Correspondence.

To the Editor:

We read with interest the recent article entitled “Pharmacokinetic study of intravitreal aflibercept in humans with neovascular age-related macular degeneration” by Do et al.[1] With the exception of an earlier (uncited) abstract by two of the authors (D.V. Do and Q.D. Nguyen),[2] there have been virtually no experimental data publicly available on the pharmacokinetics of aflibercept in human eyes. After carefully reviewing Table 1, we believe that the reported median concentrations contain multiple errors, and the half-life value of 11 days derived from them is incorrect. The erroneous median values apparently resulted from inclusion of the time points along with the concentrations.

Using the five individual subjects' data reported in Table 1, we recalculated the median values for each time point and obtained the following: 0, 104, 65.9, 51.5, 42.0, 20.3, and 9.35 mg/L, shown in the semilogarithmic plot (Figure 1). Fitting these data to an exponential function (noncompartmental analysis, Certara Phoenix software v. 6.4) yielded a decay rate of 0.072 day−1 and a corresponding half-life of 9.6 days. Similarly, an noncompartmental analysis half-life of 9.5 days was obtained using the mean concentration data. Noncompartmental analysis of individual data showed high intersubject variability, with half-life estimates ranging from 3.7 days to 13.2 days (median and mean: 8.8 days). In aggregate, our analyses suggest that the correct ocular half-life of aflibercept in the human eye is circa 9 days, consistent with what Do and Nguyen[2] reported in their earlier published abstract.

Fig. 1.

Recalculated median concentrations versus time. Dashed line corresponds to an exponential function fit to the data, excluding the earliest time point. The decay rate of 0.072 days−1 corresponds to a half-life of 9.6 days.

In view of the importance of the ocular half-life as a factor in determining the duration of biological and clinical effects after intravitreal injection,[3] it is important to obtain and report accurate pharmacokinetic data on such ophthalmic products.

Although the experimental study by Do et al[1] seems to have been well designed and performed, the analysis and conclusions reported in the article are flawed. We hope the authors will also perform a reanalysis of their data and correct this.