Search

This article provides practical recommendations developed from the International Society of Gynecological Pathologists Endometrial Carcinoma Project to address 4 issues that may arise in the diagnosis of uterine corpus low-grade endometrioid carcinoma: (1) The distinction between atypical hyperplasia and low-grade endometrioid carcinoma. (2) The distinction between low-grade endometrioid carcinoma and serous carcinoma. (3) The distinction between corded and hyalinized or spindle cell variants of low-grade endometrioid carcinoma and carcinosarcoma. (4) The diagnostic criteria for mixed endometrial carcinomas, a rare entity that should be diagnosed only after exclusion of a spectrum of tumors including morphologic variants of endometrioid carcinoma, dedifferentiated endometrial carcinoma, carcinosarcoma, and endometrial carcinomas with ambiguous morphology.

This review of challenging diagnostic issues concerning high-grade endometrial carcinomasisderivedfromtheauthors’ reviewoftheliteraturefollowedbydiscussionsatthe Endometrial Cancer Workshop sponsored by the International Society of Gynecological Pathologists in 2016. Recommendations presented are evidence-based, insofar as this is possible,giventhatthelevelsofevidenceareweakormoderateduetosmallsamplesizesand nonuniform diagnostic criteria used in many studies. High-grade endometrioid carcinomas include FIGO grade 3 endometrioid carcinomas, serous carcinomas, clear cell carcinomas, undifferentiated carcinomas, and carcinosarcomas. FIGO grade 3 endometrioid carcinoma is diagnosed when an endometrioid carcinoma exhibits >50% solid architecture (excluding squamousareas), orwhenan architecturallyFIGOgrade2endometrioid carcinomaexhibits marked cytologic atypia, provided that a glandular variant of serous carcinoma has been excluded. The most useful immunohistochemical studies to make the distinction between these 2 histotypes are p53, p16, DNA mismatch repair proteins, PTEN, and ARID1A. Endometrial clear cell carcinomas must display prototypical architectural and cytologic features for diagnosis. Immunohistochemical stains, including, Napsin A and p504s can be used as ancillary diagnostic tools; p53 expression is aberrant in a minority of clear cell carcinomas. Of note, clear cells are found in all types of high-grade endometrial carcinomas, leading to a tendency to overdiagnose clear cell carcinoma. Undifferentiated carcinoma (which when associated with a component of low-grade endometrioid carcinoma is termed “dedifferentiated carcinoma”) is composed of sheets of monotonous, typically dyscohesive cells, which can have a rhabdoid appearance; they often exhibit limited expression of cytokeratins and epithelial membrane antigen, are usually negative for PAX8 and hormone receptors, lack membranous e-cadherin and commonly demonstrate loss of expression of DNA mismatch repair proteins and SWI-SNF chromatin remodeling proteins. Carcinosarcomas must show unequivocal morphologic evidence of malignant epithelial and mesenchymal differentiation. This work was funded in part through the NIH/NCI Cancer Center Support Grant P30 CA008748.

Recent identification of somatic MED12 mutations in most uterine leiomyomas brings a new venue for the study of the tumorigenesis of leiomyomas. We are particularly interested in the correlation of MED12 and HMGA2 gene products in leiomyomas and leiomyosarcomas with and without MED12 mutations. To address these issues, in this study we examined MED12 mutations in a large cohort of usual type leiomyomas (178 cases) and uterine leiomyosarcomas (32 cases). We found that 74.7% (133/178) of leiomyomas had MED12 mutations, which was consistent with several independent studies. In contrast, only 9.7% (3/32) of leiomyosarcomas harbored MED12 mutations. Expression analysis by western blot and immunohistochemistry revealed that those leiomyomas with complex MED12 mutations had significantly lower protein products than the matched myometrium. Interestingly, most leiomyosarcomas without MED12 mutations also had very low levels of MED12 expression in comparison to the matched myometrium. These findings suggest a potential functional role of MED12 in both benign and malignant uterine smooth muscle tumors. When we further examined HMGA2 expression in all leiomyomas and leiomyosarcomas, we found that HMGA2 overexpression was exclusively present in those leiomyomas with no MED12 mutation, accounting for 10.1% (18/178) of total leiomyomas and 40% (18/45) of non-MED12 mutant leiomyomas. Twenty-five percent (8/32) of leiomyosarcomas had HMGA2 overexpression, and no MED12 mutations were found in HMGA2 positive leiomyosarcoma. These findings strongly suggest that MED12 mutations and HMGA2 overexpression are independent genetic events that occur in leiomyomas, and they may act differently in the tumorigenesis of uterine leiomyomas.

