Oropharyngeal Cancer Staging Health Record Extraction Using Artificial Intelligence.

Importance: Accurate, timely, and cost-effective methods for staging oropharyngeal cancers are crucial for patient prognosis and treatment decisions, but staging documentation is often inaccurate or incomplete. With the emergence of artificial intelligence in medicine, data abstraction may be associated with reduced costs but increased efficiency and accuracy of cancer staging. Objective: To evaluate an algorithm using an artificial intelligence engine capable of extracting essential information from medical records of patients with oropharyngeal cancer and assigning tumor, nodal, and metastatic stages according to American Joint Committee on Cancer eighth edition guidelines. Design, Setting, and Participants: This retrospective diagnostic study was conducted among a convenience sample of 806 patients with oropharyngeal squamous cell carcinoma. Medical records of patients with staged oropharyngeal squamous cell carcinomas who presented to a single tertiary care center between January 1, 2010, and August 1, 2020, were reviewed. A ground truth cancer stage dataset and comprehensive staging rule book consisting of 135 rules encompassing p16 status, tumor, and nodal and metastatic stage were developed. Subsequently, 4 distinct models were trained: model T (entity relationship extraction) for anatomical location and invasion state, model S (numerical extraction) for lesion size, model M (sequential classification) for metastasis detection, and a p16 model for p16 status. For validation, results were compared against ground truth established by expert reviewers, and accuracy was reported. Data were analyzed from March to November 2023. Main Outcomes and Measures: The accuracy of algorithm cancer stages was compared with ground truth. Results: Among 806 patients with oropharyngeal cancer (mean [SD] age, 63.6 [10.6] years; 651 males [80.8%]), 421 patients (52.2%) were positive for human papillomavirus. The artificial intelligence engine achieved accuracies of 55.9% (95% CI, 52.5%-59.3%) for tumor, 56.0% (95% CI, 52.5%-59.4%) for nodal, and 87.6% (95% CI, 85.1%-89.7%) for metastatic stages and 92.1% (95% CI, 88.5%-94.6%) for p16 status. Differentiation between localized (stages 1-2) and advanced (stages 3-4) cancers achieved 80.7% (95% CI, 77.8%-83.2%) accuracy. Conclusion and Relevance: This study found that tumor and nodal staging accuracies were fair to good and excellent for metastatic stage and p16 status, with clinical relevance in assigning optimal treatment and reducing toxic effect exposures. Further model refinement and external validation with electronic health records at different institutions are necessary to improve algorithm accuracy and clinical applicability.

Lymph node yield: Impact on oncologic outcomes in oral cavity cancer.

BACKGROUND: Lymph node metastases are associated with poor prognosis in oral cavity squamous cell carcinoma (OCSCC). In other cancers, clinical guidelines on the number of lymph nodes removed during primary surgery, lymph node yield (LNY), exist. Here, we evaluated the prognostic capacity of LNY on regional failure, locoregional recurrence, and disease-free survival (DFS) in patients with OCSCC treated by primary neck surgery. METHODS: This retrospective cohort study took place at Sunnybrook Health Sciences Centre in Toronto, Canada and involved a chart review of all adult patients with treatment-naive OCSCC undergoing primary neck dissection. For each outcome, we first used the maximally selected rank statistics and an optimism-corrected concordance to identify an optimal threshold of LNY. We then used a multivariable Cox proportional hazards model to assess the association between high LNY (>threshold) and each outcome. RESULTS: Among the 579 patients with OCSCC receiving primary neck dissection, 61.7% (n = 357) were male with a mean age of 62.9 years (standard deviation: 13.1) at cancer diagnosis. When adjusting for sociodemographic and clinical factors, LNY >15 was significantly associated with improved DFS (adjusted HR [aHR]: 0.73, 95% CI: 0.54-0.98), locoregional recurrence (aHR: 0.68, 95% CI: 0.49-0.95), and regional failure (aHR: 0.61, 95% CI: 0.39-0.93). CONCLUSIONS: Our study findings suggested high LNY to be a strong independent predictor of various patient-level quality of surgical care metrics. The optimal LNY we found (15) was lower than the conventionally recommended (18), which calls for further research to establish validity in practice.

