Multi-variant genetic panel for genetic risk of opioid addiction

Keri Donaldson, Laurence Demers, Kirk Taylor, Joe Lopez, Sherman Chang

Research output: Contribution to journalArticle

2 Citations (Scopus)

Abstract

Over 116 million people worldwide have chronic pain and prescription dependence. In the US, opioids account for the majority of overdose deaths, and in 2014, almost 2 million Americans abused or were dependent on prescription opioids. Genetic factors may play a key role in opioid prescription addiction. Herein, we describe genetic variations between opioid addicted and non-addicted populations and derive a predictive model determining risk of opioid addiction. This case cohort study compares the frequency of 16 single nucleotide polymorphisms involved in the brain reward pathways in patients with and without opioid addiction. Data from 37 patients with prescription opioid or heroin addiction and 30 age and gender matched controls were used to design the predictive score. The predictive score was then tested on an additional 138 samples to determine generalizabilty. Results for Method Derivation of Observed data: ROC statistic=0.92, sensitivity=82% (95% CI: 66-90), specificity=75% (95% CI:56-87). TreeNet "learn" data: ROC statistic=0.92, sensitivity=92%, specificity=90%,precision=92%, and overall correct=91%. Results of Generalizability data: Sensitivity=97% (95% CI: 90 to 100), specificity=87% (95% CI:86 to 93), positive likelihood ratio=7.3 (95% CI: 4.0 to 13.5), and negative likelihood ratio=0.03 (95% CI:0.01 to 0.13). This negative likelihood ratio can be used as an evidence based measure to exclude patients with a high risk of opioid addicition or substance use disorder. By identifying patients with a lower risk for opioid addiction, our model may inform therapeutic decisions.

Original languageEnglish (US)
Pages (from-to)452-456
Number of pages5
JournalAnnals of Clinical and Laboratory Science
Volume47
Issue number4
StatePublished - Jul 1 2017

Fingerprint

Opioid Analgesics
Prescriptions
Statistics
Heroin Dependence
Heroin
Polymorphism
Reward
Chronic Pain
Substance-Related Disorders
Single Nucleotide Polymorphism
Brain
Cohort Studies
Nucleotides
Sensitivity and Specificity
Population

All Science Journal Classification (ASJC) codes

  • Microbiology
  • Immunology and Allergy
  • Pathology and Forensic Medicine
  • Immunology
  • Molecular Biology
  • Hematology
  • Clinical Biochemistry
  • Medical Laboratory Technology

Cite this

Donaldson, Keri ; Demers, Laurence ; Taylor, Kirk ; Lopez, Joe ; Chang, Sherman. / Multi-variant genetic panel for genetic risk of opioid addiction. In: Annals of Clinical and Laboratory Science. 2017 ; Vol. 47, No. 4. pp. 452-456.
@article{b04af366419245acabe831826ef07402,
title = "Multi-variant genetic panel for genetic risk of opioid addiction",
abstract = "Over 116 million people worldwide have chronic pain and prescription dependence. In the US, opioids account for the majority of overdose deaths, and in 2014, almost 2 million Americans abused or were dependent on prescription opioids. Genetic factors may play a key role in opioid prescription addiction. Herein, we describe genetic variations between opioid addicted and non-addicted populations and derive a predictive model determining risk of opioid addiction. This case cohort study compares the frequency of 16 single nucleotide polymorphisms involved in the brain reward pathways in patients with and without opioid addiction. Data from 37 patients with prescription opioid or heroin addiction and 30 age and gender matched controls were used to design the predictive score. The predictive score was then tested on an additional 138 samples to determine generalizabilty. Results for Method Derivation of Observed data: ROC statistic=0.92, sensitivity=82{\%} (95{\%} CI: 66-90), specificity=75{\%} (95{\%} CI:56-87). TreeNet {"}learn{"} data: ROC statistic=0.92, sensitivity=92{\%}, specificity=90{\%},precision=92{\%}, and overall correct=91{\%}. Results of Generalizability data: Sensitivity=97{\%} (95{\%} CI: 90 to 100), specificity=87{\%} (95{\%} CI:86 to 93), positive likelihood ratio=7.3 (95{\%} CI: 4.0 to 13.5), and negative likelihood ratio=0.03 (95{\%} CI:0.01 to 0.13). This negative likelihood ratio can be used as an evidence based measure to exclude patients with a high risk of opioid addicition or substance use disorder. By identifying patients with a lower risk for opioid addiction, our model may inform therapeutic decisions.",
author = "Keri Donaldson and Laurence Demers and Kirk Taylor and Joe Lopez and Sherman Chang",
year = "2017",
month = "7",
day = "1",
language = "English (US)",
volume = "47",
pages = "452--456",
journal = "Annals of Clinical and Laboratory Science",
issn = "0091-7370",
publisher = "Association of Clinical Scientists",
number = "4",

}

Donaldson, K, Demers, L, Taylor, K, Lopez, J & Chang, S 2017, 'Multi-variant genetic panel for genetic risk of opioid addiction', Annals of Clinical and Laboratory Science, vol. 47, no. 4, pp. 452-456.

