Background
Oseltamivir (Tamiflu®, F. Hoffmann-La Roche, Ltd., Basel, Switzerland) is the orally active ester prodrug of the anti-influenza agent Ro 64-0802, a potent and selective viral neuraminidase inhibitor that is effective for the treatment of influenza A, B, and A (H1N1). Unusual neuropsychiatric events even including suicidal events in young patients taking oseltamivir have been reported in Japan [[
1]]. Such incidents have been suspected as neuropsychiatric adverse events (NPAEs) of oseltamivir, although the underlying mechanisms have not yet been clarified.
Due to such issues, great efforts have been made to elucidate the factors affecting oseltamivir penetration into the central nervous system (CNS) in rodents and nonhuman primates. It was elucidated that P-glycoprotein (P-gp) at the blood–brain barrier (BBB) limits plasma oseltamivir penetration into the brain [[
2],[
3]]. Ro 64-0802 barely diffuses into the brain, and its high hydrophilicity (cLogP −0.97) was initially considered to be the reason [[
3],[
4]]. However, a later study using mice demonstrated that active efflux at the BBB mediated by organic anion transporter 3 (Oat3) and multidrug resistance protein 4 (Mrp4) also contributes to the low brain penetration [[
5]]. In healthy adult humans, the cerebrospinal fluid/plasma ratio of oseltamivir and Ro 64-0802 were at most 2.1% and 3.5%, respectively, suggesting their low brain penetration [[
6]]. To investigate the brain penetration in humans and nonhuman primates, [
11C]oseltamivir and [
11C]Ro 64-0802 were synthesized as positron emission tomography (PET) imaging probes [[
4]]. Recent studies demonstrated that the BBB penetration of [
11C]oseltamivir was slightly higher in adolescent monkeys than that in adults [[
7],[
8]].
NPAEs have occurred in young patients during relatively early phase after the onset of influenza-like symptoms [[
9]]. Furthermore, it was reported that 30% of the abnormal behaviors of pediatric patients occurred within 2 h after the first use of oseltamivir [[
9]], which is close to the time of reaching maximum plasma concentrations of oseltamivir and Ro 64-0802 [[
6]]. It is possible that the immune response leads to the elevation of CNS oseltamivir concentration by increasing BBB permeability and/or plasma oseltamivir concentration. Influenza virus infection strongly activates the immune system and induces pro-inflammatory cytokines and chemokines systemically. Some in vitro studies have reported that pro-inflammatory cytokines alter the integrity of BBB or induce modulation of efflux transporters [[
10]]. It is reported that the expression of human carboxylesterase 1 (CES1), responsible for conversion of oseltamivir to Ro 64-0802 [[
11]], was suppressed by pro-inflammatory cytokine interleukin-6 (IL-6) in vitro [[
12]]. This highlights the importance of investigating the effect of viral infection on the CNS exposure of oseltamivir and Ro 64-8082 under therapeutic dosage in vivo.
In the present study, we attempted to evaluate the alterations of CNS uptake of [
11C]oseltamivir under similar immune activation of viral infection under therapeutic dose of oseltamivir medication in living juvenile monkeys using PET. Immune activation models of
Rhesus monkeys have been developed using synthetic double-stranded RNA, polyinosine-polycytidylic acid (poly I:C), which induces increasing IL-6 production in the blood [[
13]]. We used an immune activation model in juvenile monkeys where poly I:C was administered as an experimental model of viral infection. Stimulation by the systemic administration of poly I:C is similar to that by viral infection upstream of the innate immune response, such as the induction of pro-inflammatory cytokines [[
14]] and interferon in monkeys [[
15]].
