Guanylate cyclase-C agonists as peripherally acting treatments of chronic visceral pain

Irritable Bowel Syndrome (IBS) is a common chronic functional gastrointestinal disorder characterized by abdominal pain and altered bowel habit that affects ~11% of the global population[ 1 ]. Current diagnosis of IBS is based on symptoms enshrined in the multiple iterations of Rome criteria, and exclusion of pathophysiology based on standard clinical testing[ 2 ]. Sub classifications of IBS are determined by the nature of stool irregularities, such as those experiencing constipation (IBS-C), diarrhea (IBS-D), or alternating constipation and diarrhea (IBS-M)[ 2 ]. Additionally, chronic idiopathic constipation (CIC) affects approximately 14% of the global population. Symptoms associated with CIC diagnosis include reduced bowel movements, hard stools, straining during defecation, as well as bloating and discomfort. Thus CIC patients have some symptoms which overlap with those of IBS-C. Given the clinical burden of these conditions, efficacious therapies are necessary. There are multiple efficacious treatments for bowel dysfunction, as illustrated in systematic reviews and network meta-analyses in patients with chronic constipation[ 3 , 4 ] or IBS-C[ 5 ]. However, there is a major unmet need in the management of symptoms such as pain and bloating. Although there are many candidate approaches in development as visceral analgesics[ 6 ], one of the more attractive pharmacological approaches involves the peripherally-acting guanylate cyclase-C (GC-C) agonists (see Glossary) linaclotide and plecanatide. These agonists ‘tap’ into the intrinsic mechanisms controlling intestinal fluid homeostasis that are regulated by the endogenous hormones uroguanylin and guanylin [ Box 1 ]. Importantly, linaclotide and plecanatide have limited bioavailability [ Box 2 ] and hence have clinical safety as well as being bereft of central adverse effects, unlike traditional analgesics such as μ-opioids, or centrally acting neuromodulators and most available cannabinoid agents[ 6 ].

Numerous clinical trials have documented benefits of GC-C agonists on pain and other sensory symptoms such as bloating in patients with CIC or IBS-C. In a double-blind, placebo-controlled, parallel-group phase II study of 310 patients with CIC (assessing 75,150,300,600μg Linaclotide or placebo), Linaclotide reduced abdominal symptoms such as discomfort and bloating as well as constipation severity while improving adequate relief of constipation, treatment satisfaction and quality of life, for each dose of linaclotide[ 48 ]. A more recent phase II randomized, double-blind, placebo-controlled dose finding study of Japanese CIC patients provided similar primary endpoint efficacy in improving weekly spontaneous bowel movement (SBM) frequency as well as relief of abdominal symptoms of chronic constipation for all linaclotide doses (62.5-500μg)[ 49 ]. Furthermore, in a phase IIb randomized, double-blind, placebo-controlled study of 420 IBS-C patients meeting Rome II criteria that were given doses ranging from 75μg-600μg, the 300μg linaclotide dose most significantly reduced abdominal pain. In contrast to the rapid improvement of bowel habits (within 24hr), the improvement of abdominal pain was significant after one week of treatment and gradually increased further throughout the remaining 12-week treatment[ 50 ]. Interestingly the analgesic effect of linaclotide was most profound in patients with severe abdominal pain at baseline[ 50 ]. A randomized, double-blind, placebo-controlled, phase II dose-finding trial of a Japanese IBS-C patient cohort diagnosed using Rome III criteria evaluated once daily 62.5-500μg linaclotide during a 12-week treatment period. Monthly responder rates of complete SBMs, abdominal pain/discomfort relief, and global assessment of relief of IBS symptoms were higher than placebo during the 12-week treatment[ 51 ]. Furthermore, a phase III, randomized, double-blind, placebo-controlled trial of linaclotide has been performed in an additional patient cohort of 839 individuals in China[ 52 ]. Greater numbers of IBS-C patients randomized to once-daily oral 290μg linaclotide at centres in China, North America and Oceania met co-primary endpoints of ≥30% reduction