Immunohistochemical staining for p53 is used as a surrogate for mutational analysis in the diagnostic workup of carcinomas of multiple sites including ovarian cancers. Strong and diffuse immunoexpression of p53 is generally interpreted as likely indicating a TP53 gene mutation. The immunoprofile that correlates with wild-type TP53, however, is not as clear. In particular, the significance of completely negative immunostaining is controversial. The aim of this study was to clarify the relationship of the immunohistochemical expression of p53 with the mutational status of the TP53 gene in ovarian cancer. A total of 57 ovarian carcinomas (43 high-grade serous ovarian/peritoneal carcinomas, 2 malignant mesodermal mixed tumors (carcinosarcomas), 2 low-grade serous carcinomas, 4 clear cell carcinomas, 1 well-differentiated endometrioid carcinoma, and 5 carcinomas with mixed epithelial differentiation) were analyzed for TP53 mutations by nucleotide sequencing (exons 4-9), and subjected to immunohistochemical analysis of p53 expression. Thirty six tumors contained functional mutations and 13 had wild type TP53. Five tumors were found to harbor known TP53 polymorphism and changes in the intron region were detected in three. Tumors with wild-type TP53 displayed a wide range of immunolabeling patterns, with the most common pattern showing r10% of positive cells in 6 cases (46%). Mutant TP53 was associated with 60-100% positive cells in 23 cases (64% of cases). This pattern of staining was also seen in three cases with wildtype TP53. Tumors that were completely negative (0% cells staining) had a mutation of TP53 in 65% of cases and wild-type TP53 in 11%. Combining two immunohistochemical labeling patterns associated with TP53 mutations (0% and 60-100% positive cells), correctly identified a mutation in 94% of cases (Po0.001). Immunohistochemical analysis can be used as a robust method for inferring the presence of a TP53 mutation in ovarian carcinomas. In addition to a strong and diffuse pattern of p53 expression (in greater than 60% of cells), complete absence of p53 immunoexpression is commonly associated with a TP53 mutation. Accordingly, this latter pattern, unlike low-level expression (10-50% cells), should not be construed as indicative of wild-type TP53.

Summary It has been reported that the diagnosis of serous tubal intraepithelial carcinoma (STIC) is not optimally reproducible based only on histologic assessment. Recently, we reported that the use of a diagnostic algorithm that combines histologic features and coordinate immunohistochemical expression of p53 and Ki-67 substantially improves reproducibility of the diagnosis. The goal of the current study was to validate this algorithm by testing a group of 6 gynecologic pathologists who had not participated in the development of the algorithm (3 faculty, 3 fellows) but who were trained in its use by referring to a website designed for that purpose. They then reviewed a set of microscopic slides, which contained 41 mucosal lesions of the fallopian tube. Overall consensus (≥4 of 6 pathologists) for the 4 categories of STIC, serous tubal intraepithelial lesion (our atypical intermediate category), p53 signature, and normal/reactive was achieved in 76% of lesions with no consensus in 24%. Combining diagnoses into 2 categories (STIC vs. non-STIC) resulted in overall consensus in 93% with no consensus in 7%. The kappa value for STIC vs. non-STIC among all 6 observers was also high at 0.67 and did not significantly differ whether for faculty (κ=0.66) or fellows (κ=0.60). These findings confirm the reproducibility of this algorithm by a group of gynecologic pathologists who were trained on a website for that purpose. Accordingly, we recommend its use in research studies. Before applying it in routine clinical practice, the algorithm should be evaluated by general surgical pathologists in the community setting.