Development and validation of a novel TNM staging N-classification of oral cavity squamous cell carcinoma.

BACKGROUND: For oral cavity squamous cell carcinoma (OSCC), extent of extranodal extension (ENE) (minor, ≤2 mm; major, >2 mm) is differentially prognostic, whereas limitations exist with the 8th edition of American Joint Committee on Cancer/International Union Against Cancer TNM N-classification (TNM-8-N). METHODS: Resected OSCC patients at four centers were included and extent of ENE was recorded. Thresholds for optimal overall survival (OS) discrimination of lymph node (LN) features were established. After dividing into training and validation sets, two new N-classifications were created using 1) recursive partitioning analysis (RPA), and 2) adjusted hazard ratios (aHRs) and were ranked against TNM-8-N and two published proposals. RESULTS: A total of 1460 patients were included (pN0: 696; pN+: 764). Of the pN+ cases, 135 (18%) had bilateral/contralateral LNs; 126 (17%) and 244 (32%) had minor and major ENE, and two (0.3%) had LN(s) >6 cm without ENE (N3a). LN number (1 and >1 vs. 0: aHRs, 1.92 [95% confidence interval (CI), 1.44-2.55] and 3.21 [95% CI, 2.44-4.22]), size (>3 vs. ≤3 cm: aHR, 1.88 [95% CI, 1.44-2.45]), and ENE extent (major vs. minor: aHR, 1.40 [95% CI, 1.05-1.87]) were associated with OS, whereas presence of contralateral LNs was not (aHR, 1.05 [95% CI, 0.81-1.36]). The aHR proposal provided optimal performance with these changes to TNM-8-N: 1) stratification of ENE extent, 2) elimination of N2c and 6-cm threshold, and 3) stratification of N2b by 3 cm threshold. CONCLUSION: A new N-classification improved staging performance compared to TNM-8-N, by stratifying by ENE extent, eliminating the old N2c category and the 6 cm threshold, and by stratifying multiple nodes by size.

Utility of intraoperative pathology consultations of whipple resection specimens and their impact on final margin status.

The resection margin status is a significant surgical prognostic factor for the long-term outcomes of patients undergoing pancreaticoduodenectomy (Whipple procedure). As a result, surgeons frequently rely on intraoperative consults (IOCs) involving frozen sections to evaluate margin clearance during these resections. Nevertheless, the impact of this practice on final margin status and long-term outcomes remains a topic of debate. This study aimed to assess the impact of IOCs on the clearance rate of resection margins following Whipple procedure and distal pancreatectomy. A retrospective database review of all patients who underwent Whipple procedure or distal pancreatectomy at our institution between 2018 and 2020 was performed to evaluate the utility of IOCs by gastrointestinal surgeons and its correlation with final postoperative surgical margin status. A significant variation in the frequency of IOC requests for margins among surgeons was noted. However, the use of frozen section analysis for intraoperative margin assessment was not significantly associated with the clearance rate of final post-operative margins. More frequent use of IOC did not result in higher final margin clearance rate, an important prognostic factor following Whipple procedure.

Redefining and capturing pre-analytic deficiencies in an anatomical pathology laboratory: a quality improvement initiative.

