Comparison of MRI and PET-CT in detecting the loco-regional recurrence of soft tissue sarcomas during surveillance

To investigate the diagnostic performance of MRI and PET-CT for the detection of loco-regional recurrences after soft tissue sarcoma (STS) excision. From Dec 2003 to Aug 2014, 394 patients with STSs, who were included in the electronic patient registry for initial or repeated surgery at our hospital, were retrospectively reviewed. We identified 152 patients who underwent regular postoperative follow-ups with both MRI and PET-CT, obtained within a 3 month period of each other. We analyzed differences in the performance of MRI and PET-CT for the diagnosis of loco-regional recurrences using McNemar's test. The receiver-operating characteristic curves and calculations of the area under the curve were used. Twenty patients were found to have a loco-regional recurrence after tumor excision. For MRI and PET-CT, the sensitivities were 90.0 and 95.0 %, and the specificities 97.7 and 95.5 %, respectively, with positive predictive values of 85.7 and 76.0 % and negative predictive values of 98.5 and 99.2 %, respectively. No significant difference was detected between the sensitivities of MRI and PET-CT (p = 0.125). The area under the receiver-operating characteristic curve for PET-CT (0.952) was not significantly greater than that for MRI (0.939; p = 0.6). MRI of the area of interest is recommended for evaluation of tumor recurrence after surgical excision of STS. PET-CT was shown to be effective for detection of STS recurrence, and comparable to MRI. However, if PET-CT or MRI findings are inconclusive, the other modality may be helpful in differentiating tumor recurrence from post-therapeutic tissue change.