The Silva pattern-based classification for HPV-associated endocervical adenocarcinoma: A single institution concordance study of trainees and gynecologic pathologists.

The Silva pattern-based classification of HPV-associated endocervical adenocarcinoma has become an integral part of the histologic assessment of these tumors. Unfortunately, the Silva system reproducibility has had mixed results in past studies, and clinical practice still favors the FIGO stage assessment in directing therapeutic interventions for patients. In our study, we aimed to assess our institution's concordance including not only gynecologic pathologists, but also pathology trainees through a series of 69 cases. The grouped total kappa concordance from all participants was 0.439 (Moderate), with an overall trainee kappa of 0.417 (moderate) and an overall pathologist kappa of 0.460 (moderate). Perfect concordance among all 10 study participants was seen in 8/69 cases (11.6 %), corresponding to 5/22 Pattern A cases (22.7 %), 0/16 Pattern B cases (0 %), and 3/31 Pattern C cases (9.7 %), with similar findings between trainees and pathologists when compared within their own cohorts. Recurrence was identified in 2 Pattern A cases, indicating a potential issue with limited excisional specimens which may not fully appreciate the true biologic aggressiveness of the lesions.

Adapted generative latent diffusion models for accurate pathological analysis in chest X-ray images.

Respiratory diseases have a significant global impact, and assessing these conditions is crucial for improving patient outcomes. Chest X-ray is widely used for diagnosis, but expert evaluation can be challenging. Automatic computer-aided diagnosis methods can provide support for clinicians in these tasks. Deep learning has emerged as a set of algorithms with exceptional potential in such tasks. However, these algorithms require a vast amount of data, often scarce in medical imaging domains. In this work, a new data augmentation methodology based on adapted generative latent diffusion models is proposed to improve the performance of an automatic pathological screening in two high-impact scenarios: tuberculosis and lung nodules. The methodology is evaluated using three publicly available datasets, representative of real-world settings. An ablation study obtained the highest-performing image generation model configuration regarding the number of training steps. The results demonstrate that the novel set of generated images can improve the performance of the screening of these two highly relevant pathologies, obtaining an accuracy of 97.09%, 92.14% in each dataset of tuberculosis screening, respectively, and 82.19% in lung nodules. The proposal notably improves on previous image generation methods for data augmentation, highlighting the importance of the contribution in these critical public health challenges.

Phytoplankton detection and recognition in freshwater digital microscopy images using deep learning object detectors.

Water quality can be negatively affected by the presence of some toxic phytoplankton species, whose toxins are difficult to remove by conventional purification systems. This creates the need for periodic analyses, which are nowadays manually performed by experts. These labor-intensive processes are affected by subjectivity and expertise, causing unreliability. Some automatic systems have been proposed to address these limitations. However, most of them are based on classical image processing pipelines with not easily scalable designs. In this context, deep learning techniques are more adequate for the detection and recognition of phytoplankton specimens in multi-specimen microscopy images, as they integrate both tasks in a single end-to-end trainable module that is able to automatize the adaption to such a complex domain. In this work, we explore the use of two different object detectors: Faster R-CNN and RetinaNet, from the one-stage and two-stage paradigms respectively. We use a dataset composed of multi-specimen microscopy images captured using a systematic protocol. This allows the use of widely available optical microscopes, also avoiding manual adjustments on a per-specimen basis, which would require expert knowledge. We have made our dataset publicly available to improve the reproducibility and to foment the development of new alternatives in the field. The selected Faster R-CNN methodology reaches maximum recall levels of 95.35%, 84.69%, and 79.81%, and precisions of 94.68%, 89.30% and 82.61%, for W. naegeliana, A. spiroides, and D. sociale, respectively. The system is able to adapt to the dataset problems and improves the results overall with respect to the reference state-of-the-art work. In addition, the proposed system improves the automation and abstraction from the domain and simplifies the workflow and adjustment.

Intra- and Inter-expert Validation of an Automatic Segmentation Method for Fluid Regions Associated with Central Serous Chorioretinopathy in OCT Images.

