Basic/translational Development of Forthcoming Opioid and Non-Opioid Targeted Pain Therapeutics

Opioids

Opioids represent an efficacious therapeutic modality for some, but not all pain states. Singular reliance on opioid therapy for pain management, has limitations and abuse potential has deleterious consequences for patient and society. Our understanding of pain biology has yielded insights and opportunities for alternatives to conventional opioid agonists. The aim is to have efficacious therapies, with acceptable side effect profiles and minimal abuse potential, which is to say an absence of re-enforcing activity in the absence of a pain state. The present work provides a non-exclusive overview of current drug targets and potential future directions of research and development. We discuss channels activators and blockers, including: sodium channel blockers, potassium channel activators and calcium channel blockers; glutamate receptor targeted agents, including: NMDA, AMPA and metabotropic receptors). Further, we discuss therapeutics targeted at GABA, alpha2 adrenergic and opioid receptors. We also considered antagonists of angiotensin 2 and Toll receptors, and agonists/antagonists of adenosine, purine receptors. And cannabinoids. Novel targets considered are those focusing on lipid mediators and anti-inflammatory cytokines. Of interest is development of novel targeting strategies which produce long-term alterations in pain signaling, including viral transfection and toxins. We consider issues in the development of drugable molecules, including preclinical screening. While there are examples of successful translation, mechanistically promising pre-clinical candidates may unexpectedly fail during clinical trials because the preclinical models may not recapitulate the particular human pain condition being addressed. Molecular target characterization can diminish the disconnect between preclinical and humans’ targets, which should assist in developing non-addictive analgesics.

INTRODUCTION

The management of pain is a clinical imperative. Aside from humanistic concerns, failure to adequately control pain has negative consequences in terms of system biology. Opioids, through their potent modulatory effect mediated via canonical receptors on pain processing, have been, and remain, an essential component of pain management. Nevertheless, reliance on this therapeutic approach has limitations and deleterious consequences to the patient and society. Opioid misuse is an expanding crisis with over 36,000 deaths due to opioid overdose in 2015 alone.^1,2 Pharmaceutical companies have pursued abuse-deterrent opioid formulations.^3,4 While these formulations reduce the possibility of the content of the pills being extracted, the underlying properties of the pharmaceutical agent (i.e. opioids) remain the same and extraction deterrent systems are subject to being overcome.^5–8

Our understanding of systems that mediate and regulate nociceptive processing has yet to produce a recognized alternative to opioids. Advances in pain biology has, however, yielded remarkable insights and opportunities. We will provide an overview of salient areas of research that focus on current advances in pharmacological target. Meaningful advances in drug therapy must consider not only i) analgesic efficacy, but as well ii) therapeutic ratio (separation of pain relief from side effects); iii) constancy of response over extended use (e.g. tolerance); iv) lack of positive reinforcing properties in the absence of a pain state. Due to space restriction, this review must be considered a non-exclusive overview of advances in terms analgesic targets.

PAIN PHENOTYPES

Pain is an aversive state that reflects the perceptual covariates of events that arise from stimuli of sufficient intensity to induce tissue damage or which otherwise mimic the activity induced by such stimuli, as in nerve injury. It is heuristically useful to think of mechanisms generating the aversive condition associated with afferent stimulation as having 4 elements.

1. Acute nociception in which an acute, non-injuring, high intensity stimulus activates small unmyelinated and myelinated afferents, driving intensity linked excitation of second order dorsal horn projection neurons, leading to a stimulus-linked pain report/escape.

2. Following tissue injury and inflammation, hyperalgesia occurs at the injury site, causing an enhanced response to moderate stimuli, and an enlarging receptive field, including areas not injured, resulting in a 2^nd hyperalgesia/allodynia. This phenotype reflects a peripheral sensitization (development of ongoing activity, and a left shift in the intensity response relationship at the terminal) and a central /spinal sensitization (heightened excitability of the primary afferent terminal and second order neurons causing an enhanced discharge to a given afferent input).

3. Injury to the peripheral nerve resulting in ongoing dysesthesias and enhanced sensitivity to light touch and modest changes in temperatures (allodynia), associated with reactive changes in the afferent axon, DRG and dorsal horn (typically reflecting a loss of inhibitory regulation).

4. Following persistent inflammation and tissue injury, the evolving pain state displays characteristics suggesting the development of a nerve injury phenotype, e.g. an acute to chronic pain transition.

The biology of these above states has been reviewed in detail elsewhere.^9–11 These comments importantly emphasize that a pain condition may represent multiple mechanistic phenotypes. Accordingly, the regulation of the encoding and trafficking of the nociceptive stimulus to higher centers may reflect a role for engaging multiple targets.

ISSUES IN ANALGESIC DRUG DEVELOPMENT

Demonstration of target analgesic efficacy

Development of analgesic drugs with known targets and mechanisms of action can employ models of target engagement, such as in silico and in vitro modeling (e. g. opioids and COX inhibitors) which can move a drug with some predictability into a behavioral assessment. Novel targets often arise based on association of the target with specific systems, but their efficacy in regulating the pain state requires a sense of what role that target plays in mediating the behaviorally defined pain construct. Preclinical behavioral models provide such insights. Detailed reviews of preclinical models that focus on events secondary to inflammation (acute and chronic) and nerve injury (mono and poly neuropathies) with their strengths and shortcomings have been provided elsewhere.^12,13 While instances of failure of the predictive models have been discussed (as is true for virtually every translational system in biology), mechanistic studies have made a number of valid predictions of clinical efficacy ranging from COX inhibitors to antimigraine drugs.⁹ Several issues regarding preclinical models are noted.

1. Each behavioral model has mechanistic components particular to that system. Convergent results from multiple models and comparable dose effect relationships increase the likelihood of assessing mechanisms relevant to the human state.

2. Preclinical models have long examined a single sex, for several reasons including economy and the belief there is little difference between the sexes. Numerous instances at the behavioral and mechanistic level can now be cited to dispute this assertion.^14,15

3. Many models’ threshold measurements. Alternative models employ “spontaneous behaviors”, including general activity, rearing, weight bearing and gait as markers of an aversive condition.¹⁶ There is also an understanding that if there is an aversive condition generated by an injury, a drug that has no intrinsic rewarding property but which serves to diminish that pain state will in fact acquire a positive reinforcing property in the presence of the pain state. Such “conditioned place preference” models have an important place in current drug evaluations.^16,17

4. While preclinical analgesic drug evaluation has been largely successful in rodents, characterization of issues of analgesic efficacy and tolerability may also be achieved through naturally occurring pathologies in companion animals, notably dogs. The incidence of canine osteoarthritis and osteosarcoma provides an important way station in defining efficacy in controlled trials using validated inventories and neurological assessments.¹⁸ While safety-toxicology studies in such animals are routinely part of an IND package during drug development, there may be an advantage to pursuing efficacy studies as well. Such information is pivotal in the development of veterinary analgesic products and their approval by the US-FDA-veterinary division to manage the pain states in this patient population. The predicted spending on analgesics for pets alone was predicted to be ~$335 million in 2011, so there is a secondary market that can incentivize additional testing in the veterinary patient.¹⁹

5. Human experimental models initiating a local injury (e.g. ultraviolet B irradiation, thermode burn) or afferent stimulation (capsaicin) are increasingly used to determine efficacy of both new and existing analgesics. Their apparent ability to demonstrate efficacy with known analgesics provides some validation of their sensitivity,²⁰ and to define a drug effect and corresponding side effects at the effective dose.

