Assessment of test-retest reproducibility by [18F]Bavarostat for PET imaging of HDAC6.

Abstract

Background: Histone deacetylase 6 (HDAC6) is an enzyme pivotal for gene regulation, influencing cellular pathways through protein deacetylation. HDAC6 is a potential therapeutic target in diseases such as cancer and neurodegenerative disorders. Koole et al. investigated brain binding of [¹⁸F]Bavarostat, an HDAC6 inhibitor, in healthy participants, revealing an absolute test-retest variability (aTRV) of 7.7% (n = 4) for the distribution volume (V_T) with a 1-day interscan interval. This study aims to evaluate test-retest reproducibility with a more extended interscan interval.

Results: Six participants (3 M/3F) underwent a test-retest scan, each lasting for 120 min using a 4-ring Biograph mCT PET/CT scanner. Arterial blood sampling and metabolite analysis were performed to derive the input function. The two scans were 28 ± 12 days apart (14-43 days, n = 6). Regional time-activity curves (TACs) were generated for 15 regions of interest (ROIs). Kinetic analysis of the 120-min TACs was performed using one-tissue and two-tissue compartment models (1TC, 2TC) and multilinear analysis-1 (MA1) to quantify V_T values and compute absolute test-retest variability (aTRV). The effects of scan duration (60 to 120 min) and MA1 t* setting on aTRV and bias were investigated. Careful analysis of the plasma HPLC data was needed since metabolites eluted close in time to the parent. The MA1 model (t* = 40 min) adequately described regional TACs and produced stable kinetic parameters with good agreement to 2TC (MA1 V_T=0.98 × 2TC V_T + 0.48, bias: -0.1%) while 1TC underestimated V_T by 5.1%. Regional V_T values exhibited a relatively uniform pattern, highest in the amygdala and lowest in the centrum semiovale. Individual aTRV values ranged from 2 to 9%. Scan durations between 100 and 120 min provided the most consistent results, with minimal bias and acceptable aTRV across all tested t* values. Although a 90-minute scan with t*=10 or 20-minute balanced scan time and aTRV, optimal parameters varied by brain region. Smaller regions (e.g., amygdala) required longer scans to achieve reliable V_T quantification.

Conclusions: The test-retest variability of [¹⁸F]Bavarostat V_T values demonstrated favorable results for a one-month scan interval, comparable to the reported values.