842043 research-article20192019 TAG0010.1177/1756284819842043Therapeutic Advances in GastroenterologyN Tomsen, O Ortega Therapeutic Advances in Gastroenterology Original Research Acute and subacute effects of oropharyngeal Ther Adv Gastroenterol 2019, Vol. 12: 1–13 sensory stimulation with TRPV1 agonists DOI: 10.1177/ https://doi.org/10.1177/1756284819842043 https://doi.org/10.1177/1756284819842043 1756284819842043 in older patients with oropharyngeal © The Author(s), 2019. Article reuse guidelines: sagepub.com/journals- dysphagia: a biomechanical and permissions neurophysiological randomized pilot study Noemí Tomsen , Omar Ortega, Laia Rofes, Viridiana Arreola, Alberto Martin, Lluís Mundet and Pere Clavé Abstract Background: Older people with oropharyngeal dysphagia (OD) present a decline in pharyngeal sensory function. The aim of this proof-of-concept study was to assess the biomechanical and neurophysiological effects of acute and subacute oropharyngeal sensory stimulation with transient receptor potential vanilloid 1 (TRPV1) agonists (capsaicinoids) in older patients with OD. Methods: We studied the effect of a single dose versus multiple doses (2 weeks) of oral capsaicin treatment (10–5 M) or placebo in 28 older patients with OD (81.2 ± 4.6 years) using Correspondence to: Pere Clavé videofluoroscopy (penetration-aspiration scale [PAS], timing of swallow response) and Laboratori de Fisiologia Digestiva CIBERehd electroencephalography (EEG) (latency and amplitude of pharyngeal event-related potential CSdMUAB, Hospital de [ERP]). Mataró, Barcelona, Spain Centro de Investigación Results: Acute stimulation by capsaicinoids 10–5 M did not improve swallow function and did Biomédica en Red de not produce significant changes in pharyngeal ERP. In contrast, after 10 days of treatment, enfermedades hepáticas y digestivas (CIBERehd), patients presented a clinically relevant and statistically significant reduction in the laryngeal Instituto de Salud Carlos vestibule closure (LVC) time (22.5%, p = 0.042), and in the PAS (24.2%, p = 0.038), compared III, Barcelona, Spain Fundació Institut with the placebo group. EEG results showed a reduction in the latency of the N1 peak (28.6%, d’Investigació Germans p = 0.007) and an increase of the amplitude of the P1-N2 (59.4%, p = 0.038) and the N2- Trias i Pujol, Badalona, Spain P2 (43.6%, p = 0.050) peaks. We observed a strong and significant correlation between the pere.clave@ciberehd.org reduction in the latency of the N1 peak and change in LVC time after subacute treatment Noemí Tomsen Omar Ortega (r = 0.750, p = 0.003). Laboratori de Fisiologia Conclusions: After 2 weeks of treatment, oropharyngeal sensory stimulation with Digestiva CIBERehd CSdM-UAB, Hospital de capsaicinoids induced cortical changes that were correlated with improvements in swallowing Mataró, Barcelona, Spain biomechanics in older patients with OD. These results further show that sensory stimulation Centro de Investigación Biomédica en Red de by TRPV1 agonists can become a useful pharmacological treatment for older patients with OD. enfermedades hepáticas y digestivas (CIBERehd), Instituto de Salud Carlos III, Barcelona, Spain Keywords: deglutition disorders, neurophysiology, oropharyngeal dysphagia, therapeutics Laia Rofes Centro de Investigación Received: 30 October 2018; revised manuscript accepted: 12 March 2019. Biomédica en Red de enfermedades hepáticas y digestivas (CIBERehd), Instituto de Salud Carlos III, Barcelona, Spain Introduction by two European societies as it is highly prevalent Viridiana Arreola Alberto Martin Oropharyngeal dysphagia (OD) is a prevalent dis- in the older population, caused by multiple risk Lluís Mundet order among older people that leads to severe com- factors, associated with several comorbidities and Laboratori de Fisiologia Digestiva CIBERehd plications and impaired quality of life.1,2 OD has poor prognosis, and needs a multidisciplinary CSdM-UAB, Hospital de recently been recognized as a geriatric syndrome approach to treatment.3 The pathophysiology of Mataró, Barcelona, Spain journals.sagepub.com/home/tag 1 Creative Commons Non Commercial CC BY-NC: This article is distributed under the terms of the Creative Commons Attribution-NonCommercial 4.0 License (http://www.creativecommons.org/licenses/by-nc/4.0/) which permits non-commercial use, reproduction and distribution of the work without further permission provided the original work is attributed as specified on the SAGE and Open Access pages (https://us.sagepub.com/en-us/nam/open-access-at-sage). Therapeutic Advances in Gastroenterology 12 OD in the older population has been studied and stimulants, such as TRPV1 agonists, improved the impairment in the oropharyngeal swallow swallowing parameters.19–23 Despite these promis- response (OSR) is well known.4,5 Instrumental ing results it is not known if and how capsaicin diagnostic methods such as videofluoroscopy improves the neurophysiology of the afferent path- (VFS) have enabled the quantitative measurement way, which is crucial in order to design future ther- of the timing of the OSR4,6 to better understand its apies for older patients with OD. pathophysiology.7–9 The OSR in older patients is weak and slow, leading to impaired efficacy and We hypothesized