Substance P Antagonists for Depression
Neuropeptides and Depression
Phase 3 Failure
Merck's NK1 receptor antagonist aprepitant matched paroxetine's antidepressant efficacy in phase 2, then failed to beat placebo in five phase 3 trials. The story reshaped how pharma approaches neuropeptide targets.
Keller et al., Neuropsychopharmacology, 2006
Keller et al., Neuropsychopharmacology, 2006
View as imageIn 1998, Merck published a study that electrified psychiatry. A phase 2 trial of MK-869 (later renamed aprepitant), a substance P (neurokinin-1) receptor antagonist, produced antidepressant effects comparable to paroxetine (Paxil) in patients with major depressive disorder. The drug worked through a mechanism completely independent of the serotonin, norepinephrine, and dopamine systems that every existing antidepressant targeted. For the first time, a neuropeptide-based approach to depression had clinical proof-of-concept.
Then it fell apart. Five subsequent phase 3 trials of aprepitant in depression failed to show efficacy over placebo. Merck abandoned the psychiatric indication. Other pharmaceutical companies that had launched NK1 antagonist depression programs followed. By 2010, the substance P hypothesis of depression appeared dead. But the story is more complicated than a simple failure narrative suggests, and its lessons still shape how the field thinks about neuropeptides and mood disorders. For the broader neuropeptide landscape in depression, see Neuropeptides and Depression: The Biology Beyond Serotonin.
Key Takeaways
- Merck's 1998 phase 2 trial showed that MK-869 (aprepitant), an NK1 receptor antagonist, produced antidepressant effects comparable to paroxetine (SSRI) in patients with major depression, with a distinct side effect profile lacking the sexual dysfunction and GI effects of SSRIs (Kramer et al., Science, 1998)
- Five phase 3 trials of aprepitant subsequently failed to demonstrate efficacy over placebo, leading Merck to abandon the psychiatric indication; paroxetine (active comparator) beat placebo in the same trials, confirming assay sensitivity (Keller et al., Neuropsychopharmacology, 2006)
- The failure was likely pharmacological, not conceptual: PET imaging showed that the phase 3 doses achieved only 80-90% NK1 receptor occupancy, while later evidence suggested near-100% occupancy was required for antidepressant activity (Ratti et al., Journal of Affective Disorders, 2017)
- Substance P is released in proportion to stress intensity and frequency; it modulates serotonin neuron firing in the dorsal raphe nucleus, providing a mechanistic link between the neuropeptide stress response and monoamine-based depression biology
- NK1 antagonists have been successfully repurposed: aprepitant is FDA-approved as Emend for chemotherapy-induced nausea, validating NK1 receptor blockade as a therapeutic strategy for a different CNS application
- Covenas et al. (2023) reviewed the repurposing of NK1 receptor antagonists as antitumor drugs, showing that the substance P/NK1 receptor system is involved in cancer cell proliferation across multiple tumor types
Substance P and the NK1 receptor in the brain
Substance P is an 11-amino-acid neuropeptide (Arg-Pro-Lys-Pro-Gln-Gln-Phe-Phe-Gly-Leu-Met-NH2) belonging to the tachykinin family. It acts primarily through the neurokinin-1 (NK1) receptor, a G protein-coupled receptor expressed on neurons throughout the brain, with particularly high density in brain regions involved in mood and stress regulation: the amygdala, hippocampus, hypothalamus, periaqueductal gray, locus coeruleus, and dorsal raphe nucleus.
In the context of depression, substance P functions as a stress mediator. Stressful stimuli trigger substance P release in the amygdala and other limbic structures, and the magnitude of release correlates with stress intensity and duration. NK1 receptor activation in these regions amplifies the emotional and physiological stress response. The logic of NK1 antagonism for depression is therefore straightforward: block the receptor that amplifies stress signaling, and you should reduce the biological substrate of stress-related depression.
Harris et al. (2022) characterized the selective G protein signaling driven by substance P-neurokinin receptor dynamics, showing that NK1 receptor activation engages different downstream pathways depending on the cellular context and the duration of agonist exposure. This complexity may partly explain why simple receptor occupancy metrics failed to predict clinical efficacy.[1]
Ge et al. (2025) directly observed human substance P interactions with the NK1 receptor at the molecular level, providing structural insights into how substance P engages its receptor and how antagonists compete for the binding site.[2]
For substance P's better-known role in pain signaling, see Substance P and Pain: The Neuropeptide That Amplifies Your Suffering and How Substance P Drives Neuropathic Pain Signaling.
