1
|
Levine AB, Peng J, Farnell D, Nursey M, Wang Y, Naso JR, Ren H, Farahani H, Chen C, Chiu D, Talhouk A, Sheffield B, Riazy M, Ip PP, Parra-Herran C, Mills A, Singh N, Tessier-Cloutier B, Salisbury T, Lee J, Salcudean T, Jones SJ, Huntsman DG, Gilks CB, Yip S, Bashashati A. Synthesis of diagnostic quality cancer pathology images by generative adversarial networks. J Pathol 2020; 252:178-188. [PMID: 32686118 DOI: 10.1002/path.5509] [Citation(s) in RCA: 27] [Impact Index Per Article: 6.8] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 04/13/2020] [Revised: 07/02/2020] [Accepted: 07/10/2020] [Indexed: 12/16/2022]
Abstract
Deep learning-based computer vision methods have recently made remarkable breakthroughs in the analysis and classification of cancer pathology images. However, there has been relatively little investigation of the utility of deep neural networks to synthesize medical images. In this study, we evaluated the efficacy of generative adversarial networks to synthesize high-resolution pathology images of 10 histological types of cancer, including five cancer types from The Cancer Genome Atlas and the five major histological subtypes of ovarian carcinoma. The quality of these images was assessed using a comprehensive survey of board-certified pathologists (n = 9) and pathology trainees (n = 6). Our results show that the real and synthetic images are classified by histotype with comparable accuracies and the synthetic images are visually indistinguishable from real images. Furthermore, we trained deep convolutional neural networks to diagnose the different cancer types and determined that the synthetic images perform as well as additional real images when used to supplement a small training set. These findings have important applications in proficiency testing of medical practitioners and quality assurance in clinical laboratories. Furthermore, training of computer-aided diagnostic systems can benefit from synthetic images where labeled datasets are limited (e.g. rare cancers). We have created a publicly available website where clinicians and researchers can attempt questions from the image survey (http://gan.aimlab.ca/). © 2020 Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd.
Collapse
Affiliation(s)
- Adrian B Levine
- Department of Pathology and Laboratory Medicine, University of British Columbia, Vancouver, Canada
| | - Jason Peng
- Department of Pathology and Laboratory Medicine, University of British Columbia, Vancouver, Canada.,School of Biomedical Engineering, University of British Columbia, Vancouver, Canada
| | - David Farnell
- Department of Pathology and Laboratory Medicine, University of British Columbia, Vancouver, Canada
| | - Mitchell Nursey
- Department of Pathology and Laboratory Medicine, University of British Columbia, Vancouver, Canada.,School of Biomedical Engineering, University of British Columbia, Vancouver, Canada
| | - Yiping Wang
- Department of Pathology and Laboratory Medicine, University of British Columbia, Vancouver, Canada.,School of Biomedical Engineering, University of British Columbia, Vancouver, Canada
| | - Julia R Naso
- Department of Pathology and Laboratory Medicine, University of British Columbia, Vancouver, Canada
| | - Hezhen Ren
- Department of Pathology and Laboratory Medicine, University of British Columbia, Vancouver, Canada
| | - Hossein Farahani
- Department of Pathology and Laboratory Medicine, University of British Columbia, Vancouver, Canada.,School of Biomedical Engineering, University of British Columbia, Vancouver, Canada
| | - Colin Chen
- Department of Pathology and Laboratory Medicine, University of British Columbia, Vancouver, Canada.,School of Biomedical Engineering, University of British Columbia, Vancouver, Canada
| | - Derek Chiu
- Department of Pathology and Laboratory Medicine, University of British Columbia, Vancouver, Canada
| | - Aline Talhouk
- Department of Obstetrics and Gynecology, University of British Columbia, Vancouver, Canada
