Anesthetizing the Obese

“Thou seest I have more flesh than another man, and therefore more frailty.” —William Shakespeare, 1564–1616 Why after 25 years of clinical use is there suddenly so much interest in propofol? Paracelsus wisely recognized that “the dose makes the poison,” and it makes you wonder if recent media reportsa have stirred up niggling reminders that propofol dosing is more of an art than a science. Part of this artistic license has resulted from the milligram-per-kilogram dose,1 which attempts to normalize exposure across subjects of differing body size. However, since propofol's approval, the world's body composition has changed, with obesity prevalence in the United States' adult population (classified as a body mass index [BMI] ≥30 kg/m2) increasing from 13.4% in 19602 to 33.9% in 2008.3 What's more, we're getting bigger, prompting the World Health Organization to recategorize obesity into simple obesity (BMI 30–34.9 kg/m2), severe obesity (35–39.9 kg/m2), morbid obesity (40–49.9 kg/m2), and super morbid obesity (BMI ≥40 kg/m2).4 One must surely question whether dosing strategies developed in leaner populations remain applicable to today's demography. For propofol, and other drugs, a body of research has questioned whether drug doses based on total body weight (TBW) can be reliably transported to the obese. If so, the following assumptions would need to be true: that volume of distribution, which defines initial drug dose, is twice as large for a 200-kg subject compared with a 100-kg subject and that clearance (CL), which determines maintenance dosing, is twice as fast in the respective 200-kg subject.b The first assumption might not be unreasonable if the drug is lipid soluble and distributes into fat. However, for hydrophilic drugs, volume of distribution is more likely to be linearly related to lean body weight (LBW) rather than TBW. The second assumption, that CL is twice as fast in a 200-kg subject, seems implausible unless adipose tissue metabolizes the drug or the functional capacity of metabolizing organs, e.g., the liver and kidney, are linearly related to TBW. There is little evidence supporting this second assumption because 99% of metabolic processes occur within lean tissues.6 It is perhaps not too surprising then that CL seems best linearly related to LBW rather than TBW.7 The upshot in clinical practice is that subjects of different body compositions should expect a similar drug exposure if maintenance doses are determined on a milligram amount per kilogram of LBW. And here lies a challenge: LBW has been prohibitively difficult to use in the clinic, because accurate measurement via dual x-ray absorptiometry, for example, is not practical. Bioelectrical impedance analysis (BIA) is more readily available as demonstrated by Ingrande et al.,8 although more frequently, equations developed by James in 1976 are used to estimate LBW.9 These equations cannot, however, be extrapolated to the obese because of numerical inconsistencies whereby LBW declines at certain weights and heights.7,10 New semimechanistic LBW equations developed in 2005 are transportable to the obese, with readers directed to the article by Janmahasatian et al.11 for a complete derivation. For propofol, there is clearly much confusion when dosing the obese, with most prior work unable to help anesthetists. The reason is that the obese have often been excluded from studies, or James' LBW equations have been used to describe pharmacokinetic (PK) parameters. As such, CL has been reported to increase linearly with TBW12 or vary with some form of James' LBW13,14 (Fig. 1). The empirical suggestion by Servin et al.15 of “corrected weight” was a possible metric that might be applicable to dosing the obese, although some have questioned its suitability in clinical practice.16,17 Trial design, subject population, and James' LBW equations are responsible for the variety of relationships between TBW and CL shown in Figure 1, some of which12,13 appear in target-controlled infusion (TCI) devices to help dose propofol. To prevent overdosing using these devices, work-around solutions have been implemented by the manufacturers; e.g., TBW is capped at 150 kg in the Diprifusor system (AstraZeneca, Cheshire, UK) and LBW is capped at maximal values for the Base Primea system (Fresenius, Brezins, France).c More recently, for remifentanil TCI, which incorporates James' LBW, La Colla et al.18 developed a metric termed “fictitious height” to avoid underdosing in the obese! Does one not find it frightening that given all of today's technological advances we use fudge factors when dosing anesthetic drugs in at least one-third of the adult population?Figure 1: Propofol clearance predicted by a range of pharmacokinetic models for a 30-year-old man with a height of 1.75 m. *Unpublished data from S. C. McLeay, G. A. Morrish, C. M. Kirkpatrick, and B. Green.On a side note, it would be unusual for any clinician to dose patients without considering, for example, their age, concomitant medications, or hair color.19 Furthermore, because doses are often titrated to effect, some may argue that new dosing recommendations are unnecessary. Bouillon and Shafer10 recognized this perspective, stating that “We are clearly skilled at getting close to the right dose. Isn't ‘close' close enough?” They continue, however, to discuss the need for a sound scientific foundation to obtain the correct dose based on size. Indeed, it is this “close enough” mindset that might hinder further research on body size, PKs, and dose development for the obese. Given the mounting evidence that propofol's current dose strategies and PK models are inadequate for the obese, the Open TCI Initiative (http://opentci.org) has made data publicly available in the anticipation that universal dosing algorithms be developed. We propose that mechanistic PK models and dosing strategies use these data if necessary, but ensure models are transportable across a wide range of body compositions. More recent studies have attempted to do this,20,21 although some approaches cause confusion by failing to consider the limitations of design on covariate selection.22 It is therefore refreshing that Ingrande et al.8 are prospectively assessing BIA for dosing propofol in the obese and exploring a hypothesis based on mechanistic reasoning rather than empiricism and tradition. They also note that LBW estimated by the Janmahasatian et al. equation11 is a practical approach in the clinic where BIA is unavailable. In conclusion, we propose that it is time to move forward and update PK models in TCI devices with ones that are transportable to the obese. Let's stop making up fudge factors for TCI devices and perhaps consider Janmahasatian et al.'s LBW as a metric to dose propofol across a population that includes the obese. DISCLOSURES Name: Bruce Green, DClinPharm. Attestation: This author has prepared the manuscript and approved the final manuscript. Conflict of interest: None. Name: Sarah C. McLeay, BSc(Hons). Attestation: This author has prepared the manuscript and approved the final manuscript. Conflict of interest: SCM was supported by a grant from Pfizer Global R&D.