PDE-4 ... - ACS Publications

Jun 14, 2017 - To date, only two PDE4 inhibitors have reached marketing authorization, and only one of them, roflumilast (Daxas EU, Daliresp US), is b...
3 downloads 11 Views 3MB Size
Drug Annotation pubs.acs.org/jmc

Discovery of N‑{4-[5-(4-Fluorophenyl)-3-methyl-2methylsulfanyl‑3H‑imidazol-4-yl]-pyridin-2-yl}-acetamide (CBS-3595), a Dual p38α MAPK/PDE‑4 Inhibitor with Activity against TNFαRelated Diseases Wolfgang Albrecht,† Anke Unger,† Silke M. Bauer,‡ and Stefan A. Laufer*,‡ †

c-a-i-r biosciences GmbH, Alfred-Mendler Weg 25/1, D-89075 Ulm, Germany Eberhard Karls University Tuebingen, Department Pharmacy & Biochemistry, Pharmaceutical and Medicinal Chemistry, Auf der Morgenstelle 8, D-72076 Tuebingen, Germany



S Supporting Information *

ABSTRACT: The anti-inflammatory potential of p38 mitogen-activated protein kinase (MAPK) inhibitors was coincidentally expanded to a dual inhibition of p38α MAPK and phosphodiesterase 4 (PDE4), and the potential benefits arising from the blockage of both inflammation-related enzymes were thoroughly investigated. The most promising compound, CBS-3595 (1), was successively evaluated in in vitro experiments as well as in ex vivo and in vivo preclinical studies after administration of 1 to rodents, dogs, and monkeys. The resulting data clearly indicated a potent suppression of tumor necrosis factor alpha release. For reconfirming the findings of the animal studies when administering 1 to healthy human volunteers, a phase I clinical trial was conducted. Apart from further information regarding the pharmacokinetic and pharmacodynamic characteristics of 1, it was demonstrated that dual inhibition of p38α MAPK and PDE4 is able to synergistically attenuate the excessive anti-inflammatory response.



INTRODUCTION

Affecting approximately 1% of the adult population, RA is the consequence of a complex series of pathologic events involving cytokines like IL-1, IL-6, and TNFα leading to characteristic symptoms like synovial inflammation, cartilage damage, and bone erosion.1,2 The conventional therapy for this inflammatory and degenerative autoimmune disease applies a combination of disease-modifying drugs (DMARDs) such as methotrexate (MTX), glucocorticoids, and biologicals. Frequently, these therapeutic concepts ameliorate the inflammation processes only in an inadequate manner and are linked to adverse side effects such as serious infections and lymphoma. Without disregarding the significant improvement for treatment of these severe and debilitating diseases by biological TNFα inhibitors, there remain limitations, such as (a) a still large percentage

The role of TNFα in inflammatory, infectious, and malignant conditions has been studied for more than two decades. This 17 kDa soluble protein, predominantly produced by macrophages and T lymphocytes, is found in inflammatory and infectious conditions at elevated levels in serum and tissue. However, in a healthy state, TNFα is a modulator of an appropriate host response to bacterial, viral, and parasitic infections. TNFα blocking agents, the so-called “biologicals” Infliximab (IFX), Adalimumab, Golimumab, Certolizumab, and Etanercept, are used in chronic inflammatory diseases. These substances do not interfere with TNFα biosynthesis or signaling cascades but mask the freely circulating protein, thereby preventing it from initiating the signaling pathway, which is triggered by binding to TNFα receptors. Apart from rheumatoid arthritis (RA), Crohn’s disease (CD), psoriasis, and ankylosing spondylitis represent further indications, which are driven by inappropriate or excessive TNFα levels. © 2017 American Chemical Society

Received: November 10, 2016 Published: June 14, 2017 5290

DOI: 10.1021/acs.jmedchem.6b01647 J. Med. Chem. 2017, 60, 5290−5305

Journal of Medicinal Chemistry

Drug Annotation

(approximately 30%) of patients who do not respond to treatment with these drugs, (b) high costs, and (c) the route of administration.3 Even before the first biological TNFα inhibitor entered the market, researchers at SmithKline Beecham identified the mechanism of action of anti-inflammatory pyridinylimidazoles, which inhibited cytokine release (TNFα and IL1β) when cells were preincubated with these molecules prior to LPS stimulation. These compounds, initially denominated as cytokine-suppressive anti-inflammatory drugs (CSAIDs), were found to bind to a single 38 kDa protein kinase, implicating that the expression of pro-inflammatory cytokines can be modulated by inhibition of this unique enzyme, p38 MAP kinase (p38 MAPK). The concept of the development of an orally available anticytokine small molecule that could be used for the treatment of various inflammatory autoimmune diseases was soon adopted by essentially all big pharmaceutical companies and many biotech enterprises. Many different small molecule p38α MAPK inhibitors have been discovered; the latest generation exhibit high potency and a selectivity, which appear to be good enough to avoid the modulation of critical off-targets at therapeutically effective doses.4 To date, more than 20 candidates have entered clinical development. On the basis of the excellent efficacy of p38a MAPK inhibitors in various models of experimental arthritis and further to the successful development of biological antiTNFα therapies, most candidates were initially evaluated in RA patients. In standard 12-week clinical trials, however, despite an initial and substantial reduction of the inflammation biomarker C-reactive protein (CRP), at the end of the treatment period, the clinical efficacy was not superior to placebo and the CRP levels had returned to baseline values. The mechanism behind this intermittent but not sustained anti-inflammatory treatment remains unknown, but activation of redundant shunt or escape signaling pathways was suspected, e.g., via JNKs.5−8 As of today, in clinical phase 2 studies, p38α MAPK inhibitors have been investigated in other indications such as Crohn’s disease, neuropathic pain, coronary heart disease, atherosclerosis, depression, mild symptoms of Alzheimer’s disease, as well as in asthma and chronic obstructive pulmonary disease (COPD). In patients with asthma and COPD, proinflammatory cytokines such as TNFα, IL-1β, and IL-6 are found in increased amounts in the sputum and bronchoalveolar lavage (BAL) fluid. TNFα responses are initiated by binding of soluble TNFα to either TNF receptor 1 or 2, which leads to the activation of additional pathways and signaling proteins, including transforming growth factor-beta (TGFβ)-activated kinase (TAK)1 and the MAPKs c-Jun N-terminal kinases (JNKs) and p38.9 TNFα is expressed in various cells in asthmatic airways, particularly mast cells, and may play a key role in amplifying asthmatic inflammation through the activation of NF-κB. In preliminary small clinical studies with asthma patients, anti-TNFα therapy with either etanercept or infliximab (IFX) led to an improvement of lung function and reduced exacerbations in patients.10,11 However, in larger clinical trials, the efficacy was not confirmed.12,13 In patients with COPD, IFX failed to provide any benefit, as assessed by the effects on symptoms, lung function, and exercise performance when administered at doses that are effective in individuals with rheumatoid arthritis.14

Similarly, clinical studies with single-targeted therapy against IL-4 (dupilumab), IL-13 (tralokinumab, lebrikizumab), IL-17 (brodalumab), or chemokine receptor 2 (MK-7123; (R)-2hydroxy-N,N-dimethyl-3-((2-((1-(5-methylfuran-2-yl)propyl)amino)-3,4-dioxocyclobut-1-en-1-yl)amino)benzamide) failed.15 In 2016, two monoclonal antibodies against IL-5 (mepolizumab and reslizumab) have been approved in the US and Europe as add-on therapy for treatment of asthma patients with a specific eosinophilic phenotype. The rationale to explore the efficacy of p38α MAPK inhibitors in COPD is based on the efficacy in experimental in vivo and in vitro models and on the analysis of biopsy samples, which showed a correlation between activation of p38α MAPK and disease severity.16−18 Furthermore, in a smallscale exploratory clinical study, the intermittent short-term dosing (75 mg on days 1 and 6) with BCT197 (3-(5-amino-4(3-cyanobenzoyl)-1H-pyrazol-1-yl)-N-cyclopropyl-4-methylbenzamide) showed a marked improvement in lung function (FEV1) in COPD patients. As described in refs 19 and 20, this compound, now denominated as acumapimod, is currently being developed as first-line acute therapy for acute exacerbations of COPD. On the other hand, Watz et al. reported that treatment of COPD patients for 12 weeks with losmapimod, another selective and highly potent p38α MAPK inhibitor, did not cause an improvement in exercise tolerance or lung function despite being well-tolerated in this COPD population.21 However, Pascoes et al. performed subgroup analysis and demonstrated significant efficacy in a subpopulation (patients with low eosinophil count).22 PDE4 Inhibitors in Chronic Inflammatory Diseases. Theophylline has been used in the treatment of asthma (and COPD) since the 1930s, but its use declined with the development of newer and more efficacious drugs. Investigations to elucidate the mechanism of action of theophylline led to the identification of the protein kinase A (PKA) pathway, which is an integral part of the signaling cascade of G proteincoupled receptors. The PKA pathway depends on cellular cyclic adenosine phosphate (cAMP) levels that are controlled by the cAMP-forming adenyl cyclases and the cAMP-degrading phosphodiesterases, such as PDE4. As second messenger, cAMP holds key roles in a variety of physiological responses, including regulation of pro-inflammatory cytokine levels. Low cAMP levels are a result of highly active phosphodiesterases and lead to a low activity of PKA. As a result, the suppression of TNFα expression via PKA is repealed. Being mainly expressed in inflammatory cells, smooth muscle cells, endothelial cells and keratinocytes, inhibition of PDE4 is a reasonable strategy for impeding the pathophysiological events in chronic inflammatory lung diseases by blocking TNFα biosynthesis right at the scene of action.23 To date, only two PDE4 inhibitors have reached marketing authorization, and only one of them, roflumilast (Daxas EU, Daliresp US), is being marketed for the treatment of chronic inflammatory diseases of the lung, albeit just for a particular subgroup of COPD-affected patients. The other PDE4 inhibitor, apremilast (Otezla EU, US), is used for treatment of psoriatic arthritis and severe plaque psoriasis. Moreover, results from in vitro studies and experimental murine arthritis models suggest apremilast’s potential effectiveness in the treatment of RA, as reported by McCann et al.24 However, in a phase II RA clinical study, application of apremilast on top of a stable MTX treatment had no superior effect even though 5291

