Thoracic Spinal Nerve Neuromodulation Therapy for Diabetic Gastroparesis: A Proof-of-Concept Study

Abstract

Diabetic gastroparesis (DGp) is associated with severe symptoms, impaired quality of life, and significant health care utilization.1Sharma A. et al.Curr Gastroenterol Rep. 2020; 22: 1-10Crossref Scopus (7) Google Scholar Treatment of DGp remains unsatisfactory. Metoclopramide is the only Food and Drug Administration–approved medication for gastroparesis. Gastric electrical stimulation is approved for compassionate use in refractory patients; however, it is invasive, lacks robust supporting evidence, does not consistently improve symptoms, and is associated with serious adverse events.2Levinthal D. et al.Auton Neurosci. 2017; 202: 45-55Abstract Full Text Full Text PDF PubMed Scopus (37) Google Scholar The National Institutes of Health Gastroparesis Registry has followed patients for more than a decade with significant insight. Despite expert management, only 28% of patients had a significant reduction in their symptom scores.3Pasricha P.J. et al.Gastroenterology. 2015; 149: 1762-1774Abstract Full Text Full Text PDF PubMed Scopus (74) Google Scholar Over 4 years, approximately 40% of patients normalize their gastric emptying without improvement in symptoms.4Pasricha P.J. et al.Gastroenterology. 2021; 160: 2006-2017Abstract Full Text Full Text PDF PubMed Scopus (46) Google Scholar There is a large unmet need for therapeutic options to treat DGp.Gastroparesis symptoms are late manifestations of diabetic autonomic neuropathy. Autonomic function test results differ between DGp subjects and subjects with diabetes with normal gastric emptying.5Nguyen L. et al.Neurogastroenterol Motil. 2020; 32e13810Crossref Scopus (13) Google Scholar The autonomic nervous system modulates the stomach and foregut through the parasympathetic innervation by the vagus nerve and sympathetic innervation by the greater sympathetic nerve, originating from T5-T10 levels of the spinal cord. Abnormal afferent visceral sensory signal processing is also associated with gastroparesis symptoms.6Brock C. et al.Diabetes Care. 2013; 36: 3698-3705Crossref Scopus (40) Google Scholar Despite having severe gastroparesis symptoms, subjects with DGp have severe visceral hyposensitivity with increased sensory thresholds to electrical stimulation of the gut, implying a discordance between symptoms and visceral sensation.6Brock C. et al.Diabetes Care. 2013; 36: 3698-3705Crossref Scopus (40) Google Scholar Midthoracic spinal nerves are an ideal site for neuromodulation, because they carry preganglionic sympathetic nerves to the stomach (precursors to the greater splanchnic nerve) and general visceral afferent nerves (nerve fibers of dorsal root ganglia) from the stomach. Neuromodulation at this location could correct underlying diabetic autonomic and visceral neuropathy to improve symptoms in subjects with DGp.Our group has expertise in developing a promising neuromodulation treatment with repetitive magnetic stimulation of paraspinal regions.7Rao S.S. et al.Am J Gastroenterol. 2021; 116: 162-170Crossref PubMed Scopus (0) Google Scholar Translumbosacral neuromodulation therapy corrects underlying bidirectional gut-brain neuropathy (efferent spinoanorectal neuropathy and afferent anocortical neuropathy) and anal sphincter dysfunction by treatment of lumbosacral spinal nerve roots to improve fecal incontinence. In a similar fashion, targeting bilateral nerve roots at the midthoracic level with repetitive magnetic stimulation should reverse efferent spinogastric neuropathy and afferent gastrocortical neuropathy to improve symptoms of gastroparesis. The aim of this study was to assess safety and feasibility of a novel, noninvasive treatment, thoracic spinal nerve magnetic neuromodulation therapy (ThorS-MagNT) for refractory DGp in a proof-of-concept study. ThorS-MagNT is a novel neuromodulation treatment, which uses a magnetic coil to deliver repetitive stimulations to spinal nerves in the midthoracic, paraspinal regions of the back in a similar fashion to transcranial magnetic stimulation delivering repetitive magnetic stimulation to various regions of the skull. Transcranial magnetic stimulation is supported by strong evidence for the treatment of depression and neuropathic pain.8Lefaucheur J.