6. The following commentary considers a variety of targets and comments on systemic and neuraxial routes of delivery, reflecting the fact that drug effects upon pain processing frequently reflect an action at the first order synapse. It must be stressed that these discussions do not raise issues of safety. This commentary is particularly relevant for neuraxial drugs where appropriate assessment of neuraxial safety must be undertaken prior to such drug implementation.²¹

7. Finally, it is challenging to find a drug target that can alter a pain state with a favorable therapeutic ratio (e.g. little or no effects upon mentation, arousal or motor function). An important concern we must now consider is that the drug target is not a mediator of positive reinforcement.

Assessment of abuse liability

An important issue in developing novel analgesics is to overcome the potential for abuse. If the drug acts upon components of systems associated with positive re-enforcement or possessive of stimulant and/or sedative effects, suspicion of its potential for abuse must be elevated. Such examples, at present, include drugs interacting with opioid receptors, CNS depressants (GABAA receptors); CNS stimulants (increased dopamine release/block re-uptake; nicotine), hallucinogens, glutamate antagonists (ketamine) and cannabinoids.^22,23 It seems reasonable that a molecule lacking CNS bioavailability would show a reduced likelihood of having a reinforcing property (e.g. loperamide²⁴), but now even large molecules are considered not to be excluded as having potential liability.²⁵ To this end, locomotor, reinforcing and dependence-producing effects of the agent must be routinely assessed. A variety of strategies are considered relevant and have been effective in predicting human-drug behavior including self-administration, drug discrimination and conditioned place preference paradigms.^{26, 27,28}

SURVEY OF CURRENT TARGETS OF PAIN THERAPEUTICS

In the following sections, we will consider several current drug targets and potential future directions of research and development. Figure 1 summarizes these targets as they reflect upon actions at the level of the peripheral terminal and central sites. Table 1 summarizes those agents that moved into clinical trials.

Figure 1. Potential Opioid and Non-Opioid Targeted Drugs.

Figure showing some of the key drug targets that act as peripheral (peripheral afferent neuron – peripheral terminal) and central (dorsal horn) targets. Cav - voltage-dependent calcium channel; TRPV - transient receptor potential cation channel; Nav – voltage-gated sodium channel; K - potassium channel; GPCRs - G-protein-coupled receptors; NMDA - N-methyl-d-aspartate; AMPA - α-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid; GABA - γ-aminobutyric acid; α2 – adrenergic alpha 2 receptor; CB1 - cannabinoid receptor 1; mGlu - metabotropic glutamate receptor

Table 1.

Studies registered on Clinical Trials.gov

Name of the drug	Target	Registration Number	Testing	Indication	Phase of Clinical Trial	Status of the Study
Eslicarbazepine acetate (ESL)	Voltage-Gated Sodium Channel (VGSC) antagonist	NCT00980746	assess the efficacy of ESL	painful diabetic neuropathy	Phase 2	completed
Eslicarbazepine acetate (ESL)	Voltage-Gated Sodium Channel (VGSC) antagonist	NCT00981227	assess the efficacy of ESL	postherpetic neuralgia	Phase 2	completed
Eslicarbazepine acetate (ESL)	Voltage-Gated Sodium Channel (VGSC) antagonist	NCT01820585	3 different doses of ESL vs. placebo	fibromyalgia	Phase 2	completed
Eslicarbazepine acetate (ESL)	Voltage-Gated Sodium Channel (VGSC) antagonist	NCT01124097	assess the efficacy of ESL	postherpetic neuralgia	Phase 3	terminated
Eslicarbazepine acetate (ESL)	Voltage-Gated Sodium Channel (VGSC) antagonist	NCT01129960	assess the efficacy of ESL	painful diabetic neuropathy	Phase 3	terminated
Neosaxitoxin (NeoSTX)	Voltage Gated Sodium Channels (VGSC) antagonist	NCT01786655	Neosaxitoxin (NeoSTX) alone or in combination with bupivacaine with or without epinephrine	safety in healthy volunteers	Phase 1	completed
Tetrodotoxin (TTX)	Voltage gated sodium channels (VGSC) antagonist	NCT01655823	multiple dose levels of tetrodotoxin (TTX) vs. placebo	neuropathic pain	Phase 2	terminated
Z160	Selective N-type Calcium Channel (Cav 2.2) Blocker	NCT01757873	compare Z160 and placebo	postherpetic neuralgia	Phase 2	completed
Z160	Selective N-type Calcium Channel (Cav 2.2) Blocker	NCT01655849	compare Z160 and placebo	lumbosacral radiculopathy	Phase 2	completed
Ifenprodil Tartrate	NMDA receptor antagonist	NCT01896388	confirm whether ifenprodil tartrate is effective in the treatment of adolescents PTSD patients	posttraumatic stress disorder	Phase 1/2	currently recruiting participants
Etifoxine	GABAA receptor agonist Agonist at β2 and β3 subunit of the GABAA receptor complex	NCT02147548	Confirm the effect of etifoxine and lorazepam on vigilance and cognitive functions in the elderly	healthy volunteers	Phase 3	completed
ADL5859	delta-opioid agonist	NCT00993863	assess the efficacy and safety of ADL5859 compared with placebo and active control (ibuprofen)	acute dental pain	Phase 2	completed
ADL5859 ADL5747	delta-opioid agonist	NCT00979953	assess the efficacy of ADL5859 vs. placebo and ADL5747 vs. placebo	osteoarthritis of the knee	Phase 2	Completed
TRV130 (olicerdine)	Opioid mu receptor agonist Beta arrestin inhibitor	NCT02335294 NCT02820324	analgesic efficacy IV oliceridine compared with placebo	acute pain after abdominoplasty	Phase 2 Phase 3	completed
TRV130 (olicerdine)	Opioid mu receptor agonist Beta arrestin inhibitor	NCT02100748 NCT02815709	analgesic efficacy of IV TRV130 vs placebo	bunionectomy	Phase 2 Phase 3	completed
TRV130 (olicerdine)	Opioid mu receptor agonist Beta arrestin inhibitor	NCT02656875	safety evaluation of TRV130 in patients with acute pain	moderate to severe pain caused by medical conditions or surgery	Phase 3	currently recruiting participants
TRV130 (olicerdine)	Opioid mu receptor agonist Beta arrestin inhibitor	NCT02520297	analgesic efficacy of TRV130 for moderate to severe acute pain	fracture pain	Phase 2	terminated
EMA401	Angiotensin 2 receptor (AT2R) antagonist	NCT02435199	compare EMA401 300 mg and placebo	diabetic neuropathies	Phase 2	withdrawn
EMA401	Angiotensin 2 receptor (AT2R) antagonist	NCT02426411	compare two different doses of EMA401 and placebo	postherpetic neuralgia	Phase 2	withdrawn
EMA401	Angiotensin 2 receptor (AT2R) antagonist	NCT03094195	3 different doses vs. placebo	postherpetic neuralgia	Phase 2	not yet open for recruitment
AF219 (gefapixant)	P2X3 receptor antagonist	NCT02349425	crossover, dose escalation study of gefapixant (AF-219)	refractory chronic cough	Phase 2	completed
AF219 (gefapixant)	P2X3 receptor antagonist	NCT01432730	effectiveness of gefapixant in reducing daytime objective cough frequency	idiopathic chronic cough	Phase 2	completed
AF219 (gefapixant)	P2X3 receptor antagonist	NCT02397460	crossover study in healthy and chronic cough subjects	chronic cough	Phase 2	completed
AF219 (gefapixant)	P2X3 receptor antagonist	NCT02476890	crossover study in healthy and chronic cough subjects	chronic cough	Phase 2	completed
Resiniferatoxin (RTX)	Transient Receptor Potential Vanilloid 1 (TRPV1) receptor agonist	NCT00804154	safety of intrathecal RTX	cancer pain	Phase 1	suspended participant recruitment
Substance P-Saporin	Neurokinin 1 receptor (NK1R) antagonist	NCT02036281	8 different doses intrathecally	terminally ill patients, histologically-confirmed advanced cancer, for intractable pain	Phase 1	currently recruiting participants