Central Serous Chorioretinopathy (CSC) is a retinal disorder caused by the accumulation of fluid, resulting in vision distortion. The diagnosis of this disease is typically performed through Optical Coherence Tomography (OCT) imaging, which displays any fluid buildup between the retinal layers. Currently, these fluid regions are manually detected by visual inspection a time-consuming and subjective process that can be prone to errors. A series of six deep learning-based automatic segmentation architectural configurations of different levels of complexity were trained and compared in order to determine the best model intended for the automatic segmentation of CSC-related lesions in OCT images. The best performing models were then evaluated in an external validation study. Furthermore, an intra- and inter-expert analysis was conducted in order to compare the manual segmentation performed by expert ophthalmologists with the automatic segmentation provided by the models. Test results of the best performing configuration achieved a mean Dice of 0.868 ± 0.056 in the internal dataset. In the external validation set, these models achieved a level of agreement with human experts of up to 0.960 in terms of Kappa coefficient, contrasting with a value of 0.951 for agreement between human experts. Overall, the models reached a better agreement with either of the human experts than these experts with each other, suggesting that automatic segmentation models for the detection of CSC-related lesions in OCT imaging can be useful tools for assessing this disease, reducing the workload of manual inspection and leading to a more robust and objective diagnosis method.

Multi-task localization of the hemidiaphragms and lung segmentation in portable chest X-ray images of COVID-19 patients.

Background: The COVID-19 can cause long-term symptoms in the patients after they overcome the disease. Given that this disease mainly damages the respiratory system, these symptoms are often related with breathing problems that can be caused by an affected diaphragm. The diaphragmatic function can be assessed with imaging modalities like computerized tomography or chest X-ray. However, this process must be performed by expert clinicians with manual visual inspection. Moreover, during the pandemic, the clinicians were asked to prioritize the use of portable devices, preventing the risk of cross-contamination. Nevertheless, the captures of these devices are of a lower quality. Objectives: The automatic quantification of the diaphragmatic function can determine the damage of COVID-19 on each patient and assess their evolution during the recovery period, a task that could also be complemented with the lung segmentation. Methods: We propose a novel multi-task fully automatic methodology to simultaneously localize the position of the hemidiaphragms and to segment the lung boundaries with a convolutional architecture using portable chest X-ray images of COVID-19 patients. For that aim, the hemidiaphragms' landmarks are located adapting the paradigm of heatmap regression. Results: The methodology is exhaustively validated with four analyses, achieving an 82.31% ± 2.78% of accuracy when localizing the hemidiaphragms' landmarks and a Dice score of 0.9688 ± 0.0012 in lung segmentation. Conclusions: The results demonstrate that the model is able to perform both tasks simultaneously, being a helpful tool for clinicians despite the lower quality of the portable chest X-ray images.

Explainable artificial intelligence for the automated assessment of the retinal vascular tortuosity.

Retinal vascular tortuosity is an excessive bending and twisting of the blood vessels in the retina that is associated with numerous health conditions. We propose a novel methodology for the automated assessment of the retinal vascular tortuosity from color fundus images. Our methodology takes into consideration several anatomical factors to weigh the importance of each individual blood vessel. First, we use deep neural networks to produce a robust extraction of the different anatomical structures. Then, the weighting coefficients that are required for the integration of the different anatomical factors are adjusted using evolutionary computation. Finally, the proposed methodology also provides visual representations that explain the contribution of each individual blood vessel to the predicted tortuosity, hence allowing us to understand the decisions of the model. We validate our proposal in a dataset of color fundus images providing a consensus ground truth as well as the annotations of five clinical experts. Our proposal outperforms previous automated methods and offers a performance that is comparable to that of the clinical experts. Therefore, our methodology demonstrates to be a viable alternative for the assessment of the retinal vascular tortuosity. This could facilitate the use of this biomarker in clinical practice and medical research.

Multi-Adaptive Optimization for multi-task learning with deep neural networks.

Multi-task learning is a promising paradigm to leverage task interrelations during the training of deep neural networks. A key challenge in the training of multi-task networks is to adequately balance the complementary supervisory signals of multiple tasks. In that regard, although several task-balancing approaches have been proposed, they are usually limited by the use of per-task weighting schemes and do not completely address the uneven contribution of the different tasks to the network training. In contrast to classical approaches, we propose a novel Multi-Adaptive Optimization (MAO) strategy that dynamically adjusts the contribution of each task to the training of each individual parameter in the network. This automatically produces a balanced learning across tasks and across parameters, throughout the whole training and for any number of tasks. To validate our proposal, we perform comparative experiments on real-world datasets for computer vision, considering different experimental settings. These experiments allow us to analyze the performance obtained in several multi-task scenarios along with the learning balance across tasks, network layers and training steps. The results demonstrate that MAO outperforms previous task-balancing alternatives. Additionally, the performed analyses provide insights that allow us to comprehend the advantages of this novel approach for multi-task learning.

