Search

Since the establishment of 4 molecular subgroups of endometrial carcinoma (EC), there has been significant interest in understanding molecular classification in the context of histologic features and diagnoses. ECs with undifferentiated, spindle, and/or sarcomatous components represent a diagnostically challenging subset of tumors with overlapping clinical and histologic features. We examined the clinicopathologic, morphologic, immunohistochemical, and molecular features of these tumors identified in our institutions' pathology databases using immunohistochemistry and targeted sequencing. Disease-specific survival (DSS) and progression-free survival (PFS) were analyzed using Kaplan-Meier curves and log-rank tests. One hundred sixty-two ECs were included: carcinosarcomas (UCS; n=96), dedifferentiated/undifferentiated EC (DDEC/UDEC; n=49), and grade 3 endometrioid EC with spindled growth (GR3spEEC) (n=17). All molecular subgroups were represented in all histologic subtypes and included 12 (7%) POLE -mutated ( POLE mut), 43 (27%) mismatch repair-deficient (MMRd), 77 (48%) p53-abnormal (p53abn), and 30 (19%) no specific molecular profile (NSMP) tumors. However, the molecular classification (irrespective of histologic diagnosis) was a significant predictor for both DSS ( P =0.008) and P≤0.0001). POLE mut EC showed an excellent prognosis with no recurrences or deaths from the disease. MMRd tumors also showed better outcomes relative to NSMP and p53abn tumors. In conclusion, molecular classification provides better prognostic information than histologic diagnosis for high-grade EC with undifferentiated and sarcomatous components. Our study strongly supports routine molecular classification of these tumors, with emphasis on molecular group, rather than histologic subtyping, in providing prognostication., Competing Interests: Conflicts of Interest and Source of Funding: The authors have disclosed that they have no significant relationships with, or financial interest in, any commercial companies pertaining to this article., (Copyright © 2024 Wolters Kluwer Health, Inc. All rights reserved.)

Endometrial somatically derived yolk sac tumors are characterized by yolk sac morphology with AFP, SALL-4, and/or Glypican-3 immunoexpression. Yolk sac marker expression, however, is not limited to tumors with overt yolk sac histology. Three hundred consecutive endometrial malignancies were assessed for immunomarkers of yolk sac differentiation. Of these, 9% expressed ≥1 yolk sac marker, including 29% of high-grade tumors. Only 3 (1%) met morphologic criteria for yolk sac differentiation; these were originally diagnosed as serous, high-grade NOS, and dedifferentiated carcinoma. Two were MMR-intact and comprised exclusively of yolk sac elements, while the dedifferentiated case was MMR deficient and had a background low-grade endometrioid carcinoma; this case also showed BRG1 loss. All 3 were INI1 intact. Nonspecific yolk sac marker expression was seen in 14 carcinosarcomas, 4 endometrioid, 2 serous, 1 clear cell, 1 dedifferentiated, 1 mixed serous/clear cell, and 1 mesonephric-like carcinoma. INI1 was intact in all cases; one showed BRG1 loss. Twenty were MMR-intact, and 4 were MMR deficient. All MMR-deficient cases with yolk sac marker expression, both with and without true yolk sac morphology, had no evidence of residual disease on follow-up, whereas 82% of MMR-intact cases developed recurrent/metastatic disease. In summary, endometrial somatically derived yolk sac tumors were rare but under-recognized. While AFP immunostaining was specific for this diagnosis, Glypican-3 and SALL-4 expression was seen in a variety of other high-grade carcinomas. INI1 loss was not associated with yolk sac morphology or immunomarker expression in the endometrium, and BRG1 loss was rare. All patients with MMR-deficient carcinomas with yolk sac immunoexpression +/- morphology were disease-free on follow-up, whereas the majority of MMR-intact cancers showed aggressive disease., Competing Interests: Conflicts of Interest and Source of Funding: The authors have disclosed that they have no significant relationships with, or financial interest in, any commercial companies pertaining to this article., (Copyright © 2024 Wolters Kluwer Health, Inc. All rights reserved.)

Immunostaining of endometrial carcinomas for mismatch repair (MMR) protein loss is standard-of-care for Lynch syndrome screening, but also identifies MMR-deficient tumors without germline pathogenic variants. While the majority show MLH1 hypermethylation ( MLH1hm ), somatic MMR pathogenic variants are increasingly recognized drivers of immunohistochemistry-germline discordance. Because MMR abnormalities with both germline and somatic origins have prognostic significance and impart susceptibility to immune checkpoint inhibitors, it is important to understand how frequently tumors with MMR immunohistochemical loss and normal germline testing ("Lynch-like" tumors) have underlying somatic MMR pathogenic variants. Somatic tumor sequencing±microsatellite instability (MSI) testing was performed on 18 endometrial cancers with MMR immunohistochemical loss but negative MMR germline results and negative MLH1hm where relevant. Tumor sequencing and MSI testing were successful in 94%. Where successful, 80% were MSI-high and 94% had a molecular correlate for the initial immunohistochemical interpretation. The single case without an identified somatic pathogenic variant was MSI-low and initially showed loss of MSH6 by immunohistochemistry but with extremely limited internal control staining. On review, MSH6 immunohistochemistry was reclassified as equivocal, and repeat staining revealed improved control expression with intact MSH6. Following reclassification of this case, 100% tumors with MMR deficiency by immunohistochemistry had at least 1 confirmed somatic MMR pathogenic variant, and 86% were MSI-high. These results demonstrate that when correctly interpreted immunohistochemistry is a strong surrogate for somatic MMR pathogenic variants and support its use as the frontline MMR biomarker in endometrial cancer for heritable screening, molecular prognostic classification, and immunotherapeutic biomarker testing purposes., Competing Interests: Conflicts of Interest and Source of Funding: Funding support was provided by the UVA Department of Pathology Faculty Research Stipend. The authors have disclosed that they have no significant relationships with, or financial interest in, any commercial companies pertaining to this article., (Copyright © 2023 Wolters Kluwer Health, Inc. All rights reserved.)

The anti-programmed cell death (PD-1) checkpoint inhibitor pembrolizumab is approved for the treatment of cervical carcinoma with a programmed cell death-ligand 1 (PD-L1) Combined Positive Score (CPS) of ≥1. We assessed interobserver agreement in cervical carcinoma PD-L1 CPS to identify whether it may affect patient selection for immunotherapeutic candidacy. Twenty-nine cervical carcinomas were stained for PD-L1 (Dako 22C3), and slides were interpreted by 5 subspecialty-trained gynecologic pathologists with experience reading PD-L1 immunohistochemistry. Expression was scored using CPS and read out as positive (≥1) or negative (<1); in positive cases, a final score was assigned (1 to 100). There was consensus agreement across all 5 pathologists for 90% (26/29) (Fleiss Kappa value for interobserver agreement: 0.799). The 3 cases with disagreement were composed of 2 squamous cell carcinomas and 1 small cell carcinoma. Of the 26 with unanimous agreement, 88% (23/26) were positive and 12% (3/26) were negative. All (16/16) pure squamous cell carcinomas with full consensus were interpreted as positive, whereas tumors with glandular components were commonly consensus negative (33%, 3/9); this difference was significant ( P =0.037). Disagreements were attributable to low CPS versus negative reads (2 cases) and difficulty discerning glandular involvement from pushing invasion (1 case). In summary, experienced gynecologic pathologists showed substantial interobserver agreement in the interpretation of PD-L1 CPS at the Food and Drug Administration-approved treatment threshold, with the majority of tumors being classified as positive. Pure squamous histology was strongly associated with a consensus-positive read, whereas a subset of tumors with glandular differentiation was negative by all readers. Disagreements occurred in tumors with low versus negative CPS values and in the setting of limited invasion., Competing Interests: Conflicts of Interest and Source of Funding: The authors have disclosed that they have no significant relationships with, or financial interest in, any commercial companies pertaining to this article., (Copyright © 2023 Wolters Kluwer Health, Inc. All rights reserved.)