Sodium channel blockers/Potassium channel activators

Axon excitability depends directly upon voltage gated sodium channel while activation of potassium channels produces hyperpolarization reducing membrane excitability. These represent potential peripheral targets for altering afferent transmission.

Sodium channels

Voltage-gated sodium channel (Na_v are the target of all clinical "local anesthetics"²⁹ Nine Na_v isoforms with distinct activation properties and tissue distributions have been identified: Na_v1.1 and Na_v 1. (large DRG/axons), Na_v1.4 and Na_v 1.5 (skeletal and cardiac muscle), and Na_v 1.7,- 1.(small sensory DRGs / afferents).³⁰ Following inflammation and nerve injury, increases in small afferent Na_v (Nav1.3, 1.7, 1.8 and 1.9) expression are believed to underlie ectopic afferent traffic and increased responsiveness.^{30, 31,31,32}

Specific confirmation of the role of Nav 1.7 in human pain processing is based on the phenotype of naturally occurring gain and loss of function mutations in Nav1.7 channels wherein those expressing these mutations respectively show pronounced increased and decreased pain states.^33,34

While local anesthetics given perineurally and neuraxially produce conduction block anesthesia, systemic anesthetics such as IV lidocaine have surprisingly selective anti-hyperpathic effects in a variety of preclinical models and human pain states at concentrations that do not produce a general conduction block, suggesting a differential sensitivity of systems related to facilitated states after tissue and nerve injury.^35,36

Future development of the sodium channel blocking drugs focuses on the role of selective blockers of channels expressed on nociceptive linkages. Clinically used local anesthetics (amide and ester) do not selectively block these channels,³⁷ although several isoforms are sensitive to puffer fish toxin, tetrodotoxin (TTX) (Nav1.1–7), with the remainder resistant to TTX.^37,38 Toxin-based sodium channel blockers, neosaxitoxin and tetrodotoxin, after perineural and intrathecal delivery demonstrated long-lasting nerve blocks and surprisingly after systemic delivery in human and animal models.^39,40,38,41 (Table 1). As regards selective channel antagonists, intrathecally-delivered, toxin-based Nav1.7⁴² and 1.8⁴³ inhibitors have shown preclinical efficacy in models of inflammation and nerve injury, with a favorable therapeutic ratio. Development of systemically bioavailable, small molecule, channel selective antagonists as analgesics have faced challenges.^44,45 Clinical work with oral targeted, sodium channel selective blockade was negative,⁴⁶ although promising results from multi-center studies in post-herpetic neuralgia and primary erythromelalgia have been reported.⁴⁷ Loss of response to local anesthetics (e.g. tolerance or tachyphylaxis) has been reported after neural blocks, but the phenomena does not appear to be robust.⁴⁸

An interesting application of the specific association of TRPV1 with pain afferents has been the use of protonated local anesthetics such as QX314, which are able to enter the otherwise impermeant axon membrane through TRPV1 channels upon their activation by capsaicin, and result in function block of the sodium channel in the TRPV1 (+) afferent axon.⁴⁹ It is now understood that lidocaine by itself is a TRPV1 agonist and can promote passage of the protonated form,⁵⁰ allowing quaternary lidocaine (QX314) to enter the TRPV1-bearing axon and selectively block the Na_v channel, resulting in specific block of TRPV1 (+) primary afferents.⁵⁰

Potassium channels

There are 4 major families of K channels (Voltage-gated (Kv), Calcium-activated (K Ca); Inwardly rectifying (Kir); Two P domain (K2P) potassium channels, which, when activated lead to membrane hyperpolarization through increased potassium conductance. Genetic analyses illustrate that variations in several K⁺ channel genes are relevant to the risk for persistent pain after injury (KCNS1-Kv9.1, GIRKs-Gir, TRESK-K2P18.1), increased pain sensitivity (KCNS1, GIRKs), and analgesic efficacy of G protein coupled receptors (GIRK2).⁵¹ Inwardly rectifying, ATP sensitive potassium (K-ATP) channels are widely expressed in numerous cell types including neurons and are linked to anti-allodynic and anti-hyperalgesic activity. ATP sensitive potassium channel agonist mediated antinociceptive effects are reversed with pre-treatment with ATP-sensitive K+ channel blockers^52,53

Interestingly, autoantibodies targeting Kv channels can lead to neuronal hyperexcitability and a pain state.⁵⁴ Increasing potassium channel expression and potassium conductance via receptor channel agonists assists in hyperpolarizing (normalizing) otherwise enhanced axon, DRG and terminal excitability, resulting in anti-hyperalgesic actions.^55,56

Calcium channel blockers

Movement of calcium into the cell represents a significant source of charge leading to membrane depolarization, while increased intracellular calcium leads to activation of a variety of kinases that phosphorylate: enzymes, channels (lower threshold for activation and increasing ion permeability) and receptors, resulting in hyperalgesic states.^57–60 One source of this intracellular calcium is a variety of high and low voltage gated calcium channels (VGCC): high-VGCCs include: L-(Ca_V1.1–4), P/Q-(Ca_V2.1), N-(Ca_V2.2) and R-(Ca_V2.3) type channels; low-VGCCs include T-type (Ca_V3.1–3)^61,62 These are transmembrane channels, composed of multiple subunits endowing members of each family with distinguishing properties of voltage gating and antagonist pharmacologies. They are located on primary afferents and postsynaptic membranes in spinal dorsal horn..⁶³

N-type channel (Cav 2.2)

The N-type calcium channel is present on presynaptic nerve terminals in the superficial dorsal horn and dorsal root ganglia. Upregulation occurs following peripheral nerve injury.⁶⁴ Ziconotide, is an N-type VGCC blocker,⁶⁵ possessing potent anti-hyperpathic properties in rodents and humans when administered intrathecally as a bolus or an infusion^{66, 67} and is without tachyphylaxis (tolerance).⁶⁸ Although ziconotide remains the only approved N type channel blocker, there are efforts to develop new peptides and small molecules^65,69 and to alter nociceptive properties of N-type VGCC function by hindering its membrane trafficking.^{70, 71} In humans, a systematically active N-type calcium channel blocker (Z160) failed in Phase 2 clinical trials in treatment of post-herpetic neuralgia and lumbosacral radiculopathy.⁷² (Table 1.)

