Pro

In this issue of Anesthesia & Analgesia, Tan et al.1 report on a trial that tests the hypothesis that patients anesthetized with propofol have less pain and a better quality of recovery (as measured by a previously validated questionnaire, QoR-9) compared with patients anesthetized with sevoflurane. In this prospective, double-blind, randomized trial, the authors used a particularly clean study design in which one group had an inhaled induction with sevoflurane followed by sevoflurane maintenance, whereas the other group had an IV induction with propofol followed by propofol maintenance. The subjects were treated prophylactically during surgery with equivalent doses of alfentanil, paracetamol, and diclofenac for pain and dexamethasone and ondansetron for nausea. Pain was treated after surgery using 2-mg morphine doses until visual analog scale score was <4 and then oral oxycodone thereafter. The authors found that propofol provided a small but statistically significant (P < 0.01) difference (decrease) in postoperative pain. This study followed up on a similar study published in Anesthesia & Analgesia in January 2008 by Cheng et al.2 for which I was the senior author. Our trial was designed to test a related but subtly different hypothesis. We proposed that volatile anesthesia (in our case isoflurane) induces a hyperalgesic state, and that propofol was neutral in its modulation of pain sensitivity. We also found that patients anesthetized with a volatile anesthetic (isoflurane) reported more postoperative pain than those anesthetized with propofol. Our effect size and the degree of statistical significance (P < 0.01) were similar to those found in the current study. Unlike the current study, we also detected a difference in postoperative opioid use with more requirement in those anesthetized with isoflurane. Our article was accompanied by an editorial3 by Drs. Shafer (my husband) and Nekhendzy entitled “Anesthesia Matters: Statistical Anomaly or New Paradigm?” Although the idea that propofol creates favorable emergence conditions was not new, the idea that volatile anesthetics might make postsurgical patients more sensitive to pain was an “extraordinary finding” that should be supported by extraordinary evidence. I agree. The editorialists observed that according to Bayesian logic, one's conclusion must be influenced by both how likely they think the outcome would be before the study (the prior probability) and the quality of the data. A single study that showed that a drug class that we had been using for 50 years enhanced postsurgical pain, even with a P value <0.01, was insufficient evidence to support a change in practice. With their editorial, the authors called for new prospective studies to determine whether the findings of Cheng et al. represented the identification of a new virtue for propofol anesthesia that could be added to protection against nausea, vomiting, and peroxidative injury or was simply a statistical anomaly. The study by Tan et al. is presumably in response to their request. Now the evidence for a postoperative pain benefit from propofol anesthesia compared with a volatile anesthetic is supported as a primary outcome in 2 double-blind randomized trials. In both trials, it is a relatively small effect of 1 to 2 U on a 10-point numerical rating scale (or visual analog scale). However, in both studies, the effect persisted during the entire time period studied (24 hours in Cheng et al. and 4 hours in Tan et al.). Because isoflurane, sevoflurane, and propofol concentrations decrease to infinitesimally low levels over this time period, it seems unlikely that the persistent pain benefit is a direct drug effect. It is plausible that volatile anesthetics, including sevoflurane, have pronociceptive activity that might outlast the presence of the drug. It might exacerbate peripheral and or central sensitization in response to a surgical incision. Propofol might potentially do the reverse. Might, might, might …. We clearly have much to learn, but the speculation is not too farfetched. There is extensive animal literature that addresses the modulatory effects of anesthetics in different nociceptive paradigms. There is consensus that volatile anesthetics at low, subanesthetic concentrations cause a hyperalgesic state when tested with heat, pressure, and in a postoperative pain paradigm. The mechanism of the enhanced sensitivity is not well understood, but regulation of central nicotinic and noradrenergic facilitation has been implicated.4,5 The nociceptive consequences of propofol administration are more complicated and likely dose dependent. In the 1990s, Ewen et al.6 found that in rats a slow IV infusion of propofol resulted in “an initial decline followed by a rise in nociceptive threshold as the plasma concentration and degree of sedation increased,” suggesting that concentrations of propofol smaller than those that induce sedation are responsible for hyperalgesia. However, when we repeated similar experiments in a postoperative pain paradigm in mice, we were unable to detect any hyperalgesic phase at lower than sedative doses of propofol or on emergence.7 Other groups have found an analgesic response to propofol, particularly in inflammatory pain paradigms.8 Clearly, the animal data on nociceptive effects mediated by propofol are mixed. These types of clinical trials give us little insight into the mechanism of the effect that they demonstrate. However, this second double-blind randomized trial does give more weight to the surprising and potentially disturbing claim that the administration of a class of anesthetics that predominate in the United States for the maintenance of anesthesia might make our patients suffer more pain when they emerge from our ministrations. In contrast, total IV anesthesia with propofol may provide more advantages than previously thought. Although we still don't know whether this is a new paradigm or a statistical anomaly, the evidence for a new paradigm is accumulating.