Mesonephric-like endometrial carcinoma is a rare but frequently misclassified and aggressive malignancy. KRAS mutations, limited estrogen receptor (ER) expression, and TTF-1, GATA3, and luminal CD10 expression are described in these tumors, but an immunohistochemistry-based screening approach has not been studied. We assessed 300 endometrial carcinomas/carcinosarcomas to ascertain the specificity of TTF-1/GATA3/luminal CD10 expression with or without ER staining for this diagnosis. Next-generation sequencing and morphologic review were performed on screen-positive cases. In all, 3% (9/300) were TTF-1+; 2 coexpressed GATA3. No cases expressed luminal CD10 or GATA3 in isolation. Two TTF-1+/ER- cases, one of which was also GATA3+, were reclassified as mesonephric-like based on morphology and molecular results (KRAS mutations without mismatch repair deficiency, TP53 mutations, or PTEN mutations): these represented 0.7% of all cases (2/300). The reclassified cases were originally diagnosed as grade 1 and 2 endometrioid carcinoma, and the latter had pulmonary metastases and pelvic recurrences. Six TTF-1+ cases retained their original serous (3) and endometrioid (3) diagnoses; 1 was reclassified as dedifferentiated. All had negative or low ER. KRAS mutations were identified in 4 TTF-1+ non-mesonephric-like cases, including 1 serous and 1 grade 3 endometrioid with p53 abnormalities, 1 mismatch repair-deficient endometrioid with a complex molecular profile, and 1 endometrioid with mucinous differentiation. These findings suggest that TTF-1 and ER are good first-line screens for mesonephric-like carcinoma, but caution that a TTF-1+/ER- immunoprofile is not specific, even in the setting of KRAS mutations. A final diagnosis of mesonephric-like carcinoma requires integration of morphologic and immunohistochemical features, with molecular support when relevant., Competing Interests: Conflicts of Interest and Source of Funding: Supported by the UVA Department of Obstetrics & Gynecologic Peyton Taylor Endowed Research Scholarship Fund as well as the UVA Department of Pathology and Stanford Department of Pathology Faculty Research Funds. The authors have disclosed that they have no significant relationships with, or financial interest in, any commercial companies pertaining to this article., (Copyright © 2022 Wolters Kluwer Health, Inc. All rights reserved.)

Uterine carcinosarcoma (UCS) is an aggressive malignancy with few treatment options. A recent clinical trial has shown an increase in progression-free survival in patients with human epidermal growth factor receptor 2 (HER2)-positive serous endometrial carcinomas treated with anti-HER2-targeted therapies. Few studies have evaluated HER2 expression/amplification in UCS. Similar to serous endometrial carcinoma, the majority of UCS have TP53 mutations and a serous epithelial component, suggesting that UCS may show similar rates of HER2 positivity and therapeutic response. Therefore, we evaluated HER2 expression/amplification in a cohort of UCS over a 5-year period. HER2 immunohistochemistry (IHC) and chromogenic in situ hybridization were performed on tissue microarray and whole tissue sections and scored according to the most recent clinical trial recommendations. Three of 48 UCS (6%) had strong (3+) HER2 IHC expression, and 3 cases (6%) were equivocal (2+). Seven cases (15%) had HER2 amplification by chromogenic in situ hybridization, including all 3 with overexpression and 2 that were equivocal by IHC. Mismatch repair (MMR) protein, p53, and programmed cell death-ligand 1 (PD-L1) expression status was obtained from prior whole section analyses. All HER2-positive cases had a serous morphology and aberrant p53 expression. Only minimal PD-L1 expression was seen in the HER2-positive cases, and none had MMR loss. A subset of UCS with serous morphology have overexpression and/or amplification of HER2, which may predict response to HER2-targeted therapies. HER2-positive UCS may be less susceptible to immune checkpoint inhibition as they uncommonly show MMR deficiency and/or strong PD-L1 expression. Thus, HER2-targeted therapies could be of clinical utility in a subset of UCS without other adjuvant treatment options., Competing Interests: Conflicts of Interest and Source of Funding: The authors have disclosed that they have no significant relationships with, or financial interest in, any commercial companies pertaining to this article., (Copyright © 2022 Wolters Kluwer Health, Inc. All rights reserved.)

Uterine sarcomas with variable CD34 and S100 expression represent an emerging class of tumor in the female genital tract which commonly presents in the endocervix of premenopausal women. Initial molecular characterization identified NTRK1 and NTRK3 gene fusions as oncogenic drivers in these tumors; however, the repertoire of genetic alterations is likely more diverse given the recent discovery of PDGFB and RET gene fusions in similarly described tumors. Importantly, these fusion events lead to the aberrant activation of kinases that are potentially therapeutically targetable; therefore, recognizing this class of tumor becomes critical for initiating the molecular testing required for an accurate diagnosis and identification of clinically actionable fusions. Here, we report our institutional experience with 12 cases of uterine spindle cell sarcomas harboring kinase-related fusions. Patients ranged from 21 to 80 years old (median, 38 y) and presented either asymptomatically or with pelvic pain and/or uterine bleeding. Eleven (92%; 11/12) tumors were localized to the cervix and 1 (8%; 1/12) tumor was localized in the anterior fundus of the uterine corpus. Tumors ranged in size from 1.5 to 15.0 cm (median, 6.0 cm) and were histologically characterized by a moderately cellular, infiltrative proliferation of spindle cells with features of benign gland entrapment, stromal collagen deposition, perivascular hyalinization, occasionally myxoid stroma, a lymphocytic infiltrate, occasional nuclear pseudoinclusions, and a pseudophyllodes architecture. RNA-sequencing identified NTRK1 (8/12), NTRK3 (1/12), and PDGFB (2/12) gene fusions, which have been previously implicated in this tumor class, as well as a novel FGFR1-TACC1 (1/12) fusion. All tumors in this cohort showed coexpression of CD34 and S100 by immunohistochemistry except for those tumors with PDGFB fusions which showed solely CD34 expression. Of the 10 surgically resected tumors with follow-up, outcomes best correlated with the stage of disease. One of 4 patients with stage IA tumors (1/4) had recurrences, half of the stage IB (2/4) tumors had recurrences and all of the stage IIB tumors (2/2) had recurrences and died of disease. Future studies are still required to better understand the spectrum of genetic alterations as well as evaluate the efficacy of targeted kinase inhibitors in this class of tumor., Competing Interests: Conflicts of Interest and Source of Funding: The authors have disclosed that they have no significant relationships with, or financial interest in, any commercial companies pertaining to this article., (Copyright © 2021 Wolters Kluwer Health, Inc. All rights reserved.)