The distinction of complex atypical endometrial hyperplasia from endometrial adenocarcinoma is often problematic. Foci of back-to-back arrangement of glands or foci of cribriform arrangement of glands smaller than 2.1 mm in diameter are considered insufficient for the diagnosis of endometrial adenocarcinoma by some authors, and sufficient to be diagnosed as endometrial adenocarcinoma by other authors. We refer to these foci as endometrial adenocarcinoma in situ. In this study, we evaluated findings in subsequent hysterectomy in complex atypical endometrial hyperplasia patients with and without adenocarcinoma in situ. Follow-up findings, including the presence or absence of endometrial adenocarcinoma in the hysterectomy specimen, the grade of the carcinoma and the depth of myometrial invasion were analyzed. Of the total 87 patients with complex atypical endometrial hyperplasia, 33 patients had adenocarcinoma in situ and 54 lacked adenocarcinoma in situ. Of 33 patients 22 (66%) with adenocarcinoma in situ had endometrial adenocarcinoma on subsequent hysterectomy vs 13 of 54 (24%) patients without adenocarcinoma in situ (P=0.0001). Myoinvasive endometrial adenocarcinoma was present in 20 of 33 (61%) patients with adenocarcinoma in situ vs 8 of the 54 (15%) patients without adenocarcinoma in situ (P< or =0.0001). The depth of myometrial invasion in cases with myoinvasion was 24.5+19.4% in patients with adenocarcinoma in situ and 12.8+8.5% in patients without adenocarcinoma in situ (P=0.05). Among patients younger than age of 50, 5 of the 7 (71%) with adenocarcinoma in situ had myoinvasive carcinoma vs 2 of the 13 (15%) without adenocarcinoma in situ (P=0.02). The likelihood of finding endometrial adenocarcinoma in subsequent hysterectomy in patients with complex atypical endometrial hyperplasia is significantly increased if adenocarcinoma in situ is present in prior endometrial sampling. Endometrial adenocarcinomas in patients with adenocarcinoma in situ are far more frequently myoinvasive, and invade to a greater depth than endometrial adenocarcinomas seen in patients without adenocarcinoma in situ. Use of adenocarcinoma in situ terminology could lead to improved management of patients with complex atypical endometrial hyperplasia.

Intraoral lesions clinically suspicious for cancer/precancer should be biopsied and diagnosed histopathologically. We evaluated whether the frequency of oral cancer (OC) cases diagnosed in Puerto Rico (PR) is disproportionately high relative to the frequency of persons with histopathologic diagnoses that would have appeared clinically suspicious for OC/precancer at biopsy.All pathology reports for oral (ICD-O-3 C01-C06) soft tissue biopsies generated during 1/2004-5/2005 by seven PR and two New York City (NYC) pathology laboratories were reviewed. The analysis was restricted to persons diagnosed with invasive oral squamous cell carcinoma (OSCC), epithelial dysplasia, or hyperkeratosis/epithelial hyperplasia (HK/EH), i.e., diagnoses associated with lesions clinically suspicious for OC/precancer. The OC relative frequency measured the percentage of persons diagnosed with OSCC among persons with OSCC, dysplasia, or HK/EH. OC relative frequencies for PR and NYC laboratories were compared.Overall, the OC relative frequency was 67% in PR and 40% and 4% in the NYC general and oral pathology laboratories, respectively (each p0.001). In PR, the OC relative frequency was highest for males (80%). When OC relative frequencies were stratified by pathology laboratory type (general/oral) and compared across PR and NYC, age/gender-specific OC relative frequencies were always higher in PR; however, differences were consistently statistically significant for males only.A disparity in the OC relative frequency exists in PR vs. NYC indicating a shortfall in biopsying potentially precancerous oral lesions in PR. PR residents with intraoral lesions suspicious for oral cancer/precancer are most likely to be biopsied only after developing an invasive OC.

