Abstract
For newly diagnosed breast cancer, estrogen receptor status (ERS) is a key molecular marker used for prognosis and treatment decisions. During clinical management, ERS is determined by ...pathologists from immunohistochemistry (IHC) staining of biopsied tissue for the targeted receptor, which highlights the presence of cellular surface antigens. This is an expensive, time-consuming process which introduces discordance in results due to variability in IHC preparation and pathologist subjectivity. In contrast, hematoxylin and eosin (H&E) staining—which highlights cellular morphology—is quick, less expensive, and less variable in preparation. Here we show that machine learning can determine molecular marker status, as assessed by hormone receptors, directly from cellular morphology. We develop a multiple instance learning-based deep neural network that determines ERS from H&E-stained whole slide images (WSI). Our algorithm—trained strictly with WSI-level annotations—is accurate on a varied, multi-country dataset of 3,474 patients, achieving an area under the curve (AUC) of 0.92 for sensitivity and specificity. Our approach has the potential to augment clinicians’ capabilities in cancer prognosis and theragnosis by harnessing biological signals imperceptible to the human eye.
Because histologic types are subjective and difficult to reproduce between pathologists, tissue morphology often takes a back seat to molecular testing for the selection of breast cancer treatments. ...This work explores whether a deep-learning algorithm can learn objective histologic H&E features that predict the clinical subtypes of breast cancer, as assessed by immunostaining for estrogen, progesterone, and Her2 receptors (ER/PR/Her2). Translating deep learning to this and related problems in histopathology presents a challenge due to the lack of large, well-annotated data sets, which are typically required for the algorithms to learn statistically significant discriminatory patterns. To overcome this limitation, we introduce the concept of "tissue fingerprints," which leverages large, unannotated datasets in a label-free manner to learn H&E features that can distinguish one patient from another. The hypothesis is that training the algorithm to learn the morphological differences between patients will implicitly teach it about the biologic variation between them. Following this training internship, we used the features the network learned, which we call "fingerprints," to predict ER, PR, and Her2 status in two datasets. Despite the discovery dataset being relatively small by the standards of the machine learning community (n = 939), fingerprints enabled the determination of ER, PR, and Her2 status from whole slide H&E images with 0.89 AUC (ER), 0.81 AUC (PR), and 0.79 AUC (Her2) on a large, independent test set (n = 2531). Tissue fingerprints are concise but meaningful histopathologic image representations that capture biological information and may enable machine learning algorithms that go beyond the traditional ER/PR/Her2 clinical groupings by directly predicting theragnosis.
Mass-spectrometry-based proteomics has become an important component of biological research. Numerous proteomics methods have been developed to identify and quantify the proteins in biological and ...clinical samples
1
, identify pathways affected by endogenous and exogenous perturbations
2
, and characterize protein complexes
3
. Despite successes, the interpretation of vast proteomics datasets remains a challenge. There have been several calls for improvements and standardization of proteomics data analysis frameworks, as well as for an application-programming interface for proteomics data access
4
,
5
. In response, we have developed the ProteoWizard Toolkit, a robust set of open-source, software libraries and applications designed to facilitate proteomics research. The libraries implement the first-ever, non-commercial, unified data access interface for proteomics, bridging field-standard open formats and all common vendor formats. In addition, diverse software classes enable rapid development of vendor-agnostic proteomics software. Additionally, ProteoWizard projects and applications, building upon the core libraries, are becoming standard tools for enabling significant proteomics inquiries.
Purpose: Ansamycin antibiotics, including 17allylamino-17-demethoxygeldanamycin (17-AAG), inhibit Hsp90 function and cause the selective
degradation of signaling proteins that require this chaperone ...for folding. Because mutations in the androgen receptor (AR)
and activation of HER2 and Akt may account, in part, for prostate cancer progression after castration or treatment with antiandrogens,
we sought to determine whether an inhibitor of Hsp90 function could degrade these Hsp90 client proteins and inhibit the growth
of prostate cancer xenografts with an acceptable therapeutic index.
Experimental Design: The effect of 17-AAG on the expression of Hsp90 regulated signaling proteins in prostate cancer cells and xenografts was
determined. The pharmacodynamics of target protein degradation was associated with the toxicology and antitumor activity of
the drug.
Results: 17-AAG caused the degradation of HER2, Akt, and both mutant and wild-type AR and the retinoblastoma-dependent G 1 growth arrest of prostate cancer cells. At nontoxic doses, 17-AAG caused a dose-dependent decline in AR, HER2, and Akt expression
in prostate cancer xenografts. This decline was rapid, with a 97% loss of HER2 and an 80% loss of AR expression at 4 h. 17-AAG
treatment at doses sufficient to induce AR, HER2, and Akt degradation resulted in the dose-dependent inhibition of androgen-dependent
and -independent prostate cancer xenograft growth without toxicity.
Conclusions: These data demonstrate that, at a tolerable dose, inhibition of Hsp90 function by 17-AAG results in a marked reduction in
HER2, AR, and Akt expression and inhibition of prostate tumor growth in mice. These results suggest that this drug may represent
a new strategy for the treatment of prostate cancer.