that subacute administration of safety of swallow. Impaired efficacy (mainly oro- pharyngeal sensory stimulants, such as capsaici- pharyngeal residue) is caused by weak tongue noids, would improve cortical neuroplasticity, bolus propulsion, diminished tongue base move- measured by pharyngeal ERP to electrical stimula- ments, reduced pharyngeal muscular reserve, tion (latency/amplitude). This improvement would lower width opening of the upper oesophageal lead to a faster and greater conduction and integra- sphincter (UOS) and slow hyoid motion caused by tion of the sensory input resulting in a faster and muscular weakness and sarcopenia: impaired stronger swallow response. The aim of this proof- safety (penetrations and aspirations) is mainly due of-concept study was to assess the effect of acute to delayed laryngeal vestibule closure (LVC) and (single-dose protocol) and subacute (multiple- slow neural response.4,10–12 The motor pathway of dose protocol) treatment with capsaicinoids on the the neural control of swallow has been character- biomechanics and neurophysiology of older ized using transcranial magnetic stimulation, clari- patients with OD and to assess any possibility of fying the cortical control of swallowing muscles.13,14 desensitization after 2 weeks of treatment. However, the role of the afferent (sensory) path- way of deglutition is less well known although some studies suggest it has a key role in the patho- Methods physiology of OD.15–17 Recently we explored the pharyngeal event-related potential (ERP) to elec- Study population trical stimulation in several groups (young healthy A total of 28 older patients with OD associated volunteers and older patients with and without with aging were included in the study and patients OD) and found older people had a decline in phar- in each study were randomized with a specific yngeal sensory function, that was more severe in software (QuickCalcs 2018, GraphPad Software) older patients with OD.17 This sensory impairment (Supplementary Figure 1). Inclusion criteria were might be a critical pathophysiological element and for patients to be more than 70 years old, in a sta- a potential target for treating swallowing dysfunc- ble medical condition and to have clinical signs of tion in older patients. This hypothesis was con- OD according to the volume-viscosity swallow firmed in one of our studies where we found that test.24 The study protocol was approved by the administration of acute transient receptor potential Ethics Committee of the Hospital de Mataró vanilloid 1 (TRPV1) agonists (capsaicinoids (1.5–4 (protocol code CEIC04/12) and was conducted M)) improved swallow function (reduced penetra- according to the principles and rules laid down in tions and penetration-aspiration scale [PAS]) and the Declaration of Helsinki and its subsequent the biomechanics of the swallow response (LVC, amendments. Written informed consent was UOS opening and hyoid movement) in a group of obtained from all the participants. ClinicalTrials. older patients with OD.18 However this high con- gov registration code: NCT01762228. centration is pungent and not suitable for long- term studies. For this reason, we performed a subacute study (10 days treatment) with capsaici- Study design noids at a lower dose (10–5 M) and found an Studies were conducted in the Dysphagia Unit of improvement in the safety of swallow and OSR in the Hospital de Mataró (Barcelona, Spain). Two up to 68.42% of older patients with OD without study protocols were designed to evaluate the any complaints from the patients regarding pun- effects of acute and subacute treatment with oral gency.5 Regarding the subacute and acute treat- capsaicinoids 10–5 M on the neurophysiology of ments, it is important to state that the effect older patients with OD (Supplementary Figure 2). remained stable over the treatment period and they did not lose their effectivity due to desensitiza- Acute study, single-dose protocol. A total of 14 tion. Some other studies have found that sensory patients with OD were included in this protocol 2 journals.sagepub.com/home/tag N Tomsen, O Ortega et al. and randomized to capsaicinoids or placebo sensitization process on each patient was con- intervention. First, socio-demographic and clini- ducted 5 min before treatment by giving two cal data were collected: age, sex, Barthel index25 5 ml boluses supplemented with capsaicinoids (functional status), Mini Nutritional Assessment (10–5 M) 2 min apart (Supplementary Figure 2). short form26 (nutritional status) and Charlson index27 (comorbidities). All participants were Subacute studies. To assess the effect of suba- examined with VFS to assess their swallowing cute administrations of capsaicinoids on the bio- function and, after 5 days, underwent an electro- mechanics of swallow, we performed two VFS encephalographic (EEG) study to explore the (pre- and post-treatment). Patients were studied ERP while they received the TRPV1 agonist or during the deglutition of 5 ml, 10 ml and 20 ml placebo17 (Supplementary Figure 2). series of nectar (274.42 ± 13.14 mPa·s), liq- uid (20.40 ± 0.23 mPa·s) and pudding boluses Subacute study, multiple-dose protocol. A total of (3931.23 ± 166.15 mPa·s). Liquid viscosity was 14 patients with OD (> 70 years) were included obtained by mixing 1:1 mineral water and the in this protocol and randomly allocated to the X-ray contrast Gastrografin (Bayer Hispania SL, TRPV1 agonist or placebo group. All participants Barcelona, Spain). Nectar and pudding viscosity were studied with VFS and EEG before and 5 days were obtained by adding 3.5 g and 8 g, respec- after the treatment (Supplementary Figure 2). tively, of the thickener Resource ThickenUp (Nestlé Nutrition, Barcelona, Spain) to 100 ml of Intervention: oropharyngeal sensory stimulation liquid (1:1 water/contrast). Boluses were carefully with TRPV1 agonists. In the single-dose protocol, given to patients with a syringe.5 participants were given 10 ml of a nectar-like solution of capsaicinoids 10–5 M (active group), VFS analysis. VFS analysis was carried out blind as previously published,5 or placebo (potassium by a single observer. A good inter-rater correla- sorbate, the excipient of the capsaicinoids solu- tion has been described in the assessment of the tion) (control group). In the multiple-dose proto- signs of impaired safety of swallow we used in this col, participants were given 10 ml of a nectar-like study (κ = 0.7051) and intra-rater (0.9) and solution of capsaicinoids 10–5 M or placebo three inter-rater (0.9) reliability for the timing of times/day (before meals) for 2 weeks (5 days/ OSR.28–30 Digitization, analysis and measure- week). Capsaicinoid concentration in the capsa- ments of VFS images were made using the soft- icinoids sauce (McIlhenny Co, Avery Island, LA, ware Swallowing Observer (Image and Physiology USA) was 185.5 µg/g, measured with liquid chro- SL, Barcelona, Spain). For each swallow, we ana- matography (AOAC 995.03 method).18 lysed: (a) signs of swallowing efficacy; the pres- ence of oral or pharyngeal (vallecular and pyriform sinus) residue was assessed; (b) signs of VFS/swallow physiology swallowing safety: laryngeal vestibule penetra- VFS procedure. All patients were imaged seated, tions and tracheobronchial aspirations, classified in a lateral projection which included the oral cav- according to PAS, were assessed.4,31 Quantitative ity, pharynx, larynx and cervical oesophagus. VFS measurements of the OSR were obtained during recordings were obtained with a SuperXT-20 5 ml swallows: timing of the LVC and UOS open- Toshiba Intensifier (Toshiba Medical Systems ing were measured, the glossopalatal junction Europe, Zoetermeer, the Netherlands) using a opening was given the time value 04 (Supplemen- continuous fluoroscopy beam and recorded at 25 tary Figure 3). frames/s using a Panasonic AGDVX-100B video camera (Matsushita Electric Industrial Co, Osaka, Japan). EEG/ERP Electroencephalographic procedure. Acute studies. The single-dose protocol has been described previously.18 Patients were Single-dose protocol: the EEG consisted of two studied during the deglutition of one series of trains of 50 squared wave 0.2 ms electrical stimuli 5 ml, 10 ml and 20 ml nectar-viscosity boluses applied to the pharynx with an intrapharyngeal (274.42 ± 13.14 mPa·s) as a control and two catheter passed transnasally, with two bipolar series of 5 ml, 10 ml and 20 ml nectar boluses electrodes (Gaeltec Ltd, Dunvegan, Scotland) 1 supplemented with capsaicinoids (10–5 M). A cm apart. The electrodes were positioned to the journals.sagepub.com/home/tag 3 Therapeutic Advances in Gastroenterology 12 posterior pharyngeal wall, 14–15 cm from the ms epochs, including 100 ms of prestimulus base- nostrils and the catheter was connected to a line; a semi-automatic artefact-rejection method Digitimer DS7A current stimulator and DG2A was employed to prevent contamination from train/delay generator (Digitimer Ltd, Welwyn swallow movements; the interval from –100 ms to Garden City, UK). The stimulus intensity was –20 ms before the stimulus was used for baseline individualized for each participant by increasing correction; the epochs were averaged to obtain the electrical current intensity from 0 mA in steps the pharyngeal ERPs for each participant, and the of 0.5 mA, first determining the threshold inten- responses of the participants of each group were sity at which the participants perceived the stimu- averaged. Cortical representation of the ERPs lus (sensory threshold) and then increasing to the was shown using the registries from the software. tolerance threshold. Thresholds were determined The following time frames were used to compute by triplicate and the intensity applied to assess the cortical activity distribution for the different ERP was 75% of the tolerance threshold. The peaks: N1: 56–80 ms; P1: 120–150 ms; N2: 220– inter-stimulus interval was 5 s and the inter-set 270 ms; P2: 300–350 ms.17 We have shown the interval, 1 min. Following the two trains, partici- N1 and P2 peaks in the figures as they are the first pants were randomized into two groups. One was (afferent conduction) and last of the peaks (corti- given a single dose of 10 ml capsaicinoids 10-5 M cal integration), respectively. while the other received 10 ml placebo. Then, two new trains of stimuli with the same character- ERP source localization. To identify the topogra- istics were applied to the pharynx of all partici- phy of the brain source of each ERP component, pants (Supplementary Figure 2). the standardized low-resolution brain electromag- netic tomography software (sLORETA) (http:// Multiple-dose protocol: EEG protocol was per- www.uzh.ch/keyinst/loreta.htm) was used to com- formed as previously described after the subacute pute a standardized, discrete, three-dimensionally 10 days of treatment. During each EEG, partici- distributed, linear, minimum norm inverse solu- pants did not receive any stimulant between the tion. The particular form of standardization used four trains of stimuli (Supplementary Figure 2). provides the exact localization to test point sources, giving images of standardized current density but Pharyngeal ERP recording. Cortical responses to low spatial resolution.17 electrical stimuli were recorded through a cap with 32 scalp tin electrodes (Electro-Cap Interna- tional Inc, Eaton, OH, USA), an amplification of Safety of the treatment the 10–20 system,32 referenced to the left ear lobe During the study, no adverse events or serious and connected to a BrainAmp amplifier (Brain adverse events were recorded and reported to the Products GmbH, Gilching, Germany). The Ethics Committee of the Hospital de Mataró. ground electrode was included in the EEG cap and located just below the fronto-central elec- trode FCz. A disc electrode was placed on the Statistics skin below the left eye to record the vertical elec- Continuous variables are expressed as mean ± SD trooculogram. Electrode gel was applied to the and categorical variables as relative and absolute electrodes to keep impedance below 5 KΩ. The frequencies. Continuous variables were compared signal was digitized at a sampling rate of 500 Hz by the Student’s t test (for inter-group compari- and filtered with a 50 Hz notch. Recordings were sons) and by the paired t test (for pre/post-test performed in a quiet room with the subject seated, comparisons) or Friedman test (for control/T1 and awake and with eyes open. The subject was asked T2 treatment time points comparisons). to stay calm and relaxed. Categorical variables were compared by Fisher’s exact test. Nonparametric tests were used when Pharyngeal ERP analysis. The EEG was analysed appropriate (non-Gaussian variables). Correlation offline and processed with BrainVision Analyzer between LVC time and the latency of N1 pharyn- Software 2.0 (Brain Products GmbH) in the fol- geal ERP was determined with Spearman’s corre- lowing steps: the EEG was bandpass filtered lation coefficient. p value < 0.05 was considered between 0.5 Hz and 60 Hz; an independent com- statistically significant. After 10 days of treatment, ponent analysis was performed to correct eye patients were divided into responders and nonre- blink artefacts; the EEG was segmented into 600 sponders. Responders were defined as those 4 journals.sagepub.com/home/tag N Tomsen, O Ortega et al. Table 1. Socio-demographic and clinical characteristics of the population in the acute and subacute arms of the study. Data presented as mean ± SD unless specifically stated. Acute study Subacute study Group 1: Group 2: p value Group 1: Group 2: p value placebo capsaicinoids placebo capsaicinoids N 7 7 1.000 7 7 1.000 Age (years) 83.7 ± 3.9 83.5 ± 6.3 0.530 79.0 ± 5.7 79.8 ± 5.2 0.680 Sex (n, men) 5 4 1.000 5 4 0.550 Barthel index 79.2 ± 25.4 70 ± 33.7 0.567 55 ± 39.1 85.8 ± 16.8 0.200 MNA-sf 12.3 ± 1.9 9.5 ± 2.9 0.106 8.3 ± 3.8 11.7 ± 2.9 0.200 Charlson index 3.7 ± 2.6 3.8 ± 2.5 0.959 2.8 ± 1.9 2.0 ± 1.3 0.570 MNA-sf, Mini Nutritional Assessment short form. patients who, following treatment, achieved safe of intensity at which the stimulus was applied was swallow at a lower level of viscosity or improved at similar between groups (20.6 ± 7.8 mA for Group least one point in the PAS at the same viscosity.5 1 and 22.9 ± 12.0 mA for Group 2, p = 0.870). The sLORETA software package was used to A single dose of capsaicinoids 10–5 M did not pro- assess the differences in cortical localization duce significant changes in the pharyngeal ERP between pre/post-treatment results computed by (amplitude and latency) (Supplementary Table 1, voxel-by-voxel t tests for paired measures and cor- Figure 1). We found no correlations between the rected for multiple comparisons.17 biomechanics and neurophysiology of swallow response in the acute study. Results ERP source localization. In the acute study, no statistical differences were found. In patients who Acute stimulation: single-dose protocol received placebo, the N1 peak distribution showed Study population. A total of 14 older patients with a weak activation of the occipital lobe, the P1 OD (82.9 ± 3.2 years, 9 men) were included in peak showed a bilateral frontal distribution, the this arm of the study. Patients randomized to N2 peak had a bilateral frontotemporal cortical active (Group 2) or placebo (Group 1) treatment representation, and P2 had a right frontal activa- had similar ages, number of comorbidities, func- tion (Figure 1). In the capsaicin group, after treat- tional capacity and nutritional status (Table 1). ment, there was a centralization of the cortical activation in the N1 peak, P1 and N1 peaks had VFS results. VFS results showed a population an increased cortical representation in the left with a high prevalence of VFS signs of impaired