Bojana Djordjevic

High Frequency of PIK3R1 and PIK3R2 Mutations in Endometrial Cancer Elucidates a Novel Mechanism for Regulation of PTEN Protein Stability

We demonstrate that phosphatidylinositol 3-kinase (PI3K) pathway aberrations occur in >80% of endometrioid endometrial cancers, with coordinate mutations of multiple PI3K pathway members being more common than predicted by chance. PIK3R1 (p85α) mutations occur at a higher rate in endometrial cancer than in any other tumor lineage, and PIK3R2 (p85β), not previously demonstrated to be a cancer gene, is also frequently mutated. The dominant activation event in the PI3K pathway appears to be PTEN protein loss. However, in tumors with retained PTEN protein, PI3K pathway mutations phenocopy PTEN loss, resulting in pathway activation. KRAS mutations are common in endometrioid tumors activating independent events from PI3K pathway aberrations. Multiple PIK3R1 and PIK3R2 mutations demonstrate gain of function, including disruption of a novel mechanism of pathway regulation wherein p85α dimers bind and stabilize PTEN. Taken together, the PI3K pathway represents a critical driver of endometrial cancer pathogenesis and a novel therapeutic target.

Consequences of universal MSI/IHC in screening ENDOMETRIAL cancer patients for lynch syndrome

OBJECTIVE Determine factors impacting the uptake of genetic counseling and results of genetic testing following universal tumor testing for Lynch syndrome in patients with endometrial cancer. METHODS The study population consisted of two unselected cohorts of endometrial cancer patients, 408 identified retrospectively and 206 identified prospectively. Immunohistochemistry for mismatch repair protein expression and/or microsatellite instability analysis was performed on these tumors. MLH1 methylation analysis was performed on tumors with loss of MLH1 protein. Tumor studies were considered suggestive of Lynch Syndrome if they showed immunohistochemical loss of MSH2, MSH6 or PMS2, loss of MLH1 without MLH1 promoter methylation, and/or microsatellite instability. Participants with suggestive tumor studies were contacted and offered genetic counseling and testing. RESULTS In the retrospective cohort, 11% had tumor studies suggestive of Lynch syndrome, and 42% was seen for genetic counseling. A germline mutation was detected in 40%, and one had a variant of uncertain significance. In the prospective cohort, 8.7% of patients had tumor testing suggestive of Lynch syndrome; 72% were seen for genetic counseling. Germline mutations were found in 40%, and one had a variant of uncertain significance. Common challenges included timing of re-contact, age, perceived lack of relevance, inability to travel and limited insurance coverage. CONCLUSIONS There are several barriers to genetic counseling and testing follow-up after universal tumor testing, and uninformative genetic test results present a management challenge. It is important to consider these limitations when implementing an approach to screening endometrial cancer patients for Lynch syndrome.

Loss of switch/sucrose non-fermenting complex protein expression is associated with dedifferentiation in endometrial carcinomas.

Dedifferentiated endometrial carcinoma is an aggressive type of endometrial cancer that contains a mix of low-grade endometrioid and undifferentiated carcinoma components. We performed targeted sequencing of eight dedifferentiated carcinomas and identified somatic frameshift/nonsense mutations in SMARCA4, a core ATPase of the switch/sucrose non-fermenting (SWI/SNF) complex, in the undifferentiated components of four tumors. Immunohistochemical analysis confirmed the loss of SMARCA4 in the undifferentiated component of these four SMARCA4-mutated cases, whereas the corresponding low-grade endometrioid component showed retained SMARCA4 expression. An expanded survey of other members of the SWI/SNF complex showed SMARCB1 loss in the undifferentiated component of two SMARCA4-intact tumors, and all SMARCA4- or SMARCB1-deficient tumors showed concomitant loss of expression of SMARCA2. We subsequently examined the expression of SMARCA2, SMARCA4, and SMARCB1 in an additional set of 22 centrally reviewed dedifferentiated carcinomas and 31 grade 3 endometrioid carcinomas. Combining the results from the index and the expansion set, 15 of 30 (50%) of the dedifferentiated carcinomas examined showed either concurrent SMARCA4 and SMARCA2 loss (37%) or concurrent SMARCB1 and SMARCA2 loss (13%) in the undifferentiated component. The loss of SMARCA4 or SMARCB1 was mutually exclusive. All 31 grade 3 endometrioid carcinomas showed intact expression of these core SWI/SNF proteins. The majority (73%) of the SMARCA4/SMARCA2-deficient and half of SMARCB1/SMARCA2-deficient undifferentiated component developed in a mismatch repair-deficient molecular context. The observed spatial association between SWI/SNF protein loss and histologic dedifferentiation suggests that inactivation of these core SWI/SNF proteins may contribute to the development of dedifferentiated endometrial carcinoma.

