I am grateful for the interest shown in my TiPS article1xNK1 (substance P) receptor antagonists – why are they not analgesic in humans?. Hill, R.G. Trends Pharmacol. Sci. 2000; 21: 244–246Abstract | Full Text | Full Text PDF | PubMed | Scopus (254)See all References1; the letters from Villanueva and from Urban and Fox make pertinent and useful observations that extend the discussion initiated in my short paper.Villanueva raises the interesting prospect of developing reliable animal tests to predict analgesic efficacy in humans by avoiding the reductionist approach, which is almost universally used today (although this is usually due to the operational constraints of what is possible and practical). He advises a holistic approach using, for example, neural ensemble recording, which is a fertile area for research. Advances in the identification of which areas of the CNS are involved in the perception of different aspects of clinical pain, using functional imaging techniques2xPain-related neurons in human cingulate cortex. Hutchison, W.D. et al. Nat. Neurosci. 1999; 2: 403–405Crossref | PubMed | Scopus (377)See all References2, together with the availability of suitable hardware, suggest that these techniques should also be applied in animal experimentation. Imaging techniques are also relevant to the investigation of those intractable pain states that Villanueva mentions as having no apparent sensory system correlates and that might originate in the CNS. It is important to note, however, that treatment of clinical pain is not yet based on a full understanding of the mechanisms of pain and that although great advances are being made in developing mechanism-based pain taxonomy as a guide to treatment, it will be some time before this is possible. Max3xIs mechanism-based pain treatment attainable? Clinical trial issues. Max, M.B. J. Pain. 2000; 1: 2–9Abstract | Full Text | Full Text PDF | PubMed | Scopus (37)See all References3 recently commented that his group have failed to find any significant differences between responses of allodynia and ongoing pain in neuropathic pain patients treated with a range of drugs, despite the fact that basic science studies strongly point to differences in the mechanisms that give rise to these symptoms. When a particular drug (be it an antidepressant, opioid, adrenoceptor agonist or NMDA receptor blocker) was effective, most of the symptoms experienced by the patient were relieved3xIs mechanism-based pain treatment attainable? Clinical trial issues. Max, M.B. J. Pain. 2000; 1: 2–9Abstract | Full Text | Full Text PDF | PubMed | Scopus (37)See all References3, which suggests that mechanism-based differences in pain symptoms might often be small. This has obvious implications in deciding which animal tests might or might not be predictive.Urban and Fox re-address the issue of: (1) whether appropriate animal tests were used to select those tachykinin NK1 receptor antagonists that have so far been studied as potential analgesics in humans and (2) whether these agents were able to cross the blood-brain barrier. This second point is important and received only little attention in my recent article1xNK1 (substance P) receptor antagonists – why are they not analgesic in humans?. Hill, R.G. Trends Pharmacol. Sci. 2000; 21: 244–246Abstract | Full Text | Full Text PDF | PubMed | Scopus (254)See all References1; however, my colleagues and I have dealt with this issue extensively in other publications4.xIn vitro and in vivo predictors of the anti-emetic activity of tachykinin NK1 receptor antagonists. Rupniak, N.M.J. et al. Eur. J. Pharmacol. 1997; 326: 201–209Crossref | PubMed | Scopus (100)See all References, 5.xNeurokinin antagonists. Rupniak, N.M.J. and Hill, R.G. : 135–155See all References. I strongly agree with the general points that Urban and Fox make but disagree with the specific suggestion that the selection of MK869 for clinical study was in some way flawed because we did not establish that this compound was active in guinea-pig antinociception assays before starting human analgesia trials. Although this particular compound was not subjected to such experiments, we have shown that analogues of this compound were active in guinea-pig antinociception tests (see Fig. 2 in Ref. 6xDiscrepant results from preclinical and clinical studies on the potential of substance P–receptor antagonist compounds as analgesics. Boyce, S. and Hill, R.G. : 313–324See all References6) and that MK869 itself was brain penetrant and active in tests in a range of species, including ferret, gerbil and rabbit4.xIn vitro and in vivo predictors of the anti-emetic activity of tachykinin NK1 receptor antagonists. Rupniak, N.M.J. et al. Eur. J. Pharmacol. 1997; 326: 201–209Crossref | PubMed | Scopus (100)See all References, 5.xNeurokinin antagonists. Rupniak, N.M.J. and Hill, R.G. : 135–155See all References, 6.xDiscrepant results from preclinical and clinical studies on the potential of substance P–receptor antagonist compounds as analgesics. Boyce, S. and Hill, R.G. : 313–324See all References, 7.xThe novel NK1 receptor antagonist MK-0869 (L-754,030) and its water soluble phosphoryl prodrug, L-758,298, inhibit acute and delayed cisplatin-induced emesis in ferrets. Tattersall, F.D. et al. Neuropharmacology. 2000; 39: 652–663Crossref | PubMed | Scopus (104)See all References. Perhaps most importantly, it has been shown that similar doses of MK869 to those found to be inactive in analgesic tests in humans were effective in the treatment of cisplatin-induced emesis in humans. Such an anti-emetic action requires high levels of receptor occupancy and brain penetration4.xIn vitro and in vivo predictors of the anti-emetic activity of tachykinin NK1 receptor antagonists. Rupniak, N.M.J. et al. Eur. J. Pharmacol. 1997; 326: 201–209Crossref | PubMed | Scopus (100)See all References, 7.xThe novel NK1 receptor antagonist MK-0869 (L-754,030) and its water soluble phosphoryl prodrug, L-758,298, inhibit acute and delayed cisplatin-induced emesis in ferrets. Tattersall, F.D. et al. Neuropharmacology. 2000; 39: 652–663Crossref | PubMed | Scopus (104)See all References. I therefore do not believe that the guinea-pig is the ‘right species’ in which to evaluate NK1 receptor antagonists as potential analgesics or that agents that are active in this species will be more likely to have predictable analgesic effects in humans. However, I do agree that some NK1 receptor antagonists tested in human pain studies could have been predicted, on the basis of lack of brain penetration or short duration of action in animal tests, to be unlikely to produce analgesia in humans4xIn vitro and in vivo predictors of the anti-emetic activity of tachykinin NK1 receptor antagonists. Rupniak, N.M.J. et al. Eur. J. Pharmacol. 1997; 326: 201–209Crossref | PubMed | Scopus (100)See all References4. Reliable but negative human pain trial data exists only for very few compounds and I would be delighted if SDZNKT343 were to advance to clinical efficacy studies, as suggested by Urban and Fox, to help resolve the issue.I also agree that only a limited range of pain states has been studied so far with NK1 receptor antagonists and it would be helpful if a wider range of conditions could be examined. In particular, it can be argued that these agents might be more effective at controlling long-term hyperalgesia or progressive hypersensitivity than in reducing the intensity of existing pain8xNeurokinin antagonists as potential agents for use in pain management. Ma, Q-P. and Hill, R.G. Curr. Opin. CPNS Invest. Drugs. 1999; 1: 65–71See all References8. However, if one takes a nihilistic approach it could also be argued that the post-operative pain resulting from third molar extraction provides a reliably predictive assay and that any drug that is going to be clinically useful for treating pain should be active in this test. The choice of clinical paradigm for analgesic drug evaluation is clearly a pivotal topic and is now receiving appropriate consideration3xIs mechanism-based pain treatment attainable? Clinical trial issues. Max, M.B. J. Pain. 2000; 1: 2–9Abstract | Full Text | Full Text PDF | PubMed | Scopus (37)See all References3.