Analysis of [¹⁸F]SPA-RQ

Substance P is the preferential endogenous agonist of NK₁ receptor, and the SP-NK₁R system has been linked to affective and addictive disorders, nausea, analgesia, inflammation, pain, itching, and depression (Majkowska-Pilip et al., 2019; Schöppe et al., 2019).

[2-[¹⁸F]Fluoromethoxy-5(5-trifluoromethyl-tetrazol-1-yl)-benzyl]([2S,3S]2-phenyl-piperidin-3-yl)-amine, [¹⁸F]SPA-RQ, is a selective NK₁ receptor antagonist with a very high affinity (Solin et al., 2004), and moderate radiation dose, allowing subjects to receive multiple PET scans in a single year (Sprague et al., 2007). [¹⁸F]SPA-RQ can be labelled with very high molar activity, thus avoiding any significant receptor occupancy by the radiopharmaceutical. [¹⁸F]SPA-RQ can be used for in vivo imaging of human NK₁R availability (Bergström et al., 2004; Hietala et al., 2005; Yasuno et al., 2007; Nyman et al., 2007 and 2019; Fujimura et al., 2009).

[¹⁸F]SPA-RQ is useful for preclinical studies in guinea pigs, which have similar NK₁Rs than humans (Bergström et al., 2004; Solin et al., 2004; Grönroos et al., 2019). In addition to the brain, [¹⁸F]SPA-RQ could be used for studying peripheral NK₁ receptors, as marked specific uptake in the intestines, pancreas, uterus, urinary bladder, skin, and lungs in guinea pigs has been observed (Grönroos et al., 2019).

Data analysis methods

Metabolite correction

Fraction of parent tracer in plasma decreases relatively slowly in human studies. The percentage of unchanged [¹⁸F]SPA-RQ was ∼30-40% at 80 min after bolus administration (Hietala et al., 2005). Radio-TLC analysis revealed a hydrophilic metabolite in addition to free [¹⁸F]F^-. Radioactivity in the skull bone was apparent in some subjects (Hietala et al., 2005). Fractions of parent tracer in plasma can be fitted with Hill function (Hietala et al., 2005).

Plasma protein binding

In guinea pigs the free fraction of radioactivity in plasma was 35-41% at 15 min and ∼50% 180 after administration (Grönroos et al., 2019), suggesting lower plasma protein binding of the radioactive metabolites.

Compartmental model

Regional human brain data, collected for 90 min, can be fitted with reversible two-tissue compartmental model, using arterial plasma curve as the input function, although with high parameter variability (Hietala et al., 2005). V_B was constrained to 4%. Regional binding potential (BP) was either reported as the k₃/k₄ when K₁/k₂ was constrained to the value in the cerebellum, or, as BP_ND from distribution volumes calculated from the compartmental model parameters with cerebellum as the reference region (Hietala et al., 2005).

Simplified reference tissue model (SRTM), with cerebellum as the reference region, can be used to calculate the binding potential BP_ND (Hietala et al., 2005). MRTM2 provided better test-retest reproducibility than SRTM (Yasuno et al., 2007). The estimate of BP_ND increases with increasing scan duration in regions with high receptor density (putamen and caudate) up to 5 h after administration. Four hour scan time was found to provide stable binding potential estimates (Hietala et al., 2005; Yasuno et al., 2007).

Reference region

The lowest uptake in human brain was seen in the cerebellum, and uptake peak was observed already ∼5 min after administration, supporting its suitability as reference region; in striatum, the peak was observed at ∼210 min. Displacement study with autoradiography method in postmortem human brain confirmed that cerebellum does not have specific [¹⁸F]SPA-RQ uptake (Hietala et al., 2005). Reversible two-tissue compartmental model fitted the cerebellum curve better than one-tissue compartmental model (Hietala et al., 2005), which may be explained by partial volume effect and free fluoride. One-tissue compartmental model fit was still reasonably good, supporting the applicability of simplified reference tissue model.

Analysis of [18F]SPA-RQ