L-type channel (Cav 1)

Channels are largely present post synaptically and are considered to play a possible role in maintaining facilitated states. Intrathecal delivery preclinically of channel blockers (nifedipine, verapamil and benzothiazepines) have shown efficiency in altering injury induced hyperpathia.⁷³

T-type channel (Cav3.2)

T-type calcium channels are present in the dorsal horn and channel blockers, such as ethosuximide and mibefradil, have anti hyperalgesic effects in rodents.⁷⁴

Glutamate receptor targeted agents

Glutamate released from primary afferents, interneurons and from sequestrated stores in astrocytes may interact with a variety of receptor gated ionophores and receptors with G-protein coupling.

NMDA receptor

The NMDA-R is a calcium ionophore composed of three subunits (NR1, NR2 and NR3), each with multiple distinguishable subunits.⁷⁵ This channel is expressed on primary afferents in the dorsal horn, on second order neurons and on non-neuronal cells (oligodendroglia and astrocytes). Glutamate is released from afferents and interneurons, and binds to the NMDA receptor. At the spinal dorsal horn, high frequency C-fiber stimulation leads to post synaptic depolarization, removal of a Mg²⁺ ion blocking the pore, and, if the allosterically coupled channel binding sites for glycine and polyamines are occupied,⁷⁶ there is a channel influx of sodium and calcium⁷⁶ leading to a cascade known as wind-up.⁷⁷ NMDA-R blockade inhibits this phenomenon.

Block of NMDA receptor function is achieved by competitive glutamate binding site blockers, noncompetitive channel blockers and agents blocking associated allosteric binding sites.⁷⁸ While NMDA receptor function may be prevented by blocking any of these sites, the side effect profile (learning, memory, excitability) for these different agents vary considerably and impact upon clinical tolerability.⁷⁹

Preclinical work has demonstrated the anti-hyperpathic effects in inflammatory and nerve injury models of a variety of intrathecally and/or systemically administered competitive glutamate blockers (2 amino 5 phosphonovalorate), non-competitive NMDA channel blockers (ketamine, MK-801 and memantine, Conantokin-G; agmatine), and glycine site blocker (7 Chlorokynurenic acid, Ifenprodil).^80–82 Ifenprodil administered into the rostral cingulate cortex alleviated bone cancer pain in rats.⁸³ While there are surprisingly few high quality clinical trials, ketamine has a long clinical history of use alone and in combination with opioids in diverse pain states characterized by hyperalgesia and allodynia, including neuropathic pain, surgery, and fibromyalgia.^84–88 Ifenprodil, an inhibitor of the NMDA glycine-binding site, is currently being tested for the treatment of posttraumatic stress disorder in Phase1/2 study (Table 1).

The abuse potential of NMDA antagonists is controversial and complex.⁸⁹ Channel blockers such as ketamine have identified abuse potential. This effect may be mediated by channels associated with specific subunit constituents.⁹⁰ The role of other antagonism motifs in contributing to abuse potential is not known.

AMPA receptor

The AMPA receptor is a glutamate activated sodium selective ionophore, composed of 4 subunits (GluR1-GluR4) which plays a pivotal role in acute dorsal horn evoked excitation.⁹¹ In preclinical studies, Tezampanil (LY-293558, NGX-424) displayed efficacy in postoperative pain and spasticity.^92,93 In humans, oral administration showed efficacy upon capsaicin evoked hyperalgesia.⁹⁴ and in post-operative pain.⁹⁵

With peripheral injury, the AMPA subunit composition changes leading to a calcium permeable channel. Joro spider toxin, selective for calcium permeable AMPA site, decreases secondary mechanical allodynia development evoked in tissue injury models.⁹⁶ The abuse potential of these agents is not known.

Metabotropic receptors

Eight mGluRs (mGluR1–8) have been identified and are divided into three groups⁹⁷: Group I (mGluR1 and mGluR5) stimulates phospholipase C (PLC); Group II (mGluR2 and mGluR3) and Group III (mGluR4 – mGluR8) inhibit adenylate cyclase.^97,98 mGluRs have been localized on the primary afferents, neurons and glia within the brain and spinal cord.^99,100 Group I resides post-synaptically, and Group II and Group III are dominantly located on presynaptic terminal.¹⁰¹

Activation of Group I mGluRs is linked to central sensitization and persistent nociception, while the activation of Group II mGluRs suppresses facilitated states.¹⁰² Group I mGluR antagonists have an analgesic action by an effect upon peripheral terminal, spinally and at supraspinal sites.¹⁰³ Group II mGluR agonists regulate neurotransmitter release and depress pain transmission by acting at different levels of pain neuraxis, including nociceptors, dorsal horn and supraspinal regions such as the amygdala and PAG.¹⁰⁴ Group III mGluR agonists are also involved in the control of hyperalgesia following inflammation. As with the other metabotrophic receptors, agonist injections into a peripherally inflamed site or into the spinal dorsal horn regulates glutamatergic transmission in inflammatory and neuropathic pain.¹⁰³ Group I mGluRs antagonists and Groups II and III mGluRs agonists exhibited analgesic properties in neuropathic or inflammatory pain states^105–112 and may serve as a basis to develop future spinally-targeted agents.^113,114 Importantly, antagonists affecting Group I mGluRs, have minimal impact on fast synaptic transmission and minimal cognitive effects as compared to ionotropic glutamate antagonists.¹¹⁵

Blocking a glutamate transporter (EAAT-3), reduces intracellular glutamate, attenuates pain and decreases cellular activation. In addition to their cytoplasmic location, mGluR5 are nuclear and may mediate these effects of intracellular glutamate. Accordingly, cell permeable mGluR5 antagonists may show increased efficacy in attenuating neuropathic pain.¹¹⁶

As regards abuse potential, central Group I mGluRs appear to be substrates for stimulants.¹¹⁵ Importantly, antagonists at Group I mGluRs reduce self-administration with no alteration in motor function or the reward value of natural rewards, while agonists at Group II mGluRs prevent reinstatement of drug-seeking after abstinence.¹¹⁷

GABA receptors

GABA, a principal inhibitor transmitter, is expressed in neurons throughout brain and spinal cord. GABAergic spinal interneurons presynaptically regulate large mechanosensitive afferents and postsynaptic excitation input by a potential interaction with two GABA receptors: the GABA-A ionophore and the GABA-B metabotrophic receptor.^118,119

GABA-A

The GABA-A receptor is a GABA-gated chloride ionophore, and is composed of five subunits, each with four transmembrane spanning domains. The specific subunits define the binding of a number of molecules at the ionophore. Drugs can activate the channel (GABA. Muscimol), while others (benzodiazepines, neurosteroids, alcohol, many anesthetics) act as positive allosteric modulators at channels having specific subunit composition, stabilizing an open conformation in the presence of the agonist and at greater concentrations to directly activate the chloride channel.¹²⁰ Studies show dense and variable staining for GABA-A subunits in brain,¹²¹ and spinal dorsal horn and on primary afferent terminals,^118,122 regulating their excitability.^{123, 124} Non-spinal GABA-A ionophore activation leads to sedative, anxiolytic and amnestic effects, whereas at the spinal level increased GABA-A activity alters motor function.¹²⁵ The high degree of GABA-A receptor structure/subtype heterogeneity raises expectations for determining specific structures to target these subtypes.^{126, 127–129} Subtype specificity may exhibit different effects upon neuronal inhibition in various systems.¹³⁰

