Assessment of the performance of the Stanford online calculator for the prediction of nonsentinel lymph node metastasis in sentinel lymph node-positive breast cancer patients

Jeffrey Scow, Amy C. Degnim, Tanya L. Hoskin, Carol Reynolds, Judy C. Boughey

Research output: Contribution to journalArticle

31 Citations (Scopus)

Abstract

BACKGROUND: Several models for the prediction of nonsentinel lymph node (NSLN) metastasis in sentinel lymph node (SLN)-positive breast cancer patients have been proposed. In this study, the authors evaluate the Stanford Online Calculator (SOC), which was designed to predict the likelihood of NSLN metastasis using only 3 variables: primary tumor size, SLN metastasis size, and angiolymphatic invasion status. They compared it with the Mayo and Memorial Sloan-Kettering Cancer Center (MSKCC) nomograms. METHODS: The SOC was used to calculate the probability of NSLN metastasis in 464 breast cancer patients with SLN metastasis who underwent completion axillary lymph node dissection at the Mayo Clinic. The area under the receiver operating characteristic curve (AUC) was calculated for each model. Mean probabilities of patients with and without NSLN metastasis were compared. Patients with ≤5%, ≤10%, and 100% NSLN metastasis probabilities were examined. RESULTS: The AUCs of the Stanford, MSKCC, and Mayo models were 0.72, 0.74, and 0.77, respectively (P = .13). The mean Stanford probabilities for patients with and without NSLN metastasis were 0.75 (range, 0.06-1.0) and 0.50 (range, 0.05-1.0), respectively (P < .0001). The falsenegative rates for patients with a Stanford probability of ≤5% and ≤10% were 0% and 13%, respectively. Of the patients with a Stanford probability of 100%, 26% did not have NSLN metastasis. CONCLUSIONS: Despite using only 3 variables, the Stanford nomogram appears to perform on a par with, but not better than, the MSKCC and Mayo nomograms. Further validation in other patient populations is needed.

Original languageEnglish (US)
Pages (from-to)4064-4070
Number of pages7
JournalCancer
Volume115
Issue number18
DOIs
StatePublished - Sep 15 2009

Fingerprint

Lymph Nodes
Breast Neoplasms
Neoplasm Metastasis
Nomograms
Area Under Curve
Neoplasms
Sentinel Lymph Node
Lymph Node Excision
ROC Curve
Population

All Science Journal Classification (ASJC) codes

  • Oncology
  • Cancer Research

Cite this

@article{ae20cd911b814f1c8f845c6500a1d1b7,
title = "Assessment of the performance of the Stanford online calculator for the prediction of nonsentinel lymph node metastasis in sentinel lymph node-positive breast cancer patients",
abstract = "BACKGROUND: Several models for the prediction of nonsentinel lymph node (NSLN) metastasis in sentinel lymph node (SLN)-positive breast cancer patients have been proposed. In this study, the authors evaluate the Stanford Online Calculator (SOC), which was designed to predict the likelihood of NSLN metastasis using only 3 variables: primary tumor size, SLN metastasis size, and angiolymphatic invasion status. They compared it with the Mayo and Memorial Sloan-Kettering Cancer Center (MSKCC) nomograms. METHODS: The SOC was used to calculate the probability of NSLN metastasis in 464 breast cancer patients with SLN metastasis who underwent completion axillary lymph node dissection at the Mayo Clinic. The area under the receiver operating characteristic curve (AUC) was calculated for each model. Mean probabilities of patients with and without NSLN metastasis were compared. Patients with ≤5{\%}, ≤10{\%}, and 100{\%} NSLN metastasis probabilities were examined. RESULTS: The AUCs of the Stanford, MSKCC, and Mayo models were 0.72, 0.74, and 0.77, respectively (P = .13). The mean Stanford probabilities for patients with and without NSLN metastasis were 0.75 (range, 0.06-1.0) and 0.50 (range, 0.05-1.0), respectively (P < .0001). The falsenegative rates for patients with a Stanford probability of ≤5{\%} and ≤10{\%} were 0{\%} and 13{\%}, respectively. Of the patients with a Stanford probability of 100{\%}, 26{\%} did not have NSLN metastasis. CONCLUSIONS: Despite using only 3 variables, the Stanford nomogram appears to perform on a par with, but not better than, the MSKCC and Mayo nomograms. Further validation in other patient populations is needed.",
author = "Jeffrey Scow and Degnim, {Amy C.} and Hoskin, {Tanya L.} and Carol Reynolds and Boughey, {Judy C.}",
year = "2009",
month = "9",
day = "15",
doi = "10.1002/cncr.24469",
language = "English (US)",
volume = "115",
pages = "4064--4070",
journal = "Cancer",
issn = "0008-543X",
publisher = "John Wiley and Sons Inc.",
number = "18",

}

Assessment of the performance of the Stanford online calculator for the prediction of nonsentinel lymph node metastasis in sentinel lymph node-positive breast cancer patients. / Scow, Jeffrey; Degnim, Amy C.; Hoskin, Tanya L.; Reynolds, Carol; Boughey, Judy C.

