Based on ibrutinib pharmacokinetics and potential sensitivity towards CYP3A4‐mediated drug–drug interactions (DDIs), a physiologically based pharmacokinetic approach was developed to mechanistically describe DDI with various CYP3A4 perpetrators in healthy men under fasting conditions. These models were verified using clinical data for ketoconazole (strong CYP3A4 inhibitor) and used to prospectively predict and confirm the inducing effect of rifampin (strong CYP3A4 inducer); DDIs with mild (fluvoxamine, azithromycin) and moderate inhibitors (diltiazem, voriconazole, clarithromycin, itraconazole, erythromycin), and moderate (efavirenz) and strong CYP3A4 inducers (carbamazepine), were also predicted. Ketoconazole increased ibrutinib area under the curve (AUC) by 24‐fold, while rifampin decreased ibrutinib AUC by 10‐fold; coadministration of ibrutinib with strong inhibitors or inducers should be avoided. The ibrutinib dose should be reduced to 140 mg (quarter of maximal prescribed dose) when coadministered with moderate CYP3A4 inhibitors so that exposures remain within observed ranges at therapeutic doses. Thus, dose recommendations for CYP3A4 perpetrator use during ibrutinib treatment were developed and approved for labeling.

SUMMARY A novel CNS drug candidate demonstrated useful first-in-human safety properties but was challenged by a short half-life and high Cmax-Cmin plasma level variations. A sustained released formulation was assessed. Drug absorption in the distal bowel and ascending colon was found to be equivalent to that in the proximal small intestine. A swellable matrix tablet based on HPMC was generated and optimized such that drug release over a 6 or 12 h time frame could be achieve in vitro. Clinical assessments of these formulation suggested useful biopharmaceutical properties with prolonged absorption. INTRODUCTION In cases of drugs with short plasma half-lives or narrow therapeutic windows, the ability to sustain drug release subsequent to oral dosing could not only optimize the therapeutic index of the drug but also improve patient compliance. A number of dosage form concepts have been designed to address these needs including matrixforming tablets in which drug release is controlled by diffusion and erosion. These systems can often provide for the release characteristic required, are simple to manufacture and scale, are often very cost-effective and make use of well characterized, generally regarded as safe (GRAS) excipients. These systems can retard drug release delaying it to more distal portions of the GI tract meaning that one prerequisite for a useful system is that the drug gives good permeability in the distal small bowel and colon. Compound A is a BCS Class I drug candidate which has been successfully tested in man both as a solution and simple (immediate release) tablet. Biological half-life data suggested that a sustained release formulation might add value. In order to assess the feasibility of a controlled release dosage form, a study was completed wherein drug uptake was assessed when drug was administered in the distal small intestine or ascending colon. HPMC-based matrix tablet were then designed and prototypes generated using in vitro drug release. Selected matrix tablets were then assessed in a clinical trial to evaluate human pharmacokinetics. EXPERIMENTAL METHODS Compound A was obtained from Janssen Pharmaceutica, Beerse, Belgium and was characterized with a purity >98%. Matrix formulation were generated using HPMC at two different molecular weights (6500 and 100000), HPC and drug dried waxy maize starch. Tablets were designed to contain 40 mg eq of Compound A with a total tablet weight of approximately 350 mg. Four dissolution paradigms were applied to test the concepts generate all of which involved a USP II apparatus, thermostated to 37 o C and with a paddle speed of 75 rpm and a total media volume of 900 mL. The four media tested included (one phase) 0.01 N HCl, 0.05 M pH 6.8 phosphate buffer, (two phase) (1 h) 0,01 N HCl then (23 h) pH 6.8 phosphate buffer and (1 h) 0.01 N HCl (250 mL) and (23 h) Fessif (1000 mL). Regional intestinal absorption studies were completed in man using the Enterion® capsule in which the drug was administered as a solution in aqueous 2hydroxypropyl-cyclodextrin to either the distal small intestine or ascending colon and compared to an oral solution delivered po. Subsequently, matrix tablets based on a fast and slow release pattern were also assessed in man compared with a simple immediate release (IR) tablet. The matrix tablets contained 40 mg eq of Compound A and the IR tablets 10 mg eq (two IR tablets were dosed in the clinical evaluation). All studies were GCP-compliant with full informed consent and conducted under relevant national and international law and guidelines. The clinical investigations were completed as open label trials using 12 and 24 healthy subjects (male, Caucasians between age 18 and 55 years and within 20% of their ideal body weight) for the regional absorption and matrix tablet studies, repsectively. Each subject