From principles to practice: Implementation of entrustable professional activities (EPAs) for surgical pathology residency education in a large academic hospital.

Over the past decade, competency-based medical education (CBME) has gained momentum in the United States to develop trainees into independent and confident physicians by the end of their training. Entrustable professional activities (EPAs) are an established methodology for assessing trainee development through an outcomes-driven rather than a time-based model. While EPAs have been utilized as an assessment tool for CBME in Europe and Canada, their validation and implementation in some medical specialties has occurred more recently in the United States. Pediatrics was the first specialty in the US to conduct a large-scale UME-GME pilot. Pathology Residency EPAs were published in 2018; however, implementation in training programs has been slow. We have piloted EPAs in our residency program's surgical pathology rotation and propose a unique set of 4 surgical pathology EPAs to track trainee preparedness for independent practice.

Iatrogenic Endometriosis of the Breast Mimicking Fat Necrosis: A Case Report.

A female patient with a history of ductal carcinoma in situ in the left breast, status-post bilateral mastectomy with deep inferior epigastric perforator artery flap reconstructive surgery, presented with a right breast asymmetry concerning for fat necrosis. Histological analysis revealed the presence of benign glands and associated stroma within fibroadipose tissue, confirmed as endometriosis by immunohistochemical analysis. Further investigation revealed that the patient had a previous diagnosis of endometriosis associated with a cesarean section scar that likely seeded the ectopic endometrial glands into a tertiary site by utilizing abdominal tissue that may have harbored endometriosis.

Automatic Segmentation of Retinal Layers in Multiple Neurodegenerative Disorder Scenarios.

Retinal Optical Coherence Tomography (OCT) allows the non-invasive direct observation of the central nervous system, enabling the measurement and extraction of biomarkers from neural tissue that can be helpful in the assessment of ocular, systemic and Neurological Disorders (ND). Deep learning models can be trained to segment the retinal layers for biomarker extraction. However, the onset of ND can have an impact on the neural tissue, which can lead to the degraded performance of models not exposed to images displaying signs of disease during training. We present a fully automatic approach for the retinal layer segmentation in multiple neurodegenerative disorder scenarios, using an annotated dataset of patients of the most prevalent NDs: Alzheimer's disease, Parkinson's disease, multiple sclerosis and essential tremor, along with healthy control patients. Furthermore, we present a two-part, comprehensive study on the effects of ND on the performance of these models. The results show that images of healthy patients may not be sufficient for the robust training of automated segmentation models intended for the analysis of ND patients, and that using images representative of different NDs can increase the model performance. These results indicate that the presence or absence of patients of ND in datasets should be taken into account when training deep learning models for retinal layer segmentation, and that the proposed approach can provide a valuable tool for the robust and reliable diagnosis in multiple scenarios of ND.

Clinicopathologic and Immunophenotypic Classification of Invasive Lobular Carcinoma with Histiocytoid Features.

INTRODUCTION: Histiocytoid lobular breast carcinoma is a rare subtype of invasive lobular carcinoma characterized by relatively bland, uniform nuclei, single small eosinophilic nucleolus, and ample granular cytoplasm. These cancers are typically triple negative, show frequent androgen receptor (AR) positivity, and are therefore theorized to represent a variant of apocrine differentiation in invasive lobular carcinoma. Anecdotal evidence suggests that these tumors have excellent outcomes, though some studies suggest a variable clinical outcome. METHODS: Inclusion criteria included women with a histologic diagnosis of invasive lobular carcinoma with histiocytoid features, regardless of immunohistochemical profile, diagnosed at our institution between 2008 and 2021 with additional tissue still available for ancillary studies. We reviewed patients meeting these criteria and investigated hematoxylin and eosin-stained slides and a panel of immunohistochemical stains (estrogen receptor, progesterone receptor, human epidermal growth factor receptor 2 [HER2], AR, endothelial growth factor receptor, and keratin 5/6), as well as outcomes including survival and metastatic disease. RESULTS: Overall, 12 eligible patients were identified. The classical immunophenotype (triple negative with AR positivity) was noted in 4 out of 12 tumors. The majority of the remaining tumors (7 out of 12) showed a luminal B immunohistochemical profile, while 1 out of 12 was HER2-enriched. No patients in the cohort died from disease-related causes and 2 out of 12 presented with distant metastatic disease during their disease course. CONCLUSION: Histiocytoid lobular breast carcinoma is a morphologic variant of lobular carcinoma with apocrine features that shows a variable immunohistochemical profile and variable clinical behavior. Further subclassification and stricter diagnostic criteria may be helpful in the distinction between truly indolent tumors and those with more aggressive clinical features.

