KW-6356

KW-6356 is a non-xanthine blocker of the adenosine A2A receptor [1] from Kyowa Kirin, the company that developed istradefylline [2], the only A2A antagonist to reach the market for Parkinson's disease [3]. Its international nonproprietary name is sipagladenant [2]. Istradefylline is approved in Japan and the United States as an add-on to levodopa for people with "off" episodes [1]. Across nine trials and 2,727 patients it reduced off time, a result with signs of publication bias, but did not significantly change motor scores [7], and it had not shown a significant effect on its own [8]. KW-6356 binds about 100 times more tightly, for longer, and also suppresses the receptor's background activity [1].

Alone, in early Parkinson's disease. A phase 2a trial at 46 sites in Japan randomized 168 people with early, untreated disease to 3 mg, 6 mg or placebo once a day for 12 weeks. Motor scores (MDS-UPDRS Part III) fell by 5.37 points on 3 mg, 4.76 on 6 mg and 3.14 on placebo. The differences from placebo, −2.23 and −1.62 points, had confidence intervals that crossed zero (−4.86 to 0.40 and −4.19 to 0.94); the study was exploratory, had no formal sample-size calculation and analyzed the per-protocol set. Patients' own ratings of daily motor activities (Part II) worsened by 1.0 point on placebo and stayed level on KW-6356, a difference of −1.32 points at 3 mg whose interval just excluded zero. Clinicians rated 47% and 52% improved on 3 and 6 mg against 37% on placebo; quality of life scores barely moved [3].

As an add-on to levodopa. A phase 2b trial randomized 503 Japanese patients already on levodopa to placebo, 3 mg or 6 mg for 24 weeks. According to the conference abstract, both doses improved motor scores more than placebo (p = .006 for 3 mg, p = .049 for 6 mg) and reduced daily off time in patients with wearing-off, and in a post-hoc analysis 6 mg improved a Parkinson's sleep scale (PDSS-2). The abstract gives no effect sizes or adverse-event numbers, and the trial has not been published in full [4].

Discontinued. In July 2022 Kyowa Kirin announced that it had stopped developing KW-6356. Its notice says the phase 2 trials showed efficacy alone and with levodopa, and gives as reasons the regulatory outlook for global development, the difficulty of development and the time it would take to reach the market [6]. A 2023 review notes that it might yet prove useful in cancer, where other A2A antagonists are being tested, but no such trial has been reported [2].

Every study of KW-6356 so far comes from Kyowa Kirin or investigators it paid, and all the patients were Japanese [3][4].

Why A2A. In the striatum, A2A receptors sit on the cell bodies and terminals of the neurons of the indirect pathway, the circuit that holds movement back, and activating them makes those neurons more excitable. Losing dopamine removes the brake that D2 receptors put on the same neurons, so the pathway becomes overactive; blocking A2A receptors is meant to damp it without acting on dopamine receptors at all [1]. A2A receptor numbers also rise as Parkinson's disease progresses, particularly in patients with motor complications [3].

Binding. KW-6356 binds the human A2A receptor with a pKi of 9.93 (about 0.12 nM), and with similar affinity in mouse, rat, dog and marmoset striatum (pKi 9.49–10.00). Its affinity for human A1, A2B and A3 receptors is 270 to 3,500 times lower. At 10 µM it displaced half the binding at the dopamine D5 receptor and less than 30% at every other receptor, transporter and ion channel screened, and it did not inhibit MAO-A, MAO-B or COMT. Once bound it comes off slowly, with a dissociation half-life of about 46 minutes at 25 °C [1]. Its affinity is about 100 times that of istradefylline [1], whose Ki in a 2023 compilation is 36 nM [2].

Insurmountable antagonism and inverse agonism. In human cells that make A2A receptors naturally, istradefylline shifted the response to the agonist CGS21680 to higher concentrations without lowering its ceiling, while pretreatment with KW-6356 at 0.1–1 nM also flattened the maximal response, so that more agonist could not overcome the block. In cells engineered to express the receptor, KW-6356 lowered its baseline activity (pEC50 8.46), which makes it an inverse agonist [1].

Crystal structures explain the difference. KW-6356 makes many more contacts in the binding pocket than istradefylline, whose xanthine core sits much as caffeine's does. Its furan ring reaches the bottom of the pocket and holds two residues, His250 and Trp246, that must move for the receptor to switch on; modeled into the active receptor it clashes with His250, so it can only bind the inactive shape. Its oxane group fills a deep hydrophobic pocket and its methylnicotinamide end holds the loops that form the pocket's lid, which the authors propose keeps it bound and makes its block insurmountable [1]. Its developers proposed that the inverse agonism is why it worked on its own in early disease when istradefylline had not [3].