Overexpression of HMGA2 is common in uterine leiomyomas (ULM). The expression of HMGA2 in its malignant counterpart – uterine leiomyosarcomas (ULMS) remains undetermined. Recently it has been shown that repression of HMGA2 by microRNA let-7s is a critical molecular regulatory mechanism associated with tumour growth in many tumours and cell types, including leiomyomas. To test whether HMGA2 and let-7s play a role in ULMS, we examined the levels of endogenous HMGA2 and let-7 expression and found a significant correlation between these two molecules in a case-matched cohort of human ULMS. We found that overexpression of HMGA2 and let-7-mediated HMGA2 repression is a relevant molecular alteration in ULMS. Disrupting the control of HMGA2 and let-7 pairs promotes ULMS cell growth in vitro.

Hormones and growth factors regulate endometrial cell growth. Disrupted transforming growth factor-β (TGF-β) signaling in primary endometrial carcinoma (ECA) cells leads to loss of TGF-β–mediated growth inhibition, which we show herein results in lack of up-regulation of the cyclin-dependent kinase inhibitor p27Kip1 (p27) to arrest cells in G1 phase of the cell cycle. Conversely, in normal primary endometrial epithelial cells (EECs), TGF-β induces a dose-dependent increase in p27 protein, with a total 3.6-fold maximal increase at 100 pmol/L TGF-β, which was 2-fold higher in the nuclear fraction; mRNA levels were unaffected. In addition, ECA tissue lysates show a high rate of ubiquitin-mediated degradation of p27 compared with normal secretory-phase endometrial tissue (SE) such that 4% and 89% of recombinant p27 added to the lysates remains after 3 and 20 h, respectively. These results are reflected in vivo as ECA tissue lacks p27 compared with high expression of p27 in SE (P ≤ 0.001). Furthermore, we show that estrogen treatment of EECs causes mitogen-activated protein kinase–driven proteasomal degradation of p27 whereas progesterone induces a marked increase in p27 in both normal EECs and ECA cells. Therefore, these data suggest that TGF-β induces accumulation of p27 for normal growth regulation of EECs. However, in ECA, in addition to enhanced proteasomal degradation of p27, TGF-β cannot induce p27 levels due to dysregulated TGF-β signaling, thereby causing 17β-estradiol–driven p27 degradation to proceed unchecked for cell cycle progression. Thus, p27 may be a central target for growth regulation of normal endometrium and in the pathogenesis of ECA. [Cancer Res 2007;67(3):1007–18]

Polymorphonuclear neutrophils (PMNs) are largely considered to foster cancer development despite wielding an arsenal of cytotoxic agents. Using a mouse model of PTEN-deficient uterine cancer, we describe a surprising inhibitory role for PMNs in epithelial carcinogenesis. By inducing tumor cell detachment from the basement membrane, PMNs impeded early-stage tumor growth and retarded malignant progression. Unexpectedly, PMN recruitment and tumor growth control occurred independently of lymphocytes and cellular senescence and instead ensued as part of the tumor's intrinsic inflammatory response to hypoxia. In humans, a PMN gene signature correlated with improved survival in several cancer subtypes, including PTEN-deficient uterine cancer. These findings provide insight into tumor-associated PMNs and reveal a context-specific capacity for PMNs to directly combat tumorigenesis.

Objective To evaluate the growth pattern of the large uterine leiomyomata (ULM), we examined the spatial gene distributions, vessel density, proliferative activity, and hyaline degeneration. Design Tissue sections from three-dimensional large ULM, matched myometrium, and small ULM were collected and microarrayed. The spatial difference of the tumor activity was mapped in large ULM. Setting University clinical research laboratory. Patient(s) Hysterectomy specimens from 7 patients with large (>10 cm) ULM and 3 patients with large (>10 cm) uterine leiomyosarcomas. Intervention(s) Tissue microarray analysis by the immunohistochemistry. Main Outcome Measure(s) Selected gene products, vessel density, and the percentage of hyaline degeneration were all scored in tissue cores/sections of large and small ULM against matched myometrium. Result(s) We found that there was a spherical spatial difference of the tumor activities in large ULM. The tumor region next to the periphery, the most biologically active zone, demonstrated higher levels of gene expression, a higher density of vessels, a higher proliferative rate and a lower level of hyaline degeneration. The large ULM have higher levels of gene products (except for estrogen and progesterone receptors) than small ULM. Conclusion(s) In comparison of the spatial patterns of the gene activity between the large ULM and the large uterine leiomyosarcoma, the large ULM illustrate a growth pattern of nutritional dependence.

