Purpose: Tumor proliferation has prognostic value in resected early stage non-small cell lung cancer (NSCLC) and can, therefore, predict which NSCLCs are at high risk for recurrence after resection and would benefit from additional therapy. It may also predict which tumor will respond to cell cycle-targeted chemotherapy and help assess the tumor response, besides helping to differentiate benign from malignant lung lesions. We evaluated whether the uptake of the new positron emission tomography (PET) tracer 3'deoxy-3'-[18F]fluorothymidine (FLT) in a series of suspected NSCLCs correlated with tumor proliferation assessed by Ki-67 immunohistochemistry and flow cytometry.
Experimental design: Ten patients with 11 biopsy-proven or clinically suspected NSCLC underwent 2-h dynamic PET imaging after i.v. injection of 0.07 mCi/kg FLT. Tumor FLT uptake was quantitated with the maximum pixel standardized uptake value (maxSUV), the partial volume corrected maxSUV (PV-corr-maxSUV), the average SUV over a small region-of-interest (aveSUV) and with Patlak analysis of FLT flux (aveFLTflux). The lesion diameter from computed tomography was used to correct the maxSUV for PV effects using recovery coefficients determined for the General Electric Advance PET scanner. Two of the 11 lesions were benign inflammatory lesions and 9 were NSCLCs. Immunohistochemistry for Ki-67 (proliferation index marker) was performed on all 11 tissue specimens (10 resections, 1 NSCLC percutaneous biopsy), and the S-phase fraction (SPF) from flow cytometry could be determined for 10. The specimens were reviewed for histology and cellular differentiation (poor, moderate, well). Lesions ranged from 1.6 to 7.7 cm.
Results: Excellent correlations were found between SUV measures of FLT uptake and Ki-67 scores [percentage of positive cells; maxSUV versus Ki-67: Rho = 0.78, P = 0.0043 (n = 11); PV-corr-maxSUV versus Ki-67: Rho = 0.83, P = 0.0028 (n = 10); aveSUV versus Ki-67: Rho = 0.84, P = 0.0011 (n = 11)]. Correlation between Ki-67 proliferation scores and Patlak measures of FLT uptake were also strong: aveFLTflux versus Ki-67: Rho = 0.94, P < 0.0001 (n = 11). The correlation between the SPF and all indices of FLT uptake was weaker and reached statistical significance for only two uptake indices [maxSUV versus SPF: Rho = 0.69, P = 0.03 (n = 10); PV-corr-maxSUV versus SPF: Rho = 0.36, P = 0.35 (n = 9); aveSUV versus SPF: Rho = 0.67, P = 0.03 (n = 10); aveFLTflux versus SPF: Rho = 0.46, P = 0.18 (n = 10)].
Conclusion: FLT PET may be used to noninvasively assess proliferation rates of lung masses in vivo. Therefore, FLT PET may play a significant role in the evaluation of indeterminate pulmonary lesions, in the prognostic assessment of resectable NSCLC, and possibly in the evaluation of NSCLC response to chemotherapy.