Advertisement

The AAPS Journal

, 20:23 | Cite as

Demonstration of Direct Nose-to-Brain Transport of Unbound HIV-1 Replication Inhibitor DB213 Via Intranasal Administration by Pharmacokinetic Modeling

  • Qianwen Wang
  • Yufeng Zhang
  • Chun-Ho Wong
  • H.Y. Edwin Chan
  • Zhong Zuo
Research Article

Abstract

Intranasal administration could be an attractive alternative route of administration for the delivery of drugs to the central nervous system (CNS). However, there are always doubts about the direct transport of therapeutics from nasal cavity to the CNS since there are only limited studies on the understanding of direct nose-to-brain transport. Therefore, this study aimed to (1) investigate the existence of nose-to-brain transport of intranasally administered HIV-1 replication inhibitor DB213 and (2) assess the direct nose-to-brain transport of unbound HIV-1 replication inhibitor DB213 quantitatively by a pharmacokinetic approach. Plasma samples were collected up to 6 h post-dosing after administration via intranasal or intravenous route at three bolus doses. In the brain-uptake study, the plasma, whole brain, and cerebrospinal fluid (CSF) were sampled between 15 min and 8 h post-dosing. All samples were analyzed with LC/MS/MS. Plasma, CSF, and brain concentration versus time profiles were analyzed with nonlinear mixed-effect modeling. Structural model building was performed by NONMEM (version VII, level 2.0). Intranasal administration showed better potential to deliver HIV-1 replication inhibitor DB213 to the brain with 290-fold higher brain to plasma ratio compared with intravenous administration. Based on that, a model with two absorption compartments (nose-to-systemic circulation and nose-to-brain) was developed and demonstrated 72.4% of total absorbed unbound HIV-1 replication inhibitor DB213 after intranasal administration was transported directly into the brain through nose-to-brain pathway.

KEY WORDS

CNS targeting delivery DB213 intranasal pharmacokinetic modeling 

INTRODUCTION

Increasing evidence suggests that intranasal administration has a potential to bypass the blood-brain barrier (BBB) and can therefore be a popular noninvasive route to deliver therapeutics to the central nervous system (CNS) [1]. The in vivo fate of a drug when delivered intranasally could be described schematically as follows: First, it reaches the respiratory epithelium and from there it is absorbed into the systemic circulation. Then the drug will enter the CNS by passing through the BBB. In the past two decades, the connection between the nasal cavity and the CNS by olfactory and trigeminal neurons had been investigated extensively, adding alternative absorption routes that could bypass the BBB. In this regard, Thorne et al. has systematically demonstrated the pathways from the nasal passage to the CNS which includes three sequential transport steps: (1) transport across the epithelial barriers (olfactory or respiratory) in nasal passages through intracellular or paracellular pathways, (2) transport from the nasal mucosa to sites of brain entry near the pial brain surface in the cranial compartment, and (3) transport from the initial entry site of the brain to other sites of the tissue [1].

Several studies have reported the existence of direct nose-to-brain transport of therapeutics via intranasal route. Numerous therapeutics have been delivered to the CNS system following intranasal administration. These therapeutics vary from small molecules (e.g., tacrine [2]) to large molecules such as peptides (e.g., insulin [3]), proteins (e.g., IGF-1 [4]), and gene vectors [5], as well as stem cells [6]. To further identify the nose-to-brain pathway, some studies utilized imaging technologies such as radiographic analysis and PET scan [7, 8, 9, 10]. The imaging technologies are able to show the transport of labeled drug molecule from the nasal cavity to brain clearly. However, one important problem is that the signal detected in the CNS does not always represent the intact drug molecule [11]. Except imaging technology, some studies reported brain-drug concentration after intranasal administration at a single or multiple time points. By calculating the AUC0-∞ ratio of brain to plasma pharmacokinetics profiles, a direct nose-to-brain pathway could be demonstrated by a higher ratio via intranasal administration than that of intravenous administration [12, 13, 14]. In addition, drug targeting efficiency (%DTE) and nose-to-brain direct transport (%DTP) have been adopted to quantitatively estimate drug delivered to the brain via nasal route based on following conditions: (a) the pharmacokinetics of the drug is assumed to be linear without saturation during absorption, distribution, metabolism, and elimination; (b) both AUCbrain and AUCblood are calculated using the unbound drug concentrations which are pharmacologically relevant; (c) drug after intranasal administration is evenly distributed in the brain. However, such conditions were not fully verified in majority of the published studies using %DTE and %DTP [15, 16, 17, 18, 19]. Moreover, the calculation of %DTE and %DTP values was based on the overall drug exposure values which are highly dependent on whether the drug elimination is completed at the last sampling point [15]. For drugs that may not demonstrate homogenous distribution in the brain, instead of using %DTP, Westin et al. obtained both olfactory and brain hemisphere drug concentrations at certain time intervals to show that 95% of morphine was transported following the direct olfactory-to-brain hemisphere pathway in the first 5 min post-dosing [14]. It is also noted that estimating the amount of unbound drug delivered to the brain is crucial since only the unbound drug is best correlated with its clinical outcome and side effects [20, 21]. Recently, Stevens et al. first applied pharmacokinetic modeling approach to demonstrate both nose-to-brain and nose-to-systemic circulation of unbound remoxipride quantitatively [22].

DB213 (Fig. 1) is an HIV-1 replication inhibitor and is regarded as a promising candidate in new generation of combination antiretroviral therapy (CART) targeting HIV-associated neurocognitive disorders (HAND) [23]. HAND has been used to describe the range of neurocognitive dysfunction associated with HIV infections [24]. To date, there have been no successful clinical trials of HAND therapies. On one hand, each year around 10–30% patients receiving CART has developed resistance, which leads to CART failure in most of the cases [23]. On the other hand, the fact that HIV being persistent in the brain even when systemic virological control is achieved by CART suggests its brain concentration is not high enough to achieve virological control in CNS [24, 25]. Therefore, one of the approaches for new HAND therapy is to deliver the promising CART candidate DB213 via intranasal administration, which has a potential to bypass BBB and increase brain uptake of therapeutics. Thus, the current study is designed aiming to quantitatively assess direct transport of unbound DB213 into the brain via the intranasal route.
Fig.1

Structure of HIV-1 replication inhibitor DB213

MATERIALS AND METHODS

Materials and Chemicals

DB213 was purchased from Dieckmann Chemical Industry CO., LTD. (Hong Kong SAR, P. R. China). Benzamidine, used as the internal standard, was purchased from Sigma-Aldrich (St. Louis, MO, USA). Ammonium formate, sodium chloride, potassium chloride, calcium chloride, anhydrous magnesium sulfate, and dibasic potassium phosphate were obtained from Sigma-Aldrich (St. Louis, MO, USA). Acetonitrile was supplied by Scharlab (Barcelona, Spain), and deionized water was obtained from Millipore water purification system (Millipore, Milford, MA, USA). All the reagents used in the current study were at analytical grade.

