[18F]FLT Is Superior to [18F]FDG to Early Predict Response to Specific Inhibitors of NPM-ALK-Dependent Pathways In a Human ALCL Xenograft Model

The thymidine analogue [¹⁸F]fluorothymidine (FLT) has been shown to reflect proliferation of high-grade lymphoma cells both in preclinical and clinical studies. In this preclinical in vitro and in vivo study we assessed early FLT-uptake as an adequate and robust surrogate marker for response to inhibitors of Nucleophosmin-anaplastic lymphoma kinase (NPM-ALK)-dependent pathways in an anaplastic large cell lymphoma (ALCL) xenotransplant model.

In vitro investigations included viability assessment (MTT assay), cell cycle analysis using propidium iodide staining and western blotting to characterize response of the ALCL cell lines SUDHL-1 and Karpas299 to treatment with heat shock protein 90 (Hsp90) inhibitor NVP-AUY922, the Phosphoinositide 3-kinase (PI3K) inhibitor BGT226 or the mammalian target of rapamycin (mTOR) inhibitor RAD001. Thymidine metabolism in severe combined immunodeficient (SCID) mice bearing SUDHL-1 or Karpas 299 lymphoma xenotransplants was assessed non-invasively prior to and early in the course of therapy (48h to 7 days) by FLT and FDG positron emission tomography (FLT-PET and FDG-PET) using a dedicated small animal PET system. Tumor-to-background ratios (TBR) of FLT-PET were compared to that of PET using the standard radiotracer [¹⁸F]fluorodeoxyglucose (FDG). Reference for tumor response was local control of the tumor measured by shifting calliper and histopathological analysis of explanted lymphomas.

In vitro, SUDHL-1 cells were sensitive to all three inhibitors (IC50 AUY922= 50 nM; IC50 BGT266= 10 nM; IC50 RAD001= 1 nM). These cells showed a dose-dependent induction of cell-cycle arrest in G1-phase and reduction of S-Phase after 24 to 48 hours and - to a lesser extent - increase of apoptosis. Incubation of SUDHL-1 cells with NVP-AUY922 (50 nM) for 24 hours led to a 70% reduction of ALK level and a abrogation of Akt phosphorylation as determined by western blot analysis. Likewise, no phosphorylation of Akt was detectable after incubation with BGT266 (10 nM) already after 4 hours. RAD001 (0.1-1nM, 24h) completely inhibited phosphorylation of p70 S6K. In contrast, Karpas299 cells were only sensitive to RAD001-induced cell cycle arrest, but insensitive to NVP-AUY922 and BGT266. In vivo, we performed FLT- and FDG-PET scans to monitor inhibition of tumor growth in the course of therapy with NVP-AUY922. Tumor volume in treated animals bearing SUDHL-1 lymphomas showed modest increase within the first week (median increase= + 25%, range -30% to + 80%, n=8) as opposed to a 3.8-fold increase in untreated control animals. After 14 days a clear reduction of tumor mass could be observed (median= - 25%, range -40% to + 30%, n=4). Median TBR of FLT-PET decreased significantly to 40% compared to baseline as earlier as 5 days after initiation of therapy (range 32–67%, n=8, p=0,008). In contrast, the pattern of TBR in FDG-PET did not show any clear tendency (median TBR 79%, range 36%-161%, n=8, p=0,73). We then investigated the ability of FLT-PET to differentiate between sensitive and resistant lymphoma cells. Therefore, mice bearing Karpas299 lymphomas were treated with NVP-AUY922 (resistant in vitro) or RAD001 (sensitive in vitro). According to our in vitro results, no effect was seen during treatment with NVP-AUY299 as indicated by about 3-fold tumor growth on day 7 and increase of median TBR in FLT-PET to 162% (range 106–177%, p=0,008, n=8) on day 2. In contrast, mice receiving RAD001 showed a deceleration of tumor development with doubling of tumor volume within the first week (range -20% to + 320%, n=10) that remained fairly constant over the following weeks. FLT-PET imaging indicated a slight increase of TBR correctly reflecting tumor growth kinetics (median=126%, range 60–129%, no p-value). A larger cohort is currently investigated as well as histopathological analysis of explanted lymphomas. The updated data will be presented at the meeting.

In contrast to FDG-PET, FLT-PET is able to predict response to specific inhibitors early in the course of the therapy using a anaplastic large cell lymphoma xenograft model and is able to distinguish between sensitive and resistant lymphoma cells.

No relevant conflicts of interest to declare.