GABA-A agonists, such as muscimol or isoguavacine, display preclinical efficacy in neuropathic pain models.^131–135 Intrathecal benzodiazepines depressed nociceptive reflexes in dogs,¹³⁶ while bolus intrathecal midazolam has displayed efficacy in postoperative, low back and labor pain in humans.^137–140 At allosteric binding sites, neurosteroids, such as allopregnanolone, have shown efficacy in preclinical models of tissue and nerve injury.¹⁴¹ Of note, etifoxine promotes production of 3alpha-reduced neurosteroids and has efficacy in reducing mechanical and thermal pain symptoms in vincristine-induced neuropathic pain.¹⁴² Further, etifoxine by binding to GABA-A receptor subunits has shown to be effective in different pain disorders followed by anxiety.¹⁴³ Etifoxine was clinically tested in combination with lorazepam for cognitive improvement in elderly patients (Table 1).

The abuse potential of GABA-A targeted drugs is clearly suggested by role of the GABA-A receptor in reward circuitry. It is clear however that the potential abuse for any GABA-A targeted drug must be interpreted in terms of the subunits with which the drug interacts and the systems with which the subunits are associated.¹⁴⁴

GABA- B

Two GABA-B receptors have been cloned and are metabotrophic receptors serving to block the opening of voltage-gated Ca channel and activate inwardly rectifying K channels. These receptors are expressed peripherally and centrally, including thalamus, brain stem nuclei and spinal cord. While positive antinociceptive actions have been reported, they tend to be minimal. An important element is the potent effect upon motor neuron excitability leading to a clinically useful effect on elevated motor tone underlying spasticity occurring with neuraxial injury.¹¹⁹ Lioresal, typically employed by PO or intrathecal delivery in spasticity is not a controlled agent, but significant withdrawal can be seen with drug termination.

Opioid Receptor Targeted Drugs

Mu opioid receptor targeted agonists represent the gold standard for modifying acute nociceptive processing. This action reflects the association of these receptors i) with small afferent input that encode nociceptive processing at the spinal dorsal horn and, ii) at supraspinal levels, regulating spinal processing through descending pathways, altering perceptual processing and initiating reinforcing/reward circuit function.¹⁴⁵ Apart from their analgesic efficacy, the classic opioids display tolerance, physical dependence, respiratory depression and a high propensity for abuse.

Receptor targeting

There are three identified gene products that yield three families of opioid receptors (mu/MOR, delta /DOR and kappa/KOR)¹⁴⁶ that, when activated, alter pain processing in a naloxone-reversible fashion. More recently, identification of a receptor for the neuropeptide nociceptin has led to designation of a fourth receptor family (NOP), which is typically naloxone insensitive. While subtypes have been proposed, it appears likely that differences in pharmacology within a class may reflect on properties endowed by receptor organization and post-translational processing versus a distinctive receptor protein. These receptors are widely distributed in the brain and spinal cord and are characterized by comparable transmembrane spanning motifs and intracellular G protein coupled receptor (GPCR) signaling. At the membrane level, opioid receptors have typically been shown to be coupled, so that there is a presynaptic action reducing terminal release through a block of calcium-mediated exocytosis and membrane hyperpolarization through an increased potassium conductance.¹⁴⁷ At the spinal level, the distribution of opiate receptors on C-fiber terminals and second order neurons is consistent with the analgesic actions being mediated by a block of excitatory transmitter release from C-fibers and inhibition of 2^nd order neuron excitability.¹⁴⁵ A peripheral opiate action manifested on sensitized afferent nerve terminals is observed reflecting in part the presence of opiate receptors on the peripheral terminals of the afferent.¹⁴⁸ Supraspinal opiate actions have been identified wherein the classic descending pathways are considered to be activated by the effects of the opiate receptor on GABA interneurons in the mesencephalon removing a tonic modulation of downstream descending projections.¹⁴⁹ Higher order action on forebrain structures have additionally been identified¹⁴⁵ and likely reflect upon the effects of opioids on distress.¹⁵⁰ Recent work has suggested possible efficacy of kappa opioid antagonists as a migraine therapeutic.³⁰² Preclinical actions of opioids and their effects mediated through the several opioid receptors on pain behavior after systemic and spinal delivery have been reviewed extensively.^{145,151–153}

As noted, the common opiate target for the clinically used agents is typically the mu receptor. The possibility that amongst these receptors there may be subtypes appears likely to reflect other aspects of signaling included ligand bias and the role of heteromers (see below). Delta opioid receptors clearly exert a regulatory role.¹⁵² Intrathecal delta preferring agonist such as DADL has analgesic efficacy in humans after intrathecal delivery.¹⁵⁴ Two non-peptide molecules ADL5747 and ADL5859 were two orally bioavailable compounds¹⁵⁵ tested for acute (NCT00993863) and chronic (NCT00979953) pain management in Phase 2 clinical trials but were not more effective than placebo in osteoarthritic patients.

Kappa opioid agonists which are peripherally restricted have shown minimum abuse potential and efficacy in inflammatory and visceral pain. This along with the potential of a reduced side effect profile and lower abuse potential suggests such agonists as promising candidates for treating pain.^156,157

Interestingly while there is a typical aim to seek selective agonists, some have argued that effective improvements in efficacy side effect profiles may be achieved though ligands targeting multiple opioid receptors.^158,159

Biased ligands

One of the major strategies that is gaining interest is that GPCRs can associate with multiple second messengers (such as G alpha proteins, β-arrestin, etc.) and ligands can modulate GPCR response via one of those functional pathways, thereby exhibiting “biased agonism.”^160,161 Such biased agonists at the mu opioid receptors produce analgesia with limited side effects.¹⁶² Currently, a biased ligand (TRV-130) shows analgesia with reduced respiratory depression in phase II clinical trials.¹⁶³ (Table 1) Recently, an in silico screening approach has identified PZM21, as a mu opioid biased agonist that shows promising analgesic data with reduced side effects.¹⁶⁴

Heteromeric Receptors

Many G protein-coupled receptors couple to yield homo- and heteromers.^165–167 Such oligomerized receptors serve as targets for developing novel analgesics. For instance, a bivalent ligand containing mu agonist and delta antagonist pharmacophores linked via a spacer (MDAN-21) effectively bridges mu-delta opioid receptor heteromers and exhibits enhanced efficacy and a reduced tendency for tolerance.^168,169 Better understanding of mu and delta opioid receptor heteromers will help in understanding peripheral pain as well as development of tolerance as it has been shown that several clinically used opioids are also selective for these heteromers.^170–172 A combination of mu-receptor agonists and cannabinoid receptor agonists in rhesus monkey models, showed significant antinociception.¹⁷³ Mu opioid receptor and CB1 (cannabinoid) receptor heterodimers¹⁷⁴ and mu-mGluR5^175,176 heteromers with opioid and non-opioid binding sites, expressed strong antinociceptive effects in a range of models. In addition, a small molecule agonist for the mu-kappa opioid receptor heteromer, NNTA, is a potent antinociceptive agent with no propensity to display physical dependence or drug-seeking behavior.¹⁷⁷

