Discussion and Conclusion
A clear understanding of drug exposure at the site of action, target binding and functional pharmacological activity is key for optimal design of phase II clinical trials. (Morgan et al., 2012) MR309, a selective σ1R antagonist, was well tolerated in early clinical trials at a daily dose of 400 mg and signals of preliminary efficacy for acute CIPN were observed in patients with colorectal cancer receiving an oxaliplatin-containing regimen (Abadias, Escriche, Vaque, Sust & Encina, 2013; Bruna et al., 2018) and in patients with chronic PSNP lasting for >6 months. (EU Clinical Trials Register (EudraCT 2012-000402-30)) Here, we confirm the occupancy of brain σ1R by MR309 following administration of clinically meaningful doses, and demonstrate a quantitative relationship between plasma concentration of MR309 and target occupancy, enabling optimised dose selection in future studies.
The PK study demonstrated that BID dosing of MR309 offers the opportunity to lower fluctuation of drug plasma levels compared with QD dosing. At steady state lower peak and higher trough concentrations of MR309 were observed with BID compared with QD dosing: geometric mean [CV, %] Cmaxss (4143 [17.2] ng/mL vs 5222 [18.3] ng/mL); Cminss (1915 [29.8] ng/mL vs 1764 [29.9] ng/mL). Consequently, it is feasible that BID dosing of MR309 may be better tolerated than QD dosing, with the potential to reduce the incidence and/or severity of AEs associated with high peak plasma concentrations; this may have particular relevance as MR309 is a CNS active compound. Indeed, in this study treatment-emergent AEs (56.3% vs 37.5%) and treatment-related AEs (43.8 vs 18.8%) were more frequent with MR309 400 mg QD compared with 200 mg BID. Based on these PK findings, BID dosing of MR309 was recommended for evaluation in future studies.
In the second study, the relationship between plasma concentrations of MR309 and σ1R occupancy in the human brain was estimated successfully. Modelling the relationship between MR309 plasma concentration and σ1R occupancy indicated that for MR309 200 mg BID dose, which produces steady state plasma concentrations of approximately 2000–4000 ng/mL, brain σ1R occupancy will be in the range of 59–74%. The data suggest that meaningful levels of σ1R occupancy can be achieved with steady state dosing of 400 mg MR309 daily dose. Further, the signs of clinical efficacy in the CIPN and in the PSNP studies seen with 400 mg daily dosing could be interpreted as further support for the mechanism of action. It is apparent that higher σ1R occupancy rates will need to be investigated to explore the full potential of σ1R in neuropathic pain. Notably, nonclinical data suggest that a higher degree of brain σ1R occupancy may be required to elicit a maximal effect. When MR309 was administered to animal models of neuropathic pain, ex vivobinding in brain section autoradiograms using3H-pentazocine and analysis of dose–response curves indicated maximal antinociception was achieved at brain σ1R occupancy levels >75%. (Romero et al., 2012)
In addition to achieving higher brain σ1R occupancy, it is also important to elucidate the clinical impact of 59–74% brain σ1R occupancy by a selective σ1R antagonist. Plasma steady-state was previously reported after 8 days of MR309 QD (based on no significant difference in Cmin values on days 6, 7 and 8; p<0.05) (Abadias, Escriche, Vaque, Sust & Encina, 2013). Consequently, the 5-day dosing regimen for MR309 utilized in the Phase II study of CIPN study is anticipated to result in σ1R occupancy below 59–74%. Further studies are warranted to fully explore the potential of MR309 to mitigate signs of oxaliplatin-induced acute sensory and motor neurotoxicity, and reduce the frequency and severity of chronic sensory neuropathy. These would require higher dosing of MR309 or a pre-dosing regimen in order to achieve steady state levels resulting in σ1R occupancy >75%.
Investigation of σ1R occupancy by MR309 in peripheral nerves is also warranted, as it is feasible that receptor occupancies may differ between the peripheral and central nervous systems, given that concentrations of MR309 were approximately 2-fold greater in brain compared with plasma in animal studies. (Romero et al., 2012) To our knowledge, the only PET study analysing σ1R in peripheral tissue was a case report of a patient with chronic knee pain. Using18F-FTC-146 σ1R radioligand, σ1 receptor upregulation was reported in a subject with an inflamed intra-articular synovial lipoma. Removal of the lipoma resulted in cessation of pain. (Cipriano et al., 2018)
It is noteworthy that MR309 was well tolerated by the healthy volunteers participating in the present two studies. AEs were mild in severity, with dizziness being the most frequently reported event. Orthostatic pulse rate changes were also reported by some individuals in the PK study, although were not correlated with plasma concentrations of MR309 and were not consistent with postural orthostatic tachycardia syndrome. (Stewart, 2004) While dizziness was also a common side effect reported by healthy volunteers in a prior Phase I study of MR309, it was reported less frequently by patients with colorectal cancer in the Phase II study of MR309 compared with the present study. (Abadias, Escriche, Vaque, Sust & Encina, 2013; Bruna et al., 2018) Variations in the reported safety profile may relate in part to the different characteristics of the study participants, including underlying disease and chemotherapy.
