Reply: Simplified Methods for Quantification of ¹⁸F-Fluoromethylcholine Uptake: Is SUV_AUC,PP Actually an SUV?

REPLY: We would like to thank Laffon and coworkers for their interest in our paper (1), as well as for their valuable comments and for giving us the opportunity to further clarify our results.

In our paper we aimed to find a simplified method for quantification of ¹⁸F-fluoromethylcholine uptake. The most common simplified metric to quantify tracer uptake is the standardized uptake value (SUV), which represents tracer concentration at a certain time normalized to injected activity and a factor representing distribution volume (e.g., weight, lean body mass, or body surface area). This normalization is used as a surrogate for the integrated availability of the radiotracer to the tumor.

One of the aims of full kinetic analysis is to identify and validate these simplified methods or normalization factors. In our study, we found that commonly used normalization factors such as injected activity per weight, lean body mass, or body surface area were not appropriate surrogates for the integrated availability of the radiotracer to the tumor, that is, the integral of the input function. Therefore, we proposed a dual-scan procedure in which the first scan is used to directly measure this integral, which is then used to normalize tracer uptake measured in the second scan. We would argue that the unit of the normalization factor does not qualify or disqualify the simplified measure from being termed an SUV, since commonly used normalization factors such as body weight or body surface area also yield SUV in different units.

Nevertheless, the authors are correct in their observation that we have not consistently used the same units in nominators and denominators of the relevant equations and we could have been more transparent on the units of the variables used in the paper. K₁ is indeed expressed in mL·cm⁻³·min⁻¹ (essentially, min⁻¹), tissue activity concentration is in kBq·mL⁻¹, and AUC is in MBq·s·mL⁻¹ (i.e., 10⁶ disintegrations/mL), which makes the unit of SUV_AUC 10⁻³ s⁻¹. We have chosen these units so that the order of magnitude of the resulting values would be comparable to the commonly used SUV parameter. In this way, there would be a more intuitive understanding of high and low tracer uptake by the presented SUV_AUC,PP data for all those readers who are used to SUV data. Of course, strictly speaking, the authors are correct with respect to using units consistently.

As the authors note, SUV_AUC,PP can also be treated as a surrogate direct estimation of the rate constant K₁. Indeed, for a single-tissue-compartment irreversible model (1T1k) with negligible blood volume fraction, K₁ = SUV_AUC,PP by definition. We agree with the authors that if SUV_AUC,PP is to be used in this way, it is important to use comparable units, which can indeed be achieved by applying the correction factor of 60/1,000 as indicated by the authors.

We would like to point out, however, that in the paper, K₁ was derived with blood volume correction (1T1k + V_B), whereas SUV_AUC,PP was not. Therefore, the definition will not hold and should be replaced with , where C_WB is the whole-blood activity concentration, C_P is the metabolite-corrected activity concentration in plasma, and V_B is the blood volume parameter. We believe V_B to be responsible for part of the underestimation observed for SUV_AUC compared with K₁. We would also like to point out that in the paper, SUV_AUC,PP (and K₁) represents a surrogate measure for K_i, that is, a combination of K₁, k₂, and k₃, and should not be viewed as a simple rate constant.

Despite the confusion regarding units, we hope we have shown that commonly used SUV metrics cannot be applied to quantify ¹⁸F-fluoromethylcholine uptake reliably and that normalization of tracer uptake by the integral of the input function (as a surrogate of K_i) is, in our opinion, the best simplified approach.

• © 2015 by the Society of Nuclear Medicine and Molecular Imaging, Inc.