Uterine serous carcinomas typically have a characteristic morphology (papillary architecture, high-grade nuclei) and immunoprofile (diffuse/strong p53 expression, loss of hormone receptor expression) that distinguish them from most endometrial endometrioid carcinomas. However, glandular variants of serous carcinoma can simulate Fédération Internationale de Gynécologie et d'Obstétrique (FIGO) grade 2 endometrioid carcinomas, and some serous carcinomas lack p53 expression and retain hormone receptor expression, making classification difficult. P16 expression patterns distinguish endometrioid carcinomas (patchy) from human papillomavirus (HPV)-related endocervical adenocarcinomas (diffuse/strong) but utility for distinction of serous carcinomas from endometrioid carcinomas and endocervical adenocarcinomas has not been evaluated in a large series. Immunohistochemical analysis of p16 expression was performed on 201 uterine and endocervical adenocarcinomas in hysterectomy specimens, including 49 serous carcinomas, 101 endometrial endometrioid carcinomas (44 FIGO grade 1, 40 FIGO grade 2, and 17 FIGO grade 3), and 51 HPV-related endocervical adenocarcinomas. All serous carcinomas demonstrated diffuse/moderate-strong p16 expression, with percentage of positive tumor cells ranging from 90% to 100% (mean/median: 95%/100%). In contrast, endometrial endometrioid carcinomas exhibited less diffuse and less intense expression, with percent of positive tumor cells ranging from 10% to 90% (mean/median: 38%/30%; staining intensity: variable). Similar to serous carcinomas, all endocervical adenocarcinomas exhibited diffuse/moderate-strong p16 expression, with percentage of positive tumor cells ranging from 90% to 100% (mean/median: 94%/90%). P16 can serve as an additional diagnostic marker, used as part of an immunohistochemical panel, including p53 and hormone receptors, for distinction of uterine serous carcinomas from endometrioid carcinomas. Distinction of serous carcinomas from endocervical adenocarcinomas (HPV-related type), both of which share diffuse p16 expression and frequently lack hormone receptor expression, relies on morphology and diffuse/strong p53 expression in the former and detection of HPV in the latter.

Distinction of primary ovarian mucinous tumors from metastatic/secondary mucinous tumors involving the ovaries is often challenging, not only at the time of intraoperative assessment when requested for surgical management (staging decisions) but also for final pathologic diagnosis. Previous studies have shown that a simple algorithm using tumor size and laterality (bilateral tumors of any size, or unilateral tumor <10 cm=metastatic; unilateral tumor > or =10 cm=primary) can accurately classify a substantial majority of tumors. To assess the general utility of this algorithm for distinction of primary and secondary mucinous tumors in the ovary and address the occurrence of exceptions (large unilateral metastases), analysis of tumor size and laterality data was performed using 194 tumors (52 primary tumors and 142 metastases), with metastases subclassified by primary site [colorectum (46), appendix (28 low-grade tumors, 20 carcinomas), pancreaticobiliary tract (20), small intestine (3), stomach (5), and endocervix (20)]. Performance of the algorithm was evaluated using the originally proposed method and modified size criteria were analyzed to optimize tumor classification. The original algorithm correctly classified 84% of tumors overall, including 100% of primary ovarian tumors and 77% of all metastases (colorectal: 74%; appendiceal: 79% of low-grade tumors, 100% of carcinomas; pancreaticobiliary: 95%; small intestinal: 33%; gastric: 80%; endocervical: 55%). By adjusting the size criterion to 12 cm, performance of the algorithm was both maintained for primary ovarian tumors and improved for metastases, with correct classification of 86% of tumors overall, including 100% of primary tumors and 80% of metastases. Performance was optimized at 13 cm, with correct classification of 87% of tumors overall, including 98% of primary tumors and 82% of metastases (colorectal: 80%; appendiceal: 79% of low-grade tumors, 100% of carcinomas; pancreaticobiliary: 100%; small intestinal: 33%; gastric: 100%; endocervical: 70%). Of the more common metastases, metastatic colorectal and endocervical carcinomas provided the greatest number of exceptions, even when analyzed with the optimized size criterion. Recognition that metastatic colorectal carcinomas represent the most common metastases and have a greater tendency to violate the algorithm should prompt lowering of the threshold for suggesting the possibility of metastatic colorectal carcinoma for tumors displaying any microscopic features suggestive of that diagnosis, even when a history of primary colorectal carcinoma is lacking. Use of the algorithm is intended as an adjunct to the complete clinicopathologic evaluation that ideally should occur when problematic mucinous tumors in the ovary are encountered.