Suppression of the immune system is intimately linked to the development and progression of malignancy, and immune modulating treatment options have shown promise in a variety of tumor types, including some triple-negative breast cancers (TNBC). The most dramatic therapeutic success has been seen with immune checkpoint inhibitors targeting programmed cell death protein 1 (PD-1) and its ligand, PD-L1. Difficulty remains, however, in appropriate patient selection for treatment, as many PD-L1-positive cancers fail to show durable responses to PD-1/PD-L1 inhibition. Checkpoint inhibitor targeting of the adaptive immune response relies on the presence of major histocompatibility complex (MHC) class I molecules on the tumor cell surface for tumor antigen presentation. MHC class I loss has been previously described in breast cancer and represents a putative mechanism of immunotherapeutic resistance in this tumor type. One hundred seventeen invasive primary breast carcinomas with a range of histologic subtypes were evaluated on tissue microarrays containing formalin-fixed paraffin-embedded tissue. Loss of MHC class I expression was common among breast cancers, with greater than half of cases demonstrating either subclonal or diffuse loss. Fifty-nine percent of TNBC demonstrated loss of MHC class I, including 46% of those meeting the Food and Drug Administration-approved threshold of 1% for tumor-associated immune cell PD-L1 expression. MHC class I loss was particularly common in the apocrine subtype of TNBC (78%). MHC class I's employment as a predictive biomarker should be considered, as its loss may represent a barrier to successful enhancement of the antitumor adaptive immune response by PD-1/PD-L1 inhibition., Competing Interests: Conflicts of Interest and Source of Funding: The authors have disclosed that they have no significant relationships with, or financial interest in, any commercial companies pertaining to this article., (Copyright © 2021 Wolters Kluwer Health, Inc. All rights reserved.)

Tumor cell expression of major histocompatibility complex (MHC) class I is required for antigen presentation and adaptive immune recognition. Absent or diminished MHC class I expression is thought to contribute to immunotherapeutic resistance in some epithelial tumors but has not been previously studied in cervical and vulvar carcinoma. Given that anti-programmed cell death 1 (PD-1) checkpoint inhibition is deployed for programmed cell death ligand 1 (PD-L1)-positive recurrent and metastatic cervical squamous carcinomas, identifying tumors with loss of MHC class I is of clinical interest to optimize the selection of immunotherapeutic candidates. Immunohistochemistry for PD-L1 and MHC class I combined A, B, and C heavy chains (MHC class I) was assessed in 58 human papillomavirus-associated cervical and vulvar lesions, including 27 squamous intraepithelial lesions (SILs) and 31 invasive squamous cell carcinoma (SCC). Although 84% of SCC and 22% of SIL were PD-L1-positive, 35.5% (11/31) of SCC and 18.5% (5/27) of SIL also showed clonal or complete loss of MHC class I. Loss of MHC class I expression was more common in PD-L1-positive (10/26, 38%) versus PD-L1-negative SCC (1/5, 20%). In summary, over one third of human papillomavirus-associated cervical and vulvar SCC show clonal or complete loss of MHC class I expression, including many PD-L1-positive cases. This suggests that the efficacy of checkpoint inhibitors targeting the PD-1/PD-L1 axis may be limited in a subset of cervical and vulvar squamous neoplasms due to an impaired ability to engage with the adaptive immune system related to loss of MHC class I expression.

Immunohistochemistry (IHC) for mismatch repair (MMR) proteins is recommended in endometrial carcinomas as a screening test for Lynch syndrome, and mismatch repair deficiency (MMRd) is reported in ∼30% of cases. However, few studies have evaluated the rate of MMR loss in uterine carcinosarcomas. A 5-year retrospective database search of uterine carcinosarcomas was performed at 3 academic institutions. The histologic diagnoses, type of carcinoma present, and MMR IHC interpretations were confirmed by a gynecologic pathologist. One hundred three cases of uterine carcinosarcomas with available MMR IHC results were identified. Ninety-nine cases (96%) showed intact expression and 4 cases (4%) showed loss of MLH1/PMS2. All MMRd carcinosarcomas identified in this series had an endometrioid carcinomatous component and wild-type p53 expression. In contrast, the majority of MMR intact carcinosarcomas had a serous morphology and aberrant p53 expression. Three additional cases initially diagnosed as carcinosarcoma also revealed MMRd; however, given the lack of clear mesenchymal differentiation, these cases were reclassified as dedifferentiated endometrial carcinomas and were subsequently excluded from the carcinosarcoma category. No cases of Lynch syndrome were identified among carcinosarcoma patients, as all 4 MMRd cases were due to somatic MLH1 hypermethylation. In summary, we found that the rate of MMRd is markedly lower in uterine carcinosarcoma when compared with endometrial carcinoma. In the setting of MMR loss, a diagnosis of dedifferentiated carcinoma should be considered as almost half of the MMRd tumors which were called carcinosarcomas initially were reclassified as dedifferentiated on review. However, given the interobserver variability in the classification of carcinosarcoma versus dedifferentiated carcinoma a universal screening approach that includes uterine carcinosarcoma is still recommended.

Immunotherapies targeting the PD-1/PD-L1 checkpoint axis are of growing interest for the treatment of mesenchymal neoplasms. However, PD-L1 expression and tumor-associated lymphocytes have not been well-investigated in uterine smooth muscle tumors. Forty-nine uterine smooth muscle tumors (23 leiomyosarcomas, 8 smooth muscle tumors of uncertain malignant potential [STUMP], 7 atypical leiomyomas, and 11 benign leiomyomas) were evaluated for tumoral and tumor-associated immune PD-L1 expression and tumor-associated T-cell infiltration. ALK immunohistochemistry was performed to exclude inflammatory myofibroblastic tumors. Tumor PD-L1 expression was seen in 70% of leiomyosarcomas and 14% of atypical leiomyomas; no cases of STUMP or benign leiomyoma demonstrated tumoral PD-L1. PD-L1 positivity was seen in tumor-associated immune cells in 78% of leiomyosarcomas, 25% of STUMP, no cases of atypical leiomyomas, and 9% of benign leiomyomas. Of the 23 leiomyosarcomas, 15 (65%) had a combined positive score ≥1, while of the 26 other uterine smooth muscle tumors, only 2 (8%) had a combined positive score ≥1. Tumor-associated CD8+ cells were highest among leiomyosarcomas (mean: 87/high-power fields vs. 17/high-power fields for nonleiomyosarcomas), and were significantly associated with PD-L1 expression. One PD-L1, CD8-enriched leiomyosarcoma showed an ALK overexpression suggesting possible classification as inflammatory myofibroblastic tumor, but otherwise lacked morphologic features of this entity. Leiomyosarcomas demonstrate significantly higher PD-L1 expression and cytotoxic T-cell infiltration when compared with other uterine smooth muscle tumors. These data suggest the possibility that treatment with targeted immunotherapy may be appropriate in a selected population of patients with leiomyosarcoma and, potentially, in related tumors bearing ALK rearrangements.

Lynch syndrome (LS) is defined by germline mutations in DNA mismatch repair (MMR) genes, and affected patients are at high risk for multiple cancers. Reflexive testing for MMR protein loss by immunohistochemistry (IHC) is currently only recommended for colorectal and endometrial cancers, although upper tract urothelial carcinoma (UTUC) is the third-most common malignancy in patients with LS. To study the suitability of universal MMR IHC screening for UTUC, we investigated MMR expression and microsatellite status in UTUC in comparison to bladder UC (BUC), and evaluated the clinicopathologic features of UTUC. We found that 9% of UTUC showed MMR IHC loss (8 MSH6 alone; 1 MSH2 and MSH6; 1 MLH1 and PMS2; n=117) compared with 1% of BUC (1 MSH6 alone; n=160) (P=0.001). Of these, 4/10 (40%) of UTUC (3% overall; 3 MSH6 alone; 1 MLH1 and PMS2) and none (0%) of BUC had high microsatellite instability on molecular testing (P=0.03). The only predictive clinicopathologic feature for MMR loss was a personal history of colorectal cancer (P=0.0003). However, UTUC presents at a similar age to colon carcinoma in LS and thus UTUC may be the sentinel event in some patients. Combining our results with those of other studies suggests that 1% to 3% of all UTUC cases may represent LS-associated carcinoma. LS accounts for 2% to 6% of both colorectal and endometrial cancers. As LS likely accounts for a similar percentage of UTUC, we suggest that reflexive MMR IHC screening followed by microsatellite instability testing be included in diagnostic guidelines for all UTUC.