-P. et al.Clin Neurophysiol. 2014; 125: 2150-2206Crossref PubMed Scopus (1252) Google ScholarFor details of study design and methodology, please refer to the Supplementary Material. Seven DGp patients (6 female, 1 male) with moderate-severe gastroparesis (average baseline total ANMS Gastroparesis Cardinal Symptom Index Daily Diary [ANMS GCSI-DD] ≥2.0) were enrolled. Patient demographics, pertinent past medical history details, medication histories, and response to treatment are shown in Supplementary Table 1. All subjects completed treatment and tolerated treatment sessions. All subjects (n = 4) treated at optimal intensities of 150% above motor threshold met Food and Drug Administration responder definition of achieving greater than 30% reduction in total ANMS GCSI-DD. These responders had an average of 74.9% reduction in total ANMS GCSI-DD score from baseline (Figure 1A) versus 10.1% reduction in subjects treated with low intensities (at threshold level). None of the subjects had worsening of symptoms. Symptom improvement persisted 2 weeks posttreatment with an average of 97.9% symptom reduction in responders (Figure 1A) versus 0.3% reduction in subjects treated with low intensities. ThorS-MagNT responders had greater than 60.0% improvement in all 4 gastroparesis cardinal symptom domains (nausea, early satiety, postprandial fullness, and upper abdominal pain) and 83.3% improvement in number of vomiting episodes (Figure 1B). The most improved symptom was nausea (77.8%), as seen with prior gastric electrical stimulation studies.2Levinthal D. et al.Auton Neurosci. 2017; 202: 45-55Abstract Full Text Full Text PDF PubMed Scopus (37) Google Scholar Upper abdominal pain, a refractory gastroparesis symptom, also significantly improved by 73.8% from baseline. Three of 4 responders had a 60.4% improvement in t1/2 on gastric emptying breath test. One subject with known hypertension experienced medication-related hypotension, unrelated to study treatment. After resuscitation, this subject completed 4 remaining treatments. Another subject reported self-limited tingling and numbness, which resolved after treatment. No other adverse events occurred. All 7 subjects tolerated the treatment well and marked “agree” with the statement “ThorS-MagNT treatment was easy and convenient.”In our study, we have outlined a novel technique of delivering a promising, noninvasive, safe, and feasible neuromodulatory treatment, ThorS-MagNT for subjects with refractory DGp. Subjects treated with optimal treatment intensity demonstrate an excellent therapeutic response with a significant and sustained improvement in gastroparesis cardinal symptoms and report excellent tolerance. There is a high placebo response in gastroparesis and other disorders of gut-brain interaction.9Kaptchuk T.J. et al.BMJ. 2008; 336: 999-1003Crossref PubMed Scopus (883) Google Scholar Symptom response in our study was durable, persisting greater than 21 days posttreatment. Our clinical observation is that ThorS-MagNT responders maintain symptom relief for greater than 6 months. This durable response to treatment suggests that a placebo effect is less likely. Nonetheless, a sham comparison is necessary. Limitations of our study include its unblinded nature and small sample size. This proof-of-concept study of ThorS-MagNT in subjects with DGp offers promise of a noninvasive, safe, efficacious, and durable treatment for a debilitating condition. The anticipated results of a larger, multisite, randomized, sham-controlled trial, the TNM-DGp trial (Thoracic Neuromodulation for Diabetic Gastroparesis; Clinicaltrials.gov NCT05273788), will further shed light on the safety, feasibility, and efficacy and allow better understanding of the mechanism of action of thoracic neuromodulation. Diabetic gastroparesis (DGp) is associated with severe symptoms, impaired quality of life, and significant health care utilization.1Sharma A. et al.Curr Gastroenterol Rep. 2020; 22: 1-10Crossref Scopus (7) Google Scholar Treatment of DGp remains unsatisfactory. Metoclopramide is the only Food and Drug Administration–approved medication for gastroparesis. Gastric electrical stimulation is approved for compassionate use in refractory patients; however, it is invasive, lacks robust supporting evidence, does not consistently improve symptoms, and is associated with serious adverse events.2Levinthal D. et al.Auton Neurosci. 