Coordinate expression profiles for cytokeratins 7 and 20 (CK7 and CK20) are useful for distinguishing certain types of adenocarcinomas but use for distinction of primary and secondary mucinous tumors in the ovary is limited due to the existence of a number of tumor types exhibiting overlapping CK7/CK20 immunoprofiles; the use of staining distribution patterns in the distinction of tumors with shared profiles has not been evaluated in detail. We report analysis of both coordinate expression profiles and staining distribution in 179 rigorously classified mucinous tumors in the ovary, including 53 primary tumors [35 atypical proliferative (borderline) mucinous tumors of gastrointestinal type and 18 invasive mucinous carcinomas] and 126 secondary tumors [28 colorectal adenocarcinomas, 54 appendiceal tumors (23 adenocarcinomas, 31 low-grade adenomatous mucinous tumors associated with pseudomyxoma peritonei), 14 pancreatic adenocarcinomas, 8 endocervical adenocarcinomas, 5 gastric adenocarcinomas, 4 gallbladder/biliary tract adenocarcinomas, and 13 adenocarcinomas of unknown primary sites). A CK7+/CK20+ immunoprofile was the most common profile in primary ovarian tumors (74%), upper gastrointestinal tract tumors (78%), and endocervical tumors (88%) but was occasionally observed in lower intestinal tract tumors (colorectal: 11%; appendiceal: 13% of low-grade tumors, 35% of carcinomas). A CK7-/CK20+ immunoprofile was the most common profile in lower intestinal tract tumors (79%) and was uncommon in upper gastrointestinal tract tumors (9%), rarely seen in primary ovarian tumors (4%), and not seen in endocervical tumors. A CK7+/CK20- profile was observed in some primary ovarian (23%), upper gastrointestinal tract (13%), and endocervical tumors (13%) but not in lower intestinal tract tumors. For CK7+ tumors, staining distribution was very frequently diffuse (>50% of tumors cells positive) in primary ovarian, upper gastrointestinal tract, and endocervical tumors, whereas staining distribution was often focal (<50% of tumors cells positive) when present in colorectal and appendiceal carcinomas but not in low-grade appendiceal tumors. For CK20+ tumors, staining distribution was variable but often focal in primary ovarian tumors and nonlower intestinal tract tumors, whereas the pattern was almost always diffuse in lower intestinal tract tumors. Immunohistochemical staining distribution can supplement CK7/CK20 coordinate expression profiles to distinguish subsets of primary ovarian and metastatic lower intestinal tract mucinous tumors having overlapping immunoprofiles but neither coordinate expression profiles nor staining distribution distinguishes primary ovarian tumors from the nonlower intestinal tract metastases.