DOI: 10.1021/acs.jmedchem.6b01647 J. Med. Chem. 2017, 60, 5290−5305

Journal of Medicinal Chemistry

Drug Annotation

Figure 1. Synthetic strategy to achieve either tri- or tetrasubstitution of the 2-thioimidazole scaffold in a regioselective manner. This overview was modified from refs 45, 47 and 57. Reagents and conditions: (a) NaNO2, AcOH, room temperature; (b) H2, Pd/C, 1 atm, HCl in iPrOH, room temperature; (c) DMF, KSCN, reflux; (d) MeI, EtOH/THF, reflux, see also ref 48; (e) DMF, methyl-/ethyl-/benzylthiocyanate, reflux; see also ref 58; (f) 50% HBF4 (aq), NaNO2, 0 to 45 °C; (g) NaNO2, 70% HF in pyridine (Olah’s reagent), −15 to −10 °C to room temperature; (h) NaNO2, AcOH, room temperature; (i) (H2CNR2)3, EtOH, reflux; (j) 2,2,4,4-tetramethylcyclobutane-1,3-dithione, DCM, room temperature; (k) Hal-R3, K2CO3, MeOH, room temperature; (l) Br2, AcOH, room temperature; (m) excess H2N-R2, DCM/MeOH, −5 to 0 °C, then HCl in EtOH, Et2O/ acetone; (n) KSCN, DMF, reflux; (o) R3-SCN, DMF, reflux; (p) Hal-R3, K2CO3, MeOH, room temperature; (q) NBS, CCl4, 0 °C to room temperature; (r) Pd(PPh3)4 or Pd(PPh3)2Cl2/PPh3, Na2CO3 or K-OAc, toluene or DMF, 80 to 120 °C. Parentheses: n = 1:4-pyridinyl- or n = 0: hydrogen.

Further combinations are short- and long-acting β2 agonists or combinations with β2 agonists and muscarinic receptor antagonists.26 Although asthma is usually well controlled with ICS, approximately 20% of asthma patients are particularly difficult to manage, and 3−5% of patients have severe disease that is not controlled even by maximal inhaled treatment with a LABA/ICS combination.27 COPD has a high morbidity, and no current treatment, including corticosteroids, reduces disease progression or mortality and have relatively little effect in

administered at dosing levels that were efficacious in patients with psoriasis and psoriatic arthritis.25 An increase in the dose of PDE4 inhibitors is limited by a low therapeutic window, as PDE4 inhibition is associated with nausea and emesis as major side effects. Rationale for Dual Inhibitors of p38α MAPK and PDE4. The standard therapy for asthma and COPD is dominated by the use of inhaled fixed-dose combinations of a long-acting β2 agonist (LABA) and a corticosteroid (ICS). 5292

DOI: 10.1021/acs.jmedchem.6b01647 J. Med. Chem. 2017, 60, 5290−5305

Journal of Medicinal Chemistry

Drug Annotation

preventing exacerbations, reflecting their lack of anti-inflammatory effects in this disease. There is an enormous unmet need for the development of new treatments for COPD that target the underlying inflammatory process and aberrant repair mechanisms.28 The combination of a bronchodilator with an antiinflammatory agent as a treatment of COPD has been extensively discussed.29 The combination of roflumilast with salmeterol, formoterol, or dexamethasone on lipopolysaccharide (LPS)-induced peripheral blood mononuclear cells (PBMC) cytokine production demonstrated additive effects for the PDE4/LABA or PDE4/glucocorticosteroid combination.30,31 Because the development of fixed-dose combinations is sometimes challenging, bifunctional drugs (dual inhibitors) represent a potential alternative for the treatment of diseases with a complex and multifactorial pathophysiology such as asthma and COPD. Tannheimer et al. and Baker et al. described and characterized (R)-6-[[3-[[4-[5-[[2-hydroxy-2-(8hydroxy-2-oxo-1,2-dihydroquinolin-5-yl)ethyl]amino]pent-1ynyl]phenyl]carbamoyl]phenyl]sulfonyl]-4-[(3methoxyphenyl)amino]-8-methylquinoline-3-carboxamide (GS-5759) as a dual LABA/PDE4 inhibitor, which has been suggested to offer a molecular mechanism for additive or synergistic anti-inflammatory effects through elevation of a common second messenger.32,33 Despite promising in vitro data, this compound did not progress to clinical development. As the candidate was designed for inhalation, the localized antiinflammatory effect was probably not sufficient to improve the symptoms associated with asthma or COPD. The dual inhibition of p38α MAPK and PDE4 represents a new and as yet unexplored approach that may be useful for the treatment of pulmonary diseases as well as other inflammatory diseases. This rationale is supported by results generated with CGH2466 (6) (4-(3,4-dichlorophenyl)-5-pyridin-4-yl-1,3-thiazol-2-amine), a dual p38α MAPK/PDE4 inhibitor described by Trifilieff et al., which additionally exhibits antagonistic effect on adenosine A1, A2b, and A3 receptors.34 This compound suppressed the production of cytokines and oxygen radicals by human peripheral blood leukocytes in vitro more potently than those of the standard p38 MAP kinase inhibitor SB203580, the PDE4 inhibitor cilomilast, and the broad spectrum adenosine receptor antagonist CGS15943 (9-chloro-2-(2-furanyl)-[1,2,4]triazolo[1,5-c]quinazolin-5-amine).35 When given either orally or locally into the lungs, 6 inhibited ovalbumin- or lipopolysaccharide-induced airway inflammation in mice more potently than the single receptor antagonists or enzyme inhibitors used alone.34 This Drug Annotations reports the discovery of 1, a dual p38α MAPK/PDE4 inhibitor, and the background of the medicinal chemistry strategy as well as the synthetic routes and the resulting structure−activity relationships (SARs). Moreover, the promising synergistic efficacy of this dual p38α MAPK/PDE4 inhibitor is emphasized by exhaustive preclinical results. Furthermore, the outcome of a phase I clinical study to elucidate the anti-inflammatory efficacy as well as its toxicological profile and PK/PD data are reported. Discovery of Pyridinylimidazoles as Cytokine-Suppressive Anti-Inflammatory Drugs (CSAIDs). Chemistry. The developed and optimized synthetic routes are extremely flexible and allow regioselective synthesis, as evident from the overview in Figure 1. For more details, see synthesis refa 36−38 and 38−52 and patents WO 2002066458, 5 3 WO 2004018458,54 WO 2006089798,55 and WO 2008023066.56

The synthetic route of 1 is described in detail in the Supporting Information. Identification of CBS-3408 (3) and CBS-3435 (4) as Tool Compounds. The first CSAID prototypes, 6-(4fluorophenyl)-5-pyridin-4-yl-2,3-dihydroimidazo[2,1-b][1,3]thiazole (SKF-86002) and 4-[4-(4-fluorophenyl)-2-(4-methylsulfinylphenyl)-1H-imidazol-5-yl]pyridine (SB203580) were extremely valuable as tools and chemical probes and have emerged as frequently used reference compounds, thereby facilitating the experimental elucidation of various biological functions.59 However, clinical studies with SB203580 and other first generation p38α MAPK inhibitors revealed a dose-limiting hepatotoxicity presumably caused by interactions with the cytochrome P450 system. Second generation p38α MAPK inhibitors with improved selectivity profile within the p38 MAPK family showed a better safety profile but, in the end, lacked efficacy. Blockage of the p38 MAPK pathway leads to a positive feedback mechanism for JNK and other signaling pathways. The intention to inhibit these bypasses gave rise to the idea that multitarget drugs inhibiting both key drivers of inflammatory processes may exhibit additive or even synergistic effects in vivo. Consequently, SB203580s off-target activity was exploited for the development of dual inhibitors of p38α MAPK and JNKs. Further investigation demonstrated this dual approach to be a promising strategy for treatment of, e.g., Huntington’s disease.60 Derived from the SKF-86002 and SB203580 scaffold, the dual p38α MAPK/JNK3 inhibitors still contained the pyridinylimidazoles scaffold as the core structure. However, the thiazolering of SKF-86002 was cleaved in a manner that kept the sulfur atom at position 2 of the imidazole ring, as illustrated in Figure 2.

Figure 2. Template development of SB203580 and tri- and tetrasubstituted 2-thioimidazoles derived from the early dihydrothiazoline-derived lead compound SKF-86002.