Tissue target selective opioids

Inflamed tissues display an acidic environment as compared to a healthy tissue. NFEPP is a mu opioid agonist that displays pH-sensitive binding and is thus limited in its activity to a peripheral action at injured/inflamed tissues inflammatory, and is reported to be absent CNS effects or display addiction potential.¹⁷⁸

Abuse liability of the classical analgesic opiate agonists reflecting an effect upon higher order neuraxial function is clear. To the degree that a pain state reflects upon activity generated by a peripheral stimulus (e.g. tissue injury, inflammation, neuroma, etc.) opioids with a peripherally- restricted action acting upon systems outside the blood brain barrier offers a potential way forward. As reviewed above, there is anticipation that NOP agonists or opioid agonists restricted to a peripheral action do not have intrinsic reinforcing effects.¹⁵³ Additional work on the biased ligands and heterodimer systems is required.

Alpha2 adrenergic receptor targeted drugs

Alpha 2 adrenergic agonists have a potent analgesic action that is accompanied by sedation. The analgesic effects are mediated in large part by spinal alpha 2 receptors of which there are three subtypes (Alpha 2A,B,C).¹⁷⁹ These are G protein-coupled receptors that regulate dorsal horn excitation produced by small primary afferent input. Studies with mutations, antisense, and antagonists suggest an important role for the alpha2A subtype.^180–182 Alpha 2 agonists delivered systemically or intrathecally have significant effects upon acute, inflammatory and nerve injury hyperpathias.^183,184 In humans neuraxial alpha2 agonist (clonidine) and systemic (clonidine, tizanidine dexmedetomidine) have analgesic properties with sedation being a common sequelae of the actions of these agents. Dexmedetomidine is not a controlled substance. While the dependence potential of dexmedetomidine has not been studied in human, preclinical studies have shown, as with clonidine, withdrawal upon discontinuation.¹⁸⁵

Cannabinoids

Cannabinoids can produce strong antinociceptive results in various animal models of acute, tissue injury, and nerve injury-induced nociception.¹⁸⁶ Cannabinoid receptors (CB1 and CB2) are G-protein-bound receptors that negatively bind via Gi/o proteins.¹⁸⁷ CB1 receptors are found in spinal neurons, particularly in the dorsal root ganglia,¹⁸⁸ and its agonists decrease excitatory transmitter release, whereas CB2 receptors reside in spinal microglia and attenuate microglial activation.^189,190 Cannabinoids mediate their psychotropic effects through CB1, not CB2.^191,192 Ligands that interact with CB1 and CB2 demonstrated the ability to regulate nociceptive processing.^193,194 Agents that block the metabolism of CB1 endogenous agonists consequentially increase its concentration and may be used to activate cannabinoid receptor function.¹⁹⁵ CB1 and CB2 selective agents intrathecally-delivered decreased facilitated states such formalin model, hyperpathia in neuropathy models and in tumor bone pain in rodents.^196–198 Cannabinoid role in pain processing is based on spinal and nociceptive neuron inhibition, although peripheral sites of action have also been identified. The use of spinal cord stimulation (SCS) in a rodent neuropathic pain model revealed long-lasting and incremental reduction of hyperalgesia mediated by endocannabinoids. The effect was amplified by co-administration of LY2183240, an endocannabinoid reuptake/breakdown inhibitor, and inhibited by a CB₁ R antagonist, AM251, but not by a CB₂ R antagonist, AM630.¹⁹⁹ CB2 receptor antagonists may be a potential target in treating chronic pain of several etiologies by modifying cytokine profile when blocking peripheral immune tissue receptors, and by blocking receptors in neurons and glial cells.²⁰⁰

The anti-nociceptive effect of eslicarbazepine acetate (ESL), an anti-epileptic drug derived from carbamazepine/oxcarbazepine, has been shown to be mediated by serotonergic 5-HT1B/1D and cannabinoid CB1/CB2 receptors. ESL showed beneficial effect in different neuropathic and visceral pain models.²⁰¹ ESL has been tested clinically in different pain conditions (diabetic neuropathy, PHN, fibromyalgia, etc. (Table 1).

Abuse potential associated with CB1 receptor agonists has been well documented.²⁰² The CB2 receptors has been shown to modulate ventral tegmental dopamine neuron activity, circuitry considered pivotal in the addictive process.²⁰³

Angiotensin 2 receptor antagonist

Angiotensin may reside in primary afferents and can activate facilitatory cascades mediated through AT1 and AT2 receptors.^204,205 An AT2 antagonist, EMA401 has been tested in Phase 2 clinical trials for the treatment of post-herpetic neuralgia (PHN), and preliminary data showed that it is well tolerated and it exhibited a primary analgesic efficacy endpoint.²⁰⁶ Two phase 2b studies with EMA401 for PHN and painful diabetic neuropathy were put on hold.(207) However, recently, the new Phase 2 study for PHN was registered at clinicaltrials.gov. (Table 1)

Adenosine agonists/antagonists

In models of acute nociceptive processing,²⁰⁸ neuropathy^208–212 and inflammatory pain²¹³ administration of adenosine and related ligands yielded significant anti-hyperalgesic effects. Intrathecally administered adenosine lowered allodynia in experimental pain models,^214,215 and in patients suffering from neuropathic pain,^216,217 although negative results have also been reported.²¹⁸ Adenosine activates four G-protein-bound receptors: A1, A2A, A2B, A3.²¹⁹ A1 receptors, which pre-synaptically inhibit neurotransmitter release and post-synoptically inhibit excitatory transmission,^220,221 are found on dorsal horn neurons and on small-to-medium-sized neurons of the DRG.^{220, 222–225} A2A receptor agonists were reportedly able to cause long-term reversal of allodynia in mononeuropathies,²²⁶ while the possible explanation was the role of A2A agonists as potential glial inhibitors. In addition, A2A receptors may enhance glutamate release, and A2A antagonists may behave protectively by reducing such excitatory effect.²²⁷ A2A receptor knock-out mice showed a significant reduction of the mechanical allodynia and a suppression of thermal hyperalgesia and allodynia, and well as attenuated expression of microglia and astrocytes, confirming potential beneficial role of A2A receptor antagonists in the treatment of neuropathic pain.²²⁸ Activation of the A3 adenosine receptor (A3AR) blocked hyperalgesia in mono and polyneuropathies.²²⁹ The abuse potential of agonists at these A1–3 receptors is not known.