in average weekly abdominal pain or abdominal discomfort score for 6 out of 12-weeks and IBS degree of relief ≤2 weekly responder for 6- out of 12-weeks compared to those receiving placebo. In a phase III, double-blind, parallel-group, placebo-controlled trial, IBS-C patients were randomized to placebo or 290μg oral linaclotide once daily in a 12-week treatment period, followed by a 4-week randomized withdrawal period. Notably, greater improvements were seen in linaclotide vs. placebo patients for all secondary end points throughout the 12-weeks. During the 4-week randomized withdrawal period, patients who remained on linaclotide showed a sustained improvement in abdominal pain[ 26 ]. In contrast, patients re-randomized from linaclotide to placebo showed return of symptoms[ 26 ]. Overall, in all phase III studies conducted in IBS-C or CIC patients to date, linaclotide significantly improved abdominal pain, discomfort and bloating[ 24 - 27 , 52 ]. Importantly, on September 22, 2020 the FDA approved a supplemental New Drug Application (sNDA) for linaclotide based on data from a phase IIIb, randomized, double-blind, placebo-controlled parallel-group trial evaluating the safety and efficacy of linaclotide 290μg on multiple abdominal symptoms in 614 adult IBS-C patients over 12-weeks compared to placebo. The additional data demonstrated linaclotide improved the overall abdominal symptoms score of bloating, discomfort, and pain in adult IBS-C patients[ 53 ]( https://www.accessdata.fda.gov/drugsatfda_docs/label/2020/202811s016lbl.pdf ). This trial was designed to highlight the impact of linaclotide on a primary endpoint based on the mean abdominal score (composite of abdominal bloating, abdominal discomfort, and abdominal pain) across 12-weeks, as well as the secondary endpoint which was a responder analysis based on at least a 2.5-point improvement in the abdominal score from baseline for >6 out of 12-weeks. In addition, individual scores of bloating, discomfort, and pain were assessed over the 12- week treatment period. Linaclotide-treated patients showed a 29.7% mean decrease from baseline in their weekly abdominal score (bloating, pain and discomfort) through the 12-week treatment, compared to 18.3% for the placebo-treated patients. In addition, the data showed improvement in individual abdominal symptoms of bloating, discomfort and pain with linaclotide [ 53 ]. A 2020 study shows 26 IBS-C patients treated daily with linaclotide (290μg) for 10-weeks displayed prolongation of rectal and anal cortical evoked potential responses (using anorectal electrical stimulations and transcranial/transspinal-anorectal magnetic stimulations) compared to 13 placebo-treated IBS-C patients. These data suggest that linaclotide may improve rectal hypersensitivity in IBS-C patients[ 54 ] and supports the observations from pre-clinical studies that linaclotide inhibits nociceptors to reduce ‘peripheral drive’. Along these lines a recent report has documented that linaclotide delayed release formulations, released in the terminal ileum (LIN-DR1) or colon (LIN-DR2: MD-7246) reduces abdominal pain in IBS-C patients compared to placebo, without altering bowel function or inducing diarrhea[ 55 ]. This may suggest the predominant analgesic site of action of linaclotide occurs within the colon, as indicated by rodent studies. Two phase III identical randomized double blind placebo controlled trials of plecanatide in patients with IBS-C randomized 2189 individuals among whom 1879 completed the studies[ 29 ]. Using the overall responder definition, patients reporting ≥30% reduction from baseline in worst abdominal pain plus increase of ≥1 CSBM/week from baseline in the same week for ≥6- of 12-treatment weeks), both plecanatide 3mg and 6mg were superior to placebo, including the responses of abdominal symptoms such as pain, discomfort and bloating[ 29 ]. Overall, a systematic review and network meta-analysis[ 5 ] involving 12 trials composed of 7302 patients reported in 10 separate articles that reported dichotomous data for failure to achieve an abdominal pain response, all secretagogue treatments were more effective than placebo, with linaclotide 290μg once daily ranked as the most effective treatment based on 3 RCTs (RR 0.79; 95% CI 0.73–0.85).

