‘Cytology-on-a-chip’ based sensors for monitoring of potentially malignant oral lesions

Timothy J. Abram, Pierre N. Floriano, Nicolaos Christodoulides, Robert James, A. Ross Kerr, Martin H. Thornhill, Spencer W. Redding, Nadarajah Vigneswaran, Paul M. Speight, Julie Vick, Craig Murdoch, Christine Freeman, Anne M. Hegarty, Katy D'Apice, Joan A. Phelan, Patricia M. Corby, Ismael Khouly, Jerry Bouquot, Nagi M. Demian, Y. Etan WeinstockStephanie Rowan, Chih Ko Yeh, H. Stan McGuff, Frank R. Miller, Surabhi Gaur, Kailash Karthikeyan, Leander Taylor, Cathy Le, Michael Nguyen, Humberto Talavera, Rameez Raja, Jorge Wong, John McDevitt

Research output: Contribution to journalArticle

Abstract

Despite significant advances in surgical procedures and treatment, long-term prognosis for patients with oral cancer remains poor, with survival rates among the lowest of major cancers. Better methods are desperately needed to identify potential malignancies early when treatments are more effective. Objective To develop robust classification models from cytology-on-a-chip measurements that mirror diagnostic performance of gold standard approach involving tissue biopsy. Materials and methods Measurements were recorded from 714 prospectively recruited patients with suspicious lesions across 6 diagnostic categories (each confirmed by tissue biopsy -histopathology) using a powerful new ‘cytology-on-a-chip’ approach capable of executing high content analysis at a single cell level. Over 200 cellular features related to biomarker expression, nuclear parameters and cellular morphology were recorded per cell. By cataloging an average of 2000 cells per patient, these efforts resulted in nearly 13 million indexed objects. Results Binary “low-risk”/“high-risk” models yielded AUC values of 0.88 and 0.84 for training and validation models, respectively, with an accompanying difference in sensitivity + specificity of 6.2%. In terms of accuracy, this model accurately predicted the correct diagnosis approximately 70% of the time, compared to the 69% initial agreement rate of the pool of expert pathologists. Key parameters identified in these models included cell circularity, Ki67 and EGFR expression, nuclear-cytoplasmic ratio, nuclear area, and cell area. Conclusions This chip-based approach yields objective data that can be leveraged for diagnosis and management of patients with PMOL as well as uncovering new molecular-level insights behind cytological differences across the OED spectrum.

Original languageEnglish (US)
Pages (from-to)103-111
Number of pages9
JournalOral Oncology
Volume60
DOIs
StatePublished - Sep 1 2016

Fingerprint

Cell Biology
Cataloging
Biopsy
Mouth Neoplasms
Area Under Curve
Neoplasms
Survival Rate
Biomarkers
Sensitivity and Specificity
Therapeutics

Keywords

  • Cytology
  • High content analysis
  • LASSO
  • Machine learning
  • Microfluidic
  • Oral cancer
  • Oral epithelial dysplasia
  • Random forest

ASJC Scopus subject areas

  • Oncology
  • Oral Surgery
  • Cancer Research

Cite this

Abram, T. J., Floriano, P. N., Christodoulides, N., James, R., Kerr, A. R., Thornhill, M. H., ... McDevitt, J. (2016). ‘Cytology-on-a-chip’ based sensors for monitoring of potentially malignant oral lesions. Oral Oncology, 60, 103-111. https://doi.org/10.1016/j.oraloncology.2016.07.002

‘Cytology-on-a-chip’ based sensors for monitoring of potentially malignant oral lesions. / Abram, Timothy J.; Floriano, Pierre N.; Christodoulides, Nicolaos; James, Robert; Kerr, A. Ross; Thornhill, Martin H.; Redding, Spencer W.; Vigneswaran, Nadarajah; Speight, Paul M.; Vick, Julie; Murdoch, Craig; Freeman, Christine; Hegarty, Anne M.; D'Apice, Katy; Phelan, Joan A.; Corby, Patricia M.; Khouly, Ismael; Bouquot, Jerry; Demian, Nagi M.; Weinstock, Y. Etan; Rowan, Stephanie; Yeh, Chih Ko; McGuff, H. Stan; Miller, Frank R.; Gaur, Surabhi; Karthikeyan, Kailash; Taylor, Leander; Le, Cathy; Nguyen, Michael; Talavera, Humberto; Raja, Rameez; Wong, Jorge; McDevitt, John.

In: Oral Oncology, Vol. 60, 01.09.2016, p. 103-111.

