An ECG-based machine learning model for predicting new-onset atrial fibrillation is superior to age and clinical features in identifying patients at high stroke risk.

BACKGROUND: Several large trials have employed age or clinical features to select patients for atrial fibrillation (AF) screening to reduce strokes. We hypothesized that a machine learning (ML) model trained to predict AF risk from 12­lead electrocardiogram (ECG) would be more efficient than criteria based on clinical variables in indicating a population for AF screening to potentially prevent AF-related stroke. METHODS: We retrospectively included all patients with clinical encounters in Geisinger without a prior history of AF. Incidence of AF within 1 year and AF-related strokes within 3 years of the encounter were identified. AF-related stroke was defined as a stroke where AF was diagnosed at the time of stroke or within a year after the stroke. The efficiency of five methods was evaluated for selecting a cohort for AF screening. The methods were selected from four clinical trials (mSToPS, GUARD-AF, SCREEN-AF and STROKESTOP) and the ECG-based ML model. We simulated patient selection for the five methods between the years 2011 and 2014 and evaluated outcomes for 1 year intervals between 2012 and 2015, resulting in a total of twenty 1-year periods. Patients were considered eligible if they met the criteria before the start of the given 1-year period or within that period. The primary outcomes were numbers needed to screen (NNS) for AF and AF-associated stroke. RESULTS: The clinical trial models indicated large proportions of the population with a prior ECG for AF screening (up to 31%), coinciding with NNS ranging from 14 to 18 for AF and 249-359 for AF-associated stroke. At comparable sensitivity, the ECG ML model indicated a modest number of patients for screening (14%) and had the highest efficiency in NNS for AF (7.3; up to 60% reduction) and AF-associated stroke (223; up to 38% reduction). CONCLUSIONS: An ECG-based ML risk prediction model is more efficient than contemporary AF-screening criteria based on age alone or age and clinical features at indicating a population for AF screening to potentially prevent AF-related strokes.

rECHOmmend: An ECG-Based Machine Learning Approach for Identifying Patients at Increased Risk of Undiagnosed Structural Heart Disease Detectable by Echocardiography.

BACKGROUND: Timely diagnosis of structural heart disease improves patient outcomes, yet many remain underdiagnosed. While population screening with echocardiography is impractical, ECG-based prediction models can help target high-risk patients. We developed a novel ECG-based machine learning approach to predict multiple structural heart conditions, hypothesizing that a composite model would yield higher prevalence and positive predictive values to facilitate meaningful recommendations for echocardiography. METHODS: Using 2 232 130 ECGs linked to electronic health records and echocardiography reports from 484 765 adults between 1984 to 2021, we trained machine learning models to predict the presence or absence of any of 7 echocardiography-confirmed diseases within 1 year. This composite label included the following: moderate or severe valvular disease (aortic/mitral stenosis or regurgitation, tricuspid regurgitation), reduced ejection fraction <50%, or interventricular septal thickness >15 mm. We tested various combinations of input features (demographics, laboratory values, structured ECG data, ECG traces) and evaluated model performance using 5-fold cross-validation, multisite validation trained on 1 site and tested on 10 independent sites, and simulated retrospective deployment trained on pre-2010 data and deployed in 2010. RESULTS: Our composite rECHOmmend model used age, sex, and ECG traces and had a 0.91 area under the receiver operating characteristic curve and a 42% positive predictive value at 90% sensitivity, with a composite label prevalence of 17.9%. Individual disease models had area under the receiver operating characteristic curves from 0.86 to 0.93 and lower positive predictive values from 1% to 31%. Area under the receiver operating characteristic curves for models using different input features ranged from 0.80 to 0.93, increasing with additional features. Multisite validation showed similar results to cross-validation, with an aggregate area under the receiver operating characteristic curve of 0.91 across our independent test set of 10 clinical sites after training on a separate site. Our simulated retrospective deployment showed that for ECGs acquired in patients without preexisting structural heart disease in the year 2010, 11% were classified as high risk and 41% (4.5% of total patients) developed true echocardiography-confirmed disease within 1 year. CONCLUSIONS: An ECG-based machine learning model using a composite end point can identify a high-risk population for having undiagnosed, clinically significant structural heart disease while outperforming single-disease models and improving practical utility with higher positive predictive values. This approach can facilitate targeted screening with echocardiography to improve underdiagnosis of structural heart disease.

