Benefits of Centralized ECG Reading in Clinical Oncology Studies

Abstract

Background

Many clinical trials of investigational oncologic agents utilize electrocardiogram (ECG) machine measurements of QTc, for inclusion/exclusion and dosing decisions, though their reliability in this setting has not been established.

Methods

We compared the digital ECG machine QTc measurements with those obtained by a centralized ECG core lab on more than 270,000 consecutive ECGs collected from 299 clinical oncology trials.

Results

The mean difference between the ECG machine measurements and the central measured QTcF was 1.8 ± 15.7 milliseconds. In addition, 29.7% of ECGs with an ECG machine-measured QTcF >450 milliseconds had a centrally measured QTcF <450 milliseconds, 44.6% of ECGs with an ECG machine-measured QTcF >470 milliseconds had a centrally measured QTcF <470 milliseconds, and 77.2% of ECGs with an ECG machine-measured QTcF >500 milliseconds had a centrally measured QTcF <500 milliseconds. The likelihood of a large discrepancy between the ECG machine- and centrally measured value for QTcF increased at both the high and low ends of the range of ECG machine QTcF measurements.

Conclusions

While on average ECG machine-measured QTcF values were very similar to the central core lab measurements; there were very significant discrepancies which will have important implications for patient recruitment for clinical oncology trials as well as for patient safety during dosing with new oncologic agents. Reliance on ECG machine QTc measurements during clinical oncology trials may lead to unnecessary exclusion of patients as well as unneeded treatment interruptions.