Research output: Contribution to journalArticle

Abram, TJ, Floriano, PN, Christodoulides, N, James, R, Kerr, AR, Thornhill, MH, Redding, SW, Vigneswaran, N, Speight, PM, Vick, J, Murdoch, C, Freeman, C, Hegarty, AM, D'Apice, K, Phelan, JA, Corby, PM, Khouly, I, Bouquot, J, Demian, NM, Weinstock, YE, Rowan, S, Yeh, CK, McGuff, HS, Miller, FR, Gaur, S, Karthikeyan, K, Taylor, L, Le, C, Nguyen, M, Talavera, H, Raja, R, Wong, J & McDevitt, J 2016, '‘Cytology-on-a-chip’ based sensors for monitoring of potentially malignant oral lesions', Oral Oncology, vol. 60, pp. 103-111. https://doi.org/10.1016/j.oraloncology.2016.07.002
Abram TJ, Floriano PN, Christodoulides N, James R, Kerr AR, Thornhill MH et al. ‘Cytology-on-a-chip’ based sensors for monitoring of potentially malignant oral lesions. Oral Oncology. 2016 Sep 1;60:103-111. https://doi.org/10.1016/j.oraloncology.2016.07.002
Abram, Timothy J. ; Floriano, Pierre N. ; Christodoulides, Nicolaos ; James, Robert ; Kerr, A. Ross ; Thornhill, Martin H. ; Redding, Spencer W. ; Vigneswaran, Nadarajah ; Speight, Paul M. ; Vick, Julie ; Murdoch, Craig ; Freeman, Christine ; Hegarty, Anne M. ; D'Apice, Katy ; Phelan, Joan A. ; Corby, Patricia M. ; Khouly, Ismael ; Bouquot, Jerry ; Demian, Nagi M. ; Weinstock, Y. Etan ; Rowan, Stephanie ; Yeh, Chih Ko ; McGuff, H. Stan ; Miller, Frank R. ; Gaur, Surabhi ; Karthikeyan, Kailash ; Taylor, Leander ; Le, Cathy ; Nguyen, Michael ; Talavera, Humberto ; Raja, Rameez ; Wong, Jorge ; McDevitt, John. / ‘Cytology-on-a-chip’ based sensors for monitoring of potentially malignant oral lesions. In: Oral Oncology. 2016 ; Vol. 60. pp. 103-111.
@article{d3b8ecc761334ebd8d82a787c1fa6e80,
title = "‘Cytology-on-a-chip’ based sensors for monitoring of potentially malignant oral lesions",
abstract = "Despite significant advances in surgical procedures and treatment, long-term prognosis for patients with oral cancer remains poor, with survival rates among the lowest of major cancers. Better methods are desperately needed to identify potential malignancies early when treatments are more effective. Objective To develop robust classification models from cytology-on-a-chip measurements that mirror diagnostic performance of gold standard approach involving tissue biopsy. Materials and methods Measurements were recorded from 714 prospectively recruited patients with suspicious lesions across 6 diagnostic categories (each confirmed by tissue biopsy -histopathology) using a powerful new ‘cytology-on-a-chip’ approach capable of executing high content analysis at a single cell level. Over 200 cellular features related to biomarker expression, nuclear parameters and cellular morphology were recorded per cell. By cataloging an average of 2000 cells per patient, these efforts resulted in nearly 13 million indexed objects. Results Binary “low-risk”/“high-risk” models yielded AUC values of 0.88 and 0.84 for training and validation models, respectively, with an accompanying difference in sensitivity + specificity of 6.2{\%}. In terms of accuracy, this model accurately predicted the correct diagnosis approximately 70{\%} of the time, compared to the 69{\%} initial agreement rate of the pool of expert pathologists. Key parameters identified in these models included cell circularity, Ki67 and EGFR expression, nuclear-cytoplasmic ratio, nuclear area, and cell area. Conclusions This chip-based approach yields objective data that can be leveraged for diagnosis and management of patients with PMOL as well as uncovering new molecular-level insights behind cytological differences across the OED spectrum.",
keywords = "Cytology, High content analysis, LASSO, Machine learning, Microfluidic, Oral cancer, Oral epithelial dysplasia, Random forest",
author = "Abram, {Timothy J.} and Floriano, {Pierre N.} and Nicolaos Christodoulides and Robert James and Kerr, {A. Ross} and Thornhill, {Martin H.} and Redding, {Spencer W.} and Nadarajah Vigneswaran and Speight, {Paul M.} and Julie Vick and Craig Murdoch and Christine Freeman and Hegarty, {Anne M.} and Katy D'Apice and Phelan, {Joan A.} and Corby, {Patricia M.} and Ismael Khouly and Jerry Bouquot and Demian, {Nagi M.} and Weinstock, {Y. Etan} and Stephanie Rowan and Yeh, {Chih Ko} and McGuff, {H. Stan} and Miller, {Frank R.} and Surabhi Gaur and Kailash Karthikeyan and Leander Taylor and Cathy Le and Michael Nguyen and Humberto Talavera and Rameez Raja and Jorge Wong and John McDevitt",
year = "2016",
month = "9",
day = "1",
doi = "10.1016/j.oraloncology.2016.07.002",
language = "English (US)",
volume = "60",
pages = "103--111",
journal = "Oral Oncology",
issn = "1368-8375",
publisher = "Elsevier Limited",

}

TY - JOUR

T1 - ‘Cytology-on-a-chip’ based sensors for monitoring of potentially malignant oral lesions

AU - Abram, Timothy J.

