Prediction of cardiac mortality in hemodialysis (HD) patients based on T-wave Principal Component Analysis (PCA-T)

Almost every dialysis patient has underlying cardiac disease, placing them at risk for cardiac death. The QT interval (a measure of repolarization) is often prolonged in dialysis patients and has been found to be a predictor of sudden death as well as for the QT dispersion. Recently, a prolonged PCA-T (> 0.246 for men and > 0.320 for women) applied to 12-lead recordings is found to predict three year mortality in the general population (Okin et al 2002). The aim of this retrospective study was to explore the predictive value of PCA-T for all-cause and cardiac mortality in hemodialysis (HD) patients. We analyzed data from 164 patients (67 women and 97 men) in whom digital ECG recordings were available from previous clinical studies. Standard 12-leads Holter ECGs (H-12 Holter, Mortara Instrument Inc.) were recorded starting 10 minutes before the dialysis session and lasting for at least 4 hours. A median value of PCA-T was computed for each patient throughout the whole ECG recording. The optimal threshold for PCA-T was defined by ROC analysis as 0.366 in women (accuracy= 0.88) and as 0.320 in men (accuracy= 0.76). Deaths were classified as cardiac if caused by myocardial infarction, sudden death from CHD, or congestive heart failure. After a follow-up of 2 years, patients were censored as dead or alive considering the days from the date of the ECG recording. Patients were then divided in two groups depending on the median PCA-T value respect to the above thresholds. Endpoints were all-cause (ACM) and cardiac (CM) mortality. Survival was computed by the Kaplan-Meier analysis (significant if p <0.05). 63% percent of the patients were hypertensive and 23% were diabetics. During the followup, there were 54 deaths (33% of the population ), of which 27 (16,5%) for cardiac causes. The log rank test showed a significant difference between the two groups both for ACM and CVM (p= 0.047, p= 0.039 respectively). Repolarization abnormalities measured by PCA-T seems to predict cardiac death in HD patients. Probably only grossly abnormal values , outside the range of the general population may potentially have a practical purpose. Nonetheless, our results represents the basis for future research addressing the relationship between abnormalities of ventricular repolarization, PCA-T and underlying patients comorbidities as a risk factor for cardiac death.