported. The authors investigated the change in diagnoses from core biopsy to surgical excision according to subspecialist training in breast pathology and according to interobserver agreement between specialized breast pathologists and non-breast pathologists in classifying these lesions. Core biopsies of 281 papillary lesions from 266 patients diagnosed between 2000 and 2010 were classified by a breast pathologist and non-breast pathologist into the categories of benign, atypical, ductal carcinoma in situ/encapsulated papillary carcinoma, or invasive carcinoma. Rates of change in diagnostic category in the surgical excision specimen were calculated on the basis of the original diagnosis, diagnosis made by the breast pathologist, and diagnosis made by the non-breast pathologist. Comparisons were made using the X test. Kappa values were calculated for interobserver agreement. The authors found that of 162 lesions with subsequent excision, 90 were originally diagnosed as benign, 38 as atypical, 25 as ductal carcinoma in situ/encapsulated papillary carcinoma, and nine as invasive on core biopsy. The rate of upgrade from benign papilloma to an atypical or malignant lesion on surgical excision was 22.2 percent according to the original diagnosis. This rate fell to 16.3 percent when the breast pathologists’ diagnoses were considered, compared with 26.3 percent for the non-breast pathologists’ diagnoses. There was no significant difference between the breast and non-breast pathologists in the rate of upgrade from a benign to an atypical/malignant diagnosis, although downgrades from atypical/malignant to benign papillomas were most commonly seen among non-breast pathologists (P=.002). Overall, the breast pathologists’ diagnoses on core biopsy were less likely to differ from the excision diagnosis (P=.0001). Benign papillomas upgraded on excision were more likely to occur with larger radiologic mass size (P=.033) compared with those that were not upgraded. Of eight benign papillomas upgraded to a malignant lesion on excision, seven were discordant on radiology. Interobserver agreement between the breast and non-breast pathologists’ diagnoses was in the “fair agreement” range (K=0.38), with perfect agreement in 66.4 percent of cases. The authors concluded that correlation between core biopsy and excision diagnoses for breast papillary lesions is significantly greater for breast pathologists than for non-breast pathologists. This is largely because of non-breast pathologists’ tendency to overcall atypia or malignancy on core biopsy. However, upgrades from benign to atypical or malignant did not differ significantly based on subspecialization. With accurate pathologic assessment and radiologic-pathologic correlation, the rate of upgrade from benign papilloma to malignancy can be minimized significantly. Jakate K, De Brot M, Goldberg F, et al. Papillary lesions of the breast: impact of breast pathology subspecialization on core biopsy and excision diagnoses. Am J Surg Pathol. 2012;36(4):544–551. Correspondence: Anna Marie Mulligan at mulligana@smh.ca
The authors conducted a study to evaluate the impact of low estrogen/progesterone receptor expression and the effect of endocrine therapy on survival outcomes in human epidermal growth factor receptor 2 (HER2)-negative tumors with ER/PR of less than 10 percent that were previously labeled as triple negative. In a retrospective review, 1,257 patients were categorized into three groups based on ER/PR percentages: ER/PR of less than one percent (group A), ER/PR of one percent to five percent (group B), and ER/PR of six percent to 10 percent (group C). The Kaplan-Meier product limit method was used to estimate survival outcomes. Cox proportional hazards models were used to adjust for patient and tumor characteristics. Groups A, B, and C had 897 (71.4 percent), 241 (19.2 percent), and 119 (9.4 percent) patients, respectively. After a median followup of 40 months, there was no significant difference in three-year recurrence-free survival (64 percent, 67 percent, and 77 percent; P=.34) or overall survival (79 percent, 81 percent, and 88 percent; P=.33) for groups A, B, and C, respectively. ER/PR expression was not an independent predictor for recurrence-free survival (hazard ratio [HR], 1.10; 95 percent confidence interval [CI], 0.86–1.39; P=.46 for group B, and HR, 0.96; 95 percent CI, 0.66–1.38; P=.81 for group C, compared with group A) or overall survival (HR, 1.11; 95 percent CI, 0.84–1.46; P=.46 for group B, and HR, 0.94; 95 percent CI, 0.63–1.42; P=.78 for group C, compared with group A). Endocrine therapy had no impact on survival outcomes (recurrence-free survival: P=.10; overall survival: P=.45) among the groups. The authors concluded that in this cohort, a low ER/PR level of one percent to five percent does not appear to have any significant impact on survival outcomes. There was a tendency toward survival advantages in the ER/PR level of six percent to 10 percent. The benefit of endocrine therapy in these patients is unclear. Raghav KP, Hernandez-Aya LF, Lei X, et al. Impact of low estrogen/progesterone receptor expression on survival outcomes in breast cancers previously classified as triple negative breast cancers. Cancer. 2012;118:1498–1506. Correspondence: Dr. Ana M. Gonzalez-Angulo at agonzalez@mdanderson.org