Human uterine leiomyomata are the most common tumors in women of reproductive age. The pathogenesis of leiomyomata remains unknown. An animal model of Eker rats with deleted tuberous sclerosis complex gene 2 (tuberin) shows increased incidence of leiomyomata. The role of tuberin in human leiomyomata is unknown. In this study, we designed a tissue microarray with tissue cores of leiomyomata and the matched myometrium from 60 hysterectomy specimens. We examined the expression of tuberin and tuberous sclerosis complex gene 1 product hamartin, proteins of the insulin-signaling pathway, steroid receptors and some of their cofactors, and human mobility group gene A2 by immunohistochemistry. We found that nearly half of the cases displayed either reduction or loss of tuberin in leiomyomata compared with matched normal myometrium. No change of hamartin was noted. Furthermore, a significant reduction of glucocorticoid receptor was found in leiomyomata with reduced tuberin. The proteins insulin like growth factor 1, insulin-like growth factor receptor beta, AKT kinase, and phosphatidylinositol 3-kinase were upregulated. Nearly half of leiomyomata show upregulation of human mobility group gene A2, along with the steroid receptor cofactors. Our findings suggest that there are two broad groups of uterine leiomyomata. One group is associated with an alteration of tuberin and glucocorticoid receptor. The other group is associated with upregulation of human mobility group gene A2 and steroid receptor cofactors.

OBJECTIVE: Our purpose was to evaluate whether tamoxifen has estrogenic endometrial effects as defined by histologic study or alterations in steroid hormone receptor expression. STUDY DESIGN: Nineteen postmenopausal tamoxifen-treated breast cancer patients who also had endometrial sampling were identified from files in the Department of Obstetrics and Gynecology. To examine the subgroup of 15 polyps, age-matched, non–hormonally treated patients with polyps ( n = 8) or atrophic endometria ( n = 5) served as comparison groups. Proliferative ( n = 3) and secretory ( n = 5) endometria served as procedural controls. Immunohistochemical studies for steroid receptors (estrogen, progesterone) were performed. RESULTS: Glandular cell progesterone receptor was significantly increased and stromal cell estrogen receptor was significantly decreased in tamoxifen-treated versus atrophic endometria. Progesterone receptor staining was not significantly different in tamoxifen-treated versus control polyps, although staining was high in both groups. Stromal cell estrogen receptor staining was significantly reduced in tamoxifen-treated versus control polyps, although there were no histologic differences. Reduced stromal cell estrogen receptor and increased glandular cell progesterone receptor staining was found in all tamoxifen-treated endometria regardless of the diagnosis. CONCLUSION: The tamoxifen-associated changes in endometrial steroid receptors support an estrogenic effect that is independent of histologic diagnosis and duration of use. This may contribute to the pathogenesis of tamoxifen-associated polyps and carcinomas. (Am J Obstet Gynecol 1997;176:129-37.)

Studies using cervical carcinoma cell lines usually show mutated p53 in cases without detectable HPV, and wild-type p53 in cases with detectable HPV. These findings suggest that loss of p53 function, either by mutation or by binding to HPV E6, is required for cervical carcinogenesis. Because mutated p53 is usually detectable immunohistochemically, one would predict an inverse relationship between the presence of HPV and detectable p53. In this study we examined 88 formalin-fixed paraffin-embedded clinical specimens of cervix for the presence of HPV and p53 expression. All cases were studied for the presence of p53 using immunohistochemical methods. The antibody used was mouse monoclonal PAb1801 (Biogenex). The presence of HPV was detected by PCR. Twenty-six specimens showed foci of p53 expression (0/7 normal, 1/8 (13%) condylomas, 1/6 (17%) CIN I, 3/7 (43%) CIN II, 6/20 (30%) CIN III, 13/22 (59%) SCC, 2/5 (40%) adenosquamous carcinomas, and 0/13 adenocarcinomas). p53 expression was more frequent in SCC than with CIN (P = 0.026). HPV was present in 15 of 24 cases with detectable p53 and 22 of 48 cases without detectable p53. No correlation was seen between HPV status and detection of p53. With the exception of one case, p53 expression was seen in less than 10% of cells. p53 expression was not detected in any of the 13 adenocarcinomas examined (P = 0.0016 vs SCC). Our results show that alterations of p53 may play a role in the pathogenesis of cervical squamous carcinoma. However, p53 expression was neither sufficient nor required for cervical carcinogenesis, irrespective of HPV status.