Joint keypoint detection and description network for color fundus image registration.

Background: Retinal imaging is widely used to diagnose many diseases, both systemic and eye-specific. In these cases, image registration, which is the process of aligning images taken from different viewpoints or moments in time, is fundamental to compare different images and to assess changes in their appearance, commonly caused by disease progression. Currently, the field of color fundus registration is dominated by classical methods, as deep learning alternatives have not shown sufficient improvement over classic methods to justify the added computational cost. However, deep learning registration methods are still considered beneficial as they can be easily adapted to different modalities and devices following a data-driven learning approach. Methods: In this work, we propose a novel methodology to register color fundus images using deep learning for the joint detection and description of keypoints. In particular, we use an unsupervised neural network trained to obtain repeatable keypoints and reliable descriptors. These keypoints and descriptors allow to produce an accurate registration using RANdom SAmple Consensus (RANSAC). We train the method using the Messidor dataset and test it with the Fundus Image Registration Dataset (FIRE) dataset, both of which are publicly accessible. Results: Our work demonstrates a color fundus registration method that is robust to changes in imaging devices and capture conditions. Moreover, we conduct multiple experiments exploring several of the method's parameters to assess their impact on the registration performance. The method obtained an overall Registration Score of 0.695 for the whole FIRE dataset (0.925 for category S, 0.352 for P, and 0.726 for A). Conclusions: Our proposal improves the results of previous deep learning methods in every category and surpasses the performance of classical approaches in category A which has disease progression and thus represents the most relevant scenario for clinical practice as registration is commonly used in patients with diseases for disease monitoring purposes.

A novel 6-day cycle surgical pathology rotation improves resident satisfaction and maintains Accreditation Council for Graduate Medical Education (ACGME) milestone performance.

Surgical pathology residency training in the United States lags behind other specialties in quality control and graduated responsibility to train independent pathologists capable of seamlessly entering practice after training. We observed that our traditional 3-day-cycle surgical pathology cycle (day 1-grossing; day 2 -biopsies/frozens/preview; day 3 - sign-out) consistently and negatively impacted resident education by reducing preview time, case follow-up, immunohistochemical stain (IHC) interpretation, and molecular study integration. We aimed to create a modern surgical pathology rotation that improved performance and outcomes. We innovated our rotation to enhance resident education and ensure graduated responsibility. A novel 6-day cycle was created composed of 2 grossing days, 1 frozens/biopsies/preview days, 2 dedicated sign-out days, and 1 frozens/biopsies/case completion day. Residents completed surveys before implementing the new rotation and 6 months after implementation to track self-assessment of Accreditation Council for Graduate Medical Education (ACGME) milestone performance and internal quality control metrics. Clinical Competency Committee (CCC) annual evaluations were assessed in paired PGY levels pre- and post-intervention. After implementation, there was a statistically significant improvement in self-assessment of levels 4 and 5 of ACGME milestones and improved satisfaction of quality metrics, including time for previewing, reviewing IHC, graduated responsibility, and perceived readiness for independent practice. CCC evaluations showed overall maintained performance levels, with trends towards improvements in junior resident classes. Our 6-day cycle adequately fulfills the current demands of our sizeable academic center's surgical pathology training and can be a model for pathology residencies looking to modernize their surgical pathology rotations and resident education.

Robust multi-view approaches for retinal layer segmentation in glaucoma patients via transfer learning.

