Is quantitative analysis of planar bone scan useful in assessment of response in bone metastases?

Background: One of most common and economical method for identifying 
bone metastases in clinical workflow is a bone scan. However, its clinical 
significance for determining how well patients with advanced metastatic bone 
disease are responding to therapy is still restricted, partly due to the absence of a 
reliable approach for measuring changes in bone scans.
Objectives: using quantitative analysis of planar images as compared to 
qualitative assessment in cancer patients receiving treatment for bone 
metastasis, radiotracer uptake was observed between first and follow-up scans. 
This method of imaging interpretation can be used to produce a more reliable 
and consistent manner of patient follow-up analysis, which is essential for 
researching.
Methods: This prospective study involved 37 patients who were known to have 
primary malignant tumor and osseous metastases. They underwent two bone 
scintigraphies before and after the therapy, each involved certain osseous lesion 
for further analysis by planar images. 47 osseous lesions were interpretated 
firstly by visual assessment by 3 physicians and interpretated by (progressive, 
regressive, stationary) opinion. Secondly, the same osseous lesions analyzed by 
quantifications of planar images were added by measuring mean and maximum 
of standardized uptake value of osseous lesions (SUV mean and max). All of 
quantification is considered progressive (more than 30% rise), regressive (more 
than 30% reduction) and stationary (rest in between). The results of agreement 
between 3 readers visually and between visual and quantitative assessment was 
done using Cohen’s kappa test. 
Results: Inter-observer agreement of the visual analysis between readers 1 and 
2 was in 32 lesions of total 47 lesions, k = 0.519 (moderate degree) for planar
scintigraphy. By the same way the agreement between readers 1 and 3 was in 36
lesions, k= 0.662 (substantial degree). In addition, inter-method agreement 
between reader 2 and reader 3 was in 27 lesions k = 0.358 (fair degree). On the 
other side, 19 (40.4%) of the total 47 lesions had visual and quantitative 
assessments that were in agreement. The quantitative and visual analyses' intramethod agreement was k= 0.049. (Poor degree).
Conclusions: For up to 57% of the total lesions, visual assessment of bone 
scintigraphy for changes in tumor metabolism produced conflicting results. In 
addition, visual analysis revealed moderate to significant inter-observer 
agreement. This suggests that, to maximize consistency in treatment planning, 
continuous monitoring of bone scanning for variations in the tracer uptake of 
lesions should be carried out using quantification of tracer uptake rather than 
just by visual evaluation.