[18F]FPyGal PET TRACER DETECTS SENESCENCE IN HUMAN OSTEOARTHRITIC SPECIMENS

INTRODUCTION

Cellular senescence, a hallmark of aging, plays a key role in the development of osteoarthritis (OA). Several senolytic therapies have been developed to clear senescent cells in the joint resulting in delayed cartilage degradation and improved clinical symptoms of patients with OA. However, a critical challenge remains: Developing reliable imaging techniques to detect senescence in patients. This will be essential to effectively monitor the efficacy of senolytic therapies and personalize treatment for OA.

OBJECTIVE

Senescent cells overexpress β-galactosidase (β-gal). We have demonstrated in vitro (primary chondrocytes) and in vivo (small animal model-mice and a large animal model-pigs) that [18F]FPyGal, a β-gal targeted PET tracer can detect senescent cells (Figure 1). The objective of our study was to evaluate if [18F]FPyGal could detect senescent cells in human joint specimen from patients with OA. We hypothesized that [18F]FPyGal retention in human specimen, as measured by positron emission tomography (PET), would correlate with the Outerbridge score, determined on simultaneously acquired MRI scans.

METHODS

This study was approved by the Institutional Review Board of our Institution (IRB-62254). Written informed consent was obtained from five patients (one male and four females with an age of 63-86 years (mean 72.8 ± 8.98) to donate their knee specimens after total knee replacement with a joint endoprosthesis. The ten freshly obtained specimens were incubated with 200μCi of the radiotracer for an hour at room temperature. The specimens were washed trice with PBS and imaged in a clinical PET/MRI scanner (Signa GE Healthcare, Chicago, IL). The MRI protocol consisted of a fat-saturated proton density-weighted fast spin-echo sequence (TR = 3,345 ms, TE = 33 ms, FA = 111°, matrix size = 192 × 192 pixels, slice thickness (SL) = 1.5 mm, FOV = 8 cm, and TA = 5 min along and a LAVA sequence (TR = 3.802 ms, TE=1.674, FA=3, Matrix=192 × 192 pixels) for attenuation correction. PET images were acquired simultaneously and reconstructed using the Ordered-Subset Expectation Maximization (OSEM) algorithm with 2 iterations and 28 subsets. The PET/MRI scans were independently analyzed by one nuclear medicine physician and one radiologist. The radiologist assigned a modified Outerbridge score (1-4) of the cartilage damage of these areas, while the Nuclear Medicine physician measured the standardized uptake values (SUV) of the same areas. The SUV and Outerbridge score were correlated with Jonckheere-Terpstra test.

RESULTS

PET/MRI images of human osteoarthritic specimens demonstrated focal retention of [18F]FPyGal radiotracer in some cartilage areas and not others at 1 hour after incubation with 200μCi [18F]FPyGal radiotracer. A significantly higher radiotracer uptake was observed in cartilage areas with an Outerbridge score of 3-4 (0.45±.23μCi/ml) compared to bone marrow as an internal reference tissue (0.1±0.09μCi/ml, p=0.003) (Figure 2). There was a significant positive correlation between SUV and Outerbridge score, as indicated by the Jonckheere-Terpstra test (p<0.0001).

CONCLUSION

Areas of cartilage damage in human osteoarthritic specimens exhibited significant radiotracer uptake upon incubation with the [18F]FPyGal radiotracer as evidenced by a significant increase in SUV_max values.