The immune regulatory enzyme indoleamine-2,3-dioxygenase (IDO-1) suppresses T cell responses and may reduce efficacy of therapies targeting immune checkpoints such as programmed death receptor-1/programmed death ligand-1 (PD-1/PD-L1). Early phase clinical trials combining IDO-1 and PD-1/PD-L1 inhibitors have shown some promise in non-small cell lung cancers (NSCLCs). However, the coexpression of IDO-1 and PD-L1 has not been thoroughly investigated, and the potential for IDO-1 immunohistochemical expression as a therapeutic biomarker is unknown. One hundred two cases of NSCLC (51 adenocarcinomas, 9 adenosquamous carcinomas, and 42 squamous cell carcinomas) were evaluated for IDO-1 and PD-L1 expression by immunohistochemistry. IDO-1 expression was identified in 43% of NSCLC (42% of adenocarcinomas, 44% of adenosquamous carcinomas, and 43% of squamous cell carcinomas). Coexpression with PD-L1 (≥1%) was common (27% overall; 27% of adenocarcinomas, 33% of adenosquamous carcinomas, and 26% of squamous cell carcinomas). A smaller population of tumors showed isolated PD-L1 (25% overall; 16% of adenocarcinomas, 44% of adenosquamous carcinomas, and 33% of squamous cell carcinomas) or IDO-1 expression (15% overall; 14% of adenocarcinomas, 11% of adenosquamous carcinomas, and 17% of squamous cell carcinomas). In summary, IDO-1 is commonly expressed by NSCLC, and its frequent coexpression with PD-L1 may account for the increased efficacy seen with dual blockade of PD-1/PD-L1 and IDO in clinical studies. IDO-1 immunohistochemistry may be a useful biomarker for selection of patients who could benefit from dual-agent therapy and should be evaluated in prospective clinical trials using PD-1/PD-L1 and IDO inhibitors.

Mismatch repair (MMR) deficiency in solid tumors has recently been linked to susceptibility to immunotherapies targeting the programmed cell death-1 (PD-1)/programmed cell death-1 ligand (PD-L1) axis. Loss of MMR proteins has been shown to correlate with tumoral PD-L1 expression in colorectal and endometrial carcinomas, but the association between expression of MMR proteins and PD-L1 has not previously been studied in breast carcinoma, where MMR deficiency is less common. We assessed the relationship between PD-L1 and MMR protein expression by immunohistochemistry in 245 primary and 40 metastatic breast carcinomas. Tumoral staining for PD-L1 was positive in 12% of all cases, including 32% of triple-negative cancers. MMR deficiency was observed in 0.04% of breast cancers; the single MMR-deficient case was a high-grade, triple-negative ductal carcinoma which showed dual loss of MLH1 and PMS2 proteins and expressed PD-L1. Two ER carcinomas initially were scored with MMR protein loss in tissue microarray format but were subsequently shown to be MMR-intact on whole sections. Analysis of MMR gene mutation in The Cancer Genome Atlas corroborates low frequency of MMR deficiency for invasive breast cancer. MMR protein expression is therefore unlikely to show utility as a screen for immunotherapeutic vulnerability in this tumor type, and may provoke unwarranted genetic testing in patients unlikely to have a heritable cancer syndrome. PD-L1 may be a more clinically relevant biomarker for anti-PD-1/PD-L1 therapies in this setting.

Cervical low-grade squamous intraepithelial lesions (LSIL) (aka cervical intraepithelial neoplasia, grade 1 [CIN1]) can present considerable diagnostic challenges and are associated with poor interobserver reproducibility and overdiagnosis. Furthermore, ancillary studies such as p16 immunohistochemistry have shown little utility in resolving the LSIL versus negative/reactive differential. Human papillomavirus (HPV) RNA in situ hybridization (ISH) has shown promise as a diagnostic aid in this setting, but has not been studied in a large case series. We herein investigate high-risk and low-risk HPV RNA ISH in 126 cervical biopsies originally diagnosed as LSIL/CIN1 and compare HPV RNA ISH results to expert-adjudicated morphologic diagnosis to assess whether this assay can help routine cases attain the existing "gold standard" of morphologic consensus diagnosis. We also assess whether this criterion standard can be further improved by integration of HPV RNA ISH results. A consensus diagnosis of intraepithelial lesion (CIN1) was confirmed in 61% of cases, whereas 57% were HPV RNA. HPV-RNA positivity was 84% sensitive and 86% specific for an expert-adjudicated diagnosis of CIN1. Conversely, consensus diagnosis was 90% sensitive and 78% specific for the presence of HPV RNA. Integrating RNA ISH into morphologic review led to further reclassification of 10% of cases, resulting in 95% sensitivity and 98% specificity of HPV RNA ISH for a CIN1 diagnosis and 98% sensitivity and 92% specificity of CIN1 for the presence of HPV RNA. These findings suggest that judicious use of HPV RNA ISH can improve the accuracy of LSIL/CIN1 diagnosis for morphologically ambiguous cases.

Most prior studies of primary diagnosis in surgical pathology using whole slide imaging (WSI) versus microscopy have focused on specific organ systems or included relatively few cases. The objective of this study was to demonstrate that WSI is noninferior to microscopy for primary diagnosis in surgical pathology. A blinded randomized noninferiority study was conducted across the entire range of surgical pathology cases (biopsies and resections, including hematoxylin and eosin, immunohistochemistry, and special stains) from 4 institutions using the original sign-out diagnosis (baseline diagnosis) as the reference standard. Cases were scanned, converted to WSI and randomized. Sixteen pathologists interpreted cases by microscopy or WSI, followed by a wash-out period of ≥4 weeks, after which cases were read by the same observers using the other modality. Major discordances were identified by an adjudication panel, and the differences between major discordance rates for both microscopy (against the reference standard) and WSI (against the reference standard) were calculated. A total of 1992 cases were included, resulting in 15,925 reads. The major discordance rate with the reference standard diagnosis was 4.9% for WSI and 4.6% for microscopy. The difference between major discordance rates for microscopy and WSI was 0.4% (95% confidence interval, -0.30% to 1.01%). The difference in major discordance rates for WSI and microscopy was highest in endocrine pathology (1.8%), neoplastic kidney pathology (1.5%), urinary bladder pathology (1.3%), and gynecologic pathology (1.2%). Detailed analysis of these cases revealed no instances where interpretation by WSI was consistently inaccurate compared with microscopy for multiple observers. We conclude that WSI is noninferior to microscopy for primary diagnosis in surgical pathology, including biopsies and resections stained with hematoxylin and eosin, immunohistochemistry and special stains. This conclusion is valid across a wide variety of organ systems and specimen types.

Prior work has shown that digital images and microscopic slides can be interpreted with comparable diagnostic accuracy. Although accuracy has been well-validated, the interpretative time for digital images has scarcely been studied and concerns about efficiency remain a major barrier to adoption. We investigated the efficiency of digital pathology when compared with glass slide interpretation in the diagnosis of surgical pathology biopsy and resection specimens. Slides were pulled from 510 surgical pathology cases from 5 organ systems (gastrointestinal, gynecologic, liver, bladder, and brain). Original diagnoses were independently confirmed by 2 validating pathologists. Diagnostic slides were scanned using the Philips IntelliSite Pathology Solution. Each case was assessed independently on digital and optical by 3 reading pathologists, with a ≥6 week washout period between modalities. Reading pathologists recorded assessment times for each modality; digital times included time to load the case. Diagnostic accuracy was determined based on whether a rendered diagnosis differed significantly from the original diagnosis. Statistical analysis was performed to assess for differences in interpretative times across modalities. All 3 reading pathologists showed comparable diagnostic accuracy across optical and digital modalities (mean major discordance rates with original diagnosis: 4.8% vs. 4.4%, respectively). Mean assessment times ranged from 1.2 to 9.1 seconds slower on digital versus optical. The slowest reader showed a significant learning effect during the course of the study so that digital assessment times decreased over time and were comparable with optical times by the end of the series. Organ site and specimen type did not significantly influence differences in interpretative times. In summary, digital image reading times compare favorably relative to glass slides across a variety of organ systems and specimen types. Mean increase in assessment time is 4 seconds/case. This time can be minimized with experience and may be further balanced by the improved ease of electronic chart access allowed by digital slide viewing, as well as quantitative assessments which can be expedited on digital images.