Animals

Sprague-Dawley (SD) rats (male, 180–200 g) were supplied by the Laboratory Animal Services Centre at The Chinese University of Hong Kong. The animal studies were conducted under the approval of the Animal Ethics Committee of The Chinese University of Hong Kong (animal ethical approval number: 14/040/GRF-4-B).

Plasma Pharmacokinetic Profiles of DB213 in SD Rats

The rats were randomly assigned to two groups (n = 6–8 per group) to receive DB213 deionized water solution via intravenous or intranasal administration. One day before the experiment, the rats were anesthetized with an intramuscular injection of ketamine (60 mg/kg) and xylazine (6 mg/kg) and received a minor surgery of cannulation with a polythene tube (i.d.0.4 mm × o.d.0.8 mm, Harvard Apparatus, USA) in the left jugular vein followed by overnight recovery with free access to food and water.

In the following day, each rat was weighed and administered with DB213 deionized water solution at a bolus dose of 1, 10, or 50 mg/kg via intravenous or intranasal route. To facilitate the intranasal administration, rats were anesthetized by inhalation of 95% CO2 for a short period of time (recovery within 5 min) and kept at a supine position during intranasal administration with 10 μL of DB213 deionized water solution pipetting into each nostril as we did previously [26]. For intravenous administration, to minimize the bias from anesthesia effects, short-term anesthesia by 95% CO2 was performed prior to administration of 0.2 mL DB213 deionized water solution via the jugular vein catheter. The catheter was then rinsed with 0.5 mL of blank blood withdrawn prior to drug administration and 1 mL of saline (0.9% sodium chloride solution containing 50 IU heparin) to ensure the complete dosing.

To avoid the animal going into shock, not more than 10% of the total blood volume (6.40 mL/100 g [27]) was collected during the study period [28]. Around 100 μL blood samples were collected at different time intervals (2, 5, 10, 15, 20, 30, 45, 60, 90, 120, 240, and 360 min for intravenous route and at 5, 10, 15, 20, 30, 45, 60, 90, 240, and 360 min for intranasal route) from the jugular vein catheter followed by obtaining the plasma samples via centrifugation at 8000 rpm for 3.5 min. After last blood sampling, subsequent cardiac infusion using 200-mL saline to each rat was performed to rinse out the blood residue from brain followed by collecting whole brain (including olfactory bulb). Both the collected plasma and brain samples were stored at − 80 °C until analyses of DB213 content by LC/MS/MS.

Brain Uptake of DB213 in SD Rats

Another group of rats was used to conduct the brain-uptake study. Rats were randomly divided into six groups (n = 30–40 per group) to receive DB213 water solution either intravenously or intranasally at a bolus dose of 1, 10, or 50 mg/kg. One day before the experiment, rats received jugular vein cannulation surgery as described above. On the following day, each rat was administered with DB213 deionized water solution through the intravenous or intranasal route. At 15, 30, 60, 90, 120, 240, 360, and 480 min post-dosing, rats were anesthetized by inhalation of 95% CO2 followed by collection of blood samples via cardiac puncture and cerebrospinal fluid (CSF) samples withdrawn from the cisterna magna and then were put to euthanasia. Plasma was obtained from blood samples by centrifugation at 8000 rpm for 3.5 min. Subsequent cardiac infusion using 200 mL saline to each rat was performed to rinse out the blood residue from the brain then the whole brain (including olfactory bulb) was collected. All collected plasma, CSF, and brain samples were stored at − 80°C until analyses.

Determination of Unbound Fractions of DB213 in Biomatrices

The unbound fractions of DB213 in plasma and brain homogenates were determined by the 96-well Thermo ScientificTM Single-Use Rapid Equilibrium Dialysis unit using a previously reported method [29, 30]. Briefly, the fresh plasma and brain were obtained from drug-naïve rats at the same day of the study. DB213 was added to plasma to achieve a final concentration of 0.3, 3, 30, and 300 μg/mL. Two hundred-microliter aliquots (n = 3) were loaded to the equilibrium dialysis unit and were dialyzed against 350 μL phosphate-buffered saline (containing 100 mM sodium phosphate and 150 mM sodium chloride, pH = 7.4 at 37°C) [31]. The brain tissue was diluted threefold with artificial extracellular fluid (aECF) [32] and homogenized with an ultrasonic probe. Then DB213 was added to the brain homogenate to final concentrations of 50, 500, and 5000 ng/g. Two hundred-microliter aliquots (n = 3) were loaded to the equilibrium dialysis unit and dialyzed against 350 μL aECF (pH = 7.6 at 37°C). The equilibrium dialysis unit was covered with sealing tape and incubated at 37°C on an orbital shaker at 250 rpm for 2, 4, or 17 h [33]. After incubation, 100 μL of matrix sample and 100 μL of buffer sample were withdrawn and stored at − 80°C until analysis. The DB213-unbound fraction in plasma or diluted brain homogenate was calculated using Eq. 1. Then, the unbound DB213 in the undiluted brain was calculated using Eq. 2 [30].

$$ \%\mathrm{Free}=\frac{\mathrm{DB}213\ \mathrm{concentration}\ \mathrm{in}\ \mathrm{buffer}}{\mathrm{DB}213\ \mathrm{concentration}\ \mathrm{in}\ \mathrm{matrix}\ \mathrm{sample}}\times 100\% $$
(1)
$$ \%{\mathrm{Free}}_{\mathrm{undiluted}\ \mathrm{brain}}=\frac{1}{1+D\left(\frac{100}{\%\mathrm{Free}}-1\right)}\left(D\ \mathrm{is}\ \mathrm{dilute}\ \mathrm{factor}\right) $$
(2)

Sample Analyses

The quantification of DB213 in rat plasma and brain was conducted by our previously reported assay [23]. Briefly, for the measurement of DB213 in plasma and CSF, DB213 was extracted by protein precipitation followed by LC/MS/MS analysis (Agilent 6430 triple quadrupole mass spectrometer with an electrospray ionization source (ESI), Agilent 1290 pump, and auto-sampler). For the measurement of DB213 in the brain, 1 g brain tissue together with 2 mL saline (0.9% sodium chloride solution) was homogenized via ultrasonic probe homogenization (OMNI Sonic Ruptor 400, OMNI International, Kennesaw, GA, USA) in an ice bath, followed by extraction of DB213 with acetonitrile for LC/MS/MS analysis. Data acquisition and processing were performed by the Agilent Mass Hunter Quantitative Analysis (version B.03.01) software. The extraction recovery of DB213 in both biological matrices remained consistent. For plasma samples, the extraction recovery of DB213 was 42.5 ~ 56.7%. For brain samples, the extraction recovery of DB213 was 47.0~52.7%. The lower limits of quantification were determined to be 1.96, 0.98, and 0.98 ng/mL for DB213 in plasma, CSF, and brain homogenate, respectively.

Calculation of Pharmacokinetic Parameters and Kp,uu/K’p,uu of DB213

Pharmacokinetic parameters were calculated based on the obtained DB213 plasma and brain concentration versus time profiles by fitting the data with a non-compartmental model. The Kp,uu,brain/CSF of DB213 is the ratio of area under the curve (AUC0-∞) of unbound DB213 in brain/CSF to that of unbound drug in plasma via intravenous administration [34], while K’p,uu,brain/CSF is adopted in the current study for intranasal administration using the similar calculation method as Kp,uu,brain/CSF. The unbound plasma and brain concentrations were determined using the unbound fraction investigated previously.

