Cardiovascular Events Before and After Initiation of Immune Checkpoint Inhibitor Therapy for Cancer: Agreement Between ICD-Coded and -Adjudicated Events

Get Permission

In a study reported in JACC: CardioOncology, Kondapalli et al found that cardiovascular events common in the general population were observed both prior to and after initiation of immune checkpoint inhibitor therapy for cancer; myocarditis was more common posttreatment, but infrequent. Agreement between International Classification of Diseases (ICD) code and adjudication using established definitions was good for some cardiovascular events and poorer for others.

Study Details

The study included all 1,813 patients aged > 18 years who received immune checkpoint inhibitors within the University of Colorado Health System from January 2011 to April 2019. Electronic medical records (EMRs) were used to identify potential cardiovascular events; those of interest included:

  • Myocardial infarction (MI)
  • Hospitalization for unstable angina
  • Hospitalization for heart failure or heart failure exacerbation requiring treatment
  • Transient ischemic attack (TIA)
  • Stroke
  • Hypertensive emergency
  • Noncoronary (peripheral) vascular events
  • Venous thromboembolism (VTE), including deep vein thrombosis (DVT) and pulmonary embolism (PE)
  • Immune checkpoint inhibitor–related myocarditis.

For potential cardiovascular events identified by ICD code, all EMR data—including encounter notes, specialist assessments, diagnostic testing, and medications—were independently reviewed by two cardiologists and adjudicated using prespecified definitions; discrepancies were resolved by consensus. Agreement between ICD code and adjudicated diagnoses was assessed using the kappa statistic.

Patients received immune checkpoint inhibitor treatment for both solid tumors and hematologic malignancies, with the most common cancer types in the population being melanoma (35.5%), lung cancer (25.8%), and kidney/urinary tract cancers (12.4%). Cardiovascular risk factors common at baseline (pre–immune checkpoint inhibitor) included hypertension (48.2%), coronary revascularization (17.0%), diabetes (16.1%), chronic kidney disease (11.4%) and current smoking (11.3%). The most common immune checkpoint inhibitors received were pembrolizumab (46.9%), nivolumab (42.4%), and ipilimumab (26.1%), irrespective of combination treatment. 

Cardiovascular Events

Mean follow-up was 3.2 years pre–immune checkpoint inhibitor and 1.4 years post–immune checkpoint inhibitor. For adjudicated events, incidence rates pre- vs posttreatment were: 11.4% (n = 206) vs 11.3% (n = 205) for VTE, with PE in 4.9% (n = 89) vs 5.4% (n = 97) and DVT in 8.1% (n = 146) vs 8.2% (n = 148); MI in 1.8% (n = 33) vs 3.0% (n = 54); heart failure in 2.2% (n = 40) vs 2.8% (n = 50); stroke in 1.8% (n = 33 ) vs 1.6% (n = 29); and myocarditis in one patient vs six patients (0.3%). Adjudicated hypertensive emergency occurred in one vs three patients and noncoronary vascular events occurred in five vs two patients.


  • Common cardiovascular events occurred both prior to and after initiation of immune checkpoint inhibitor treatment; myocarditis was more common posttreatment, but infrequent (0.3%).
  • Agreement between ICD coding and adjudication were good for VTE and MI, but poorer for other cardiovascular events.

ICD-Coded vs -Adjudicated Events

Apart from MI, ICD coding overestimated the occurrence of cardiovascular events relative to adjudication for each event type. Overall, the numbers of ICD-coded vs -adjudicated events were: 476 vs 381 for VTE; 74 vs 86 for MI; 93 vs 59 for stroke; 10 vs 6 for myocarditis; 215 vs 86 for heart failure; 22 vs 4 for hypertensive emergency; 47 vs 3 for TIA; and 25 vs 1 for unstable angina.  

ICD codes correlated well with adjudicated diagnosis for VTE (kappa = 0.82, 95% confidence interval [CI] = 0.79–0.85) and MI (kappa = 0.74, 95% CI = 0.66–0.82); less well for heart failure (kappa = 0.47, 95% CI = 0.40–0.54) and myocarditis (kappa = 0.50, 95% CI = 0.20–0.80); and poorly for hypertensive emergency (kappa = 0.23, 95% CI = 0.01–0.44), TIA (kappa = 0.12,  95% CI = –0.00 to 0.24), and unstable angina (kappa = 0.08, 95% CI = –0.06 to 0.22).

The investigators concluded, “ICD codes correlated well with adjudicated events for VTE and MI, but correlation was worse for heart failure and myocarditis. Adjudication with standardized definitions can enhance understanding of the incidence of cardiovascular events related to [immune checkpoint inhibitor] therapy.

Lavanya Kondapalli, MD, of the Division of Cardiology, University of Colorado School of Medicine, Aurora, is the corresponding author for the JACC: CardioOncology article.

Disclosure: The authors reported that there was no funding for the study. For full disclosures of the study authors, visit

The content in this post has not been reviewed by the American Society of Clinical Oncology, Inc. (ASCO®) and does not necessarily reflect the ideas and opinions of ASCO®.