Variability in Pathologists' Interpretations of Individual Breast Biopsy Slides: A Population Perspective

Federated Fusion of Magnified Histopathological Images for Breast Tumor Classification in the Internet of Medical Things

Breast tumor detection and classification on the Internet of Medical Things (IoMT) can be automated with the potential of Artificial Intelligence (AI). Deep learning models rely on large datasets, however, challenges arise when dealing with sensitive medical data. Restrictions on sharing these medical data result in limited publicly available datasets thereby impacting the performance of the deep learning models. To address this issue, we propose an approach that combines different magnification factors of histopathological images using a residual network and information fusion in Federated Learning (FL). FL is employed to preserve the privacy of patient data, while enabling the creation of a global model. Using the BreakHis dataset, we compare the performance of FL with centralized learning (CL). We also performed visualizations for explainable AI. The final models obtained become available for deployment on internal IoMT systems in healthcare institutions for timely diagnosis and treatment. Our results demonstrate that the proposed approach outperforms existing works in the literature on multiple metrics.

Pathology Residents' Perceptions and Attitudes Toward Breast Pathology: A National Survey

CONTEXT.—

Breast pathology (BP) is considered to be subject to interobserver variability among pathologists, emphasizing the need for adequate training. However, specifics of BP residency training have not been elucidated.

OBJECTIVE.—

To assess the characteristics of BP residency training in the United States.

DESIGN.—

A Qualtrics-managed online survey was emailed to program directors of all US pathology residency programs, requesting them to forward the survey link to their pathology residents.

RESULTS.—

One hundred seventeen residents' survey responses were evaluable. Most responses (92; 79%) came from residents in university hospital-based programs. Thirty-five respondents (30%) had a dedicated BP rotation in their program. Most respondents believed that BP was an important part of training (96 of 100; 96%) and pathology practice (95 of 100; 95%). Seventy-one respondents believed that their BP training was adequate overall (71 of 100; 71%). Forty-one percent of respondents indicated that they would not like BP to be a significant part of their future practice. The main reasons given were that they had a different preferred area of interest, that they lacked interest in BP, or that breast cases were time-consuming to sign out.

CONCLUSIONS.—

Our results show that in the United States, most programs do not offer a dedicated BP rotation, but breast cases are signed out by subspecialized or experienced breast pathologists. In addition, most respondents believed that they received adequate training and would be competent to independently sign out BP in the future. Additional studies addressing new-in-practice pathologists' proficiency in BP will further help elucidate the quality of BP training in the United States.

AI-driven estimation of O6 methylguanine-DNA-methyltransferase (MGMT) promoter methylation in glioblastoma patients: a systematic review with bias analysis

BACKGROUND

Accurate and non-invasive estimation of MGMT promoter methylation status in glioblastoma (GBM) patients is of paramount clinical importance, as it is a predictive biomarker associated with improved overall survival (OS). In response to the clinical need, recent studies have focused on the development of non-invasive artificial intelligence (AI)-based methods for MGMT estimation. In this systematic review, we not only delve into the technical aspects of these AI-driven MGMT estimation methods but also emphasize their profound clinical implications. Specifically, we explore the potential impact of accurate non-invasive MGMT estimation on GBM patient care and treatment decisions.

METHODS

Employing a PRISMA search strategy, we identified 33 relevant studies from reputable databases, including PubMed, ScienceDirect, Google Scholar, and IEEE Explore. These studies were comprehensively assessed using 21 diverse attributes, encompassing factors such as types of imaging modalities, machine learning (ML) methods, and cohort sizes, with clear rationales for attribute scoring. Subsequently, we ranked these studies and established a cutoff value to categorize them into low-bias and high-bias groups.

RESULTS

By analyzing the 'cumulative plot of mean score' and the 'frequency plot curve' of the studies, we determined a cutoff value of 6.00. A higher mean score indicated a lower risk of bias, with studies scoring above the cutoff mark categorized as low-bias (73%), while 27% fell into the high-bias category.

CONCLUSION

Our findings underscore the immense potential of AI-based machine learning (ML) and deep learning (DL) methods in non-invasively determining MGMT promoter methylation status. Importantly, the clinical significance of these AI-driven advancements lies in their capacity to transform GBM patient care by providing accurate and timely information for treatment decisions. However, the translation of these technical advancements into clinical practice presents challenges, including the need for large multi-institutional cohorts and the integration of diverse data types. Addressing these challenges will be critical in realizing the full potential of AI in improving the reliability and accessibility of MGMT estimation while lowering the risk of bias in clinical decision-making.

From Image to Diagnosis: Characterizing Sources of Error in Histopathologic Interpretation

An accurate histopathologic diagnosis on surgical biopsy material is necessary for the clinical management of patients and has important implications for research, clinical trial design/enrollment, and public health education. This study used a mixed methods approach to isolate sources of diagnostic error while residents and attending pathologists interpreted digitized breast biopsy slides. Ninety participants, including pathology residents and attending physicians at major United States medical centers reviewed a set of 14 digitized whole-slide images of breast biopsies. Each case had a consensus-defined diagnosis and critical region of interest (cROI) representing the most significant pathology on the slide. Participants were asked to view unmarked digitized slides, draw their participant region of interest (pROI), describe its features, and render a diagnosis. Participants' review behavior was tracked using case viewer software and an eye-tracking device. Diagnostic accuracy was calculated in comparison to the consensus diagnosis. We measured the frequency of errors emerging during 4 interpretive phases: (1) detecting the cROI, (2) recognizing its relevance, (3) using the correct terminology to describe findings in the pROI, and (4) making a diagnostic decision. According to eye-tracking data, trainees and attending pathologists were very likely (∼94% of the time) to find the cROI when inspecting a slide. However, trainees were less likely to consider the cROI relevant to their diagnosis. Pathology trainees (41% of cases) were more likely to use incorrect terminology to describe pROI features than attending pathologists (21% of cases). Failure to accurately describe features was the only factor strongly associated with an incorrect diagnosis. Identifying where errors emerge in the interpretive and/or descriptive process and working on building organ-specific feature recognition and verbal fluency in describing those features are critical steps for achieving competency in diagnostic decision making.

Spatial mapping of the DNA adducts in cancer

DNA adducts and strand breaks are induced by various exogenous and endogenous agents. Accumulation of DNA damage is implicated in many disease processes, including cancer, aging, and neurodegeneration. The continuous acquisition of DNA damage from exogenous and endogenous stressors coupled with defects in DNA repair pathways contribute to the accumulation of DNA damage within the genome and genomic instability. While mutational burden offers some insight into the level of DNA damage a cell may have experienced and subsequently repaired, it does not quantify DNA adducts and strand breaks. Mutational burden also infers the identity of the DNA damage. With advances in DNA adduct detection and quantification methods, there is an opportunity to identify DNA adducts driving mutagenesis and correlate with a known exposome. However, most DNA adduct detection methods require isolation or separation of the DNA and its adducts from the context of the nuclei. Mass spectrometry, comet assays, and other techniques precisely quantify lesion types but lose the nuclear context and even tissue context of the DNA damage. The growth in spatial analysis technologies offers a novel opportunity to leverage DNA damage detection with nuclear and tissue context. However, we lack a wealth of techniques capable of detecting DNA damage in situ. Here, we review the limited existing in situ DNA damage detection methods and examine their potential to offer spatial analysis of DNA adducts in tumors or other tissues. We also offer a perspective on the need for spatial analysis of DNA damage in situ and highlight Repair Assisted Damage Detection (RADD) as an in situ DNA adduct technique with the potential to integrate with spatial analysis and the challenges to be addressed.

