The signal averaged P wave : a non-invasive marker of atrial electrophysiological substrate

The technological advances made over the last century have afforded the clinician an array of sophisticated tests to aid the diagnostic process. Much of the knowledge gained on the pathophysiology of cardiac disease has been from invasive assessment, often in animals, but also in human subjects. Application of this knowledge to patient care is limited by the need for invasive studies that present some risk of harm to the patient. Non-invasive assessment reduces risk of harm significantly whilst providing information equivalent to invasive assessment. The best example of this is the insight delivered by technological advances in imaging of the heart. Ultrasound echocardiography, radio-isotope imaging, computerised tomography and magnetic resonance imaging have all excelled expectations in delivering accurate anatomic and functional information non-invasively. Assessment of electrophysiologic function began non-invasively with the recording of surface potentials by Augustus D Waller1 and the development of the electrocardiogram (ECG) by William Einthoven2 (who was later awarded the Nobel Prize in 1924 for his endeavours). Recognition of pathology from the surface ECG was hypothesis generating. In order to explore the heart's electrical system further electrophysiological assessment was made invasively to supplement the information obtained from the surface ECG. This information proved favourable and when combined with pacing stimulus protocols provided the clinician with detailed information on conduction properties that could be measured in a reproducible and reliable way to reflect the impact of drugs or disease in detail that the surface ECG could not. Moreover, the invasively measured properties could be linked with changes at the cellular level and thus the effect of changes in ion channel density, for example, on electrophysiologic properties could be predicted. It would obviously be beneficial to somehow gather the information non-invasively, but this has proved more challenging. Firstly, much of the information obtained invasively is the product of pacing protocols that cannot be reproduced non-invasively. Secondly, the detailed assessment of cardiac electrophysiology from surface electrograms is hampered by multiple factors pertaining to the intervening tissue, i.e. body habitus and electrical interference (noise). Given these factors the expectations of non-invasive assessment of cardiac electrophysiology must be limited and cannot be compared to imaging. The utility of non-invasive tests must be in the 'broad stokes' rather than the fine detail. However it is not beyond expectations to provide useful insights that may be employed in the investigation of disease trends and or the impact of intervention. The trade off for lack of detailed information is the safety, low cost and general applicability to a large patient population. The difficulty in gathering further information from the surface ECG has been alluded to briefly above. Digital techniques are used to overcome some of the difficulties such as amplification and noise reduction. Digital applications are then often used to analyse the data gathered. It is useful to be familiar with some of the concepts involved in digital signal processing as it pertains to cardiac signals and thus a brief outline is presented in appendix 1. This thesis begins with a detailed review of the surface P wave in health and disease, and a review of atrial fibrillation (AF) - the most common arrhythmia encountered in clinical practice.