Pharmacokinetic Model Building

The structural model building was conducted by using nonlinear mixed-effect modeling (NONMEM) software (Version VII, level 2.0). The control stream was written using ADVAN6, and the first-order conditional method with interaction (FOCE + I) was used for estimation with a convergence criterion of two digits in the parameter estimates.

The pharmacokinetic parameters clearance (CL), volume of distribution (V), and inter-compartmental clearance (Q) were estimated based on parameter estimates (θ). The transport of DB213 over time between different compartments was defined as rate constants (Eq. 3):

$$ {k}_{\mathrm{x},\mathrm{y}}={\mathrm{CL}}_{\mathrm{x},\mathrm{y}}/{V}_{\mathrm{x}} $$
(3)
In model U1, only unbound plasma data after intravenous administration of DB213 was modeled in a structural model with a central and a peripheral compartment (Fig. 2). In model U2, both plasma and brain concentration of DB213 via intravenous route were modeled and an additional brain compartment was added (Fig. 2). Based on studies by Stevens et al., the elimination of remoxipride from brain was underestimated, and an additional first-order elimination rate constant helped the model building [22]. An additional brain elimination rate constant (k40) was designed in model U3 to investigate the removal of DB213 from brain (Fig. 2). Both unbound plasma and brain data were included in model U3 as well. The value for k40 was assumed to be smaller than total elimination rate constant from systemic circulation (k30) and therefore calculated as a fraction of k30 (Eq. 4):
Fig. 2

Proposed compartmental model structures of model U1, U2, U3, U4, U5, and U6

$$ {k}_{40}=f\times {k}_{30} $$
(4)

Models U1-U3 served as the basis for the identification of more complicated model structures including the intranasal route data set. In model U4, U5, and U6, unbound plasma and brain data via both intravenous and intranasal route were modeled. For the inclusion of intranasal route data, an absorption compartment with an absorption rate constant (ka13) was added to model U4, U5, and U6. In order to investigate the existence of direct nose-to-brain absorption of DB213, a second absorption compartment with absorption rate constant (ka24) was added to model U4, U5, and U6. Furthermore, a brain elimination rate constant k40 was added to model U5 to investigate the elimination of DB213 from brain (Fig. 2). Studies also support that CSF clearance of drugs into venous circulation is one important pathway for drugs to be excreted from the CNS system [35]. Therefore, a compartment representing CSF absorption and elimination was added to model U6 (Fig. 2). The total bioavailability (FTOT) was defined to be the sum of bioavailability of nose-to-systemic circulation (F1) and nose-to-brain (F2). The change of DB213 amount over time in each compartment was defined using Eqs. 410. To monitor the random variability, both additive and proportional residual variability models were investigated. Log-normal distribution of the inter-individual variability (IIV) was assumed, and coefficient of variation (CV) was calculated and monitored as well.

$$ \mathrm{Nose}-\mathrm{to}-\mathrm{systemic}\ \mathrm{circulation}:\mathrm{d}{A}_{\mathrm{a}\mathrm{bsorption}1}/\mathrm{d}t=-{A}_{\mathrm{a}\mathrm{bsorption}1}\times {k}_{\mathrm{a}13} $$
(5)
$$ \mathrm{Nose}-\mathrm{to}-\mathrm{brain}:\mathrm{d}{A}_{\mathrm{a}\mathrm{bsorption}2}/\mathrm{d}t=-{A}_{\mathrm{a}\mathrm{bsorption}2}\times {k}_{\mathrm{a}24} $$
(6)
$$ \mathrm{Central}\ \mathrm{distribution}\ \mathrm{and}\ \mathrm{elimination}:\mathrm{d}{A}_{\mathrm{central}}/\mathrm{d}t={A}_{\mathrm{a}\mathrm{bsorption}1}\times {k}_{\mathrm{a}13}-{A}_{\mathrm{central}}\times \left({k}_{30}+{k}_{35}+{k}_{34}\ \right)+{A}_{\mathrm{central}}\times \left({k}_{53}+{k}_{43}+{k}_{63}\ \right) $$
(7)
$$ \mathrm{Peripheral}\ \mathrm{distribution}\ \mathrm{and}\ \mathrm{elimination}:\mathrm{d}{A}_{\mathrm{peri}}/\mathrm{d}t=-{A}_{\mathrm{peri}}\times {k}_{53}+{A}_{\mathrm{central}}\times {k}_{35} $$
(8)
$$ \mathrm{Brain}\ \mathrm{uptake}\ \mathrm{and}\ \mathrm{elimination}\ \mathrm{for}\ \mathrm{U}5:\mathrm{d}{A}_{\mathrm{brain}}/\mathrm{d}t=-{A}_{\mathrm{brain}}\times \left({k}_{40}+{k}_{43}\right)+{A}_{\mathrm{a}\mathrm{bsorption}2}\times {k}_{\mathrm{a}24+}\ {A}_{\mathrm{central}}\times {k}_{34} $$
(9)
$$ \mathrm{Brain}\ \mathrm{uptake}\ \mathrm{and}\ \mathrm{elimination}\ \mathrm{for}\ \mathrm{U}6:\mathrm{d}{A}_{\mathrm{brain}}/\mathrm{d}t=-{A}_{\mathrm{brain}}\times \left({k}_{43}+{k}_{46}\right)+{A}_{\mathrm{a}\mathrm{bsorption}2}\times {k}_{\mathrm{a}24+}\ {A}_{\mathrm{central}}\times {k}_{34} $$
(10)
$$ \mathrm{CSF}\ \mathrm{distribution}\ \mathrm{and}\ \mathrm{elimination}\ \mathrm{for}\ \mathrm{U}6:\mathrm{d}{A}_{\mathrm{CSF}}/\mathrm{d}t=-{A}_{\mathrm{CSF}}\times {k}_{36}+{A}_{\mathrm{CSF}}\times {k}_{46} $$
(11)

Evaluations of the Pharmacokinetic Models

NONMEM reports an objective function value (OFV) after each run. Since the model hypothesis testing was done using the likelihood ratio test under the assumption that the difference in the – 2 × log likelihood is Χ2 distributed, a decrease in the OFV of at least 3.84 points (p < 0.05) indicates the model with an additional parameter is preferable. Since model U2 and U3 and model U4, U5, and U6 are nested (based on identical data sets), respectively, their OFVs could be used to identify the model that best describes the data.

The coefficient of variation (CV) was used to identify the precision of each pharmacokinetic parameter estimated and was considered acceptable when being lower than 30%. The goodness of fit plots which include the following: (1) observations (DV) versus population predictions (PRED); (2) observations (DV) versus individual predictions (IPRED); (3) absolute individual-weighed residuals (|IWRES|) versus IPRED; (4) weighed population residuals (CWRES) versus independent variable, were obtained by Xpose (version 4.5.3) to assess the fitting of each model.

