Author Interviews, Dermatology, Genetic Research, Melanoma / 19.11.2020

MedicalResearch.com Interview with: Sarah I. Estrada, M.D., FCAP  Laboratory Director Affiliated Dermatology® www.affderm.com MedicalResearch.com: What is the background for this study? What are the main findings? Response: As a dermatopathologist who makes diagnoses on lesions that may be melanoma, I’m faced with the reality that my accurate interpretation of biopsy tissue is key for the patient to be treated most effectively. Often histopathological evaluation is straightforward but not as often as I would like. The study presented here offers a new test that can be used in conjunction with my evaluation to determine if a questionable lesion is in fact melanoma. The test was developed to take into account the gene expression of the lesion which may factor in characteristics that I cannot visually observe. The test was validated and has shown very promising accuracy metrics. (more…)
Author Interviews, Breast Cancer, JAMA, UCLA / 12.08.2019

MedicalResearch.com Interview with: Joann G. Elmore, MD, MPH Professor of Medicine, Director of the UCLA National Clinician Scholars Program David Geffen School of Medicine at UCLA  MedicalResearch.com: What is the background for this study? What are the main findings?
  • A pathologist makes the diagnosis of breast cancer versus non-cancer after reviewing the biopsy specimen. Breast biopsies are performed on millions of women each year and It is critical to get a correct diagnosis so that we can guide patients to the most effective treatments.
  • Our prior work (Elmore et al. 2015 JAMA) found significant levels of disagreement among pathologists when they interpreted the same breast biopsy specimen. We also found that pathologists would disagree with their own interpretations of breast biopsies when they where shown the same biopsy specimen a year later.
  • In this study, 240 breast biopsy images were fed into a computer, training it to recognize patterns associated with several types of breast lesions, ranging from benign (noncancerous) to invasive breast cancer. We compared the computer readings to independent diagnoses made by 87 practicing U.S. pathologists and found that while our artificial intelligence program came close to performing as well as human doctors in differentiating cancer from non-cancer, the AI program outperformed doctors when differentiating ductal carcinoma in situ (DCIS) from atypia, which is considered the greatest diagnostic challenge.
(more…)
Author Interviews, Lung Cancer, Nature, Technology / 05.03.2019

MedicalResearch.com Interview with: Saeed Hassanpour, PhD Assistant Professor Departments of Biomedical Data Science, Computer Science, and Epidemiology Geisel School of Medicine at Dartmouth Lebanon, NH 03756 MedicalResearch.com: What is the background for this study? What are the main findings? Response: Lung cancer is the deadliest cancer for both men and women in the western world. The most common form, lung adenocarcinoma, requires pathologist’s visual examination of resection slides to determine grade and treatment. However, this is a hard and tedious task. Using new technologies in artificial intelligence and deep learning, we trained a deep neural network to classify lung adenocarcinoma subtypes on histopathology slides and found that it performed on par with three practicing pathologists. (more…)
Author Interviews, Lung Cancer, Nature, NYU, Technology / 17.09.2018

MedicalResearch.com Interview with: Aristotelis Tsirigos, Ph.D. Associate Professor of Pathology Director, Applied Bioinformatics Laboratories New York University School of Medicine MedicalResearch.com: What is the background for this study? What are the main findings? Response: Pathologists routinely examine slides made from tumor samples to diagnose cancer types. We studied whether an AI algorithm can achieve the same task with high accuracy. Indeed, we show that such an algorithm can achieve an accuracy of ~97%, slightly better than individual pathologists. In addition, we demonstrated that AI can be used to predict genes that are mutated in these tumors, a task that pathologists cannot do. Although the accuracy for some genes is as high as 86%, there is still room for improvement. This will come from collecting more training data and also from improvement in the annotations of the slides by expert pathologists.   (more…)
Author Interviews, Technology / 21.08.2018