Dysregulated expression of oncogenic types of E6 and E7 is necessary for human papillomavirus (HPV)-driven carcinogenesis. An HPV E6/E7 mRNA in situ hybridization (ISH) assay covering 18 common high-risk types ("HR-RISH," aka HR-HPV RNA18 ISH) has not been extensively studied in the anogenital tract or validated on automated technology. We herein compare HR-RISH to DNA polymerase chain reaction (PCR), p16 immunohistochemistry, and a previously available HPV DNA ISH assay in HPV-related anogenital and head and neck (H&N) neoplasia. A total of 102 squamous intraepithelial lesions (16 CIN1, 25 CIN3, 3 AIN1, 12 AIN3, 9 VIN3)/invasive squamous cell carcinomas (17 cervical, 2 anal, 18 H&N) as well as 10 normal and 15 reactive cervix samples were collected. HR-RISH, DNA ISH, and p16 immunohistochemistry were performed on whole formalin-fixed, paraffin-embedded sections. RNA ISH for 6 low-risk HPV types (LR-RISH) was also performed. RNA and DNA ISH assays used automated systems. HR-HPV PCR was performed on morphology-directed formalin-fixed, paraffin-embedded punches. HR-RISH was ≥97% sensitive for PCR+ and p16+ neoplasia, as well as morphologically defined anogenital high grade squamous intraepithelial lesion/invasive squamous cell carcinoma. HR-RISH was also positive in 78% of anogenital low grade squamous intraepithelial lesion, including 81% of CIN1. Furthermore, a subset of PCR-negative/invalid and p16-negative lesions was positive for HR-RISH. Only 1 problematic reactive cervix sample and no normal cervix samples stained. These results demonstrate that HR-RISH is a robust method for the detection of HR-HPV-related neoplasia and provides insight into HPV pathobiology. Performance meets or exceeds that of existing assays in anogenital and H&N lesions and may play a role in resolving diagnostically challenging CIN1 versus reactive cases.

Tumor expression of programmed cell death ligand 1 (PD-L1) is associated with immune evasion in a variety of malignancies, including a subset of triple-negative breast carcinomas, and may mark cancers as susceptible to PD-1/PD-L1 inhibitor therapies. We herein characterize PD-L1 expression in breast cancers across the full range of histomorphologies and investigate its intratumoral heterogeneity and fidelity across primaries and metastases. A total of 245 primary and 40 metastatic (20 nodal, 20 distant) breast carcinomas were evaluated with PD-L1 immunohistochemistry on tissue microarray. Tumor PD-L1 staining was seen in 12% of all primaries including 32% of triple-negative cancers. Staining was common in ductal cancers with medullary (54%), apocrine (27%), and metaplastic features (40%). However, diffuse (>50%) staining was rare (2% of all cancers and 5% of triple negatives). Immune staining was seen in 29% of all primaries and 61% of triple negatives. Tumor expression of PD-L1 was conserved in 94% of matched primary/metastasis pairs, while immune staining showed fidelity in 71%; the remaining cases acquired PD-L1 immune cell expression in the metastasis. Only half of cases with positive tumor staining showed concordance across all analyzed cores. These data demonstrate that PD-L1 expression is prevalent among high-grade, hormone receptor-negative breast cancers with a range of histomorphologies and shows fidelity between primary and metastatic sites in treatment-naive cancers, although acquisition of immune PD-L1 staining in metastases is not uncommon. There is considerable intratumoral heterogeneity in PD-L1 expression, undermining the suitability of core biopsy in the determination of PD-L1 status. Clinical trials are needed to determine PD-L1 staining thresholds required for therapeutic response, as diffuse staining is rare.

Mismatch repair (MMR)-deficient endometrial carcinomas (ECs) bearing Lynch syndrome (LS)-associated germline mutations or sporadic MLH1 promoter hypermethylation (MLH1hm) are highly immunogenic and may represent excellent candidates for therapies targeting the programmed cell death (PD)/programmed cell death ligand-1 (PD-L1) immune checkpoint pathway. This study evaluates PD-L1 expression in MMR-deficient ECs including LS-associated and MLH1hm cases, in comparison with MMR-intact tumors. Immunohistochemistry for PD-L1/CD274 was performed on 38 MMR-deficient and 29 MMR-intact ECs. Staining was scored in the tumor and the peritumoral immune compartment. The majority of MMR-deficient tumors were PD-L1 positive (53%) in at least a subset of tumor cells. LS-associated tumors were more likely to be PD-L1 positive relative to MLH1hm tumors (70% vs. 33%, P=0.05). Only 10% of MMR-intact ECs demonstrated any tumoral PD-L1 expression; this was significantly lower than was observed in MMR-deficient tumors (P=0.0005). When reviewed by histologic grade, PD-L1 expression remained highest in LS-associated ECs followed by MLH1hm and MMR-intact carcinomas, respectively. The MMR immunohistochemical pattern most uniformly associated with PD-L1 expression was MSH6 loss. Immune PD-L1 expression was seen in 100% of MMR-deficient and 66% of MMR-intact cases. This study represents the first to characterize differences in PD-L1 expression between LS-associated and MLH1hm endometrial cancers. It demonstrates that tumoral PD-L1 expression is more common in LS-associated endometrial cancers relative to MLH1hm and MMR-intact tumors, although sporadic cancers often show PD-L1 positive immune staining. These data suggest that MMR deficiency may be a better predictor of response to PD-1/PD-L1 inhibitor therapy than tumor grade in EC, and that potential benefit may vary based on the molecular mechanism of MMR defects.

Cervical high-grade squamous intraepithelial lesion (CIN2-3) is thought to arise from a distinct population of cells at the squamocolumnar junction (SCJ). Immunohistochemical (IHC) biomarkers that characterize the SCJ phenotype, including CK7, have been proposed as tools to separate the subset of low-grade squamous intraepithelial lesions (LSILs) (CIN1) that will progress to high-grade squamous intraepithelial lesion from the majority of cases, which will resolve without further intervention. We conducted a retrospective study of CK7 IHC on adjudicated CIN1 tissue from women in the placebo arm of the quadrivalent human papillomavirus (HPV) vaccine trials. Tissue sections were stained with CK7 IHC and scored as negative, patchy, gradation (ie, top-down), or full-thickness pattern. Results were assessed for the prediction of future diagnosis of CIN2-3/AIS (eg, CIN2+ progression) along with p16 IHC, antecedent high-grade cytology, and HPV16 status. A total of 517 patients with CIN1 biopsies and complete data were identified, 12% of whom showed CIN2+ progression on follow-up. Full-thickness CK7 staining showed the highest correlation with CIN2+ progression (odds ratio [OR] 2.8, P=0.021) relative to the other risk factors (HPV16: OR 2.0, P=0.035; antecedent high-grade cytology: OR 2.2, P=0.028; p16 IHC: OR 1.5, P=0.16). Inclusion of the gradation/"top-down" CK7 pattern resulted in a less robust association with progression (CIN2+: OR 2.0, P=0.028; CIN3+: OR 1.3, P=0.74). Interobserver variability ranged from slight to substantial and was not contingent on gynecologic pathology training experience (κ=0.7078 for negative/patchy vs. gradation/full thickness; κ=0.5672 for negative/patchy/gradation vs. full thickness). These data support the theory that SCJ-derived LSILs are precursors to a potentially aggressive subset of cervical SILs and that CK7 staining may inform risk stratification for LSIL (CIN1). However, clinical utility is significantly tempered by the relatively low amplitude of the risk increase, interpretative variability, and limitations of colposcopic sampling.