In addition, the validity of the final pharmacokinetic model was investigated by a visual predictive check (VPC) using the final pharmacokinetic parameter estimates and a simulation of 1000 curves [36]. Since there has been no published DB213 plasma or brain pharmacokinetics data by others, we used another set of data collected by us at a higher dose of 50 mg/kg to validate the developed pharmacokinetic model with pharmacokinetics data from 1 to 10 mg/kg via VPC. The median and the fifth and 95th percentile of the predicted concentration, which represent 90% prediction interval, were calculated. The predicted concentrations were then compared with the observed values.

RESULTS

Unbound Fractions of DB213 in Plasma and Brain

The unbound fraction of DB213 in plasma was close to 100% (97 ± 3 to 102 ± 4%) at concentrations from 0.3 to 300 μg/mL. After 2, 4, or 17 h of incubation, the unbound fraction of DB213 concentration did not show significant difference (p > 0.05), thus, 2 h of incubation was sufficient. The mean value of the unbound fraction of DB213 after 2 h incubation (98%) was used to calculate the unbound DB213 plasma concentration.

The unbound fraction of DB213 in brain homogenate was lower than 10%. Compared with different incubation time, the unbound fraction at 2 h was significantly different from that at 4 and 17 h (p < 0.05). The unbound fractions at 4 and 17 h were similar at the same spiked concentration (p > 0.05). Therefore, it was concluded that 4 h of incubation was sufficient for determining unbound fraction of DB213 in brain homogenate. After 4 h of incubation, when the spiked concentration of DB213 increased from 50 to 5000 ng/g, the unbound fractions of DB213 also showed a trend to increase from 4.7 ± 0.3 to 5.2 ± 0.4%. The mean value of unbound fraction (4.9%) after 4-h incubation was used to calculate the unbound brain concentration of DB213.

Unbound Plasma, CSF, and Brain Concentration Versus Time Profiles of DB213

The unbound plasma concentration versus time profiles of DB213 after 1, 10, and 50 mg/kg via intravenous and intranasal administration are shown in Fig. 3a, d. The pharmacokinetic parameters of plasma are summarized in Table I. After intravenous administration, the back-extrapolated plasma concentration at time 0 (C0) and area under the curve (AUCplasma) increased proportionally with increasing dose from 1 to 50 mg/kg, from 7 ± 1 to 310 ± 57 μg/mL and from 0.18 ± 0.04 to 11.8 ± 0.9 mg min/mL, respectively. However, with the 50-fold increase in intranasally administered dose, peak plasma concentration (Cmax, plasma) of DB213 only increased tenfold from 0.12 ± 0.07 to 1.16 ± 0.13 μg/mL while the time to reach plasma Cmax (Tmax, plasma), ranged from 40 ± 9 to 45 ± 10 min and was comparable between doses. Systemic elimination half-life (t1/2, plasma) was comparable from 87 ± 24 to 92 ± 39 min at different doses after either intravenous or intranasal administration.
Fig. 3

Plasma (a, d), brain (b, e) and CSF (c, f) concentration versus time profiles of DB213 at doses of 1, 10, and 50 mg/kg via (IV, left panel) or intranasal (IN, right panel) administration in SD rats

Table I

Summary of DB213 plasma and brain pharmacokinetic parameters after intravenous (IV) and intranasal (IN) administration at dose of 1, 10, and 50 mg/kg in SD rats (n = 6–8 per group)

 

Plasma

Brain

Route

Dose (mg/kg)

C0(IV)/Cmax(IN), plasma (μg/mL)

Tmax, plasma (min)

t1/2, plasma (min)

AUC0-∞, plasma (mg·min/mL)

Cmax, brain (μg/g)

Tmax, brain (min)

t1/2, brain (min)

AUC0-∞, brain (μg·min/g)

IV

1

7 ± 1

N.A.

78 ± 17

0.18 ± 0.04

0.011 ± 0.004

42 ± 16

66 ± 24

0.78 ± 0.06

10

64 ± 11

N.A.

83 ± 9

2.05 ± 0.13

0.222 ± 0.047

60

60 ± 13

21 ± 5

50

310 ± 57

N.A.

71 ± 16

11.8 ± 0.9

1.22 ± 0.24

60

72 ± 24

133 ± 23

IN

1

0.12 ± 0.07

41 ± 15

92 ± 39

0.016 ± 0.011

0.022 ± 0.007

30

72 ± 27

1.6 ± 0.3

10

0.31 ± 0.08

40 ± 9

87 ± 24

0.068 ± 0.009

0.785 ± 0.127

30

69 ± 22

64.9 ± 9.8

50

1.16 ± 0.13

45 ± 10

90 ± 19

0.139 ± 0.020

5.17 ± 1.47

30

77 ± 11

428 ± 20

The unbound brain concentration versus time profiles are shown in Fig. 3b, e with their relevant pharmacokinetic parameters summarized in Table I. Non-linear DB213 brain pharmacokinetics were observed in rats via both intravenous and intranasal routes. When dose increased from 1 to 10 mg/kg, peak brain concentration (Cmax, brain) and area under the curve (AUCbrain) increased around 20-fold (from 0.011 ± 0.004 to 0.222 ± 0.047 μg/g and from 0.78 ± 0.06 to 21 ± 5 μg min/g) via intravenous administration and Cmax, brain and AUCbrain increased more than 30-fold (from 0.022 ± 0.007 to 0.785 ± 0.127 μg/g and from 1.6 ± 0.3 to 64.9 ± 9.8 μg min/g) via intranasal administration. When dose increased further from 10 to 50 mg/kg, Cmax, brain and AUCbrain increased propotionally after both intravenous and intranasal administrations. In addition, time to reach brain Cmax (Tmax, brain) was different between different routes with 30 min for intranasal and more than 40 min for intravenous administrations. Moreover, brain elimination half-life (t1/2, brain) ranging from 60 ± 13 to 72 ± 24 min was comparable at different doses after either intravenous or intranasal administration. The subsequently calculated Kp,uu,brain (ranging from 0.00038 to 0.00055) also demonstrated comparable values after intravenous administration at three dose levels. However, after the intranasal administration, the Kp,uu,brain value increased dramatically from 0.0049 to 0.064 when the dose increased from 1 to 10 mg/kg and further increased to 0.16 for the 50 mg/kg dose.

The CSF concentration versus time profiles of DB213 at doses of 1, 10, and 50 mg/kg via intravenous and intranasal administration are shown in Fig. 3c, f. After intravenous administration, DB213 in CSF peaked at 30 min (1 mg/kg) or 60 min (10 and 50 mg/kg). The CSF concentrations were comparable with DB213-unbound brain concentration (p > 0.05) at the same sampling time points. However, when DB213 was delivered intranasally, the CSF concentrations kept increasing until 120 min. For intravenous administration, the Kp,uu,CSF was comparable between doses, ranging from 0.00042 to 0.00048 whereas after intranasal administration, the Kp,uu,CSF was much higher than that after intravenous administration and was not comparable at 1, 10, and 50 mg/kg (0.0090, 0.054, and 0.061, respectively).

