To the Editor:
We would like to acknowledge the contribution of Cai et al. in their January 2012 publication “Association between apolipoprotein E4 and postoperative cognitive dysfunction in elderly patients undergoing intravenous anesthesia and inhalation anesthesia.”1Prospective studies in this realm are sorely lacking, and a study of this magnitude is indeed an achievement. However, we do have some questions about the methodology and external validity of their findings.
In a prospective cohort of elderly patients undergoing general anesthesia and major noncardiac surgery, Cai et al. demonstrated enormously disparate postanesthesia Mini-Mental State Examination (MMSE) scores between patients who received total intravenous anesthesia and those who received isoflurane. The MMSE includes questions about orientation to time and place (which may include specific information like the floor of the building the patient is in), immediate and delayed recall of words, attention and calculation, language skills, and ability to follow complex commands, such as copying a figure. Although the MMSE is not a very descriptive tool for perioperative cognition, extreme scores have value. It is unlikely that a patient with a perfect score (30) is disoriented, but patients in the range of 25–30 may comprise a broad range of cognitive ability. In their analysis of apolipoprotein E genotype effects, there was a relatively higher incidence of apolipoprotein E ε4 allele in patients who received isoflurane (but not total intravenous anesthesia) and had MMSE scores less than 25 at any time-point postoperatively. The authors conclude this suggests that apolipoprotein E ε4 allele is a risk factor for cognitive dysfunction in patients who receive inhalational anesthesia. The possibility remains that significant variance may be explained by baseline MMSE status in these groups, as patients who move from a score of 25 preanesthesia to 24 postanesthesia would be classified the same as patients who moved from a score of 30 to 10. The causal link between exposure to an anesthetic and cognitive dysfunction would be more clearly represented as change from preoperative baseline score.
Irrespective of the effects of apolipoprotein E genotype, the effects of isoflurane anesthesia on postoperative MMSE reported by Cai et al. are not subtle. In figure 2, it appears that profound cognitive impairment persisted to at least 3 days after surgery in the group that received isoflurane.†Regarding the clinical correlate of the MMSE scores seen, patients with early Alzheimer's disease may score 24 or less, scores between 10 and 19 indicate moderate dementia, and scores less than 10 generally denote severe impairment. In this study patients did not recover to a score above 20, which still represents significant impairment, until somewhere between 3 and 10 days postoperatively. In other words, patients who were generally healthy with MMSE of 25 or more (and mean more than 27) and no major heart, lung, or renal dysfunction became floridly demented for at least 3 days after surgery following isoflurane anesthesia. We find this extremely unusual. In many cases we would expect patients to be conversational and ambulatory by the end of 48 h after procedures that included nephrectomy, gastrectomy, and esophagectomy. If this type of major impairment persisted at our institution, where more than 80% of elderly patients receive isoflurane for major surgery, it would prevent mobilization of a massive number of postsurgical patients and create a huge public health problem.
Very little is discussed regarding the anesthetic technique in the study by Cai et al. Depth of anesthesia is described as only actively managed before intubation and otherwise is noted to have been “normal.” It is unclear if the two groups have equivalent anesthetic depths throughout the procedure, which could have accounted for at least the early difference but not the prolonged recovery. Alarmingly, the authors state that patients in the inhalation group, after a loading dose of propofol, received “continuous inhalation [of] 2–3% end-tidal concentration of isoflurane, which was used for maintenance of anesthesia.” Using equation 3 of Nickalls and Mapleson,2this expired concentration of isoflurane represents a range of 2.06 to 3.09 minimum alveolar concentration of isoflurane for a 70-yr-old (the mean age of patients in their study). The carrier gas is not specified, but if it contained nitrous oxide, the effective minimum alveolar concentration would be much higher. The minimum alveolar concentration value of isoflurane in 100% O2in this age group is 0.97; with 50% N2O, the minimum alveolar concentration of isoflurane is 0.41. Thus, their results may reflect a toxic effect of administration of a very high dose of isoflurane. In contrast, the group that received intravenous anesthesia had age-appropriate serum concentration of propofol,3around 1 μg/ml. Postoperative visual analog scale score, incidence of delirium, and postoperative complications by group were also not mentioned. Intraoperative physiologic parameters, such as blood pressure and bispectral index, are also not mentioned.
Cai et al. admit that there are issues of external validity in comparing the results of their studies of elderly Han Chinese patients with other populations. We are concerned that given the magnitude of the cognitive decline seen in patients who received inhalational anesthesia, and the apparently exceptionally high levels of inhalation anesthesia that were used, that there are significant differences other than ethnicity between patients in the study and the standard of anesthetic care for elderly patients. In practice, we administer general anesthesia with isoflurane to millions of elderly patients per year, and we are not on average seeing 3 days of profound cognitive impairment in previously healthy patients. We suggest that duplication of this study in other patients would not yield similar results and do not believe that there is sufficient cause to suspend use of isoflurane maintenance anesthetic in elderly patients.