In this way, a compound library (CBS library) comprising more than 2500 tri- and tetrasubstituted imidazoles was synthesized over the years. Table 1 provides an overview of the structures that are mentioned in the present work. Extensive structure−activity relationships (SARs) of the CBS library and p38α MAPK and JNK3 were determined in the course of various in vitro screenings using the isolated enzymes for determination of IC50 values and kinome selectivity. Optimizing the substituents at positions R1−R4 led to IC50 values as low as the single-digit nanomolar range for both kinases. Apart from optimized interactions, the decoration of the pyridinyl-imidazole pharmacophore with appropriate substituents modulates the selectivity and potency toward p38α MAPKs and JNKs, results in a substantial reduction in 5293

DOI: 10.1021/acs.jmedchem.6b01647 J. Med. Chem. 2017, 60, 5290−5305

Journal of Medicinal Chemistry

Drug Annotation

Table 1. Summary of Pharmacological Profiling Data of Selected Tool Compounds 3 and 4 assay potency p38 MAPKαa whole blood assay IL-1β TNFα kinase selectivityb (>70% inhibition @ 10 μM) CYP-inhibition @ 10 μMc [%] in vivo endotoxemiad

compound 3 2.35

Table 2. Pharmacokinetic Parameters (Mean of Males and Females) of 3 and 4 after Single Oral (Gavage) Administration of 30 mg/kg to Male and Female Wistar Rats

compound 4

Tmax

5.8

1.5 2.1 JNK3, p38β

3.2 5.4 JNK1, -2, -3, p38β

18, 46, 4, −7, 45 rat: 97% @ 10 mg/kg

72, 60, 90, 16.5, 74.5 mouse: 60% @ 10 mg/kg

Cmax

T1/2

AUC0−t −1

compound

h

ng/mL

μM

ng h ml

μM h

h

3 4

1.0 2.0

738 608

2.07 1.61

4928 3713

13.83 9.86

3.0 2.8

to 12.5- and 24-fold IC50 for inhibition of TNFα release in LPSstimulated human whole blood. On the basis of the rapid systemic elimination for the CIA-study, a twice daily dosing regimen was suggested. Furthermore, both compounds exhibited gender-specific pharmacokinetics with approximately 2-fold higher Cmax and AUC values determined in females. In experimental CIA in male DBA/1 mice, treatment of animals with compounds 3 and 4 (50 mg/kg b.i.d.) commenced on day 22, i.e., 1 day after the second (boost) injection of type II collagen. The development of the arthritic score, which represents a quantification of both the number of inflamed joints and the severity of inflammation, is illustrated in Figure 3. Treatment of animals with compounds 3 and 4 was associated with a substantial reduction of the disease severity, which was quantified by scoring of the number of inflamed joints and the severity of inflammation. In the CIA study, except for body weight loss essentially identical to that observed for control animals, no signs of toxicity were recorded. For the maximum tolerated dose in Wistar rats to be estimated, a two-step approach was followed. In the first setting, single doses of 100 (n = 1), 500 (n = 1), and 2500 mg/kg (n = 3) were administered, and the clinical signs were observed for 48 h. Subsequently, 300 mg/kg (n = 5) was administered twice daily for 7 days. The second dose was given 8 h after the first dose. During the 7 day study, 3 was tolerated well, whereas one animal receiving 4 was found dead on day 6, and two further animals had to be sacrificed due to poor conditions. After 7 days, treatment-related effects on body weight gain, number and composition of blood cells, liver enzymes, and macroscopic appearance of liver and spleen were recorded. In addition, an hERG-assay (manual patch-clamp technique) was performed, and both compounds were found to inhibit the hERG-mediated potassium current in a concentration-dependent manner with IC50 values of 28.6 (3) and 2.76 μM (4). On the basis of these data, it was concluded that the tool compounds were useful for demonstrating that inhibition of p38α MAPK and TNFα release in LPS-stimulated whole blood translated into a relevant efficacy in acute (endotoxemia) and chronic (CIA) inflammation in vivo. However, the observed gender-specific pharmacokinetics, the signs of toxicity after treatment of Wistar rats for 7 days, and the results of the hERG assay did not allow for considering either compound as a clinical candidate. Irrespective of this conclusion, the genderspecific pharmacokinetics of 3, which was also observed for further tetrasubstituted imidazoles, was investigated. In vitro metabolism studies with liver microsomes from rats and humans demonstrated a fast oxidation of the methylsulfanyl group to the corresponding sulfoxide. Except for traces of the sulfone, no additional metabolites were detected. Thus, on the basis of in vitro metabolism studies, the sulfoxide represents a metabolically stable compound and represented another candidate for pharmacological testing. Compound CBS-3595 (1) (Figure 4), bearing a chiral sulfoxide

a

IC50 (SB203580)/IC50 (test). bKinase profiler (Upstate): CaMKII, cRaf, GSK3β, JNK1, -2, -3, Lck, MAPK1, MAPK2, MEK-1, PKA, PKB, PKC, Rock-II, p38β, -γ, -δ. cCYP1A2, CYP2C9, CYP2C19, CYP3A4, CYP2D6. dNinety minutes after administration of 0.1 mg/kg iv to rats or 1.5 μg/kg LPS+500 mg/kg GalN i.p. to mice, TNFα plasma concentrations were used as read-out.

cytochrome P450 interactions, and consequently reduced liver toxicity potential.48 Primary screening of test compounds included the inhibition of p38α MAPK-catalyzed phosphorylation of the transcription factor ATF-2 and the suppression of TNFα/IL-1β release in LPS-stimulated human whole blood. In each enzyme assay or whole blood testing sequence, SB203580 was included as reference, and the potency was expressed as the ratio of IC50 (SB) vs IC50 (test). For the top 100 trisubstitued candidates, the values varied between 2.0 and 28.5 (enzyme assay) and between 10 2.4 0.7 5296

DOI: 10.1021/acs.jmedchem.6b01647 J. Med. Chem. 2017, 60, 5290−5305

Journal of Medicinal Chemistry

Drug Annotation

Figure 5. Inhibition of selected PDE4 splice variants by 1 and 2. Results are shown as mean ± SE from n = 3−4 experiments.

rats, and Cynomolgus monkeys for 1, 10, and 30 μM of compound 1. Low protein binding was confirmed across all species, as shown in Table 5. In comparison, at 5 μM in human plasma, the unbound fraction of 3 was 3.7%.

Table 4. Effect of 1 on the Activity of Selected ATPDependent Protein Kinases kinase

inhibition [%]

kinase

inhibition [%]

c-RAF JNK2α2 JNK3 p38δ CK1δ CK1 MKK4

36 51 76 35 94 97 37

MKK6 MKK7β Lyn EphB2 RIPK2 CDK3/cyclinE cyclinH/MAT1 p70S6K

62 101 31 60 33 55 54 32

Table 5. Plasma Protein Binding of 1 unbound fraction [%] human Sprague−Dawley rats Cynomolgus monkey

1 μM

10 μM

30 μM

90.0 ± 0.0 86.3 ± 12.2 66.7 ± 4.1

80.6 ± 0.9 63.1 ± 8.9 61.1 ± 1.7

78.1 ± 2.8 75.4 ± 5.2 58.3 ± 3.1

For investigation of 1’s in vitro metabolism, concentrations of 10, 30, and 100 μM of the compound were incubated with liver microsomes from either rats, Cynomolgus monkeys, or humans. Two major metabolites were detected: 2, deacetylated 1, and the oxidized sulfone metabolite CBS-3733 (5), as shown in Figure 7. Investigation of the biotransformation of 1 in primary human hepatocytes revealed 2 to be the main metabolite. The sulfone metabolite 5 and its deacetylated product CBS-3846 (7) were also detected but at substantially lower concentrations. 6. Efficacy of 1 on TNFα Release in LPS-Induced Endotoxemia in Rats and Mice. First, the effect of orally administered 1 was assessed in Sprague−Dawley rats. Five treatment groups (8 animals/group) received either vehicle (control) or 1, 3, or 10 mg/kg of compound 1 or 3 mg/kg of SB203580, respectively. One hour after dosing, 0.1 mg/kg of LPS was administered by intravenous injection. Two hours later (i.e., 3 h after dosing), blood samples were collected, and TNFα serum levels were determined. In comparison to the mean TNFα concentration determined in vehicle-treated animals, 1, 3, and 10 mg/kg of 1 inhibited the TNFα release by an average of 43, 88, and 96%, respectively. Regarding the benchmark of 80% of TNFα inhibition, a dose of 3 mg/kg of compound 1, reduced TNFα serum levels below this value in seven out of the eight animals, and 10 mg/kg of 1 was sufficient to diminish the TNFα concentrations in all animals below this 80% limit. Administration of 3 mg/kg of SB203580 decreased the TNFα serum levels by an average of 37%. Next, the effect of the time period between administration of 1 and LPS challenge on TNFα release and the existence of gender-specific efficacy were investigated. To this end, female Lewis rats (n = 8/group) received 10 mg/kg 1, 3, and 5 h prior to the LPS challenge (relative time points: −1, −3, and −5 h) . In addition, one group of male Lewis rats was treated with 10 mg/kg at time point −1 h. As controls, each of one group with male and female animals received vehicle and one group of

Figure 6. Effect of orally administered 1 on TNFα synthesis in ex vivo LPS-stimulated whole blood from Cynomolgus monkeys.

