Reply: Quantification of ¹⁸F-DCFPyL Uptake: TBR Versus Patlak’s Analysis

REPLY: We thank Drs. Laffon, de Clermont, and Marthan for their positive letter on our paper (1,2). They effectively argue that the tumor-to-blood ratio (TBR) can be seen as a simplification of a Patlak’s analysis—which in turn should be seen as a simplification of full pharmacokinetic analysis (based on nonlinear regression analysis with arterial blood sampling as input function). As suggested by Laffon et al., we have provided the correlation between Patlak’s K_i (net-influx-rate constant) and our proposed image-based TBR (Fig. 1) (2). Indeed, a good correlation was observed (R² = 0.91), supporting the suggestion of Laffon et al. that TBR could be considered as a surrogate for Patlak’s analysis that is suitable for daily clinical practice.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 1.

Correlation of image-based tumor-to-blood ratios and K_i derived from Patlak’s graphical analysis.

Overall, our study and the letter by Laffon et al. (1,2) emphasize the need for technical validation of simplified metrics to quantify tracer uptake of novel tracers. In particular, our study demonstrated that SUV, normalized to injected activity over body weight, is not a robust and valid simplified method for ¹⁸F‐DCFPyL uptake quantification—despite its widespread (often unvalidated) use. The main reason for the invalidity of SUV is that the input function, that is, the bioavailability of the tracer in plasma to tissue, is not comparable between subjects. Normalizing tracer uptake by injected activity over body weight (or lean body mass) assumes that the input function of individual patients is simply a scaled version of a population curve. When this assumption is violated, quantitative kinetic approaches that include an individually measured input function, such as Patlak analysis, are required. A simplification to the Patlak approach could be to normalize the tumor uptake to blood activity concentrations, as was shown in our paper. The use of TBR, at least partly, compensates for changes in the input functions that are not explained by variation in injected activity and weight alone. In our specific case, the overall mass of the disease affected the shape and amplitude of the input function, and thus normalizing tumor uptake by injected activity over weight, that is, SUV, should not be used for the quantification of ¹⁸F‐DCFPyL uptake.

Lastly, we agree with Laffon et al. that not only the repeatability of tumor uptake (SUV) should be evaluated, but also the repeatability of the TBR (and the blood–activity concentration itself) must be understood. The same differences in tracer bioavailability that were observed between patients can develop within patients over time, in the case of disease progression or treatment response. Results on our repeatability study are expected shortly.

• © 2019 by the Society of Nuclear Medicine and Molecular Imaging.