received a test formulation as described (either fasted or after a high fat breakfast). Blood was sampled at 0, 1, 2, 3, 4, 5, 6, 8 and 24 h after dosing and analysed using a fully validated HPLC analytical method. Pharmacokinetic parameters including Tmax, Cmax and AUC24h were determined by standard model-independent methods using actual times of blood sampling. RESULTS AND DISCUSSION Drug absorption was assessed using the Enterion® capsule 1 which was loaded with a solution of Compound A in a cyclodextrin vehicle with the drug administered either to the distal small intestine (at doses of 2 and 10 mg) or in the ascending colon (at a dose of 10 mg). This was compared to an oral dose of the 10 mg cyclodextrin solution. The AUC of the drug administered to the distal small bowel and ascending colon was similar to that of the orally dosed medication. The Cmax was reduced by approximately 30% when the oral solution was compared to the colon however Cmax were comparable when oral and distal small bowel dosing were assessed. These data suggested that a controlled release dosage form may be feasible. Controlled release tablet was considered with the following properties including a total dose of 40 mg and a drug release window of between 6 and 12 h. A significant design concern related to the sensitivity of solubility on pH which decreased rapidly with increasing pH. Based on drug release from a swellable matrix, this may negatively impact the release mechanism as the tablet transitions from the stomach to the intestine. To proactively address this, solubility differences as a function of pH were attenuated by adding 2-hydroxypropyl-cyclodextrin to the matrix such that diffusion would continue to drive drug release over the life time of the matrix. 2 The excipient space of the tablet therefore included the active API, HP CD and three polymers including HPMC at various molecular weights/viscosity grads, HPC and drum dried waxy maize starch. Screening experiments found that the optimal amount of the matrix-forming polymers was ~30% with little change in drug release as this percent changes from 20 to 40%. In the case of HP CD, concentrations between 20 and 50% yielded similar release profiles however excluding the cyclodextrin generated systems that poorly released drug at pH 6.8 and which were also highly dependent on ionic strength of the release media. Two prototype, matrix-based tablets were therefore designed based on high and low viscosity HPMC. A “fast” releasing matrix was prepared based on HPMC 6500 mPa.s and a slow releasing tablet based on 100000 mPa.s material. Four dissolution assessments were completed with the two matrix tablets as is illustrated in Figures 1 and 2. Figure 1. Drug Release Dynamics for a “Fast” system in Four Media including 0.01N HCl (green), 0,05 M pH 6.8 Phosphate Buffer (blue), 1h 0.01 N HCl + 23 h 0.05 M, pH 6.8 Phosphate Buffer (red) and 1 h 0.01 N HCl + 23 h Fessif (black, dotted line).. Figure 2. Drug Release Dynamics for a “Slow” system in Four Media including 0.01N HCl (green), 0,05 M pH 6.8 Phosphate Buffer (blue), 1h 0.01 N HCl + 23 h 0.05 M, pH 6.8 Phosphate Buffer (red) and 1 h 0.01 N HCl + 23 h Fessif (black, dotted line). These tablets were then examined in a clinical assessment under fasted conditions and compared to the simple IR tablets. Two IR tablets (10 mg eq. of Compound A per tablet) versus 40 mg eq in the controlled release tablets were evaluated (Table I). Table I. Pharmacokinetic Parameters in Man (Fasted) Associated with Compound A Dosing as Either an Immediate Release Tablet, a Slow Matrix and a Fast Matrix Tablet

Aims: Drug release from paliperidone ER (P-ER) occurs in gradual way with ascending pharmacokinetic profile reaching peak at 24h and small peak-to-trough fluctuations at steady-state compared to immediate-release (IR) risperidone (RIS-IR). Methods: Comparison of D2-occupancies and plasma-concentrations of P-ER, P-IR and/or RIS-IR (active moiety) in four studies with schizophrenic patients and healthy subjects: 1: n=113 patients, double-blind, randomized to (a) placebo Day 1, P-ER 12mg Days 2–6; (b) 12mg P-ER Days 1–6; (c) 2mg RIS-IR Day 1, 4mg Days 2–6; 2: Positron Emission Tomography (PET) in 3 healthy volunteers following single dose of P-IR 1mg. 3: PET in 4 healthy volunteers following single oral dose P-ER 6mg. 4: PET in 8 patients following once-daily RIS-IR 6mg. Results: Concentrations of paliperidone and risperidone active moiety increased dose proportionally across the dose range studied. RIS active moiety mean peak (trough) plasma concentrations were between 15 (8) ng/mLfor 1mg and 107 (40) ng/mL for 6mg (fluctuation index (FI):125%). FI of P-ER was lower (38%) with mean plasma concentrations between ˜8ng/mL (3mg) and ˜43ng/mL (15mg). D2-occupancy of RIS-IR fluctuated between 40–70%(1mg), 60–85%(3mg), 80–90%(4mg) and 85–93% (6mg). D2-occupancies with P-ER were fairly constant: e.g 75%(6mg). Conclusions: Similar occupancies with lower plasma concentrations while showing more stable D2-binding profile achievable with P-ER compared to RIS-IR.