In: Cancer, Vol. 115, No. 18, 15.09.2009, p. 4064-4070.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Assessment of the performance of the Stanford online calculator for the prediction of nonsentinel lymph node metastasis in sentinel lymph node-positive breast cancer patients

AU - Scow, Jeffrey

AU - Degnim, Amy C.

AU - Hoskin, Tanya L.

AU - Reynolds, Carol

AU - Boughey, Judy C.

PY - 2009/9/15

Y1 - 2009/9/15

N2 - BACKGROUND: Several models for the prediction of nonsentinel lymph node (NSLN) metastasis in sentinel lymph node (SLN)-positive breast cancer patients have been proposed. In this study, the authors evaluate the Stanford Online Calculator (SOC), which was designed to predict the likelihood of NSLN metastasis using only 3 variables: primary tumor size, SLN metastasis size, and angiolymphatic invasion status. They compared it with the Mayo and Memorial Sloan-Kettering Cancer Center (MSKCC) nomograms. METHODS: The SOC was used to calculate the probability of NSLN metastasis in 464 breast cancer patients with SLN metastasis who underwent completion axillary lymph node dissection at the Mayo Clinic. The area under the receiver operating characteristic curve (AUC) was calculated for each model. Mean probabilities of patients with and without NSLN metastasis were compared. Patients with ≤5%, ≤10%, and 100% NSLN metastasis probabilities were examined. RESULTS: The AUCs of the Stanford, MSKCC, and Mayo models were 0.72, 0.74, and 0.77, respectively (P = .13). The mean Stanford probabilities for patients with and without NSLN metastasis were 0.75 (range, 0.06-1.0) and 0.50 (range, 0.05-1.0), respectively (P < .0001). The falsenegative rates for patients with a Stanford probability of ≤5% and ≤10% were 0% and 13%, respectively. Of the patients with a Stanford probability of 100%, 26% did not have NSLN metastasis. CONCLUSIONS: Despite using only 3 variables, the Stanford nomogram appears to perform on a par with, but not better than, the MSKCC and Mayo nomograms. Further validation in other patient populations is needed.

AB - BACKGROUND: Several models for the prediction of nonsentinel lymph node (NSLN) metastasis in sentinel lymph node (SLN)-positive breast cancer patients have been proposed. In this study, the authors evaluate the Stanford Online Calculator (SOC), which was designed to predict the likelihood of NSLN metastasis using only 3 variables: primary tumor size, SLN metastasis size, and angiolymphatic invasion status. They compared it with the Mayo and Memorial Sloan-Kettering Cancer Center (MSKCC) nomograms. METHODS: The SOC was used to calculate the probability of NSLN metastasis in 464 breast cancer patients with SLN metastasis who underwent completion axillary lymph node dissection at the Mayo Clinic. The area under the receiver operating characteristic curve (AUC) was calculated for each model. Mean probabilities of patients with and without NSLN metastasis were compared. Patients with ≤5%, ≤10%, and 100% NSLN metastasis probabilities were examined. RESULTS: The AUCs of the Stanford, MSKCC, and Mayo models were 0.72, 0.74, and 0.77, respectively (P = .13). The mean Stanford probabilities for patients with and without NSLN metastasis were 0.75 (range, 0.06-1.0) and 0.50 (range, 0.05-1.0), respectively (P < .0001). The falsenegative rates for patients with a Stanford probability of ≤5% and ≤10% were 0% and 13%, respectively. Of the patients with a Stanford probability of 100%, 26% did not have NSLN metastasis. CONCLUSIONS: Despite using only 3 variables, the Stanford nomogram appears to perform on a par with, but not better than, the MSKCC and Mayo nomograms. Further validation in other patient populations is needed.

UR - http://www.scopus.com/inward/record.url?scp=70149086903&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=70149086903&partnerID=8YFLogxK

U2 - 10.1002/cncr.24469

DO - 10.1002/cncr.24469

M3 - Article

C2 - 19517477

AN - SCOPUS:70149086903

VL - 115

SP - 4064

EP - 4070

JO - Cancer

JF - Cancer

SN - 0008-543X

IS - 18

ER -