Expression of c-kit in fibroepithelial lesions of the breast is a mast cell phenomenon

The expression of c-kit, a protooncogene tyrosine kinase receptor (CD117), in phyllodes tumors of the breast has been the subject of recent investigations. We examined stromal c-kit expression by immunohistochemistry in 68 cases comprising fibroadenomas, fibroadenomas with cellular stroma, and benign, borderline, and malignant phyllodes tumors. Membrane staining was identified in the epithelium of 82% of cases, representing all diagnostic categories in the study. Membrane and cytoplasmic staining was detected in scant cells in the stroma in 74% of the cases in the study, again, across all diagnostic entities. However, when toluidine blue and tryptase staining was performed, the staining pattern matched that of c-kit in the number of cells and their distribution, thereby confirming the presence of mast cells and excluding any appreciable level of true stromal c-kit staining. One borderline and one malignant phyllodes tumor showed a diffuse weak stromal signal, which could not be accounted for by toluidine blue and tryptase. These cases were further tested for the presence of known activating mutations of c-kit and PDGFR-α, and yielded negative results. Our findings indicate that c-kit is an unlikely player in the pathogenesis of fibroepithelial lesions of the breast. The positive stromal staining suggested previously by other authors may be attributed to mast cells. C-kit, therefore, has neither a diagnostic nor a prognostic role in phyllodes tumors, and there is no rationale for the treatment of recurrent of malignant phyllodes tumor patients with tyrosine kinase inhibitors.

The pattern of myometrial invasion as a predictor of lymph node metastasis or extrauterine disease in low-grade endometrial carcinoma.

The purpose of this study was to examine predictors of lymph node (LN) metastases or extrauterine disease (ED) in low-grade (FIGO grade 1 or 2) endometrioid carcinoma (LGEC) in a multi-institutional setting. For LGEC with and without LN metastasis or ED, each of the 9 participating institutions evaluated patients’ age, tumor size, myometrial invasion (MI), FIGO grade, % solid component, the presence or absence of papillary architecture, microcystic, elongated, and fragmented glands (MELF), single-cell/cell-cluster invasion (SCI), lymphovascular invasion (LVI), lower uterine segment (LUS) and cervical stromal (CX) involvement, and numbers of pelvic and para-aortic LNs sampled. A total of 304 cases were reviewed: LN+ or ED+, 96; LN−/ED−, 208. Patients’ ages ranged from 23 to 91 years (median 61 y). Table 1 summarizes the histopathologic variables that were noted for the LN+ or ED+ group: tumor size ≥2 cm, 93/96 (97%); MI>50%, 54/96 (56%); MELF, 67/96 (70%); SCI, 33/96 (34%); LVI, 79/96 (82%); >20% solid, 65/96 (68%); papillary architecture present, 68/96 (72%); LUS involved, 64/96 (67%); and CX involved, 41/96 (43%). For the LN−/ED− group, the results were as follows: tumor size ≥2 cm, 152/208 (73%); MI>50%, 56/208 (27%); MELF, 79/208 (38%); SCI, 19/208 (9%); LVI, 56/208 (27%); >20% solid, 160/208 (77%); papillary architecture present, 122/208 (59%); LUS involved, 77/208 (37%); CX involved, 24/208 (12%). There was no evidence of a difference in the number of pelvic or para-aortic LNs sampled between groups (P=0.9 and 0.1, respectively). After multivariate analysis, the depth of MI, CX involvement, LVI, and SCI emerged as significant predictors of advanced-stage disease. Although univariate analysis pointed to LUS involvement, MELF pattern of invasion, and papillary architecture as possible predictors of advanced-stage disease, these were not shown to be significant by multivariate analysis. This study validates MI, CX involvement, and LVI as significant predictors of LN+ or ED+. The association of SCI pattern with advanced-stage LGEC is a novel finding.

Utility of cytokeratin 5/6, cytokeratin 20, and p16 in the diagnosis of reactive urothelial atypia and noninvasive component of urothelial neoplasia.

