Postoperative residual tumour imaged by contrast-enhanced computed tomography and 201Tl single photon emission tomography: can they predict progression-free survival in high-grade gliomas?

Abstract

To evaluate if postoperative residual tumour imaged by either computed tomography or 201Tl single photon emission tomography (SPECT) carried out postoperatively could predict progression-free survival (PFS) in high-grade malignant gliomas. Thirty-three patients with high-grade malignant gliomas underwent both contrast-enhanced CT scan and 201Tl-SPECT postoperatively before receiving radiotherapy. The PFS was evaluated against the individual reports of the above two imaging studies by univariate analysis. CT and 201Tl-SPECT were carried out within a median interval of 17 days after surgery. Of the 33 patients, CT and 201Tl-SPECT were reported as positive for residual tumours in 23 (69.7%) and 30 (91%) patients, respectively. Sensitivity, specificity and overall accuracy were 71.4%, 40% and 66.6% for CT, and 96.4%, 40% and 87.8% for 201Tl-SPECT, respectively, and were based on their last follow-up status (P = 0.627 for CT; P = 0.053 for 201Tl-SPECT). The median PFS for patients reported to be positive or negative on CT scan was 4 and 5 months, respectively (P = 0.202). With 201Tl-SPECT, although the median PFS for patients with a positive 201Tl uptake was also 4 months, it had not even reached for those reported having a negative 201Tl uptake (cumulative survival 66.7% at last follow-up) (P = 0.198). However, Karnofsky performance status (KPS) was the only significant predictor on univariate analysis (KPS: < 80 vs. > or = 80; P < 0.001) for PFS. Although both the imaging modalities have a poor specificity, postoperative 201Tl-SPECT had a significantly better accuracy to predict the status at last follow-up than contrast-enhanced CT. Nevertheless, KPS remained the most significant outcome predictor for PFS in high-grade malignant gliomas.