BACKGROUND Despite widespread use of risperidone for disruptive behavior disorders in children and adolescents no pharmacokinetic (PK) data for this age group are available. The objective of the study was to determine the PK of RIS and its active metabolites (+)- and (−)- 9-hydroxy-RIS in pediatric patients. METHODS Patients participated in an open-label, Phase-1 trial. PK samples were drawn at steady-state at pre-dose, 1, 2, 4 and 7h post dose. RIS and (+)- and (−)- 9-hydroxy-RIS plasma concentrations were analyzed by LC-MS/MS. PK data analysis was performed using standard non-compartmental methods (WinNonlin Professional). CYP2D6 genotyping was performed according to published methods. RESULTS 19 patients, mean age 9.9y (range 4.2–15.9y) weighing 42.7 ± 17.3 Kg, were enrolled. Mean RIS Cmax concentrations were 15.9 ± 22.2 μg/L (range 2.4–104 μg/L) and pre-dose troughs 8.3 ± 20.4 μg/L (range 0.4–78.4 μg/L). Hydroxy-metabolites maximum concentrations were 13.24 ± 9.98 μg/L for (+)-9-OH-RIS (at 2h) and 4.64 ± 2.87 μg/L for (−)-9-OH-RIS (at 4h), respectively. Pre-dose concentrations were 6.6 ± 4.3 μg/L for (+)-OH-RIS and 3.81 ± 2.39 μg/L for (−)-OH-RIS, respectively. PK see Table 1. Parameter Mean SD Range Clearance/F (L/h/Kg) 0.649 0.629 0.033 2.195 Volume/F (L/Kg) 1.60 1.09 0.37 4.89 Half-life (h) 2.97 2.28 1.08 7.69 AUC-inf (μg*h/L) 125.61 308.02 8.92 1378.43 CONCLUSIONS Mean RIS oral clearance was higher than reported in adults. Large inter-individual variability in PK profiles and dose normalized AUCs was noted. PM status resulted in 8–10 fold higher RIS concentrations and AUC. These population PK-PG data will aid in the better design of age appropriate dosing regimens. Clinical Pharmacology & Therapeutics (2005) 79, P73–P73; doi: 10.1016/j.clpt.2005.12.264

Several recent studies have demonstrated that the combination of antipsychotics, including risperidone, with a mood stabilizer, e.g. lithium, is effective and well tolerated in patients with bipolar disorders and mania [4,13,14,15,18] including some pediatric cases [11]. Since the first reports by Böszörményi [3] and Cohen and Cohen [7], a series of case reports and metaanalyses have been published describing neurotoxic effects during the course of treatment with lithium in combination with different antipsychotics. For risperidone, single case reports on neurotoxic interactions with lithium have also been published [6,10]. Because the neurotoxic effects described have been attributed to a reciprocal pharmacokinetic influence, the pharmacokinetics of lithium combined with psychotropics [1, 2, 5,12,16] and non-psychotropics [17] have been studied. The pharmacokinetics of lithium in combination with risperidone has not been reported to date. We have now investigated the specific pharmacokinetics of a combined lithium and risperidone therapy in 13 psychiatric patients.

