Use of Artificial Intelligence and Machine Learning Algorithms with Gene Expression Profiling to Predict Recurrent Nonmuscle Invasive Urothelial Carcinoma of the Bladder

Georg Bartsch, Anirban P. Mitra, Sheetal A. Mitra, Arpit A. Almal, Kenneth E. Steven, Donald G. Skinner, David W. Fry, Peter F. Lenehan, William P. Worzel, Richard J Cote

Research output: Contribution to journalArticle

14 Citations (Scopus)

Abstract

Purpose Due to the high recurrence risk of nonmuscle invasive urothelial carcinoma it is crucial to distinguish patients at high risk from those with indolent disease. In this study we used a machine learning algorithm to identify the genes in patients with nonmuscle invasive urothelial carcinoma at initial presentation that were most predictive of recurrence. We used the genes in a molecular signature to predict recurrence risk within 5 years after transurethral resection of bladder tumor. Materials and Methods Whole genome profiling was performed on 112 frozen nonmuscle invasive urothelial carcinoma specimens obtained at first presentation on Human WG-6 BeadChips (Illumina®). A genetic programming algorithm was applied to evolve classifier mathematical models for outcome prediction. Cross-validation based resampling and gene use frequencies were used to identify the most prognostic genes, which were combined into rules used in a voting algorithm to predict the sample target class. Key genes were validated by quantitative polymerase chain reaction. Results The classifier set included 21 genes that predicted recurrence. Quantitative polymerase chain reaction was done for these genes in a subset of 100 patients. A 5-gene combined rule incorporating a voting algorithm yielded 77% sensitivity and 85% specificity to predict recurrence in the training set, and 69% and 62%, respectively, in the test set. A singular 3-gene rule was constructed that predicted recurrence with 80% sensitivity and 90% specificity in the training set, and 71% and 67%, respectively, in the test set. Conclusions Using primary nonmuscle invasive urothelial carcinoma from initial occurrences genetic programming identified transcripts in reproducible fashion, which were predictive of recurrence. These findings could potentially impact nonmuscle invasive urothelial carcinoma management.

Original languageEnglish (US)
Pages (from-to)493-498
Number of pages6
JournalJournal of Urology
Volume195
Issue number2
DOIs
StatePublished - Feb 1 2016

Fingerprint

Artificial Intelligence
Gene Expression Profiling
Urinary Bladder
Carcinoma
Recurrence
Genes
Politics
Sensitivity and Specificity
Polymerase Chain Reaction
Machine Learning
Urinary Bladder Neoplasms
Gene Frequency
Theoretical Models
Genome

Keywords

  • algorithms
  • genome
  • local
  • neoplasm recurrence
  • software
  • urinary bladder neoplasms

ASJC Scopus subject areas

  • Urology

Cite this

Use of Artificial Intelligence and Machine Learning Algorithms with Gene Expression Profiling to Predict Recurrent Nonmuscle Invasive Urothelial Carcinoma of the Bladder. / Bartsch, Georg; Mitra, Anirban P.; Mitra, Sheetal A.; Almal, Arpit A.; Steven, Kenneth E.; Skinner, Donald G.; Fry, David W.; Lenehan, Peter F.; Worzel, William P.; Cote, Richard J.

In: Journal of Urology, Vol. 195, No. 2, 01.02.2016, p. 493-498.

Research output: Contribution to journalArticle

Bartsch, Georg ; Mitra, Anirban P. ; Mitra, Sheetal A. ; Almal, Arpit A. ; Steven, Kenneth E. ; Skinner, Donald G. ; Fry, David W. ; Lenehan, Peter F. ; Worzel, William P. ; Cote, Richard J. / Use of Artificial Intelligence and Machine Learning Algorithms with Gene Expression Profiling to Predict Recurrent Nonmuscle Invasive Urothelial Carcinoma of the Bladder. In: Journal of Urology. 2016 ; Vol. 195, No. 2. pp. 493-498.
@article{9494eaf146b54501bbaa783d5c373079,
title = "Use of Artificial Intelligence and Machine Learning Algorithms with Gene Expression Profiling to Predict Recurrent Nonmuscle Invasive Urothelial Carcinoma of the Bladder",
abstract = "Purpose Due to the high recurrence risk of nonmuscle invasive urothelial carcinoma it is crucial to distinguish patients at high risk from those with indolent disease. In this study we used a machine learning algorithm to identify the genes in patients with nonmuscle invasive urothelial carcinoma at initial presentation that were most predictive of recurrence. We used the genes in a molecular signature to predict recurrence risk within 5 years after transurethral resection of bladder tumor. Materials and Methods Whole genome profiling was performed on 112 frozen nonmuscle invasive urothelial carcinoma specimens obtained at first presentation on Human WG-6 BeadChips (Illumina{\circledR}). A genetic programming algorithm was applied to evolve classifier mathematical models for outcome prediction. Cross-validation based resampling and gene use frequencies were used to identify the most prognostic genes, which were combined into rules used in a voting algorithm to predict the sample target class. Key genes were validated by quantitative polymerase chain reaction. Results The classifier set included 21 genes that predicted recurrence. Quantitative polymerase chain reaction was done for these genes in a subset of 100 patients. A 5-gene combined rule incorporating a voting algorithm yielded 77{\%} sensitivity and 85{\%} specificity to predict recurrence in the training set, and 69{\%} and 62{\%}, respectively, in the test set. A singular 3-gene rule was constructed that predicted recurrence with 80{\%} sensitivity and 90{\%} specificity in the training set, and 71{\%} and 67{\%}, respectively, in the test set. Conclusions Using primary nonmuscle invasive urothelial carcinoma from initial occurrences genetic programming identified transcripts in reproducible fashion, which were predictive of recurrence. These findings could potentially impact nonmuscle invasive urothelial carcinoma management.",
keywords = "algorithms, genome, local, neoplasm recurrence, software, urinary bladder neoplasms",
author = "Georg Bartsch and Mitra, {Anirban P.} and Mitra, {Sheetal A.} and Almal, {Arpit A.} and Steven, {Kenneth E.} and Skinner, {Donald G.} and Fry, {David W.} and Lenehan, {Peter F.} and Worzel, {William P.} and Cote, {Richard J}",
year = "2016",
month = "2",
day = "1",
doi = "10.1016/j.juro.2015.09.090",
language = "English (US)",
volume = "195",
pages = "493--498",
journal = "Journal of Urology",
issn = "0022-5347",
publisher = "Elsevier Inc.",
number = "2",

