LAUSR

We read with interest the work by Spyker et al., that simulated overdose of different acetaminophen formulations [1]. We compare their model of 100 g MR preparations in overdose, to actual acetaminophen concentrations of MR overdose patients not included in their study [2]. ATOM-3 reported acetaminophen concentrations of 116 MR overdose patients [2], and we show in the figure acetaminophen concentrations of single MR only ingestions of >35 g overlaid on the prediction limits of the 100 g MR simulated data from their study (Figure 1). The ATOM-3 data demonstrate the erratic pharmacokinetics of MR acetaminophen in overdose, and patients in the range of 35–50 g mostly fell within the upper part of the predicted range for 100 g overdoses. Patients with overdoses larger than 50 g were all above the modeled median and 7 of the 24 patients in the range of 35–100 g were outside their upper 95% prediction interval. ATOM data show frequent persistently high acetaminophen concentrations, double peaks, and ongoing absorption for over 24 h. Their model clearly is a poor predictor of concentrations with overdoses of the MR formulation. Why did their model perform poorly? They incorrectly stated that the MR products have a ratio of “69% IR formulation and 31% ER formulation.” In fact, the opposite is true. These bilayer 665mg tablets contain a layer of 31% IR acetaminophen and a second layer of 69% ER acetaminophen [2–5]. Their simulation results reflect the incorrect proportions. They predicted a shorter Tmax for the MR product than for the ER product (650mg: 50% ER and 50% IR) at all simulated doses (Table 1). This result is inconsistent to what has been found in simulated human volunteer studies of supratherapeutic (subtoxic) doses of ER and MR acetaminophen products (Table 1) [3,4,6,7]. The two human volunteer studies of the ER product showed differing results despite similar methods [6,7]. Douglas et al., found a lower area under the curve (AUC) and peak concentration (Cmax) when compared with IR acetaminophen [7]. Stork et al., found similar AUC values for both preparations [6]. Both studies reported the Tmax was not delayed with the ER preparation (Table 1). This contrasts with two human volunteer studies of the MR preparation. Both found a reduced AUC and Cmax and a significantly later Tmax 2.8 h with the MR preparation compared to 1.3 h with the IR acetaminophen [3,4]. The Tmax values simulated by Spyker et al. for 10 g MR ingestions are lower than these observed Tmax despite the lower