Apocrine carcinomas comprise ∼1% of all breast cancers and are characterized by large cells bearing abundant eosinophilic granular cytoplasm, round nuclei, and prominent nucleoli. They are typically estrogen receptor/progesterone receptor/HER2 negative, making them unresponsive to typical hormonal or HER2-based chemotherapy. However, this subtype of triple-negative breast cancers expresses androgen receptor (AR), a feature not shared by most nonapocrine triple-negative cancers (NA-TNCs). AR therefore represents a potential diagnostic tool and therapeutic target for apocrine breast carcinoma. All pure apocrine carcinomas diagnosed during a 10-year period were reviewed, and clinicopathologic characteristics were compared with a control group of 26 NA-TNC cases. Twenty apocrine carcinomas were identified (∼0.8% of all breast cancers). The mean age at diagnosis was 69.3 years for apocrine carcinomas and 56.7 years for NA-TNC. All apocrine carcinomas and no NA-TNC were AR positive. The proportions of apocrine carcinoma grades varied, with G1 being seen in 15% of patients, G2 in 55%, and G3 in 30%. In contrast, 100% of NA-TNC cases were G3. The majority of apocrine carcinomas presented at low T stage (T1: 70%; T2: 20%; T3: 10%; T4: 0%), whereas NA-TNC cases more often presented at T2 or higher (T1: 46.2%; T2: 30.8%; T3: 11.5%; T4: 11.5%). Thirty percent of apocrine carcinomas and 30.8% of NA-TNCs had nodal metastases at presentation. Apocrine carcinomas had a favorable clinical prognosis, with 80% of patients showing no evidence of disease-related morbidity or mortality (mean follow-up: 45.2 mo). Pure apocrine carcinomas represent a clinicopathologically distinct subgroup of triple-negative breast cancer characterized by AR positivity. When compared with NA-TNC, apocrine carcinomas more often present in older women with lower grade and T stage, a group in which a more conservative treatment regimen is often desired.

Lynch syndrome underlies approximately 5% of endometrial cancers and ∼1% of ovarian cancers. Gynecologic malignancies are often the presenting cancer in these patients. Therefore, there is considerable benefit to identifying these patients and enrolling them and affected family members in surveillance programs for secondary malignancies. The molecular basis for Lynch syndrome is a defect in the DNA mismatch repair (MMR) system. Tumors can be screened for these defects using immunohistochemistry to identify loss of MMR proteins or by enlisting polymerase chain reaction to identify the microsatellite instability that attends dysfunctional MMR. However, diagnostic confirmation of Lynch syndrome requires germline mutational testing. The algorithm for screening endometrial carcinomas for Lynch syndrome remains a subject of debate, with some studies supporting universal screening and others proposing a hybrid approach informed by clinicopathologic features. This review discusses the rationales and relative merits of current Lynch syndrome-screening approaches for endometrial and ovarian cancers and provides pathologists with an informed approach to Lynch syndrome testing in gynecologic cancers. It also addresses the clinical difficulties presented by cases with discordant screening and germline results (Lynch-like cancers) and emphasizes the critical role of strong communication with clinician and genetic counseling colleagues to ensure that the significance of a positive screening test is appropriately conveyed to patients. Finally, it discusses the need for more nuanced cost-effective analyses and the potential role for next-generation sequencing panels in future screening efforts.

The squamocolumnar junction (SCJ) cell population of the uterine cervix is a discrete epithelial area and the putative source of the majority of high-grade squamous intraepithelial lesions (HSIL). The SCJ cells can be identified by immunohistochemical (IHC) stains including cytokeratin 7 (CK7). Others have theorized that an SCJ marker-positive low-grade squamous intraepithelial lesion (LSIL) has a higher risk for future HSIL compared with an SCJ marker-negative LSIL. This study has 2 aims: first, to refine the definition of a positive CK7 immunostaining pattern in cervical lesions, and, second, to test the hypothesis that CK7 positivity in LSIL indicates higher risk for future HSIL, with both questions addressed using a data set with consensus diagnoses. One hundred cases each of LSIL, moderate HSIL (CIN2), and severe HSIL (CIN3) were stained for CK7, with positivity defined as a diffuse cytoplasmic staining pattern (>5 to 6 contiguous cells); all others were considered negative. Using this model, 34% of CIN1, 45% of CIN2, and 60% of CIN3 were CK7 positive. With follow-up, CK7-positive LSILs were more likely to progress to HSIL compared with CK7-negative LSIL (32% vs. 11%, P=0.05), in concordance with the results of other researchers. This study simplifies cervical CK7 IHC grading into a reproducible system and supports the thesis that CK7 positivity in LSIL is associated with increased risk for future HSIL. Larger cohorts using consensus-diagnosed LSIL are needed to confirm these results before CK7 may be considered for clinical validation.

Expanded testing for Lynch syndrome (LS) is increasingly recommended for patients with endometrial carcinomas, and immunohistochemistry (IHC) for tumor loss of mismatch-repair (MMR) protein expression is the most common primary screen. This has led to the recognition of MMR-IHC-deficient cases without identifiable mutations on directed germline sequencing. The clinical implications of such "Lynch-like" (LL) cancers are unclear. We here report the clinicopathologic features of putative familial endometrial carcinoma identified on universal MMR-IHC screening with attention to cases with discordant IHC and germline results. The files of the University of Virginia Pathology Department were retrospectively searched for all MMR-deficient endometrial carcinomas identified on screening. Cases were categorized as likely sporadic (MLH1/PMS2 loss, evidence of MLH1 promoter hypermethylation) or putative LS (PLS) (loss of MSH2/MSH6, MSH6, or PMS2). PLS cases were further subdivided into LS and LL groups on the basis of the presence or absence of a confirmatory mutation by germline testing, and the clinicopathologic features of these cases were compared. A deficiency of ≥1 MMR protein was observed in 31.4% (66/210) of endometrial carcinomas, including 26 PLS cases, 15 of which had germline testing. Directed germline sequencing confirmed LS in 46.7% (7/15); the remaining cases were classified as LL. High-grade and/or biphasic morphology was seen in 42.9% (3/7) of LS and 62.5% (5/8) of LL cases; the remaining cases showed low-grade, conventional endometrioid morphology. High level microsatellite instability was observed in 71.4% (5/7) of LL cases. The majority of cases from both groups (LS: 85.7% [6/7]; LL: 87.5% [7/8]) were low-stage (T1a/T1b). Endometrial carcinoma was the presenting malignancy in 85.7% (6/7) of LS patients and 87.5% (7/8) of LL patients. Family history was suggestive of LS in 28.5% (2/7) of LS patients and 12.5% (1/8) of LL patients. Screening algorithms based on age and cancer history would have failed to identify LS patients in 57.1% (4/7) of cases. Although universal MMR-IHC identifies endometrial carcinoma patients with LS who would have been missed using targeted screening algorithms, it also identifies cancers with discordant IHC and germline results for which the somatic versus germline origin of the MMR defect is unclear. Further study of this LL group is required before drawing definitive conclusions about their familial cancer risk.