Background:The utility of cytokeratin (CK)5/6 in distinguishing reactive urothelial atypia (RA) from urothelial carcinoma in situ (CIS) and from noninvasive component of high-grade papillary urothelial carcinoma (PUC) is unknown. Design:Twenty RA with or without papillary hyperplasia and 90 noninvasive components of neoplastic urothelial lesions were submitted for immunostaining with CK5/6, CK20, and p16. Results:Diffuse and strong reactivity involving the full or nearly full thickness of urothelium was observed with CK5/6 in 90% of RA cases. CK20 and p16 were negative in 90% and 85% of the RA cases, respectively. For CIS and noninvasive components of high-grade PUC without squamous differentiation, there was no CK5/6 staining or reactivity in the basal layer only. CK20 and p16 showed strong positivity in full thickness of urothelium in 75% to 85% of cases. CIS with weak/focal or negative reactivity for p16 or CK20 exhibited moderate cytologic atypia. Low-grade PUC displayed variable reactivity for CK5/6, CK20, and p16. Urothelial lesions with squamoid or basaloid features showed positive reactivity for CK5/6. Urothelial lesions with glandular differentiation showed negative reactivity for CK5/6. Conclusions:Diffuse CK5/6 reactivity in RA and negative CK5/6 reactivity in CIS and PUC may be helpful in distinguishing between these 2 entities.

The Search Continues

Objective PI3K/mammalian target of rapamycin (mTOR) pathway aberrations occur in 40% to 80% of endometrial cancer. Prior studies suggest KRAS mutations are associated with resistance to mTOR inhibitors in solid tumors. The objective of this study was to determine if biomarker expression in the PI3K/mTOR pathway or KRAS mutations would predict response to therapy with everolimus, an mTOR inhibitor. Methods Specimens from a phase II study of everolimus in recurrent endometrioid endometrial cancer were utilized. The primary end point was clinical benefit rate (CBR: objective response and nonprogression at 20 weeks). Correlative studies evaluating PTEN expression and phospho-S6 ribosomal protein (pS6rp) status by immunohistochemistry and KRAS mutational analysis were performed. Results Six of 28 evaluable patients achieved prolonged stable disease (SD) at 20 weeks (CBR, 21%). Loss of PTEN expression did not predict CBR (P = 0.62) with a positive predictive value (PPV) of 0.13. Five (83%) of 6 patients with SD maintained PTEN expression. Neither pS6rp expression (P = 0.65) nor KRAS mutation (P = 0.99) predicted CBR; the PPV was 0.14 for each. Eighty percent (4/5) of those with SD were KRAS wild type. Combined analysis of pS6rp expression and KRAS mutation provided 100% PPV (95% confidence interval, 39.6%–100%), suggesting no chance of CBR for these individuals with 100% specificity (95% confidence interval, 46.3%–100%). Conclusions S6rp phosphorylation, loss of PTEN expression, and presence of KRAS mutations alone did not correlate with CBR. However, positive pS6rp staining combined with KRAS mutation performed with 100% PPV and specificity to predict nonresponse. Identifying patients who will not benefit from mTOR inhibitors can direct therapy and reduce exposure to agents that add toxicity without clinical benefit.

Histopathology of Placenta Creta: Chorionic Villi Intrusion into Myometrial Vascular Spaces and Extravillous Trophoblast Proliferation are Frequent and Specific Findings With Implications for Diagnosis and Pathogenesis