The advantages of the use of monkeys over rodents are the following: (
1) their CNS developmental changes are similar to those in humans; (2) oseltamivir is converted to Ro 64-8082 by hepatic CES in monkeys as in humans, whereas serum CES activity is high in rodents [[
3]]; and (3) monkeys are capable of undergoing serial blood samplings required for estimation of BBB permeability. To achieve therapeutically relevant plasma concentrations of oseltamivir and Ro 64-0802 during PET scan, unlabeled oseltamivir was administered because the dose of [
11C]oseltamivir for PET imaging was approximately 10 μg/kg. This study is the first to report the impact of immune activation on oseltamivir brain uptake in living juvenile monkeys under therapeutic oseltamivir dosage.
Discussion
In the present study, we examined the effect of innate immune activation on the CNS uptake of oseltamivir at its clinically relevant plasma concentrations using [
11C]oseltamivir and PET in living juvenile monkeys. We used intravenous poly I:C administration to activate the immune response to simulate viral infection. Poly I:C stimulates toll-like receptor 3 (TLR3), one of the innate immune-recognition receptors, whereas influenza virus is recognized by TLR7. Both TLRs are expressed in endosome, and their stimulations result in the induction of pro-inflammatory cytokines via different signaling pathways [[
21]]. This study demonstrated an immediate increase of plasma IL-6 after poly I:C treatment, while no alterations were observed in TNF-α and IL-1β in the first 2 h (Table
2). The elevation of plasma IL-6 has been reported as the common feature in influenza patients including children [[
22],[
23]], and therefore, the immune responses to poly I:C and influenza virus are similar despite differences in the signaling pathways. From the rapid elevation of plasma IL-6 throughout the experimental period, the immune activation model in this study was considered to simulate a part of the innate immune response to influenza virus infection. The lack of elevation of body temperature is considered to be due to the hypothermic effect of the applied anesthesia.
In the quantification of the CNS concentration of low BBB permeable radiolabeled ligand, the linearity of measured radioactivity concentration from a low level and the contribution of intravascular radioactivity must be considered significant. The lowest decay-uncorrected radioactivity concentration in the brain was approximately 0.2 kBq/cm
3. The PET scanner used in this study has a linearity between 0.05 and 50 kBq/cm
3. Therefore, the brain concentration was accordingly quantified with the PET scanner. Nevertheless, in consideration of the noise, we placed ROI on the central semiovale where the largest ROI was available for better statistics. Regarding intravascular radioactivity, both the TACs of blood and brain peaked at about 1 min simultaneously, as shown in Figures
4b and
5a. At this point, the peak of blood radioactivity was approximately 1% ID/mL × kg (Figure
4b). Using
Vb of 0.025 mL/cm
3 estimated by integration plot analysis, the vascular radioactivity at the peak was estimated to be about 0.025% ID/cm
3 × kg. This value is close to the peak values of the brain TACs (Figure
5a). At later times, between 40 and 60 min post injection, the mean normalized blood concentration of poly I:C treatment condition was 0.025% ID/mL × kg. Using
Vb of 0.025 mL/cm
3, the vascular radioactivity in the brain ROI can be estimated as 0.000625% ID/cm
3 × kg, corresponding to 18% of
Cbrain,40-60min (0.0035% ID/cm
3 × kg). This indicates that the contribution of vascular radioactivity at later times became less than in the first few minutes. To remove the contribution of intravascular radioactivity from the measured brain concentration, Equation
1 can be applied to estimate the net brain concentration
Ct(
t) using
Vb obtained from integration plot analysis (Equation
3). By this intravascular radioactivity correction (CBV-correction),
Cbrain,40-60min and
Kp, brain overall became 15% less than the uncorrected values, but the changes of
Cbrain,40-60min and
Kp,brain values by poly I:C treatment were very similar. The individual CBV-corrected results are shown in Additional file
1.
With respect to the plasma-to-brain transfer rate, the estimated
K1 values were considerably lower than monkey CBF, which is approximately 0.5 mL/min/g [[
20]], and therefore, the extraction fraction (
E) is approximated to 0.01 using the relationship
K1 =
E∙CBF. According to the Crone-Renkin equation [[
24],[
25]], the penetration of [
11C]oseltamivir across the BBB is in a diffusion-limited manner, and therefore,
K1 is not considered to reflect CBF.
