Validation of a Transcriptome-Based Assay for Classifying Cancers of Unknown Primary Origin

INTRODUCTION

Cancers assume a variety of distinct histologies, and may originate from a myriad of sites including solid organs, hematopoietic cells, and connective tissue. Clinical decision-making based on consensus guidelines such as the National Comprehensive Cancer Network (NCCN) is often predicated on a specific histologic and anatomic diagnosis, supported by clinical features and pathologist interpretation of morphology and immunohistochemical (IHC) staining patterns. However, in patients with nonspecific morphologic and IHC findings-in addition to ambiguous clinical presentations such as recurrence versus new primary-a definitive diagnosis may not be possible, resulting in the patient being categorized as having a cancer of unknown primary (CUP). Therapeutic options and clinical outcomes are poor for patients with CUP, with a median survival of 8-11 months.

METHODS

Here, we describe and validate the Tempus Tumor Origin (Tempus TO) assay, an RNA-sequencing-based machine learning classifier capable of discriminating between 68 clinically relevant cancer subtypes. Model accuracy was assessed using primary and/or metastatic samples with known subtype.

RESULTS

We show that the Tempus TO model is 91% accurate when assessed on both a retrospectively held out cohort and a set of samples sequenced after model freeze that collectively contained 9210 total samples with known diagnoses. When evaluated on a cohort of CUPs, the model recapitulated established associations between genomic alterations and cancer subtype.

DISCUSSION

Combining diagnostic prediction tests (e.g., Tempus TO) with sequencing-based variant reporting (e.g., Tempus xT) may expand therapeutic options for patients with cancers of unknown primary or uncertain histology.

Adrenal gland cytology reporting: a multi-institutional proposal for a standardized reporting system

BACKGROUND

With the development of new technologies and the changing patient profiles, cytopathology departments receive increasing numbers of adrenal gland cytology specimens. In this study, the authors analyzed archival adrenal gland cytology cases and attempted to implement a diagnostic reporting system.

DESIGN

Retrospective electronic medical record search was performed for adrenal gland cytology specimens in seven tertiary care centers. The cytology diagnoses were grouped in 7 categories: nondiagnostic, nonneoplastic, benign adrenal cortical elements (BACE), primary neoplasm of noncortical origin (NONC), atypia of undetermined significance (AUS), suspicious for malignancy (SM), and malignant (MAL). If available, histopathology results of concurrent and/or follow-up biopsies and/or resections were documented.

RESULTS

A total of 473 adrenal gland cytology cases were included. BACE cases comprised 21.8%, whereas MAL cases were 57.5% of all cases. For BACE and MAL categories, there were 100% and 98.9% correlation, respectively, in the cases with histopathology follow-up. Six of 10 NONC cases had histopathology diagnoses and there were 3 pheochromocytomas and 3 schwannomas. Twenty-one AUS cases had histology follow-up and 10 (47.6%) of them were malignant. Six cases of SM had histopathology follow-up, and all of them were malignant on the follow-up.

CONCLUSIONS

The authors propose a 7-tier diagnostic scheme for adrenal gland cytology. The risk of malignancy was 98.9% in MAL cases (87/88) in the cohort. The only case with discordance was reported as "adrenal cortical adenoma with marked atypia"' on resection. There was no difference between endoscopic ultrasound-guided and percutaneous methods. Further studies are needed to validate and make this approach universal.

Cytologic and Histologic Features of Pleomorphic Undifferentiated Sarcoma Arising in a Hybrid Hemosiderotic Fibrolipomatous Tumor and Pleomorphic Hyalinizing Angiectatic Tumor: Report of an Unusual Case with a Literature Review

BACKGROUND

Pleomorphic hyalinizing angiectatic tumor (PHAT) and hemosiderotic fibrolipomatous tumor (HFLT) are low-grade neoplasms that share clinicopathologic features and recurring translocation t(1;10)(p22;q24) involving the TGFBR3 and MGEA5 genes. Coexistence of these tumors with a high-grade sarcoma is exceedingly rare and the cytologic features have not been widely described in the literature.

CASE

A 55-year-old female presented with a soft tissue tumor on the dorsum of the foot. Cytologic smears and corresponding core biopsies were composed of a population of markedly pleomorphic spindle cells seen singly and in loose clusters within a myxofibrous matrix and infiltrating fat, with coarse chromatin, prominent nucleoli, irregular nuclear contours and delicate to vacuolated cytoplasm. Intracytoplasmic hemosiderin granules and rare intranuclear cytoplasmic pseudoinclusions were identified. The histologic features of the excisional biopsy mirrored those of the cytologic preparations, but also demonstrated cellular foci of higher-grade sarcoma composed of markedly pleomorphic tumor cells with large vesicular nuclei and prominent nucleoli, exhibiting a mitotic index of 12 mitotic figures per 10 high-powered fields.