A test to predict distant metastases would be valuable for prognostication in colon cancer. In previous studies, colon cancer with microsatellite instability (MSI) showed a reduced risk of distant metastases. High expression of CD133 and β-catenin, both related to cancer stem cell phenotypes, might predict metastasis. The authors conducted a study with the aim of developing a simple and robust test to assess the risk of distant metastases in colon cancer. In a case-control study, the authors matched 57 right-sided colon cancer specimens with synchronous distant metastases to 57 colon cancer specimens without distant metastases. They used immunohistochemistry to assess MLH1, CD133, and nuclear β-catenin. To define the diagnostic algorithm, the tumors were stratified according to MLH1 expression. Loss of MLH1 expression correlated significantly with very low risk of distant metastases (5.3 percent; P=.00003). In MLH1-positive cases, combined high scores of CD133 and β-catenin were associated with a very high rate of distant metastases (94.4 percent), whereas the risk was intermediate for carcinomas with either low CD133 or low β-catenin expression (P=.0007), or both. A validation study using an independent set of 68 right-sided colon cancer specimens showed a clear trend toward risk stratification according to the algorithm. However, sample sizes were small and associations were not statistically significant. The authors concluded that by using three markers, this algorithm allowed identification of subgroups of right-sided colon cancer patients with extremely high or extremely low risk of distant metastases. Neumann J, Horst D, Kriegl L, et al. A simple immunohistochemical algorithm predicts the risk of distant metastases in right-sided colon cancer. Histopathology. 2012;60(3):416–426. Correspondence: J. Neumann at jens.neumann@med.uni-muenchen.de
The authors undertook a study with the hypothesis that certain common morphologic features of ovarian carcinomas are predictably associated with BRCA1 and BRCA2 deficiencies. They selected 43 high-grade serous carcinomas diagnosed at Memorial Sloan-Kettering Cancer Center that were studied as part of The Cancer Genome Atlas pilot project. The study included 12 randomly selected nonfamilial BRCA-unassociated cases and 31 Memorial Sloan-Kettering Cancer Center cases with BRCA1 or BRCA2 abnormality (n=43). Slides were examined to assess tumor architecture, mitotic index, tumor-infiltrating lymphocytes, nuclear pleomorphism, necrosis, and involvement of fallopian tube epithelium. Comparing BRCA1-associated cases (BRCA1 germline mutation, n=4; BRCA1 somatic mutation, n=6; BRCA1 promoter methylation, n=13) with unassociated cases (n=12) identified statistically significant differences in morphology. BRCA1-associated high-grade serous carcinomas had more frequent solid, pseudoendometrioid, and transitional cell carcinoma-like morphology (SET features) (P=.0045), higher mitotic indeces (P=.012), more tumor-infiltrating lymphocytes (P=.034), as well as geographic or comedo necrosis (P=.034). BRCA2-associated cases (germline mutation, n=4; somatic mutation, n=4) tended to show SET features, but they were relatively deficient in tumor-infiltrating lymphocytes and necrosis. Two algorithms incorporating tumor architecture, necrosis, and either mitotic indeces or tumor-infiltrating lymphocytes separated cases that showed two of three features (BRCA1 associated) from those with zero of three features (BRCA unassociated; P=.0016 and P=.0033). A test set comprising nine BRCA1 germline mutations and 14 high-grade serous carcinoma controls lacking BRCA1 and BRCA2 germline mutation was used to validate the algorithms, with specific emphasis on the ability to detect cases with BRCA1 germline mutation. The best results were obtained with the algorithm that incorporated SET features, necrosis, and mitotic index (P=.0072; sensitivity of 1.0 [95 percent confidence interval [CI], 0.66–1.0]; specificity of 0.57 [95 percent CI, 0.29–0.82]; positive predictive value of 0.60 [95 percent CI, 0.32–0.84]; and negative predictive value of 1.0 [95 percent CI, 0.63–1.0]). These preliminary data indicate potential strong associations between morphology and genotype in high-grade serous carcinomas. Soslow RA, Han G, Park KJ, et al. Morphologic patterns associated with BRCA1 and BRCA2 genotype in ovarian carcinoma. Mod Pathol. 2012;25:625–636. Correspondence: Dr. R. A. Soslow at gynbreast@mskcc.org
Eosinophils are often present in the inflammatory infiltrate of interface dermatitis, but the diagnostic specificity of eosinophils in interface dermatitis has not been formally evaluated. The authors retrospectively identified 97 examples of interface dermatitis with clinically confirmed diagnoses, including lupus erythematosus, lichen planus, pityriasis lichenoides, graft-versus-host disease, dermatomyositis, and drug reaction. At least two dermatologists clinically confirmed the diagnoses, and at least two dermatopathologists reviewed slides in a blinded fashion. The average eosinophil count per 10 × 200 (×20 objective) fields was lowest for pityriasis lichenoides (0.2), dermatomyositis (0.3), graft-versus-host disease (0.4), and lupus erythematosus (0.5) (defined as group one) and was higher for lichen planus, drug reactions, erythema multiforme (major and minor), and viral exanthems (defined as group two). Distinction between groups one and two was maximized using an eosinophil count cutoff of 1.1. The authors concluded that eosinophils are usually rare to absent in pityriasis lichenoids, dermatomyositis, most forms of lupus erythematosus, and graft-versus-host disease. While final interpretation requires a composite assessment of all features, the results suggest that the presence of even a single eosinophil within nine or 10 ×20 fields argues against a diagnosis of pityriasis lichenoides, dermatomyositis, or lupus erythematosus. Sharon VR, Konia TH, Barr KL, et al. Assessment of the ‘no eosinophils’ rule: are eosinophils truly absent in pityriasis lichenoides, connective tissue disease, and graft-versus-host disease? J Cutan Pathol. 2012;39:413–418. Correspondence: Maxwell A. Fung at maxwell.fung@ucdmc.ucdavis.edu