Comparison of gene fusion detection methods in salivary gland tumors.

Salivary gland neoplasms may pose diagnostic difficulties due to overlapping morphologic features. Recently, specific gene fusions have been discovered that correspond to particular tumor types, and can aid in accurate diagnosis. Gene rearrangements are commonly assessed by fluorescence in situ hybridization (FISH), although use of next-generation sequencing is increasing. However, there is no "gold standard" for fusion detection. We determined the concordance between FISH and a targeted RNA sequencing panel in gene fusion detection across twenty-two salivary gland tumors, including five mucoepidermoid carcinomas, four acinic cell carcinomas, four pleomorphic adenomas, two adenoid cystic carcinomas, two NUT carcinomas, and one each of basal cell adenoma, salivary duct carcinoma ex-pleomorphic adenoma, salivary duct carcinoma, clear cell carcinoma, and secretory carcinoma. Directed FISH testing based on the diagnosis was performed on cases that did not already have FISH conducted during clinical workup. Targeted RNA sequencing of 507 genes and their partners (using the Illumina TruSight Fusion Panel) was completed. Six of twenty-two (27.3%) cases had discordant results. In three cases, FISH results were negative while RNA sequencing results found fusion transcripts, which were all confirmed with RT-PCR and Sanger sequencing. In three cases, RNA sequencing results were negative while FISH results were positive for a gene rearrangement. Thus, if fusion analysis results are conflicting with the morphologic impression, a second mode of fusion detection may be warranted. Although both methods have advantages and drawbacks, RNA sequencing provides additional information about novel fusion partners and fusions that may not have been originally considered.

Expression of Potential Biomarker Targets by Immunohistochemistry in Cervical Carcinomas.

There have been few clinically useful targetable biomarkers in uterine cervical carcinomas. Estrogen receptor (ER), HER2, and fibroblast activation protein (FAP) are potential therapeutic or theranostic targets in other gynecologic and genitourinary carcinoma types. We determined the immunohistochemical expression patterns of these markers in treatment-naive cervical carcinoma, and whether expression correlated with clinical outcomes after definitive chemoradiation therapy. Tissue microarrays were created from 71 patient samples taken before therapy (57 squamous cell carcinomas and 14 nonsquamous cell carcinomas) and stained for ER, HER2, and FAP. ER was positive in 25/70 cases (36%). Of 66 tumors with evaluable HER2 staining, only 1 had positive (3+) staining (3%, positive for HER2 amplification by fluorescence in situ hybridization), and 1 had equivocal (2+) staining (negative for amplification by fluorescence in situ hybridization). The remainder were negative for HER2 overexpression. FAP expression was widely variably in the tumor stroma. ER positivity and FAP expression did not correlate with cervical recurrence, pelvic recurrence, distant recurrence, or cancer death. In conclusion, HER2 amplification is very rare in nonmetastatic treatment-naive cervical carcinomas, but if present, could represent a target for antibody therapy. ER and FAP were expressed in a subset of tumors, but expression did not correlate with clinical outcomes. These immunohistochemical markers do not demonstrate prognostic significance in treatment-naive cervical cancer, but they may have utility in targeted therapy or imaging.

A Framework of Critical Considerations in Clinical Exome Reanalyses by Clinical and Laboratory Standards Institute.

Exome reanalysis is useful for providing molecular diagnoses for previously uninformative samples. However, challenges exist in implementing a practical solution for clinicians and laboratories. This study complements the current literature by providing practical considerations for patient-level and cohort-level reanalyses. The Clinical and Laboratory Standards Institute assembled the Document Development Committee and an interpretation working group that developed the framework for reevaluation of exome-based data. We describe two distinct but complementary approaches toward exome reanalyses: clinician-initiated patient-level reanalysis, and laboratory-initiated cohort-level reanalysis. We highlight the advantages and constraints for both approaches, and provide a high-level conceptual guide for ordering clinicians and laboratories through the critical decision pathways. Because clinical exome sequencing continues to be the standard of care in genetics, exome reanalysis would be critical in increasing the overall diagnostic yield. A systematic guide will facilitate the efficient adoption of reevaluation of exome data for laboratories, health care professionals, genetic counselors, and clinicians.

Laboratory and Clinical Implications of Incidental and Secondary Germline Findings During Tumor Testing.

