Table 2.

Changes in parameter values for PD and non-PD after 8 weeks of treatment on [¹⁸F]fluoride PET/CT with associated optimum and 25 % cut-off values and AUC

	PD (20 lesions/4 patients)							Non-PD (52 lesions, 12 patients)					P value for PD vs non-PD
	Median change (%)	Optimal cut-off (ROC curve analysis)		Number of lesions identified (↑ or ↓)		Number of patients correctly identified (↑ or ↓)		Median change (%)	Number of lesions identified (↑ or ↓)		Number of patients identified (↑ or ↓)
		Value	AUC	Opt cut-off	25 % cut-off	Opt cut-off	25 % cut-off		Opt cut-off	25 % cut-off	Opt cut-off	25 % cut-off
SUVmax (%)	− 3.0	− 15.1	0.67	15 ↑	7 ↑	3 ↑	2 ↑	− 20.1	29 ↓	47 ↓	5 ↓	9 ↓	0.021
SUVmean (%)	15.1	− 2.6	0.75	14 ↑	6 ↑	3 ↑	2 ↑	− 16.0	38 ↓	46 ↓	8 ↓	9 ↓	0.001
SUVpeak (%)	3.0	− 3.9	0.72	13↑	8 ↑	3 ↑	2 ↑	− 17.6	31 ↓	46 ↓	7 ↓	9 ↓	0.003
MTV (%)	− 2.5	19.0	0.51	14 ↓		0		1.1	44 ↓		10 ↓		0.806
TLM (%)	9.2	− 19.0	0.67	18 ↑		4 ↑		− 11.2	26 ↑		8 ↑		0.023
SD (%)	4.9	9.2	0.67	10 ↑		2 ↑		− 25.4	43↓		8 ↓		0.022
Ent (%)	1.3	0.6	0.67	16 ↑		3 ↑		− 0.5	37 ↓		5 ↓		0.020
Unif (%)	− 6.6	1.8	0.69	16 ↓		3 ↓		2.9	31 ↑		7 ↑		0.012
Kurt (abs)	− 0.1	− 0.005	0.66	14 ↓		4 ↓		0.04	31 ↑		8 ↑		0.034
Skew (abs)	− 0.05	0.1	0.5	18 ↓		3 ↓		0.01	37 ↓		11 ↓		0.277

MTV metabolic tumour volume, TLM tumour lesion metabolism, SD standard deviation, Ent entropy, Unif uniformity, Kurt kurtosis, Skew skewness, abs absolute change, % percentage change, ↑ ≥ the cut-off value, ↓ < the cut-off value