An interpretable decision-support model for breast cancer diagnosis using histopathology images.

Microscopic examination of biopsy tissue slides is perceived as the gold-standard methodology for the confirmation of presence of cancer cells. Manual analysis of an overwhelming inflow of tissue slides is highly susceptible to misreading of tissue slides by pathologists. A computerized framework for histopathology image analysis is conceived as a diagnostic tool that greatly benefits pathologists, augmenting definitive diagnosis of cancer. Convolutional Neural Network (CNN) turned out to be the most adaptable and effective technique in the detection of abnormal pathologic histology. Despite their high sensitivity and predictive power, clinical translation is constrained by a lack of intelligible insights into the prediction. A computer-aided system that can offer a definitive diagnosis and interpretability is therefore highly desirable. Conventional visual explanatory techniques, Class Activation Mapping (CAM), combined with CNN models offers interpretable decision making. The major challenge in CAM is, it cannot be optimized to create the best visualization map. CAM also decreases the performance of the CNN models. To address this challenge, we introduce a novel interpretable decision-support model using CNN with a trainable attention mechanism using response-based feed-forward visual explanation. We introduce a variant of DarkNet19 CNN model for the classification of histopathology images. In order to achieve visual interpretation as well as boost the performance of the DarkNet19 model, an attention branch is integrated with DarkNet19 network forming Attention Branch Network (ABN). The attention branch uses a convolution layer of DarkNet19 and Global Average Pooling (GAP) to model the context of the visual features and generate a heatmap to identify the region of interest. Finally, the perception branch is constituted using a fully connected layer to classify images. We trained and validated our model using more than 7000 breast cancer biopsy slide images from an openly available dataset and achieved 98.7% accuracy in the binary classification of histopathology images. The observations substantiated the enhanced clinical interpretability of the DarkNet19 CNN model, supervened by the attention branch, besides delivering a 3%-4% performance boost of the baseline model. The cancer regions highlighted by the proposed model correlate well with the findings of an expert pathologist. The coalesced approach of unifying attention branch with the CNN model capacitates pathologists with augmented diagnostic interpretability of histological images with no detriment to state-of-art performance. The model's proficiency in pinpointing the region of interest is an added bonus that can lead to accurate clinical translation of deep learning models that underscore clinical decision support.

Unchartered waters: Significance of fall in Ki67 index after short-term preoperative endocrine therapy in early breast cancers.

BACKGROUND: Endocrine treatment for breast cancer acts largely by inhibiting tumor cell proliferation. The biomarker Ki67 is linked to the proliferative index of the tumour. OBJECTIVE: To identify the factors affecting the fall in Ki67 value in early-stage hormone receptor (HR) positive breast cancer patients receiving short-term preoperative endocrine therapy in an Indian cohort. METHODS: Women with hormone receptor positive, invasive, nonmetastatic, and early breast cancer (4 week) did not affect the fall in Ki67. CONCLUSION: Preoperative therapy with Letrozole resulted in a more significant fall in Ki67, as compared to therapy with Tamoxifen. Determining the fall in Ki67 value in response to preoperative endocrine therapy could provide an insight into the response to endocrine therapy in luminal breast cancer.

Fall in Ki67 Index After Short-Term Preoperative Letrozole: a Gateway to Assess the Response in Hormone-Positive Early Breast Cancers.

Endocrine treatment for breast cancer acts largely by inhibiting tumor cell proliferation. The study aimed to explore the fall in proliferative marker Ki67 in patients receiving preoperative endocrine therapy and the factors associated with it. A prospective series of hormone receptor-positive postmenopausal women with early N0/N1 breast cancer were enrolled. Patients were requested to take letrozole OD while they await surgery. The fall in Ki67 after the endocrine therapy was defined as the percentage of the difference between the pre-and postoperative Ki67 value with the preoperative Ki67. Sixty cases matched the criteria of which 41 (68.3%) of women showed a good response to preoperative letrozole (fall in Ki67 > 50%; p-value < 0.001). The average mean fall in Ki67 was 57.083 ± 37.97. Postoperative Ki67 after the therapy was less than 10% in 39 (65%) patients. Ten patients (16.6%) had a low Ki67 index at baseline, which continued to remain low after preoperative endocrine therapy. The duration of the therapy did not affect the percentage of Ki67 fall in our study. Short-term changes in the Ki67 index in the neoadjuvant settings may predict outcomes during adjuvant use of the same treatment. Proliferation index on residual tumor holds prognostic importance, and our results reflect that greater attention should be given to the percentage of reduction of Ki67, rather than focusing purely on a fixed value. This could help predict patients who respond well to endocrine therapy, while those who respond poorly may require further adjuvant treatment.