CONTEXT.­: Next-generation sequencing is a powerful clinical tool for cancer management but can produce incidental/secondary findings that require special consideration. OBJECTIVE.­: To discuss clinical and laboratory issues related to incidental or secondary germline findings in the clinical setting of tumor testing and inform future guidelines in this area. DESIGN.­: A College of American Pathologists workgroup including representation from the American Society of Clinical Oncology, the Association for Molecular Pathology, and the American College of Medical Genetics and Genomics created a review of items that should be considered when developing guidelines for incidental or secondary findings when performing clinical tumor testing. RESULTS.­: Testing recommendations should be cognizant of the differences among anticipated incidental, unanticipated incidental, and secondary findings, and whether normal tissue is also tested. In addition to defining which variants will be reported, robust recommendations must also take into account test design and validation, reimbursement, cost, infrastructure, impact on reflex testing, and maintenance of proficiency. Care providers need to consider the potential of a test to uncover incidental or secondary findings, the recommendation of upfront counseling, the need for consent, the timing of testing and counseling, and that the exact significance of a finding may not be clear. CONCLUSIONS.­: As clinical oncology testing panels have become a mainstay of clinical cancer care, guidelines addressing the unique aspects of incidental and secondary findings in oncology testing are needed. This paper highlights clinical and laboratory considerations with regard to incidental/secondary findings and is a clarion call to create recommendations.

Histopathology of Third Trimester Placenta from SARS-CoV-2-Positive Women.

Background: This study aims to investigate whether maternal SARS-CoV-2 status affects placental pathology. Methods: A retrospective case-control study was conducted by reviewing charts and slides of placentas delivered between April 1 to July 24, 2020. Clinical history of "COVID-19" was searched in Pathology Database (CoPath). Controls were matched with SARS-CoV-2-negative women with singleton deliveries in the 3rd-trimester. Pathological features were extracted from placental pathology reports. Results: Twenty-one 3rd trimester placentas from SARS-CoV-2-positive women were identified and compared to 20 placentas from SARS-CoV-2-negative women. There were no significant differences in individual or group gross or microscopic pathological features. Within the SARS-CoV-2+ group, there are no differences between symptomatic and asymptomatic women. Conclusion: Placentas from SARS-CoV-2-positive women do not demonstrate a specific pathological pattern. Pregnancy complicated with COVID-19 during the 3rd trimester does not have a demonstrable effect on placental structure and pathology.

Reference Samples to Compare Next-Generation Sequencing Test Performance for Oncology Therapeutics and Diagnostics.

OBJECTIVES: Diversity of laboratory-developed tests (LDTs) using next-generation sequencing (NGS) raises concerns about their accuracy for selection of targeted therapies. A working group developed a pilot study of traceable reference samples to measure NGS LDT performance among a cohort of clinical laboratories. METHODS: Human cell lines were engineered via CRISPR/Cas9 and prepared as formalin-fixed, paraffin-embedded cell pellets ("wet" samples) to assess the entire NGS test cycle. In silico mutagenized NGS sequence files ("dry" samples) were used to assess the bioinformatics component of the NGS test cycle. Single and multinucleotide variants (n = 36) of KRAS and NRAS were tested at 5% or 15% variant allele fraction to determine eligibility for therapy with the EGFR inhibitor panitumumab in the setting of metastatic colorectal cancer. RESULTS: Twenty-one (21/21) laboratories tested wet samples; 19 of 21 analyzed dry samples. Of the laboratories that tested both the wet and dry samples, 7 (37%) of 19 laboratories correctly reported all variants, 3 (16%) of 19 had fewer than five errors, and 9 (47%) of 19 had five or more errors. Most errors were false negatives. CONCLUSIONS: Genetically engineered cell lines and mutagenized sequence files are complementary reference samples for evaluating NGS test performance among clinical laboratories using LDTs. Variable accuracy in detection of genetic variants among some LDTs may identify different patient populations for targeted therapy.

Clinical Implications of a Targeted RNA-Sequencing Panel in the Detection of Gene Fusions in Solid Tumors.