Identifying Good Candidates for Active Surveillance of Ductal Carcinoma In Situ: Insights from a Large Neoadjuvant Endocrine Therapy Cohort

Ductal carcinoma in situ (DCIS) is a biologically heterogenous entity with uncertain risk for invasive ductal carcinoma (IDC) development. Standard treatment is surgical resection often followed by radiation. New approaches are needed to reduce overtreatment. This was an observational study that enrolled patients with DCIS who chose not to pursue surgical resection from 2002 to 2019 at a single academic medical center. All patients underwent breast MRI exams at 3- to 6-month intervals. Patients with hormone receptor-positive disease received endocrine therapy. Surgical resection was strongly recommended if clinical or radiographic evidence of disease progression developed. A recursive partitioning (R-PART) algorithm incorporating breast MRI features and endocrine responsiveness was used retrospectively to stratify risk of IDC. A total of 71 patients were enrolled, 2 with bilateral DCIS (73 lesions). A total of 34 (46.6%) were premenopausal, 68 (93.2%) were hormone-receptor positive, and 60 (82.1%) were intermediate- or high-grade lesions. Mean follow-up time was 8.5 years. Over half (52.1%) remained on active surveillance without evidence of IDC with mean duration of 7.4 years. Twenty patients developed IDC, of which 6 were HER2 positive. DCIS and subsequent IDC had highly concordant tumor biology. Risk of IDC was characterized by MRI features after 6 months of endocrine therapy exposure; low-, intermediate-, and high-risk groups were identified with respective IDC rates of 8.7%, 20.0%, and 68.2%. Thus, active surveillance consisting of neoadjuvant endocrine therapy and serial breast MRI may be an effective tool to risk-stratify patients with DCIS and optimally select medical or surgical management.

Significance

A retrospective analysis of 71 patients with DCIS who did not undergo upfront surgery demonstrated that breast MRI features after short-term exposure to endocrine therapy identify those at high (68.2%), intermediate (20.0%), and low risk (8.7%) of IDC. With 7.4 years mean follow-up, 52.1% of patients remain on active surveillance. A period of active surveillance offers the opportunity to risk-stratify DCIS lesions and guide decisions for operative management.

The g3mclass is a practical software for multiclass classification on biomarkers

The analytes qualified as biomarkers are potent tools to diagnose various diseases, monitor therapy responses, and design therapeutic interventions. The early assessment of the diverseness of human disease is essential for the speedy and cost-efficient implementation of personalized medicine. We developed g3mclass, the Gaussian mixture modeling software for molecular assay data classification. This software automates the validated multiclass classifier applicable to single analyte tests and multiplexing assays. The g3mclass achieves automation using the original semi-constrained expectation-maximization (EM) algorithm that allows inference from the test, control, and query data that human experts cannot interpret. In this study, we used real-world clinical data and gene expression datasets (ERBB2, ESR1, PGR) to provide examples of how g3mclass may help overcome the problems of over-/underdiagnosis and equivocal results in diagnostic tests for breast cancer. We showed the g3mclass output's accuracy, robustness, scalability, and interpretability. The user-friendly interface and free dissemination of this multi-platform software aim to ease its use by research laboratories, biomedical pharma, companion diagnostic developers, and healthcare regulators. Furthermore, the g3mclass automatic extracting information through probabilistic modeling is adaptable for blending with machine learning and artificial intelligence.

Role of Artificial Intelligence in Radiogenomics for Cancers in the Era of Precision Medicine

Simple Summary

Recently, radiogenomics has played a significant role and offered a new understanding of cancer’s biology and behavior in response to standard therapy. It also provides a more precise prognosis, investigation, and analysis of the patient’s cancer. Over the years, Artificial Intelligence (AI) has provided a significant strength in radiogenomics. In this paper, we offer computational and oncological prospects of the role of AI in radiogenomics, as well as its offers, achievements, opportunities, and limitations in the current clinical practices.

Abstract

Radiogenomics, a combination of “Radiomics” and “Genomics,” using Artificial Intelligence (AI) has recently emerged as the state-of-the-art science in precision medicine, especially in oncology care. Radiogenomics syndicates large-scale quantifiable data extracted from radiological medical images enveloped with personalized genomic phenotypes. It fabricates a prediction model through various AI methods to stratify the risk of patients, monitor therapeutic approaches, and assess clinical outcomes. It has recently shown tremendous achievements in prognosis, treatment planning, survival prediction, heterogeneity analysis, reoccurrence, and progression-free survival for human cancer study. Although AI has shown immense performance in oncology care in various clinical aspects, it has several challenges and limitations. The proposed review provides an overview of radiogenomics with the viewpoints on the role of AI in terms of its promises for computational as well as oncological aspects and offers achievements and opportunities in the era of precision medicine. The review also presents various recommendations to diminish these obstacles.

Preneoplastic Low-Risk Mammary Ductal Lesions (Atypical Ductal Hyperplasia and Ductal Carcinoma In Situ Spectrum): Current Status and Future Directions

Intraepithelial mammary ductal neoplasia is a spectrum of disease that varies from atypical ductal hyperplasia (ADH), low-grade (LG), intermediate-grade (IG), to high-grade (HG) ductal carcinoma in situ (DCIS). While ADH has the lowest prognostic significance, HG-DCIS carries the highest risk. Due to widely used screening mammography, the number of intraepithelial mammary ductal neoplastic lesions has increased. The consequence of this practice is the increase in the number of patients who are overdiagnosed and, therefore, overtreated. The active surveillance (AS) trials are initiated to separate lesions that require active treatment from those that can be safely monitored and only be treated when they develop a change in the clinical/radiologic characteristics. At the same time, the natural history of these lesions can be evaluated. This review aims to evaluate ADH/DCIS as a spectrum of intraductal neoplastic disease (risk and histomorphology); examine the controversies of distinguishing ADH vs. DCIS and the grading of DCIS; review the upgrading for both ADH and DCIS with emphasis on the variation of methods of detection and the definitions of upgrading; and evaluate the impact of all these variables on the AS trials.

Breast cancer or surrounding normal tissue? A successful discrimination by FTIR or Raman microspectroscopy

Breast cancer is a type of cancer with the highest incidence worldwide in 2021, with early diagnosis and rapid treatment intervention being the reasons for the decreasing mortality rate associated with the disease.

Improving Medical Image Decision-Making by Leveraging Metacognitive Processes and Representational Similarity

Improving the accuracy of medical image interpretation can improve the diagnosis of numerous diseases. We compared different approaches to aggregating repeated decisions about medical images to improve the accuracy of a single decision maker. We tested our algorithms on data from both novices (undergraduates) and experts (medical professionals). Participants viewed images of white blood cells and made decisions about whether the cells were cancerous or not. Each image was shown twice to the participants and their corresponding confidence judgments were collected. The maximum confidence slating (MCS) algorithm leverages metacognitive abilities to consider the more confident response in the pair of responses as the more accurate "final response" (Koriat, 2012), and it has previously been shown to improve accuracy on our task for both novices and experts (Hasan et al., 2021). We compared MCS to similarity-based aggregation (SBA) algorithms where the responses made by the same participant on similar images are pooled together to generate the "final response." We determined similarity by using two different neural networks where one of the networks had been trained on white blood cells and the other had not. We show that SBA improves performance for novices even when the neural network had no specific training on white blood cell images. Using an informative representation (i.e., network trained on white blood cells) allowed one to aggregate over more neighbors and further boosted the performance of novices. However, SBA failed to improve the performance for experts even with the informative representation. This difference in efficacy of the SBA suggests different decision mechanisms for novices and experts.

A New Look into Cancer-A Review on the Contribution of Vibrational Spectroscopy on Early Diagnosis and Surgery Guidance

Simple Summary

Cancer is a leading cause of death worldwide, with the detection of the disease in its early stages, as well as a correct assessment of the tumour margins, being paramount for a successful recovery. While breast cancer is one of most common types of cancer, head and neck cancer is one of the types of cancer with a lower prognosis and poor aesthetic results. Vibrational spectroscopy detects molecular vibrations, being sensitive to different sample compositions, even when the difference was slight. The use of spectroscopy in biomedicine has been extensively explored, since it allows a broader assessment of the biochemical fingerprint of several diseases. This literature review covers the most recent advances in breast and head and neck cancer early diagnosis and intraoperative margin assessment, through Raman and Fourier transform infrared spectroscopies. The rising field of spectral histopathology was also approached. The authors aimed at expounding in a more concise and simple way the challenges faced by clinicians and how vibrational spectroscopy has evolved to respond to those needs for the two types of cancer with the highest potential for improvement regarding an early diagnosis, surgical margin assessment and histopathology.