}

TY - JOUR

T1 - Use of Artificial Intelligence and Machine Learning Algorithms with Gene Expression Profiling to Predict Recurrent Nonmuscle Invasive Urothelial Carcinoma of the Bladder

AU - Bartsch, Georg

AU - Mitra, Anirban P.

AU - Mitra, Sheetal A.

AU - Almal, Arpit A.

AU - Steven, Kenneth E.

AU - Skinner, Donald G.

AU - Fry, David W.

AU - Lenehan, Peter F.

AU - Worzel, William P.

AU - Cote, Richard J

PY - 2016/2/1

Y1 - 2016/2/1

N2 - Purpose Due to the high recurrence risk of nonmuscle invasive urothelial carcinoma it is crucial to distinguish patients at high risk from those with indolent disease. In this study we used a machine learning algorithm to identify the genes in patients with nonmuscle invasive urothelial carcinoma at initial presentation that were most predictive of recurrence. We used the genes in a molecular signature to predict recurrence risk within 5 years after transurethral resection of bladder tumor. Materials and Methods Whole genome profiling was performed on 112 frozen nonmuscle invasive urothelial carcinoma specimens obtained at first presentation on Human WG-6 BeadChips (Illumina®). A genetic programming algorithm was applied to evolve classifier mathematical models for outcome prediction. Cross-validation based resampling and gene use frequencies were used to identify the most prognostic genes, which were combined into rules used in a voting algorithm to predict the sample target class. Key genes were validated by quantitative polymerase chain reaction. Results The classifier set included 21 genes that predicted recurrence. Quantitative polymerase chain reaction was done for these genes in a subset of 100 patients. A 5-gene combined rule incorporating a voting algorithm yielded 77% sensitivity and 85% specificity to predict recurrence in the training set, and 69% and 62%, respectively, in the test set. A singular 3-gene rule was constructed that predicted recurrence with 80% sensitivity and 90% specificity in the training set, and 71% and 67%, respectively, in the test set. Conclusions Using primary nonmuscle invasive urothelial carcinoma from initial occurrences genetic programming identified transcripts in reproducible fashion, which were predictive of recurrence. These findings could potentially impact nonmuscle invasive urothelial carcinoma management.

AB - Purpose Due to the high recurrence risk of nonmuscle invasive urothelial carcinoma it is crucial to distinguish patients at high risk from those with indolent disease. In this study we used a machine learning algorithm to identify the genes in patients with nonmuscle invasive urothelial carcinoma at initial presentation that were most predictive of recurrence. We used the genes in a molecular signature to predict recurrence risk within 5 years after transurethral resection of bladder tumor. Materials and Methods Whole genome profiling was performed on 112 frozen nonmuscle invasive urothelial carcinoma specimens obtained at first presentation on Human WG-6 BeadChips (Illumina®). A genetic programming algorithm was applied to evolve classifier mathematical models for outcome prediction. Cross-validation based resampling and gene use frequencies were used to identify the most prognostic genes, which were combined into rules used in a voting algorithm to predict the sample target class. Key genes were validated by quantitative polymerase chain reaction. Results The classifier set included 21 genes that predicted recurrence. Quantitative polymerase chain reaction was done for these genes in a subset of 100 patients. A 5-gene combined rule incorporating a voting algorithm yielded 77% sensitivity and 85% specificity to predict recurrence in the training set, and 69% and 62%, respectively, in the test set. A singular 3-gene rule was constructed that predicted recurrence with 80% sensitivity and 90% specificity in the training set, and 71% and 67%, respectively, in the test set. Conclusions Using primary nonmuscle invasive urothelial carcinoma from initial occurrences genetic programming identified transcripts in reproducible fashion, which were predictive of recurrence. These findings could potentially impact nonmuscle invasive urothelial carcinoma management.

KW - algorithms

KW - genome

KW - local

KW - neoplasm recurrence

KW - software

KW - urinary bladder neoplasms

UR - http://www.scopus.com/inward/record.url?scp=84953746356&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84953746356&partnerID=8YFLogxK

U2 - 10.1016/j.juro.2015.09.090

DO - 10.1016/j.juro.2015.09.090

M3 - Article

C2 - 26459038

AN - SCOPUS:84953746356

VL - 195

SP - 493

EP - 498

JO - Journal of Urology

JF - Journal of Urology

SN - 0022-5347

IS - 2

ER -