MedicalResearch.com Interview with:| Dr. Wendy L. Frankel, MD. Kurtz Chair and Distinguished Professor and Dr. Anil Parwani, MD, PhD, MBA, Associate Professor Wexner Medical Center The Ohio State University MedicalResearch.com: What is the background for this work? How does digital pathology differ from traditional H/E specimens?  Is there is different processing method?  Difference in prep time or costs? Response: Traditional pathology involves patient tissue coming to the lab and being processed. The end result is a glass slide with a stained tissue that pathologists use under a microscope. The process in digital pathology is the same, up until the point right after when the glass slide is made. In digital pathology, we put the glass slide under a scanner instead of under a microscope. The scanner creates a large file image that can be reviewed remotely by pathologists around the world. The advantage of digital pathology, and the reason we are doing this at The Ohio State University Comprehensive Cancer Center – Arthur G. James Cancer Hospital and Richard J. Solove Research Institute (OSUCCC - James), is because when the slide is digitized, the image can be rapidly shared with an expert for review, or another institute that the patient may be going to. In addition, I can look at the image and ask the computer to quantitate different types of features that are present in the sample. While this has historically been done manually with a microscope, it’s been a more subjective process that is open to human error. On top of that, we now have computer programs that allow us to ask very specific questions about the sample. For example, we can ask how many nuclei are in the field, how many of the nuclei show signs of cancer, and the size and color of the nucleus. These programs make the whole diagnostic process more objective and standardized. This is something we just can’t do by looking at a glass slide under a microscope. Finally, you can also use these images for presentations at clinical conferences or for teaching residents, fellows or other pathologists. You now have the means to create an archive of patient slides and have it instantaneously available. (more…)
Author Interviews, JAMA, Melanoma, UCLA / 24.05.2018

MedicalResearch.com Interview with: Joann G. Elmore, MD, MPH Professor of Medicine David Geffen School of Medicine at UCLA Director of the UCLA National Clinician Scholars Program Affiliate Professor of Medicine, University of Washington School of Medicine MedicalResearch.com: What is the background for this study? What are the main findings?  Response: In a recent study published in 2017 in the British Medical Journal, our team found that pathologists disagreed on their diagnoses of some melanocytic skin biopsy lesions and early stage invasive melanoma more than 50% of the time. This concerning level of disagreement was particularly true for diagnoses in the middle of the disease spectrum, such as atypical lesions and melanoma in situ.  For example, Figure 1 from this paper shows the diagnoses of 36 pathologists who interpreted the same glass slide of a skin biopsy using their own microscopes; the diagnoses ranged from a benign lesion to invasive melanoma. Since that study, the American Joint Committee on Cancer has released new guidelines for melanoma staging. Given this change, we wanted to examine whether the updated guidelines improved the reliability of melanoma diagnosis. We found that using the new guidelines improved the accuracy of pathologists’ diagnoses for invasive melanoma (Elmore J, et al, JAMA Network Open 2018).  (more…)
Author Interviews, Brain Cancer - Brain Tumors, Emory, PNAS, Technology / 16.03.2018

MedicalResearch.com Interview with: Lee Cooper, Ph.D. Assistant Professor of Biomedical Informatics Assistant Professor of Biomedical Engineering Emory University School of Medicine - Georgia Institute of Technology MedicalResearch.com: What is the background for this study? What are the main findings?  Response: Gliomas are a form of brain tumor that are often ultimately fatal, but patients diagnosed with glioma may survive as few as 6 months to 10 or more years. Prognosis is an important determinant in selecting treatment, that can range from simply monitoring the disease to surgical removal followed by radiation treatment and chemotherapy. Recent genomic studies have significantly improved our ability to predict how rapidly a patient's disease will progress, however a significant part of this determination still relies on the visual microscopic evaluation of the tissues by a neuropathologist. The neuropathologist assigns a grade that is used to further refine the prognosis determined by genomic testing. We developed a predictive algorithm to perform accurate and repeatable microscopic evaluation of glioma brain tumors. This algorithm learns the relationships between visual patterns presented in the brain tumor tissue removed from a patient brain and the duration of that patient's survival beyond diagnosis. The algorithm was demonstrated to accurately predict survival, and when combining images of histology with genomics into a single predictive framework, the algorithm was slightly more accurate than models based on the predictions of human pathologists. We were also able to identify that the algorithm learns to recognize some of the same tissue features used by pathologists in evaluating brain tumors, and to appreciate their prognostic relevance. (more…)
Author Interviews, Breast Cancer, Cancer Research, JAMA, Technology / 13.12.2017