frontotemporal area and finally, P2 had a moder- efficacy of swallow (mainly pharyngeal residue) ate increase in the activation of the right frontal and low prevalence of VFS signs of impaired lobe. safety of swallow (penetrations) with a moder- ately delayed timing of OSR. After the single-dose Using sLORETA, we found the basal anatomical treatment with capsaicinoids, patients did not activation in both groups of treatment, showing show any significant improvement when com- different localizations according to each one of pared with the pretreatment values (Table 2). the peaks from the pharyngeal ERPs: N1 peak activation was found in the middle temporal gyrus EEG results. The sensory threshold to electrical (Brodmann area [BA] 21); P1 and P2 in the stimulation was 13.0 ± 5.4 mA for Group 1 and medial frontal gyrus (BA10); N2 in the inferior 11.3 ± 4.5 mA for Group 2 (p = 0.410). The level frontal gyrus (BA47). journals.sagepub.com/home/tag 5 Therapeutic Advances in Gastroenterology 12 Table 2. Swallowing characteristics of the seven patients that received the active single-dose treatment. Data presented as % except for PAS score, LVC and UOS opening time (mean ± SD). Pretreatment Post-treatment p value (n = 7) T1 T2 Impaired efficacy (%) 85.7 57.1 57.1 0.424 Oral residue (%) 42.9 42.9 42.9 1.000 Pharyngeal residue (%) 85.7 42.9 42.9 0.174 Impaired safety (%) 28.6 28.6 0.0 0.291 Penetrations (%) 28.6 28.6 0.0 0.291 Aspirations (%) 0.0 0.0 0.0 1.000 PAS score 1.6±1.1 2±1.4 1±0.0 0.296 LVC time (ms) 251.4±72.0 245.7±106.9 211.43±44.5 0.302 UOS opening time (ms) 228.6±64.1 228.6±72.0 211.4±59.8 0.539 LVC, laryngeal vestibule closure; PAS, penetration-aspiration scale; UOS, upper oesophageal sphincter. Compared with basal activation, patients that found a responder rate of 71.43% (5) in Group received the active treatment showed a signifi- 2 (capsaicinoids group) and of 28.57% in Group cant reduction in cortical activity at the N1, P1 1 (2), but there were no significant differences in and N2 peaks (p = 0.0002) distributed in the responder rate between capsaicinoids and control following way: the N1 peak showed less activa- groups (p = 0.2861). tion at the anterior cingulate (BA24); P1 and N2 peaks at the paracentral lobule (BA6, premotor Efficacy signs. We observed alterations in the cortex); P2 at the cuneus (BA17, primary visual efficacy of 100% of the participants. Neither the cortex). patients treated with capsaicinoids nor the con- trol group presented changes after treatment (Table 3). Subacute stimulation: multiple-dose protocol Study population. A total of 14 additional older Oropharyngeal swallow response. Patients patients with OD were included in this arm of the in the control group (Group 1) did not present study (79.4 ± 5.2 years, 8 men), with similar changes in the timing of the OSR. In contrast, socio-demographic and clinical characteristics patients treated with capsaicinoids (Group 2), between patients that received the placebo (Group showed a statistically significant reduction in 1) and patients that receive capsaicinoids (Group the LVC time from 457.3 ± 46.8 ms to 354.3 2) (Table 1). ± 53.8 ms (p = 0.042) and the UOS opening time from 348.6 ± 88.6 ms to 285.7 ± 78.66 ms VFS results (p = 0.125) (Table 3). Safety signs. All patients showed impaired safety of swallow (PAS > 2) before receiving any EEG results. The sensory threshold of patients treatment. After 10 days of treatment with cap- included in Groups 1 and 2 was 11.0 ± 3.1 mA saicinoids, patients presented a significant reduc- and 7.0 ± 5.4 mA, respectively (p = 0.173). The tion in the PAS from 4.14 ± 0.4 to 3.14 ± 0.9 intensity level at which the stimulus was applied (p = 0.038) without changes in the prevalence of did not change between sessions: Group 1 aspirations and penetrations. Patients in Group received a stimulation intensity of 24.5 ± 8.1 mA 1 (control group) did not show changes (Table before treatment and 24.2 ± 8.5 mA after treat- 3). According to the established definition we ment (p = 1.00) and Group 2 received a 6 journals.sagepub.com/home/tag N Tomsen, O Ortega et al. Figure 1. Pharyngeal event-related potential (ERP) and scalp density maps to pharyngeal ERPs for acute (a) and subacute (b) studies. At the top of each treatment, ERP traces obtained at the Cz electrode for pretreatment (blue line) and post-treatment (red line) from the placebo group and capsaicinoids group after pharyngeal electrical stimulation are shown. Deflection at time point 0 corresponds to stimulus artefact. At the bottom, current scalp density maps at each ERP peak time point for each group are shown. journals.sagepub.com/home/tag 7 Therapeutic Advances in Gastroenterology 12 Table 3. Swallowing characteristics and latency and peak-to-peak amplitude of the pharyngeal ERP components at the Cz electrode of the placebo and capsaicinoids groups. Group 1: placebo p value Group 2: capsaicinoids p value Pretreatment Post-treatment Pretreatment Post- treatment n=7 n=7 n=7 n=7 Swallowing characteristics (VFS) Impaired efficacy (%) 85.7 85.7 1.000 100 85.7 1.000 Oral residue (%) 85.7 85.7 1.000 85.7 71.4 1.000 Pharyngeal residue (%) 71.4 57.1 1.000 85.7 71.4 1.000 Impaired safety (%) 100 100 1.000 100 71.4 0.462 Penetrations (%) 71.4 100 0.462 