Multi-variant genetic panel for genetic risk of opioid addiction. / Donaldson, Keri; Demers, Laurence; Taylor, Kirk; Lopez, Joe; Chang, Sherman.

In: Annals of Clinical and Laboratory Science, Vol. 47, No. 4, 01.07.2017, p. 452-456.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Multi-variant genetic panel for genetic risk of opioid addiction

AU - Donaldson, Keri

AU - Demers, Laurence

AU - Taylor, Kirk

AU - Lopez, Joe

AU - Chang, Sherman

PY - 2017/7/1

Y1 - 2017/7/1

N2 - Over 116 million people worldwide have chronic pain and prescription dependence. In the US, opioids account for the majority of overdose deaths, and in 2014, almost 2 million Americans abused or were dependent on prescription opioids. Genetic factors may play a key role in opioid prescription addiction. Herein, we describe genetic variations between opioid addicted and non-addicted populations and derive a predictive model determining risk of opioid addiction. This case cohort study compares the frequency of 16 single nucleotide polymorphisms involved in the brain reward pathways in patients with and without opioid addiction. Data from 37 patients with prescription opioid or heroin addiction and 30 age and gender matched controls were used to design the predictive score. The predictive score was then tested on an additional 138 samples to determine generalizabilty. Results for Method Derivation of Observed data: ROC statistic=0.92, sensitivity=82% (95% CI: 66-90), specificity=75% (95% CI:56-87). TreeNet "learn" data: ROC statistic=0.92, sensitivity=92%, specificity=90%,precision=92%, and overall correct=91%. Results of Generalizability data: Sensitivity=97% (95% CI: 90 to 100), specificity=87% (95% CI:86 to 93), positive likelihood ratio=7.3 (95% CI: 4.0 to 13.5), and negative likelihood ratio=0.03 (95% CI:0.01 to 0.13). This negative likelihood ratio can be used as an evidence based measure to exclude patients with a high risk of opioid addicition or substance use disorder. By identifying patients with a lower risk for opioid addiction, our model may inform therapeutic decisions.

AB - Over 116 million people worldwide have chronic pain and prescription dependence. In the US, opioids account for the majority of overdose deaths, and in 2014, almost 2 million Americans abused or were dependent on prescription opioids. Genetic factors may play a key role in opioid prescription addiction. Herein, we describe genetic variations between opioid addicted and non-addicted populations and derive a predictive model determining risk of opioid addiction. This case cohort study compares the frequency of 16 single nucleotide polymorphisms involved in the brain reward pathways in patients with and without opioid addiction. Data from 37 patients with prescription opioid or heroin addiction and 30 age and gender matched controls were used to design the predictive score. The predictive score was then tested on an additional 138 samples to determine generalizabilty. Results for Method Derivation of Observed data: ROC statistic=0.92, sensitivity=82% (95% CI: 66-90), specificity=75% (95% CI:56-87). TreeNet "learn" data: ROC statistic=0.92, sensitivity=92%, specificity=90%,precision=92%, and overall correct=91%. Results of Generalizability data: Sensitivity=97% (95% CI: 90 to 100), specificity=87% (95% CI:86 to 93), positive likelihood ratio=7.3 (95% CI: 4.0 to 13.5), and negative likelihood ratio=0.03 (95% CI:0.01 to 0.13). This negative likelihood ratio can be used as an evidence based measure to exclude patients with a high risk of opioid addicition or substance use disorder. By identifying patients with a lower risk for opioid addiction, our model may inform therapeutic decisions.

UR - http://www.scopus.com/inward/record.url?scp=85027303714&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85027303714&partnerID=8YFLogxK

M3 - Article

C2 - 28801372

AN - SCOPUS:85027303714

VL - 47

SP - 452

EP - 456

JO - Annals of Clinical and Laboratory Science

JF - Annals of Clinical and Laboratory Science

SN - 0091-7370

IS - 4

ER -