MedicalResearch.com Interview with: Babak Ehteshami Bejnordi Department of Radiology and Nuclear Medicine Radboud University medical center, NijmegenBabak Ehteshami Bejnordi Department of Radiology and Nuclear Medicine Radboud University medical center, Nijmegen MedicalResearch.com: What is the background for this study? Response: Artificial intelligence (AI) will play a crucial role in health care. Advances in a family of AI popularly known as deep learning have ignited a new wave of algorithms and tools that read medical images for diagnosis. Analysis of digital pathology images is an important application of deep learning but requires evaluation for diagnostic performance. Accurate breast cancer staging is an essential task performed by the pathologists worldwide to inform clinical management. Assessing the extent of cancer spread by histopathological analysis of sentinel lymph nodes (SLN) is an important part of breast cancer staging. Traditionally, pathologists endure time and labor-intensive processes to assess tissues by reviewing thousands to millions of cells under a microscope. Using computer algorithms to analyze digital pathology images could potentially improve the accuracy and efficiency of pathologists. In our study, we evaluated the performance of deep learning algorithms at detecting metastases in lymph nodes of patients with breast cancer and compared it to pathologist’s diagnoses in a diagnostic setting. (more…)
Author Interviews, Technology / 26.11.2017

MedicalResearch.com Interview with: <a href="https://www.flickr.com/photos/fotologic/3862190141">“Virtual Reality”</a> by <i> <a href="https://www.flickr.com/people/fotologic/">fotologic</a> </i> is licensed under <a href="https://creativecommons.org/licenses/by/2.0"> CC BY 2.0</a>Professor Robert G. Parton FAA Institute for Molecular Bioscience, University of Queensland Brisbane, Australia MedicalResearch.com: What is the background for this study? What are the main findings? Response: We have combined whole cell microscopy with image analysis and virtual reality visualization to allow a user to explore the surface of a cancer cell and then to move inside the cell and interact with the organelles vital for cellular function. The study utilized serial blockface electron microscopy data of a migratory breast cancer cell (combining thousands of electron microscopic images into one 3D dataset) together with to-scale animations of endocytic processes at the cell surface. Through the use of low-cost commercial VR headsets, the user could then ‘walk’ over the alien landscape of the cell surface, observe endocytic processes occurring around them in real-time, and then enter a portal into the cell interior. We provide guidelines for the use of VR in biology and describe pilot studies illustrating the potential of this new immersive experience as an educational too (more…)
Author Interviews, Brigham & Women's - Harvard, Cancer Research, Dermatology, JAMA / 07.07.2017

MedicalResearch.com Interview with: Dr. Chrysalyne D. Schmults, MD, MSCE Associate Professor of Dermatology, Harvard Medical School Director, Mohs and Dermatologic Surgery Center and Mr. Pritesh S. Karia, MPH Department of Epidemiology Johns Hopkins Bloomberg School of Public Health, Baltimore, Maryland Department of Dermatology Brigham and Women's Faulkner Hospital Harvard Medical School, Boston, Massachusetts Jamaica Plain, MA 02130-3446  MedicalResearch.com: What is the background for this study? Response: Perineural nerve invasion (PNI) is a well-recognized risk factor for poor prognosis in patients with cutaneous squamous cell carcinoma (CSCC). Most cases of CSCC with PNI are identified on histologic examination at the time of surgery and the patient has no clinical symptoms or radiologic evidence of PNI. These cases are classified as incidental PNI (IPNI). However, some patients with PNI present with clinical symptoms and/or radiologic evidence of PNI. These cases are classified as clinical PNI (CPNI). A few studies have shown differences in disease-related outcomes between CSCC patients with IPNI and CPNI but consensus regarding adjuvant treatment and detailed guidelines on follow-up schedules have not yet materialized. (more…)
Author Interviews, BMJ, Dermatology / 03.07.2017

