The recent publication by Felip et al1 describing the dose optimization of subcutaneous atezolizumab concerns us due to the continued missed opportunity to lower an unnecessarily high atezolizumab dose.2 Felip et al cite the 2016 article by Genentech authors that identified a target clinical serum concentration of 6 μg/mL (4 μg/mL without bevacizumab).3 Yet at the currently approved intravenous dose of 1200 mg every 3 weeks (equivalent to 15 mg/kg every 3 weeks), the steady-state trough concentration is approximately 140 μg/mL, which is >20-fold higher than the stated target trough of 6 μg/mL.4 Furthermore, doses as low as 1 mg/kg have been demonstrated to have clinical activity.5, 6

Calls for dose reoptimization are not limited to individuals or small groups. The Food and Drug Administration recently emphasized the importance of dose optimization for oncology drugs,7 which is relevant to atezolizumab, given the known relationship of exposure to adverse events of significant interest (AESIs) and the lack of relationship of exposure (in the range of concentrations achieved at the labeled doses) to efficacy.4 In this context, we do not believe that the dose of atezolizumab has been optimized and suggest that less frequent dosing could not only reduce costs and improve convenience but also reduce the incidence and/or severity of AESIs.8-10

An additional concern with the subcutaneous approach is high interpatient variability (124%) in bioavailability, partly due to injection site variability (thigh vs abdomen). Since bioavailability was estimated as low as 30%, simulations revealed that a 56% dose increase (1875 mg) was required to achieve the same exposure range as intravenous administration. Currently, Genentech manufactures only vial sizes of 840 and 1200 mg. To implement the subcutaneous dose of 1875 would require additional manufacturing burdens. A more convenient approach, which would also reduce administration and manufacturing costs, would be to simply administer intravenous atezolizumab at the currently approved dose amounts (840, 1200, or 1680 mg) less often, as we previously suggested.2

Atezolizumab is already available at 1680 mg every 4 weeks, but publicly available data suggest longer intervals (every 8 weeks or more) will likely maintain efficacy while lowering the risk of occurrence of AESIs.3, 4, 11 An interval of every 8 weeks is conservative at best, given the ≥20-fold excess in steady-state serum concentrations with standard dosing relative to the stated target. In a similar approach our group took for reoptimizing nivolumab and pembrolizumab, dose regimen simulation12 is the initial step to identify potential modified regimens, which can then be evaluated prospectively in a clinical trial,13 aiming to achieve the therapeutic concentration identified by Genentech.

Daniel Goldstein has institutional research funding from Merck, BMS, and Janssen and has received consulting fees from Vivio Health and owns stock in Vivio Health and TailorMed. Mark Ratain has served as a patent litigation consultant and expert witness on behalf of multiple generic pharmaceutical companies; has received consulting fees from Ascentage, Aptevo, Genentech, Pneuma Respiratory, and Shionogi; clinical trial funding from AbbVie, Dicerna, and Genentech; and other support from BeiGene. All other authors have declared no competing interests for this work.

This project has been funded in whole or in part with federal funds from the National Cancer Institute, National Institutes of Health Grant ZIC SC 006537, and from the National Institute of General Medical Sciences, Grant T32 GM007019.