Summary Background Treatment with bisphosphonates decreases bone loss and can increase disease-free survival in patients with breast cancer. The aim of our study was to assess the effect of zoledronic acid on clearance of disseminated tumour cells (DTCs) from the bone marrow in women undergoing neoadjuvant chemotherapy for breast cancer. Methods Patients were recruited for this open-label, phase 2 randomised trial between March 17, 2003, and May 19, 2006, at a single centre. Eligible patients had clinical stage II–III (≥T2 and/or ≥N1) newly diagnosed breast cancer, Eastern Cooperative Oncology Group performance status of 0 or 1, and normal cardiac, renal, and liver function. 120 women were randomly assigned, using allocation concealment, to receive 4 mg zoledronic acid intravenously every 3 weeks (n=60), or no zoledronic acid (n=60), for 1 year concomitant with four cycles of neoadjuvant epirubicin (75 mg/m²) plus docetaxel (75 mg/m²) and two cycles of adjuvant epirubicin plus docetaxel. The primary endpoint was the number of patients with detectable DTCs at 3 months. Final analysis was done 1 year after the last patient was enrolled. Analyses were done for all patients with available data at 3 months. This study is registered with ClinicalTrials.gov, number NCT00242203. Findings Of the 120 patients initially enrolled, one withdrew after signing consent and one patient’s baseline bone marrow was not available. Both of these patients were in the control group. At 3 months, 109 bone-marrow samples were available for analysis. In the zoledronic acid group, bone marrow was not collected from one patient because of disease progression, one patient was taken off study because of severe diarrhoea, and two patients had not consented at the time of surgery. In the control group, bone marrow was not collected from two patients because of disease progression, one patient withdrew consent, and three patients were not consented at the time of surgery. At baseline, DTCs were detected in 26 of 60 patients in the zoledronic acid group and 28 of 58 patients in the control group. At 3 months, 17 of 56 patients receiving zoledronic acid versus 25 of 53 patients who did not receive zoledronic acid had detectable DTCs (p=0·054). The most common grade 3–4 toxicities were infection (five of 60 patients in the zoledronic acid group and six of 59 in the control group) and thrombosis (five of 60 in the zoledronic acid and two of 59 in the control group). There was one documented case of osteonecrosis in the zoledronic acid group. Interpretation Zoledronic acid administered with chemotherapy resulted in a decreased proportion of patients with DTCs detected in the bone marrow at the time of surgery. Our study supports the hypothesis that the antimetastatic effects of zoledronic acid may be through effects on DTCs. Funding Novartis Pharmaceuticals and Pfi zer Inc.

Artificial intelligence (AI) algorithms continue to rival human performance on a variety of clinical tasks, while their actual impact on human diagnosticians, when incorporated into clinical workflows, remains relatively unexplored. In this study, we developed a deep learning-based assistant to help pathologists differentiate between two subtypes of primary liver cancer, hepatocellular carcinoma and cholangiocarcinoma, on hematoxylin and eosin-stained whole-slide images (WSI), and evaluated its effect on the diagnostic performance of 11 pathologists with varying levels of expertise. Our model achieved accuracies of 0.885 on a validation set of 26 WSI, and 0.842 on an independent test set of 80 WSI. Although use of the assistant did not change the mean accuracy of the 11 pathologists (p = 0.184, OR = 1.281), it significantly improved the accuracy (p = 0.045, OR = 1.499) of a subset of nine pathologists who fell within well-defined experience levels (GI subspecialists, non-GI subspecialists, and trainees). In the assisted state, model accuracy significantly impacted the diagnostic decisions of all 11 pathologists. As expected, when the model's prediction was correct, assistance significantly improved accuracy (p = 0.000, OR = 4.289), whereas when the model's prediction was incorrect, assistance significantly decreased accuracy (p = 0.000, OR = 0.253), with both effects holding across all pathologist experience levels and case difficulty levels. Our results highlight the challenges of translating AI models into the clinical setting, and emphasize the importance of taking into account potential unintended negative consequences of model assistance when designing and testing medical AI-assistance tools.npj Digital Medicine (2020) 3:23 ; https://doi.