No abstract

Recent studies have demonstrated conflicting results regarding the value of CDX2 for distinguishing primary ovarian mucinous tumors from metastatic mucinous carcinomas in the ovary. Utility of coordinate expression of cytokeratins 7 and 20 is restricted to distinction of ovarian mucinous tumors from lower gastrointestinal tract metastases and data comparing coordinate expression of all three markers is limited. Immunohistochemical studies were performed to compare expression of CDX2 and cytokeratin 20, both markers of intestinal differentiation, in conjunction with coordinate expression of cytokeratin 7, in 90 mucinous tumors involving the ovary: 42 primary ovarian mucinous tumors (31 atypical proliferative (borderline) mucinous tumors (gastrointestinal type), 11 mucinous carcinomas) and 48 metastatic mucinous carcinomas of upper (pancreaticobiliary tract: 14; stomach: five) and lower (colon and rectum: 25; appendix: four) gastrointestinal tract origin. Primary ovarian tumors expressed CDX2 (40%) less frequently than cytokeratin 20 (83%) (Po0.0001). CDX2 expression in primary ovarian tumors (40%) was lower than CDX2 expression in metastatic carcinomas of both upper (74%; P ¼ 0.016) and lower gastrointestinal tract origin (90%; Po0.0001). Cytokeratin 20 expression was similar in primary ovarian tumors (83%) and metastases of upper (89%; P ¼ 0.071) and lower gastrointestinal tract origin (93%; P ¼ 0.29). Thus, as a single marker CDX2 offers some advantage over cytokeratin 20 because it is less frequently positive in primary ovarian tumors. In the almost universally cytokeratin 7-positive primary ovarian tumors and metastases of upper gastrointestinal tract origin, CDX2 coordinate expression was less common in primary ovarian tumors (36%) than in metastases of upper gastrointestinal tract origin (63%) (P ¼ 0.022) whereas cytokeratin 20 coordinate expression was identical in both tumor types (79%). In the almost universally cytokeratin 7-negative metastases of lower gastrointestinal tract origin, coordinate expression of CDX2 (83%) and cytokeratin 20 (86%) were equivalent (P ¼ 1.00). CDX2 was comparable to cytokeratin 20 in distinguishing metastases of lower gastrointestinal tract origin (usually cytokeratin 7-negative and CDX2/cytokeratin 20 positive) from primary ovarian tumors and metastases of upper gastrointestinal tract origin (usually cytokeratin 7-positive and CDX2/cytokeratin 20 variable). CDX2 provided some advantage over cytokeratin 20 for distinguishing primary ovarian mucinous tumors from metastases of upper but not lower gastrointestinal tract origin; however, the advantage in the former was limited due to the occurrence of shared coordinate expression profiles in both tumor types. Cytokeratin 7 provides the predominant discriminatory value among these markers yet is limited to distinction of primary ovarian tumors from metastases of lower gastrointestinal tract origin.

Most endocervical adenocarcinomas ( approximately 90%) are high-risk human papillomavirus (HPV)-related neoplasms, with the remainder being unrelated to HPV; both types infrequently metastasize to the ovaries. Clinicopathologic features of 29 cases of synchronous and metachronous endocervical and ovarian tumors (26 HPV-related, 3 unrelated to HPV) were analyzed. In 18 cases, the cervical tumors were clearly invasive; these included 5 clinically evident tumors diagnosed before the ovarian metastases (immediately preoperatively to 7 y), 11 clinically unsuspected tumors diagnosed concurrently in specimens obtained for evaluation of ovarian/pelvic masses, 1 case with concurrent clinically evident cervical and ovarian masses, and 1 clinically occult tumor diagnosed subsequent to the ovarian metastasis. In 11 cases, the cervical tumors were more limited; these included 5 tumors comprised predominantly of adenocarcinoma in situ with small foci of superficial invasion ("microinvasive carcinomas") diagnosed before the ovarian metastases (3 mo to 7 y) and 6 tumors comprised of extensive adenocarcinoma in situ lacking unequivocally recognizable stromal invasion diagnosed before (9 mo to 7 y, n=4), concurrently with (n=1), or subsequent to (n=1) the ovarian metastases. Fifteen cervical tumors involved lower uterine segment corpus endometrium or endomyometrium, including 4 tumors that were minimally invasive or not recognizably invasive in the cervix. The ovarian tumors ranged in size from 2.1 to 30.0 cm (mean/median=12.7/13.5); they were unilateral in 19 cases (65.5%) and 12 of these were unilateral and 10 cm or greater. In 26 cases, including the 19 unilateral tumors, the ovarian tumors exhibited "borderlinelike," confluent glandular, cribriform, and/or villoglandular patterns simulating primary ovarian atypical proliferative (borderline) tumors or well-differentiated carcinomas; these patterns were pure in 24 and admixed with minor infiltrative foci in 2. The ovarian tumors had features typical of metastases (bilateral and infiltrative) in only 3 cases. In all HPV-related cases the paired endocervical and ovarian tumors contained identical HPV types, establishing the ovarian tumors as metastases. Endocervical adenocarcinomas, including microinvasive forms and some not recognizably invasive, have the potential to metastasize to the ovaries; extension into the lower uterine segment/corpus endometrium may be a risk factor, with retrograde uterine/transtubal spread as a possible mechanism.