Abstract

In 2020, approximately 10 million people died of cancer, rendering this disease the second leading cause of death worldwide. Detecting cancer in its early stages is paramount for patients’ prognosis and survival. Hence, the scientific and medical communities are engaged in improving both therapeutic strategies and diagnostic methodologies, beyond prevention. Optical vibrational spectroscopy has been shown to be an ideal diagnostic method for early cancer diagnosis and surgical margins assessment, as a complement to histopathological analysis. Being highly sensitive, non-invasive and capable of real-time molecular imaging, Raman and Fourier transform infrared (FTIR) spectroscopies give information on the biochemical profile of the tissue under analysis, detecting the metabolic differences between healthy and cancerous portions of the same sample. This constitutes tremendous progress in the field, since the cancer-prompted morphological alterations often occur after the biochemical imbalances in the oncogenic process. Therefore, the early cancer-associated metabolic changes are unnoticed by the histopathologist. Additionally, Raman and FTIR spectroscopies significantly reduce the subjectivity linked to cancer diagnosis. This review focuses on breast and head and neck cancers, their clinical needs and the progress made to date using vibrational spectroscopy as a diagnostic technique prior to surgical intervention and intraoperative margin assessment.

More scanning, but not zooming, is associated with diagnostic accuracy in evaluating digital breast pathology slides

Diagnoses of medical images can invite strikingly diverse strategies for image navigation and visual search. In computed tomography screening for lung nodules, distinct strategies, termed scanning and drilling, relate to both radiologists' clinical experience and accuracy in lesion detection. Here, we examined associations between search patterns and accuracy for pathologists (N = 92) interpreting a diverse set of breast biopsy images. While changes in depth in volumetric images reveal new structures through movement in the z-plane, in digital pathology changes in depth are associated with increased magnification. Thus, "drilling" in radiology may be more appropriately termed "zooming" in pathology. We monitored eye-movements and navigation through digital pathology slides to derive metrics of how quickly the pathologists moved through XY (scanning) and Z (zooming) space. Prior research on eye-movements in depth has categorized clinicians as either "scanners" or "drillers." In contrast, we found that there was no reliable association between a clinician's tendency to scan or zoom while examining digital pathology slides. Thus, in the current work we treated scanning and zooming as continuous predictors rather than categorizing as either a "scanner" or "zoomer." In contrast to prior work in volumetric chest images, we found significant associations between accuracy and scanning rate but not zooming rate. These findings suggest fundamental differences in the relative value of information types and review behaviors across two image formats. Our data suggest that pathologists gather critical information by scanning on a given plane of depth, whereas radiologists drill through depth to interrogate critical features.

Characterization of multiple diagnostic terms in melanocytic skin lesion pathology reports

BACKGROUND

Histopathologically ambiguous melanocytic lesions lead some pathologists to list multiple diagnostic considerations in the pathology report. The frequency and circumstance of multiple diagnostic considerations remain poorly characterized.

METHODS

Two hundred and forty skin biopsy samples were interpreted by 187 pathologists (8976 independent diagnoses) and classified according to a diagnostic/treatment stratification (MPATH-Dx).

RESULTS

Multiple diagnoses in different MPATH-Dx classes were used in n = 1320 (14.7%) interpretations, with 97% of pathologists and 91% of cases having at least one such interpretation. Multiple diagnoses were more common for intermediate risk lesions and are associated with greater subjective difficulty and lower confidence. We estimate that 6% of pathology reports for melanocytic lesions in the United States contain two diagnoses of different MPATH-Dx prognostic classes, and 2% of cases are given two diagnoses with significant treatment implications.

CONCLUSIONS

Difficult melanocytic diagnoses in skin may necessitate multiple diagnostic considerations; however, as patients increasingly access their health records and retrieve pathology reports (as mandated by US law), uncertainty should be expressed unambiguously.

Accuracy of Preoperative Breast MRI Versus Conventional Imaging in Measuring Pathologic Extent of Invasive Lobular Carcinoma

OBJECTIVE

To determine whether invasive lobular carcinoma (ILC) extent is more accurately depicted with preoperative MRI (pMRI) than conventional imaging (mammography and/or ultrasound).

METHODS

After IRB approval, we retrospectively identified women with pMRIs (February 2005 to January 2014) to evaluate pure ILC excluding those with ipsilateral pMRI BI-RADS 4 or 5 findings or who had neoadjuvant chemotherapy. Agreement between imaging and pathology sizes was summarized using Bland-Altman plots, absolute and percent differences, and the intraclass correlation coefficient (ICC). Rates of underestimation and overestimation were evaluated and their associations with clinical features were explored.

RESULTS

Among the 56 women included, pMRI demonstrated better agreement with pathology than conventional imaging by mean absolute difference (1.6 mm versus -7.8 mm, P < 0.001), percent difference (10.3% versus -16.4%, P < 0.001), and ICC (0.88 versus 0.61, P = 0.019). Conventional imaging more frequently underestimated ILC span than pMRI using a 5 mm difference threshold (24/56 (43%) versus 10/56 (18%), P < 0.001), a 25% threshold (19/53 (36%) versus 10/53 (19%), P = 0.035), and T category change (17/56 (30%) versus 7/56 (13%), P = 0.006). Imaging-pathology size concordance was greater for MRI-described solitary masses than other lesions for both MRI and conventional imaging (P < 0.05). Variability of conventional imaging was lower for patients ≥ the median age of 62 years than for younger patients (SD: 12 mm versus 22 mm, P = 0.012).

CONCLUSION

MRI depicts pure ILC more accurately than conventional imaging and may have particular value for younger women.

Classifying Breast Histopathology Images with a Ductal Instance-Oriented Pipeline

In this study, we propose the Ductal Instance-Oriented Pipeline (DIOP) that contains a duct-level instance segmentation model, a tissue-level semantic segmentation model, and three-levels of features for diagnostic classification. Based on recent advancements in instance segmentation and the Mask RCNN model, our duct-level segmenter tries to identify each ductal individual inside a microscopic image; then, it extracts tissue-level information from the identified ductal instances. Leveraging three levels of information obtained from these ductal instances and also the histopathology image, the proposed DIOP outperforms previous approaches (both feature-based and CNN-based) in all diagnostic tasks; for the four-way classification task, the DIOP achieves comparable performance to general pathologists in this unique dataset. The proposed DIOP only takes a few seconds to run in the inference time, which could be used interactively on most modern computers. More clinical explorations are needed to study the robustness and generalizability of this system in the future.

Image-Based Machine Learning Algorithms for Disease Characterization in the Human Type 1 Diabetes Pancreas

Emerging data suggest that type 1 diabetes affects not only the β-cell-containing islets of Langerhans, but also the surrounding exocrine compartment. Using digital pathology, machine learning algorithms were applied to high-resolution, whole-slide images of human pancreata to determine whether the tissue composition in individuals with or at risk for type 1 diabetes differs from those without diabetes. Transplant-grade pancreata from organ donors were evaluated from 16 nondiabetic autoantibody-negative controls, 8 nondiabetic autoantibody-positive subjects with increased type 1 diabetes risk, and 19 persons with type 1 diabetes (0 to 12 years' duration). HALO image analysis algorithms were implemented to compare architecture of the main pancreatic duct as well as cell size, density, and area of acinar, endocrine, ductal, and other nonendocrine, nonexocrine tissues. Type 1 diabetes was found to affect exocrine area, acinar cell density, and size, whereas the type of difference correlated with the presence or absence of insulin-positive cells remaining in the pancreas. These changes were not observed before disease onset, as indicated by modeling cross-sectional data from pancreata of autoantibody-positive subjects and those diagnosed with type 1 diabetes. These data provide novel insights into anatomic differences in type 1 diabetes pancreata and demonstrate that machine learning can be adapted for the evaluation of disease processes from cross-sectional data sets.