Development of Pharmacokinetic Model for Intravenous Administration of DB213

Models U1, U2, and U3 were first proposed for analysis of data obtained from intravenous dosing of DB213 (Fig. 2). Table II summarizes the OFVs, parameter estimates, and their CVs. Model U1, which consists of a central compartment and a peripheral compartment, predicted pharmacokinetic parameters with CVs lower than 40%. An additional brain compartment in model U2 improved the CV of CL3. Since models U2 and U3 were nested, increase of OFV of model U3 by 69.124 points showed that an additional brain elimination rate constant (k40) was preferred. Thus, model U3 with an addition of a brain compartment and a brain elimination rate constant was most favorable. However, when evaluating model U3 from the goodness of fit plots (Fig. 4a), the non-horizontal trend line at the bottom left panel ((|IWRES|) vs. IPRED) suggested that the residual error model should be changed from an additive to a proportional model.
Table II

Summary of objective function values (OFVs) and parameter estimates with coefficients of variants and calculated pharmacokinetic parameter for proposed models U1, U2, U3, U4, and U5

 

Model U1

Model U2

Model U3

Model U4

Model U5

Data set

IV plasma at 1, 10, and 50 mg/kg

IV plasma and brain at 1, 10, and 50 mg/kg

IV plasma and brain at 1, 10, and 50 mg/kg

IV and IN plasma and brain at 1 and 10 mg/kg

IV and IN plasma and brain at 1 and 10 mg/kg

OFV

1476.374

513.519

444.213

89.214

− 43.671

Parameter estimate*

CL3

(L·min−1·kg−1)

2.7 (35)

2.9 (23)

3.1 (31)

3.1 (25)

3.1(21)

V3 (mL·kg−1)

320 (29)

290 (26)

290 (41)

280 (38)

280 (26)

Q4

(L·min−1·kg−1)

N.A.

0.88 (26)

1.1 (24)

3.2 (28)

4.6 (32)

V4 (L·kg−1)

N.A.

3.1 (28)

2.4 (25)

2.4 (23)

2.4 (26)

Q5

(L·min−1·kg−1)

0.45 (25)

0.43 (33)

0.24 (29)

0.35 (23)

0.38 (25)

V5 (L·kg−1)

0.83 (36)

0.81 (32)

0.77 (19)

0.85 (21)

0.85 (29)

k a13 (min−1)

N.A.

N.A.

N.A.

18 (24)

18 (21)

k a24 (min−1)

N.A.

N.A.

N.A.

22 (26)

26 (20)

FTOT (%)

N.A.

N.A.

N.A.

7.9

7.6

F1 (%)

N.A.

N.A.

N.A.

2.3 (25)

2.1 (29)

k 40 (min−1)

N.A.

N.A.

3.8 (21)

N.A.

3.8 (27)

Parameter calculated

F2 (%)

N.A.

N.A.

N.A.

5.6

5.5

*Parameter estimates are given with their compartment number in Fig. 2 and are presented as technical efficiency

N.A. not applicable

Fig. 4

Goodness of fit plots for model U3 (a) and U5 (b)

Further Development of Pharmacokinetic Model for Intravenous and Intranasal Administrations of DB213

Models U4, U5, and U6 were further proposed to fit data obtained from both intravenous and intranasal of DB213. Table II summarizes the OFVs, parameter estimates, and their CVs of models U4 and U5. When including CSF data into model U6 (with an additional CSF compartment, Fig. 2), the minimization was terminated. Therefore, there is no parameter estimate reported in Table II for model U6.

After intranasal administration, the addition of two absorption compartments in models in U4 and U5 (Fig. 2) improved the estimation of parameters with CVs lower than 40% (Table II). The fact that V4 was almost three times higher than V5 in models U4 and U5, together with acceptable CVs for both V4 and V5, indicated that the distribution of DB213 in brain was dominantly taking place after being delivered intranasally.

Compared to model U4, the additional brain elimination rate constant (k40) in model U5 (Fig. 2) led to a decrease of OFV with 125.083 points in comparison to that of model U4, which is consistent to what was observed in model U3. The uncertainties in the parameter estimates (CVs) in model U5 were also found to be lower than that in model U4, and the precision of the parameter estimates CL3, V3, and ka24 was significantly improved.

To further evaluate model U5 from the goodness of fit plots (Fig. 4b), the horizontal trend line in the bottom left panel ((|IWRES|) vs. IPRED) indicates that the proportional model could best describe the residual error in both plasma and brain compartments. In the visual predictive check of the final model U5 (Fig. 5), it can be seen that the final model describes the data well with most of the observations falling in the 90% predictive interval. Therefore, model U5 was considered to be the final integrated model.
Fig. 5

Visual predictive check of DB213 plasma and brain concentration versus time profiles at 50 mg/kg via intranasal and prediction from model U5

Since the unbound DB213 concentrations and all pharmacokinetic parameters were accurately predicted (CV values no greater than 35%) in model U5, the absorption rate constant (ka13, ka24) and the bioavailability (FTOT, F1) could also be regarded as accurate. ka24 was slightly higher compared with ka13, explaining a slightly faster absorption DB213 directly from nose-to-brain compared to nose-to-systemic circulation. The bioavailability of nose-to-brain transport was calculated to be 5.3% (FTOTF1), which indicated 72.4% of the total drug absorption was transported directly into the brain.

DISCUSSIONS

The current study not only demonstrated the nose-to-brain transport of DB213 but also quantitatively assess its nose-to-brain transport via intranasal administration. Studies showed that HIV infection occurs in perivascular macrophages, microglia, and astrocytes which can produce and export non-structural proteins and promote inflammation as well as neuronal damage leading to HAND [37, 38]. Therefore, currently DB213 has no specific target in CNS for the treatment of HAND and we determined its whole brain concentrations in the current study. In addition, diseases like HAND which requires long-term treatment, in comparison to intravenous administration, the benefits of intranasal administration are both non-invasive and producing more efficient brain target delivery of DB213. Furthermore, olfactory bulb is one of the proposed CNS entry points for intranasally administered therapeutics [1]. We also investigated brain distribution kinetics of DB213 in rats via both intranasal routes in different regions including olfactory bulb, frontal cortex, medium cortex, posterior cortex, hippocampus, striatum, cerebellum and tissues containing mid brain, thalamus, and hypothalamus. It was found that intranasally delivered DB213 was detectable at all brain regions up to 8-h post-dosing. Although concentration in olfactory bulb was around 40-fold higher than other brain regions, given that the weight of olfactory bulb was around 50 mg, which was only around 3% to a rat brain (around 1.5 g), the contribution of DB213 in olfactory bulb to whole brain uptake is not significant.