100 71.4 0.462 Aspirations (%) 28.6 14.3 1.000 0 0 1.000 PAS Score 4.9 ± 1.2 5 ± 1.0 1.000 4.14 ± 0.4 3.14 ± 0.9 0.038 LVC time (ms) 342.9 ± 142.1 388.6 ± 88.6 0.469 457.1 ± 46.8 354.3 ± 53.8 0.042 UOS opening time (ms) 274.3 ± 90.7 274.3 ± 81.4 0.936 348.6 ± 88.6 285.7 ± 74.6 0.125 Pharyngeal ERP (EEG) N1 latency (ms) 85.7 ± 15.4 81.7 ± 16.9 0.463 90 ± 18.6 64.3 ± 16.8 0.007 P1 latency (ms) 153.7 ± 35.0 157.4 ± 31.8 0.403 128 ± 20.6 112.3 ± 24.5 0.325 N2 latency (ms) 234.9 ± 37.4 218.9 ± 38.5 0.219 241.7 ± 68.9 257.4 ± 44.9 0.295 P2 latency (ms) 341.7 ± 35.3 333.4 ± 58.7 0.437 374 ± 70.9 388.6 ± 66.0 0.623 P1-N1 amplitude (μV) 3.6 ± 1.3 3.5 ± 1.3 0.850 2.2 ± 1.7 3.2 ± 1.7 0.311 P1-N2 amplitude (μV) 0.5 ± 1.1 1.1 ± 2.5 0.504 1.3 ± 1.4 3.2 ± 1.9 0.038 N2 -P2 amplitude (μV) 3.2 ± 4.0 3.4 ± 3.6 0.784 3.1 ± 1.0 5.5 ± 2.3 0.050 EEG, electroencephalography; ERP, event-related potential: LVC, laryngeal vestibule closure; PAS, penetration-aspiration scale; UOS, upper oesophageal sphincter; VFS, videofluoroscopy. stimulation of 16.08 ± 5.1 mA before treatment treatment (Figure 1, Table 3). Pretreatment and and 15.7 ± 5.6 mA after treatment (p = 0.872). post-treatment scalp distributions of the pharyn- geal ERP showed no statistically significant differ- A 2-week treatment with capsaicinoids induced sig- ences in Group 1 (Table 3). nificant changes in the pharyngeal ERP; a reduc- tion in the latency of the N1 peak (from 90.0 ± Correlation between VFS and EEG results. The 18.6 ms to 64.3 ± 16.8 ms, p = 0.007) and an reduction observed in the latency of the N1 com- increase in the amplitude of the P1-N2 (from 1.3 ± ponent of the pharyngeal ERP after the subacute 1.4 mV to 3.2 ± 1.9 mV, p = 0.038) and N2-P2 treatment strongly and significantly correlated (r = (from 3.1 ± 1.0 mV to 5.5 ± 2.3 mV, p = 0.050) 0.750; p = 0.003) with the reduction observed in peaks (Figure 1, Table 3). In contrast, patients in the LVC time, the main airway protection mecha- Group 1 did not show any change in their pharyn- nism during swallow. In contrast there was no cor- geal ERP or their cortical activation after the 2-week relation when we analysed the reduction observed 8 journals.sagepub.com/home/tag N Tomsen, O Ortega et al. Figure 2. Correlation between the change in LVC time (left) and UOS closure time (right) and the reduction in the latency of N1 peak after the multiple-dose treatment. LVC, laryngeal vestibule closure; pERP, pharyngeal event-related potential; UOS: upper oesophageal sphincter. in the latency of the same peak with that of the Safety of the treatment overall duration of the swallow response until the During the study there were no adverse or serious UOS closure time (r = −0.06026; p = 0.7503) adverse events. Thus, we concluded that our treat- (Figure 2). ments were safe for our older patients with OD. ERP source localization. Patients that received the placebo in the subacute study did not show Discussion changes in cortical activation distribution. The Results from this study further confirm that the bio- N1 peak had a bilateral cortical representation; mechanical airway protection mechanisms during the P1 peak had a frontoparietal and left frontal the swallow response (mainly LVC) are delayed in lobe representation; the N2 peak had a right older patients with OD and the cortical activation to temporal and frontoparietal distribution; the P2 pharyngeal sensory stimuli at ERP is also delayed, peak had a wide centroparietal representation impaired and reduced in this population. We also (Figure 1). After 2 weeks of treatment with cap- found that acute treatment with low doses of capsai- saicinoids, we also found statistically significant cinoids (10 µM) did not have any effect on the OSR differences (Table 3). N1 cortical activation or the pharyngeal ERP. In contrast, subacute treat- changed from bilateral frontal distribution to a ment with the same concentration of capsaicinoids frontoparietal and temporal distribution; P1 cor- induced significant cortical changes that correlated tical representation (not shown) changed from a with significant improvements in the OSR of older unilateral frontal representation to a more bilat- patients with OD, further suggesting that the use of eral frontal distribution; N2 temporal bilateral sensory stimulation by TRPV1 agonists could be a activity was reduced; P2 activity was strongly valid pharmacological strategy for these patients. increased with higher representation in the fron- However, although our results concur with our ini- toparietal lobes (Figure 1). tial hypothesis, we need to perform further studies and a randomized clinical trial with more patients to The basal anatomical activation was similar to confirm that these neuroplastic and biomechanical that described in the acute study. The localization changes in swallow function are caused specifically of each peak was: BA21 in the N1 peak, BA10 in by subacute TRPV1 stimulation. the P1 and P2 peak, and BA47 in the N2 peak. Our research strategy first characterized the After 2 weeks of treatment with capsaicinoids, we impaired biomechanics of the OSR in older people observed an increase in cortical activation at the using VFS and then the therapeutic effect of oro- N1, N2 and P2 peaks (p < 0.0001) represented