An update on the management of breast atypical ductal hyperplasia

Among lesions with uncertain malignant potential found at percutaneous breast biopsy, atypical ductal hyperplasia (ADH) carries both the highest risk of underestimation and the closest and most pathologist-dependent differential diagnosis with ductal carcinoma in situ (DCIS), matching the latter's features save for size only. ADH is therefore routinely surgically excised, but single-centre studies with limited sample size found low rates of upgrade to invasive cancer or DCIS. This suggests the possibility of surveillance over surgery in selected subgroups, considering the 2% threshold allowing for follow-up according to the Breast Imaging Reporting and Data System. A recent meta-analysis on 6458 lesions counters this approach, confirming that, surgically excised or managed with surveillance, ADH carries a 29% and 5% upgrade rate, respectively, invariably higher than 2% even in subgroups considering biopsy guidance and technique, needle calibre, apparent complete lesion removal. The high heterogeneity (I2 = 80%) found in this meta-analysis reaffirmed the need to synthesise evidence from systematic reviews to achieve generalisable results, fit for guidelines development. Limited tissue sampling at percutaneous biopsy intrinsically hampers the prediction of ADH-associated malignancy. This prediction could be improved by using contrast-enhanced breast imaging and applying artificial intelligence on both pathology and imaging results, allowing for overtreatment reduction.

Raman spectroscopy for rapid intra-operative margin analysis of surgically excised tumour specimens

Raman spectroscopy, a form of vibrational spectroscopy, has the ability to provide sensitive and specific biochemical analysis of tissue. This review article provides an in-depth analysis of the suitability of different Raman spectroscopy techniques in providing intra-operative margin analysis in a range of solid tumour pathologies. Surgical excision remains the primary treatment of a number of solid organ cancers. Incomplete excision of a tumour and positive margins on histopathological analysis is associated with a worse prognosis, the need for adjuvant therapies with significant side effects and a resulting financial burden. The provision of intra-operative margin analysis of surgically excised tumour specimens would be beneficial for a number of pathologies, as there are no widely adopted and accurate methods of margin analysis, beyond histopathology. The limitations of Raman spectroscopic studies to date are discussed and future work necessary to enable translation to clinical use is identified. We conclude that, although there remain a number of challenges in translating current techniques into a clinically effective tool, studies so far demonstrate that Raman Spectroscopy has the attributes to successfully perform highly accurate intra-operative margin analysis in a clinically relevant environment.

Machine Learning in oncology: A clinical appraisal

Machine learning (ML) is a branch of artificial intelligence centered on algorithms which do not need explicit prior programming to function but automatically learn from available data, creating decision models to complete tasks. ML-based tools have numerous promising applications in several fields of medicine. Its use has grown following the increased availability of patient data due to technological advances such as digital health records and high-volume information extraction from medical images. Multiple ML algorithms have been proposed for applications in oncology. For instance, they have been employed for oncological risk assessment, automated segmentation, lesion detection, characterization, grading and staging, prediction of prognosis and therapy response. In the near future, ML could become essential part of every step of oncological screening strategies and patients' management thus leading to precision medicine.

Inter-observer reproducibility of classical lobular neoplasia (B3 lesions) in preoperative breast biopsies: a study of the Swiss Working Group of breast and gynecopathologists

PURPOSE

Classical type of lobular neoplasia (LN) spans a spectrum of disease, including atypical lobular hyperplasia (ALH) and lobular carcinoma in situ (LCIS), classical lobular neoplasia (LN), and the three-tiered classification of lobular intraepithelial neoplasia (LIN-1, -2, -3). This study addressed inter-observer variability of classical lobular neoplasias (LN) (B3 lesions) in preoperative breast biopsies among breast and gynecopathologists METHODS: A retrospective, observational, cross-sectional study was conducted. 40 preoperative digital images of breast core/vacuum biopsies were analyzed by eight experienced breast- and gynecopathologists. Evaluation criteria were ALH, LCIS, LN classic, LIN-1, LIN-2, LIN-3, focal B3 (one focus), extensive B3 (> one focus). Kappa-index and Chi-square tests were used for statistics. Digital scanned slides were provided to each participant. Agreement between the categories was defined as at least six of eight (cut-off 75%) concordant diagnoses.

RESULTS

The highest agreement between eight pathologists was reached using the category lobular neoplasia (LN, classical), 26/40 (65%) cases were diagnosed as such. Agreements in other categories was low or poor: 12/40 (30%) (ALH), 9/40 (22%) (LCIS), 8/40 (20%) (LIN-1), 8/40 (20%) (focal B3), 3/40 (7.5%) (LIN-2), and 2/40 (5%) (extensive B3). Chi-square-test (classical LN versus the other nomenclatures) was significant (p = 0.001137).

CONCLUSION

Our data suggest that among Swiss breast pathologists, the most reproducible diagnosis for B3 lobular lesions is the category of classical LN. These data further support lack of consistent data in retrospective studies using different terminologies. Validation of reproducible nomenclature is warranted in further studies. This information is useful especially in view of retro- and prospective data analysis with different diagnostic categories.

Digital Whole Slide Imaging Compared With Light Microscopy for Primary Diagnosis in Surgical Pathology

Context.—

The adoption of digital capture of pathology slides as whole slide images (WSI) for educational and research applications has proven utility.

Objective.—

To compare pathologists' primary diagnoses derived from WSI versus the standard microscope. Because WSIs differ in format and method of observation compared with the current standard glass slide microscopy, this study is critical to potential clinical adoption of digital pathology.

Design.—

The study enrolled a total of 2045 cases enriched for more difficult diagnostic categories and represented as 5849 slides were curated and provided for diagnosis by a team of 19 reading pathologists separately as WSI or as glass slides viewed by light microscope. Cases were reviewed by each pathologist in both modalities in randomized order with a minimum 31-day washout between modality reads for each case. Each diagnosis was compared with the original clinical reference diagnosis by an independent central adjudication review.

Results.—

The overall major discrepancy rates were 3.64% for WSI review and 3.20% for manual slide review diagnosis methods, a difference of 0.44% (95% CI, −0.15 to 1.03). The time to review a case averaged 5.20 minutes for WSI and 4.95 minutes for glass slides. There was no specific subset of diagnostic category that showed higher rates of modality-specific discrepancy, though some categories showed greater discrepancy than others in both modalities.

Conclusions.—

WSIs are noninferior to traditional glass slides for primary diagnosis in anatomic pathology.

Integrating Artificial and Human Intelligence: A Partnership for Responsible Innovation in Biomedical Engineering and Medicine

Historically, the term "artificial intelligence" dates to 1956 when it was first used in a conference at Dartmouth College in the US. Since then, the development of artificial intelligence has in part been shaped by the field of neuroscience. By understanding the human brain, scientists have attempted to build new intelligent machines capable of performing complex tasks akin to humans. Indeed, future research into artificial intelligence will continue to benefit from the study of the human brain. While the development of artificial intelligence algorithms has been fast paced, the actual use of most artificial intelligence (AI) algorithms in biomedical engineering and clinical practice is still markedly below its conceivably broader potentials. This is partly because for any algorithm to be incorporated into existing workflows it has to stand the test of scientific validation, clinical and personal utility, application context, and is equitable as well. In this context, there is much to be gained by combining AI and human intelligence (HI). Harnessing Big Data, computing power and storage capacities, and addressing societal issues emergent from algorithm applications, demand deploying HI in tandem with AI. Very few countries, even economically developed states, lack adequate and critical governance frames to best understand and steer the AI innovation trajectories in health care. Drug discovery and translational pharmaceutical research stand to gain from AI technology provided they are also informed by HI. In this expert review, we analyze the ways in which AI applications are likely to traverse the continuum of life from birth to death, and encompassing not only humans but also all animal, plant, and other living organisms that are increasingly touched by AI. Examples of AI applications include digital health, diagnosis of diseases in newborns, remote monitoring of health by smart devices, real-time Big Data analytics for prompt diagnosis of heart attacks, and facial analysis software with consequences on civil liberties. While we underscore the need for integration of AI and HI, we note that AI technology does not have to replace medical specialists or scientists and rather, is in need of such expert HI. Altogether, AI and HI offer synergy for responsible innovation and veritable prospects for improving health care from prevention to diagnosis to therapeutics while unintended consequences of automation emergent from AI and algorithms should be borne in mind on scientific cultures, work force, and society at large.