The detection of recurrent gene fusions can help confirm diagnoses in solid tumors, particularly when the morphology and staining are unusual or nonspecific, and can guide therapeutic decisions. Although fluorescence in situ hybridization and PCR are often used to identify fusions, the rearrangement must be suspected, with only a few prioritized probes run. It was hypothesized that the Illumina TruSight RNA Fusion Panel, which detects fusions of 507 genes and their partners, would uncover fusions with greater sensitivity than other approaches, leading to changes in diagnosis, prognosis, or therapy. Targeted RNA sequencing was performed on formalin-fixed, paraffin-embedded sarcoma and carcinoma cases in which fluorescence in situ hybridization, RT-PCR, or DNA-based sequencing was conducted during the diagnostic workup. Of the 153 cases, 138 (90%) were sequenced with adequate quality control metrics. A total of 101 of 138 (73%) cases were concordant by RNA sequencing and the prior test method. RNA sequencing identified an additional 30 cases (22%) with fusions that were not detected by conventional methods. In seven cases (5%), the additional fusion information provided by RNA sequencing would have altered diagnosis and management. A total of 19 novel fusion pairs (not previously described in the literature) were discovered (14%). Overall, the findings show that a targeted RNA-sequencing method can detect gene fusions in formalin-fixed, paraffin-embedded specimens with high sensitivity.

Programmed Death Ligand 1 Expression and Related Markers in Pleuropulmonary Blastoma.

INTRODUCTION: Pleuropulmonary blastoma (PPB), a rare childhood neoplasm of the lung, is linked to pathogenic DICER1 variants. We investigated checkpoint inhibitor markers including Programmed Death Ligand 1 (PD-L1), PD1, CD8 and tumor mutational burden (TMB) in PPB. MATERIAL AND METHODS: Cases were collected from departmental archives and the International PPB/DICER1 Registry. Immunohistochemistry (IHC) for PD-L1, PD-1, CD8 and DNA mismatch repair (MMR) genes were performed. In addition, normal-tumor paired whole exome sequencing (WES) was performed in two cases. RESULTS: Twenty-five PPB cases were studied, consisting of Type I (n = 8, including 2 Ir), Type II (n = 8) and Type III (n = 9). PD-L1 combined positive score (CPS) of 1, 4 and 80 was seen in three (3/25, 12.0%) cases of Type II PPB with negative staining in the remaining cases. PD-1 and CD8 stains demonstrated positive correlation (P < .05). The density of PD1 and CD8 in the interface area was higher than within tumor (P < .05). The MMR proteins were retained. TMB was 0.65 mutations/Mb in type II PPB with high expression of PD-L1, and 0.94 mutations/Mb in one negative PD-L1 case with metastatic tumor. CONCLUSION: A small subpopulation of PPB patient might benefit from checkpoint immunotherapy due to positive PD-L1 staining.

A Curriculum for Genomic Education of Molecular Genetic Pathology Fellows: A Report of the Association for Molecular Pathology Training and Education Committee.

Molecular genetic pathology (MGP) is a subspecialty of pathology and medical genetics and genomics. Genomic testing, which is defined as that which generates large data sets and interrogates large segments of the genome in a single assay, is increasingly recognized as essential for optimal patient care through precision medicine. The most common genomic testing technologies in clinical laboratories are next-generation sequencing and microarray. It is essential to train in these methods and to consider the data generated in the context of the diagnosis, medical history, and other clinical findings of individual patients. Accordingly, updating the MGP fellowship curriculum to include genomics is timely, important, and challenging. At the completion of training, an MGP fellow should be capable of independently interpreting and signing out results of a wide range of genomic assays and, given the appropriate context and institutional support, of developing and validating new assays in compliance with applicable regulations. The Genomics Task Force of the MGP Program Directors, a working group of the Association for Molecular Pathology Training and Education Committee, has developed a genomics curriculum framework and recommendations specific to the MGP fellowship. These recommendations are presented for consideration and implementation by MGP fellowship programs with the understanding that MGP programs exist in a diversity of clinical practice environments with a spectrum of available resources.

Odontoblast TRPC5 channels signal cold pain in teeth.

Teeth are composed of many tissues, covered by an inflexible and obdurate enamel. Unlike most other tissues, teeth become extremely cold sensitive when inflamed. The mechanisms of this cold sensation are not understood. Here, we clarify the molecular and cellular components of the dental cold sensing system and show that sensory transduction of cold stimuli in teeth requires odontoblasts. TRPC5 is a cold sensor in healthy teeth and, with TRPA1, is sufficient for cold sensing. The odontoblast appears as the direct site of TRPC5 cold transduction and provides a mechanism for prolonged cold sensing via TRPC5's relative sensitivity to intracellular calcium and lack of desensitization. Our data provide concrete functional evidence that equipping odontoblasts with the cold-sensor TRPC5 expands traditional odontoblast functions and renders it a previously unknown integral cellular component of the dental cold sensing system.