A concentration of 12 mg/kg of 1 was able to reduce the release of TNFα by more than 50% when compared to the TNFα achieved with vehicle only. The dose level of 20 mg/kg completely inhibited TNFα release in this experiment. 5. ADME Experiments: Membrane Permeability, Plasma Protein Binding, and in Vitro Metabolism of 1. The compound’s membrane permeability was investigated in Caco-2 cell monolayers at the final test concentration of 10 μM and the obtained results were compared to those of the reference compounds (atenolol for low, propranolol for high permeability, and talinolol as P-glycoprotein substrate). Compound 1 demonstrated a fairly good permeability (Papp= 18.5 ± 6.2 × 10−6 cm/s compared to 8.7 ± 2.1 and 71.1 ± 6.8 × 10−6 cm/s for atenolol and propranolol, respectively), and Pglycoprotein-mediated transport could be excluded. The extent of plasma protein binding was determined by equilibrium dialysis in plasma from humans, Sprague−Dawley 5297

DOI: 10.1021/acs.jmedchem.6b01647 J. Med. Chem. 2017, 60, 5290−5305

Journal of Medicinal Chemistry

Drug Annotation

Figure 7. Metabolic pathways of the dual p38α MAPK/PDE-4 inhibitor 1.

Figure 8. Effect of orally administered 1, the reference p38 MAPK-inhibitor (8), the highly potent tetrasubstituted imidazole p38 MAPK inhibitor (9), and the PDE4 inhibitor roflumilast on TNFα release in Lewis rats following LPS challenge.

p38α MAPK inhibitor CBS-3835 (9) (15-times as potent as SB203580), and roflumilast in a head-to-head setting in eight Lewis rats. One hour prior to LPS challenge, 10 mg/kg of each compound and vehicle were orally administered. Then 1.5 h after LPS challenge, the resulting TNFα plasma levels were determined. Compound 1’s reduction of TNFα release in an in vivo rat LPS model was superior when compared to the selective inhibitors of p38α MAPK or PDE-4. Its performance was comparable to that of the very potent p38α MAPK inhibitor or roflumilast, which featured an IC50 value of 0.2 μM in this experiment. As illustrated in Figure 8, TNFα synthesis was inhibited by almost 80% after administration of 1 when compared to the vehicle group. For roflumilast, a mean inhibition of 67% was calculated, demonstrating that at the same dose level, 1 and roflumilast exhibit comparable effects toward the inhibition of TNFα release. In male BALB/c mice (n = 6/group), endotoxemia was induced by intraperitoneal injection of 500 mg/kg of D-GalN and 1.5 μg/kg of LPS, and TNFα serum concentrations were determined 90 min later. The effects of 3, 10, or 30 mg/kg of 1 administered by oral gavage 1 h prior to LPS/GalN injection were compared to vehicle, and another group received only vehicle (control). Then 1.5 h after dosing, blood samples were collected by cardiac puncture, out of which plasma levels of

females received 30 mg/kg of SB203580, all administered at time point −1 h. In addition, aliquots of the serum samples were used for determination of drug concentrations. The extent of TNFα suppression was dependent on the relative time point of dosing. The closer the time of dosing to LPS challenge, the higher the plasma concentration at 1.5 h post dose, and the higher the inhibition of TNFα release. Administration of 10 mg/kg of 1 5 h before LPS challenge resulted in comparable TNFα plasma levels than those achieved with 30 mg/kg of SB203580 administered 1 h prior to LPS challenge. The comparison of data obtained with male and female animals indicated the absence of gender-specific PK or efficacy. However, application of 1 3 or 1 h before injection of LPS was more effective than the reference compound SB203580 in blocking TNFα release in male and female animals. Serum levels of 237 ± 20, 526 ± 40, and 871 ± 21 ng/mL of 1 were measured 1.5 h after LPS administration in female animals. In the group of male rats that received 10 mg/kg of 1 1 h before LPS administration, the serum concentration was 689 ± 20 ng/mL and thus in a similar range when compared to the female animals. Third, the effect of the p38α MAPK inhibitor properties of 1 was compared to the effects of an intermediate p38α MAPK inhibitor (8) (4-times as potent as SB203580), a very potent 5298

DOI: 10.1021/acs.jmedchem.6b01647 J. Med. Chem. 2017, 60, 5290−5305

Journal of Medicinal Chemistry

Drug Annotation

Figure 9. Effect of 1 (20 and 50 mg kg−1 day−1 p.o.) and IFX on paw swelling in hTNFtg mice with inflammatory arthritis.

out. Histological readouts were inflammation, bone erosion, osteoclast number, and cartilage damage. As illustrated in Figure 9, both dose regimens of 1 showed a similar reduction in paw swelling compared to IFX after week ten. Compound 1 also demonstrated beneficial effects on synovitis, erosion, and cartilage damage. 8. PK in Vivo Studies of 1. Plasma/concentration profiles of 1 after oral administration were determined in different animal species either as part of dedicated pharmacokinetic studies or from satellite animals in general toxicity studies. For the characteristics of 1 in different species to be compared, results obtained with similar doses are presented. After oral administration of 30 mg/kg to two mongrel dogs, blood samples were collected after 0.25 to 24 h, and subsequently, plasma concentrations of 1 were determined. In a similar experiment with four Beagle dogs orally receiving 40 mg/kg of 1 to determine the maximum tolerated dose (MTD), the toxicokinetic findings featured comparable results. In this experiment, the concentration of 2 was also monitored, revealing a mean Cmax of 5.1 μM, a Tmax of 4.0 h, and an AUC0−t of 86.5 μM h. Further PK studies in rats demonstrated that 1’s PK properties do not differ in male and female animals, as comparable results were obtained for animals of both genders. All results are summarized in Table 6.

TNFα as well as concentrations of 1 and its predominant metabolite 2 were determined. The dose-dependent diminution of TNFα levels already observed in previous experiments was reconfirmed. In comparison to the control group, the mean TNFα release was reduced by 82, 84, and 95% for the administered doses of 3, 10, and 30 mg/kg of 1, respectively. These values are in good accordance with the measured plasma concentrations of 1 and its metabolite 2. 7. Efficacy of 1 in Experimental Arthritis Models. In collagen-induced arthritis in the rat, male Lewis rats were assigned to four treatment groups, a control group of five animals received vehicle alone and a group of five animals received 1 mg/kg of dexamethasone. Compound 1 was administered as a once daily treatment of 50 mg/kg to a group of six animals and a twice daily treatment of 25 mg/kg to a group of five animals. The development of disease was monitored by assessment of body weight, arthritic score (AS), and paws volume. In addition, on an exploratory level, at the end of the 10-day treatment period, blood samples were collected for determination of the collagen oligomeric matrix protein (COMP), a biomarker for cartilage degradation. On day 10, the observed ASs were 7.5, 6.0, 4.0, and 0 for the control group, twice daily dosing regimen, single dose, and dexamethasone, respectively. Regarding the paw volumes between the vehicle group and the groups treated with 1, there were no relevant differences observable. However, when compared to the COMP serum concentration of the control of 0.17 ± 0.02 U/l, the COMP serum concentration of animals treated with 1 was significantly lower with 0.08 ± 0.07 U/l for both dosing regimens. The lowest COMP serum concentration of 0.03 ± 0.02 U/l was achieved with 1 mg/kg of dexamethasone. Above that, the effect of a ten week treatment of 1 was also studied in a human TNF transgenic mouse model and directly compared with the efficacy of an anti-TNFα antibody (infliximab, IFX). Heterozygous human TNF transgenic (hTNFtg) mice develop chronic inflammatory and destructive polyarthritis 4−6 weeks after birth. Three treatment groups comprising eight animals each receiving either 20 or 50 mg kg−1 day−1 of compound 1 or 10 mg kg−1 week−1 of IFX were compared to a control group of eight animals receiving only vehicle. While 1 was administered orally as methylcellulose suspension, IFX was given intraperitoneally. AS benchmarks served joint swelling and grip strength, which was monitored in all animals according to a weekly schedule. At the end of the study, histological examination of the hind paws was carried

Table 6. Main PK Parameters of 1 in Different Species mongrel dog oral dose [mg/kg] Cmax [μM] Tmax [h] T1/2 [h] AUC0−t [μM h] AUC0−∞ [μM h]

Beagle dog

female rats

male rats

male monkeys

30

40

30

30

20

14.3 2 3.85 95.05 96.85

15.6 1.5 7.8 93.4 120.2

4.8 2 2.2 40.9 40.9

3.4 0.5 1.8 29.8 29.8

8.4 4 4.8 79.6

A 4 week toxicity study in Sprague−Dawley rats demonstrated that the systemic exposure increased with the administered dose. There are no indications of a substantially overproportional increase in systemic exposure. During the administration of 15 and 50 mg kg−1 day−1 over a period of 4 weeks, the systemic exposure of 1 remained essentially unchanged, and thus the influence of subacute treatment on the disposition of 1 can be excluded. In a 4 week toxicity study, Cynomolgus monkeys received 6, 12, and 20 mg/kg of 1. On days 1 and 27, blood samples were collected at time points 0.5 5299

DOI: 10.1021/acs.jmedchem.6b01647 J. Med. Chem. 2017, 60, 5290−5305

Journal of Medicinal Chemistry

Drug Annotation

Table 7. Mean PK Parameters of 1 after Single Oral Administration of Different Doses to Healthy Subjects dose 10 50 10 10 20 30

mg mg mg (fasting) mg (fed) mg mg

Tmax [h] 2.20 1.50 3.30 3.70 2.90 2.30

± ± ± ± ±

1.30 0.70 1.60 1.40 0.80

Cmax [μM] 0.20 0.59 0.14 0.14 0.31 0.42

± ± ± ± ±

AUC0−t [μM h]