Importantly, the analgesic efficacy of linaclotide occurs across a broad spectrum of distension paradigms, which are usually stepwise, graded distensions called ‘ascending method of limits’ or ‘rapid phasic’ distensions. Overall, these distension paradigms would simulate rapid mechanical stimulation and pressure levels observed during high amplitude colonic contractions. VMR studies of controlled isobaric colonic distensions 0-60mmHg at 15mmHg steps[ 20 ] or controlled isobaric colonic distensions 0-80mmHg at 10-20mmHg steps[ 22 ] have shown orally administered linaclotide reduces visceral hypersensitivity in rodents in vivo when performing CRD with the ascending method of limits. Moreover, controlled isobaric colonic distensions at randomized pressures of 0,20,40 and 60mmHg, also shows orally administered linaclotide reduces visceral hypersensitivity in rats[ 23 , 36 ]. Studies have used single intra-colonic [ 19 ], daily chronic oral (14 days)[ 22 ] or chronic intra-colonic (14 days)[ 43 , 47 ] linaclotide treatment followed by rapid phasic CRD and then examined dorsal horn neuron activation within the spinal cord. In all of these studies, acute or chronic linaclotide reduces the number of CRD-activated dorsal horn neurons[ 19 , 22 , 43 , 47 ]. These studies show that irrespective of the method of distension (ascending method of limits/randomized phasic distensions) that linaclotide is analgesic in models of rat and mouse acute and CVH. Studies also show a single oral dose of plecanatide in healthy naïve rats did not alter the rate of abdominal contractions to CRD. However, in rats 4 days post-TNBS lower doses (0.01 and 0.05mg/kg), but not higher doses (>0.1mg/kg) of plecanatide attenuated the TNBS-induced increase in the number of abdominal contractions with increasing distending pressures up to 60mmHg[ 28 ]. The precise mechanisms by which plecanatide induces visceral hyperalgesia in these studies have not been specifically investigated.

Pre-clinical and clinical evidence supports the concept that GC-C agonists act as peripherally acting visceral analgesics. Preclinical studies show that linaclotide is a peripherally acting drug with cGMP acting as a neuromodulator and downstream effector of linaclotide that directly inhibits colonic nociceptors via a membrane target accessed from an extracellular site. However, the precise identity of this molecular target remains to be identified. Randomized controlled clinical studies of GC-C agonists have demonstrated beneficial treatment effects on abdominal symptoms including pain and bloating in adult IBS-C patients. Whether changes in the GC-C/cGMP pathway are apparent in patients and whether such changes contribute to patient symptomatology warrants further investigation. Finally, recent preclinical studies also provide the prospect that linaclotide may offer a novel therapeutic option, not only for the treatment of chronic abdominal pain in IBS-C, but also co-morbid bladder dysfunction and endometriosis-associated pain.

Recent studies demonstrate linaclotide itself does not inhibit nociceptors, as linaclotide applied directly to isolated DRG neurons does not alter their excitability[ 22 ]. This functional evidence confirms expression studies using in situ hybridization showing GC-C is absent from the DRG and spinal cord[ 19 ]. Furthermore, linaclotide treatment also has no effect on colonic compliance during graded CRD[ 22 ], nor affects electric field stimulation-induced colonic contractions[ 19 ], which is consistent with the absence of GC-C on smooth muscle cells. These expression profiles are in contrast to the colonic epithelium where GC-C is extensively expressed in rodents and humans [ 19 , 22 , 44 ]. It was hypothesized that cGMP produced in intestinal epithelial cells upon GC-C activation could mediate the analgesic actions of linaclotide by acting on sensory afferents in the colon wall. It is known