Pre-analytical deficiencies (PADs) are a major source of errors in anatomical pathology, accounting for about 70% of laboratory deficiencies. These can lead to incorrect diagnoses, delayed treatments, and increased healthcare costs. As part of a quality improvement initiative, we retrospectively identified and characterized 237 PADs documented over a 1-year period in a tertiary care academic center. The most common PADs were errors in specimen procurement (56%), handling of samples within the lab (16%), accessioning (10%), incomplete requisitions (9%), and transportation-related issues (7%). Strategies were then devised to mitigate these errors. Categorization of pre- and intra-laboratory PADs was refined into eight categories (collection, requisition, specimen container, transportation, receiving, accessioning, preparation, and communications) in the laboratory information system. Mandatory PAD documentation was implemented for accessioning staff. Post-implementation, prospective analysis identified that the most common PADs were related to surgical requisitions (75%). Among these, missing ordering physician's signature was the most common, accounting for 67.7% of requisition-related PADs and 50.8% of all PADs. Other common PADs included incomplete information of specimens, clinical information, patient information, physician information, source location, collection time, incorrect requisition forms, and illegible handwritten information. This study highlights the importance of identifying and addressing PADs in the anatomical pathology laboratory setting as well as the potential benefits of implementing standardized documentation and quality improvement processes to address these deficiencies.

Nuclear expression of AFF2 C-terminus is a sensitive and specific ancillary marker for DEK::AFF2 carcinoma of the sinonasal tract.

DEK::AFF2 carcinoma of the sinonasal tract is an emerging entity. The tumor is typically characterized by papillary proliferation of non-keratinizing squamous epithelial cells with monotonous cytologic features, which may mimic other sinonasal tumors. The confirmation of this gene fusion has thus far relied solely on next-generation sequencing, fluorescence in situ hybridization (FISH), or reverse transcription polymerase chain reaction (RT-PCR). This current study aimed to validate an immunohistochemical assay for AFF2 C-terminus as an ancillary marker. We first analyzed publicly available RNA sequencing data of sinonasal tumors from the national center for biotechnology information (NCBI) sequence read archive and identified 3 DEK::AFF2 carcinomas out of 28 sinonasal tumors. The gene expression of AFF2 was significantly higher in the fusion-positive cases compared to the wild-type tumors (p < 0.001), while DEK was not. We then optimized an immunohistochemical assay with an anti-AFF2 C-terminus antibody for ancillary diagnosis. Seventeen DEK::AFF2 carcinomas, including 11 cases with predominantly low-grade morphology and one showing glandular differentiation, as well as 78 DEK FISH-negative sinonasal tumors were evaluated by AFF2 immunohistochemistry (IHC). Sixteen of the 17 DEK::AFF2 carcinomas showed nuclear AFF2 expression in ≥30% of tumor cells, including one decalcified case that failed FISH and RT-PCR confirmation. The one case that was negative for AFF2 IHC in the tumor cells also lacked expression in the internal positive control. It was thus considered a failure of the IHC rather than a truly negative case and was excluded from the statistical analysis. All DEK FISH-negative sinonasal tumors were negative for nuclear AFF2 expression. The nuclear expression of AFF2 IHC showed 100% sensitivity and specificity for DEK::AFF2 carcinoma. Accordingly, AFF2 IHC is a highly sensitive and specific ancillary marker that distinguishes DEK-AFF2 carcinoma from the other sinonasal tumors with overlapping morphological features and may be an especially useful alternative for decalcified specimens.

Radiologic-pathologic correlation of major versus minor extranodal extension in oral cavity cancer.

BACKGROUND: To evaluate the diagnostic performance of radiologic extranodal extension (rENE) in predicting major (>2 mm) and minor (≤2 mm) pathologic ENE (pENE). METHODS: All oral cavity squamous cell carcinoma patients who underwent neck dissection with pathological nodal disease (pN+) between 2010 and 2015 were reviewed. Preoperative computed tomography and/or magnetic resonance imaging were reviewed by two head and neck neuroradiologists. RESULTS: Three hundred and thirty-four patients were included. The sensitivity and specificity of rENE were 37% [95% CI 29-44] and 98% [95% CI 96-100], respectively. Sensitivity for pENE improved in the subset of patients with major ENE (48% [95% CI 38-57]). The presence of rENE was associated with inferior 3-year overall survival: 26% [95% CI 17-41] versus 60% [95% CI 54-67]. CONCLUSIONS: This large cohort study demonstrates high specificity, but low sensitivity for preoperative imaging in the detection of pENE in OCSCC. Patients with rENE demonstrated poor OS. pENE in the absence of rENE is still an adverse risk factor.