0.06 0.08 0.03 0.02 0.09

0.95 3.40 0.73 0.79 1.99 2.65

to 24 h post-dosing. Individual concentration/time profiles of 1 and its metabolite 2 were used to determine the parameters T max , C max , T 1/2 , and AUC 0−24 . In summary, plasma concentrations and derived toxicokinetic parameters demonstrated that (a) all animals were continuously exposed to 1 and, clearly to a lower extent, to metabolite 2, (b) on both days 1 and 27 the systemic exposure increased with increasing doses in a slightly subproportional manner, (c) there were no genderspecific toxicokinetic characteristics, and (d) the daily dosing over a period of 4 weeks led to a substantially lower extent of bioavailability. 9. Safety Pharmacology Studies of 1. Safety pharmacology studies included the investigation of the effect of 1 on the central nervous system and respiratory system both in rats as well as the cardiovascular system in Cynomolgus monkeys. In addition, an hERG assay was performed with both 1 and its predominant metabolite 2. After single oral administration of 30, 100, and 300 mg/kg of 1, no treatment-related effects on the central nervous system were recorded. In addition, as investigated at the same dose levels, 1 did not modify the respiratory rate. However, a significant and dose-related increase in peak inspiratory flow was reported after dosing with 100 and 300 mg/kg of 1, respectively. When compared to predose values, a significant increase in peak expiratory flow was observed at the same dose levels. A significant increase of the minute volume was reported after dosing with 300 mg/kg of 1. This could be a consequence of the peak inspiratory flow’s augmentation. Apart from that, excessive defecation and soft feces were observed after treatment with the test item at dose levels of 30 and 300 mg/kg. Furthermore, loud breathing and excessive salivation were noted after treatment with 300 mg/kg of 1 in 1/8 and 2/8 animals 4 and 6 h after dosing, respectively. The hERG-mediated potassium current in HEK 293 cells was influenced by 1 and 2 at very high concentrations. IC50 values of 213 and 83 μM indicate a very low probability of a drug-induced cardiovascular effect at therapeutic doses. The in vivo study was performed in six Cynomolgus monkeys. Dose levels of 6, 12, and 20 mg/kg of either compound 1 or vehicle were orally administered according to a crossover design, respecting a wash-out period of at least 3 days between two treatments. At all doses of 1, traces of whitish vomiting were reported. Regarding the cardiovascular function, induced tachycardia from 0.5 h after dosing until the end of the recording period after 6 h was observed. This was associated with a decrease in PQ and QT durations. The mean blood pressure remained unaffected at all dose levels. Dose-related periods of sinusal brady-arrhythmias were reported, possibly related to the demonstrated emetic properties of the test item. 10. General Toxicity of 1. Acute toxicity studies in the rats as well as dose-finding studies and four week toxicity studies in the rats and Cynomolgus monkeys were performed. In the acute toxicity setting, five male and five female Sprague−Dawley rats

± ± ± ± ±

0.31 0.53 0.25 0.19 0.50

T1/2 [h] 2.7 3.2 2.8 2.8 3.3 3.4

± ± ± ± ±

0.5 0.9 0.4 0.4 0.4

AUC0−∞ [μM h] 1.04 3.58 0.82 0.88 2.24 3.05

± ± ± ± ±

0.35 0.52 0.25 0.23 0.59

received a single oral dose of 2,000 mg/kg (gavage) of 1, and effects were observed for a period of 14 days. No mortalities were observed. As clinical signs, piloerection, hypoactivity, and dyspnea were noted on days 1 and 2. No clinical signs were noted from day 3. The outcome of the dose-range-finding study in Sprague−Dawley rats was a maximum tolerated dose (MTD) of 150 mg/kg. On the basis of this information, 144 Sprague−Dawley rats were allocated to four treatment groups of 10−16 animals per group. During the 28 day treatment period, the control group received only vehicle, whereas the three groups received 15, 50, or 150 mg/kg of test item as low, intermediate, or high dose, respectively. After these 28 days, six male and female animals of the control and high dose group each were observed for a further 2-week treatment-free period. There were no premature deaths during this study. As expected, clinical signs such as an increase in abdomen size, piloerection, hunched posture, or ptyalism and histopathological changes were only observed after daily administration of the MTD of 150 mg/kg. The MTD study in Cynomolgus monkeys suggested that the highest dose of the subsequent 4 week toxicity should not exceed 20 mg kg−1 day−1. On the basis of this information, 32 Cynomolgus monkeys were allocated to four treatment groups of three to five animals per group. During the 28 day treatment period, the control group received only vehicle, and the other three groups received 6, 12, or 20 mg/kg of test item as low, intermediate, or high dose, respectively. After 1 week, the highdose level was increased to 25 mg kg−1 day−1. After 28 days, two male and female animals of the control and high dose group each were observed for a further 2 week treatment-free period. There were no unscheduled deaths and no treatmentrelated clinical signs at any time during the study. Under the conditions of this study, a no observable effect level (NOEL) was not established for males or females and is considered to be lower than 6 mg kg−1 day−1. However, as none of the observed changes were considered to be adverse or of toxicological importance, the no observable adverse effect level (NOAEL) is considered to be greater than 25 mg kg−1 day−1 under the conditions of the study. 11. Genotoxicity. The mutagenic and carcinogenic potency was evaluated with a bacterial reverse mutation assay (AMES test) and also cytogenetic assay in human lymphocytes and an in vivo micronucleus test in mice. In the AMES test, 1 gave a negative, i.e., nonmutagenic, response in all S. typhimurium and E. coli strains investigated in both the presence and absence of S9 mix. The mouse bone marrow micronucleus test demonstrated that 1 was not clastogenic. In the cytogenetic assay with cultured human lymphocytes, 1 was not clastogenic in vitro in either the presence or absence of S9 mix. 12. First-In-Human (FIH) Study to Investigate the Safety, Tolerability, Pharmacokinetics, and ex Vivo Pharmacodynamics of 1. In a placebo-controlled, single-blind (volunteerblind), randomized, consecutive dosing FIH study, a starting 5300

DOI: 10.1021/acs.jmedchem.6b01647 J. Med. Chem. 2017, 60, 5290−5305

Journal of Medicinal Chemistry

Drug Annotation

Table 8. Mean PK Parameters of 2 after Single Oral Administration of Different Doses to Healthy Subjects dose 10 50 10 10 20 30

mg mg mg (fasting) mg (fed) mg mg

Tmax [h] 4.70 5.00 6.00 6.70 5.00 6.00

± ± ± ± ±

2.40 1.20 2.40 2.00 1.10

Cmax [μM] 0.19 0.60 0.15 0.14 0.36 0.42

± ± ± ± ±

AUC0−t [μM h]

0.05 0.12 0.03 0.03 0.07

2.95 8.73 2.82 2.60 6.38 7.37

± ± ± ± ±

1.37 1.57 0.85 0.90 0.26

T1/2 [h] 9.5 9.0 11.3 10.7 10.9 11.3

± ± ± ± ±

3.0 0.5 2.0 2.4 1.6

AUC0−∞ [μM h] 3.43 9.40 3.29 3.07 7.17 8.55

± ± ± ± ±

1.69 1.64 1.03 1.09 0.30

Figure 10. Results of ex vivo LPS stimulation of blood samples withdrawn from healthy subjects at different time points after single administration of 30 mg of 1. The relative TNFα release was determined by comparison of actual concentrations compared to the TNFα concentration determined in LPS-stimulated pre-dose blood samples. The PD effect is compared to plasma concentrations of 1 and its metabolite 2.

dose of 10 mg was selected, and in case of the absence of adverse events, 50 mg was chosen as the next dose level. The dose of 10 mg was tolerated well, but after administration of 50 mg of compound 1, all four subjects reported nausea and malaise, and some vomiting events were recorded. In a subsequent setting, doses of 20 and 30 mg were applied, which confirmed a clear relationship between dose and adverse events, which were characteristic for PDE4 inhibitors. Because of the mild to severe adverse reactions (nausea, vomiting, and malaise), the study was interrupted and not continued. For determination of PK and metabolism in humans, drug plasma concentrations of 1 and its metabolite 2 were measured predose and between 0.25 and 36 h post dose. Mean pharmacokinetic parameters of 1 and 2, determined from individual plasma concentration time profiles, are summarized in Tables 7 and 8, respectively. Both when administered under fasting conditions or after intake of a standard breakfast, 1 was readily absorbed from the gastrointestinal tract. Recorded adverse reactions, especially those observed after treatment of subjects with 30 and 50 mg of 1, which included vomiting events, had no impact on the PK characteristics. In the 30 treated subjects, quantifiable plasma concentrations of 1 were detectable between 0.5 and 3.0 h as earliest time points. In 72% of the subjects, the first quantifiable concentration occurred at 0.50 and 0.75 h post dosing; 75% of subjects achieved peak plasma concentrations of the parent drug between 1.5 and 4.0 h. In comparison to that, the systemic exposure of 2 was slightly delayed. The calculated mean

elimination half-lives were between 2.7 and 3.4 h for 1 and between 9.0 and 11.3 h for 2. For the route of the elimination of 1 to be assessed, urine samples were collected and analyzed. Urine collection for clinical laboratory evaluation was performed at intervals of 0−6, 6−12, 12−24, and 24−36 h post dose. Results suggested that the relative urinary excreted amount of 1 and its metabolite 2 was ∼6% of the administered dose. The determined PK parameters of 1 and 2 are listed in Tables 7 and 8. Adverse events (AEs), vital signs, physical findings, and other related observations: The AEs observed were vomiting, retching, nausea, and malaise, which can be attributed to PDE4 inhibition. Generally, they had an onset of 1−2 h postdose with a mild-to-moderate intensity that correlated with the administered dose. No major CNS effects were observed at the dose of 50 mg. Subjects receiving matching placebo did not experience any AEs post-administration. All subjects enrolled concluded the study as defined in the protocol. For investigating if the tolerability of 1 might be improved by performing the administration under fed conditions, a crossover design at a dose of 10 mg of 1 was applied as an amendment to the study protocol. Treatment with a 10 mg single dose was generally well tolerated, and no clear conclusion could be drawn. Because of the significant AEs observed at doses of 30 and 50 mg, which were considered to be probably treatmentrelated, the study was prematurely interrupted and finally terminated for ethical reasons. The observations made after administration of the 30 mg single dose confirmed the previous findings made after 50 mg dosing and proved that the use of 1 5301

