Assessment of cellular proliferation in recurrent primary brain cancer with 3'-deoxy-3'-18F-fluorothymidine PET imaging.

366 Objectives 39-Deoxy-39-18F-fluorothymidine (FLT) uptake within the CNS is restricted by the blood-brain barrier, and occurs only in areas of vascular leakage. We have previously shown that the cellular proliferation rate can be separated from transport of FLT in brain cancer patients through compartmental modeling using dynamic FLT images, although transport effects invariably dominate the imaging signal [Muzi 2006, Spence 2009]. Thus, overall uptake represents a combination of transport and proliferation when reported as FLT SUV or SUVmax, and these variables have been shown to correlate to Ktrans, a measure of blood-brain-barrier permeability from gadolinium–enhanced MRI [Muzi 2014]. Our objective was to determine whether FLT PET measures of proliferation correlate with survival in patients with suspected primary brain cancer recurrence. Methods Our cohort includes 34 primary brain cancer patients who were imaged following radiation treatment and at additional times up to the time of recurrence. They underwent dynamic FLT scanning for 90-120 minutes after injection with concomitant blood sampling and plasma metabolite assessment. Compartmental modeling provided estimates of transport (K1) and proliferation (Ki, Flux), residue analysis provided image maps of similar parameters in an objective non-modeling approach and conventional SUV parameters were also included. We tested the predictive value of FLT imaging for this cohort using several parameters of FLT uptake, including FLT extraction (Flux/K1 = k3/(k2+k3)), with respect to survival from the time of FLT imaging (FS) and overall survival from the time of initial diagnosis (OS). In recurrent glioma, FS is more clinically relevant than OS. Results Our multivariate Cox model analysis with covariates (age, KPS, gender, MRI gadolinium-enhancement and extent of surgical resection) finds that the combination of FLT SUV (Figure 1) or FLT Flux (not shown) (p<.001), FLT extraction (p=.023) and gender (p=.048) are the key variables associated with the length of survival after the FLT PET study (FS). OS results were similar. In addition, FLT extraction (the retention quantity recovered by kinetic analysis) was also significant, and shows particular promise for differentiating true tumor recurrence from other causes of BBB breakdown, including radionecrosis and pseudoprogression (p=.022); no other factor showed potential for providing this differentiation in our cohort. CONCLUSIONS: For FLT PET imaging of primary brain cancer, FLT SUV is a parameter affected by tissue perfusion as well as cellular proliferation. Perfusion has previously shown significance for similar cohorts of glioma patients imaged at recurrence in MRI studies [Hoefnagels 2009, Hu 2012]. The estimation of FLT flux, although more purely reflective of cellular proliferation, produces low values that are sensitive to noise due to the tracer perfusion effects for regions only accessible through blood-brain-barrier breakdown. Bevacizumab treatment during imaging in some of these patients may have complicated interpretation of uptake due to normalization of the vasculature, thus restricting FLT access. Assessing tissue perfusion and cellular proliferation in brain cancer imaged at the time of recurrence using dynamic FLT PET shows value for survival prediction. [Supported by NIH/NCI P01-CA42045 (Krohn), U01-CA148131 (Kinahan); SFI-PI 11/1027 (O’Sullivan)] . Muzi M, Fink JR, Richards TL, et al. J Nucl Med 55(S1):1512-. 2014. Spence AM, Muzi M, Link JM, et al. Mol Imaging Biol 11(5):343-355. 2009. Muzi M, Spence AM, O9Sullivan F, et al. J Nucl Med 47(10):1612-1621. 2006. Hoefnagels FW, Lagerwaard FJ, Sanchez E, et al. J Neurol 256(6):878-887. 2009. Hu LS, Eschbacher JM, Heiserman JE, et al. Neuro Oncol 14(7):919-930. 2012. . .