The CNS radioactivity concentration measured with PET includes both [
11C]oseltamivir and [
11C]Ro 64-0802. The latter is less BBB-permeable than oseltamivir, so the radioactive metabolite [
11C]Ro 64-0802 penetration into the CNS is less likely. But, the fact that the immunoreactivity of CES1 was detected in human brain endothelial cells [[
26]] suggests that [
11C]Ro 64-0802 can be formed from [
11C]oseltamivir in brain capillary endothelial cells. Although Oat3 and Mrp4 facilitate removal of Ro 64-0802, which is formed in the endothelial cells from the brain to the blood [[
5]], part of the radioactivity in the brain may be attributed to [
11C]Ro 64-0802, but the major portion of CNS radioactivity can be ascribed to [
11C]oseltamivir.
With the use of CNS concentration measured with PET, one can roughly estimate the CNS drug concentration under therapeutic dose. Using
Cbrain,40-60min of 0.0035% ID/cm
3 × kg, the CNS oseltamivir concentration at a therapeutic dose of 2 mg/kg can be approximated as 0.22 μM. Or, with the use of plasma-to-brain ratio
Kp,brain,40-60min and the reported
Cmax in a clinical study of 115 ng/mL [[
6]], the CNS concentration can be estimated as high as 0.37 μM. It has been reported that no inhibitory effect against recombinant human neuramidases was detected up to 1 mM of both oseltamivir and Ro 64-0802 [[
24]]. This report also demonstrated that no relevant inhibitory effect against various molecular targets of neurotransmitter system was observed up to 3 μM for both oseltamivir and Ro 64-0802 [[
27]]. Alternatively, a previous study of electrophysiology experiments using mouse brain reported that the pharmacological effect was detected at ED
50 of 10.2 μM for oseltamivir and 0.7 μM for Ro 64-0802 [[
28]]. The CNS concentrations estimated with the PET data are lower than these reported concentrations.
The normalized brain concentration at later times and the transfer rate (
K1 or CL
uptake, brain) were about half of those of a previous report using adolescent monkeys [[
8]]. Several factors including different anesthesia can be considered. First, the monkeys used in this study were in a later developmental stage with almost double the body weight of those used in the previous report. Therefore, the different developmental stage would be a reason for the difference. Second, the values of the parameters are considered to depend on the placing of the ROI. We chose the central semiovale. As described previously, the radioactivity was considered relatively lower than in other regions. Third, the number of blood samples for input function used for integration plot analysis can be a factor that causes different results in calculated transfer rates. In this study, we took samples 12 times, with the shortest interval being 10 s, between 10 s and 2.5 min, whereas four samples with the shortest interval of 30 s were collected between 30 s and 2.5 min in the previous study.
One of the limitations of this study was the small sample size and gender heterogeneity. For this reason, statistical evaluation of changes in oseltamivir brain uptake by poly I:C treatment could not be strictly performed.
There have been no studies to test the alterations of a drug penetration into the CNS by immune activation in vivo in nonhuman primates under therapeutic dose. PET imaging of radiolabeled drug in a monkey model of immune activation is advantageous for assessing the changes in drug distribution by viral infection directly in vivo.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made.
The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
CS, AON, YN, TM, SO, MI, and TS conceived of and contributed to the study design. CS, AON, YN, and SO performed the experiments and acquired the PET data. CS analyzed the data and drafted the manuscript. CS, AON, YN, and TM participated in interpretation of the data. AON, MRZ, MH, and HK revised the manuscript. MT, KF, TI, and MRZ carried out the radiochemical synthesis and radio-HPLC metabolite analysis. SI and HK quantified the plasma unlabeled drug concentrations. HI and YS reviewed the manuscript. TS supervised the study. All authors read and approved the final manuscript.