Immunohistochemistry has recently emerged as a powerful ancillary tool for differentiating lung adenocarcinoma from squamous cell carcinoma—a distinction with important therapeutic implications. Although the most frequently recommended squamous marker, p63, is extremely sensitive, it suffers from low specificity due to its reactivity in a substantial proportion of lung adenocarcinomas and other tumor types, particularly lymphomas. P40 is a relatively unknown antibody that recognizes ΔNp63, a p63 isoform suggested to be highly specific for squamous/basal cells. The authors compared the standard p63 antibody (4A4) and p40 in a series of 470 tumors from the archives of Memorial Sloan-Kettering Cancer Center and Johns Hopkins Hospital. The tumors included lung squamous cell carcinomas (n=81), adenocarcinomas (n=237), and large cell lymphomas (n=152). P63 was positive in 100 percent of squamous cell carcinomas, 31 percent of adenocarcinomas, and 54 percent of large cell lymphomas (sensitivity, 100 percent; specificity, 60 percent). In contrast, although p40 was also positive in 100 percent of squamous cell carcinomas, only three percent of adenocarcinomas and no large cell lymphomas had p40 labeling (sensitivity, 100 percent; specificity, 98 percent). The mean percentage of p63- versus p40-immunoreactive cells in squamous cell carcinomas was equivalent (97 percent versus 96 percent, respectively; P=.73). Rare adenocarcinomas with p40 labeling had reactivity in no more than five percent of tumor cells, whereas the mean (range) of p63-positive cells in adenocarcinomas and lymphomas was 26 percent (one percent to 90 percent) and 48 percent (two percent to 100 percent), respectively. In summary, p40 is equivalent to p63 in sensitivity for squamous cell carcinoma, but it is markedly superior to p63 in specificity, which eliminates a potential pitfall of misinterpreting a p63-positive adenocarcinoma or unsuspected lymphoma as squamous cell carcinoma. These findings strongly support the routine use of p40 in lieu of p63 for diagnosing pulmonary squamous cell carcinoma. Bishop JA, Teruya-Feldstein J, Westra WH, et al. p40 (ΔNp63) is superior to p63 for the diagnosis of pulmonary squamous cell carcinoma. Mod Pathol. 2012;25:405–415. Correspondence: Dr. N. Rekhtman at rekhtman@mskcc.org
A minimum of 12 dissected lymph nodes has been recommended as a consensus guideline for resection in colon cancer patients. The authors conducted a study to assess the influence of socioeconomic status and hospital type on compliance with this colon lymph node dissection guideline and to examine the time trend for 12 or more lymph nodes dissected. The authors obtained from the Louisiana Tumor Registry stage I to III incident colon cancer cases diagnosed from 1996 to 2007. A composite census tract-level socioeconomic status score was created to serve as a surrogate for individual-level socioeconomic status. Hospitals performing colon resections were categorized into five groups according to the Commission on Cancer Accreditation Program. Multiple logistic regression analyses were used. The authors found that of 10,460 colon cancer cases diagnosed during the study period, 43.9 percent had 12 or more lymph nodes dissected. Patients residing in less affluent areas were less likely to receive a dissection of 12 or more nodes than those residing in more affluent areas. Socioeconomic status was no longer significant after adjusting for race, gender, age, stage, grade, anatomic subsite, diagnosis year, and hospital type. In contrast, hospital type was significantly associated with the number of lymph nodes dissected, even after adjusting for other factors. Patients diagnosed from 2002 to 2007 were twice as likely (95 percent confidence interval, 1.84–2.17) to have 12 or more lymph nodes dissected than those diagnosed from 1996 to 2001 after adjustment. The authors concluded that in Louisiana, hospital type is a significant independent predictor of adequate lymph node evaluation for colon cancer. Training and education are needed to reduce this disparity in facilities with consistently lower lymph node yield in their dissections. Hsieh MC, Velasco C, Wu XC, et al. Influence of socioeconomic status and hospital type on disparities of lymph node evaluation in colon cancer patients. Cancer. 2012;118:1675–1683. Correspondence: Mei-Chin Hsieh at mhsieh@lsuhsc.edu Anatomic pathology abstracts editors: Michael Cibull, MD, professor and vice chair, Department of Pathology and Laboratory Medicine, University of Kentucky College of Medicine, Lexington; Rouzan Karabakhtsian, MD, assistant professor of pathology and laboratory medicine, University of Kentucky College of Medicine; and Thomas Cibull, MD, dermatopathologist, Evanston Hospital, NorthShore University HealthSystem, Evanston, Ill. |