Cancer diagnosis through a tandem of classifiers for digitized histopathological slides

The current research study is concerned with the automated differentiation between histopathological slides from colon tissues with respect to four classes (healthy tissue and cancerous of grades 1, 2 or 3) through an optimized ensemble of predictors. Six distinct classifiers with prediction accuracies ranging from 87% to 95% are considered for the task. The proposed method of combining them takes into account the probabilities of the individual classifiers for each sample to be assigned to any of the four classes, optimizes weights for each technique by differential evolution and attains an accuracy that is significantly better than the individual results. Moreover, a degree of confidence is defined that would allow the pathologists to separate the data into two distinct sets, one that is correctly classified with a high level of confidence and the rest that would need their further attention. The tandem is also validated on other benchmark data sets. The proposed methodology proves to be efficient in improving the classification accuracy of each algorithm taken separately and performs reasonably well on other data sets, even with default weights. In addition, by establishing a degree of confidence the method becomes more viable for use by actual practitioners.

High risk (B3) breast lesions: What is the incidence of malignancy for individual lesion subtypes? A systematic review and meta-analysis

INTRODUCTION

Provide evidence to support evolving management strategies for high-risk (B3) breast lesions by assessing risk of carcinoma in subgroups of B3 lesions using systematic review and meta-analysis.

METHODS

Databases identified observational studies between 1980 and 2015 that reported on underestimation of malignancy following B3 lesion diagnosis at core needle biopsy. Critical appraisal, quality assessment, data extraction and meta-analysis was undertaken to calculate rate of malignancy of the whole B3 group and individual lesions. Study heterogeneity and association between variables and underestimation rate was investigated using random effects logistic modelling.

RESULTS

Meta-analysis, using data from 129 studies, assessed 11 423 lesions of which 2160 were upgraded to malignancy after surgical excision biopsy (17% malignancy rate, 95% CI 15-19%). Malignancy rates varied from 6% in radial scars with no atypia (95% CI 2-13%, I2 72.8%), to 32% in papillomas with atypia (95% CI 23-41%, I2 57.4%). Differences in upgrade rates between atypical and non-atypical lesions were statistically significant (p < 0.05). Study heterogeneity could not be explained by differences in core biopsy size or year of publication.

CONCLUSIONS

This comprehensive, inclusive assessment of all published literature, provides an accurate estimate of malignancy risk in subgroups of B3 lesions, to guide tailored management strategies. Some lesions have a high risk of malignancy, while others have a much lower risk, and could be safely managed with surveillance strategies rather than surgery.

Intratumoral Heterogeneity in Ductal Carcinoma In Situ: Chaos and Consequence

Ductal carcinoma in situ (DCIS) is a non-invasive proliferative growth in the breast that serves as a non-obligate precursor to invasive ductal carcinoma. The widespread adoption of screening mammography has led to a steep increase in the detection of DCIS, which now comprises approximately 20% of new breast cancer diagnoses in the United States. Interestingly, the intratumoral heterogeneity (ITH) that has been observed in invasive breast cancers may have been established early in tumorigenesis, given the vast and varied ITH that has been detected in DCIS. This review will discuss the intratumoral heterogeneity of DCIS, focusing on the phenotypic and genomic heterogeneity of tumor cells, as well as the compositional heterogeneity of the tumor microenvironment. In addition, we will assess the spatial heterogeneity that is now being appreciated in these lesions, and summarize new approaches to evaluate heterogeneity of tumor and stromal cells in the context of their spatial organization. Importantly, we will discuss how a growing understanding of ITH has led to a more holistic appreciation of the complex biology of DCIS, specifically its evolution and natural history. Finally, we will consider ways in which our knowledge of DCIS ITH might be translated in the future to guide clinical care for DCIS patients.

Mammography-detected ultrasound-negative asymptomatic micro-calcifications in Chinese women: Would it be safe to watch and wait?

Although mammography (MG) has been widely used for breast cancer screening in the western world, over-diagnosis remains controversial. Milestone studies showed that ultrasound (US) was an effective primary screening test for breast cancer both in the western world and in China. US improves the sensitivity of screening in Chinese women who have denser breasts and develop breast cancer earlier than Caucasian counterparts, and is used as the primary imaging test in the hospital-based opportunistic screening among asymptomatic self-referred women. Our previous work showed that US result might further differentiate the MG-detected breast cancers into low risk (US+) and ultra-low risk (US-). Indeed, most of the MG+/US- breast cancers would be ultra-low risk cancers and almost always present as MG micro-calcifications. Furthermore, majority of the commonest MG+/US- abnormal finding of micro-calcification is usually benign. Biopsy of benign breast disease increases not only the risk of breast cancer, but the expenses of screening and healthcare. Our hypothesis proposes that mammography-positive ultrasound-negative (MG+/US-) asymptomatic micro-calcifications might not need immediate invasive procedures and be safe to observe until the micro-calcifications increase significantly or become US-positive. If this hypothesis is proved, US would serve as the primary imaging test for breast cancer screening in China, with MG as the selective screening test and diagnostic tool for surgical plan. Unnecessary biopsy or surgery might be avoided with screening expenses considerably decrease.

Complexities of perceived and actual performance in pathology interpretation: A comparison of cutaneous melanocytic skin and breast interpretations

BACKGROUND

Little is known about how pathologists process differences between actual and perceived interpretations.

OBJECTIVE

To compare perceived and actual diagnostic agreement before and after educational interventions.

METHODS

Pathologists interpreted test sets of skin and/or breast specimens that included benign, atypical, in situ and invasive lesions. Interventions involved self-directed learning, one skin and one breast, that showed pathologists how their interpretations compared to a reference diagnoses. Prior to the educational intervention, participants estimated how their interpretations would compare to the reference diagnoses. After the intervention, participants estimated their overall agreement with the reference diagnoses. Perceived and actual agreements were compared.

RESULTS

For pathologists interpreting skin, mean actual agreement was 52.4% and overall pre- and postinterventional mean perceived agreement was 72.9% vs 54.2%, an overestimated mean difference of 20.5% (95% confidence interval [CI] 17.2% to 24.0%) and 1.8% (95% CI -0.5% to 4.1%), respectively. For pathologists interpreting breast, mean actual agreement was 75.9% and overall pre- and postinterventional mean perceived agreement was 81.4% vs 76.9%, an overestimation of 5.5% (95% CI 3.0% to 8.0%) and 1.0% (95% CI 0.0% to 2.0%), respectively.

CONCLUSIONS

Pathologists interpreting breast tissue had improved comprehension of their performance after the intervention compared to pathologists interpreting skin lesions.

