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

TO THE EDITOR: Verwer et al. (1) recently presented a study aimed at validating the use of simplified methods for quantification of ¹⁸F-fluoromethylcholine uptake in a routine clinical setting of prostate cancer patients. The authors nicely demonstrated that ¹⁸F-fluoromethylcholine uptake should be quantified using full kinetic modeling involving a single-tissue-compartment model with irreversible trapping and a blood volume parameter, in combination with a metabolite-corrected plasma input function based on invasive arterial blood sampling. The authors proposed—as a noninvasive simplified method based on 2 consecutive static PET scans—the use of the ratio (SUV_AUC,PP) of lesion activity concentrations (A_L(t), assessed 30–40 min after injection) normalized to the area under the curve of the metabolite-corrected plasma input function (AUC_PP, computed over 0–30 min after injection). This ratio provided an excellent correlation to the uptake rate constant of the full kinetic modeling (Fig. 6C; SUV_AUC,PP = 14.54 × K₁ + 0.02; R² = 0.91) (1).

We would like to point out that the slope of the fit reveals a discrepancy of 14.54 between SUV_AUC,PP and K₁, whereas SUV_AUC,PP should be considered as a noninvasive surrogate for K₁ and a slope around 1 should be expected. Indeed, as previously shown by Patlak (2), K₁ = A_L(t)/AUC_PP , which is actually the SUV_AUC,PP definition. Therefore, corrections to the SUV_AUC,PP outcomes reported by Verwer et al. may be proposed for a better comparison with K₁. For this comparison, an analytic expression for AUC_PP and hence for SUV_AUC,PP , as simple as possible, is needed to clarify the unit of each parameter. Let us assume that the metabolite-corrected plasma input function monoexponentially decays with a (decay-corrected) time constant α: then AUC_PP = A₀/α × [1 – exp(–αT)], with T = 30 min and A₀ the initial (virtual) metabolite-corrected plasma activity concentration (3). AUC_PP is the total number of disintegrations per milliliter (of blood) that have occurred over the time range 0–T; A₀ is expressed in Bq/mL, that is, number of disintegrations per second and per milliliter; [1 – exp(–αT)] has no dimension; α is expressed in s⁻¹ because A₀ involves becquerels (i.e., equivalent to s⁻¹). Finally, SUV_AUC,PP is expressed in s⁻¹ because of the A_L(t) unit, which is Bq/mL. To consistently compare SUV_AUC,PP and K₁, we suggest that 2 corrective factors should be applied. First, because in current practice A_L(t) is usually expressed in kBq/mL rather than in Bq/mL, A₀ should then be expressed in kBq/mL instead of in MBq/mL, as indicated in Figure 6C (and in Supplemental Fig. 2C) (1): the corrective factor is 1/1,000. Second, because K₁ is usually expressed in min⁻¹ rather than in s⁻¹ (the axis units in Fig. 6C and supplemental Fig. 2C are not clearly indicated), the corrective factor is 60. As a result, we suggest that the SUV_AUC,PP outcomes reported by Verwer et al. should be multiplied by a corrective factor of 60/1,000, leading to a further slope of 0.87 instead of 14.54 in Figure 6C.

In conclusion, Verwer et al. convincingly demonstrated that, instead of SUV, SUV_AUC,PP could be used in current clinical practice to noninvasively quantify ¹⁸F-fluoromethylcholine uptake in prostate cancer patients. We further suggest that SUV_AUC,PP is actually an uptake rate constant rather than an SUV (usually expressed in min⁻¹ and g/mL, respectively) and that the above-proposed correction strengthens its relevance.

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