| | - Brandon Sheffield
- Department of Pathology, William Osler Health Centre-Brampton Civic Hospital, Brampton, Canada
| | - Maziar Riazy
- Department of Pathology and Laboratory Medicine, University of British Columbia, Vancouver, Canada
| | - Philip P Ip
- Department of Pathology, University of Hong Kong, Hong Kong SAR, PR China
| | - Carlos Parra-Herran
- Laboratory Medicine and Pathobiology, University of Toronto, Toronto, Canada
| | - Anne Mills
- Department of Pathology, University of Virginia, Charlottesville, VA, USA
| | - Naveena Singh
- Barts and the London School of Medicine and Dentistry, Queen Mary University of London, London, UK
| | - Basile Tessier-Cloutier
- Department of Pathology and Laboratory Medicine, University of British Columbia, Vancouver, Canada
| | - Taylor Salisbury
- Department of Pathology and Laboratory Medicine, University of British Columbia, Vancouver, Canada
| | - Jonathan Lee
- Department of Pathology and Laboratory Medicine, University of British Columbia, Vancouver, Canada
| | - Tim Salcudean
- Electrical & Computer Engineering, University of British Columbia, Vancouver, Canada
| | - Steven Jm Jones
- Canada's Michael Smith Genome Sciences Centre, Vancouver, Canada
| | - David G Huntsman
- Department of Pathology and Laboratory Medicine, University of British Columbia, Vancouver, Canada
| | - C Blake Gilks
- Department of Pathology and Laboratory Medicine, University of British Columbia, Vancouver, Canada
| | - Stephen Yip
- Department of Pathology and Laboratory Medicine, University of British Columbia, Vancouver, Canada
| | - Ali Bashashati
- Department of Pathology and Laboratory Medicine, University of British Columbia, Vancouver, Canada.,School of Biomedical Engineering, University of British Columbia, Vancouver, Canada.,Electrical & Computer Engineering, University of British Columbia, Vancouver, Canada
| |
Collapse
|
2
|
Li SS, Ip CKM, Tang MYH, Tang MKS, Tong Y, Zhang J, Hassan AA, Mak ASC, Yung S, Chan TM, Ip PP, Lee CL, Chiu PCN, Lee LTO, Lai HC, Zeng JZ, Shum HC, Wong AST. Sialyl Lewis x-P-selectin cascade mediates tumor-mesothelial adhesion in ascitic fluid shear flow. Nat Commun 2019; 10:2406. [PMID: 31160622 PMCID: PMC6547673 DOI: 10.1038/s41467-019-10334-6] [Citation(s) in RCA: 26] [Impact Index Per Article: 5.2] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 04/03/2018] [Accepted: 04/29/2019] [Indexed: 01/01/2023] Open
Abstract
Organ-specific colonization suggests that specific cell–cell recognition is essential. Yet, very little is known about this particular interaction. Moreover, tumor cell lodgement requires binding under shear stress, but not static, conditions. Here, we successfully isolate the metastatic populations of cancer stem/tumor-initiating cells (M-CSCs). We show that the M-CSCs tether more and roll slower than the non-metastatic (NM)-CSCs, thus resulting in the preferential binding to the peritoneal mesothelium under ascitic fluid shear stress. Mechanistically, this interaction is mediated by P-selectin expressed by the peritoneal mesothelium. Insulin-like growth factor receptor-1 carrying an uncommon non-sulfated sialyl-Lewisx (sLex) epitope serves as a distinct P-selectin binding determinant. Several glycosyltransferases, particularly α1,3-fucosyltransferase with rate-limiting activity for sLex synthesis, are highly expressed in M-CSCs. Tumor xenografts and clinical samples corroborate the relevance of these findings. These data advance our understanding on the molecular regulation of peritoneal metastasis and support the therapeutic potential of targeting the sLex-P-selectin cascade. Tumor cell in the peritoneum are often exposed to shear forces generated by ascitic flow during metastasis. Here, the authors show that metastatic cancer stem cells tether more and roll slower than the non-metastatic counterparts, and that sialyl-Lewisx -P-selectin axis mediates peritoneal metastasis.