pharyngeal sensory stimulation using TRP ago- as increased activity in the N1 and P2 peaks on nists. We found that impaired safety of deglutition the cingulate gyrus (BA31), in P1 at the paracen- and aspirations in older people are mainly caused tral lobule (BA5, somatosensory association cor- by delayed LVC.4 We also found that acute oro- tex), and N2 at the medial frontal gyrus pharyngeal sensory stimulation with natural capsai- (BA8 secondary motor cortex) (Figure 3). cinoids (150 µM) had the strongest therapeutic journals.sagepub.com/home/tag 9 Therapeutic Advances in Gastroenterology 12 Figure 3. Differences in LORETA source activity after acute stimulation (top) and subacute stimulation (bottom) compared with basal cortical activity. Coloured voxels represent areas of significant differences in activation (blue, increase; red, decrease) after correction for multiple comparisons. effect over other TRP agonists by significantly processing was closely associated with the reducing the prevalence of penetrations, pharyn- pathophysiology of OD in older patients.17 The geal residue, LVC time and increasing bolus veloc- preservation of the afferent pathway is essential ity in older patients with OD.18,23 However, in the for a safe and effective swallow, allowing con- present study we did not find such an improvement tinuous oropharyngeal feedback to higher-level in VFS results when applying acute stimulation cerebral centres and activation of sensorimotor with capsaicinoids at lower doses. This lower con- integration processes.33 In previous studies with centration (10 µM) compared with the previous a similar methodology, we found that older study (150 µM)18 was due to the poor palatability patients with OD had an impaired cortical felt by patients at this concentration due to pun- response to the pharyngeal electrical stimulus gency, and by the good results found in a previous compared with older people without dysphagia. subacute study with a 10 µM dose in older patients The amplitude of all peaks was clearly inferior with OD with a positive responder rate of 68.42%.5 in older patients with OD, while only the latency Moreover, the application of the same reduced of the N1 and N2 peaks was delayed,17 with val- concentration of capsaicinoids for 10 days signifi- ues quite similar to those observed in the pre- cantly reduced the severity of impaired safety alter- sent study. We did not find any significant effect ation measured by the PAS in 24.15% (p = 0.038) of acute stimulation on the latency or amplitude and the most relevant biomechanical element of the of ERP. In contrast, when we analysed pharyn- reconfiguration phase of the OSR, the LVC (p = geal ERP in the subacute study, we also found a 0.042) (~100 ms reduction). In addition we found significant reduction in the latency of the N1 a responder rate of 71.43%, similar to that of our peak, and an increase in the amplitude of the previous subacute study,5 further suggesting the P1-N2 and N2-P2 and changes in cortical acti- strong therapeutic effect of this sensory-stimulation vation after capsaicinoids treatment, indicating approach to developing future pharmacological an improvement in conduction (N1 and P1 treatments for OD. peaks) and integration (N2 and P2 peaks) of sensory information into the cortex. This sug- We previously found that the impaired conduc- gests that capsaicinoids treatment induces plas- tion/integration of the afferent pathway in cere- tic cortical changes that are translated into a bral structures involved in sensory pharyngeal safer and faster OSR. 10 journals.sagepub.com/home/tag N Tomsen, O Ortega et al. As we reported in a previous study, older patients also important to note that there were no desensi- with OD show an increase in prefrontal and asso- tization effects after 2 weeks of treatment. ciated-area activation to compensate for the impaired activation of the other brain areas Finally, there are some studies that correlate the involved in the OSR.17 When we analysed brain increase in substance P (SP) level in saliva after activity with sLORETA, we found a similar pharyngeal sensory stimulation.37 A recent publi- increase in frontal activation prior to the stimula- cation showed that after oral capsaicin treatment tion treatment. After acute stimulation, we found (0.7 µg during 7 days), there was an increase in a reduction in brain activity in nonrelevant areas salivary concentration of SP that was associated for deglutition. However, after subacute capsaici- with an improvement in the safety and efficacy of noids stimulation, we observed more activity in swallowing, and with a shortening in the OSR in the cingulate gyrus (limbic lobe) at the N1 and P2 older patients with OD.19 These results suggest peaks and in the medial frontal gyrus (frontal that SP can be used as a biomarker for neuro- lobe) at the N2 peak. Both areas, especially the stimulation treatment response to stimulants such cingulate gyrus, are mainly activated during swal- as capsaicin and will be an interesting factor to lowing preparation and play a major role in the take into account in future studies. perception of stimulus and go/no-go decisions,34 suggesting a neuroplasticity potential for future This study has some limitations, the