In order to investigate the brain uptake of DB213, both Kp,uu,brain and Kp,uu,brain of DB213 were calculated and compared. Fridén et al. [39] reported the Kp,uu,brain of 69 drugs in rat. Among the drugs reported, the lowest Kp,uu,brain value was found for sucrose, being 0.0027 [40]. The Kp,uu,brain of DB213 was even lower than that of sucrose, ranging from 0.00038 to 0.00055, indicating that the ability of passing through the BBB is very poor for DB213. The increased Kp,uu,brain of DB213 of 0.16 when changing the route of administration to intranasal route suggests that it can be a good alternative to deliver drugs with poor BBB permeability to the CNS. In theory, Kp,uu,brain represents the potential of a systemically administered drug to cross the BBB and a change in dose should not influence the value of Kp,uu,brain if the pharmacokinetics is linear. Therefore, comparable Kp,uu,brain values of DB213 after intravenous administration of the three doses were found. Compared with the intravenous route, the Kp,uu,brain of DB213 after intranasal administration was up to 290-fold higher. The significant increased Kp,uu,brain after intranasal route compared to that from intravenous route indicated that DB213 could be transported to the brain not only via the BBB. In order to understand more details on the uptake and disposition of DB213 from the brain, we also obtained CSF concentration versus time profiles [41]. CSF composition is similar to that in plasma except for very low protein content [42]. With the detected unbound fraction of DB213 in plasma being close to 100%, we expect all DB213 to be in unbound form in the CSF. Kp,uu,CSF represents the potential of a systemic administered drug to cross the blood-cerebrospinal fluid barrier (BCSFB). We found that the Kp,uu,CSF of DB213 via the intravenous route did not change as a function of dose, thus indicating linear transport at the BCSFB. The much higher Kp,uu,CSF of DB213 after the intranasal administration indicated that there may be other transport pathways of DB213 into the CSF. In addition to a direct nose-to-brain pathway, studies have indicated a direct nose-to-CSF pathway [10, 43], it was suggested that detection of DB213 in CSF could result from both of its nose-to-brain transport and elimination in the brain tissue.

Stevens et al. has reported direct nose-to-brain transport of remoxipride by pharmacokinetic modeling based on data from microdialysis after an intranasal infusion [22]. In our study, the rats received the intranasal bolus dose by pipetting the drug into the nostrils. The advantage of using intranasal bolus is that it mimics the real situation of intranasal administration in humans, such as nasal drops. However, the potential drawback of such operation is the increase of animal stress and therefore changing the rat system processes (i.e., blood flow in the brain and systemic circulation) compared with intravenous administration. Thus, we compared the plasma pharmacokinetics and brain uptake of rats receiving intranasal vehicle (deionized water) prior to intravenous administration of DB213 water solution with the ones without prior intranasal vehicle administrations and found no statistically significant difference, suggesting that the potential stress caused by intranasal administration could be minimized by the short-term anesthesia prior to drug administration. In addition, the intranasally administered DB213 was dissolved in deionized water, which was also used by several other published studies [32, 33, 34]. However, using water solution for nasal administration could influence the potential drug transport process, which is dependent on the volume of deionized water, duration of treatment, and nature of specific formulation etc. Local ion homeostasis disruption was reported when SD rats were exposed to 3000 μg/m3 PM2.5 3 h/day for 30 consecutive days [24]. In our current study, a bolus dose of 20-μL DB213 deionized water solution was intranasally administered to rat. Since the nasal cavity volume of a male SD rat (200 g) is around 170 μL, the administered DB213 solution is just about 12% of its nasal cavity. We also found that the retention time of deionized water formulation in nasal cavity is less than a minute. Moreover, no statistical difference was noticed after comparing the pharmacokinetics and brain uptake of intranasal administered DB213 in deionized water solution at 10 mg/kg versus that in saline (data not shown). Therefore, considering limited volume and time of contact with the nasal mucosa for DB213 deionized water solution, we believe that the risk of causing local ion homeostasis is minimal under our current experimental conditions.

Intranasal administration instead of intravenous administration of DB213 was found to follow non-linear pharmacokinetics (based on AUC0-∞, plasma). Thus, we use data set from intravenous route to build our base model (model U1) and to generate major PK parameters including CL3, V3, and V5. Based on model U1, we gradually add different data set and compartment in order to illustrate direct nose-to-brain transport of DB213, leading to model U5. From VPC, we could see there are still observations fell out of 90% predictive interval, which could be due to the partially non-linear PK data set used in building model U5. In addition, since each rat only represented one point in the brain concentration versus time profile, no inter-individual variability could be calculated. The drawback of our experimental design could be a large variability in data obtained which could also influence the accuracy of the pharmacokinetic parameter estimates [25].

In our study, a multi-compartment pharmacokinetic model was developed to quantitatively assess the amount of DB213 that was delivered directly from nose to brain. Based on the above identified pharmacokinetic model U5, the non-linear plasma pharmacokinetics after intranasal administration could be due to its two absorption pathways while the non-linear brain pharmacokinetics after both intravenous and intranasal administrations may result from the brain tissue bindings. In addition, the quick equilibrium as well as elimination of DB213 in plasma and brain suggests the possibility of transporter involvement. We studied the possible influx and efflux transporters on brain slices model and transfected cell models (MDCK II-WT, MDCK II-MRP2, MDCK II MRP3, MDCK II-MDR1, HEK-293 MOCK, and HEK-293 OAT3). DB213 was only confirmed to be a substrate of amino acid transporter (data not shown), suggesting DB213 could be transported in to brain cells by amino acid transporter which may be one of the explanations of its quick equilibrium between plasma and brain.

CONCLUSION

Compared to intravenous administration, intranasal administration showed better potential to deliver HIV-1 replication inhibitor DB213 to the brain with an up to 290-fold higher partitioning into brain. Based on that, a multi-compartment pharmacokinetic model with two absorption compartments (nose-systemic circulation, nose-brain) was developed and further demonstrated a direct nose-to-brain transport of HIV-1 replication inhibitor DB213.

Notes

Acknowledgments

This work was generously supported by the Lui Che Woo Institute of Innovative Medicine BRAIN Initiative (Project Number 8303404) and Gerald Choa Neuroscience Centre (Project Number 7105306), Faculty of Medicine, The Chinese University of Hong Kong. The authors are grateful to Prof. Margareta Hammarlund-Udenaes from Translational PKPD Research Group, Department of Pharmaceutical Biosciences, Uppsala University, Uppsala, Sweden, for her valuable suggestions to the data analyses and manuscript.