Background: Glaucoma is the leading global cause of irreversible blindness. Glaucoma patients experience a progressive deterioration of the retinal nervous tissues that begins with a loss of peripheral vision. An early diagnosis is essential in order to prevent blindness. Ophthalmologists measure the deterioration caused by this disease by assessing the retinal layers in different regions of the eye, using different optical coherence tomography (OCT) scanning patterns to extract images, generating different views from multiple parts of the retina. These images are used to measure the thickness of the retinal layers in different regions. Methods: We present two approaches for the multi-region segmentation of the retinal layers in OCT images of glaucoma patients. These approaches can extract the relevant anatomical structures for glaucoma assessment from three different OCT scan patterns: circumpapillary circle scans, macular cube scans and optic disc (OD) radial scans. By employing transfer learning to take advantage of the visual patterns present in a related domain, these approaches use state-of-the-art segmentation modules to achieve a robust, fully automatic segmentation of the retinal layers. The first approach exploits inter-view similarities by using a single module to segment all of the scan patterns, considering them as a single domain. The second approach uses view-specific modules for the segmentation of each scan pattern, automatically detecting the suitable module to analyse each image. Results: The proposed approaches produced satisfactory results with the first approach achieving a dice coefficient of 0.85±0.06 and the second one 0.87±0.08 for all segmented layers. The first approach produced the best results for the radial scans. Concurrently, the view-specific second approach achieved the best results for the better represented circle and cube scan patterns. Conclusions: To the extent of our knowledge, this is the first proposal in the literature for the multi-view segmentation of the retinal layers of glaucoma patients, demonstrating the applicability of machine learning-based systems for aiding in the diagnosis of this relevant pathology.

Comprehensive analysis of clinical data for COVID-19 outcome estimation with machine learning models.

COVID-19 is a global threat for the healthcare systems due to the rapid spread of the pathogen that causes it. In such situation, the clinicians must take important decisions, in an environment where medical resources can be insufficient. In this task, the computer-aided diagnosis systems can be very useful not only in the task of supporting the clinical decisions but also to perform relevant analyses, allowing them to understand better the disease and the factors that can identify the high risk patients. For those purposes, in this work, we use several machine learning algorithms to estimate the outcome of COVID-19 patients given their clinical information. Particularly, we perform 2 different studies: the first one estimates whether the patient is at low or at high risk of death whereas the second estimates if the patient needs hospitalization or not. The results of the analyses of this work show the most relevant features for each studied scenario, as well as the classification performance of the considered machine learning models. In particular, the XGBoost algorithm is able to estimate the need for hospitalization of a patient with an AUC-ROC of 0 . 8415 ± 0 . 0217 while it can also estimate the risk of death with an AUC-ROC of 0 . 7992 ± 0 . 0104 . Results have demonstrated the great potential of the proposal to determine those patients that need a greater amount of medical resources for being at a higher risk. This provides the healthcare services with a tool to better manage their resources.

Computer-assisted evaluation of retinal vessel tortuosity in moderate-to-late preterm children.

PURPOSE: Since very preterm children often have increased retinal tortuosity that may indicate decisive architectural changes in the systemic microvascular network, we used a new semi-automatic software to measure retinal vessel tortuosity on fundus digital images of moderate-to-late preterm (MLP) children. METHODS: In this observational case-control study, the global and local tortuosity parameters of retinal vessels were evaluated on fundus photographs of 36 MLP children and 36 age- and sex-matched controls. The associations between birth parameters and parameters reflecting retinal vessel tortuosity were evaluated using correlation analysis. RESULTS: Even after incorporation of anatomical factors, the global and local tortuosity parameters were not significantly different between groups. The MLP group showed a smaller arteriolar caliber (0.53 ± 0.2) than the controls (0.56 ± 0.2; p = 0.013). Other local tortuosity parameters, such as vessel length, distance to fovea, and distance to optic disc, were not significantly different between arteries and veins. Tortuosity in both groups was higher among vessels closer to the fovea (r = -0.077, p < 0.001) and the optic disc (r = -0.0544, p = 0.009). Global tortuosity showed a weakly positive correlation with gestational age and a weakly negative correlation with birth weight in both groups. CONCLUSION: MLP patients did not display increased vessel tortuosity in comparison with the controls; however, the arteriolar caliber in the MLP group was smaller than that in children born full-term. Larger studies should confirm this finding and explore associations between cardiovascular and metabolic status and retinal vessel geometry in MLP children.

A new generative approach for optical coherence tomography data scarcity: unpaired mutual conversion between scanning presets.

