Predicting Persistent Opioid Use, Abuse and Toxicity Among Cancer Survivors

Abstract

While opioids play a critical role in the management of cancer pain, the ongoing opioid epidemic has raised concerns regarding their persistent use and abuse. An individualized assessment of benefits and risks is required for optimal management with opioids. Current guidelines for risk stratification, however, are based on expert opinion or instruments validated in non-oncology cohorts that omit numerous relevant factors for cancer patients. The purpose of this study was to develop and validate a model to predict the risk of persistent opioid use, abuse and toxicity among cancer survivors. This study used a retrospective cohort of 106,732 non-metastatic cancer survivors with non-hematologic or cutaneous malignancies (bladder, breast, colon, esophagus, stomach, head and neck, kidney, liver, lung, pancreas, prostate, or rectal) from the Veterans Affairs (VA) Informatics and Computing Infrastructure (VINCI) database. We determined rates of persistent opioid use, diagnoses of opioid abuse or dependence, and admissions for opioid toxicity by cancer type. A multivariable logistic regression model was used to identify patient, cancer, and treatment risk factors for persistent opioid use. Predictive models for persistent opioid use, opioid abuse, and opioid related admissions were developed using least absolute shrinkage and selection operator (LASSO) regression and tested in an independent validation cohort. The prevalence of persistent opioid use in cancer survivors was 8.3% overall and 6.2% among opioid-naïve patients with a new prescription during treatment. Rates of opioid abuse or dependence and opioid related admissions were 2.9 and 2.1%, respectively. Advanced age (HR: 0.71 [per 10 years], 95% CI: 0.6-0.74), black race (0.76, 0.71-0.81), employment (0.67, 0.61-0.74) and living in a zip code with a higher median income (0.97, 0.97-0.99) were associated with a decreased adjusted risk for persistent use. Prior diagnoses of alcohol abuse (1.11, 1.03-1.19), non-opioid drug abuse (1.07, 0.97-1.18), opioid drug abuse (1.39, 1.2-1.6), depression (1.46, 1.37-1.55) and increased preexisting comorbidities (1.21 [CCI ≥ 3], 1.21-1.3) were associated with increased adjusted risk. Validation of the predictive models in an independent test set showed a high level of accuracy in predicting persistent opioid use (AUC: 0.869), future diagnoses of opioid abuse or dependence (0.883) and admission for opioid abuse or toxicity (0.785). Patient-specific risks of persistent opioid use, opioid abuse, and opioid related admissions can be estimated for cancer survivors. Personalized risk stratification should be used to guide management when prescribing opioids in cancer patients.