Purine agonist/antagonists

Adenosine triphosphate, widely present in the CNS,²³⁰ reacts with P2 receptor family with several subtypes. The P2X ligand-gated ionotropic receptors (consisting of 7 subtypes) and P2Y\G-protein-coupled receptors (divided into 8 subtypes). P2Y receptors may signal either via Gq/G11 to initiate the phospholipase C/inositol triphosphate (InsP3) endoplasmic reticulum Ca²⁺-release pathway (the P2Y1, P2Y2, P2Y4, P2Y6, and P2Y11 receptors) or via Gi/o, blocking adenylate cyclase and modulating ion channel function.²³¹ Both P2X and P2Y receptors reside in dorsal root ganglia, spinal neurons and glia.²³²These receptors as glia activators, lead to spinal release of proinflammatory proteins and cytokines, associated to commencing facilitated pain states.^233–237 Transient reversal of hyperpathia after nerve injury was achieved via intrathecal administration of P2X and P2Y inhibitors.^238–241 The P2X3 subtype is predominantly on C- and Aδ-fiber primary afferent neurons. P2X3 antagonists have shown efficacy in inflammatory and in mono and poly neuropathic pain states.²⁴² P2X4 subtype is important in spinal facilitation that originated from tissue and nerve injury.²⁴³ P2X4R antisense oligodeoxynucleotide intrathecal delivery prevented P2X4R protein expression and restrained mechanical allodynia development.²³⁸ P2X4R by modulating neuroimmune interactions in the spinal cord and DRG could have an important role development of neuropathic pain, signifying potential therapeutic effects of P2X4 receptor antagonists.²⁴⁴ Electroacupuncture showed beneficial effect in neuropathic pain models by attenuating IFN-γ release and reduced expression of P2X4R in microglia.²⁴⁵ Furthermore, duloxetine, a serotonin and noradrenaline reuptake inhibitor, showed results in neuropathic pain models by inhibition of P2X4 receptors.²⁴⁶ AF-219, a P2X3 antagonist is in clinical development as an antitussive. The abuse potential of purine receptor agonists and antagonists is unknown.²⁴⁷ (Table 1)

Innate immune signaling

Toll-like receptors (TLRs), a key sensory component in innate immune function are found on neuronal and non-neuronal cells in the spinal cord and function by recognizing injury-associated molecular structures, while being strongly associated with proalgesic/inflammatory cytokines (DRG).^248,249 Intrathecal TLR4 antagonist administration resulted in improved effects within inflammatory and neuropathic pain states²⁵⁰ and was associated with opiate-induced hyperalgesia phenomenon.²⁵¹ Another perspective on the role of TLR4 signaling was noted when it was found that the spinal delivery of a TLR4 antagonist (LPS-RS) would prevent the transition from an acute inflammatory state to chronic post inflammatory state with neuropathic pain phenotype²⁵² while a small molecule TLR4 antagonist (TAK242) would prevent the onset of late phase allodynia after intraplantar formalin.²⁵³ Repeated intrathecal administration of LPS-RS (TLR2 and TLR4 antagonist) and LPS-RS Ultrapure (TLR4 antagonist) attenuated allodynia and hyperalgesia and potentiated the effect of buprenorphine but not morphine²⁵⁴ Effort has been put into developing new structures to block TLR activation by interacting with the TLR4 ligand or downstream signaling^255–257 as shown by the antihyperpathic effects achieved by inhibition of MyD88 signaling.²⁵⁸

Lipid mediators

Prostaglandins

The role of lipid mediators, such as the omega 6 derived prostaglandins, which produce a sensitized primary afferent and is centrally facilitated and mediated by eponymous receptors, has been long appreciated. Discovery of cyclooxygenase isoforms led to the rational development of prostanoid receptor antagonists and isoform specific inhibitors, which were shown to have both a peripheral anti-inflammatory and a central action on spinal facilitatory processing.²⁵⁹ Non selective and COX-2 inhibitors have been shown to have significant anti-hyperpathic effects in a variety of tissue injury pain states in animal models²⁵⁹ and in humans.²⁶⁰ Unfortunately, typical limiting issues involve target related actions on cyclooxygenase (GI, platelet function and cardiovascular) side effects.²⁶¹ An interesting parallel to the NSAIDs is the actions of acetaminophen.²⁶² This molecule has been shown to be efficacious in a variety of preclinical models and in clinical pain states associated with inflammation and tissue injury and in mono and polyneuropathies, with dose dependent effects upon hyperalgesia and allodynia.^263–265 Available as an OTC, this agent, in us for more than 100 years, has revealed no abuse liability. In spite of its utility, its mechanism of action is at best controversial.²⁶⁴

Soluble epoxide hydrolases

The epoxidized metabolites obtained from omega-3 long chain fatty acids show anti-inflammatory and an anti-hyperpathic effect in a variety of preclinical models. However, they are being rapidly metabolized by enzymatic hydrolysis by soluble epoxide hydrolases (SHI). Of interest inhibitors of SHI have been shown to have significant anti-hyperalgesic actions in a variety of preclinical models.²⁶⁶

Proresolvins

Inflammatory cascades are typically self-limited leading to the healing phase of an injury. One of the mechanisms of this resolution has been a variety of lipid mediators referred to as proresolvins. These endogenous mediators include omega 3 (resolvins, protectins, and maresins) and omega 6 derived lipoxins. It is increasingly recognized that anti-inflammation and proresolution cascades represent distinct mechanisms for controlling the inflammatory response.²⁶⁷ Delivery of a variety of these proresolvin molecules has shown to have significant anti-hyperpathic actions in a variety of inflammatory, mono and polyneuropathic models.^268,269 The abuse potential of these lipid mediators is not known.

Anti-inflammatory Cytokines

Upon activation of various glial signaling cascades, numerous cytokines (including activation of NF-κB) influence the proinflammatory mediators’ production (e.g. TNF, IL-6, and IL-1β), which, in turn, activate pro-algesic cascades.²⁷⁰ Additionally, such cascades can aid in the release of anti-inflammatory products (e.g. IL-4, IL-6, IL-10, IL-11, IL-13, TGF-β) and soluble cytokine receptors,²⁷¹ which control the inflammatory cascade.²⁷²

IL-10

It has been shown that IL-10 is one of the most powerful endogenous anti-inflammatory cytokines in nervous system.²⁷³ In animal models, IL-10 intrathecal delivery demonstrated therapeutic efficacy in various chronic pain models, primarily in treating different types of neuropathic pain.²⁷³ In different animal models, viral vector-mediated expression of IL-10 in DRGs prevented development of painful diabetic neuropathy²⁷⁴ and helped in treatment of HIV-induced neuropathy.²⁷⁵

IL4

Another anti-inflammatory cytokine IL-4 showed beneficial role in treating different types of neuropathic pain in different animal models.^276–278 An interesting variation is the intrathecal transfection of an IL4/IL10 fusion protein leading to a potent and persistent antihyperalgesia.²⁷⁹

Toxins

There is an increasing interest in potential of producing long term changes in neuraxial pain processing by the peripheral or spinal delivery of agents which target the functionality of systems processing pain information. Here the consideration is for the treatment of persisting pain states. The role of these therapeutic approaches is not clear at present time. For those approaches leading to permanent loss of cells such as the saporin conjugates or the TRPV1 agonists, it appears less likely that they would be employed outside of the terminal patient (as in cancer). Toxins that result in long lasting but irreversible effects such as the botulinum toxins might betherepeutic approach for persistent pain states in a non-terminal patient.