Collapse
Affiliation(s)
- Shan-Shan Li
- School of Biological Sciences, University of Hong Kong, Pokfulam, Hong Kong
| | - Carman K M Ip
- School of Biological Sciences, University of Hong Kong, Pokfulam, Hong Kong
| | - Matthew Y H Tang
- Department of Mechanical Engineering, University of Hong Kong, Pokfulam, Hong Kong
| | - Maggie K S Tang
- School of Biological Sciences, University of Hong Kong, Pokfulam, Hong Kong
| | - Yin Tong
- School of Biological Sciences, University of Hong Kong, Pokfulam, Hong Kong
| | - Jiangwen Zhang
- School of Biological Sciences, University of Hong Kong, Pokfulam, Hong Kong
| | - Ayon Ahmed Hassan
- School of Biological Sciences, University of Hong Kong, Pokfulam, Hong Kong
| | - Abby S C Mak
- School of Biological Sciences, University of Hong Kong, Pokfulam, Hong Kong
| | - Susan Yung
- Department of Medicine, University of Hong Kong, Queen Mary Hospital, Pokfulam, Hong Kong
| | - Tak-Mao Chan
- Department of Medicine, University of Hong Kong, Queen Mary Hospital, Pokfulam, Hong Kong
| | - Philip P Ip
- Department of Pathology, University of Hong Kong, Queen Mary Hospital, Pokfulam, Hong Kong
| | - Cheuk Lun Lee
- Department of Obstetrics and Gynecology, University of Hong Kong, Queen Mary Hospital, Pokfulam, Hong Kong
| | - Philip C N Chiu
- Department of Obstetrics and Gynecology, University of Hong Kong, Queen Mary Hospital, Pokfulam, Hong Kong
| | - Leo Tsz On Lee
- Centre of Reproduction Development and Aging, Faculty of Health Sciences, University of Macau, Taipa, Macau, China
| | - Hung-Cheng Lai
- Department of Obstetrics and Gynecology, School of Medicine, College of Medicine, Taipei Medical University, Taipei, 23561, Taiwan.,Department of Obstetrics and Gynecology, Shuang Ho Hospital, Taipei Medical University, Taipei, 23561, Taiwan
| | - Jin-Zhang Zeng
- State Key Laboratory of Cellular Stress Biology and Fujian Provincial Key Laboratory of Innovative Drug Target Research, School of Pharmaceutical Sciences, Xiamen University, Xiamen, 361102, China
| | - Ho Cheung Shum
- Department of Mechanical Engineering, University of Hong Kong, Pokfulam, Hong Kong.
| | - Alice S T Wong
- School of Biological Sciences, University of Hong Kong, Pokfulam, Hong Kong.
| |
Collapse
|
3
|
Abstract
A key to successful metastasis is the formation of new vasculature, known as angiogenesis. Therefore, it is of great interest to unravel the underlying molecular mechanisms of tumor angiogenesis. Cadherins are a major class of cell surface receptors. The loss of cadherins, especially E-cadherin, is a well-established marker for tumor metastasis. Loss of E-cadherin is also a defining characteristic of several carcinomas, such as lobular carcinoma of the breast, and de-differentiated endometrioid carcinoma of the endometrium and ovary, which are known to be associated with more aggressive tumor behavior. Although E-cadherin is synthesized as a transmembrane molecule, its extracellular domain can be enzymatically cleaved off and released as a soluble E-cadherin (sE-cad), and this accounts for the loss of E-cadherin function or expression that has been implicated in tumor progression and metastasis. Importantly, sE-cad is present at high levels in the serum and malignant ascites of ovarian carcinoma patients. Nevertheless, little is known about how this essential protein dictates metastasis. Hitherto, many studies have given attention only to the dominant negative role of the loss of E-cadherin in weakening cell-cell adhesion, however, it is not known if sE-cad has biological activity in itself. In addition, the release mechanism of sE-cad has remained elusive. Here we show for the first time that sE-cad is a pivotal regulator of angiogenesis. We further show that exosomes are a novel major platform for the cleavage and release of sE-cad in vitro, in vivo and in patients’ derived samples (Nat Commun, 9: 2270).
