Improved detection of prostate cancer using classification and regression tree analysis

Mark Garzotto, Tomasz (Tom) Beer, R. Guy Hudson, Laura Peters, Yi Ching Hsieh, Eduardo Barrera, Thomas Klein, Motomi (Tomi) Mori

Research output: Contribution to journalArticle

77 Citations (Scopus)

Abstract

Purpose: To build a decision tree for patients suspected of having prostate cancer using classification and regression tree (CART) analysis. Patients and Methods: Data were uniformly collected on 1,433 referred men with a serum prostate-specific antigen (PSA) levels of ≤ 10 ng/mL who underwent a prostate biopsy. Factors analyzed included demographic, laboratory, and ultrasound data (ie, hypoechoic lesions and PSA density [PSAD]). Twenty percent of the data was randomly selected and reserved for study validation. CART analysis was performed in two steps, initially using PSA and digital rectal examination (DRE) alone and subsequently using the remaining variables. Results: CART analysis selected a PSA cutoff of more than 1.55 ng/mL for further work-up, regardless of DRE findings. CART then selected the following subgroups at risk for a positive biopsy: (1) PSAD more than 0.165 ng/mL/cc; (2) PSAD <0.165 ng/mL/cc and a hypoechoic lesion; (3) PSAD ≤ 0.165 ng/mL/cc, no hypoechoic lesions, age older than 55.5 years, and prostate volume ≤ 44.0 cc; and (4) PSAD ≤ 0.165 ng/mL/cc, no hypoechoic lesions, age older than 55.5 years, and 50.25 cc less than prostate volume ≤ 80.8 cc. In the validation data set, specificity and sensitivity were 31.3% and 96.6%, respectively. Cancers that were missed by the CART were Gleason score 6 or less in 93.4% of cases. Receiver operator characteristic curve analysis showed that CART and logistic regression models had similar accuracy (area under the curve = 0.74 v 0.72, respectively). Conclusion: Application of CART analysis to the prostate biopsy decision results in a significant reduction in unnecessary biopsies while retaining a high degree of sensitivity when compared with the standard of performing a biopsy of all patients with an abnormal PSA or DRE.

Original languageEnglish (US)
Pages (from-to)4322-4329
Number of pages8
JournalJournal of Clinical Oncology
Volume23
Issue number19
DOIs
StatePublished - 2005

Fingerprint

Prostatic Neoplasms
Prostate-Specific Antigen
Regression Analysis
Digital Rectal Examination
Prostate
Biopsy
Logistic Models
Decision Trees
Neoplasm Grading
Validation Studies
Area Under Curve
Demography
Sensitivity and Specificity
Serum
Neoplasms

ASJC Scopus subject areas

  • Cancer Research
  • Oncology

Cite this

Improved detection of prostate cancer using classification and regression tree analysis. / Garzotto, Mark; Beer, Tomasz (Tom); Hudson, R. Guy; Peters, Laura; Hsieh, Yi Ching; Barrera, Eduardo; Klein, Thomas; Mori, Motomi (Tomi).

In: Journal of Clinical Oncology, Vol. 23, No. 19, 2005, p. 4322-4329.

Research output: Contribution to journalArticle

Garzotto, Mark ; Beer, Tomasz (Tom) ; Hudson, R. Guy ; Peters, Laura ; Hsieh, Yi Ching ; Barrera, Eduardo ; Klein, Thomas ; Mori, Motomi (Tomi). / Improved detection of prostate cancer using classification and regression tree analysis. In: Journal of Clinical Oncology. 2005 ; Vol. 23, No. 19. pp. 4322-4329.
@article{b77b0baf7a2f42a7aca23e6c5627c426,
title = "Improved detection of prostate cancer using classification and regression tree analysis",
abstract = "Purpose: To build a decision tree for patients suspected of having prostate cancer using classification and regression tree (CART) analysis. Patients and Methods: Data were uniformly collected on 1,433 referred men with a serum prostate-specific antigen (PSA) levels of ≤ 10 ng/mL who underwent a prostate biopsy. Factors analyzed included demographic, laboratory, and ultrasound data (ie, hypoechoic lesions and PSA density [PSAD]). Twenty percent of the data was randomly selected and reserved for study validation. CART analysis was performed in two steps, initially using PSA and digital rectal examination (DRE) alone and subsequently using the remaining variables. Results: CART analysis selected a PSA cutoff of more than 1.55 ng/mL for further work-up, regardless of DRE findings. CART then selected the following subgroups at risk for a positive biopsy: (1) PSAD more than 0.165 ng/mL/cc; (2) PSAD <0.165 ng/mL/cc and a hypoechoic lesion; (3) PSAD ≤ 0.165 ng/mL/cc, no hypoechoic lesions, age older than 55.5 years, and prostate volume ≤ 44.0 cc; and (4) PSAD ≤ 0.165 ng/mL/cc, no hypoechoic lesions, age older than 55.5 years, and 50.25 cc less than prostate volume ≤ 80.8 cc. In the validation data set, specificity and sensitivity were 31.3{\%} and 96.6{\%}, respectively. Cancers that were missed by the CART were Gleason score 6 or less in 93.4{\%} of cases. Receiver operator characteristic curve analysis showed that CART and logistic regression models had similar accuracy (area under the curve = 0.74 v 0.72, respectively). Conclusion: Application of CART analysis to the prostate biopsy decision results in a significant reduction in unnecessary biopsies while retaining a high degree of sensitivity when compared with the standard of performing a biopsy of all patients with an abnormal PSA or DRE.",
author = "Mark Garzotto and Beer, {Tomasz (Tom)} and Hudson, {R. Guy} and Laura Peters and Hsieh, {Yi Ching} and Eduardo Barrera and Thomas Klein and Mori, {Motomi (Tomi)}",
year = "2005",
doi = "10.1200/JCO.2005.11.136",
language = "English (US)",
volume = "23",
pages = "4322--4329",
journal = "Journal of Clinical Oncology",
issn = "0732-183X",
publisher = "American Society of Clinical Oncology",
number = "19",