In optical coherence tomography (OCT), there is a trade-off between the scanning time and image quality, leading to a scarcity of high quality data. OCT platforms provide different scanning presets, producing visually distinct images, limiting their compatibility. In this work, a fully automatic methodology for the unpaired visual conversion of the two most prevalent scanning presets is proposed. Using contrastive unpaired translation generative adversarial architectures, low quality images acquired with the faster Macular Cube preset can be converted to the visual style of high visibility Seven Lines scans and vice-versa. This modifies the visual appearance of the OCT images generated by each preset while preserving natural tissue structure. The quality of original and synthetic generated images was compared using BRISQUE. The synthetic generated images achieved very similar scores to original images of their target preset. The generative models were validated in automatic and expert separability tests. These models demonstrated they were able to replicate the genuine look of the original images. This methodology has the potential to create multi-preset datasets with which to train robust computer-aided diagnosis systems by exposing them to the visual features of different presets they may encounter in real clinical scenarios without having to obtain additional data. Graphical Abstract Unpaired mutual conversion between scanning presets. Two generative adversarial models are trained for the conversion of OCT images into images of another scanning preset, replicating the visual features that characterise said preset.

Fully automatic segmentation and monitoring of choriocapillaris flow voids in OCTA images.

Optical coherence tomography angiography (OCTA) is a non-invasive ophthalmic imaging modality that is widely used in clinical practice. Recent technological advances in OCTA allow imaging of blood flow deeper than the retinal layers, at the level of the choriocapillaris (CC), where a granular image is obtained showing a pattern of bright areas, representing blood flow, and a pattern of small dark regions, called flow voids (FVs). Several clinical studies have reported a close correlation between abnormal FVs distribution and multiple diseases, so quantifying changes in FVs distribution in CC has become an area of interest for many clinicians. However, CC OCTA images present very complex features that make it difficult to correctly compare FVs during the monitoring of a patient. In this work, we propose fully automatic approaches for the segmentation and monitoring of FVs in CC OCTA images. First, a baseline approach, in which a fully automatic segmentation methodology based on local contrast enhancement and global thresholding is proposed to segment FVs and measure changes in their distribution in a straightforward manner. Second, a robust approach in which, prior to the use of our segmentation methodology, an unsupervised trained neural network is used to perform a deformable registration that aligns inconsistencies between images acquired at different time instants. The proposed approaches were tested with CC OCTA images collected during a clinical study on the response to photodynamic therapy in patients affected by chronic central serous chorioretinopathy (CSC), demonstrating their clinical utility. The results showed that both approaches are accurate and robust, surpassing the state of the art, therefore improving the efficacy of FVs as a biomarker to monitor the patient treatments. This gives great potential for the clinical use of our methods, with the possibility of extending their use to other pathologies or treatments associated with this type of imaging.

The role of PRAME and NY-ESO-1 as potential therapeutic and prognostic biomarkers in triple-negative breast carcinomas.

PRAME and NY-ESO-1 are cancer-testis antigens (CTAs) reported to be highly enriched in triple-negative breast cancers (TNBCs), against which vaccines and immunotherapies are currently being developed. This study aims to analyze PRAME and NY-ESO-1 expression in TNBCs and their correlation with clinical outcomes. This is a retrospective cohort study of TNBC patients who have undergone neoadjuvant chemotherapy. PRAME and NY-ESO-1 expression were assessed on pre-therapy biopsies as H-scores (percentage x intensity) with final H scores of 2-3 considered as positive. Association between expression and pathologic complete response (pCR), metastasis, and residual cancer burden (RCB) were assessed via logistic regression. Cox proportional hazards models were used to assess the association with progression-free survival. P-values < 0.05 were considered statistically significant. Sixty-three percent of 76 patients were positive for PRAME. In contrast, only 5 % were positive for NY-ESO-1. PRAME positivity was significantly associated with a lower likelihood of early metastatic disease (OR = 0.24, 95 % CI 0.08-0.62; P = 0.005). However, it was not significantly associated with pCR, RCB category, or progression-free survival. NY-ESO1 score was not significantly associated with early metastatic disease, pCR, RCB category, or progression-free survival. Our results suggest that PRAME positivity may be associated with a lower risk of early metastasis in TNBCs, but not with response to neoadjuvant chemotherapy or progression-free survival. The high expression of PRAME in TNBCs makes it a potential therapeutic target, while NY-ESO1 appears to be a less useful marker. However, further larger studies are needed to ascertain the utility of these markers.