Objective: Fine needle aspiration (FNA) is a reliable method in the initial assessment of thyroid nodules. The purpose of this study was to evaluate the causes for discordance between the interpretation on FNA and the pathologic findings in the resected thyroid. Methods: A computer search of all thyroidectomy specimens with previous FNA from January 1998 to December 2001 was obtained from the files of the Lauren V. Ackerman laboratory of surgical pathology, Barnes-Jewish Hospital. Excluded from the study were those FNAs performed for suspected and confirmed metastatic disease to the thyroid as well as those cases unavailable for review. A total of 45 FNA cases were identified with cytologic and histologic discrepancies. Results: Of the 1253 individual thyroid FNA performed during the study period, 255 patients (20%) subsequently had an open surgical procedure on the thyroid. Of those who underwent surgery, 196 cases (77%) were concordant, whereas 45 patients (18%) were discordant, and 14 cases were excluded due to unavailability of slides for review (for example, returned consult slides). The causes of the 45 discordant cases were: 20 cases (44%) were unsatisfactory for diagnosis, 14 cases (31%) were due to interpretation error (false positive), and 11 cases (24%) were due to sampling error (false negative). Conclusions: The most common causes of our discrepant cases are those whose FNA diagnosis was interpreted as "unsatisfactory for diagnosis," in 20 (7.8%) of 255 surgical cases. The false negative rate due to sampling error in 11 (4%) of 255 cases was mainly due to the presence of microscopic papillary thyroid carcinoma (PTC); the false positive rate was due to interpretation error in 14 (6%) of 255 cases, and those were explained by the occurrence of overlapping cytologic features among adenomatous nodules, follicular neoplasms, the follicular variant of PTC, and Hashimoto's thyroiditis. 35

Purpose: Micrometastatic cells detected in the bone marrow have prognostic significance in breast cancer.These cells are heterogeneous and likely do not exhibit uniform biological behavior.To understand the molecular diversity of disseminated cancer cells that reside inbone marrow, we enriched this cell population and did global gene expression profiling in the context of a prospective clinical trial involving women with clinical stage II/IIIbreast cancer undergoing neoadjuvant chemotherapy. Experimental Design: Enrichment of TACSTD1 (EpCAM)^expressing cells from bone marrow of breast cancer patients was achieved using immunomagnetic beads. Gene expression profiles were compared between enriched cell populations and whole bone marrow from 5 normal volunteers and 23 breast cancer patients after neoadjuvant chemotherapy treatment. Enriched cells from bone marrow samples of breast cancer patients before treatment or at 1year follow-up were also analyzed (total of 87 data sets). The expression of transcripts specifically detected in enriched cell populations from breast cancer patients was correlated with 1-year clinical outcome using quantitative reverse transcription-PCR in an independent cohort of bone marrow samples. Results: Analysis of EpCAM-enriched bone marrow cells revealed specific expression of a subgroup of transcripts, including the metastasis regulator, TWIST1. Most transcripts identified, including TWIST1, were not expressed in enriched populations of bone marrow from normal volunteers, suggesting that this expression profile reflects a signature of breast cancer bone marrow micrometastases that persist after chemotherapy. In an independent set of bone marrow samples obtained before any treatment,TWIST1 expression correlated with early disease relapse. Conclusions: Disseminated breast cancer cells present in bone marrow after chemotherapy possess unique transcriptional signatures. Genes whose expression is overrepresented in these cell populations, such asTWIST1, may prove to be excellent markers of early distant relapse in breast cancer patients.