The substance P-serotonin connection
NK1 antagonism does not bypass serotonin signaling; it modulates it through an upstream mechanism. Substance P neurons in the dorsal raphe nucleus (where most brain serotonin neurons originate) tonically inhibit serotonin neuron firing. When an NK1 antagonist blocks substance P's action at the dorsal raphe, serotonin neurons are disinhibited, increasing serotonergic transmission in forebrain targets.
This disinhibition mechanism was demonstrated through in vivo microdialysis studies showing that NK1 antagonists increase extracellular serotonin in the frontal cortex and hippocampus. The effect is mechanistically distinct from SSRIs (which block serotonin reuptake at the synapse) but converges on the same endpoint: more serotonin availability in mood-regulating circuits.
When combined with an SSRI, NK1 antagonists produce additive increases in serotonin release, suggesting that the two mechanisms are complementary rather than redundant. This finding, first reported in mouse microdialysis studies, raised the possibility that NK1 antagonists could augment SSRI therapy rather than replacing it.
Kertes et al. (2024) studied the role of pallidal substance P and neurokinin receptors in memory consolidation, demonstrating that the SP/NK1 system operates across cognitive and emotional circuits in ways that extend beyond simple stress amplification.[3]
The rise: Merck's phase 2 results
The 1998 Science paper by Kramer et al. reported the results of a randomized, double-blind, placebo-controlled trial comparing MK-869 (aprepitant 300 mg/day) with paroxetine (20 mg/day) and placebo in outpatients with moderate to severe major depressive disorder.
The results were striking:
- MK-869 reduced Hamilton Depression Rating Scale (HAM-D) scores by a magnitude comparable to paroxetine
- The antidepressant onset was similar in timing to paroxetine (2-4 weeks)
- MK-869 had a distinct side effect profile: no sexual dysfunction, minimal GI effects, and no weight gain, all advantages over SSRIs
- MK-869 also showed significant anxiolytic effects, consistent with substance P's role in stress and anxiety circuits
The paper concluded that substance P antagonism represented "a conceptually novel approach to the therapy of depression and anxiety." It was published in Science (not a psychiatry journal), reflecting the broader significance of the finding for neuroscience.
The fall: five phase 3 failures
Merck launched a series of large, multicenter phase 3 trials of aprepitant in major depression. All five failed to demonstrate efficacy.
Keller et al. (2006) reported the results: no statistically significant differences from placebo on the HAM-D at week 8 for either dose of aprepitant (160 mg or 80 mg daily) in any of the trials. Paroxetine (included as an active comparator) beat placebo in each of the three trials where it was included, confirming that the trials were properly designed and that the patient population was responsive to antidepressant treatment. The problem was not trial design or patient selection. Aprepitant simply did not work at the tested doses.
The failure was devastating for the field. Multiple pharmaceutical companies had initiated NK1 antagonist programs based on the phase 2 data. GlaxoSmithKline, Sanofi, Pfizer, and others had compounds in development. The phase 3 failures triggered widespread abandonment of NK1 antagonists for psychiatric indications.
What went wrong: the receptor occupancy hypothesis
The most convincing explanation for the phase 3 failure involves receptor occupancy.
The phase 2 trial used aprepitant at 300 mg/day in an older formulation that achieved high drug exposure. Between phase 2 and phase 3, Merck reformulated aprepitant into a more bioavailable capsule (the formulation eventually approved for nausea as Emend). The new formulation at 160 mg was expected to provide equivalent exposure to the old 300 mg. But PET imaging studies later showed that 160 mg of the new formulation achieved only 80-90% NK1 receptor occupancy in the brain.
Ratti et al. (2017) published a comprehensive analysis arguing that NK1 receptor antagonists were abandoned prematurely. They presented evidence that:
- The original phase 2 success at 300 mg achieved near-complete (approaching 100%) NK1 receptor occupancy
- The phase 3 doses (80-160 mg reformulated) achieved incomplete occupancy (80-90%)
- A subsequent compound (casopitant, from GlaxoSmithKline) showed antidepressant efficacy at doses that achieved near-complete occupancy, but was discontinued for unrelated hepatotoxicity concerns
- The threshold for antidepressant effect appears to require near-100% receptor occupancy, a much higher bar than the 50-80% occupancy sufficient for the anti-emetic effect
This pharmacological analysis suggests that the concept was valid but the execution was flawed. The dose was too low. The receptor needed to be nearly completely blocked, not partially blocked, for the antidepressant mechanism to work. This is pharmacologically unusual, as most GPCR-targeted drugs produce clinical effects at 50-80% receptor occupancy, but not unprecedented.