AU - Floriano, Pierre N.

AU - Christodoulides, Nicolaos

AU - James, Robert

AU - Kerr, A. Ross

AU - Thornhill, Martin H.

AU - Redding, Spencer W.

AU - Vigneswaran, Nadarajah

AU - Speight, Paul M.

AU - Vick, Julie

AU - Murdoch, Craig

AU - Freeman, Christine

AU - Hegarty, Anne M.

AU - D'Apice, Katy

AU - Phelan, Joan A.

AU - Corby, Patricia M.

AU - Khouly, Ismael

AU - Bouquot, Jerry

AU - Demian, Nagi M.

AU - Weinstock, Y. Etan

AU - Rowan, Stephanie

AU - Yeh, Chih Ko

AU - McGuff, H. Stan

AU - Miller, Frank R.

AU - Gaur, Surabhi

AU - Karthikeyan, Kailash

AU - Taylor, Leander

AU - Le, Cathy

AU - Nguyen, Michael

AU - Talavera, Humberto

AU - Raja, Rameez

AU - Wong, Jorge

AU - McDevitt, John

PY - 2016/9/1

Y1 - 2016/9/1

N2 - Despite significant advances in surgical procedures and treatment, long-term prognosis for patients with oral cancer remains poor, with survival rates among the lowest of major cancers. Better methods are desperately needed to identify potential malignancies early when treatments are more effective. Objective To develop robust classification models from cytology-on-a-chip measurements that mirror diagnostic performance of gold standard approach involving tissue biopsy. Materials and methods Measurements were recorded from 714 prospectively recruited patients with suspicious lesions across 6 diagnostic categories (each confirmed by tissue biopsy -histopathology) using a powerful new ‘cytology-on-a-chip’ approach capable of executing high content analysis at a single cell level. Over 200 cellular features related to biomarker expression, nuclear parameters and cellular morphology were recorded per cell. By cataloging an average of 2000 cells per patient, these efforts resulted in nearly 13 million indexed objects. Results Binary “low-risk”/“high-risk” models yielded AUC values of 0.88 and 0.84 for training and validation models, respectively, with an accompanying difference in sensitivity + specificity of 6.2%. In terms of accuracy, this model accurately predicted the correct diagnosis approximately 70% of the time, compared to the 69% initial agreement rate of the pool of expert pathologists. Key parameters identified in these models included cell circularity, Ki67 and EGFR expression, nuclear-cytoplasmic ratio, nuclear area, and cell area. Conclusions This chip-based approach yields objective data that can be leveraged for diagnosis and management of patients with PMOL as well as uncovering new molecular-level insights behind cytological differences across the OED spectrum.

AB - Despite significant advances in surgical procedures and treatment, long-term prognosis for patients with oral cancer remains poor, with survival rates among the lowest of major cancers. Better methods are desperately needed to identify potential malignancies early when treatments are more effective. Objective To develop robust classification models from cytology-on-a-chip measurements that mirror diagnostic performance of gold standard approach involving tissue biopsy. Materials and methods Measurements were recorded from 714 prospectively recruited patients with suspicious lesions across 6 diagnostic categories (each confirmed by tissue biopsy -histopathology) using a powerful new ‘cytology-on-a-chip’ approach capable of executing high content analysis at a single cell level. Over 200 cellular features related to biomarker expression, nuclear parameters and cellular morphology were recorded per cell. By cataloging an average of 2000 cells per patient, these efforts resulted in nearly 13 million indexed objects. Results Binary “low-risk”/“high-risk” models yielded AUC values of 0.88 and 0.84 for training and validation models, respectively, with an accompanying difference in sensitivity + specificity of 6.2%. In terms of accuracy, this model accurately predicted the correct diagnosis approximately 70% of the time, compared to the 69% initial agreement rate of the pool of expert pathologists. Key parameters identified in these models included cell circularity, Ki67 and EGFR expression, nuclear-cytoplasmic ratio, nuclear area, and cell area. Conclusions This chip-based approach yields objective data that can be leveraged for diagnosis and management of patients with PMOL as well as uncovering new molecular-level insights behind cytological differences across the OED spectrum.

KW - Cytology

KW - High content analysis

KW - LASSO

KW - Machine learning

KW - Microfluidic

KW - Oral cancer

KW - Oral epithelial dysplasia

KW - Random forest

UR - http://www.scopus.com/inward/record.url?scp=84978699068&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84978699068&partnerID=8YFLogxK

U2 - 10.1016/j.oraloncology.2016.07.002

DO - 10.1016/j.oraloncology.2016.07.002

M3 - Article

C2 - 27531880

AN - SCOPUS:84978699068

VL - 60

SP - 103

EP - 111

JO - Oral Oncology

JF - Oral Oncology

SN - 1368-8375

ER -