Abstract
609
Aim: To compare F-18 FDG PET/CT to Tc-99m MDP bone scan in assessment of metastatic osseous burden in breast cancer patients with bone metastases and to test the relation to tumor marker CA15-3 and a bone turnover marker alkaline-phosphatase (ALP).
Methods: We reviewed 37 patients (mean age of 55.38±13.08 y; 36 females) with known metastatic breast cancer to bone. Semi-quantitative scores was developed to reflect the extent and intensity of metastatic osseous burden. Skeleton divided into 7 zones: 1-Skull and mandible; 2-Cervicothoracic spine; 3-Lumbosacral spine; 4-Sternum, ribs, scapulae and clavicles; 5-Pelvis; 6-Upper arms; 7-thighs. Zones were scored using a 4 point scale: 0=no uptake; 1=single lesion; 2=multiple lesions <50% of the zone; 3=multiple lesions >50% of the zone. To assess intensity of uptake a 4 point score was used for the most intense lesion in each zone (no uptake=0, mild=1, moderate=2, severe=3). Overall metastatic burden was calculated as the sum of products of each zone score and its intensity score both for PET/CT (PM score) and for bone scans (BM score). We used student T-test to compare means, ROC analysis to define serum CA15-3 best cutoff to predict more lesions on PET/CT, McNemar test to assess lesional matching as paired data, Chi square test to assess difference in frequencies and Pearson correlation for correlations between scores, CA15-3 and ALP.
Results: Mean disease duration since diagnosis was 4.89±4.16 years (range 5 days to 17.9 years). Prior surgery was reported in 25 patients (67.6%), chemotherapy in 33 patients (89.2%), radiotherapy in 29 patients (78.4%) and hormonal therapy in 26 patients (70.3%). Mean CA15-3 level was 118.38±207.23 U/ml (range 8-1016). Time interval between PET/CT and bone scan was 23.3±21.9 days. Comparing PET/CT and bone scans, metastatic osseous burden was almost matched in 20 patients (54.1%), more on PET/CT in 13 patients (35.1%), of these patients 6 (16.2%) had no obvious lesions on bone scan (normal bone scan) while more osseous burden on bone scan was detected in 4 patients (10.8%) (p=0.049). Most of the additional lesions detected on PET/CT were osteolytic and predominantly medullary, suggesting early-developing new lesions, while lesions with higher uptake on bone scans showed more sclerotic changes with low FDG activity on PET/CT, suggesting non-active healed/healing lesions. Eight patients (21.6%) had better visualization of skull lesions on bone scan compared to PET/CT while 1 patient had bilateral tibial metastases that can not be visualized on PET/CT due to imaging field limitations. The CA15-3 was positively correlated to PM score (r=0.386; p=0.018) but not to BM score (r=-0.027; p=0.874). On the other hand serum ALP was positively correlated to both PM score (r=0.389; p=0.017) and BM score (r=0.363; p=0.027). A cutoff value for serum CA15-3 of >47 U/ml was found on ROC analysis (AUC=0.708; p=0.0261) to best predict additional findings on PET/CT imaging compared to bone scan with sensitivity, specificity, PPV, NPV and accuracy of 61.5%, 79.2%, 61.5%, 79.2% and 73% respectively (p=0.013). Conclusion: FDG-PET/CT is more sensitive and specific in detecting metastatic osseous lesions and tends to better identify new osteomedullary lesions. Bone scan is less specific due to osteoblastic activity in healing sclerotic lesions that could persist for several months, rendering it suboptimal for short-term follow-up. The metastatic osseous burden of PET/CT is correlated to CA15-3 serum level and a higher level of CA15-3 could predict more lesions on PET/CT but not on bone scan. Nevertheless, bone scans can still detect lesions in areas not covered by the PET/CT imaging field; skull lesions that are masked by the high brain metabolic activity on PET/CT or metastatic lesions from histopathological variants with inherently low FDG avidity.
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.