164 Part II Chapter 6 Correlation between imaging and biochemical parameters The correlation of imaging parameters for FD-disease activity (SBS, TLF, SUVpeak) to biochemical disease activity were determined (Table 2): TLF correlated to ALP, P1NP and FGF-23. The SBS did not correlate with ALP or P1NP, but was correlated to FGF-23. SUVpeak was not correlated to any of the serumbiomarkers of bone metabolism. None of the imaging parameters correlated to CTX. Table 2. Squared Spearman’s rho correlation coefficients (R2) between either classical skeletal burden score on planar bone scintigraphy (SBS) or parameters on Na[18F]F-PET/CT versus 4 serum biomarkers for FD measured around the time of imaging. The asterisk (*) denotes a significant correlation with a p <0.05, and a double asterisk (**) after adjustment for multiple comparisons (pcritical = 0.013). ALP (n=20) P1NP (n=20) FGF-23 (n=14) CTX (n=20) SBS R2=0.351 p= 0.065 R2=0.337 p=0.073 R2=0.596 ** p= 0.007 R2=0.158 p= 0.253 TLF R2=0.571 ** p= 0.004 R2=0.621 ** p= 0.002 R2=0.541 * p= 0.015 R2=0.321 p= 0.084 SUVpeak R2=0.381 * p= 0.049 R2=0.111 p= 0.320 R2=0.385 p= 0.070 R2=0.015 p= 0.475 TLF was higher in patients using bisphosphonates, when compared to those without. This is reflected in Figure D with two dot plots (log10-scale) concerning SBS vs bisphosphonate use (panel A) and TLF vs bisphosphonate use (panel B). As one of the groups consisted of only two patients, we refrained from statistical analysis (Figure D). Correlation of SBS and Na[18F]F-PET/CT with clinical and biochemical parameters Average BPI-scores correlated to increased values of FGF-23 (R2=0.535, p=0.045 with pcritical=0.006; Spearman, n=15), suggesting a relation between disease burden/activity and pain, although SBS (R2=0.043, p=0.439 with p critical=0.006; Spearman, n=15) and TLF (R2=0.345, p=0.104 with p critical=0.006; Spearman, n=15) did not correlate with average BPI.