It is uncertain whether uterine leiomyosarcoma arises de novo or in preexisting leiomyoma. Leiomyoma-like areas can be seen associated with uterine leiomyosarcoma, raising the possibility of precursor lesions for uterine leiomyosarcoma. In this study, we examined cases of uterine leiomyosarcoma associated with leiomyoma-like areas at the histological, immunohistochemical and DNA level to further evaluate if benign-looking leiomyoma-like and uterine leiomyosarcoma areas are related. Cases of uterine leiomyosarcoma observed at the New York University Medical Center from 1994 to 2007 were reviewed for the presence of leiomyoma-like areas. Of the 26 cases of uterine leiomyosarcoma observed during this period, 18 cases had an associated leiomyoma-like area (five cellular leiomyoma, four symplastic leiomyoma, four cellular and symplastic leiomyoma and five usual type leiomyoma). Sixteen of the 18 cases were examined immunohistochemically for Ki-67, for estrogen receptor, progesterone receptor and for p53. Immunohistochemical profiles were as expected for leiomyoma-like (the mean expression of p53, ER, PR and Ki-67 at 0.3, 63, 75 and 0.6%, respectively), symplastic leiomyoma-like areas (the mean expression of p53, ER, PR and Ki-67 at 0.6, 85, 89 and 5.5%, respectively) and uterine leiomyosarcoma areas (the mean expression of p53, ER, PR and Ki-67 at 52, 38, 39 and 61%, respectively). In six cases, the leiomyoma-like and uterine leiomyosarcoma areas from each case were examined using high-density oligonucleotide array-CGH to determine genetic aberrations in the two areas. Nearly all the genetic aberrations found in leiomyoma-like areas were also found in the corresponding uterine leiomyosarcoma areas. In addition, uterine leiomyosarcoma areas had additional genetic aberrations. The immunohistochemical profiles and genetic aberrations of the examined cases suggest that uterine leiomyosarcoma could arise from the preexisting leiomyoma-like areas that often have a symplastic or cellular morphology.

The pathology of term placentas from seropositive human immunodeficiency virus (HIV)-infected and seronegative women was investigated by routine histologic, immunocytochemical, and in situ hybridization techniques. Placentas were evaluated for evidence of villitis, chorioamnionitis, and funisitis. Membranes, trophoblast, and decidua were also examined by immunohistochemistry using monoclonal HIV p24 antibody. Twenty placentas were evaluated by combined immunochemical and in situ hybridization techniques, using a 35S-labeled RNA probe complementary to the 3' long terminal repeat and envelope region of HIV-1. HIV-seropositive placentas did not show significant villitis; however, the incidence of chorioamnionitis increased (P less than .01). HIV antigens and nucleic acids were identified in the trophoblast of 10% of the placentas that also showed chorionitis. Term HIV-positive placentas may show histologic changes that may or may not be directly related to the virus. Analysis of tissues from earlier gestational placentas may prove more informative in clarifying the mechanism of maternal-fetal HIV transmission.

High-mobility group A2 (HMGA2) is commonly overexpressed in large leiomyomas. HMGA2 is an important regulator of cell growth, differentiation, apoptosis, and transformation. As a predicted target of Let-7 microRNAs (Let-7s), HMGA2 can be repressed by Let-7s in vitro. MicroRNA profiling analysis revealed that Let-7s were significantly dysregulated in uterine leiomyomas: high in small leiomyomas and lower in large leiomyomas. To evaluate whether Let-7 repression of HMGA2 plays a major role in leiomyomas, we analyzed the molecular relationship of HMGA2 and Let-7s, both in vitro and in vivo. We first characterized that exogenous Let-7 microRNAs could directly repress the dominant transcript of HMGA2, HMGA2a. This repression was also identified for two cryptic HMGA2 transcripts in primary leiomyoma cultures. Second, we found that the endogenous Let-7s were biologically active and played a major role in the regulation of HMGA2. Then, we illustrated that Let-7 repression of HMGA2 inhibited cellular proliferation. Finally, we examined the expression levels of Let-7c and HMGA2 in a large cohort of leiomyomas (n = 120), and we found high levels of Let-7 and low levels of HMGA2 in small leiomyomas, and low levels of Let-7 and high levels of HMGA2 in large leiomyomas. Our findings suggest that the Let-7–mediated repression of HMGA2 mechanism can be an important molecular event in leiomyoma growth. (Mol Cancer Res 2008;6(4):663–73)

