Development of a cytology-based multivariate analytical risk index for oral cancer

Timothy J. Abram, Pierre N. Floriano, Robert James, A. Ross Kerr, Martin H. Thornhill, Spencer W. Redding, Nadarajah Vigneswaran, Rameez Raja, Michael P. McRae, John McDevitt

Research output: Contribution to journalArticle

Abstract

Objectives: The diagnosis and management of oral cavity cancers are often complicated by the uncertainty of which patients will undergo malignant transformation, obligating close surveillance over time. However, serial biopsies are undesirable, highly invasive, and subject to inherent issues with poor inter-pathologist agreement and unpredictability as a surrogate for malignant transformation and clinical outcomes. The goal of this study was to develop and evaluate a Multivariate Analytical Risk Index for Oral Cancer (MARIO) with potential to provide non-invasive, sensitive, and quantitative risk assessments for monitoring lesion progression. Materials and methods: A series of predictive models were developed and validated using previously recorded single-cell data from oral cytology samples resulting in a “continuous risk score”. Model development consisted of: (1) training base classification models for each diagnostic class pair, (2) pairwise coupling to obtain diagnostic class probabilities, and (3) a weighted aggregation resulting in a continuous MARIO. Results and conclusions: Diagnostic accuracy based on optimized cut-points for the test dataset ranged from 76.0% for Benign, to 82.4% for Dysplastic, 89.6% for Malignant, and 97.6% for Normal controls for an overall MARIO accuracy of 72.8%. Furthermore, a strong positive relationship with diagnostic severity was demonstrated (Pearson's coefficient = 0.805 for test dataset) as well as the ability of the MARIO to respond to subtle changes in cell composition. The development of a continuous MARIO for PMOL is presented, resulting in a sensitive, accurate, and non-invasive method with potential for enabling monitoring disease progression, recurrence, and the need for therapeutic intervention of these lesions.

Original languageEnglish (US)
Pages (from-to)6-11
Number of pages6
JournalOral Oncology
Volume92
DOIs
StatePublished - May 1 2019

Fingerprint

Mouth Neoplasms
Cell Biology
Uncertainty
Mouth
Disease Progression
Biopsy
Recurrence

Keywords

  • Cytology
  • Model ensembles
  • Multi-class classification
  • Oral cancer
  • Risk assessment

ASJC Scopus subject areas

  • Oral Surgery
  • Oncology
  • Cancer Research

Cite this

Abram, T. J., Floriano, P. N., James, R., Kerr, A. R., Thornhill, M. H., Redding, S. W., ... McDevitt, J. (2019). Development of a cytology-based multivariate analytical risk index for oral cancer. Oral Oncology, 92, 6-11. https://doi.org/10.1016/j.oraloncology.2019.02.011

Development of a cytology-based multivariate analytical risk index for oral cancer. / Abram, Timothy J.; Floriano, Pierre N.; James, Robert; Kerr, A. Ross; Thornhill, Martin H.; Redding, Spencer W.; Vigneswaran, Nadarajah; Raja, Rameez; McRae, Michael P.; McDevitt, John.

In: Oral Oncology, Vol. 92, 01.05.2019, p. 6-11.

Research output: Contribution to journalArticle

Abram, TJ, Floriano, PN, James, R, Kerr, AR, Thornhill, MH, Redding, SW, Vigneswaran, N, Raja, R, McRae, MP & McDevitt, J 2019, 'Development of a cytology-based multivariate analytical risk index for oral cancer', Oral Oncology, vol. 92, pp. 6-11. https://doi.org/10.1016/j.oraloncology.2019.02.011
Abram TJ, Floriano PN, James R, Kerr AR, Thornhill MH, Redding SW et al. Development of a cytology-based multivariate analytical risk index for oral cancer. Oral Oncology. 2019 May 1;92:6-11. https://doi.org/10.1016/j.oraloncology.2019.02.011
Abram, Timothy J. ; Floriano, Pierre N. ; James, Robert ; Kerr, A. Ross ; Thornhill, Martin H. ; Redding, Spencer W. ; Vigneswaran, Nadarajah ; Raja, Rameez ; McRae, Michael P. ; McDevitt, John. / Development of a cytology-based multivariate analytical risk index for oral cancer. In: Oral Oncology. 2019 ; Vol. 92. pp. 6-11.
@article{064b1d5514d3478498b6140f8e46637a,
title = "Development of a cytology-based multivariate analytical risk index for oral cancer",
abstract = "Objectives: The diagnosis and management of oral cavity cancers are often complicated by the uncertainty of which patients will undergo malignant transformation, obligating close surveillance over time. However, serial biopsies are undesirable, highly invasive, and subject to inherent issues with poor inter-pathologist agreement and unpredictability as a surrogate for malignant transformation and clinical outcomes. The goal of this study was to develop and evaluate a Multivariate Analytical Risk Index for Oral Cancer (MARIO) with potential to provide non-invasive, sensitive, and quantitative risk assessments for monitoring lesion progression. Materials and methods: A series of predictive models were developed and validated using previously recorded single-cell data from oral cytology samples resulting in a “continuous risk score”. Model development consisted of: (1) training base classification models for each diagnostic class pair, (2) pairwise coupling to obtain diagnostic class probabilities, and (3) a weighted aggregation resulting in a continuous MARIO. Results and conclusions: Diagnostic accuracy based on optimized cut-points for the test dataset ranged from 76.0{\%} for Benign, to 82.4{\%} for Dysplastic, 89.6{\%} for Malignant, and 97.6{\%} for Normal controls for an overall MARIO accuracy of 72.8{\%}. Furthermore, a strong positive relationship with diagnostic severity was demonstrated (Pearson's coefficient = 0.805 for test dataset) as well as the ability of the MARIO to respond to subtle changes in cell composition. The development of a continuous MARIO for PMOL is presented, resulting in a sensitive, accurate, and non-invasive method with potential for enabling monitoring disease progression, recurrence, and the need for therapeutic intervention of these lesions.",
keywords = "Cytology, Model ensembles, Multi-class classification, Oral cancer, Risk assessment",
author = "Abram, {Timothy J.} and Floriano, {Pierre N.} and Robert James and Kerr, {A. Ross} and Thornhill, {Martin H.} and Redding, {Spencer W.} and Nadarajah Vigneswaran and Rameez Raja and McRae, {Michael P.} and John McDevitt",
year = "2019",
month = "5",
day = "1",
doi = "10.1016/j.oraloncology.2019.02.011",
language = "English (US)",
volume = "92",
pages = "6--11",
journal = "Oral Oncology",
issn = "1368-8375",
publisher = "Elsevier Limited",