Mammography screening: A major issue in medicine

Breast cancer mortality is declining in most high-income countries. The role of mammography screening in these declines is much debated. Screening impacts cancer mortality through decreasing the incidence of number of advanced cancers with poor prognosis, while therapies and patient management impact cancer mortality through decreasing the fatality of cancers. The effectiveness of cancer screening is the ability of a screening method to curb the incidence of advanced cancers in populations. Methods for evaluating cancer screening effectiveness are based on the monitoring of age-adjusted incidence rates of advanced cancers that should decrease after the introduction of screening. Likewise, cancer-specific mortality rates should decline more rapidly in areas with screening than in areas without or with lower levels of screening but where patient management is similar. These two criteria have provided evidence that screening for colorectal and cervical cancer contributes to decreasing the mortality associated with these two cancers. In contrast, screening for neuroblastoma in children was discontinued in the early 2000s because these two criteria were not met. In addition, overdiagnosis - i.e. the detection of non-progressing occult neuroblastoma that would not have been life-threatening during the subject's lifetime - is a major undesirable consequence of screening. Accumulating epidemiological data show that in populations where mammography screening has been widespread for a long time, there has been no or only a modest decline in the incidence of advanced cancers, including that of de novo metastatic (stage IV) cancers at diagnosis. Moreover, breast cancer mortality reductions are similar in areas with early introduction and high penetration of screening and in areas with late introduction and low penetration of screening. Overdiagnosis is commonplace, representing 20% or more of all breast cancers among women invited to screening and 30-50% of screen-detected cancers. Overdiagnosis leads to overtreatment and inflicts considerable physical, psychological and economic harm on many women. Overdiagnosis has also exerted considerable disruptive effects on the interpretation of clinical outcomes expressed in percentages (instead of rates) or as overall survival (instead of mortality rates or stage-specific survival). Rates of radical mastectomies have not decreased following the introduction of screening and keep rising in some countries (e.g. the United States of America (USA)). Hence, the epidemiological picture of mammography screening closely resembles that of screening for neuroblastoma. Reappraisals of Swedish mammography trials demonstrate that the design and statistical analysis of these trials were different from those of all trials on screening for cancers other than breast cancer. We found compelling indications that these trials overestimated reductions in breast cancer mortality associated with screening, in part because of the statistical analyses themselves, in part because of improved therapies and underreporting of breast cancer as the underlying cause of death in screening groups. In this regard, Swedish trials should publish the stage-specific breast cancer mortality rates for the screening and control groups separately. Results of the Greater New York Health Insurance Plan trial are biased because of the underreporting of breast cancer cases and deaths that occurred in women who did not participate in screening. After 17 years of follow-up, the United Kingdom (UK) Age Trial showed no benefit from mammography screening starting at age 39-41. Until around 2005, most proponents of breast screening backed the monitoring of changes in advanced cancer incidence and comparative studies on breast cancer mortality for the evaluation of breast screening effectiveness. However, in an attempt to mitigate the contradictions between results of mammography trials and population data, breast-screening proponents have elected to change the criteria for the evaluation of cancer screening effectiveness, giving precedence to incidence-based mortality (IBM) and case-control studies. But practically all IBM studies on mammography screening have a strong ecological component in their design. The two IBM studies done in Norway that meet all methodological requirements do not document significant reductions in breast cancer mortality associated with mammography screening. Because of their propensity to exaggerate the health benefits of screening, case-control studies may demonstrate that mammography screening could reduce the risk of death from diseases other than breast cancer. Numerous statistical model approaches have been conducted for estimating the contributions of screening and of patient management to reductions in breast cancer mortality. Unverified assumptions are needed for running these models. For instance, many models assume that if screening had not occurred, the majority of screen-detected asymptomatic cancers would have progressed to symptomatic advanced cancers. This assumption is not grounded in evidence because a large proportion of screen-detected breast cancers represent overdiagnosis and hence non-progressing tumours. The accumulation of population data in well-screened populations diminishes the relevance of model approaches. The comparison of the performance of different screening modalities - e.g. mammography, digital mammography, ultrasonography, magnetic resonance imaging (MRI), three-dimensional tomosynthesis (TDT) - concentrates on detection rates, which is the ability of a technique to detect more cancers than other techniques. However, a greater detection rate tells little about the capacity to prevent interval and advanced cancers and could just reflect additional overdiagnosis. Studies based on the incidence of advanced cancers and on the evaluation of overdiagnosis should be conducted before marketing new breast-imaging technologies. Women at high risk of breast cancer (i.e. 30% lifetime risk and more), such as women with BRCA1/2 mutations, require a close breast surveillance. MRI is the preferred imaging method until more radical risk-reduction options are eventually adopted. For women with an intermediate risk of breast cancer (i.e. 10-29% lifetime risk), including women with extremely dense breast at mammography, there is no evidence that more frequent mammography screening or screening with other modalities actually reduces the risk of breast cancer death. A plethora of epidemiological data shows that, since 1985, progress in the management of breast cancer patients has led to marked reductions in stage-specific breast cancer mortality, even for patients with disseminated disease (i.e. stage IV cancer) at diagnosis. In contrast, the epidemiological data point to a marginal contribution of mammography screening in the decline in breast cancer mortality. Moreover, the more effective the treatments, the less favourable are the harm-benefit balance of screening mammography. New, effective methods for breast screening are needed, as well as research on risk-based screening strategies.

Characteristics associated with requests by pathologists for second opinions on breast biopsies

AIMS

Second opinions in pathology improve patient safety by reducing diagnostic errors, leading to more appropriate clinical treatment decisions. Little objective data are available regarding the factors triggering a request for second opinion despite second opinion consultations being part of the diagnostic system of pathology. Therefore we sought to assess breast biopsy cases and interpreting pathologists characteristics associated with second opinion requests.

METHODS

Collected pathologist surveys and their interpretations of 60 test set cases were used to explore the relationships between case characteristics, pathologist characteristics and case perceptions, and requests for second opinions. Data were evaluated by logistic regression and generalised estimating equations.

RESULTS

115 pathologists provided 6900 assessments; pathologists requested second opinions on 70% (4827/6900) of their assessments 36% (1731/4827) of these would not have been required by policy. All associations between case characteristics and requesting second opinions were statistically significant, including diagnostic category, breast density, biopsy type, and number of diagnoses noted per case. Exclusive of institutional policies, pathologists wanted second opinions most frequently for atypia (66%) and least frequently for invasive cancer (20%). Second opinion rates were higher when the pathologist had lower assessment confidence, in cases with higher perceived difficulty, and cases with borderline diagnoses.

CONCLUSIONS

Pathologists request second opinions for challenging cases, particularly those with atypia, high breast density, core needle biopsies, or many co-existing diagnoses. Further studies should evaluate whether the case characteristics identified in this study could be used as clinical criteria to prompt system-level strategies for mandating second opinions.

Second opinion strategies in breast pathology: a decision analysis addressing over-treatment, under-treatment, and care costs

PURPOSE

To estimate the potential near-term population impact of alternative second opinion breast biopsy pathology interpretation strategies.

METHODS

Decision analysis examining 12-month outcomes of breast biopsy for nine breast pathology interpretation strategies in the U.S. health system. Diagnoses of 115 practicing pathologists in the Breast Pathology Study were compared to reference-standard-consensus diagnoses with and without second opinions. Interpretation strategies were defined by whether a second opinion was sought universally or selectively (e.g., 2nd opinion if invasive). Main outcomes were the expected proportion of concordant breast biopsy diagnoses, the proportion involving over- or under-interpretation, and cost of care in U.S. dollars within one-year of biopsy.

RESULTS

Without a second opinion, 92.2% of biopsies received a concordant diagnosis. Concordance rates increased under all second opinion strategies, and the rate was highest (95.1%) and under-treatment lowest (2.6%) when all biopsies had second opinions. However, over-treatment was lowest when second opinions were sought selectively for initial diagnoses of invasive cancer, DCIS, or atypia (1.8 vs. 4.7% with no 2nd opinions). This strategy also had the lowest projected 12-month care costs ($5.907 billion vs. $6.049 billion with no 2nd opinions).

CONCLUSIONS

Second opinion strategies could lower overall care costs while reducing both over- and under-treatment. The most accurate cost-saving strategy required second opinions for initial diagnoses of invasive cancer, DCIS, or atypia.

Paradigm Shift toward Reducing Overtreatment of Ductal Carcinoma In Situ of Breast

The prevalence of ductal carcinoma in situ (DCIS) of the breast has increased substantially after the introduction of breast cancer screening programs, although the clinical effects of early DCIS detection and treatment remain unclear. The standard treatment for DCIS has involved local breast-conserving surgery (BCS) followed by radiotherapy (RT) or total mastectomy with/without endocrine therapy, and the choice of local treatment is not usually based on clinicopathologic or biological factors. However, we have investigated the effectiveness of local treatment using breast surgery and RT using Surveillance, Epidemiology, and End Results data, and found that the effectiveness of breast surgery was modified by the nuclear grade. Furthermore, breast cancer-specific survival was identical between patients with low-grade DCIS who did and did not undergo surgery. Moreover, we found that RT after BCS for DCIS was only associated with a survival benefit among patients with risk factors for local recurrence, such as nuclear grade, age, and tumor size. Ongoing clinical trials and translational research have attempted to develop a treatment strategy that prevents the overdiagnosis and overtreatment of low-risk DCIS, as well as a biology-based treatment strategy for using targeted therapy. Therefore, to develop a tailored treatment strategy for DCIS, we need to identify molecular and biological classifications based on the results from translational research, national databases, and clinical trials.

