2. DOES BEATING HEART SURGEY AFFORD ANY CEREBRAL PROTECTION?

Off-pump coronary artery bypass (OPCAB) is currently very topical in anaesthesia literature. One of the many problems associated with coronary artery bypass grafting (CABG) is the so-called "cognitive carnage" associated with cardiopulmonary bypass (CPB). Studies suggest that this may be due to the high number of emboli associated with CPB. In a recent study by Watters et al, OPCAB was shown to be associated with fewer cerebral emboli than with CPB, but does this mean less cognitive deficit?

Watters et al demonstrate in this study that OPCAB is associated with fewer emboli as detected by transcranial doppler (TCD) , and suggest that future studies may show that it is associated with a better neurological outcome

Study design:

• Comparative study
• Non randomised
• Prospective
• Non blinded

Methods:
20 patients were scheduled for CABG, and choice of technique (OPCAB vs CPB) was at the surgeons' discretion(n = 10 in both groups). Anaesthesia was the same in both groups (fentanyl and propofol). A 2 MHz Bidirectional TCD was applied to the right temple to detect emboli in the region of the middle cerebral artery. Emboli were recorded from the time of pericardial incision to last aortic manipulation. Data was analysed using the Mann-Whitney U-test

Results:
Microembolus count (median and range) in the OPCAB group was 3 (0-18), compared to 79 (38-876) in the CPB group (p <0.001), demonstrating that OPCAB was associated with fewer emboli

Study criticisms:

For:

• Measurable endpoints (emboli counts)
• Elegant, simple design
• Appropriate use of statistics (Mann-Whitney U test for nonparametric data)
• Prospective

  Against:
  

• Non-randomised – biggest criticism! (although some may argue that surgeons’ choice is sufficient randomisation!)
• Although it answers the question it sets out to, it does not give us information regarding neurological outcome – no doubt a trial for later consideration!

3. PROPOFOL AS AN ANTIOXIDANT?

Propofol has a structure similar to known antioxidants (such as tocopherol). The ability of propofol to inhibit the formation of lipid peroxides has been found in various media in which free radicals are produced. De La Cruz et al take an in depth look at propofol’s antioxidant properties…

Propofol can protect against lipid peroxidation, claims Cruz et al. Lets take a closer look at their study.

Study design:

• Prospective
• Randomised (at least partly)
• Non blinded

Methods:
60 Patients (ASA 1 and 2) undergoing surgery (not open abdominal surgery) and 12 healthy volunteers were studied. The 60 patients undergoing surgery were allocated to 1 of 3 groups. Allocation to groups 1 and 2 was random, while allocation to group 3 was based upon length of surgery (greater than 1 hour)

Group 1:
Induction with thiopentone 4mg/kg

Group 2:
Induction with propofol 2mg/kg

Group 3:

• TIVA
• Induction propofol 2mg/kg
• Infusion: 10mg/kg for first 10min, 8mg/kg for next 10 mins then 6mg/kg for rest of operation

Group 4:

• Healthy volunteers
• No anaesthetic
• Intralipid 10% iv

Measurements:
The following were measured as indicators of free radical damage within platelets:

• Thiobarbituric acid (measure of lipid peroxidation)
• Glutathione, reduced and oxidised
• Glutathione peroxidase
• Glutathione reductase
• Glutathione transferase

Samples were drawn pre-induction, 5 mins post-induction and, in the case of group 3, 60 mins after the first dose of propofol and analysed for the above substances. All statistical analyses done with Social Program for Statistical Sciences (SPSS). One way analysis of variance and post hoc Bonferroni adjustment used to compare differences within groups. Kruskal-Wallis test used to compare differences among groups. Considered significant when p < 0.05

Results:
No difference between all groups preinduction. No statistical differences between group 1 and 4 post induction, but groups 2 and 3 were significantly different from groups 1 and 4. Propofol significantly inhibited lipid peroxidation in platelets, whereas administration of intralipid (group 4) had no significant on this value. Platelet glutathione was significantly higher in patients given bolus propofol Oxidised glutathione was significantly lower after propofol was given whereas intralipid had no effect on this value.

Glutathione peroxidase levels were lower, G. transferase levels higher after propofol administration. G reductase levels not significantly different. Propofol infusion was associated with even lower levers of oxidised glutathione than bolus alone.

Conclusion:
Propofol does have antioxidant effects in human platelets. This effect may be beneficial in protection against free radical injury. Although this study looked only at platelets (specifically, platelet membranes), in vitro studies demonstrate that propofol has antioxidant effects in other organs too (brain, liver, lung, vessel wall, kidney and heart) Whether this is of clinical value remains to be seen

Study criticisms:

For

• Prospective
• Some randomisation
• Clearly defined exclusion criteria
• Measurable endpoints

  Against:

• Non blinded. Probably not necessary, as analysis done by laboratory, who will be blinded.
• Statistical analysis. This is the part that bothers me most. ANOVA is a good start, to detect if there is a difference between groups, but the Bonferroni puzzles me. It is a little unclear as to which comparisons were made, but appears that all possible comparisons between groups were made. This renders the Bonferroni insenstive, as the value to reach significance decreases as the number of comparisons increases, and results may be called insignificant when in fact they actually are. A better test here may be the Student-Newman-Keuls test, as this gives a more accurate result when considering multiple comparisons. They then use the Kruskal-Wallis test for between group comparisons – previously they have used a parametric test and now they are using a non parametric test. Lastly, they are using repeated measurements on the various groups - would not a repeated measures (2 way analysis of variance) test be more appropriate? Perhaps our editor could help me untangle this conundrum.