The pivot: aprepitant for nausea
While the depression program failed, Merck successfully developed aprepitant (Emend) for chemotherapy-induced nausea and vomiting (CINV). The anti-emetic application requires lower NK1 receptor occupancy than the antidepressant application, explaining why the same drug succeeded for one indication and failed for another at similar doses.
Aprepitant was FDA-approved in 2003 for CINV and is now widely used in oncology supportive care. The irony is significant: the same molecular target that was "validated" for depression and then "invalidated" by failed trials was simultaneously validated for emesis. The NK1 receptor is a legitimate therapeutic target. The challenge was always achieving sufficient receptor blockade in the right brain regions for the right indication.
What happened next: the field today
Covenas et al. (2023) reviewed the repurposing of non-peptide NK1 receptor antagonists as antitumor drugs, demonstrating that the substance P/NK1 receptor system plays a role in cancer cell proliferation, migration, and angiogenesis across multiple tumor types. This represents yet another application of NK1 antagonism that emerged after the depression program's collapse.[4]
Kokabi et al. (2023) reviewed the substance P/NK1 receptor system in diabetes, showing that this neuropeptide pathway influences pancreatic function, insulin secretion, and diabetic complications through mechanisms independent of its CNS roles.[5]
Kriska et al. (2024) discovered that cellular metabolism of substance P produces neurokinin-1 receptor peptide agonists, revealing that SP is not just one ligand but a family of metabolites, each with potentially different receptor activation profiles. This finding adds complexity to the simple "block substance P with an antagonist" model and may inform future drug design.[6]
Dionysakopoulou et al. (2023) studied the role of substance P, neurokinin A, neuropeptide Y, and cortisol in assessing stress and mood, providing a multi-neuropeptide framework that positions substance P within a network of stress-responsive peptides rather than as an isolated target.[7]
For another neuropeptide approach to anxiety that reached clinical use, see Selank: The Anxiolytic Peptide and What Research Shows. For the CRF/stress hormone axis in depression, see CRF (Corticotropin-Releasing Factor) and Depression: The Stress Hormone Connection.
Lessons for neuropeptide drug development
The substance P depression story is taught in pharmacology programs as a cautionary tale about premature enthusiasm and dose-response relationships. Several specific lessons have shaped subsequent neuropeptide drug development:
Lesson 1: Receptor occupancy thresholds vary by indication. The same receptor may require 50% occupancy for one effect and 100% for another. PET imaging to verify central receptor occupancy should precede phase 3 trials, not follow them post hoc.
Lesson 2: Reformulation can change everything. Changing a drug's formulation between phases of clinical development is common but can alter bioavailability in ways that affect efficacy. Bioequivalence for systemic exposure does not guarantee bioequivalence for central receptor occupancy.
Lesson 3: Phase 2 success does not guarantee phase 3 success. This is broadly true across drug development but was particularly stark for aprepitant, where a well-powered phase 2 trial with active comparator validation was followed by five consecutive phase 3 failures.
Lesson 4: Abandoning a mechanism after one drug's failure may be premature. The NK1 antagonist concept for depression was not disproven by the phase 3 data; it was under-dosed. Ratti et al.'s 2017 analysis suggests that adequately dosed NK1 antagonists may indeed have antidepressant effects, but no company has been willing to reinvest in the mechanism after the high-profile failure.
Lesson 5: Neuropeptide targets are not serotonin. The monoamine hypothesis of depression led to drugs (SSRIs, SNRIs) that work at 50-80% transporter occupancy. Neuropeptide receptor occupancy requirements may be fundamentally different, requiring new pharmacological frameworks rather than assumptions borrowed from monoamine pharmacology.
The Bottom Line
The substance P/NK1 receptor antagonist depression program remains one of the most instructive failures in neuropeptide drug development. Phase 2 data showed antidepressant efficacy comparable to SSRIs through a completely novel mechanism. Phase 3 trials failed, likely because reformulation between phases reduced receptor occupancy below the near-100% threshold required for antidepressant effect. The drug succeeded for nausea (lower occupancy threshold) and the receptor has since found applications in oncology and metabolic research. The story's central lesson, that neuropeptide targets may require near-complete receptor blockade for CNS effects, remains relevant for every subsequent neuropeptide psychiatric program.