178 Part II Chapter 7 our cohort (up to 57%): in our opinion an inter-individual variability in Na18F uptake that is far too large for a single value to be used [6]. To summarize, we acknowledge the suggested method to normalize our SUV measurements by introducing skeletal volume measurements. However, in our cohort, this did not lead to independence of our SUV measurements to its corresponding volume of distribution. Also, despite normalizing for SV, the measurements of both normal bone based on the cutoff used to discriminate normal bone fromdiseased bone still yielded a rangemuch larger than 10% from themedian. This finding was similar to a paper on SUV-SV by the authors of the letter [6]. This confirms our finding that an individual cutoff to discern HB from FD should be used. Apart from the observation that the method described by Duarte and Sapienza did deteriorate Na18F uptake normalization in our cohort, their method does require specific measuring tools that are not available in many clinically used software packages, restricting the availability of the technique. Lastly, it merely introduces the possibility of error by the introduction of an extra parameter (SV) measured on CT. In conclusion this means that in measuring FD burden, the volumetric parameter fluoride tumor volume (FTV) is superior to SUV or its derivative intensity measures as stated in our paper [2]. Patient- and scan-specific thresholds to segment FD (‘internal normalization’) are preferably used, based on a large intra-individual variation and confirmed in current additional analyses. This volumetric measure of FD burden is unrelated to the distribution volume of Na18F. Our published method to measure FD burden on Na18F-PET/CT and its usefulness for clinical evaluation therefore remains our method of choice, irrespective of the normalization procedure used. Conflict of interest The authors declare that they have no conflict of interest.