Potency, Impotency, and Importance!

Abstract

In Response: We would like to thank Drs. Polley and Columb for their comments and appreciate the opportunity to respond. Before addressing the specific issues central to this discussion, we would like to comment on the up-down sequential allocation study design. Drs. Polley and Columb use this study design to estimate the 50% effective dose (ED50) concentrations of local anesthetics and indirectly assess relative potencies (1–3). Although similar study designs were used to define minimum alveolar concentration for volatile anesthetics, up-down study designs can theoretically be used to compare any class of drugs. In fact, we have utilized the up-down study design to estimate the ED50 values of intrathecal fentanyl and sufentanil (4,5). The primary benefit of the up-down study design is that all data points concentrate around the ED50 and by open-label design, each subsequent patient’s dose varies according to the previous patient’s response. As a result, fewer patients are needed to estimate the ED50 than are needed with traditional dose-response studies. Although traditional dose response studies enroll many more patients and are time consuming, the entire dose-response curve is estimated, not only an isolated point along the curve. Up-down studies only estimate the ED50 and nothing more. In fact, the ED50 estimated in an up-down study can vary considerably because it is directly related to the study’s methodology as the endpoints are arbitrarily chosen by the investigators. For example, the endpoint criteria for a “success” in the local anesthetic up-down studies is a visual analog scale (VAS) score of <10 (on a 0–100 scale). Choosing any other VAS endpoint (or even qualifying the VAS to include duration, for example a VAS of <10 lasting at least 30 min) will yield a different ED50 estimation. Nevertheless, the ED50 estimations from these studies provide insight into the relative differences in ED50 values between the two drugs, regardless of initial success criteria. Thus, there is no question that the concentration of ropivacaine that produces a VAS score <10 in 50% of laboring patients is 40% higher than the concentration of bupivacaine which does the same. However, contrary to Drs. Polley and Columb’s assertion, information about the slope or shape of the respective dose response curves cannot be made with any statistical certainty from an up-down study. A minimum of three points along the slope is needed to construct a dose response curve, and then only if the points happen to fall near both ends and the middle (approximating the 5%, 50%, and 95% effective doses) of the slope. A major assumption when extrapolating up-down study design findings to clinical practice is that the slope and shape of the respective dose response curves are similar between drugs, which is not necessarily true. ED50 relative potency may not correlate with relative potency at the higher concentrations administered in clinical practice (physicians typically administer doses nearer the ED95 rather than the ED50). Two drugs with similar ED50 values can have markedly different dose response curves. Now let us specifically discuss ropivacaine and bupivacaine. Because traditional dose response studies comparing these two local anesthetics do not exist, can the ED50 estimations from up-down studies be extrapolated into clinical potency? If the respective dose-response curves are parallel, 40% more ropivacaine should be required than bupivacaine to produce labor analgesia, which is not the case. In separate patient-controlled epidural analgesia studies at our institution, patients self-administered similar hourly amounts of 0.125% ropivacaine and bupivacaine (6,7). It can be argued that this concentration lies near the upper end of the dose response curve, and differences in potency may have been masked. In addition, and as Drs. Polley and Columb note, the findings may also reflect the minimal volume required to spread local anesthetic within the epidural space. To circumvent these problems, we repeated the study using 0.075% concentrations and yet once again found no differences in hourly drug requirements between the two local anesthetics (8). Regardless of concentration ropivacaine and bupivacaine are clinically equipotent using the patient-controlled epidural analgesia technique. Of what benefit are the up-down ED50 estimations for ropivacaine and bupivacaine? Because differences in relative potency exist at the up-down ED50 concentration, the issue of potency must now be at least considered when designing future studies comparing the two local anesthetics, and previous study findings should be critically reexamined because similar concentrations of each local anesthetic were almost universally administered. In summary, the ED50 estimations from up-down studies provide an indication of relative potency of two drugs at one particular concentration but do not necessarily correlate with clinical potency and certainly do not “define and quantify the pharmacodynamics of the agents we use for labor analgesia” as Drs. Polley and Columb suggest. The up-down sequential allocation study design cannot substitute for traditional dose response studies that do define and quantify the pharmacodynamics and, along with clinical studies, actually guide clinical practice. And clinical practice is the bottom line: how should physicians administer a drug to their patients? From the existing literature, we would argue that similar concentrations of ropivacaine and bupivacaine should be administered to provide labor analgesia because they are clinically indistinguishable, regardless of the ED50 estimations. Robert D’Angelo MD Medge D. Owen MD