MedicalResearch.com Interview with: Joann G. Elmore M.D., M.P.H. Professor of Medicine, Adjunct Professor of Epidemiology, University of Washington School of Medicine Harborview Medical Center Seattle, WA 98104-2499 MedicalResearch.com: What is the background for this study? JE: Previous studies on diagnostic accuracy in interpreting melanocytic lesions exist but have small sample size, inclusion of experts only, or small numbers of specimens. We sought to examine accuracy and reproducibility in melanocytic skin lesions by improving upon the methodological limitations of previous studies. Specifically, we recruited a large national sample of practicing community and academic pathologists with a wide range of experience, and we utilized a large sample of biopsy cases that were carefully selected. Given that diagnostic errors can lead to patient deaths and invasive melanoma kills more than 9,000 Americans each year, we wanted to study the issue of diagnostic accuracy in interpreting melanocytic skin lesions in a very robust fashion. (more…)
Author Interviews, Dermatology, Melanoma / 15.11.2016

MedicalResearch.com Interview with: Dr. Mario Mandalà, MD Division of Oncology, Department of Oncology and Hematology Papa Giovanni XXIII Hospital Bergamo, Italy.  MedicalResearch.com: What is the background for this study? Response: The 7th edition of the TNM AJCC classification incorporated mitotic rate (MR) only for primary cutaneous melanoma with Breslow thickness ≤1 mm. We investigated whether and to what extent MR is able to predict sentinel lymph node (SLN) status and clinical outcome of  primary cutaneous melanoma (PCM) patients with BT >1 mm. (more…)
Author Interviews, Biomarkers / 09.09.2016

MedicalResearch.com Interview with: Stefan Enroth, Associate Professor, PhD Dept. of Immunology, Genetics & Pathology Uppsala University MedicalResearch.com: What is the background for this study? Response: One basic requirement of life science research is the quality of samples. Proper handling and rigorous biobanking of clinical samples is very important when for instance collecting samples for rare diseases, for monitoring individual variation in longitudinal studies and when conducting prospective studies of biomarkers and risk of developing for instance cardiovascular disease. In epidemiological studies using case and control cohorts, great care is taken to ensure that the cases and controls are matched in terms of for instance age, anthropometrics and lifestyle exposures such as smoking or alcohol consumption. Technical factors and sampling handling history are not as commonly used. There has been a lack of studies that systematically investigated the effects of for instance storage-time on a larger set of plasma proteins. With emerging high-throughput technologies enabling measurements of a high number of proteins simultaneously on a population level, biomarker research will enter a new era and the more knowledge we have on what factors that influence circulating biomarker levels - such as plasma proteins, the higher the chances are of finding new clinically important biomarkers for disease. (more…)
Author Interviews, Breast Cancer, Cost of Health Care, Johns Hopkins / 11.08.2016