In parkinsonian marmosets. In marmosets made parkinsonian with MPTP, oral KW-6356 reversed motor disability in a dose-dependent way up to 1 mg/kg, by more than istradefylline did, and repeated doses caused little dyskinesia in animals already primed to show it [8]. At 1 mg/kg, a single dose strengthened the motor effect of every L-DOPA dose tested (2.5–10 mg/kg). In six marmosets primed with L-DOPA to show dyskinesia, KW-6356 alone raised activity with almost no dyskinesia (score 0.5 against 0.17 on vehicle, on a 0–4 scale), and with the optimal L-DOPA dose it left dyskinesia where L-DOPA alone put it (3.67). Taken daily for 21 days with a low L-DOPA dose, it raised that dose's dyskinesia from 0.5 to 2.0, still below the 3.67 of the optimal dose [9].

KW-6356 was given by mouth once a day in every human study [3][5].

Single doses. In two phase 1 studies in healthy men, 1 to 60 mg on an empty stomach peaked within about 0.75 to 3 hours, and peak levels and total exposure rose in proportion to dose, from 122 ng/mL after 1 mg to 5,460 ng/mL after 60 mg. Levels fell fast over the first six hours and then slowly, with a terminal half-life of 16 to 30 hours [5].

Daily dosing. Taken after breakfast, it peaked at about 4 hours and reached steady state within a week, at levels about 2 to 2.5 times those after the first dose; on 6 mg a day, peak levels were about 850 to 1,180 ng/mL and troughs about 450 to 650 ng/mL. Japanese and White men had similar KW-6356 levels on 6 mg a day, though Japanese men had more of the metabolite M6 [5].

The metabolite M6. M6 peaks 12 to 24 hours after a dose and is as potent an inverse agonist as KW-6356 in vitro, but its exposure is 4 to 16 times lower, and its developers judged that it adds little to the effect [5]. A population model built from healthy volunteers and Parkinson's patients described KW-6356 and M6 with one compartment each; food changed the rate of absorption, serum albumin the clearance of KW-6356, and body weight its volume of distribution and the clearance of M6, but none made a clinically meaningful difference to exposure [10].

Healthy volunteers. Single doses up to 60 mg and 24 mg a day for 14 days were judged well tolerated, and most adverse events were mild or moderate, but most people given KW-6356 had one. Insomnia dominated: after single doses it affected 4 of 6 men at 1 mg, 4 of 6 at 3 mg and all 6 at 10 mg, with none on placebo, and 50%, 83% and 100% at 21, 42 and 60 mg against 33% on placebo; on 6 mg a day for a week, 5 of 6 men had it. The investigators linked it to a wake-promoting effect seen in rats. Fever, feeling hot, sweating and palpitations also occurred. One of six men on 24 mg a day developed rhabdomyolysis, a serious event judged drug-related, and one White man on 6 mg a day stopped because of anxiety and a panic reaction, judged unrelated [5].

Patients. In the 12-week monotherapy trial, 47.3% and 50.0% of patients on 3 and 6 mg had at least one adverse event, against 43.6% on placebo, and drug-related events were about as common as on placebo (34.5%, 27.6% and 29.1%). Constipation was the most frequent (7.3% and 10.3%, against 1.8%), followed by nasopharyngitis, which was equally common on placebo; nausea, sleepiness and insomnia each affected two patients on 6 mg. Three patients on 3 mg had serious events, pneumonia, rhabdomyolysis and a brain hemorrhage, and the investigators could not rule out that the last two were caused by the drug. Five patients on KW-6356 and none on placebo stopped because of adverse events. There were no deaths and no meaningful changes in laboratory tests, vital signs, ECGs, cognitive screening or suicidality ratings [3]. The add-on trial's abstract describes it as well tolerated without giving figures [4]; the phase 1 report says insomnia, nasopharyngitis and constipation were among the common adverse events in both phase 2 trials [5].

Drug interactions. In company laboratory studies, CYP3A4/5 did most of the work of breaking KW-6356 down, and KW-6356 itself inhibited CYP2C8, CYP2B6 and CYP3A4/5, lowered CYP1A2 and tended to induce several other CYP enzymes in human liver cells [5]. The patient trials excluded people taking drugs that act on CYP3A4/5 [3]. A clinical interaction study with midazolam, caffeine and rosuvastatin, probes for CYP3A4/5, CYP1A2 and the BCRP transporter, was completed in 2019 but its results have not been posted [11].