Previous studies of p16 immunohistochemistry (IHC) on CIN1 have suggested the likely utility of p16 in stratification of women at risk for subsequent CIN2/3. But those studies had limitations in statistical power, histologic diagnosis, and disease ascertainment. We conducted a retrospective study of p16 IHC on adjudicated CIN1 tissue diagnosed in young women participating in the placebo arm of the quadrivalent human papillomavirus (HPV) vaccine trials. Tissue sections were stained with p16 IHC and hematoxylin and eosin. p16 IHC was scored using LAST criteria, and hematoxylin and eosin-stained sections were reviewed for concordance with the adjudicated diagnosis. p16 IHC, antecedent high-grade cytology, review diagnosis, and HPV16 detection were assessed as independent risk factors for subsequent CIN2/3. Five hundred twenty-four patients with CIN1 biopsies and complete data were identified, 63 (12.0%) of whom developed CIN2/3 in follow-up. p16 positivity (P=0.06), review diagnosis of CIN2/3 (P=0.04), HPV16 positivity (P=0.01), and antecedent high-grade cytology (P=0.02) were (marginally) associated with CIN2/3. In a logistic regression model, the associations with CIN2/3 (vs. other), expressed as odds ratios (95% confidence intervals), were 1.6 (0.91-2.8) for p16, 2.0 (1.0-3.7) for HPV16, and 2.2 (1.1-2.4) for antecedent high-grade cytology. The mean risks for CIN2/3 estimated from the model ranged from 7.6% for negative for all markers to 36.3% for positive for all 3 markers. p16 IHC does not risk stratify CIN1 patients in a manner that would alter recommended management for CIN1. This reinforces the LAST recommendations that p16 should only be used selectively for problematic scenarios, such as CIN2 because of its inherent lack of reproducibility, cases in which one is struggling between CIN1 and CIN2, and benign mimics of CIN3.

Lynch syndrome (LS) is an autosomal dominant inherited disorder caused by germline mutations in DNA mismatch repair (MMR) genes. Mutation carriers are at substantially increased risk of developing cancers of the colorectum and endometrium, among others. Given recent recommendations for universal, cost-effective screening of all patients with newly diagnosed colorectal cancer using MMR protein immunohistochemistry, we evaluated MMR protein expression in a series of endometrial cancers in the general population. A total of 605 consecutive cases of primary endometrial cancer at a single institution (1997 to 2013) were evaluated regardless of age, family history, or histologic features. Evaluation methods consisted of immunohistochemistry for the MMR proteins MLH1, MSH2, MSH6, and PMS2, followed by DNA methylation analysis for cases with MLH1/PMS2 deficiency. Germline mutation testing was performed on a subset of cases. Forty MMR-deficient, nonmethylated endometrial cancers were identified: 3 MLH1/PMS2 and 37 MSH6/MSH2 protein deficiencies. Only 25% occurred in women below 50 years of age (range, 39 to 88 y), 1 of which was in a risk-reducing hysterectomy specimen. Only 15% of patients had a prior history of carcinoma, including only 2 patients with prior colorectal carcinoma. Most (80%) of the endometrial cancers were purely endometrioid; there were 2 mixed endometrioid/mucinous, 1 mucinous, 1 serous, 2 clear cell, and 2 carcinosarcoma cases. When grading was applicable, 40% of the endometrial malignancies were FIGO grade 1, 34% grade 2, and 26% grade 3. Thirteen percent arose in the lower uterine segment, and 23% had tumor-infiltrating lymphocytes. Of the tumors with known germline testing, 41% with a LS-associated germline mutation were not associated with any of the traditional indicators that have been recommended for LS screening (ie, age 50 y or younger, personal/family cancer pedigree that meets Bethesda guideline criteria, presence of MMR-associated tumor morphology, or location in the lower uterine segment). These data suggest that a significant number of LS-associated endometrial carcinomas are missed using clinical, histologic, and locational screening parameters and provide support for universal screening of all newly diagnosed endometrial cancers.

Cytomegalovirus (CMV) can precipitate and exacerbate gastrointestinal (GI) mucosal injury. The gold standard for CMV detection in formalin-fixed, paraffin-embedded (FFPE) tissue is immunohistochemistry (IHC). Although CMV polymerase chain reaction (PCR) on fresh tissue may be a valuable adjunct to IHC, its utility is unknown for FFPE tissues. We therefore evaluated quantitative, real-time CMV PCR in a total of 102 FFPE GI biopsy specimens from 74 patients with a history of hematopoietic stem cell or solid organ transplant, inflammatory bowel disease, human immunodeficiency virus infection, or unspecified colitis. CMV DNA was detected by PCR in 90.9% (30/33) of IHC-positive, 14.5% (8/55) of IHC-negative, and 20.0% (1/5) of IHC-equivocal FFPE tissues. Quantitation of CMV DNA copies normalized to β-globin demonstrated a wide range of values (median 0.276; range, 0.0004 to 144.50). Importantly, 93.3% (14/15) of patients with IHC-positive, active colitis showed no evidence of CMV in matched concurrent, histologically normal biopsies tested by PCR. These results suggest that CMV PCR on FFPE GI biopsies complements IHC and has the potential to identify additional patients who may benefit from anti-CMV therapy.

Cell cycle regulatory protein expression by immunohistochemical assay may have diagnostic utility in the distinction of uterine leiomyosarcoma from leiomyoma variants. p16, p21, p27, and p53 protein expression was evaluated by immunohistochemistry on 44 atypical leiomyomas (mean follow-up, 50.8 mo), 16 leiomyosarcomas (mean follow-up, 29.7 mo), and 8 cellular leiomyomas (mean follow-up, 22.6 mo). Nuclear staining was semiquantitatively scored on 1 representative section per case as negative (0%), focal (>0% to 33%), patchy (>33% to 66%), or diffuse (>66%). In addition, staining intensity was noted as weak, moderate, or strong. Proliferative index was gauged by Ki-67 and PHH3 immunohistochemical staining. One of 35 atypical leiomyoma patients with follow-up data developed an extrauterine recurrence 25.7 months after hysterectomy, whereas a second had intrauterine recurrence 24.9 months after myomectomy. Seven of 8 patients with leiomyosarcoma with follow-up had recurrence within the follow-up period, whereas there were no recurrences in patients with cellular leiomyoma. The Ki-67 proliferation index ranged from 0% to 25% in atypical leiomyoma (mean, 2%) and 6% to 50% in leiomyosarcoma (mean, 25%) with 0% to 10% in cellular leiomyoma (mean, 3%), whereas the PHH3 proliferation index ranged from 0% to 3% in atypical leiomyoma (mean, <1%) and 0% to 10% in leiomyosarcoma (mean, 2%) with 0% to 2% in cellular leiomyoma (mean, <1%). The atypical leiomyoma with extrauterine recurrence was diffusely positive for p21, but showed only weak focal (<33%) staining for all other cell cycle markers. Uterine atypical leiomyomas, cellular leiomyomas, and leiomyosarcomas demonstrate a heterogenous pattern of cell cycle regulatory protein expression. Caution should be exercised in distinguishing leiomyosarcoma from atypical leiomyoma variants on the basis of cell cycle protein expression alone. In our study, cell cycle markers were not useful for predicting recurrence in atypical leiomyoma.