2017; 202: 45-55Abstract Full Text Full Text PDF PubMed Scopus (37) Google Scholar The National Institutes of Health Gastroparesis Registry has followed patients for more than a decade with significant insight. Despite expert management, only 28% of patients had a significant reduction in their symptom scores.3Pasricha P.J. et al.Gastroenterology. 2015; 149: 1762-1774Abstract Full Text Full Text PDF PubMed Scopus (74) Google Scholar Over 4 years, approximately 40% of patients normalize their gastric emptying without improvement in symptoms.4Pasricha P.J. et al.Gastroenterology. 2021; 160: 2006-2017Abstract Full Text Full Text PDF PubMed Scopus (46) Google Scholar There is a large unmet need for therapeutic options to treat DGp. Gastroparesis symptoms are late manifestations of diabetic autonomic neuropathy. Autonomic function test results differ between DGp subjects and subjects with diabetes with normal gastric emptying.5Nguyen L. et al.Neurogastroenterol Motil. 2020; 32e13810Crossref Scopus (13) Google Scholar The autonomic nervous system modulates the stomach and foregut through the parasympathetic innervation by the vagus nerve and sympathetic innervation by the greater sympathetic nerve, originating from T5-T10 levels of the spinal cord. Abnormal afferent visceral sensory signal processing is also associated with gastroparesis symptoms.6Brock C. et al.Diabetes Care. 2013; 36: 3698-3705Crossref Scopus (40) Google Scholar Despite having severe gastroparesis symptoms, subjects with DGp have severe visceral hyposensitivity with increased sensory thresholds to electrical stimulation of the gut, implying a discordance between symptoms and visceral sensation.6Brock C. et al.Diabetes Care. 2013; 36: 3698-3705Crossref Scopus (40) Google Scholar Midthoracic spinal nerves are an ideal site for neuromodulation, because they carry preganglionic sympathetic nerves to the stomach (precursors to the greater splanchnic nerve) and general visceral afferent nerves (nerve fibers of dorsal root ganglia) from the stomach. Neuromodulation at this location could correct underlying diabetic autonomic and visceral neuropathy to improve symptoms in subjects with DGp. Our group has expertise in developing a promising neuromodulation treatment with repetitive magnetic stimulation of paraspinal regions.7Rao S.S. et al.Am J Gastroenterol. 2021; 116: 162-170Crossref PubMed Scopus (0) Google Scholar Translumbosacral neuromodulation therapy corrects underlying bidirectional gut-brain neuropathy (efferent spinoanorectal neuropathy and afferent anocortical neuropathy) and anal sphincter dysfunction by treatment of lumbosacral spinal nerve roots to improve fecal incontinence. In a similar fashion, targeting bilateral nerve roots at the midthoracic level with repetitive magnetic stimulation should reverse efferent spinogastric neuropathy and afferent gastrocortical neuropathy to improve symptoms of gastroparesis. The aim of this study was to assess safety and feasibility of a novel, noninvasive treatment, thoracic spinal nerve magnetic neuromodulation therapy (ThorS-MagNT) for refractory DGp in a proof-of-concept study. ThorS-MagNT is a novel neuromodulation treatment, which uses a magnetic coil to deliver repetitive stimulations to spinal nerves in the midthoracic, paraspinal regions of the back in a similar fashion to transcranial magnetic stimulation delivering repetitive magnetic stimulation to various regions of the skull. Transcranial magnetic stimulation is supported by strong evidence for the treatment of depression and neuropathic pain.8Lefaucheur J.-P. et al.Clin Neurophysiol. 