DOI: 10.1007/s00259-016-3459-5Pages: 441-448

Predictive value of [18F]-fluoride PET for monitoring bone remodeling in patients with orthopedic conditions treated with a Taylor spatial frame

1. Karolinska University Hospital, Department of Hospital Physics, Nuclear medicine

2. Karolinska University Hospital

3. Nyköpings Hospital, Department of Radiology

4. Karolinska Institutet, Department of Molecular Medicine and Surgery

5. Uppsala University Hospital, Department of Radiology and Molecular Imaging

Correspondence to:
Alejandro Sanchez-Crespo




The Taylor Spatial Frame (TSF) is used to correct orthopedic conditions such as correction osteotomies in delayed fracture healing and pseudarthrosis. Long-term TSF-treatments are common and may lead to complications. Current conventional radiological methods are often unsatisfactory for therapy monitoring. Hence, an imaging technique capable of quantifying bone healing progression would be advantageous.


A cohort of 24 patients with different orthopedic conditions, pseudarthrosis (n = 10), deformities subjected to correction osteotomy (n = 9), and fracture (n = 5) underwent dynamic [18F]-fluoride (Na18F) PET/CT at 8 weeks and 4 months, respectively, after application of a TSF. Parametric images, corresponding to the net transport rate of [18F]-fluoride from plasma to bone, Ki were calculated. The ratio of the maximum Ki at PET scan 2 and 1 ( K¯i,max $$ {\overline{K}}_{i, \max } $$) as well as the ratio of the maximum Standard Uptake Value at PET scan 2 and 1 ( SUV¯max $$ {\overline{SUV}}_{\max } $$) were calculated for each individual. Different treatment end-points were scored, and the overall treatment outcome score was compared with the osteoblastic activity progression as scored with K¯i,max $$ {\overline{K}}_{i, \max } $$ or SUV¯max $$ {\overline{SUV}}_{\max } $$.


K¯i,max $$ {\overline{K}}_{i, \max } $$ and SUV¯max $$ {\overline{SUV}}_{\max } $$ were not correlated within each orthopedic group (p > 0.1 for all groups), nor for the pooled population (p = 0.12). The distribution of K¯i,max $$ {\overline{K}}_{i, \max } $$ was found significantly different among the different orthopedic groups (p = 0.0046) -also for SUV¯max $$ {\overline{SUV}}_{\max } $$ (p = 0.022). The positive and negative treatment predictive values for K¯i,max $$ {\overline{K}}_{i, \max } $$ were 66.7 % and 77.8 %, respectively. Corresponding values for SUV¯max $$ {\overline{SUV}}_{\max } $$ were 25 % and 33.3 %


The K¯i,max $$ {\overline{K}}_{i, \max } $$ obtained from dynamic [18F]-fluoride-PET imaging is a promising predictive factor to evaluate changes in bone healing in response to TSF treatment.

To access the full text, please Sign in

If you have institutional access, please click here

  • Accepted: Jul 5, 2016
  • Online: Jul 23, 2016

Article Tools