BACKGROUND Endoscopic ultrasound‐guided fine‐needle aspiration biopsy (EUS‐FNAB) of small pancreatic lesions that are undetectable by computed tomography has gained wide acceptance for the procurement of cells for diagnostic purposes. However, this technique is not without difficulty. The authors examined the sensitivity, specificity, and positive and negative predictive values (PPV and NPV, respectively) of this technique in the evaluation of patients with pancreatic biliary duct strictures/masses. The authors were interested in reviewing their cases of pancreatic adenocarcinoma of ductal type and finding the sources of their false‐negative cases. METHODS A computer search was performed between January 1998 and July 2001. For the last 3 years, a total of 80 cases of suspected ductal adenocarcinoma of the pancreas was identified. Thirty‐four patients (42%) underwent a subsequent Whipple procedure or biopsy. Cytologic and histologic correlation was performed in these cases. The rest of the 23 patients (29%) considered to be positive and the 23 patients (29%) considered to be negative underwent no subsequent biopsy and were followed clinically. Cases termed “suspicious” on cytology were considered positive and those termed “atypical cytology” were considered negative in the authors' final calculation. The causes of the false‐negative diagnoses were evaluated carefully. RESULTS Of the 34 cases followed with subsequent tissue biopsy or surgery; 12 were confirmed to be positive, 12 were confirmed to be negative, and 10 were considered to be false‐negative. Previously identified cytomorphologic features of malignancy were used to review all cases. These features were: loss of the honeycomb pattern (100%), anisonucleosis (100%), nuclear contour irregularity (100%), a high nuclear/cytoplasmic ratio (100%), paranuclear chromatin clearing (77%), and the presence of prominent nucleoli in the absence of inflammatory cells (77%). The causes of the 10 false‐negative cases were technical difficulty of procuring material in 6 cases, the nature of the lesion in 2 cases, and the scarcity of lesional tissue in 2 cases. CONCLUSIONS Using strict cytoarchitectural and cytomorphologic criteria of malignancy for ductal pancreatic lesions previously described in the literature, the sensitivity of this technique at the study institution was 78% with a specificity of 100%. The PPV and NPV of this technique were 100% and 78%, respectively. The most common causes of the false‐negative results in descending order were the technical aspect of the procedure, the size and nature of the lesion, and the scarcity of lesional tissue. Cancer (Cancer Cytopathol) 2002;96:000–000. © 2002 American Cancer Society. DOI 10.1002/cncr.10759

No abstract

Background In all, 20% of fine‐needle aspiration (FNA) biopsies of thyroid nodules have an indeterminate diagnosis; of these, 80% are found to be benign after thyroidectomy. Some previous reports indicate that positron emission tomography (PET) with 18F‐fluorodeoxyglucose (FDG) imaging may predict malignancy status. We now report results on the first 51 patients in the largest prospective study of FDG‐PET in patients with an indeterminate thyroid nodule FNA. Methods Eligible patients had a dominant thyroid nodule that was palpable or ≥1 cm in greatest dimension as seen by ultrasonography, and indeterminate histology of the FNA biopsy specimen. Participants underwent preoperative neck FDG‐PET alone or FDG‐PET with computed tomography (FDG‐PET/CT). Images were evaluated qualitatively and semiquantitatively using the maximum standardized uptake value (SUVmax). Final diagnosis was determined by histopathologic analysis after thyroidectomy. Descriptive statistical analysis was performed. Results A total of 51 patients underwent preoperative FDG‐PET or FDG‐PET/CT. Studies without focally increased uptake localized to the lesion were considered negative. For all lesions (10 malignant, 41 benign), the sensitivity, specificity, positive‐predictive value (PPV), and negative‐predictive value (NPV) were 80%, 61%, 33%, and 93%, respectively. Postoperatively, two malignant and six benign lesions were found to be <1 cm by pathology examination; one lesion was not measured. When these lesions were excluded, the sensitivity, specificity, PPV, and NPV were 100%, 59%, 36%, and 100%, respectively. Conclusions Based on these preliminary data, FDG‐PET may have a role in excluding malignancy in thyroid nodules with an indeterminate FNA biopsy. This finding justifies ongoing accrual to our target population of 125 participants.

BackgroundFocal adhesion Kinase (FAK) is a nonreceptor protein tyrosine kinase that is overexpressed in tumors and plays a significant role in tumor survival and metastasis. The purpose of the study is to perform correlation of FAK expression with patient prognostic factors using tissue microarrays (TMA) samples.MethodsWe analyzed FAK expression by immunohistochemical staining in 196 breast primary tumor samples from stage II-IV patients and in 117 metastatic tissues matched to the primary tumors using TMA that were stained with FAK monoclonal antibody.ResultsHigh FAK expression in primary tumors was associated with a younger age of patients (p = 0.033), lymphovascular invasion (p = 0.001) and with the triple-negative phenotype (p = 0.033). FAK expression in 117 metastatic tissues positively correlated with FAK expression in matched primary tumors by Spearman correlation analysis. In addition, a strong positive correlation was observed between high FAK expression and shorter overall survival and progression free survival in patients with metastatic tumors.ConclusionsThe data demonstrate a high potential for FAK as a therapeutic target, especially in triple-negative breast cancer patients with high FAK expression.