that cGMP is actively transported out of cells through membrane-bound transport proteins (cGMP transporters or efflux pumps)[ 45 ] and that in the CNS, extracellular cGMP has direct inhibitory effects on neurons causing reduced excitability and neurotransmitter release. Accumulating in vitro , ex vivo and in vivo evidence support the concept that the anti-nociceptive actions of linaclotide are mediated by activation of the GC-C/cGMP pathway and extracellular cGMP secreted from intestinal epithelial cells acting on afferent endings. Firstly, in vitro exposure of intestinal Caco-2 cells to GC-C agonists stimulates bidirectional, active extracellular transport of cGMP into luminal and basolateral spaces that are significantly and concentration-dependently decreased by probenecid, an inhibitor of cGMP efflux pumps[ 18 , 19 , 45 ]. This cGMP secretion after GC-C activation is also observed from the basolateral side of rat colonic epithelium into the submucosal space[ 18 ]. This implicates energy dependent transport of cGMP by cGMP efflux pumps, such as the multidrug resistant proteins 4 (MRP4) or 5 (MRP5). Under normal conditions, cGMP phosphodiesterases are considered the major elimination pathway for intracellular cGMP. However, MRP4/5-mediated extracellular cGMP transport is consistent with their function as overflow pumps, allowing cGMP to be released extracellularly when concentrations are supra-physiological, such as during activation of GC-C by linaclotide. For example, MRP4 is expressed on the apical membrane of intestinal epithelial cells and inhibition of MRP4 suppresses apical but not basolateral cGMP efflux[ 11 , 46 ]. As discussed above, the cGMP transporter inhibitor probenecid prevents linaclotide-induced inhibition of colonic nociceptors[ 19 ], implicating cGMP in linaclotide’s inhibitory effect. Confirmation of cGMP-induced modulation of colonic hyperalgesia comes from studies directly dosing or applying cGMP in animal models. In rats with TNBS-induced visceral hypersensitivity, oral cGMP administration prior to CRD dose-dependently increases colonic distension thresholds, and decreases the number of abdominal contractions to CRD in a partial restraint-induced stress model[ 18 ]. These effects of cGMP mimicked the anti-hyperalgesic effects of uroguanylin. Furthermore, mice with visceral hypersensitivity treated intra-duodenally with cGMP exhibited decreased colonic afferent firing to increasing CRD pressures[ 18 ]. The use of ex vivo single-unit recordings of colonic nociceptors confirmed the anti-nociceptive effects of cGMP and linaclotide[ 19 ]. These studies showed cGMP applied to the surface of the colonic epithelium concentration-dependently decreased nociceptor activity with greater efficacy in CVH mice, mimicking the effects of linaclotide[ 18 , 19 ]. Importantly, the anti-nociceptive effect of cGMP was greater following removal of the colonic mucosa, allowing cGMP to bypass the epithelial barrier. These findings provide further evidence that cGMP is acting directly on the underlying afferent nerve endings following basolateral release from epithelial cells[ 19 ]. cGMP also reduces the number of action potentials fired by low-threshold stretch-sensitive pelvic afferents [ 16 ],[ 18 ]. Most recently, patch-clamp recordings of isolated DRG neurons, completely independent of any colonic tissue, demonstrated cGMP causes dose-dependent inhibition of action potential firing. This occurs in populations of colon-innervating DRG neurons from mice, as well as a mixed population of DRG neurons from human donors[ 22 ]. Fluorescence resonance energy transfer (FRET)/cGMP measurements showed conclusively for the first time that cGMP acts on a membrane target that is accessed from an extracellular site to inhibit nociceptors[ 22 ]. The precise molecular target that cGMP acts on extracellularly remains elusive but is the subject of continued investigation.