}

TY - JOUR

T1 - Improved detection of prostate cancer using classification and regression tree analysis

AU - Garzotto, Mark

AU - Beer, Tomasz (Tom)

AU - Hudson, R. Guy

AU - Peters, Laura

AU - Hsieh, Yi Ching

AU - Barrera, Eduardo

AU - Klein, Thomas

AU - Mori, Motomi (Tomi)

PY - 2005

Y1 - 2005

N2 - Purpose: To build a decision tree for patients suspected of having prostate cancer using classification and regression tree (CART) analysis. Patients and Methods: Data were uniformly collected on 1,433 referred men with a serum prostate-specific antigen (PSA) levels of ≤ 10 ng/mL who underwent a prostate biopsy. Factors analyzed included demographic, laboratory, and ultrasound data (ie, hypoechoic lesions and PSA density [PSAD]). Twenty percent of the data was randomly selected and reserved for study validation. CART analysis was performed in two steps, initially using PSA and digital rectal examination (DRE) alone and subsequently using the remaining variables. Results: CART analysis selected a PSA cutoff of more than 1.55 ng/mL for further work-up, regardless of DRE findings. CART then selected the following subgroups at risk for a positive biopsy: (1) PSAD more than 0.165 ng/mL/cc; (2) PSAD <0.165 ng/mL/cc and a hypoechoic lesion; (3) PSAD ≤ 0.165 ng/mL/cc, no hypoechoic lesions, age older than 55.5 years, and prostate volume ≤ 44.0 cc; and (4) PSAD ≤ 0.165 ng/mL/cc, no hypoechoic lesions, age older than 55.5 years, and 50.25 cc less than prostate volume ≤ 80.8 cc. In the validation data set, specificity and sensitivity were 31.3% and 96.6%, respectively. Cancers that were missed by the CART were Gleason score 6 or less in 93.4% of cases. Receiver operator characteristic curve analysis showed that CART and logistic regression models had similar accuracy (area under the curve = 0.74 v 0.72, respectively). Conclusion: Application of CART analysis to the prostate biopsy decision results in a significant reduction in unnecessary biopsies while retaining a high degree of sensitivity when compared with the standard of performing a biopsy of all patients with an abnormal PSA or DRE.

AB - Purpose: To build a decision tree for patients suspected of having prostate cancer using classification and regression tree (CART) analysis. Patients and Methods: Data were uniformly collected on 1,433 referred men with a serum prostate-specific antigen (PSA) levels of ≤ 10 ng/mL who underwent a prostate biopsy. Factors analyzed included demographic, laboratory, and ultrasound data (ie, hypoechoic lesions and PSA density [PSAD]). Twenty percent of the data was randomly selected and reserved for study validation. CART analysis was performed in two steps, initially using PSA and digital rectal examination (DRE) alone and subsequently using the remaining variables. Results: CART analysis selected a PSA cutoff of more than 1.55 ng/mL for further work-up, regardless of DRE findings. CART then selected the following subgroups at risk for a positive biopsy: (1) PSAD more than 0.165 ng/mL/cc; (2) PSAD <0.165 ng/mL/cc and a hypoechoic lesion; (3) PSAD ≤ 0.165 ng/mL/cc, no hypoechoic lesions, age older than 55.5 years, and prostate volume ≤ 44.0 cc; and (4) PSAD ≤ 0.165 ng/mL/cc, no hypoechoic lesions, age older than 55.5 years, and 50.25 cc less than prostate volume ≤ 80.8 cc. In the validation data set, specificity and sensitivity were 31.3% and 96.6%, respectively. Cancers that were missed by the CART were Gleason score 6 or less in 93.4% of cases. Receiver operator characteristic curve analysis showed that CART and logistic regression models had similar accuracy (area under the curve = 0.74 v 0.72, respectively). Conclusion: Application of CART analysis to the prostate biopsy decision results in a significant reduction in unnecessary biopsies while retaining a high degree of sensitivity when compared with the standard of performing a biopsy of all patients with an abnormal PSA or DRE.

UR - http://www.scopus.com/inward/record.url?scp=22344444621&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=22344444621&partnerID=8YFLogxK

U2 - 10.1200/JCO.2005.11.136

DO - 10.1200/JCO.2005.11.136

M3 - Article

C2 - 15781880

AN - SCOPUS:22344444621

VL - 23

SP - 4322

EP - 4329

JO - Journal of Clinical Oncology

JF - Journal of Clinical Oncology

SN - 0732-183X

IS - 19

ER -