}

TY - JOUR

T1 - Development of a cytology-based multivariate analytical risk index for oral cancer

AU - Abram, Timothy J.

AU - Floriano, Pierre N.

AU - James, Robert

AU - Kerr, A. Ross

AU - Thornhill, Martin H.

AU - Redding, Spencer W.

AU - Vigneswaran, Nadarajah

AU - Raja, Rameez

AU - McRae, Michael P.

AU - McDevitt, John

PY - 2019/5/1

Y1 - 2019/5/1

N2 - Objectives: The diagnosis and management of oral cavity cancers are often complicated by the uncertainty of which patients will undergo malignant transformation, obligating close surveillance over time. However, serial biopsies are undesirable, highly invasive, and subject to inherent issues with poor inter-pathologist agreement and unpredictability as a surrogate for malignant transformation and clinical outcomes. The goal of this study was to develop and evaluate a Multivariate Analytical Risk Index for Oral Cancer (MARIO) with potential to provide non-invasive, sensitive, and quantitative risk assessments for monitoring lesion progression. Materials and methods: A series of predictive models were developed and validated using previously recorded single-cell data from oral cytology samples resulting in a “continuous risk score”. Model development consisted of: (1) training base classification models for each diagnostic class pair, (2) pairwise coupling to obtain diagnostic class probabilities, and (3) a weighted aggregation resulting in a continuous MARIO. Results and conclusions: Diagnostic accuracy based on optimized cut-points for the test dataset ranged from 76.0% for Benign, to 82.4% for Dysplastic, 89.6% for Malignant, and 97.6% for Normal controls for an overall MARIO accuracy of 72.8%. Furthermore, a strong positive relationship with diagnostic severity was demonstrated (Pearson's coefficient = 0.805 for test dataset) as well as the ability of the MARIO to respond to subtle changes in cell composition. The development of a continuous MARIO for PMOL is presented, resulting in a sensitive, accurate, and non-invasive method with potential for enabling monitoring disease progression, recurrence, and the need for therapeutic intervention of these lesions.

AB - Objectives: The diagnosis and management of oral cavity cancers are often complicated by the uncertainty of which patients will undergo malignant transformation, obligating close surveillance over time. However, serial biopsies are undesirable, highly invasive, and subject to inherent issues with poor inter-pathologist agreement and unpredictability as a surrogate for malignant transformation and clinical outcomes. The goal of this study was to develop and evaluate a Multivariate Analytical Risk Index for Oral Cancer (MARIO) with potential to provide non-invasive, sensitive, and quantitative risk assessments for monitoring lesion progression. Materials and methods: A series of predictive models were developed and validated using previously recorded single-cell data from oral cytology samples resulting in a “continuous risk score”. Model development consisted of: (1) training base classification models for each diagnostic class pair, (2) pairwise coupling to obtain diagnostic class probabilities, and (3) a weighted aggregation resulting in a continuous MARIO. Results and conclusions: Diagnostic accuracy based on optimized cut-points for the test dataset ranged from 76.0% for Benign, to 82.4% for Dysplastic, 89.6% for Malignant, and 97.6% for Normal controls for an overall MARIO accuracy of 72.8%. Furthermore, a strong positive relationship with diagnostic severity was demonstrated (Pearson's coefficient = 0.805 for test dataset) as well as the ability of the MARIO to respond to subtle changes in cell composition. The development of a continuous MARIO for PMOL is presented, resulting in a sensitive, accurate, and non-invasive method with potential for enabling monitoring disease progression, recurrence, and the need for therapeutic intervention of these lesions.

KW - Cytology

KW - Model ensembles

KW - Multi-class classification

KW - Oral cancer

KW - Risk assessment

UR - http://www.scopus.com/inward/record.url?scp=85062810898&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85062810898&partnerID=8YFLogxK

U2 - 10.1016/j.oraloncology.2019.02.011

DO - 10.1016/j.oraloncology.2019.02.011

M3 - Article

VL - 92

SP - 6

EP - 11

JO - Oral Oncology

JF - Oral Oncology

SN - 1368-8375

ER -