Response Assessment Using Early and End-of-Treatment 18F-FLT PET Can Predict Outcome in Patients with Non-Hodgkin Lymphoma,

Abstract 3679

Background Fluorine-18-fluorodeoxyglucose (FDG) positron emission tomography (PET) scan has been known as a useful modality for response assessment in malignant lymphoma. However, FDG is not tumor-specific and can be false positive in inflammatory lesions. To overcome these limitations, a new PET tracer, thymidine analog 3'-deoxy-3'-¹⁸F-fluorothymidine (FLT), was introduced recently. Preliminary data showed close correlation between FLT uptake and tumor cell proliferation in lymphoma, suggesting the possibility of noninvasive tumor grading and early response assessment. Therefore, we performed a prospective trial to evaluate the feasibility of FLT-PET in risk stratification and prediction for treatment outcome, especially in early interim analysis, in patients with non-Hodgkin lymphoma (NHL).

Methods Seventy-five patients newly diagnosed with NHL were prospectively enrolled at National Cancer Center, Korea, from Oct 2005 to Oct 2008. All received standard chemotherapy for their pathologic classifications. Patients were evaluated with FLT-PET at baseline (FLT0), after 1 cycle of chemotherapy (FLT1, early), and after completion of the 1^st line chemotherapy (FLTE, end-of-treatment). FLT-PET results were assessed according to the International Workshop Criteria (IWC). Maximum standardized uptake values (SUVmax) of each FLT-PET were calculated to evaluate its correlation with the clinical characteristics and treatment outcome. Treatment outcome was estimated using 3-year progression-free survival (3yr-PFS) and overall survival (3yr-OS).

Results Of the 75 enrolled patients, 63 (84%) had diffuse large B-cell lymphoma. Median age at diagnosis was 57 years (range, 29–87). Twenty-eight (37.3%) presented with stage III or IV diseases and 20 (26.7%) showed more than 3 IPI scores. Median follow up duration was 4.5 years (range, 3.5–5.8). Five (6.7%) patients underwent hematopoietic stem cell transplantation at last. Three-year PFS and OS rates for all enrolled patients were 68% and 78.7%. Seventy-three (97.3%) had their FLT-PET at baseline, 69 (92%) after 1 cycle of chemotherapy, and 66 (88%) at the end of the 1^st line treatment. By IWC, 50 (66.7%) patients achieved complete remission (CR) on FLT1 and 56 (74.7%) had CR on FLTE. Positive predictive values (PPV) of residual uptake on FLT1 and FLTE for relapse or disease progression were 83.3% (95%CI 57.7–95.6) and 80% (95%CI 44.2–96.5), respectively. Negative predictive values (NPV) of them were 88% (95%CI 75.0–95.0) and 82.1% (95%CI 69.2–90.7). Sensitivity and specificity were 71.4% (47.7–87.8) and 93.6% (81.4–98.3) for FLT1 and 44.4% (22.4–67.8) and 95.8% (84.6–99.3) for FLTE, respectively. Complete disappearance of uptake on FLT1 was significantly associated with better PFS compared to residual uptake on FLT1 (3yr-PFS rates, 87.5% and 12.2%, p<0.001). Three-year OS rates according to CR achievement on FLT1 were 96.0% and 27.8%, significantly lower in patients with residual disease after 1 cycle of chemotherapy (p<0.001). SUVmax of FLT0 correlated with LDH level significantly (p=0.044), but not with age (p=0.214), Ki-67 index (p=0.073), IPI score (p=0.270), and Ann Arbor stage (p=0.089). SUVmax of FLT0 were not associated with survival outcomes, however, residual SUVmax of FLT1 reflecting early response to treatment was significantly associated with poor survival outcome (PFS, HR 1.29, 95%CI 1.14–1.47; OS, HR 1.27, 95%CI 1.08–1.49). In multivariate analysis, SUVmax of FLT1 remained as an independent predictive factor for PFS (HR 1.63, 95%CI 1.25–2.13) and for OS (HR 1.89, 95%CI 1.38–2.58). Residual SUVmax of FLTE also revealed to be significantly associated with PFS (HR 1.43, 95%CI 1.05–1.94) and OS (HR 1.59, 95%CI 1.12–2.26) in the same multivariate model.

Conclusion Response assessment in cooperation with FLT-PET provided accurate prediction for clinical outcome including PFS and OS in patients with NHL. Especially, early FLT-PET result after 1 cycle of chemotherapy was an independent predictive factor for survival as well as relapse or disease progression, with comparable performance with end-of-treatment FLT-PET.