Metatypical Adenoid Cystic Carcinoma: A Variant Showing Prominent Squamous Differentiation With a Predilection for the Sinonasal Tract and Skull Base.

Adenoid cystic carcinoma is a malignant salivary gland neoplasm, commonly involving the major and minor salivary glands. Adenoid cystic carcinoma arising in the skull base region is considerably less common and is characterized by aggressive clinical behavior, perineural invasion, and intracranial extension. Classically, these tumors are composed of ductal and myoepithelial cells, arranged as tubules and cribriform structures, as well as in a solid pattern when higher in grade. The distinctive molecular findings in this tumor are the gene fusions involving the MYB/MYBL1 and NFIB genes. Squamous differentiation, trabecular, and macrocystic growth patterns are exceedingly rare in these tumors and when present can cause significant diagnostic challenges. Squamous differentiation, in particular, is considered by many to be an exclusion criterion for adenoid cystic carcinoma outside of cases with high-grade transformation. In addition, a similar-appearing tumor with squamous differentiation, namely human papillomavirus-related multiphenotypic sinonasal carcinoma, has recently been defined, further complicating this differential diagnosis. Recently, we have come across 3 cases of adenoid cystic carcinomas involving the sinonasal tract and skull base having extensive interconnecting trabecular growth, macrocysts, and squamous differentiation, yet demonstrating the signature fusions involving MYB-NFIB and MYBL1-NFIB by RNA sequencing. In this article, we describe the clinical, histomorphologic, and imaging findings of these cases and propose the appellation "metatypical adenoid cystic carcinoma" for this uncommon variant morphology.

DEK-AFF2 Carcinoma of the Sinonasal Region and Skull Base: Detailed Clinicopathologic Characterization of a Distinctive Entity.

A novel DEK-AFF2 fusion was recently reported in 4 nonkeratinizing squamous cell carcinomas of the sinonasal region and skull base, including 1 with exceptional response to immunotherapy, but it is not yet clear if this rearrangement defines a unique clinicopathologic category or represents a rare event. This study aims to characterize a larger cohort of carcinomas with DEK-AFF2 fusions to assess whether they truly constitute a distinctive entity. Among 27 sinonasal and skull base nonkeratinizing squamous cell carcinoma that were negative for human papillomavirus and Epstein-Barr virus, RNA sequencing identified DEK-AFF2 fusions in 13 cases (48%). Nine were centered in the nasal cavity, 2 in the middle ear/temporal bone, 1 in the nasopharynx, and 1 in the orbit. These tumors displayed recurrent histologic features including (1) complex endophytic and exophytic, frequently papilloma-like growth, (2) transitional epithelium with eosinophilic to amphophilic cytoplasm, (3) absent or minimal keratinization with occasional compact keratin pearls, (4) monotonous nuclei, and (5) prominent tumor-infiltrating neutrophils or stromal lymphocytes. This appearance not only overlaps with high-grade basaloid sinonasal carcinomas but also with benign papillomas and tumors reported as low-grade papillary Schneiderian carcinoma. However, DEK-AFF2 carcinomas showed frequent local recurrence, cervical lymph node metastases, and distant metastasis with 2 deaths from disease, confirming they are aggressive malignancies despite relatively bland histology. Overall, the distinctive molecular, histologic, and clinical features of DEK-AFF2 carcinomas suggest they represent a unique entity in the sinonasal region. This tumor merits increased pathologic recognition to better understand its prognostic and therapeutic implications.

Expanding Awareness of the Distribution and Biologic Potential of Ectomesenchymal Chondromyxoid Tumor.

Ectomesenchymal chondromyxoid tumor is a rare neoplasm of uncertain histogenesis that typically occurs in the anterior dorsal tongue. Recent reports in the literature have described rare examples of gingival, palatal and tonsillar lesions. Histologically, ectomesenchymal chondromyxoid tumors are typically well-circumscribed, lacking overtly aggressive features. Herein we report a tumor arising in the right mandible that is morphologically and molecularly consistent with ectomesenchymal chondromyxoid tumor. This case furthers awareness of the extra-glossal distribution of this neoplasm; moreover, it suggests that a subset of these tumors have the potential for locally aggressive behaviour.