References

  1. 1.
    Lochhead JJ, Thorne RG. Intranasal delivery of biologics to the central nervous system. Adv Drug Deliv Rev. 2012;64(7):614–28.  https://doi.org/10.1016/j.addr.2011.11.002.CrossRefPubMedGoogle Scholar
  2. 2.
    Qian S, Wo SK, Zuo Z. Pharmacokinetics and brain dispositions of tacrine and its major bioactive monohydroxylated metabolites in rats. J Pharm Biomed Anal. 2012;61:57–63.  https://doi.org/10.1016/j.jpba.2011.11.025.CrossRefPubMedGoogle Scholar
  3. 3.
    Freiherr J, Hallschmid M, Frey WH II, Brünner YF, Chapman CD, Hölscher C, et al. Intranasal insulin as a treatment for Alzheimer’s disease: a review of basic research and clinical evidence. CNS drugs. 2013;27(7):505–14.  https://doi.org/10.1007/s40263-013-0076-8.CrossRefPubMedPubMedCentralGoogle Scholar
  4. 4.
    Thorne R, Pronk G, Padmanabhan V, Frey W II. Delivery of insulin-like growth factor-I to the rat brain and spinal cord along olfactory and trigeminal pathways following intranasal administration. Neuroscience. 2004;127(2):481–96.  https://doi.org/10.1016/j.neuroscience.2004.05.029.CrossRefPubMedGoogle Scholar
  5. 5.
    Draghia R, Caillaud C, Manicom R, Pavirani A, Kahn A, Poenaru L. Gene delivery into the central nervous system by nasal instillation in rats. Gene Ther. 1995;2(6):418–23.PubMedGoogle Scholar
  6. 6.
    Danielyan L, Schäfer R, von Ameln-Mayerhofer A, Buadze M, Geisler J, Klopfer T, et al. Intranasal delivery of cells to the brain. Eur J Cell Biol. 2009;88(6):315–24.  https://doi.org/10.1016/j.ejcb.2009.02.001.CrossRefPubMedGoogle Scholar
  7. 7.
    Dahlin M, Bergman U, Jansson B, Björk E, Brittebo E. Transfer of dopamine in the olfactory pathway following nasal administration in mice. Pharm Res. 2000;17(6):737–42.  https://doi.org/10.1023/A:1007542618378.CrossRefPubMedGoogle Scholar
  8. 8.
    Wall A, Kågedal M, Bergström M, Jacobsson E, Nilsson D, Antoni G, et al. Distribution of zolmitriptan into the CNS in healthy volunteers. Drugs in R & D. 2005;6(3):139–47.  https://doi.org/10.2165/00126839-200506030-00002.CrossRefGoogle Scholar
  9. 9.
    Walker LC, Price DL, Voytko ML, Schenk DB. Labeling of cerebral amyloid in vivo with a monoclonal antibody. J Neuropathol Exp Neurol. 1994;53(4):377–83.  https://doi.org/10.1097/00005072-199407000-00009.CrossRefPubMedGoogle Scholar
  10. 10.
    Mittal D, Ali A, Md S, Baboota S, Sahni JK, Ali J. Insights into direct nose to brain delivery: current status and future perspective. Drug Deliv. 2014;21(2):75–86.  https://doi.org/10.3109/10717544.2013.838713.CrossRefPubMedGoogle Scholar
  11. 11.
    Pardridge WM. The blood-brain barrier: bottleneck in brain drug development. NeuroRx. 2005;2(1):3–14.  https://doi.org/10.1602/neurorx.2.1.3.CrossRefPubMedPubMedCentralGoogle Scholar
  12. 12.
    Kugel D, Kochs G, Obojes K, Roth J, Kobinger GP, Kobasa D, et al. Intranasal administration of alpha interferon reduces seasonal influenza a virus morbidity in ferrets. J Virol. 2009;83(8):3843–51.  https://doi.org/10.1128/JVI.02453-08.CrossRefPubMedPubMedCentralGoogle Scholar
  13. 13.
    Eriksson C, Bergman U, Franzen A, Sjöblom M, Brittebo E. Transfer of some carboxylic acids in the olfactory system following intranasal administration. J Drug Target. 1999;7(2):131–42.  https://doi.org/10.3109/10611869909085497.CrossRefPubMedGoogle Scholar
  14. 14.
    Westin UE, Boström E, Gråsjö J, Hammarlund-Udenaes M, Björk E. Direct nose-to-brain transfer of morphine after nasal administration to rats. Pharm Res. 2006;23(3):565–72.  https://doi.org/10.1007/s11095-006-9534-z.CrossRefPubMedGoogle Scholar
  15. 15.
    Kozlovskaya L, Abou-Kaoud M, Stepensky D. Quantitative analysis of drug delivery to the brain via nasal route. J Control Release. 2014;189:133–40.  https://doi.org/10.1016/j.jconrel.2014.06.053.CrossRefPubMedGoogle Scholar
  16. 16.
    Bagger MA, Bechgaard EA. Microdialysis model to examine nasal drug delivery and olfactory absorption in rats using lidocaine hydrochloride as a model drug. Int J Pharm. 2004;269(2):311–22.  https://doi.org/10.1016/j.ijpharm.2003.09.017.CrossRefPubMedGoogle Scholar
  17. 17.
    Bagger MA, Bechgaard E. The potential of nasal application for delivery to the central brain—a microdialysis study of fluorescein in rats. Eur J Pharm Sci. 2004;21(2-3):235–42.  https://doi.org/10.1016/j.ejps.2003.10.012.CrossRefPubMedGoogle Scholar
  18. 18.
    Charlton ST, Whetstone J, Fayinka ST, Read KD, Illum L, Davis SS. Evaluation of direct transport pathways of glycine receptor antagonists and an angiotensin antagonist from the nasal cavity to the central nervous system in the rat model. Pharm Res. 2008;25(7):1531–43.  https://doi.org/10.1007/s11095-008-9550-2.CrossRefPubMedGoogle Scholar
  19. 19.
    Uchida M, Katoh T, Mori M, Maeno T, Ohtake K, Kobayashi J, et al. Intranasal administration of milnacipran in rats: evaluation of the transport of drugs to the systemic circulation and central nervous system and the pharmacological effect. Biol Pharm Bull. 2011;34(5):740–7.  https://doi.org/10.1248/bpb.34.740.CrossRefPubMedGoogle Scholar
  20. 20.
    Hammarlund-Udenaes M. Active-site concentrations of chemicals–are they a better predictor of effect than plasma/organ/tissue concentrations? Basic Clin Pharmacol Toxicol. 2010;106(3):215–20.  https://doi.org/10.1111/j.1742-7843.2009.00517.x.CrossRefPubMedGoogle Scholar
  21. 21.
    Kalvass JC, Olson ER, Cassidy MP, Selley DE, Pollack GM. Pharmacokinetics and pharmacodynamics of seven opioids in P-glycoprotein-competent mice: assessment of unbound brain EC50,u and correlation of in vitro, preclinical, and clinical data. J Pharmacol Exp Ther. 2007;323(1):346–55.  https://doi.org/10.1124/jpet.107.119560.CrossRefPubMedGoogle Scholar
  22. 22.
    Stevens J, Ploeger BA, van der Graaf PH, Danhof M, de Lange EC. Systemic and direct nose-to-brain transport pharmacokinetic model for remoxipride after intravenous and intranasal administration. Drug Metab Dispos. 2011;39(12):2275–82.  https://doi.org/10.1124/dmd.111.040782.CrossRefPubMedGoogle Scholar