Placenta creta is characterized by invasion of placental villi into the myometrium in the setting of a dysfunctional or absent decidua. Histopathologic diagnosis of placenta creta is important, particularly in cases of hysterectomy because of unanticipated intractable postpartum hemorrhage. Previous studies have documented a higher amount and depth of myometrial infiltration by the implantation site intermediate trophoblast compared with controls. In addition, we have anecdotally observed chorionic villi in myometrial vascular spaces in specimens with placenta creta. The aim of this study was to explore the prevalence and specificity of these features. Sixty-one postpartum hysterectomies, 44 with placenta creta and 17 without were reviewed. Villous intrusion into vascular spaces was recorded. Using immunohistochemistry for GATA3, the amount of intermediate trophoblast (number of positive cells in five 40× fields) and depth of trophoblast myometrial infiltration were assessed. Mean gestational ages of the creta group (34.4 yr; range, 20–43 yr) and control group (35 yr; range, 25–51 yr) were comparable. Presence of chorionic villi in myometrial vascular spaces was frequent in placenta creta: 31/44 versus 1/17 controls (70.4% vs. 5.8%, P<0.0001). This finding was more common in the percreta (87.5%) and increta (84%) than in the accreta (27.2%, P=0.0008). Mean depth of trophoblast myometrial invasion was greater in cretas (47.9%) than in controls (14.5%, P=0.004). Likewise, mean distance of deepest trophoblast to serosa was shorter in the cretas (7.3 mm) than in controls (23.8 mm, P<0.0001). These differences were, however, attributable to placentas increta and percreta. When only accretas and controls were compared, the myometrial depth of trophoblast was similar. The mean intermediate trophoblast cell count in the placental bed was greater in cretas (664) than in controls (288, P<0.0001). Such difference was seen in all creta cases despite the type (accreta 639, increta 676, percreta 661). A trophoblast count of ≥100 cells/high-power field was seen in 75.8% of cretas and 11.1% of controls (P=0.0009). For the first time, we document the finding of chorionic villi intrusion into myometrial vascular spaces, which is highly specific of placenta creta. In addition, assessment of the amount of intermediate trophoblast using GATA3 immunohistochemistry can assist in the diagnosis. We hypothesize that placental invasion in placenta creta is due, at least partially, to transformation of low-resistance myometrial vessels leading to subsequent protrusion of villi into their lumens, in the context of absent decidua.

Molecular pathogenesis of endometrial cancers in patients with Lynch syndrome

The authors hypothesized that Lynch syndrome (LS)‐associated endometrial cancer (EC) develops from morphologically normal endometrium that accumulates enough molecular changes to progress through a continuum of hyperplasia to carcinoma, similar to sporadic EC. The primary objective of the current study was to determine whether LS‐associated EC involves progression through a preinvasive lesion. The secondary objective was to identify molecular changes that contribute to endometrial carcinogenesis in patients with LS.

Immunohistochemistry in the Diagnosis of Mucinous Neoplasms Involving the Ovary: The Added Value of SATB2 and Biomarker Discovery Through Protein Expression Database Mining.

Immunohistochemistry is frequently used to identify ovarian mucinous neoplasms as primary or metastatic; however, there is significant overlap in expression patterns. We compared traditional markers (CK7, CK20, CDX2, PAX8, estrogen receptor, &bgr;-catenin, MUC1, MUC2, and MUC5AC) to 2 novel proteins identified through mining of the Human Protein Atlas expression database: SATB2 and POF1B. The study cohort included 49 primary gastrointestinal (GI) mucinous adenocarcinomas (19 colorectal, 15 gastric, 15 pancreatobiliary), 60 primary ovarian mucinous neoplasms (19 cystadenomas, 21 borderline tumors, 20 adenocarcinomas), and 19 metastatic carcinomas to the ovary (14 lower and 5 upper GI primaries). Immunohistochemistry was performed on tissue microarrays, scored and interpreted as negative (absent or focal/weak) or positive. Metastatic tumors were frequently unilateral (42.8% of tumors from lower and 40% of tumors from upper tract) and ≥10 cm (85.7% of tumors from lower and 80% of tumors from upper tract). CK7 was positive in 88.5% upper GI and 88.3% primary ovarian compared with 24.3% lower GI neoplasms. CK20 and CDX2 were positive in 84.8% and 100% of lower GI tumors, respectively; however, expression was also common in upper GI (CK20 42.8%, CDX2 50%) and primary ovarian neoplasms (CK20 65.7%, CDX2 38.3%). Conversely, SATB2 was more specific for lower GI origin, being positive in 78.8% lower GI but only 11.5% upper GI and 1.7% primary ovarian neoplasms. PAX8 expression was common in primary ovarian neoplasms (75% of all neoplasms, 65% of carcinomas); only 1 (1.5%) GI tumor was positive. MUC2 and &bgr;-catenin were frequently positive in lower GI tumors (96.9% and 51.5%, respectively). Estrogen receptor expression was only seen in primary ovarian neoplasms (13.3%). Nuclear premature ovarian failure 1B (POF1B) expression was seen in malignant tumors regardless of their origin. A panel including CK7, SATB2, and PAX8 separated primary from secondary GI neoplasms with up to 77.1% sensitivity and 99% specificity, outperforming tumor laterality and size. Second-line markers such as CDX2, MUC2, estrogen receptor, MUC1, and &bgr;-catenin increased the sensitivity of immunohistochemistry in excluding lower GI origin. Biomarker search using proteomic databases has a value in diagnostic pathology, as shown with SATB2; however, as seen with POF1B, expression profiles in these databases are not always reproduced in larger cohorts.