Consistency of BRCA1 and BRCA2 Variant Classifications Among Clinical Diagnostic Laboratories

BACKGROUND

Genetic tests of the cancer predisposition genes BRCA1 and BRCA2 inform significant clinical decisions for both physicians and patients. Most uncovered variants are benign, and determining which few are pathogenic (disease-causing) is sometimes challenging and can potentially be inconsistent among laboratories. The ClinVar database makes de-identified clinical variant classifications from multiple laboratories publicly available for comparison and review, per recommendations of the American Medical Association (AMA), the American College of Medical Genetics (ACMG), the National Society for Genetic Counselors (NSGC), and other organizations.

METHODS

Classifications of more than 2000 BRCA1/2 variants in ClinVar representing approximately 22,000 patients were dichotomized as clinically actionable or not actionable and compared across up to seven laboratories. The properties of these variants and classification differences were investigated in detail.

RESULTS

Per-variant concordance was 98.5% (CI 97.9%-99.0%). All discordant variants were rare; thus, per patient concordance was estimated to be higher: 99.7%. ClinVar facilitated resolution of many of the discordant variants, and concordance increased to 99.0% per variant and 99.8% per patient when reclassified (but not yet resubmitted) variants and submission errors were addressed. Most of the remaining discordances appeared to involve either legitimate differences in expert judgment regarding particular scientific evidence, or were classifications that predated availability of important scientific evidence.

CONCLUSIONS

Significant classification disagreements among the professional clinical laboratories represented in ClinVar are infrequent yet important. The unrestricted sharing of clinical genetic data allows detailed interlaboratory quality control and peer review, as exemplified by this study.

Pathologists' diagnosis of invasive melanoma and melanocytic proliferations: observer accuracy and reproducibility study

Objective To quantify the accuracy and reproducibility of pathologists' diagnoses of melanocytic skin lesions.Design Observer accuracy and reproducibility study.Setting 10 US states.Participants Skin biopsy cases (n=240), grouped into sets of 36 or 48. Pathologists from 10 US states were randomized to independently interpret the same set on two occasions (phases 1 and 2), at least eight months apart.Main outcome measures Pathologists' interpretations were condensed into five classes: I (eg, nevus or mild atypia); II (eg, moderate atypia); III (eg, severe atypia or melanoma in situ); IV (eg, pathologic stage T1a (pT1a) early invasive melanoma); and V (eg, ≥pT1b invasive melanoma). Reproducibility was assessed by intraobserver and interobserver concordance rates, and accuracy by concordance with three reference diagnoses.Results In phase 1, 187 pathologists completed 8976 independent case interpretations resulting in an average of 10 (SD 4) different diagnostic terms applied to each case. Among pathologists interpreting the same cases in both phases, when pathologists diagnosed a case as class I or class V during phase 1, they gave the same diagnosis in phase 2 for the majority of cases (class I 76.7%; class V 82.6%). However, the intraobserver reproducibility was lower for cases interpreted as class II (35.2%), class III (59.5%), and class IV (63.2%). Average interobserver concordance rates were lower, but with similar trends. Accuracy using a consensus diagnosis of experienced pathologists as reference varied by class: I, 92% (95% confidence interval 90% to 94%); II, 25% (22% to 28%); III, 40% (37% to 44%); IV, 43% (39% to 46%); and V, 72% (69% to 75%). It is estimated that at a population level, 82.8% (81.0% to 84.5%) of melanocytic skin biopsy diagnoses would have their diagnosis verified if reviewed by a consensus reference panel of experienced pathologists, with 8.0% (6.2% to 9.9%) of cases overinterpreted by the initial pathologist and 9.2% (8.8% to 9.6%) underinterpreted.Conclusion Diagnoses spanning moderately dysplastic nevi to early stage invasive melanoma were neither reproducible nor accurate in this large study of pathologists in the USA. Efforts to improve clinical practice should include using a standardized classification system, acknowledging uncertainty in pathology reports, and developing tools such as molecular markers to support pathologists' visual assessments.

Surgical implications and variability in the use of the flat epithelial atypia diagnosis on breast biopsy specimens

OBJECTIVES

Flat epithelial atypia (FEA) is a relatively new diagnostic term with uncertain clinical significance for surgical management. Any implied risk of invasive breast cancer associated with FEA is contingent upon diagnostic reproducibility, yet little is known regarding its use.

MATERIALS AND METHODS

Pathologists in the Breast Pathology Study interpreted one of four 60-case test sets, one slide per case, constructed from 240 breast biopsy specimens. An electronic data form with standardized diagnostic categories was used; participants were instructed to indicate all diagnoses present. We assessed participants' use of FEA as a diagnostic term within: 1) each test set; 2) 72 cases classified by reference as benign without FEA; and 3) six cases classified by reference as FEA. 115 pathologists participated, providing 6900 total independent assessments.

RESULTS

Notation of FEA ranged from 0% to 35% of the cases interpreted, with most pathologists noting FEA on 4 or more test cases. At least one participant noted FEA in 34 of the 72 benign non-FEA cases. For the 6 reference FEA cases, participant agreement with the case reference FEA diagnosis ranged from 17% to 52%; diagnoses noted by participating pathologists for these FEA cases included columnar cell hyperplasia, usual ductal hyperplasia, atypical lobular hyperplasia, and atypical ductal hyperplasia.

CONCLUSIONS

We observed wide variation in the diagnosis of FEA among U.S. pathologists. This suggests that perceptions of diagnostic criteria and any implied risk associated with FEA may also vary. Surgical excision following a core biopsy diagnosis of FEA should be reconsidered and studied further.

Characteristics of a Breast Pathology Consultation Practice

CONTEXT

- Intradepartmental consultation is a routine practice commonly used for new diagnoses. Expert interinstitutional case review provides insight into particularly challenging cases.

OBJECTIVE

- To investigate the practice of breast pathology consultation at a large tertiary care center.

DESIGN

- We reviewed breast pathology cases sent for private consultation and internal cases reviewed by multiple pathologists at a tertiary center. Requisitions and reports were evaluated for diagnostic reason for consultation, rate of multiple pathologist review at the tertiary center, use of immunohistochemistry, and, for private consultation cases, type of sender and concordance with the outside diagnosis.

RESULTS

- In the 985 private consultation cases, the most frequent reasons for review were borderline atypia (292 of 878; 33.3%), papillary lesion classification (151 of 878; 17.2%), evaluating invasion (123 of 878; 14%), subtyping carcinoma (75 of 878; 8.5%), and spindle cell (67 of 878; 7.6%) and fibroepithelial (65 of 878; 7.4%) lesion classification. Of 4981 consecutive internal cases, 358 (7.2%) were reviewed, most frequently for borderline atypia (90 of 358; 25.1%), subtyping carcinoma (63 of 358; 17.6%), staging/prognostic features (59 of 358; 16.5%), fibroepithelial lesion classification (45 of 358; 12.6%), evaluating invasion (37 of 358; 10.3%), and papillary (20 of 358; 5.6%) and spindle cell (18 of 358; 5.0%) lesion classification. Of all internal cases, those with a final diagnosis of atypia had a significantly higher rate of review (58 of 241; 24.1%) than those with benign (119 of 2933; 4.1%) or carcinoma (182 of 1807; 10.1%) diagnoses. Immunohistochemistry aided in diagnosis of 39.7% (391 of 985) and 21.2% (76 of 359) of consultation and internally reviewed cases, respectively.

CONCLUSIONS

- This study confirms areas of breast pathology that represent diagnostic challenge and supports that pathologists are appropriately using expert consultation.

A Randomized Study Comparing Digital Imaging to Traditional Glass Slide Microscopy for Breast Biopsy and Cancer Diagnosis

Background:

Digital whole slide imaging may be useful for obtaining second opinions and is used in many countries. However, the U.S. Food and Drug Administration requires verification studies.