BACKGROUND: Black ethnicity is one of the risk factors for uterine leiomyomata (ULM). Little is known about the ethnic differences in leiomyoma-associated gene products in women with uterine leiomyomata. METHODS: A total of 120 hysterectomies with ULM were collected from black, Asian, Hispanic and white women (30 cases from each group). Twenty-two gene products were selected for the study. The expressions of the selected dysregulated gene products were measured by the semiquantification and the immunoscores were normalized by matched myometrium. RESULTS: The relative expressions of progesterone receptor A (PR-A) (up-regulation), retinoid acid receptor (down-regulation), and retinoid X receptor (RXR) (no change) in leiomyomata compared to normal myometrium in black women were significantly different compared to other ethnic groups (P < 0.05). About one-third of ULM from black women subclustered together in association with a group of up-regulated gene products. Many other gene products, including local growth factors, insulin-like growth factor (IGF)-signalling proteins, and cell proliferation markers, were dysregulated in ULM but showed non-significant differences between the ethnic groups. CONCLUSIONS: There are substantial differences of the sex steroid receptors and other nuclear receptors between black women and other ethnic groups. Based on tissue microarray data, there are at least two broad groups of leiomyomata presented by the dysregulation of different groups of gene products. One is dominated by up-regulation of amplified in breast cancer 1, CD24, hamartin, human mobility group gene 2, IGF2, PR-A and RXR, and the other is characterized by up-regulation of epithelial growth factor receptor, down-regulation of hamartin, PR-A and tuberin.

Objective To use tissue microarray in combination with dendrogram cluster analysis to characterize some potential tumorigenic factors in association with tumor size and sex steroid hormone status in uterine leiomyomata. Design Expression analysis of 21 selected potential tumorigenic factors in 60 patients with uterine leiomyomata. Setting University clinical research laboratory. Patient(s) Hysterectomy specimens from 60 patients with uterine leiomyomata of different ages and tumor sizes. Intervention(s) Tissue cores from normal myometrium and leiomyomata were examined by immunohistochemistry. Main Outcome Measure(s) Semiquantitative immunointensity was scored and analyzed by net gain and loss between normal myometrium and leiomyomata and integrated into dendrogram cluster tree view. Result(s) We found that upregulation of estrogen and progesterone receptors was reverse associated with tumor size. Upregulation of some factors (HMGA2, sex steroid receptor cofactors, proteins in insulin pathway, and CD24) were identified in a group of patients in their later forties, were associated with large fibroids, and were weakly affected by the SSH status (illustrated by endometrial phases and menopause). Downregulation of tuberin and glucocorticoid receptor was mostly isolated in a second group of women at their late reproductive age. Conclusion(s) Analyses of the sex steroid hormone receptors and the nonsex steroid hormone factors in leiomyomata and the matched myometrium showed different expression patterns in different tumor sizes and patients' ages. A group of patients in their late forties with the larger leiomyomata contributes largely by upregulation of nonsex steroid hormone factors. Adenomyosis is a protective factor preventing large leiomyomata.

In a retrospective study to determine the accuracy of frozen section diagnoses in ovarian neoplasms, the results of consecutive frozen section diagnoses of 311 ovarian neoplasms from two institutions, New York University Medical Center and State University of New York Medical Center at Brooklyn, from 1980 through 1989 were compared with the final diagnosis results following extensive sampling on permanent sections. The final diagnosis was assumed to be correct for purposes of this study. Ovarian neoplasms were correctly diagnosed on frozen section as either benign or malignant in 292 patients (accuracy of 93.8%). Frozen section diagnoses were incorrect in 11 patients (3.5%). Frozen section diagnosis was deferred in 8 instances (2.6%). The positive predictive value was 100%. The negative predictive value was 95.3%, specificity 100%, and sensitivity 86%. There were no false positives. Of the 11 false negative frozen section diagnoses, 9 (82%) were due to limited sampling for frozen section. We therefore suggest that careful examination with sampling of any suspicious lesions be carried out at the time of surgery for patients with benign frozen section diagnosis, since this may avoid a second staging laparotomy, if the final diagnosis is malignant.