2014; 125: 2150-2206Crossref PubMed Scopus (1252) Google Scholar For details of study design and methodology, please refer to the Supplementary Material. Seven DGp patients (6 female, 1 male) with moderate-severe gastroparesis (average baseline total ANMS Gastroparesis Cardinal Symptom Index Daily Diary [ANMS GCSI-DD] ≥2.0) were enrolled. Patient demographics, pertinent past medical history details, medication histories, and response to treatment are shown in Supplementary Table 1. All subjects completed treatment and tolerated treatment sessions. All subjects (n = 4) treated at optimal intensities of 150% above motor threshold met Food and Drug Administration responder definition of achieving greater than 30% reduction in total ANMS GCSI-DD. These responders had an average of 74.9% reduction in total ANMS GCSI-DD score from baseline (Figure 1A) versus 10.1% reduction in subjects treated with low intensities (at threshold level). None of the subjects had worsening of symptoms. Symptom improvement persisted 2 weeks posttreatment with an average of 97.9% symptom reduction in responders (Figure 1A) versus 0.3% reduction in subjects treated with low intensities. ThorS-MagNT responders had greater than 60.0% improvement in all 4 gastroparesis cardinal symptom domains (nausea, early satiety, postprandial fullness, and upper abdominal pain) and 83.3% improvement in number of vomiting episodes (Figure 1B). The most improved symptom was nausea (77.8%), as seen with prior gastric electrical stimulation studies.2Levinthal D. et al.Auton Neurosci. 2017; 202: 45-55Abstract Full Text Full Text PDF PubMed Scopus (37) Google Scholar Upper abdominal pain, a refractory gastroparesis symptom, also significantly improved by 73.8% from baseline. Three of 4 responders had a 60.4% improvement in t1/2 on gastric emptying breath test. One subject with known hypertension experienced medication-related hypotension, unrelated to study treatment. After resuscitation, this subject completed 4 remaining treatments. Another subject reported self-limited tingling and numbness, which resolved after treatment. No other adverse events occurred. All 7 subjects tolerated the treatment well and marked “agree” with the statement “ThorS-MagNT treatment was easy and convenient.” In our study, we have outlined a novel technique of delivering a promising, noninvasive, safe, and feasible neuromodulatory treatment, ThorS-MagNT for subjects with refractory DGp. Subjects treated with optimal treatment intensity demonstrate an excellent therapeutic response with a significant and sustained improvement in gastroparesis cardinal symptoms and report excellent tolerance. There is a high placebo response in gastroparesis and other disorders of gut-brain interaction.9Kaptchuk T.J. et al.BMJ. 2008; 336: 999-1003Crossref PubMed Scopus (883) Google Scholar Symptom response in our study was durable, persisting greater than 21 days posttreatment. Our clinical observation is that ThorS-MagNT responders maintain symptom relief for greater than 6 months. This durable response to treatment suggests that a placebo effect is less likely. Nonetheless, a sham comparison is necessary. Limitations of our study include its unblinded nature and small sample size. This proof-of-concept study of ThorS-MagNT in subjects with DGp offers promise of a noninvasive, safe, efficacious, and durable treatment for a debilitating condition. The anticipated results of a larger, multisite, randomized, sham-controlled trial, the TNM-DGp trial (Thoracic Neuromodulation for Diabetic Gastroparesis; Clinicaltrials.gov NCT05273788), will further shed light on the safety, feasibility, and efficacy and allow better understanding of the mechanism of action of thoracic neuromodulation. Supplementary MaterialRefractory diabetic gastroparesis (DGp) subjects were recruited into our study approved by the Augusta University Institutional Review Board (IRB No. 1623939) and registered in Clinicaltrials.gov (NCT04706832). Inclusion criteria were previously diagnosed DGp in subjects less than 85 years with refractory symptoms despite treatment or intolerance to treatment. Exclusion criteria were postsurgical or idiopathic gastroparesis, gastrointestinal obstruction or presence of gastric bezoar, active inflammatory bowel disease or other mucosal disease, use of opioids, active depression, severe cardiac disease and arrhythmias, presence of metal implants (gastric electrical stimulators, deep brain stimulators, sacral nerve stimulators, or cardiac pacemakers), and pregnant women or nursing mothers. If dosage was stable for greater than 3 months, patients were permitted to continue medications affecting gastric emptying. If subjects met inclusion criteria, they signed an informed consent and were enrolled into the study. Subjects kept a 7-day baseline ANMS Gastroparesis Cardinal Symptom Index Daily Diary (ANMS GCSI-DD) before treatment and for an additional 21 days, which was completed every evening 1 hour after dinner. The ANMS GCSI-DD is a validated patient-reported outcome measure of cardinal symptoms of gastroparesis, including nausea, early satiety, postprandial fullness and upper abdominal pain (0–4, increasing severity), and number of vomiting episodes (0–4, with greater than 4 episodes recorded as 4).1Revicki D.A. et al.Neurogastroenterol Motil. 