Atypical leiomyoma is a well-described smooth muscle neoplasm of the uterus. Only 1 study has addressed long-term clinical follow-up in a large series, and little is known about the adequacy of treatment by myomectomy. The surgical pathology archives were searched for consecutive cases of uterine atypical leiomyoma from 1992 to 2003. Glass slides were reviewed to confirm the diagnoses, and patient age, treatment modality, and clinical follow-up data were recorded. Fifty-one atypical leiomyomas with available glass slides and clinical follow-up data were identified. Thirty tumors exhibited diffuse, moderately to severely atypical cells, whereas 21 showed atypical cells in a more focal or patchy distribution. Twelve had ischemic-type necrosis. By the highest count method, 37 cases were found to have ≤1 MF/10 HPF, 13 showed 1 to 3 MF/10 HPF, and 1 was nearly entirely necrotic precluding mitotic assessment. Among cases in which adjacent non-neoplastic tissue was well visualized, all were found to have pushing margins (46 cases). The average tumor size was 6.8 cm (median 6.5 cm; range, 0.7 to 14 cm). The average patient age was 42.5 years (median 42 y; range, 21 to 72 y). In all cases, the initial diagnostic procedure was hysterectomy (34) or myomectomy (17). Average follow-up was 42 months (range, 0.3 to 121.8 mo). Of those treated with hysterectomy, 1 had recurrent atypical leiomyoma in the retroperitoneum at 87.5 months, 1 died of other causes, and the remaining 32 (94%) were free of disease. Of the myomectomy group, 82% had no evidence of recurrent disease on follow-up: 2 had residual atypical leiomyoma in the subsequent hysterectomy specimen; and 1 underwent second myomectomy for atypical leiomyoma with 2 subsequent successful pregnancies. Atypical leiomyoma has a low rate of extrauterine, intra-abdominal recurrence (<2%) with a negligible risk for distant metastasis. Patients may be treated by myomectomy alone with successful pregnancy, but should be monitored for local intrauterine residual/recurrent disease.

ProExC expression has been shown to perform similarly to p16 as an aid in the diagnosis of cervical dysplasia but has not been well characterized in head and neck squamous cell carcinomas (SCC). The purpose of this study is to determine whether ProExC performs similarly to p16 as a prognostic marker in oropharyngeal SCC and to evaluate the threshold of ProExC and p16 staining that correlates with survival. ProExC, p16, and human papillomavirus DNA in situ hybridization were performed on tissue microarray (TMA) cores and whole sections from 62 patients with oropharyngeal SCC. Sensitivity and specificity for high-risk HPV and correlation with overall survival (OS), cancer-specific survival (CSS), and time to distant metastasis (TDM) were calculated for ProExC and p16 at different thresholds. ProExC did not prove to be a robust marker. It showed strong correlation with OS at a 66% threshold on TMA cores, but correlation with OS was lost on whole sections. It also exhibited low sensitivity (53.7%) on TMA cores and low specificity on whole sections (65%). ProExC at a 33% threshold exhibited unacceptably low specificity and did not correlate with OS, CSS, or TDM. Sensitivity and specificity of p16 varied predictably with threshold: higher sensitivity and lower specificity with lower thresholds and vice versa for higher thresholds. p16 at a 50% threshold offers a balance between sensitivity and specificity, and correlates with OS, CSS, and TDM on whole sections; correlation with TDM is lost on TMA cores. These findings indicate that ProExC does not perform well enough to be used as a prognostic marker in oropharyngeal SCC. p16 should be used and scored as positive when at least half the tumor is strongly stained.

Leiomyosarcomas (LMSs) constitute approximately one quarter of all sarcomas and are usually defined by morphologic criteria and/or immunoreactivity for actin or desmin. Among high-grade lesions, the distinction from undifferentiated pleomorphic sarcoma (UPS) can be problematic, and previous studies have shown that a significant number of LMS cases may be hiding under the diagnosis of UPS. We recently described 3 novel molecular LMS subtypes that are distributed similarly over LMSs of gyneocologic and non-gyneocologic origins. The group 1 subtype shows an improved disease-specific survival compared with the other 2 groups that is independent of histologic grade. Group 1 comprises approximately 25% of all LMSs, and is defined by a shared pattern of gene expression, a distinct pattern of genomic changes, and reactivity for at least 3 of 5 immunohistochemistry (IHC) markers (smooth muscle gamma actin, calsequestrin 2, human muscle cofilin2, myosin light chain kinase, and sarcolemmal membrane associated protein), as tested on 271 cases of LMS in tissue microarrays. These IHC markers have not been well characterized in non-LMS sarcomas. Here we provide a characterization of these 5 markers across normal tissues, an additional 59 cases of LMS, and a wide range of 565 non-LMS soft tissue tumors from 44 diagnostic categories, with a focus on UPS. When analyzed individually, the 5 markers were found to be expressed in many sarcomas other than LMSs. However, when analyzed by the same criteria used for the recognition of group 1 LMSs, in which a case is scored positive when at least 3 of 5 markers reacted, coordinate expression was seen in significant numbers of cases from only 3 diagnostic groups that included 22% of leiomyomas (n=22), 16% of gastrointestinal stromal tumors (n=43), and 18% of endometrial stromal sarcomas (n=11). In addition, 5% (n=57) of UPSs showed a staining pattern similar to that seen in group 1 LMSs. To further examine the possibility that group 1 LMS constitutes a small part of cases diagnosed as UPS, we examined the expression of the top 500 genes from the group 1 LMS expression signature in 29 UPSs by complementary DNA microarray. Unsupervised hierarchical clustering of 29 UPS expression showed that 2 (7%) had coordinated high levels of expression of genes from the group 1 LMS signature, a rate similar to that seen by IHC analysis. These findings show that group 1 LMS IHC markers smooth muscle gamma actin, calsequestrin 2, human muscle cofilin2, myosin light chain kinase, and sarcolemmal membrane associated protein when coordinately expressed have specificity for a subset of LMS when compared with other sarcomas, and may be useful for the recognition of group 1 LMS cases within cases diagnosed as UPS.

Primary cervical stromal sarcomas are rare neoplasms that have been poorly characterized. We report the clinical, histologic, and immunohistologic features of 3 primary endocervical S100 protein (S100p)-positive and CD34-positive sarcomas, herein designated as fibroblastic malignant peripheral nerve sheath sarcoma (endocervical neurofibrosarcoma), 2 of which occurred in women younger than 35 years of age. All tumors presented as a cervical polyp or mass lesion; 1 extended into the pelvic side wall and vaginal soft tissue. The tumors measured 2.0 to 8.0 cm, and were composed of compact fascicles of spindled cells arranged in herringbone, loose fascicular, or ill-defined storiform patterns. A focal whorled architecture was identified in all 3 tumors, but distinct Antoni A areas and Verocay bodies were absent. Ultrastructural examination in 1 case confirmed the presence of fibrocyte-like differentiation. Strong, diffuse, and in 1 case, patchy S100p expression was seen in all cases; strong and diffuse CD34 expression was also present in all tumors. Adjacent uninvolved endocervical stroma also showed CD34 positivity but expression was much less dramatic than in tumor cells. All other markers of neural, melanocytic, smooth muscle, endometrial stromal, and epithelial differentiation were negative. One of the tumors behaved extremely aggressively with extensive pelvic involvement, resulting in patient death within 16 months of diagnosis; another tumor was associated with pelvic recurrence 13 months after diagnosis; and the third tumor had an indolent course with no evidence of recurrence at 33 months after complete excision and local radiotherapy, although follow-up was limited. Review of large numbers of mesenchymal tumors in the uterus did not show similar tumors. Endocervical neurofibrosarcoma should be distinguished from solitary fibrous tumor, endometrial stromal sarcoma, leiomyosarcoma, melanoma, and other spindle cell neoplasms. The prominent fibroblastic endoneurial-like differentiation seen in this peripheral nerve sheath tumor may be related to the presence of a rich mucosal stromal fibrocyte network in the endocervix.