Pertuzumab, a recombinant humanized monoclonal antibody (2C4), binds to extracellular domain II of the HER-2 receptor and blocks its ability to dimerize with other HER receptors. Pertuzumab ...represents a new class of targeted therapeutics known as HER dimerization inhibitors. A clinical study was conducted to investigate safety and pharmacokinetics of pertuzumab and to perform a preliminary assessment of HER dimerization inhibition as a treatment strategy.
Patients with incurable, locally advanced, recurrent or metastatic solid tumors that had progressed during or after standard therapy were recruited to a dose-escalation study of pertuzumab (0.5 to 15 mg/kg) given intravenously every 3 weeks.
Twenty-one patients received pertuzumab and 19 completed at least two cycles. Pertuzumab was well tolerated. Overall, 365 adverse events were reported and 122 considered to be possibly drug related. Of these, 116 were of grade 1 to 2 intensity. The pharmacokinetics of pertuzumab were similar to other humanized immunoglobulin G antibodies, supporting a 3-week dosing regimen. Trough plasma concentrations were in excess of target concentrations at doses greater than 5 mg/kg. Two patients, one with ovarian cancer (5.0 mg/kg) and one with pancreatic islet cell carcinoma (15.0 mg/kg), achieved a partial response. Responses were documented by Response Evaluation Criteria in Solid Tumors after 1.5 and 6 months of pertuzumab therapy, and lasted for 11 and 10 months, respectively. Stable disease lasting for more than 2.5 months (range, 2.6 to 5.5 months) was observed in six patients.
These results demonstrate that pertuzumab is well tolerated, has a pharmacokinetic profile which supports 3-week dosing, and is clinically active, suggesting that inhibition of dimerization may be an effective anticancer strategy.
Coding mutations in the AR (androgen receptor) gene have been identified in tissue samples from patients with advanced prostate cancer and represent a possible mechanism underlying the development of ...castration-resistant prostate cancer (CRPC). There is a paucity of tumor-derived tissue available for molecular studies of CRPC patients. Circulating tumor cells (CTCs) in the blood of CRPC patients represent a possible avenue for interrogating the disease of such patients.
Circulating tumor cells were captured with the CellSearch Circulating Tumor Cell (CTC) Kit and with the CellSearch Profile Kit plus Qiagen's AllPrep DNA/RNA Micro Kit for the measurement of the CTC count per 7.5 mL of blood and for the isolation of nucleic acids, respectively. The AR gene was amplified by the PCR, and mutation status and relative abundance were analyzed by applying Transgenomic's WAVE denaturing HPLC technology followed by direct sequencing.
AR mutations were detected in 20 of 35 CRPC patients; 19 missense mutations, 2 silent mutations, 5 deletions, and 1 insertion were observed. The relative abundance of the mutants in the amplified products ranged from 5% to 50%. Many of the AR mutations were identified in surgical biopsies or at autopsy and were associated with resistance to androgen-directed therapies.
AR mutations can be identified in CTC-enriched peripheral blood samples from CRPC patients. This approach has the potential to open new perspectives in understanding CTCs and the mechanisms for tumor progression and metastasis in CRPC.
ErbB2 is a ligand-less member of the ErbB receptor family that functions as a coreceptor with EGFR, ErbB3, and ErbB4. Here, we describe an approach to target ErbB2's role as a coreceptor using a ...monoclonal antibody, 2C4, which sterically hinders ErbB2's recruitment into ErbB ligand complexes. Inhibition of ligand-dependent ErbB2 signaling by 2C4 occurs in both low- and high-ErbB2-expressing systems. Since the ErbB3 receptor contains an inactive tyrosine kinase domain, 2C4 is very effective in blocking heregulin-mediated ErbB3-ErbB2 signaling. We demonstrate that the in vitro and in vivo growth of several breast and prostate tumor models is inhibited by 2C4 treatment.
We are in the midst of a technological revolution that is providing new insights into human biology and cancer. In this era of big data, we are amassing large amounts of information that is ...transforming how we approach cancer treatment and prevention. Enactment of the Cancer Moonshot within the 21st Century Cures Act in the USA arrived at a propitious moment in the advancement of knowledge, providing nearly US$2 billion of funding for cancer research and precision medicine. In 2016, the Blue Ribbon Panel (BRP) set out a roadmap of recommendations designed to exploit new advances in cancer diagnosis, prevention, and treatment. Those recommendations provided a high-level view of how to accelerate the conversion of new scientific discoveries into effective treatments and prevention for cancer. The US National Cancer Institute is already implementing some of those recommendations. As experts in the priority areas identified by the BRP, we bolster those recommendations to implement this important scientific roadmap. In this Commission, we examine the BRP recommendations in greater detail and expand the discussion to include additional priority areas, including surgical oncology, radiation oncology, imaging, health systems and health disparities, regulation and financing, population science, and oncopolicy. We prioritise areas of research in the USA that we believe would accelerate efforts to benefit patients with cancer. Finally, we hope the recommendations in this report will facilitate new international collaborations to further enhance global efforts in cancer control.