MedicalResearch.com Interview with: Pedram Argani, M.D. Professor of Pathology and Principal consultant of the Breast Pathology Service Johns Hopkins Medicine MedicalResearch.com: What is the background for this study? What are the main findings? Response: Most pathology laboratories, at the request if clinicians, automatically (reflexively) test needle core biopsies containing ductal carcinoma in situ (DCIS) for estrogen receptor (ER) and progesterone receptor (PR). The logic for testing DCIS for these hormone receptors is that, for patients who have pure DCIS that is ER positive after surgical excision, treatment with estrogen blockers like Tamoxifen can decrease the recurrence of DCIS by a small amount, though overall survival (which is excellent) is not impacted. However, there are several factors which suggest that this reflex testing unnecessarily increases costs. • First, the ER/PR results on core needle biopsy do not impact the next step in therapy; namely, surgical excision. • Second, a subset of excisions performed for DCIS diagnosed on core needle biopsy will harbor invasive breast carcinoma, which would than need to be retested for ER/PR. • Third, because ER and PR labeling is often variable in DCIS, negative results for ER/PR in a small core biopsy specimen should logically be repeated in a surgical excision specimen with larger amounts of DICS to be sure that the result is truly negative. • Fourth, many patients with pure DCIS which is ER/PR positive after surgical excision will decline hormone therapy, so any ER/PR testing of their DCIS is unnecessary. • Fifth, PR status in DCIS has no independent value. We reviewed the Johns Hopkins experience with reflex ER/PR testing of DCIS on core needle biopsies over 2 years. We found that reflex core needle biopsy specimen testing unnecessarily increased costs by approximately $140.00 per patient. We found that ER/PR testing in the excision impacted management in only approximately one third of cases, creating an unnecessary increased cost of approximately $440.00 per patient. Extrapolating the increased cost of reflex ER/PR testing of DCIS to the 60,000 new cases of DCIS in the United States each year, reflex core needle biopsy ER/PR testing unnecessarily increased costs by approximately 35 million dollars. (more…)
Author Interviews, Breast Cancer, JAMA / 18.03.2015

Joann G. Elmore M.D., M.P.H. Professor of Medicine, Adjunct Professor of Epidemiology, University of Washington School of Medicine Harborview Medical Center Seattle, WA 98104-2499MedicalResearch.com Interview with: Joann G. Elmore M.D., M.P.H. Professor of Medicine, Adjunct Professor of Epidemiology, University of Washington School of Medicine Harborview Medical Center Seattle, WA 98104-2499 MedicalResearch: What is the background for this study? What are the main findings? Dr. Elmore: It is estimated that 1.6 million women in the United States each year undergo a breast biopsy. By interpreting these biopsies under the microscope, pathologists provide diagnoses on a spectrum from benign, to atypia, to ductal carcinoma in situ (DCIS), to invasive cancer. Using these diagnostic classifications, clinical doctors work with their patients to decide if they are at increased risk of developing breast cancer in the future, which can lead to additional surveillance, or how to treat them when the diagnosis is invasive breast cancer. As misclassification of breast lesions by pathologists may contribute to overtreatment and undertreatment of breast disease, we decided to study the accuracy of breast pathology diagnoses in the U.S. In the Breast Pathology (B-Path) Study, we used a set of 240 breast biopsy cases to evaluate the interpretive accuracy of 115 U.S. pathologists who were actively interpreting breast biopsies in their clinical practices. Their diagnoses were compared with reference diagnoses established by a consensus panel of experienced breast pathologists. When the panel members each independently diagnosed the slides pre-consensus, they agreed unanimously on 75 percent of their diagnoses; ninety percent of the panel members’ initial independent diagnoses agreed with the final consensus-derived reference diagnoses. When comparing participating pathologists’ diagnoses to the reference diagnoses, we found overall agreement for 75 percent of interpretations. The concordance rate for invasive breast cancer was reassuringly high at 96 percent, and fairly high for benign findings without atypia at 87 percent. However, concordance was lower for atypia at 48 percent and for DCIS at 84 percent. This means that nearly one out of five pathologists disagreed on the diagnosis of DCIS. We found disagreement with the reference diagnosis to be statistically more frequent when pathologists had lower weekly case volumes or worked in smaller labs. Disagreement was also statistically significantly more likely when the patient had dense breast tissue on mammogram; however, the absolute difference was small. Our accuracy findings were not altered when we used different methods of defining the reference diagnosis. (more…)