Only a single clinical study, with CP-99,994 in post-operative dental pain, has demonstrated analgesic activity in man (Dionne et al. 1999). In this study, CP-99,994 was administered as an intravenous infusion over 5 h, starting
30 min prior to surgery (total dose 0.75 mg/kg). CP-99,994 had comparable clinical efﬁcacy to ibuprofen (Dionne et al. 1998). In contrast, the long-acting
orally active NK1 receptor antagonist aprepitant (300 mg p.o. given 2 h prior to surgery) was ineffective in post-operative dental pain (Reinhardt et al.
1998). The dental pain model is unlikely to be ideal for testing the analgesic effects of NK1 receptor antagonists, especially as these agents are not effective against acute noxious stimuli in rodents (Urban and Fox 2000). A recent study which has utilised NK1 receptor internalisation to determine the extent of substance P release in the spinal cord following dental tooth extraction in rodents (Sabino et al. 2002) demonstrated NK1 receptor internalisation in neurons in the trigeminal nucleus caudalis and cervical spinal cord (100% of neurons within lamina I and 65% in lamina II–V) within 5 min following incisor extraction. The effect was relatively short-lived with only few neurons showing internalised receptor at 60 min. This suggests that substance P may play only a minor role in mediating pain associated with tooth extraction, limited to the initial phase of pain, which is not typically assessed in the clinical studies. A more appropriate model to evaluate NK1 receptor antagonists may be dental pain associated with pulpal inﬂammation where elevated levels of substance P are found in dental pulp (Awawdeh et al. 2002).
NK1 receptor antagonists have also been evaluated in patients with neu- ropathic pain and aprepitant (300 mg p.o. for 2 weeks) was ineffective in patients with established post-herpetic neuralgia (PHN; duration of 6 months to 6 years; Block et al. 1998). Lanepitant (LY303870; 50, 100 or 200 mg p.o. b.i.d. for 8 weeks) had no signiﬁcant effect on pain intensity (daytime or night time) when compared to placebo in patients with painful diabetic neuropathy (Goldstein et al. 1999). Lanepitant (10, 30, 100, or 300 mg p.o. for 3 weeks) was also without effect in patients with moderate to severe osteoarthritis (Gold- stein et al. 1998). To date, lanepitant and aprepitant are the only NK1 receptor antagonists which have been evaluated in chronic pain conditions. The nega- tive data obtained with aprepitant in PHN is not due to insufﬁcient dose, as the same dose was shown by positron emission tomography (PET) to block central NK1 receptors in man (Hargreaves 2002). Destruction of peptidergic C-ﬁbre function has been reported in patients with long-established disease, and this loss in substance P function could therefore account, at least in part, for the failure of aprepitant to produce analgesia in patients with PHN.
Finally, clinical trials with NK1 receptor antagonists for acute migraine and migraine prophylaxis have also been disappointing. L-758,298, an intravenous pro-drug of aprepitant (20, 40 or 60 mg i.v) failed to abort migraine pain as
measured either by the time to meaningful relief or the number of patients reporting pain relief within 4 h (Norman et al. 1998). Similarly, GR205171 (25 mg i.v.; Connor et al. 1998) and lanepitant (30, 80 or 240 mg p.o., Goldstein et al. 1997) were ineffective as abortive treatments for migraine headache. Furthermore, prophylactic administration of lanepitant (200 mg p.o. per day) for 1 month had no effect on migraine frequency and severity compared to placebo (Goldstein et al. 1999). The hypothesis that NK1 receptor antagonists would be effective anti-migraine agents was based on their ability to inhibit dural plasma extravasation (Shepheard et al. 1993), an effect shared with known anti-migraine drugs such as the triptans (Williamson et al. 1997; Williamson and Hargreaves 2001) but not established as the reason for their efﬁcacy in migraine treatment.
The lack of clinical efﬁcacy of aprepitant or GR205171 in pain or migraine trials is not due to insufﬁcient dose or lack of brain penetration. At the dose used in the analgesia trials, aprepitant was found to produce more than 90% NK1 receptor occupancy using PET (Hargreaves 2002) and is anti-emetic in cancer patients following chemotherapy (Navari et al. 1999). Similarly, the dose of GR205171 employed in the migraine trial was based on NK1 receptor adequate occupancy calculated from PET studies (Connor et al. 1998). In view of these largely negative ﬁndings it has been concluded that NK1 receptor antagonists are not effective as analgesic agents in man.