Methods:

Pathologists were randomized to interpret one of four sets of breast biopsy cases during two phases, separated by ≥9 months, using glass slides or digital format (sixty cases per set, one slide per case, n = 240 cases). Accuracy was assessed by comparing interpretations to a consensus reference standard. Intraobserver reproducibility was assessed by comparing the agreement of interpretations on the same cases between two phases. Estimated probabilities of confirmation by a reference panel (i.e., predictive values) were obtained by incorporating data on the population prevalence of diagnoses.

Results:

Sixty-five percent of responding pathologists were eligible, and 252 consented to randomization; 208 completed Phase I (115 glass, 93 digital); and 172 completed Phase II (86 glass, 86 digital). Accuracy was slightly higher using glass compared to digital format and varied by category: invasive carcinoma, 96% versus 93% (P = 0.04); ductal carcinoma in situ (DCIS), 84% versus 79% (P < 0.01); atypia, 48% versus 43% (P = 0.08); and benign without atypia, 87% versus 82% (P < 0.01). There was a small decrease in intraobserver agreement when the format changed compared to when glass slides were used in both phases (P = 0.08). Predictive values for confirmation by a reference panel using glass versus digital were: invasive carcinoma, 98% and 97% (not significant [NS]); DCIS, 70% and 57% (P = 0.007); atypia, 38% and 28% (P = 0.002); and benign without atypia, 97% and 96% (NS).

Conclusions:

In this large randomized study, digital format interpretations were similar to glass slide interpretations of benign and invasive cancer cases. However, cases in the middle of the spectrum, where more inherent variability exists, may be more problematic in digital format. Future studies evaluating the effect these findings exert on clinical practice and patient outcomes are required.

Estimate of false-positive breast cancer diagnoses from accuracy studies: a systematic review

Background

False-positive histological diagnoses have the same consequences of overdiagnosis in terms of unnecessary treatment. The aim of this systematic review is to assess their frequency at needle core biopsy (CB) and/or surgical excision of the breast.

Methods

PubMed, Embase, Cochrane Library were systematically searched up to 30 October 2015. Eligibility criteria: cross-sectional studies assessing diagnostic accuracy of CB compared with surgical excision; studies assessing reproducibility of pathologists reading the same slides. Outcomes: false-positive rates; Misclassification of Benign as Malignant (MBM) histological diagnosis; K statistic. Independent reviewers extracted data and assessed quality using an adapted QUADAS-2 tool.

Results

Sixteen studies assessed CB false-positive rates. In 10 studies (41 989 screen-detected lesions), the range of false-positive rates was 0%–7.1%. Twenty-seven studies assessed pathologists' reproducibility. Studies with consecutive, random or stratified samples of all the specimens: at CB the MBM range was 0.25%–2.4% (K values 0.83–0.98); at surgical excision, it was 0.67%–1.2% (K values 0.86–0.94). Studies with enriched samples: the MBM range was 1.4%–6.2% (K values 0.57–0.86). Studies of cases selected for second opinion: the MBM range was 0.29%–12.2% (K values 0.48 and 0.50).

Conclusions

High heterogeneity of the included studies precluded formal pooling estimates. When considering studies of higher sample size or methodological quality, false-positive rates and MBM are around 1%. The impact of false-positive histological diagnoses of breast cancer on unnecessary treatment, as well as that of overdiagnosis, is not negligible and is of importance in clinical practice.

DATA SHARING AND REPRODUCIBLE CLINICAL GENETIC TESTING: SUCCESSES AND CHALLENGES

Open sharing of clinical genetic data promises to both monitor and eventually improve the reproducibility of variant interpretation among clinical testing laboratories. A significant public data resource has been developed by the NIH ClinVar initiative, which includes submissions from hundreds of laboratories and clinics worldwide. We analyzed a subset of ClinVar data focused on specific clinical areas and we find high reproducibility (>90% concordance) among labs, although challenges for the community are clearly identified in this dataset. We further review results for the commonly tested BRCA1 and BRCA2 genes, which show even higher concordance, although the significant fragmentation of data into different silos presents an ongoing challenge now being addressed by the BRCA Exchange. We encourage all laboratories and clinics to contribute to these important resources.

Review of the national external quality assessment (EQA) scheme for breast pathology in the UK

Background

The National Health Service Breast Screening Programme (NHSBSP; pathology) external quality assurance (EQA) scheme aims to provide a mechanism for examination and monitoring of concordance of pathology reporting within the UK. This study aims to review the breast EQA scheme performance data collected over a 24-year period following its introduction.

Methods

Data on circulations, number of cases and diagnosis were collected. Detailed analyses with and without combinations of certain diagnostic entities, and over different time periods were performed.

Results

Overall, of 576 cases (172 benign, 11 atypical hyperplasia, 98 ductal carcinoma in situ/microinvasive and 295 invasive disease), consistency of assessment of diagnostic parameters was very high (overall k=0.80; k for benign diagnosis=0.79; k for invasive disease=0.91). For distinguishing benign versus malignant lesions, no further improvement is considered possible in view of the limitations of the scheme methodology. Although diagnostic consistency of atypical hyperplasia remains at a low level, combining it with the benign category results in a high level of agreement (k=0.93). The level of consistency of reporting prognostic information is variable and some items such as lymphovascular invasion and tumour size measurement may need further intervention to improve their reporting consistency. Although the level of consistency of reporting of histological grade remained at a moderate level overall (k=0.48), it was variable among cases and appears to have levelled off; no further significant improvement is expected and no significant impact of the previous publication of guidelines is observed.

Conclusions

These results provide further evidence to indicate the value of the breast EQA scheme in monitoring performance and the identification of specific areas where improvement or new approaches are required. For most parameters, the concordance of reporting reached a plateaux a few years after the introduction of the EQA scheme. It is important to maintain this high level and also to tackle specific low-performance areas innovatively.

Evaluation of 12 strategies for obtaining second opinions to improve interpretation of breast histopathology: simulation study

OBJECTIVE

 To evaluate the potential effect of second opinions on improving the accuracy of diagnostic interpretation of breast histopathology.

DESIGN

 Simulation study.

SETTING

 12 different strategies for acquiring independent second opinions.

PARTICIPANTS

 Interpretations of 240 breast biopsy specimens by 115 pathologists, one slide for each case, compared with reference diagnoses derived by expert consensus.

MAIN OUTCOME MEASURES

 Misclassification rates for individual pathologists and for 12 simulated strategies for second opinions. Simulations compared accuracy of diagnoses from single pathologists with that of diagnoses based on pairing interpretations from first and second independent pathologists, where resolution of disagreements was by an independent third pathologist. 12 strategies were evaluated in which acquisition of second opinions depended on initial diagnoses, assessment of case difficulty or borderline characteristics, pathologists' clinical volumes, or whether a second opinion was required by policy or desired by the pathologists. The 240 cases included benign without atypia (10% non-proliferative, 20% proliferative without atypia), atypia (30%), ductal carcinoma in situ (DCIS, 30%), and invasive cancer (10%). Overall misclassification rates and agreement statistics depended on the composition of the test set, which included a higher prevalence of difficult cases than in typical practice.

RESULTS

 Misclassification rates significantly decreased (P<0.001) with all second opinion strategies except for the strategy limiting second opinions only to cases of invasive cancer. The overall misclassification rate decreased from 24.7% to 18.1% when all cases received second opinions (P<0.001). Obtaining both first and second opinions from pathologists with a high volume (≥10 breast biopsy specimens weekly) resulted in the lowest misclassification rate in this test set (14.3%, 95% confidence interval 10.9% to 18.0%). Obtaining second opinions only for cases with initial interpretations of atypia, DCIS, or invasive cancer decreased the over-interpretation of benign cases without atypia from 12.9% to 6.0%. Atypia cases had the highest misclassification rate after single interpretation (52.2%), remaining at more than 34% in all second opinion scenarios.

CONCLUSION

 Second opinions can statistically significantly improve diagnostic agreement for pathologists' interpretations of breast biopsy specimens; however, variability in diagnosis will not be completely eliminated, especially for breast specimens with atypia.