2019; 31e13553Crossref Scopus (5) Google Scholar The Food and Drug Administration responder definition for gastroparesis clinical trials is a 30% or greater reduction in 7-day total ANMS GCSI-DD score. Subjects also completed a device tolerability questionnaire during treatment.ThorS-MagNT treatment was administered twice daily for 5 consecutive days, following evidence-based transcranial magnetic stimulation treatment schedules for depression and neuropathic pain.2Lefaucheur J.-P. et al.Clin Neurophysiol. 2014; 125: 2150-2206Crossref PubMed Scopus (1257) Google Scholar The inferior angles of the scapula served as landmarks for the T7 level. The resting motor thresholds were obtained using a single-pulse, circular, 90-mm, stimulation coil (Magstim 200, MAGSTIM, Whitland, UK) to determine the minimum intensity of stimulation required to achieve a motor response of 10 μV with 50% of trials in the upper rectus abdominis and external oblique muscles. If obtained, the maximum intensity used for ThorS-MagNT treatment did not exceed 150% greater than this motor threshold, in accordance with repetitive magnetic stimulation safety guidelines. Thoracic stimulations were administered with a 70-mm double air-film self-cooling coil (MAGSTIM Rapid2, MAGSTIM) positioned on marked bilateral locations, held in place by a coil fixator to deliver, as shown in Supplementary Figure 1. Stimulation settings were 2 trains of 300 pulses at 1 Hz frequency with an intermittent 5-minute rest period (total dose 1200 stimulations per session) for 10 total sessions over 5 days. Treatment frequency was derived from our laboratory’s prior repetitive magnetic treatment experience of the lumbosacral region in a dose-ranging study for fecal incontinence, demonstrating that the 1 Hz was the most efficacious frequency setting.3Rao S.S. et al.Am J Gastroenterol. 2021; 116: 162-170Crossref PubMed Scopus (7) Google Scholar Patients were continuously monitored via cardiorespiratory monitoring throughout the study and closely observed for any adverse events. Gastric emptying was assessed through gastric emptying breath tests 1 week before and 2 weeks after treatment, as shown in Supplementary Figure 1.Supplementary Table 1Patient Demographics, Pertinent Past Medical History Details, Medication Histories, Responder Status, and Magnitude of Response Defined as Percent Symptom (Total ANMS GCSI-DD score) Improvement From BaselineSubject 1Subject 2Subject 3Subject 4Subject 5Subject 6Subject 7Age24485748255660GenderFFFFFFMType of diabetesType 1Type 2Type 2Type 2Type 1Type 2Type 2Duration of diabetic gastroparesis3 y2 y6 y8 y3 y3 y1 yHbA1c97.46.45.69.56.16.8ComorbiditiesAdjustment disorderEndometriosisAnxietyHyperlipidemiaHypertensionCADMigrainesHypertriglyceridemiaObstructive sleep apneaAdrenocortical insufficiencyHypothyroidismGoutDepressionHypothyroidismOsteochondritis dissecansHypertensionBipolar 1 disorderHypertensionMedications (stable dosage >3 mo)CarafateCarvedilolFerrous sulfateGabapentinHydroxyzineInsulinLosartanPantoprazolePromethazinePrucaloprideAlbuterolBudesonide-formoterol inhalationBuprenorphineCitalopramEstradiolInsulinNortriptylinePantoprazolePhenazopyridinePregabalinPromethazineSimvastatinSpironolactoneTrazodoneValsartanAspirinAtorvastatinBrexpiprazoleCarvedilolCetirizineCholecalciferolCholestyramineErenumab-aooe injectionFluticasone nasalFurosemideIsosorbide mononitrateLevomilnacipranLevothyroxineLosartanNitroglycerinPantoprazolePioglitazoneTizanidineTopiramateTrazodoneValbenazineAllopurinolAtorvastatinBuspironeColchicineDapagliflozinDextroamphetamineDuloxetineFamotidineFlonaseGabapentinInsulinLevothyroxineLinaclotideLoratadineLosartanMetforminOndansetronPolyethylene