The reported studies are associated with some limitations. Both studies were exploratory and the PK– σ1R model fitting did not account for potential variation in the precision with which individual data points were determined, or that the data set comprised multiple points for each participant rather than independent samples. Consequently, while the choice of model used in the present study had only a modest effect on EC50 estimates (ranged from 1268–1380 ng/mL), it is feasible that more sophisticated model fitting may have produced a different EC50 estimate. Also, based on data that were available at the time the PK study was designed, 9 days of dosing with MR309 was considered sufficient to achieve steady-state PK. However, in the present study small increases in trough concentrations of MR309 were observed prior to Day 10, when full PK sampling was obtained (data not shown). This was not anticipated, given the previously reported half-life of MR309 was approximately 18 hours. (Abadias, Escriche, Vaque, Sust & Encina, 2013) As trough concentrations were not formally compared to establish whether steady-state had been achieved in the present study, further evaluation of the time to establish steady state is recommended in future studies of MR309. With regard to the PET study, an underestimation of receptor occupancy achieved with MR309 may occur if SA4503 had limited selectivity for the σ1R. SA4503 has been reported to have σ2R / σ1R ratio of approximately 100-fold, with no significant affinity to 36 other receptors, transporters or channels. (Kawamura et al., 2000; Matsuno, Nakazawa, Okamoto, Kawashima & Mita, 1996) However, using masking protocols with selective ligands, the σ2R / σ1R ratio was reported to be 14-fold in guinea pig brain membranes, although such masking protocols were shown to carry their own challenges and these data should be treated with caution. (Abbas, Borde, Willars, Ferry & Safrany, 2020; Lever, Gustafson, Xu, Allmon & Lever, 2006). The magnitude of the PET signal is proportional to the ratio of target density and affinity (Bmax/Kd). The σ2R, a transmembrane ER-resident regulator of the Nieman-Pick-Carrier protein, (Alon, Schmidt, Wood, Sahn, Martin & Kruse, 2017) contributed to the PET signals obtained in this study. While sigma receptor expression is qualitatively similar between various animal species and man, there are notable species-dependent differences in expression levels. For example, Bmax for σ1R in cerebellum differs between s mouse strains and is lower for Rhesus compared with guinea pig and mice. (Brust, Deuther-Conrad, Lehmkuhl, Jia & Wunsch, 2014) While human data on brain σ1R binding has not been published, σ2R expression was found to be much lower in human cortical membranes than in rat brain. (Brust, Deuther-Conrad, Lehmkuhl, Jia & Wunsch, 2014) Data so far suggest comparable expression levels of σ1R and σ2R. Based on the assumption of comparable σ1R and σ2R expression levels and using the lowest reported σ2R / σ1R selectivity of 14, the contribution of non- σ1R component to the total [11C]SA4503 will be at most ~7%. Assuming that MR309 has no affinity for σ2R, and use this estimate to adjust for the potential bias in the estimation of σ1R occupancy, we find a minor effect with the expected EC50 at σ1R being 1225 ng/mL and a range of 62-77% receptor occupancy expected in the plasma concentration range of 2000-4000 ng/mL.
In conclusion, our data indicate MR309 200 mg BID dosing facilitates less fluctuation of plasma levels of this selective σ1R antagonist compared with 400 mg QD dosing. However, PK– σ1R occupancy modelling indicated that at steady state a daily dose of MR309 200 mg BID would be associated with a level of brain σ1R occupancy <75%, which may be insufficient to elicit maximal efficacy in neuropathic pain. Furthermore, the 5-day MR309 regimen utilized in the previous study of CIPN would likely be associated with lower brain σ1R occupancy compared to that observed in the present study of healthy volunteers. (Bruna et al., 2018) Consequently, further clinical studies of MR309 in neuropathic pain indications are needed to inform efficacy and tolerability across a broad range of brain σ1R occupancy levels in order to establish the optimal dose ranges in different patient populations.