Introduction

Each year more than half a million US individuals undergo coronary-artery bypass grafting (CABG), and the procedure is established as being effective in revascularizing the heart. With improving anesthetic and surgical techniques the age of candidates for CABG has risen, so that older patients with multiple health problems often undergo the procedure.

The most serious complication of CABG is stroke, reported variously in 1.5% to 5.2% of cases. More common, although less serious, is postoperative cognitive decline, which is described in 33% to 82% of patients on discharge from hospital; the wide variability in these figures depends largely on the test methods used to evaluate cognition.

Cognitive decline improves in the majority of patients in the months following CABG, so that by six months the rate is between 10% and 30%. However, a new study, summarized here, has determined the course of cognitive changes over a 5-year period, with some disturbing results.

Method

This study was conducted at Duke Medical Center, N Carolina, USA. 261 patients who underwent CABG were enrolled; they had no symptomatic cerebrovascular disease, psychiatric, renal or liver disease. A battery of neurocognitive tests was administered preoperatively, before discharge, and 6 weeks, 6 months and 5 years after the surgery.¹

A decline in cognitive function was defined as a drop of at least one Standard Deviation (SD) in the test scores for any of these 4 domains: verbal memory and language comprehension; abstraction and visuospatial orientation; attention, psychomotor processing speed and concentration; and visual memory. A reduction of one SD represented a decline of about 20% in function.

To assess overall cognitive function across all domains, a composite cognitive index was established by adding the four domain scores. Various predictors of change in cognitive function were analysed by logistic regression.

Results

Of the 261 patients enrolled, 172 provided full testing data that included the 5-year follow-up assessment. The three main reasons for loss to follow-up were: inability to contact the patient, death, and ill health of the patient. Demographic data at enrolment and duration of the surgical interventions were the same for 'completers' and 'non-completers'. Their mean age was 61 years, and 72% were male.

Cognitive decline, as defined above, was found in 53% of the patients at discharge from hospital. This rate decreased to 36% at 6 weeks, and 24% at 6 months. Five years after CABG the rate of cognitive decline was 42%.

The composite cognitive index showed a similar gradual improvement up to 6 months. However, the 5-year values were different for those subjects with cognitive decline at discharge from hospital compared to those without such decline. In patients without evidence of early cognitive decline, the scores at 5 years were near the baseline scores. For those with a decline evident at discharge, the 5-year scores, after the 6-week and 6-month improvements, were found to be similar to the discharge scores.

Older age, a lower level of education, a higher preoperative score for cognitive function, and the presence of cognitive decline at discharge were all predictors of cognitive decline at 5 years after CABG. Of these, the most significant predictor was a decline in cognition seen at discharge. Other factors analysed which were found not to be predictive of 5-year cognitive decline, were: gender, duration of cardiopulmonary bypass, and the duration of aortic cross-clamping; the left ventricular ejection fraction approached significance in this analysis, suggesting that cardiac dysfunction may be a significant predictor.

Comment

What conclusions are to be drawn from this? The authors of the study and an accompanying editorial in the New England Journal of Medicine²have commented on the possible implications. First, there was no control group in the study, so it is possible that the decline seen was representative of the 'normal' decline with age. (The improvement in test results seen at 6 weeks and 6 months may have been because of test-learning processes - this would also be revealed in a controlled study).

Second, short-term declines in cognitive function are reported in elderly subjects after non-cardiac surgery, and this can persist in a proportion of these patients - in fact, in 10% after 2 years.³

Third, what is the mechanism of the effect - microembolism, hypoperfusion, or both? Can anesthetic or surgical techniques be modified to reduce one or both of these potential factors? 'Beating heart' surgery represents one such approach at present under evaluation.

Finally, should all patients who exhibit a post-CABG decline in cognitive function be treated with intensive measures in an attempt to restore, or maintain, this function? Such changes include modifications in lifestyle (smoking, drinking, exercise), diet, and medications that may delay the vasculopathy.