Treatment implications of postoperative chemoradiotherapy for squamous cell carcinoma of the oral cavity with minor and major extranodal extension.

OBJECTIVES: To evaluate adjuvant chemoradiotherapy (CRT) for patients with oral cavity squamous cell carcinoma (OSCC) with minor or major extranodal extension (ENE). MATERIALS AND METHODS: Surgically resected OSCC with pathologically involved lymph node(s) (pN+) between 2006 and 2017. Sections of pN+ were re-reviewed and classified as no, minor (≤2 mm), or major (>2 mm) ENE. Patterns of failure and survival were compared between the groups and stratified by adjuvant treatment. Multivariable (MVA) analysis assessed the value of adjuvant treatment for minor and major ENE. RESULTS: Total of 384 patients, 62 had minor and 114 had major ENE. Adjuvant CRT was delivered in 32(15%), 21(34%), and 45(39%) of patients with no, minor and major ENE, respectively. Patients with minor ENE had similar 5-year loco-regional control (LRC) and distant control (DC) but lower disease-free survival (DFS) (38% vs. 51%, p = 0·02) compared to patients with no ENE, while patients with major ENE had marginally lower LRC (59% vs 74%, p = 0·07), lower DC (58% vs 82%,p = 0·005) and DFS (13% vs. 38%, p=·001) compared to those with minor. On MVA, adjuvant chemotherapy was associated with improved DFS for major ENE (adjusted HR = 0·49; 95% CI 0·29-0·85, p = 0·01) but not for minor ENE after adjusting for age, ECOG status, T-, N-category, margin status, and radiotherapy. CONCLUSIONS: Adjuvant chemoradiotherapy improves outcomes in patients with major ENE, but the benefit is unclear in patients with minor ENE. Future trials should focus on intensification of treatment for patients with major ENE and alternative adjuvant strategies for patients with minor ENE.

Middle Ear and Temporal Bone Nonkeratinizing Squamous Cell Carcinomas With DEK-AFF2 Fusion: An Emerging Entity.

Primary squamous cell carcinomas (SCCs) of the middle ear and temporal bone are rare and usually keratinizing by morphology. Nonkeratinizing, basaloid SCCs arising in this area are exceedingly rare, and, due to the anatomic proximity to the skull base, nasopharynx, and nasal sinuses, the differential diagnosis is broad. Most tumors with squamous differentiation arising in these subsites are either viral-induced (human papillomavirus/Epstein-Barr virus) or rarely may have specific molecular alterations (BRD4-NUT, EWSR1-FLI translocations). Occasional tumors are negative for these findings, and their pathogenesis is unknown. A recently discovered DEK-AFF2 fusion was clinically detected in a series of 2 cases known to the authors. This fusion has been previously reported in the literature in a patient with a base of skull tumor who was an exceptional responder to programmed cell death protein 1 inhibitor therapy. We examine here the histomorphologic and molecular findings of 2 additional cases of an emerging entity. Two male patients were identified. Each had a primary middle ear/temporal bone mass with locally advanced disease. The histology was reviewed, and immunohistochemistry was performed. RNA-based next-generation sequencing was performed for clinical detection of diagnostic or actionable fusions. Both patients had basaloid/nonkeratinizing tumors on biopsy. They were positive for markers of squamous differentiation (HMWK, CK5, and p40). By RNA sequencing, they demonstrated the presence of a DEK-AFF2 fusion and were negative for EWSR1 and NUT translocations. The DEK-AFF2 fusion may define a novel diagnostic category of middle ear and temporal bone nonkeratinizing/basaloid SCCs. This fusion also may represent a potential avenue for immunotherapy in these patients. Further studies are needed to fully explore whether this fusion defines a location-specific clinicopathologic entity.