TRPV1 receptors

TRPV1 channels are with few exceptions located on the central and peripheral terminals of high threshold primary afferents. Topical²⁸⁰ and spinal delivery²⁸¹ of TRPV1 agonists such as capsaicin or analogues such as resiniferatoxin (RTX) desensitize the TRPV1 (+) afferent and destroy the DRG terminal²⁸² by calcium cytotoxicity.²⁸³ and analgesia. The effects after topical delivery has led to the approval of transdermal capsaicin.²⁸⁴ Neuraxial delivery of TRPV1 agonists has been shown to result in robust anti-nociception in dogs.^{285, 286} Intrathecal RTX showed potent and persistent anti-hyperalgesic effects refractory bone cancer pain in canines, without evidence of deafferentation sequelae.²⁸⁷ One clinical trial testing the use of intrathecal resiniferatoxin for intractable cancer pain was begun and is currently on hold. (Table 1)

Saporin conjugates

G-protein-coupled receptors undergo internalization when occupied by their respective agonists.²⁸⁸ Appropriate linking of a G-protein targeted agonist such as substance P (SP) and a toxin such as saporin (plant product from Saponaria officinalis which is not otherwise taken up by the cell) will result after agonist binding to the neurokinin 1 receptor (NK1r), internalization of the agonist and toxin complex into the cell expressing that receptor.²⁸⁹ Saporin blocks riboslyation and protein synthesis, resulting in cell death. The neurokinin 1 (NK1) receptor, a GPCR, found on postsynaptic second order dorsal horn nociceptive neurons²⁹⁰ is taken up into the neuron and the neuron dies. Intrathecally administered sP-saporin, but not saporin, robustly destroys NK1(+) dorsal horn neurons and attenuates pain states in rodents and bone cancer pain in dogs^{291, 292,293} Intrathecal administration of sP-Saporin is currently in Phase 1 clinical trial for the treatment of intractable cancer pain (Table 1). Importantly, this functional coupling of a ligand to saporin is effective for any ligand for any G protein coupled receptor that displays internalization.²⁸⁹

Botulinum toxin

These toxins are composed of a heavy chain (HC) and a light chain (LC). The HC portion enables the toxin to be taken into the cell. Once inside, the complex is cleaved, freeing LC, which serves as enzyme cleaving SNARES.^{294, 295} SNAREs mobilize vesicles for transmitter release and aid in the transport of GLUA1 AMPA receptor subunits to the membrane.²⁹⁶ In case of SNARE cleavage, transmitter release is blocked. Preclinically, intrathecally administered BoNTs produced anti-hyperalgesic effects in various inflammatory and neuropathic hyperpathia.^297–300 The BoNT uptake is ubiquitous and the potent effects upon transmitter release may include inhibitory interneurons and motor neurons.^301,302 Several BoNT serotypes have been shown after topical application to be taken up and to block both local (peripheral) release from a nociceptor and to be transported centrally to inhibit downstream nociceptive processing, with indications of a possible pre- and post-synaptic effect.³⁰³ Intravesical injections of onabotulinumtoxin-A (BoNT-A) showed significant pain reduction in patients with interstitial cystitis/bladder pain syndrome (IC/BPS) refractory to other treatments suggesting a local effect upon the urothelium.³⁰⁴ Coupling of the light chain of BoNT-A with substance P showed a beneficial role in treating chronic pain after intrathecal delivery.³⁰⁵

Transfection targets

The use of viral transection at the spinal level represents an exciting approach to modify spinal function. Intrathecal delivery of various transfection systems has been used to increase the expression of cytokines,^302,303 knock down of pivotal targets with shRNAs,(304,305) expression of transcription factor decoy proteins,³⁰⁶ over-expression of microRNAs are among many spinal targets that have been successfully manipulated through transfection approaches. Technically, it is clear that while intrathecal delivery of AAV may transfect ganglion neurons, parenchymal transfection may be limited, in part by the diffusion barrier presented by the pia and transfection enhanced by subpial delivery.³⁰⁷ Intraganglionic injections have also been suggested as an efficient tool to alter afferent function.³⁰⁸ The ability to produce long-term, regulated changes in processing offer a potential to modify the pathological expression of pain by modifying system function.

Conclusions

The FDA has been mandated to address the national epidemic of opioid abuse with policies aimed at reversing the epidemic. One element of this plan, recognizing the pivotal role opiate receptors play in pain management, is to stimulate development of more effective pain medications with abuse-deterrent (AD) properties and abuse-deterrent formulations (ADFs) of opioids. The rational way forward is to develop analgesics that minimize abuse potential. In order to improve the translations, the FDA launched the Analgesic, Anesthetic, and Addiction Clinical Trial Translations, Innovations, Opportunities, and Networks (ACTTION) that issued recommendations to improve the reproducibility of research pertaining to pain studies.⁽³⁰⁶⁾

In the last three decades, by virtue of funding from the national funding agencies and by pharma, we have obtained an increased understanding of pain mechanisms and accordingly an abundance of relevant targets for which we must develop drugable molecules. The development of novel targets and the implementation of approaches that can alter processing for extended periods (transfection and toxins) represent exciting advances in managing the chronic condition.

An important issue in analgesic drug discovery is that apparently promising pre-clinical candidates can fail during clinical trials. Several reasons for this may be entertained. It is straightforward to model human conditions for which the initiating mechanisms are likely known, as for example in chemotherapy-induced neuropathy. Conversely, it is difficult, if not impossible, to rationally define surrogate models for a pain state such as fibromyalgia, where the mechanisms of hyperpathia observed in the human conditions are not known.³⁰⁷ Thus, preclinical models may fail to recapitulate the human pain condition being studied. Research into mechanisms and the appreciation of the role played by innate and adaptive immunity are likely to shed light on these complex problems, revealing novel, mechanistically-defined targets that lack congruence of the clinical and preclinical target, where minor species differences in a receptor sequence may yield a drug that does not engage the human target. Modern molecular techniques and target sequencing makes this disconnect less likely. Further research into biological biomarkers (exosomes) and genetic characterization will provide an important link to define human and animal covariates.^308,309

It is interesting to note, as outlined in Figure 1, that a preponderance of the targets producing therapeutic efficacy as analgesics (vs. anesthetics) display a robust effect upon primary afferent and dorsal horn processing that leads to surprisingly specific changes in pain behavior, denoting the role played by the content of the ascending message in characterizing components of the aversive nature of the stimulus event. This emphasis does not exclude the likelihood that many agents, notably opioids, can exert a potent effect upon pain behavior after supraspinal action with such actions accounting for changes in the affective-motivational component of the pain state. While it appears likely that specific supraspinal systems may be found that possess a pharmacology specifically targeting the pain state, current research has provided little evidence that what affects pain processing/behavior at supraspinal sites does not also have pronounced effects upon behavior and perception, aspects of which are associated with positive reward and the addictive potential. These results suggesting the interdigitation of these affective components are in parallel with the early work involving surgical resection of limbic and forebrain structures. While such interventions were reported to lead to a loss of the affective components of the pain state, they also led to profound changes in personality and judgement.³¹⁰

Finally, the translational development of analgesics for the clinic must increasingly consider the issues of drug abuse. The aim is to address the specific management of pain and suffering. Clearly, agents minimizing the psychological underpinnings of suffering may well display a positive reinforcing component in the absence of pain. This conflation emphasizes the complexity of the problem and the challenges to selectively modify one of the most basic cognitive elements, the pain experience.