Deep Neural Networks Can Predict New-Onset Atrial Fibrillation From the 12-Lead ECG and Help Identify Those at Risk of Atrial Fibrillation-Related Stroke.

BACKGROUND: Atrial fibrillation (AF) is associated with substantial morbidity, especially when it goes undetected. If new-onset AF could be predicted, targeted screening could be used to find it early. We hypothesized that a deep neural network could predict new-onset AF from the resting 12-lead ECG and that this prediction may help identify those at risk of AF-related stroke. METHODS: We used 1.6 M resting 12-lead digital ECG traces from 430 000 patients collected from 1984 to 2019. Deep neural networks were trained to predict new-onset AF (within 1 year) in patients without a history of AF. Performance was evaluated using areas under the receiver operating characteristic curve and precision-recall curve. We performed an incidence-free survival analysis for a period of 30 years following the ECG stratified by model predictions. To simulate real-world deployment, we trained a separate model using all ECGs before 2010 and evaluated model performance on a test set of ECGs from 2010 through 2014 that were linked to our stroke registry. We identified the patients at risk for AF-related stroke among those predicted to be high risk for AF by the model at different prediction thresholds. RESULTS: The area under the receiver operating characteristic curve and area under the precision-recall curve were 0.85 and 0.22, respectively, for predicting new-onset AF within 1 year of an ECG. The hazard ratio for the predicted high- versus low-risk groups over a 30-year span was 7.2 (95% CI, 6.9-7.6). In a simulated deployment scenario, the model predicted new-onset AF at 1 year with a sensitivity of 69% and specificity of 81%. The number needed to screen to find 1 new case of AF was 9. This model predicted patients at high risk for new-onset AF in 62% of all patients who experienced an AF-related stroke within 3 years of the index ECG. CONCLUSIONS: Deep learning can predict new-onset AF from the 12-lead ECG in patients with no previous history of AF. This prediction may help identify patients at risk for AF-related strokes.

Deep-learning-assisted analysis of echocardiographic videos improves predictions of all-cause mortality.

Machine learning promises to assist physicians with predictions of mortality and of other future clinical events by learning complex patterns from historical data, such as longitudinal electronic health records. Here we show that a convolutional neural network trained on raw pixel data in 812,278 echocardiographic videos from 34,362 individuals provides superior predictions of one-year all-cause mortality. The model's predictions outperformed the widely used pooled cohort equations, the Seattle Heart Failure score (measured in an independent dataset of 2,404 patients with heart failure who underwent 3,384 echocardiograms), and a machine learning model involving 58 human-derived variables from echocardiograms and 100 clinical variables derived from electronic health records. We also show that cardiologists assisted by the model substantially improved the sensitivity of their predictions of one-year all-cause mortality by 13% while maintaining prediction specificity. Large unstructured datasets may enable deep learning to improve a wide range of clinical prediction models.

Cancer of the Vulva: A Review.

BACKGROUND: Vulvar cancers, although rare, are becoming an increasingly serious threat to women's health. Cancer of the vulva accounted for 0.3% of all new cancers in the United States in 2019, with 6,070 newly diagnosed cases. This review details the epidemiology, pathogenesis, diagnosis, staging, and treatment of vulvar malignancies. OBJECTIVE: To review cancer entities of the vulva, including vulvar intraepithelial neoplasms, squamous cell carcinoma (SCC), malignant melanoma, basal cell carcinoma, neuroendocrine tumors, and adenocarcinomas. MATERIALS AND METHODS: Literature review using PubMed search for articles related to cancer of the vulva. RESULTS: Vulvar intraepithelial neoplasms represent premalignant precursors to SCC of the vulva. There are several different histopathologic subtypes of SCC, and treatment is dependent on characteristics of primary tumor and lymph node involvement. Melanoma is the second most common cancer to affect the vulva, and staging is based on tumor, node, and metastatic spread. CONCLUSION: Vulvar malignancies are rare, and diagnosis is dependent on biopsy and pathologic evaluation. Treatment for vulvar malignancies depends on histopathologic diagnosis but ranges from wide local excision with or without lymph node biopsy or dissection to radiation therapy with chemo- or immunotherapy. Overall survival varies by diagnosis.

Clinicopathologic and molecular features of six cases of phosphaturic mesenchymal tumor.