glycolSumatriptanTrazodoneZofranInsulinPantoprazoleAtorvastatinCyclobenzaprineDexlansoprazoleDulaglutideDuloxetineEmpagliflozinGabapentinGlimepirideLamotrigineLinaclotideLurasidonePromethazineSitagliptinTrazodoneAmlodipineCarvedilolGabapentinGlipizideInsulinLubiprostoneNortriptylinePantoprazolePromethazineTadalafilOutcomeNonresponderNonresponderNonresponderResponderResponderResponderResponder% improvement15.620.57.475.676.966.653.8 Open table in a new tab Refractory diabetic gastroparesis (DGp) subjects were recruited into our study approved by the Augusta University Institutional Review Board (IRB No. 1623939) and registered in Clinicaltrials.gov (NCT04706832). Inclusion criteria were previously diagnosed DGp in subjects less than 85 years with refractory symptoms despite treatment or intolerance to treatment. Exclusion criteria were postsurgical or idiopathic gastroparesis, gastrointestinal obstruction or presence of gastric bezoar, active inflammatory bowel disease or other mucosal disease, use of opioids, active depression, severe cardiac disease and arrhythmias, presence of metal implants (gastric electrical stimulators, deep brain stimulators, sacral nerve stimulators, or cardiac pacemakers), and pregnant women or nursing mothers. If dosage was stable for greater than 3 months, patients were permitted to continue medications affecting gastric emptying. If subjects met inclusion criteria, they signed an informed consent and were enrolled into the study. Subjects kept a 7-day baseline ANMS Gastroparesis Cardinal Symptom Index Daily Diary (ANMS GCSI-DD) before treatment and for an additional 21 days, which was completed every evening 1 hour after dinner. The ANMS GCSI-DD is a validated patient-reported outcome measure of cardinal symptoms of gastroparesis, including nausea, early satiety, postprandial fullness and upper abdominal pain (0–4, increasing severity), and number of vomiting episodes (0–4, with greater than 4 episodes recorded as 4).1Revicki D.A. et al.Neurogastroenterol Motil. 2019; 31e13553Crossref Scopus (5) Google Scholar The Food and Drug Administration responder definition for gastroparesis clinical trials is a 30% or greater reduction in 7-day total ANMS GCSI-DD score. Subjects also completed a device tolerability questionnaire during treatment. ThorS-MagNT treatment was administered twice daily for 5 consecutive days, following evidence-based transcranial magnetic stimulation treatment schedules for depression and neuropathic pain.2Lefaucheur J.-P. et al.Clin Neurophysiol. 2014; 125: 2150-2206Crossref PubMed Scopus (1257) Google Scholar The inferior angles of the scapula served as landmarks for the T7 level. The resting motor thresholds were obtained using a single-pulse, circular, 90-mm, stimulation coil (Magstim 200, MAGSTIM, Whitland, UK) to determine the minimum intensity of stimulation required to achieve a motor response of 10 μV with 50% of trials in the upper rectus abdominis and external oblique muscles. If obtained, the maximum intensity used for ThorS-MagNT treatment did not exceed 150% greater than this motor threshold, in accordance with repetitive magnetic stimulation safety guidelines. Thoracic stimulations were administered with a 70-mm double air-film self-cooling coil (MAGSTIM Rapid2, MAGSTIM) positioned on marked bilateral locations, held in place by a coil fixator to deliver, as shown in Supplementary Figure 1. Stimulation settings were 2 trains of 300 pulses at 1 Hz frequency with an intermittent 5-minute rest period (total dose 1200 stimulations per session) for 10 total sessions over 5 days. Treatment frequency was derived from our laboratory’s prior repetitive magnetic treatment experience of the lumbosacral region in a dose-ranging study for fecal incontinence, demonstrating that the 1 Hz was the most efficacious frequency setting.3Rao S.S. et al.Am J Gastroenterol. 2021; 116: 162-170Crossref PubMed Scopus (7) Google Scholar Patients were continuously monitored via cardiorespiratory monitoring throughout the study and closely observed for any adverse events. Gastric emptying was assessed through gastric emptying breath tests 1 week before and 2 weeks after treatment, as shown in Supplementary Figure 1.

Topics & Concepts