DOI: 10.1021/acs.jmedchem.6b01647 J. Med. Chem. 2017, 60, 5290−5305

Journal of Medicinal Chemistry

Drug Annotation

revealed that 2 potently inhibits PDE4 isoform D3, which is not effected to such an extent by 1. This PDE4 isoform is considered to be mainly responsible for the vomiting and nausea side effects of PDE4 inhibitors. Thus, the lesson learned is that isoform selectivity is an essential aspect to consider as it might decide the fate of a clinical candidate, and its prediction during the process of rational inhibitor design is still not possible. However, besides RA, COPD and psoriasis are other indications for dual p38α MAPK/PDE4 inhibitors. Both COPD and psoriasis allow topical or local application. In this way, the emetic effect via systemic application can be bypassed. For these new application routes, other candidates of the CBS library with distinct p38α MAPK/PDE4 inhibition profiles have to be investigated with regard to their physicochemical properties, isoform selectivity, and metabolic stability.

is limited by a clear emetic effect of the compound. Concerning body weight, blood pressure, and heart rate, no significant changes from the clinical point of view were observed in any of the treatment groups. As a pharmacodynamical parameter, the efficacy of 1 was assessed in an ex vivo whole blood assay. Therefore, LPSinduced TNFα release was induced in blood samples of the time points pre-dose, 1, 2, 4, 9, 24, and 36 h post-dose. Compared to placebo data in subjects treated with single dose levels of 10, 20, or 30 mg of the test item, a clear inhibition of TNFα release was achieved. Because of the use of an erroneous anticoagulant at the beginning of the study (dosing with 10 and 50 mg), the ex vivo assay could not be performed because the addition of LPS did not result in an induction of TNFα release. Figure 10 shows the mean TNFα release determined in blood samples within 36 h after administration of 30 mg of test item. Compared to placebo data, a clear inhibition of TNFα release was achieved in subjects treated with 1. The outcome of the FIH study clearly demonstrated that the tolerated dose is determined by the well-known PDE4 inhibition side effect. To evaluate whether a clinically relevant anti-inflammatory effect can be expected when patients are treated with a tolerated dose of 1, the pharmacokinetics of 1 and its pharmacologically active metabolite 2 were compared to those of published pharmacokinetic data of roflumilast and roflumilast N-oxide.63 This comparison was performed with consideration of plasma protein binding data of all analytes, i.e., fu = 0.5 for both 1 and 2 as well as fu = 0.02 and 0.04 for roflumilast and roflumilast-N-oxide, respectively.64 Pharmacokinetic parameters Cmax and AUC (unbound fractions) are summarized in Table 9.



CONCLUSIONS In summary, 1, a dual p38α MAPK/PDE4 inhibitor, was discovered by coincidence in quest of a dual p38α MAPK/ JNK3 inhibitor. The compound features moderate inhibition of p38α MAPK and PDE-4 in the higher nanomolar range in combination with excellent kinetic properties and low plasma protein binding contributing to the outstanding results of the efficacy experiments performed in animal models and humans. In in vitro, ex vivo, and in vivo pharmacology studies, a substantial suppression of TNFα release was demonstrated. The inhibition of TNFα release was achieved at dose levels and systemic concentrations that were well below levels related to toxicity. Therefore, 1 is considered as a candidate drug for the treatment of diseases related to an overproduction of TNFα. Examples are rheumatoid arthritis, psoriatic arthritis, psoriasis, COPD, asthma, and CD. In this first administration to humans, the MTD of 1 was determined to be 20 mg. A further increase of the dose was limited by emesis, a known side effect of all PDE-4 inhibitors. A clear benefit of 1 over roflumilast is its additional p38α MAPK inhibitory property, thereby providing an additional antiinflammatory effect.

Table 9. Comparison of Cmax and AUCinf Data Obtained Following Single Oral Administration of 0.5 mg Roflumilast and 20 mg of 1 dose

analyte

0.5 mg of roflumilast

roflumilast

Cmax [nM] AUCinf [nM h] dose

13.1 86.8

roflumilast N-oxide 22.4 837.9 analyte

sum



35.5 924.7

20 mg of 1

1

2

sum

Cmax [nM] AUCinf [nM h]

313.9 2242.9

308.7 6102.2

622.6 8345.1

ASSOCIATED CONTENT

* Supporting Information S

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.jmedchem.6b01647. Information regarding physical, chemical, and pharmaceutical properties and formulation of 1 used for conduction of the described studies, synthesis of 1, and a detailed list of the conducted studies as well as the related study declarations (PDF) Molecular formula strings (CSV)

These data revealed 9- and 17-fold higher plasma concentrations of 1 and 2 than those determined for roflumilast and its active N-oxide metabolite. Consequently, the ratio of systemic exposure increases to approximately 110- and 220fold. Thus, with regard to PDE4 inhibition, the substantially higher potency of roflumilast compared to 1 is at least compensated by the significantly higher systemic exposure of unbound drug obtained for 1 and its active metabolite 2. With regard to the emetic effect of 1, it must be noted that these study-limiting AEs occurred in much lower doses for humans than for monkeys. One reasonable explanation, which might be attributed to differences in metabolism, is the actual observed plasma level of 2. The concentration in human plasma was much higher than expected based on preliminary in vitro and in vivo metabolism studies. A retrospective analysis of the inhibitory effect of 2 toward the distinct PDE4 subtypes



AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. ORCID

Stefan A. Laufer: 0000-0001-6952-1486 Notes

The authors declare no competing financial interest. 5302

DOI: 10.1021/acs.jmedchem.6b01647 J. Med. Chem. 2017, 60, 5290−5305

Journal of Medicinal Chemistry



Drug Annotation

(12) Erin, E. M.; Leaker, B. R.; Nicholson, G. C.; Tan, A. J.; Green, L. M.; Neighbour, H.; Zacharasiewicz, A. S.; Turner, J.; Barnathan, E. S.; Kon, O. M.; Barnes, P. J.; Hansel, T. T. The effects of a monoclonal antibody directed against tumor necrosis factor-alpha in asthma. Am. J. Respir. Crit. Care Med. 2006, 174 (7), 753−762. (13) Morjaria, J. B.; Chauhan, A. J.; Babu, K. S.; Polosa, R.; Davies, D. E.; Holgate, S. T. The role of a soluble TNFalpha receptor fusion protein (etanercept) in corticosteroid refractory asthma: a double blind, randomised, placebo controlled trial. Thorax 2008, 63 (7), 584− 591. (14) Rennard, S. I.; Fogarty, C.; Kelsen, S.; Long, W.; Ramsdell, J.; Allison, J.; Mahler, D.; Saadeh, C.; Siler, T.; Snell, P.; Korenblat, P.; Smith, W.; Kaye, M.; Mandel, M.; Andrews, C.; Prabhu, R.; Donohue, J. F.; Watt, R.; Lo, K. H.; Schlenker-Herceg, R.; Barnathan, E. S.; Murray, J.; Investigators, C. The safety and efficacy of infliximab in moderate to severe chronic obstructive pulmonary disease. Am. J. Respir. Crit. Care Med. 2007, 175 (9), 926−934. (15) Durham, A. L.; Caramori, G.; Chung, K. F.; Adcock, I. M. Targeted anti-inflammatory therapeutics in asthma and chronic obstructive lung disease. Transl. Res. 2016, 167 (1), 192−203. (16) Renda, T.; Baraldo, S.; Pelaia, G.; Bazzan, E.; Turato, G.; Papi, A.; Maestrelli, P.; Maselli, R.; Vatrella, A.; Fabbri, L. M.; Zuin, R.; Marsico, S. A.; Saetta, M. Increased activation of p38 MAPK in COPD. Eur. Respir. J. 2008, 31 (1), 62−69. (17) Chung, K. F. p38 mitogen-activated protein kinase pathways in asthma and COPD. Chest 2011, 139 (6), 1470−1479. (18) Alevy, Y. G.; Patel, A. C.; Romero, A. G.; Patel, D. A.; Tucker, J.; Roswit, W. T.; Miller, C. A.; Heier, R. F.; Byers, D. E.; Brett, T. J.; Holtzman, M. J. IL-13-induced airway mucus production is attenuated by MAPK13 inhibition. J. Clin. Invest. 2012, 122 (12), 4555−4568. (19) De Buck, S.; Hueber, W.; Vitaliti, A.; Straube, F.; Emotte, C.; Bruin, G.; Woessner, R. Population PK-PD model for tolerance evaluation to the p38 MAP kinase inhibitor BCT197. CPT: Pharmacometrics Syst. Pharmacol. 2015, 4 (12), 691−700. (20) Fryszman, O. M.; Lang, H.; Lan, J.; Chang, E.; Fang, Y. 5Membered Heterocycle-Based p38 Kinase Inhibitors. WO 2005/ 009973, 2005. (21) Watz, H.; Barnacle, H.; Hartley, B. F.; Chan, R. Efficacy and safety of the p38 MAPK inhibitor losmapimod for patients with chronic obstructive pulmonary disease: a randomised, double-blind, placebo-controlled trial. Lancet Respir. Med. 2014, 2 (1), 63−72. (22) Pascoe, S. J., Shore, A. D., Losmapimod for Treating COPD. WO 2016/066687, 2016. (23) Schett, G.; Sloan, V. S. Apremilast: A Novel PDE4 Inhibitor in the Treatment of Autoimmune and Inflammatory Diseases. Ther. Adv. Musculoskeletal Dis. 2010, 2 (5), 271−278. (24) McCann, F. E.; Palfreeman, A. C.; Andrews, M.; Perocheau, D. P.; Inglis, J. J.; Schafer, P.; Feldmann, M.; Williams, R. O.; Brennan, F. M. Apremilast, a novel PDE4 inhibitor, inhibits spontaneous production of tumour necrosis factor-alpha from human rheumatoid synovial cells and ameliorates experimental arthritis. Arthritis Res. Ther. 2010, 12 (3), R107. (25) Genovese, M. C.; Jarosova, K.; Cieslak, D.; Alper, J.; Kivitz, A.; Hough, D. R.; Maes, P.; Pineda, L.; Chen, M.; Zaidi, F. Apremilast in patients with active rheumatoid arthritis: a phase II, multicenter, randomized, double-blind, placebo-controlled, parallel-group study. Arthritis Rheumatol. 2015, 67 (7), 1703−1710. (26) Page, C.; Cazzola, M. Bifunctional drugs for the treatment of respiratory diseases. Handb. Exp. Pharmacol. 2016, 237, 197−212. (27) Barnes, P. J. Kinases as novel therapeutic targets in asthma and chronic obstructive pulmonary disease. Pharmacol. Rev. 2016, 68 (3), 788−815. (28) Barnes, P. J. Development of new drugs for COPD. Curr. Med. Chem. 2013, 20 (12), 1531−1540. (29) Phillips, G. S. M. Bifunctional compounds for the treatment of COPD. Annu. Rep. Med. Chem. 2012, 47, 209−221. (30) Seldon, P. M.; Meja, K. K.; Giembycz, M. A. Rolipram, salbutamol and prostaglandin E2 suppress TNFalpha release from

