Article Text

Effects of cardiac resynchronisation on maximal and submaximal exercise performance in advanced heart failure patients with conduction abnormality
  1. D BIRNIE,
  2. L P SOUCIE,
  3. S SMITH,
  4. A S L TANG
  1. Ottawa Heart Institute
  2. 40 Ruskin Road
  3. Ottawa, Ontario K1Y 4W7
  4. Canada
  1. Dr Birnie; davidbirnie{at}

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Atriobiventricular pacing (cardiac resynchronisation) is a promising treatment for patients with advanced heart failure and conduction abnormality.1 These patients have severely impaired exercise tolerance and this has been shown to be multifactorial. The present study examined, in detail, the effect of cardiac resynchronisation on maximal and submaximal exercise capacity as evaluated by metabolic exercise testing.

A consecutive series of patients from a single centre with cardiac resynchronisation therapy, able to perform baseline and follow up treadmill tests, were enrolled in the study. Inclusion criteria were New York Heart Association (NYHA) functional class III/IV heart failure on maximal medical treatment, left ventricular ejection fraction < 35%, left ventricular end diastolic dimension > 60 mm, and QRS duration > 130 ms. Patients with chronic atrial fibrillation were excluded. The protocol was approved by the hospital research and ethics committee and all subjects gave written informed consent.

Left ventricular (LV) stimulation was accomplished by a lead inserted into an LV branch of the coronary sinus. Specially designed leads were used (Medtronic 2187 or 2188, Medtronic Inc, Minneapolis, Minnesota, USA). Right atrial and ventricular leads were positioned in the right atrial appendage and right ventricular apex, respectively. All patients received cardiac resynchronisation devices (InSync, Medtronic Inc) with one atrial and two ventricular ports. Optimal atrioventricular delays were individually determined by analysing transmitral flow using Doppler echocardiography.

Exercise testing was performed before and 1–3 months after implantation. Patients underwent symptom limited maximal breath-by-breath cardiopulmonary exercise testing at baseline and follow up using the Modified Naughton protocol with continuous electrocardiographic recording. Oxygen consumption (V˙o 2, ml/min), carbon dioxide production (V˙co 2, ml/min) and minute ventilation (V˙E, l/min) were continuously measured using an automated breath by breath system (Med Graphics CPX-D, Medical Graphics Co, St Paul, Minnesota, USA). Anaerobic threshold was determined by the V slope method and confirmed by the ventilatory equivalence method. The patients were all encouraged to exercise until the respiratory exchange ratio (that is,V˙co 2/V˙o 2) was greater than 1.0 which is a good indication that near maximal effort has been reached. Maximal exercise capacity was assessed by exercise duration and peak oxygen consumption (peakV˙o 2). Submaximal exercise capacity was assessed in three ways: by measuring oxygen consumption at anaerobic threshold (V˙o 2@AT)2; by examining continuous oxygen uptake using the oxygen uptake efficiency slope (OUES)3; and by measuring the respiratory response to exercise (V˙E/V˙co 2slope).4

We studied 15 patients (13 males and two females); 14 had NYHA class III and one had class IV heart failure. The aetiology was ischaemic in 10 patients, valvar in two, and idiopathic in three. Mean (SD) age was 66.2 (9.6) years. Baseline ejection fraction was 23.6 (7.5)%, left ventricular end diastolic dimension was 71.9 (9.5) mm, QRS duration was 169.3 (27.6) ms, and PR interval was 226.1 (44.8) ms. Fourteen patients had left bundle branch block and one had a pre-existing standard dual chamber pacemaker. QRS duration was reduced to 148.1 (13.5) ms after atriobiventricular pacing. The reduction in QRS duration was 21.1 (22.4) ms.

All 15 patients completed the metabolic exercise test at baseline and at follow up. Anaerobic threshold was identifiable in all exercise tests. All patients reached a respiratory exchange ratio of ⩾ 1.0 with a mean (SD) of 1.17 (0.09) at baseline. At follow up, the respiratory exchange ratio was 1.17 (0.12); only one patient had a respiratory exchange ratio of < 1.0.

The exercise results are listed in table 1. Both maximal parameters and two out of three submaximal parameters improved after pacing. There was no significant correlation between baseline QRS or QRS width reduction and any of the parameters.

Table 1

Comparison of exercise parameters before and after cardiac resynchronisation (n=15)

Maximal exercise capacity as assessed by peakV˙o 2 is a strong independent predictor in a broad range of heart failure patients. Indeed peakV˙o 2 is widely accepted as the best single measure of prognosis in ambulatory patients with severe heart failure.5 However, peak V˙o 2is poorly correlated with everyday physical activity which is largely a series of submaximal exercises. This may explain why there is a poor correlation between self reported exertional symptoms and maximal exercise capacity. In this regard, submaximal exercise parameters may provide better estimation of functional capacity of daily living.

Oxygen consumption at anaerobic threshold (V˙o 2@AT) has been used as an estimate of submaximal capacity in heart failure patients.2 There was a small increase in this variable from 10.0 (1.8) ml/min/kg before to 10.8 (1.6) ml/min/kg with cardiac resynchronisation, although this did not reach significance (p = 0.13). HoweverV˙o 2@AT has recognised limitations.2

The kinetics of oxygen consumption at the start and end of exercise are delayed in patients with heart failure. Therefore an index of oxygen kinetics before anaerobic threshold might better reflect submaximal exercise capacity and thus patient's exertional symptoms. The OUES is one such index and was first proposed and validated as a new index of submaximal cardiorespiratory reserve by Baba and colleagues in 1996.3 Also, OUES is effort independent and thus less likely to be affected by subjective factors including training effects or intra-observer variability. For these reasons, OUES may be the most valid assessment of submaximal exercise capacity. In this study OUES showed a major improvement from 517.6 (148.2) at baseline to 612.0 (210.2) with cardiac resynchronisation (p = 0.02).

Respiratory efficiency (V˙E/V˙co 2slope) is also a measure of submaximal exercise capacity. Heart failure patients often have an excessive ventilatory response to exercise4 and this can be quantified by analysing the relation between carbon dioxide output (V˙co 2) and minute ventilation (V˙E). In our study there was a clear and significant improvement in this variable with cardiac resynchronisation (41.8 (7.1) beforev 38.1 (9.0) after, p = 0.04).

In conclusion, this is the first detailed analysis of the effects of cardiac resynchronisation on exercise performance and our results show that atriobiventricular pacing significantly improved maximal and submaximal exercise capacity. The study has significant limitations in being small and uncontrolled; however, larger randomised studies are underway to investigate whether these positive effects of cardiac resynchronisation on exercise performance are real and can be translated into an impact on mortality.


We are grateful for the assistance of our research nurses Patricia Theorot-Patrick and Clare Carey.