Utility of Fluorine-18-Labeled Fluorodeoxyglucose Positron Emission Tomography/Computed Tomography in Diagnosis of Erdheim-Chester Disease with Multisystemic Involvement.

Erdheim‒Chester disease (ECD) is a rare non-Langerhans' cell histiocytic proliferative disorder of unknown origin with multisystemic predilection. It commonly affects adults in the fifth-seventh decades of life, with male preponderance, and has nonspecific clinical manifestations. Presence of characteristic radiological findings and demonstration of CD68 positive xanthogranulomatous infiltrates in histology clinches the diagnosis. Nevertheless, being a nonmalignant condition, it might be fatal due to multiorgan dysfunction. Hence, timely diagnosis and initiation of treatment with corticosteroids, immunosuppressants, or tyrosine kinase inhibitors are of paramount importance. We present a case of ECD with multisystemic involvement, who was initially evaluated for the left lung mass and treated as tuberculosis, where fluorine-18-labeled fluorodeoxyglucose positron emission tomography/computed tomography aided in targeting the metabolically active site for biopsy as well as assessing the multisystemic involvement.

Recurrent DHA nephropathy in renal allograft-revisiting clinicopathological aspects of a rare entity.

BACKGROUND: Adenine phosphoribosyltransferase (APRT) enzyme deficiency is a rare autosomal recessive disorder of purine metabolism affecting mainly the kidneys. It can present at any age with varying degrees of acute and chronic renal damage. Though xanthine dehydrogenase inhibitors offer effective control over the disease process, delay in diagnosis and treatment often lead to compromised function of native and even graft kidneys. METHODS: We have done a retrospective search of records of renal biopsies reported at our center during the 5-year period from 2014 to 2018 to identify biopsies with 2,8-dihydroxyadenine crystal deposits. The demographic, clinical, and histopathological findings in these cases were studied and reviewed in the light of available literature. RESULTS: Of 9059 renal biopsies received during the study period, 3 cases had the rare 2,8- dihydroxyadenine (DHA) crystals. All of them were diagnosed for the first time on allograft biopsies. CONCLUSION: A high index of clinical suspicion together with the characteristic microscopic appearance of crystals on renal biopsy and urine microscopy can clinch the diagnosis of this rare disease. Hence, improving awareness about this entity among clinicians and pathologists is extremely important.

Primary granulosa cell tumor of retroperitoneal origin: A rare presentation with emphasis on cytomorphology.

The most frequently occurring retroperitoneal tumors are those of the kidneys, adrenal glands, and the pancreas. A primary retroperitoneal tumor composed of granulosa cells and developing far away from the normal location of the ovary is less frequently observed. A 69-year-old female patient presented with abdominal discomfort. Computerized tomography (CT) of the abdomen revealed a solid heterogeneous mass lesion measuring 11.2 cm × 8 cm × 12 cm consistent with retroperitoneal hematoma. Ultrasonography (USG)-guided aspiration smears revealed cytological features suggestive of adult-type granulosa cell tumor (AGCT). As the patient had a history of hysterectomy with bilateral salpingo-oophorectomy 22 years ago for leiomyoma, a diagnosis of extraovarian AGCT was made. Intraoperatively, the tumor was removed in piecemeal that showed yellowish areas with extensive necrosis and hemorrhage. Histopathological examination of the excised mass and inhibin positivity confirmed the diagnosis. Primary retroperitoneal extraovarian GCT is a rare tumor with only 12 cases reported in medical literature in English.

Disseminated gastric carcinoma in disguise-presentation as microangiopathic haemolytic anemia with bone marrow necrosis.

Cancer related microangiopathic hemolytic anemia (MAHA) and bone marrow necrosis (BMN) are two rare and distinct clinicopathological entities reported in patients with disseminated malignancy. When they emerge as initial findings in a previously undetected case of malignancy, the diagnosis is often missed resulting in inappropriate management. We report the extremely rare association of cancer related MAHA and BMN as the first presentation of unsuspected disseminated gastric carcinoma in a 63 years old male. Early institution of chemotherapy is the only effective option for both these conditions, though with little improvement in outcome. This emphasises the need for increased awareness of such presentations in advanced malignancies.

Primary cardiac sarcoma with metastases to thyroid and brain.

Primary cardiac sarcomas are extremely rare with only a few large case series and isolated case reports in the literature. In spite of their aggressive nature with high chances of local recurrence and distant metastases, these tumors and their treatment strategies remain incompletely defined. We report an unsuspected case of primary cardiac pleomorphic undifferentiated sarcoma in a 52-year-old female who presented with progressive dyspnea and palpitation. In the postoperative period, she was detected to have secondaries in brain and thyroid, confirmed thereafter, by fine-needle aspiration cytology.