ACKNOWLEDGMENTS Stefanie Bü hler, Dominik Hauser, Roland Selig: CAIR Biosciences, Hans-Günther Striegel: former Merckle GmbH, Gernot-Sebastian Haase, Heike Wentsch. BioProfile, BMBF, Grant Number: 0313857A



ABBREVIATIONS AE, adverse event; AS, arthritic score; cAMP, cyclic adenosine phosphate; CD, Crohn’s disease; CNS, central nervous system; COMP, collagen oligomeric matrix protein; COPD, chronic obstructive pulmonary disease; CSAID, cytokine-suppressive anti-inflammatory drug; D-GalN, D-Galactosamine; DMARDs, disease-modifying drugs; GPCR, G protein-coupled receptor; hERG, human-ether-à-go-go-related gene; hTNFtg, heterozygous human TNF transgenic; IFX, infliximab; JNK, c-Jun Nterminal kinase; LPS, lipopolysaccharide; LTB4, leukotriene B4; MAPK, p38α mitogen-activated protein kinase; MTD, maximum tolerated dose; MTX, methotrexate; NOAEL, no observable adverse effect level; NOEL, no observable effect level; PDE-4, phosphodiesterase 4; PD, pharmacodynamic; PK, pharmacokinetic; PKA, protein kinase A; RA, rheumatoid arthritis; SAR, structure−activity relationship; SD, standard deviation; SEM, standard error of the mean; SPR, surface plasmon resonance; TAK, TGFβ-activated kinase; TGFβ, transforming growth factor beta; TNFα, tumor necrosis factor alpha



REFERENCES

(1) Bradley, J. R. TNF-mediated inflammatory disease. J. Pathol. 2008, 214 (2), 149−160. (2) Zwerina, J.; Hayer, S.; Redlich, K.; Bobacz, K.; Kollias, G.; Smolen, J. S.; Schett, G. Activation of p38 MAPK is a key step in tumor necrosis factor-mediated inflammatory bone destruction. Arthritis Rheum. 2006, 54 (2), 463−472. (3) Albrecht, W.; Laufer, S. Dual Inhibition of Phosphodiesterase-4 and p38 MAP Kinase: a Strategy for Treatment of Chronic inflammatory Diseases. In RSC Drug Discov., 1st ed.; Levin, J., Laufer, S., Eds.; Royal Society of Chemistry: Cambridge, UK, 2012; pp 137−157. (4) Dominguez, C.; Powers, D. A.; Tamayo, N. p38 MAP kinase inhibitors: many are made, but few are chosen. Curr. Opin. Drug Discovery Dev. 2005, 8 (4), 421−430. (5) Cheung, P. C. F.; Campbell, D. G.; Nebreda, A. R.; Cohen, P. Feedback control of the protein kinase TAK1 by SAPK2a/p38 alpha. EMBO J. 2003, 22 (21), 5793−5805. (6) Wang, Y. J.; Singh, R.; Lefkowitch, J. H.; Rigoli, R. M.; Czaja, M. J. Tumor necrosis factor-induced toxic liver injury results from JNK2dependent activation of caspase-8 and the mitochondrial death pathway. J. Biol. Chem. 2006, 281 (22), 15258−15267. (7) Sweeney, S. E. The as-yet unfulfilled promise of p38 MAPK inhibitors. Nat. Rev. Rheumatol. 2009, 5 (9), 475−477. (8) Hammaker, D.; Firestein, G. S. ″Go upstream, young man″: lessons learned from the p38 saga. Ann. Rheum. Dis. 2010, 69 (Suppl 1), i77−82. (9) Mukhopadhyay, S.; Hoidal, J. R.; Mukherjee, T. K. Role of TNFalpha in pulmonary pathophysiology. Respir. Res. 2006, 7, 125. (10) Howarth, P. H.; Babu, K. S.; Arshad, H. S.; Lau, L.; Buckley, M.; McConnell, W.; Beckett, P.; Al Ali, M.; Chauhan, A.; Wilson, S. J.; Reynolds, A.; Davies, D. E.; Holgate, S. T. Tumour necrosis factor (TNFalpha) as a novel therapeutic target in symptomatic corticosteroid dependent asthma. Thorax 2005, 60 (12), 1012−1018. (11) Berry, M. A.; Hargadon, B.; Shelley, M.; Parker, D.; Shaw, D. E.; Green, R. H.; Bradding, P.; Brightling, C. E.; Wardlaw, A. J.; Pavord, I. D. Evidence of a role of Tumor Necrosis Factor α in refractory asthma. N. Engl. J. Med. 2006, 354 (7), 697−708. 5303