OBJECTIVES Assess pharmacokinetics, dopamine D2 and serotonin 5-HT2A receptor occupancy of paliperidone immediate-release 1mg (paliperidone IR; Study 1) and paliperidone extended-release tablets 6mg (paliperidone ER; Study 2). METHODS Blood samples were collected pre-dose and ≤24h (Study 1) and ≤48h (Study 2). Striatal D2 receptor binding was measured using PET and 11C-raclopride pre-dose and post-dose at 2.5h (Study 1) and at 22h and 46h (predicted Cmax; Study 2). Frontal cortex 5-HT2A receptor occupancy (Study 1) was assessed using[11C]M100,907 ≥1 week pre-dose and 4.5h post-dose (Cmax). The apparent dissociation constant (KDapp=plasma concentration at which 50% of target receptor is occupied) was estimated using an Emax model. RESULTS Formulation differences were reflected in tmax and Cmax. Paliperidone ER 6mg corresponds to median D2 occupancy of 64% at 22h post-dose. Using the KDapp, plasma concentrations corresponding to 70–80% D2 occupancy were estimated: 15–25ng/mL (Study 1); 10–17ng/mL (Study 2). Using pooled data from both studies the in vivo KDapp is estimated to be 4.9ng/mL. CONCLUSION Paliperidone occupies central D2 and 5HT2A receptors. D2 receptor occupancy with paliperidone ER suggests 6mg will be an effective dose in the treatment of schizophrenia. (See Table) Measure Paliperidone IR 1mg*. Study 1 (n=3) Paliperidone ER 6mg*. Study 2 (n=4) Median Cmax (ng/mL) (range) 6.02 (5.34–6.14) 11.3 (7.73–16.5) Median tmax (h) (range) 4.2 (4.1–8.1) 24.1 (23.1–29.0) Median % D2 receptor occupancy (range) 48 (35–51)[2.5h post-dose] 64 (56–79)[22h post-dose]; 53 (40–62)[46h post-dose] Calculated KDapp for D2-receptor occupancy (ng/mL) 6.4 4.4 Median % 5-HT2A-receptor occupancy (range) 65 (65–71)[4.5h post-dose] Not measured Corresponding plasma concentration range (mg/mL) 5.1–6.0[4.0h post-dose] Not measured a Initial studies have shown that paliperidone ER has a bioavailability of approximately 33% of that of paliperidone IR. Clinical Pharmacology & Therapeutics (2005) 79, P74–P74; doi: 10.1016/j.clpt.2005.12.265

Two open-label studies assessed pharmacokinetics, dopamine D2 and serotonin 5-HT2A receptor occupancy for paliperidone immediate-release 1mg (IR; Study 1) and extended-release 6mg (ER; Study 2). Blood was collected pre-dose and ≤24h (Study 1, n=3) and ≤48h (Study 2, n=4). Striatal D2 receptor binding measured using PET and 11C-raclopride pre-dose and post-dose at 2.5h (Study 1) and at 22h and 46h (predicted Cmax; Study 2). Frontal cortex 5-HT2A receptor occupancy (Study 1) assessed using [11C]M100,907 ≥1 week pre-dose and 4.5h post-dose (Cmax). The apparent dissociation constant (KDapp=plasma concentration at which 50% of target receptor is occupied) estimated via Emax model. Formulation differences reflected in tmax (median 4.2h for paliperidone IR 1mg vs. 24.1h for paliperidone ER 6mg) and Cmax (median 6.02ng/mL vs. 11.3ng/mL). Paliperidone ER 6mg had a median D2 occupancy of 64% at 22h post-dose (for paliperidone IR 1mg 48% at 2.5h post-dose). Calculated KDapp from pooled data was 4.9ng/mL. Using the KDapp, calculated plasma concentrations for 70–80%, D2 occupancy is 11–20/mL. The median 5HT2A occupancy for paliperidone IR 1mg was 65% at 4.5h post-dose, with a corresponding plasma concentration of 5.1–6.0mg/Ml, which indicates that the 5HT2A occupancy is substantially higher than D2at therapeutic doses. Paliperidone highly occupies central D2 and 5HT2A receptors. The D2 occupancy with paliperidone ER suggests that 6mg will be an effective dose in schizophrenia treatment.