Phosphaturic mesenchymal tumors (PMT) are rare neoplasms characterized by secretion of FGF23, resulting in renal phosphate wasting and osteomalacia. This tumor-induced osteomalacia (TIO) is cured by complete resection; thus, diagnosis is important, particularly on biopsy. Although PMT have a classic histologic appearance of bland spindled cells with conspicuous vascular network and characteristic smudgy basophilic matrix, there is a broad histologic spectrum and variant histologic patterns can make recognition difficult. Recent studies have demonstrated FN1-FGFR1 and FN1-FGF1 gene fusions in PMT; however, approximately 50% of cases are negative for these fusions. We sought to characterize 6 cases of PMT in-depth, compare fusion detection methods, and determine whether alternative fusions could be uncovered by targeted RNA sequencing. Of the 6 cases of PMT in our institutional archive, 3 were not given diagnoses of PMT at the time of initial pathologic examination. We characterized the immunoprofile (SMA, D2-40, CD56, S100 protein, desmin, SATB2, and ERG) and gene fusion status (FN1 and FGFR1 rearrangements by fluorescent in situ hybridization (FISH) and two targeted RNA sequencing approaches) in these cases. Tumors were consistently positive for SATB2 and negative for desmin, with 5/6 cases expressing ERG and CD56. One specimen was acid-decalcified and failed FISH and RNA sequencing. We found FN1 gene rearrangements by FISH in 2/5 cases, and a FN1-FGFR1 fusion by targeted RNA sequencing. No alternative gene fusions were identified by RNA sequencing. Our findings suggest that IHC and molecular analysis can aid in the diagnosis of PMT, guiding excision of the tumor and resolution of osteomalacia.

A Next-Generation Sequencing Test for Severe Congenital Neutropenia: Utility in a Broader Clinicopathologic Spectrum of Disease.

Severe congenital neutropenia (SCN) is a collection of diverse disorders characterized by chronically low absolute neutrophil count in the peripheral blood, increased susceptibility to infection, and a significant predisposition to the development of myeloid malignancies. SCN can be acquired or inherited. Inherited forms have been linked to variants in a group of diverse genes involved in the neutrophil-differentiation process. Variants that promote resistance to treatment have also been identified. Thus, genetic testing is important for the diagnosis, prognosis, and management of SCN. Herein we describe clinically validated assay developed for assessing patients with suspected SCN. The assay is performed from a whole-exome backbone. Variants are called across all coding exons, and results are filtered to focus on 48 genes that are clinically relevant to SCN. Validation results indicated 100% analytical sensitivity and specificity for the detection of constitutional variants among the 48 reportable genes. To date, 34 individuals have been referred for testing (age range: birth to 67 years). Several pathogenic and likely pathogenic variants have been identified, including one in a patient with late-onset disease. The pattern of cases referred for testing suggests that this assay has clinical utility in a broader spectrum of patients beyond those in the pediatric population who have classic early-onset symptoms characteristic of SCN.

Histopathologic, immunophenotypic, and proteomics characteristics of low-grade phyllodes tumor and fibroadenoma: more similarities than differences.

Distinguishing low-grade phyllodes tumor from fibroadenoma is practically challenging due to their overlapping histologic features. However, the final interpretation is essential to surgeons, who base their management on the final pathology report. Patients who receive a diagnosis of fibroadenoma might not undergo any additional intervention while lumpectomy with wide margins is the standard of care for phyllodes tumor, which can have significant cosmetic consequences. We studied the clinical, immunophenotypic, and proteomics profiles of 31 histologically confirmed low-grade phyllodes tumor and 30 fibroadenomas. Matrix-assisted laser desorption ionization (MALDI) imaging mass spectrometry (IMS) and immunohistochemistry for Ki-67, p53, ß-catenin, and E-cadherin were performed on all cases. After the mass spectra for all 31 cases of low-grade phyllodes tumor and 30 cases of fibroadenoma were collected, an average peak value for all cases was generated. There was no significant difference in the overall mass spectra pattern in any of the peaks identified. There was also overlap in the percentage of cells staining positive for Ki-67, p53, ß-catenin, and E-cadherin. The two groups of patients showed no statistically significant difference in age, tumor size, or disease-free survival. Neither group developed malignant transformation, distant metastases, or disease-related mortality. We have demonstrated low-grade phyllodes tumor and fibroadenoma to show significant overlapping clinical and proteomics features.