Neuroanatomical studies performed in rodents show sensory afferents innervating the urogenital organs and colon are predominantly peptidergic (containing substance P, calcitonin gene related peptide)[ 56 - 58 ]. There is also strong evidence for convergence of bladder and colon sensory innervation at the level of primary afferents, and peripheral sensitization of urinary bladder and pelvic nerve afferents in the setting of colonic irritation, which potentially contribute to disorders characterized by chronic pelvic pain[ 57 , 59 ]. Consistent with the mechanisms underlying the anti-nociceptive actions of linaclotide, recent studies suggest linaclotide may be useful in the treatment of bladder hypersensitivity disorders occurring concurrently with gastrointestinal disorders[ 31 , 59 ]. Neonatal rats exposed to unpredictable ELS exhibit increased abdominal contractions to CRD, but also exhibit bladder hypersensitivity[ 36 ]. Linaclotide (3μg/kg daily for 7 days) but not vehicle, reduced hypersensitivity from both the colon and bladder, suggesting the anti-nociceptive actions of linaclotide may extend beyond the colon. Protamine sulfate infused into the bladder, which transiently enhances mucosal permeability, induced mechanical hypersensitivity of both the bladder and colon. The 7-day linaclotide, but not vehicle, treatment attenuated bladder and colonic hyperalgesia to control levels[ 23 ]. Therefore, these studies suggest linaclotide impacts visceral hypersensitivity in adjacent pelvic organs through a novel mechanism that is likely via common shared sensory pathways. Indeed, recent evidence shows linaclotide is able to exploit the neuronal pathways responsible for visceral organ ‘cross-talk’, and resolve bladder afferent hypersensitivity that normally occurs following TNBS-induced colitis[ 22 ]( Figure 2 ). The study first identified CVH mice, which exhibit colonic afferent sensitization ex vivo and hyperalgesia to CRD in vivo , also exhibit simultaneous bladder afferent hypersensitivity, increased excitability of bladder-innervating DRG neurons and altered urinary voiding patterns, suggestive of overactive bladder. Following chronic daily oral linaclotide (3μg/kg) administration for 2-weeks colonic hypersensitivity was reversed and bladder afferent sensitization resolved[ 22 ]. This study also demonstrated linaclotide had no direct effect on bladder afferents, with GC-C absent from bladder urothelial cells, suggesting linaclotide acts via an indirect mechanism to normalize bladder hypersensitivity. Notably, these beneficial effects of linaclotide on bladder function also occur following 2-week intra-colonic linaclotide administration, to mimic a delayed release formulation[ 47 ]. Cross-organ sensitization of the bladder via afferent mechanisms has been well characterized[ 31 , 57 , 59 ]. This includes the presence of dichotomizing neurons that have a single cell body with axons that innervate both the colon and bladder[ 22 ]. The central terminals of distinct bladder- or colon-innervating afferents also reside in close proximity to one another within the spinal cord[ 22 ]. Studies using a mouse CVH model show colonic afferent central terminals are denser and display sprouting into other regions of the dorsal horn[ 60 ], thereby increasing potential interaction between colon and bladder afferent terminals[ 22 ]. These studies suggest such interconnected neuronal pathways can be exploited to reduce comorbid hypersensitivity of adjacent organs, in particular the bladder ( Figure 2 ). Such mechanisms were recently exploited in a rat model of endometriosis whereby 8-weeks after endometrium transplantation into the intestinal mesentery, rats developed endometrial lesions as well as vaginal hyperalgesia to distension and decreased mechanical hind paw withdrawal thresholds[ 21 ]. Daily oral linaclotide (3μg/kg) administration increased pain thresholds to vaginal distension and increased mechanical paw withdrawal thresholds, indicating analgesic actions of linaclotide. Furthermore, using a cross-over design, administering linaclotide to rats previously administered vehicle resulted in increased paw withdrawal thresholds (reduced pain), whereas replacing linaclotide with vehicle decreased paw withdrawal thresholds (increased pain)[ 21 ]( Figure 3 ). Retrograde tracing studies from ileum, colon and vagina identified dichotomizing dual-labelled ileal/colon innervating afferents as well as colon/vaginal dual-labelled neurons and a rare population of triple traced ileal/colon/vaginal neurons within thoracolumbar DRG [ 21 ]. The central terminals of these sensory afferents also lie in close apposition to one another within the spinal cord [ 21 ]. Notably, GC-C expression is absent in the vagina and endometrial cysts, suggesting the actions of linaclotide are shared through nerve pathways between these organs [ 21 ]. These pre-clinical studies suggest linaclotide may offer a novel therapeutic option not only for treatment of chronic abdominal pain and co-morbid bladder dysfunction, but also endometriosis-associated pain and concurrent treatment of comorbid chronic pelvic pain syndromes.