DOI: 10.1021/acs.jmedchem.6b01647 J. Med. Chem. 2017, 60, 5290−5305

Journal of Medicinal Chemistry

Drug Annotation

with the enzyme’s surface-exposed front region. J. Med. Chem. 2008, 51 (14), 4122−4149. (46) Laufer, S.; Koch, P. Towards the improvement of the synthesis of novel 4(5)-aryl-5(4)-heteroaryl-2-thio-substituted imidazoles and their p38 MAP kinase inhibitory activity. Org. Biomol. Chem. 2008, 6 (3), 437−439. (47) Laufer, S. A.; Zimmermann, W.; Ruff, K. J. Tetrasubstituted imidazole inhibitors of cytokine release: probing substituents in the N1 position. J. Med. Chem. 2004, 47 (25), 6311−6325. (48) Laufer, S. A.; Wagner, G. K.; Kotschenreuther, D. A.; Albrecht, W. Novel substituted pyridinyl imidazoles as potent anticytokine agents with low activity against hepatic cytochrome P450 enzymes. J. Med. Chem. 2003, 46 (15), 3230−3244. (49) Laufer, S. A.; Striegel, H. G.; Wagner, G. K. Imidazole inhibitors of cytokine release: probing substituents in the 2 position. J. Med. Chem. 2002, 45 (21), 4695−4705. (50) Laufer, S. A.; Wagner, G. K. From imidazoles to pyrimidines: new inhibitors of cytokine release. J. Med. Chem. 2002, 45 (13), 2733− 2740. (51) Abu Thaher, B.; Koch, P.; Schattel, V.; Laufer, S. Role of the hydrogen bonding heteroatom-Lys53 interaction between the p38alpha mitogen-activated protein (MAP) kinase and pyridinylsubstituted 5-membered heterocyclic ring inhibitors. J. Med. Chem. 2009, 52 (8), 2613−2617. (52) Kammerer, B.; Scheible, H.; Albrecht, W.; Gleiter, C. H.; Laufer, S. Pharmacokinetics of ML3403 ({4-[5-(4-Fluorophenyl)-2-methylsulfanyl-3H-imidazol-4-yl]-pyridin-2-yl}-(1-phenylethyl)-amine), a 4pyridinylimidazole-type p38 mitogen-activated protein kinase inhibitor. Drug Metab. Dispos. 2007, 35 (6), 875−883. (53) Laufer, S.; Kotschenreuther, D.; Merckle, P.; Tollmann, K.; Striegel, H. G. 2-Thio-substituierte Imidazolderivate und ihre Verwendung in der Pharmazie. WO 2004/018458, 2002. (54) Laufer, S.; Striegel, H. G.; Tollmann, K.; Albrecht, W. 2-Thiosubstituierte Imidazolderivate und ihre Verwendung in der Pharmazie. WO 2004/018458, 2004. (55) Albrecht, W.; Greim, C.; Striegel, H. G.; Tollmann, K.; Merckle, P.; Laufer, S. 2-Sulfinyl- and 2-Sulfonyl-Substituted Imidazole Derivatives and Their Use as Cytokine. WO 2006/089798, 2006. (56) Albrecht, W.; Hauser, D.; Laufer, S.; Striegel, H. G.; Tollmann, K. Imidazole Compounds Having an Antiinflammatory Effect. WO 2008/023066, 2008. (57) Laufer, S.; Wagner, G.; Kotschenreuther, D. Ones, thiones, and N-oxides: An exercise in imidazole chemistry. Angew. Chem., Int. Ed. 2002, 41 (13), 2290−2293. (58) Laufer, S. A.; Liedtke, A. J. A concise and optimized four-step approach toward 2-(aryl-)alkylsulfanyl-, 4(5)-aryl-, 5(4)-heteroarylsubstituted imidazoles using alkyl- or arylalkyl thiocyanates. Tetrahedron Lett. 2006, 47 (40), 7199−7203. (59) Griswold, D. E.; Marshall, P. J.; Webb, E. F.; Godfrey, R.; Newton, J.; DiMartino, M. J.; Sarau, H. M.; Gleason, J. G.; Poste, G.; Hanna, N. SK&F 86002: A structurally novel anti-inflammatory agent that inhibits lipoxygenase- and cyclooxygenase-mediated metabolism of arachidonic acid. Biochem. Pharmacol. 1987, 36 (20), 3463−3470. (60) Muth, F.; Gunther, M.; Bauer, S. M.; Doring, E.; Fischer, S.; Maier, J.; Druckes, P.; Koppler, J.; Trappe, J.; Rothbauer, U.; Koch, P.; Laufer, S. A. Tetra-substituted pyridinylimidazoles as dual inhibitors of p38 alpha mitogen-activated protein kinase and c-Jun N-terminal kinase 3 for potential treatment of neurodegenerative diseases. J. Med. Chem. 2015, 58 (1), 443−456. (61) Koch, D. A.; Silva, R. B. M.; de Souza, A. H.; Leite, C. E.; Nicoletti, N. F.; Campos, M. M.; Laufer, S.; Morrone, F. B. Efficacy and gastrointestinal tolerability of ML3403, a selective inhibitor of p38 MAP kinase and CBS-3595, a dual inhibitor of p38 MAP kinase and phosphodiesterase 4 in CFA-induced arthritis in rats. Rheumatology 2014, 53 (3), 425−432. (62) Bauer, S. M.; Kubiak, J. M.; Rothbauer, U.; Laufer, S. From enzyme to whole blood: sequential screening procedure for identification and evaluation of p38 MAPK inhibitors. In Kinase

human monocytes by activating Type II cAMP-dependent protein kinase. Pulm. Pharmacol. Ther. 2005, 18 (4), 277−284. (31) Tannheimer, S. L.; Sorensen, E. A.; Haran, A. C.; Mansfield, C. N.; Wright, C. D.; Salmon, M. Additive anti-inflammatory effects of beta 2 adrenoceptor agonists or glucocorticosteroid with roflumilast in human peripheral blood mononuclear cells. Pulm. Pharmacol. Ther. 2012, 25 (2), 178−184. (32) Tannheimer, S. L.; Sorensen, E. A.; Cui, Z. H.; Kim, M.; Patel, L.; Baker, W. R.; Phillips, G. B.; Wright, C. D.; Salmon, M. The in vitro pharmacology of GS-5759, a novel bifunctional phosphodiesterase 4 inhibitor and long acting beta2-adrenoceptor agonist. J. Pharmacol. Exp. Ther. 2014, 349 (1), 85−93. (33) Baker, W. R.; Cai, S.; Kaplan, J. A.; Kim, M.; Loyer-Drew, J. A.; Perreault, S.; Phillips, G.; Purvis, L. J. I.; Stasiak, M.; Stevens, K. L. Bifunctional Quinoline Derivatives. WO 2011/143105, 2011. (34) Trifilieff, A.; Keller, T. H.; Press, N. J.; Howe, T.; Gedeck, P.; Beer, D.; Walker, C. CGH2466, a combined adenosine receptor antagonist, p38 mitogen-activated protein kinase and phosphodiesterase type 4 inhibitor with potent in vitro and in vivo anti-inflammatory activities. Br. J. Pharmacol. 2005, 144 (7), 1002−1010. (35) Cuenda, A.; Rouse, J.; Doza, Y. N.; Meier, R.; Cohen, P.; Gallagher, T. F.; Young, P. R.; Lee, J. C. SB203580 is a specific inhibitor of a MAP kinase homologue which is stimulated by cellular stresses and interleukin-1. FEBS Lett. 1995, 364 (2), 229−233. (36) Muth, F.; Gunther, M.; Bauer, S. M.; Doring, E.; Fischer, S.; Maier, J.; Druckes, P.; Koppler, J.; Trappe, J.; Rothbauer, U.; Koch, P.; Laufer, S. A. Tetra-substituted pyridinylimidazoles as dual inhibitors of p38 alpha mitogen-activated protein kinase and c-Jun N-Terminal Kinase 3 for potential treatment of neurodegenerative diseases. J. Med. Chem. 2015, 58 (5), 2567−2567. (37) Dimova, D.; Iyer, P.; Vogt, M.; Totzke, F.; Kubbutat, M. H. G.; Schachtele, C.; Laufer, S.; Bajorath, J. Assessing the target differentiation dotential of imidazole-based protein kinase inhibitors. J. Med. Chem. 2012, 55 (24), 11067−11071. (38) Buhler, S.; Goettert, M.; Schollmeyer, D.; Albrecht, W.; Laufer, S. A. Chiral sulfoxides as metabolites of 2-thioimidazole-based p38alpha mitogen-activated protein kinase inhibitors: enantioselective synthesis and biological evaluation. J. Med. Chem. 2011, 54 (9), 3283− 3297. (39) Laufer, S.; Hauser, D.; Stegmiller, T.; Bracht, C.; Ruff, K.; Schattel, V.; Albrecht, W.; Koch, P. Tri- and tetrasubstituted imidazoles as p38alpha mitogen-activated protein kinase inhibitors. Bioorg. Med. Chem. Lett. 2010, 20 (22), 6671−6675. (40) Koch, P.; Laufer, S. Unexpected reaction of 2-alkylsulfanylimidazoles to imidazol-2-ones: pyridinylimidazol-2-ones as novel potent p38alpha mitogen-activated protein kinase inhibitors. J. Med. Chem. 2010, 53 (12), 4798−4802. (41) Bracht, C.; Hauser, D. R.; Schattel, V.; Albrecht, W.; Laufer, S. A. Synthesis and biological testing of N-aminoimidazole-based p38alpha MAP kinase inhibitors. ChemMedChem 2010, 5 (7), 1134−1142. (42) Ziegler, K.; Hauser, D. R.; Unger, A.; Albrecht, W.; Laufer, S. A. 2-Acylaminopyridin-4-ylimidazoles as p38 MAP kinase inhibitors: Design, synthesis, and biological and metabolic evaluations. ChemMedChem 2009, 4 (11), 1939−1948. (43) Peifer, C.; Abadleh, M.; Bischof, J.; Hauser, D.; Schattel, V.; Hirner, H.; Knippschild, U.; Laufer, S. 3,4-Diaryl-isoxazoles and -imidazoles as potent dual inhibitors of p38alpha mitogen activated protein kinase and casein kinase 1delta. J. Med. Chem. 2009, 52 (23), 7618−7630. (44) Koch, P.; Bauerlein, C.; Jank, H.; Laufer, S. Targeting the ribose and phosphate binding site of p38 mitogen-activated protein (MAP) kinase: synthesis and biological testing of 2-alkylsulfanyl-, 4(5)-aryl-, 5(4)-heteroaryl-substituted imidazoles. J. Med. Chem. 2008, 51 (18), 5630−5640. (45) Laufer, S. A.; Hauser, D. R.; Domeyer, D. M.; Kinkel, K.; Liedtke, A. J. Design, synthesis, and biological evaluation of novel Triand tetrasubstituted imidazoles as highly potent and specific ATPmimetic inhibitors of p38 MAP kinase: focus on optimized interactions 5304

DOI: 10.1021/acs.jmedchem.6b01647 J. Med. Chem. 2017, 60, 5290−5305

Journal of Medicinal Chemistry

Drug Annotation

Screening and Profiling: Methods and Protocols; Zegzouti, H.; Goueli, S. A., Eds.; Springer: New York, NY, 2016; pp 123−148. (63) Bethke, T. D.; Bohmer, G. M.; Hermann, R.; Hauns, B.; Fux, R.; Morike, K.; David, M.; Knoerzer, D.; Wurst, W.; Gleiter, C. H. Doseproportional intraindividual single-and repeated-dose pharmacokinetics of roflumilast, an oral, once-daily phosphodiesterase 4 inhibitor. J. Clin. Pharmacol. 2007, 47 (1), 26−36. (64) Hauns, B.; Hermann, R.; Hunnemeyer, A.; Herzog, R.; Hauschke, D.; Zech, K.; Bethke, T. D. Investigation of a potential food effect on the pharmacokinetics of roflumilast, an oral, once-daily phosphodiesterase 4 inhibitor, in healthy subjects. J. Clin. Pharmacol. 2006, 46 (10), 1146−1153.

5305

DOI: 10.1021/acs.jmedchem.6b01647 J. Med. Chem. 2017, 60, 5290−5305