CONCLUSION

While HFLT/PHAT generally can be managed by wide local excision, it is important to be aware of their capacity to harbor higher-grade lesions with metastatic potential which may require more radical surgical excision.

Atypical urine cytology and the Johns Hopkins Hospital template: the University of Chicago experience

INTRODUCTION

The rate of atypical diagnoses in urine cytology can be high depending on the screening population. Unlike thyroid and cervical cytology, there is a lack of standardized criteria to stratify them into more clinically meaningful categories.

MATERIALS AND METHODS

A set of diagnostic criteria described by Johns Hopkins Hospital (JHH) provided a tool to divide atypical urine specimens into those that were low risk and those likely to be predictive of high-grade urothelial carcinoma (HGUC). In this study, the JHH template was applied to a cohort of atypical urine cytology specimens from the University of Chicago (U of C) to compare it to existing U of C terminology and determine whether it should be formally adopted.

RESULTS

Sixty-eight percent of patients classified as atypical urothelial cells, favor high-grade lesion (AUC-H) were diagnosed with HGUC during the study. Correlation was noted between the JHH diagnostic categories and the U of C diagnostic categories, with 49% of patients reclassified as AUC-H being diagnosed with atypical urothelial cells, suspicious for neoplasia and 83% of cases of patients reclassified as atypical urothelial cells of unknown significance being diagnosed as atypical, urothelial cells. The JHH category of AUC-H had a higher positive predictive value for HGUC than the U of C category atypical urothelial cells, suspicious for neoplasm did (69% versus 58%, p = 0.0087). Unlike the JHH study, AUC-H showed higher correlation with HGUC in the hematuria group (90%) than in the surveillance group (66%).

CONCLUSIONS

JHH criteria demonstrated a higher rate of predicting HGUC than U of C diagnostic categories, supporting the adoption of these criteria at U of C.

The "drunken honeycomb" feature of pulmonary mucinous adenocarcinoma: a diagnostic pitfall of bronchial brushing cytology

Pulmonary mucinous adenocarcinoma (PMA) is the terminology recently proposed in the new International Association for the Study of Lung Cancer/American Thoracic Society/European Respiratory Society (IASLC/ATS/ERS) International Multidisciplinary Classification of Lung Adenocarcinoma Guidelines for most tumors previously classified as mucinous bronchioloalveolar carcinomas (mBACs). PMA is histologically characterized by lepidic growth and at least some degree of invasive growth of goblet or columnar neoplastic cells with abundant intracytoplasmic mucin. We report here the cytologic features of PMA in a bronchial brushing specimen. The patient is an 84-year-old woman with a persistent dense consolidation in the right middle lobe of the lung found on non-contrast computed tomography (CT) scan. Bronchial brushing smears showed a monotonous population of columnar neoplastic cells forming "drunken honeycomb"-like cell clusters. The neoplastic cells displayed inconspicuous cytologic atypia. The concurrent transbronchial tissue biopsy and the resection specimen confirmed the diagnosis of PMA. Due to the bland nuclear features, the neoplastic cells in the bronchial brushing specimen were interpreted as benign at the time of the initial diagnosis, demonstrating a diagnostic pitfall of bronchial brushing cytology. A high index of suspicion is recommended when a lung lesion with "drunken honeycomb"-like cell clusters is encountered in bronchial brushing specimens. The review of the literature regarding the recently designated PMA is presented.

The protein ICP0 of herpes simplex virus type 1 is targeted to nucleoli of infected cells

This study describes the nucleolar localization of the viral protein ICP0 of herpes simplex virus type 1. We show that the RING finger domain of ICP0 is essential for ICP0 to localize in nucleoli of transfected and 4 hour-infected cells. ICP0 forms particular intranucleolar domains that do not correspond to any known nucleolar domains. This distribution was confirmed by immunoblots performed on fractionated infected cells. Quantitative RT-PCR experiments indicated that ICP0 did not increase the transcription from the RNA polymerase I (Pol I) promoter in transfected cells, an effect opposite to that observed on viral and cellular Pol II promoters. Nucleoli are thus, after PML bodies and centromeres, a novel nuclear structure targeted by ICP0.