Collapse
Affiliation(s)
- Maggie K S Tang
- School of Biological Sciences, University of Hong Kong, Pokfulam Road, Hong Kong
| | - Philip P Ip
- Department of Pathology, University of Hong Kong, Pokfulam Road, Hong Kong
| | - Alice S T Wong
- School of Biological Sciences, University of Hong Kong, Pokfulam Road, Hong Kong
| |
Collapse
|
5
|
Slatter TL, Hsia H, Samaranayaka A, Sykes P, Clow WB, Devenish CJ, Sutton T, Royds JA, Ip PP, Cheung AN, Hung NA. Loss of ATRX and DAXX expression identifies poor prognosis for smooth muscle tumours of uncertain malignant potential and early stage uterine leiomyosarcoma. J Pathol Clin Res 2015; 1:95-105. [PMID: 27499896 PMCID: PMC4858134 DOI: 10.1002/cjp2.11] [Citation(s) in RCA: 26] [Impact Index Per Article: 2.9] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Figures] [Subscribe] [Scholar Register] [Received: 09/05/2014] [Accepted: 10/22/2014] [Indexed: 12/13/2022]
Abstract
Uterine smooth muscle tumours of uncertain malignant potential (STUMP) are diagnostically and clinically challenging. The alternative lengthening of telomeres (ALT) telomere maintenance mechanism is associated with poor survival in soft tissue leiomyosarcoma. Time to first recurrence and survival were known for 18 STUMP and 43 leiomyosarcomata (LMS). These were screened for ALT telomere maintenance by the presence of ALT‐associated PML bodies (APBs) and for changes associated with the ALT phenotype, namely aberrant p53 expression, isocitrate dehydrogenase 1 mutation (R132H substitution) expression, mutant ATRX (αthalassemia/mental retardation syndrome X‐linked) expression and mutant DAXX (death‐domain‐associated protein) expression by immunohistochemistry (IHC). Overexpression of p16INK4A was examined immunohistologically in a subset of cases. Many of the tumours associated with death or recurrence demonstrated APBs commensurate with ALT telomere maintenance. However, all uterine STUMP (4/4), and vaginal STUMP (2/2) patients, and almost all LMS patients (88.4%, 23/26, including 90% (9/10) of stage 1 LMS cases), who had died of disease or who had recurrent disease, displayed loss of ATRX or DAXX expression. Loss of ATRX or DAXX expression identified poor prognosis (95% CI 2.1 to 40.8, p < 0.003), in the LMS group. Thus, loss of ATRX or DAXX expression in uterine smooth muscle tumours identifies a clinically aggressive molecular subtype of early stage LMS and when histopathological features are problematic such as in STUMP. As ATRX and DAXX IHC is readily performed in diagnostic laboratories these are potentially useful for routine histopathological classification and management.
Collapse
Affiliation(s)
- Tania L Slatter
- Department of Pathology Dunedin School of Medicine University of Otago Dunedin New Zealand
| | - Howard Hsia
- Department of Pathology Dunedin School of Medicine University of Otago Dunedin New Zealand
| | - Ari Samaranayaka
- Department of Preventive and Social Medicine Dunedin School of Medicine University of Otago Dunedin New Zealand
| | - Peter Sykes
- Department of Obstetrics and Gynaecology University of Otago Christchurch New Zealand
| | - William Bill Clow
- Department of Women's and Children's Health Dunedin School of Medicine University of Otago Dunedin New Zealand
| | - Celia J Devenish
- Department of Women's and Children's Health Dunedin School of Medicine University of Otago Dunedin New Zealand
| | - Tim Sutton
- Anatomical Pathology Service Pathlab Bay of Plenty Tauranga New Zealand
| | - Janice A Royds
- Department of Pathology Dunedin School of Medicine University of Otago Dunedin New Zealand
| | - Philip P Ip
- Anatomical Pathology Services Department of Pathology The University of Hong Kong, Queen Mary Hospital Hong Kong SAR
| | - Annie N Cheung
- Anatomical Pathology Services Department of Pathology The University of Hong Kong, Queen Mary Hospital Hong Kong SAR
| | - Noelyn Anne Hung
- Department of Pathology Dunedin School of Medicine University of Otago Dunedin New Zealand
| |
Collapse
|