  23. 23.
    Wang Q, Zhang Y, Qian S, Peng S, Zhang Q, Wong C, et al. Pharmacokinetics and brain uptake of HIV-1 replication inhibitor DB213 in Sprague-Dawley rats. J Pharm Biomed Anal. 2016;125:41–7.  https://doi.org/10.1016/j.jpba.2016.03.025.CrossRefPubMedGoogle Scholar
  24. 24.
    Guo Z, Hong Z, Dong W, Deng C, Zhao R, Xu J, et al. PM2. 5-induced oxidative stress and mitochondrial damage in the nasal mucosa of rats. Int J Environ Res Public Health. 2017;14(2):134.  https://doi.org/10.3390/ijerph14020134.
  25. 25.
    Yamamoto Y, Välitalo PA, van den Berg D, Hartman R, van den Brink W, Wong YC, et al. A generic multi-compartmental CNS distribution model structure for 9 drugs allows prediction of human brain target site concentrations. Pharm Res. 2016:1–19.Google Scholar
  26. 26.
    Wong YC, Qian S, Zuo Z. Pharmacokinetic comparison between the long-term anesthetized, short-term anesthetized and conscious rat models in nasal drug delivery. Pharm Res. 2014;31(8):2107–23.  https://doi.org/10.1007/s11095-014-1312-8.CrossRefPubMedGoogle Scholar
  27. 27.
    Lee HB, Blaufox MD. Blood volume in the rat. J Nucl Med. 1985;26(1):72–6.PubMedGoogle Scholar
  28. 28.
    Stypinski D, Wiebe L, Tam Y, Mercer J, McEwan A. Effects of methoxyflurane anesthesia on the pharmacokinetics of 125 I-IAZA in sprague-dawley rats. Nucl Med Biol. 1999;26(8):959–65.  https://doi.org/10.1016/S0969-8051(99)00071-2.CrossRefPubMedGoogle Scholar
  29. 29.
    Kalvass JC, Maurer TS, Pollack GM. Use of plasma and brain unbound fractions to assess the extent of brain distribution of 34 drugs: comparison of unbound concentration ratios to in vivo p-glycoprotein efflux ratios. Drug Metab Dispos. 2007;35(4):660–6.  https://doi.org/10.1124/dmd.106.012294.CrossRefPubMedGoogle Scholar
  30. 30.
    Cory Kalvass J, Maurer TS. Influence of nonspecific brain and plasma binding on CNS exposure: implications for rational drug discovery. Biopharm Drug Dispos. 2002;23(8):327–38.  https://doi.org/10.1002/bdd.325.CrossRefPubMedGoogle Scholar
  31. 31.
    Brouwer E, Verweij J, De Bruijn P, Loos WJ, Pillay M, Buijs D, et al. Measurement of fraction unbound paclitaxel in human plasma. Drug Metab Dispos. 2000;28(10):1141–5.PubMedGoogle Scholar
  32. 32.
    Friden M, Gupta A, Antonsson M, Bredberg U, Hammarlund-Udenaes M. In vitro methods for estimating unbound drug concentrations in the brain interstitial and intracellular fluids. Drug Metab Dispos. 2007;35(9):1711–9.  https://doi.org/10.1124/dmd.107.015222.CrossRefPubMedGoogle Scholar
  33. 33.
    Modi ME, Majchrzak MJ, Fonseca KR, Doran A, Osgood S, Vanase-Frawley M, et al. Peripheral administration of a long-acting peptide oxytocin receptor agonist inhibits fear-induced freezing. J Pharmacol Exp Ther. 2016;358(2):164–72.  https://doi.org/10.1124/jpet.116.232702.CrossRefPubMedPubMedCentralGoogle Scholar
  34. 34.
    Hammarlund-Udenaes M. Pharmacokinetic concepts in brain drug delivery. In: Hammarlund-Udenaes M, de Lange E, Thorne RG, editors. Drug Delivery to the Brain. New York: Springer. 2014:127–161.Google Scholar
  35. 35.
    Johanson CE, Duncan JA, Klinge PM, Brinker T, Stopa EG, Silverberg GD. Multiplicity of cerebrospinal fluid functions: new challenges in health and disease. Cerebrospinal Fluid Res. 2008;5(1):10.  https://doi.org/10.1186/1743-8454-5-10.CrossRefPubMedPubMedCentralGoogle Scholar
  36. 36.
    Bergstrand M, Hooker AC, Wallin JE, Karlsson MO. Prediction-corrected visual predictive checks for diagnosing nonlinear mixed-effects models. AAPS J. 2011;13(2):143–51.  https://doi.org/10.1208/s12248-011-9255-z.CrossRefPubMedPubMedCentralGoogle Scholar
  37. 37.
    Lindl KA, Marks DR, Kolson DL, Jordan-Sciutto KL. HIV-associated neurocognitive disorder: pathogenesis and therapeutic opportunities. J NeuroImmune Pharmacol. 2010;5(3):294–309.  https://doi.org/10.1007/s11481-010-9205-z.CrossRefPubMedPubMedCentralGoogle Scholar
  38. 38.
    Nightingale S, Winston A, Letendre S, Michael BD, McArthur JC, Khoo S, et al. Controversies in HIV-associated neurocognitive disorders. Lancet Neurol. 2014;13(11):1139–51.  https://doi.org/10.1016/S1474-4422(14)70137-1.CrossRefPubMedPubMedCentralGoogle Scholar
  39. 39.
    Fridén M, Winiwarter S, Jerndal G, Bengtsson O, Wan H, Bredberg U, et al. Structure− brain exposure relationships in rat and human using a novel data set of unbound drug concentrations in brain interstitial and cerebrospinal fluids. J Med Chem. 2009;52(20):6233–43.  https://doi.org/10.1021/jm901036q.CrossRefPubMedGoogle Scholar
  40. 40.
    Terasaki T, Deguchi Y, Sato H, Hirai KI, Tsuji A. In vivo transport of a dynorphin-like analgesic peptide, E-2078, through the blood–brain barrier: an application of brain microdialysis. Pharm Res. 1991;8(7):815–20.  https://doi.org/10.1023/A:1015882924470.
  41. 41.
    de Lange EC. Utility of CSF in translational neuroscience. J Pharmacokinet Pharmacodyn. 2013;40(3):315–26.  https://doi.org/10.1007/s10928-013-9301-9.CrossRefPubMedPubMedCentralGoogle Scholar
  42. 42.
    Somjen GG. Ions and water in the brain. In: Somjen GG, editor. Ions in the brain: normal function, seizures, and stroke. Oxford: Oxford University Press. 2004:6–7.Google Scholar
  43. 43.
    Merkus FWHM, van den Berg MP. Can nasal drug delivery bypass the blood-brain barrier? Drugs R D. 2007;8(3):133–44.  https://doi.org/10.2165/00126839-200708030-00001.

Copyright information

© American Association of Pharmaceutical Scientists 2017

Authors and Affiliations

  • Qianwen Wang
    • 1
  • Yufeng Zhang
    • 1
  • Chun-Ho Wong
    • 2
  • H.Y. Edwin Chan
    • 2
    • 3
  • Zhong Zuo
    • 1
  1. 1.School of PharmacyThe Chinese University of Hong KongHong KongPeople’s Republic of China
  2. 2.School of Life SciencesThe Chinese University of Hong KongHong KongPeople’s Republic of China
  3. 3.Gerald Choa Neuroscience CentreThe Chinese University of Hong KongHong KongPeople’s Republic of China

Personalised recommendations