main one OD treatments based on sensory stimulation. being that this is a proof of concept with few patients in each group. Further studies with larger We also found a strongly significant correlation sample sizes will be needed in order to confirm between LVC reduction time and N1 pharyngeal these results and with a second post-treatment ERP peak latency reduction, indicating that bio- neurophysiological evaluation at a longer follow- mechanical improvements seen with VFS corre- up time period to assess whether the observed late with neurophysiological improvements on neuroplasticity is maintained over time. In addi- the sensory side. On the other hand, we did not tion, patients from the acute study had a lower find this correlation when we compared the total severity of OD compared with those in the suba- duration of swallow with the same peak latency. cute study and this could have affected the results This result agrees with a previous publication of of the study as those patients from the acute study our group, which found that, after the application had less potential improvement margin than those of acute capsaicinoids treatment, the LVC time from the subacute. To solve this limitation, in was significantly reduced but the treatment did future studies with a similar design, we will rand- not modify the total duration of OSR.18 This indi- omize patients to the acute or subacute group cates that this stimulant affects the first phase of instead of randomizing only for treatment or pla- pharyngeal swallow (airway protection mecha- cebo to balance the severity of OD between both nisms and reconfiguration from a digestive to a therapeutic groups. respiratory pathway). Other studies have also found a close relationship Conclusion between sensory deficits and impaired OSR, indi- We have shown that impaired pharyngeal sensory cating that reduced sensory input is translated function is a key element in the pathophysiology into an impaired motor response in patients with of dysphagia in older patients and treatments OD.16,35,36 This pathophysiologically relevant fac- increasing the sensory input, such as TRPV1 ago- tor in OD has been shown to be reversed or nists, will play a major role in the future treatment improved with our treatment, suggesting the rel- of dysphagia. The acute effect of capsaicin and evance of sensory stimulation in the management capsaicinoids has been well studied by our group and treatment of older patients with OD and indi- but the chronic effect is not well known and this cating that subacute administration of capsaici- study shows an improvement in this direction. noids is a valid strategy to induce cortical changes that have a positive impact on swallow physiol- Future studies with larger patient samples, and ogy. Despite these good results, we still need to including salivary SP measurement, are needed to adjust the dosage to obtain the highest biome- better learn the chronic effects of these stimulants chanical improvements with the minimal-dose and to select the most appropriate dose to improve effect due to the pungency of capsaicinoids. It is dysphagia and avoid nutritional and respiratory journals.sagepub.com/home/tag 11 Therapeutic Advances in Gastroenterology 12 complications among older patients with OD. stimulation strategies after two weeks treatment The therapeutic paradigm for older patients with on older patients with oropharyngeal dysphagia. OD is now changing from compensatory to Dysphagia 2016; 31: 706–716. improving brain and swallow function. 6. Logemann JA, Rademaker AW, Pauloski BR, et al. Normal swallowing physiology as viewed Acknowledgements by videofluoroscopy and videoendoscopy. Folia We thank Mrs Jane Lewis for English editing Phoniatr Logop 1998; 50: 311–319. assistance and Dr Pascal Faber from the 7. Logemann J. Evaluation and treatment of University of Zurich for sLORETA support. swallowing disorders. Austin, TX: College-Hill Press, 1983. Funding This work was supported by Fundació Agrupació 8. Kahrilas PJ, Lin S, Rademaker AW, et al. Impaired deglutitive airway protection: a Mútua, Barcelona, CIBERehd, Instituto de videofluoroscopic analysis of severity and Salud Carlos III (ISCIII), Barcelona, Spain, mechanism. Gastroenterology 1997; 113: 1457– Fondo de Investigaciones Sanitarias, ISCIII 1464. (14/00453; 18/00241)/CIBERehd, and Programa de Estabilización de Investigadores/Programa de 9. Robbins J, Langmore S, Hind JA, et al. Intensificación de la Actividad Investigadora en Dysphagia research in the 21st century and el Sistema Nacional de Salud (INT15/00026; beyond: proceedings from Dysphagia Experts Meeting, August 21, 2001. J Rehabil Res Dev INT16/00111). 2002; 39: 543–548. Conflict of interest statement 10. Logemann JA, Pauloski BR, Alfred W, et al. The authors declare no conflicts of interest in Temporal and biomechanical characteristics of preparing this article. oropharyngeal swallow in younger and old men. J Speech Lang Hear Res 2000; 43: 1264–1674. Supplementary material 11. Logemann JA, Pauloski BR, Alfred W, et al. Supplementary material for this article is available Oropharyngeal swallow in younger and older online. women: videofluoroscopic analysis. J Speech Lang Hear Res 2002; 45: 434–445. ORCID iD 12. Robbins J, Humpal NS, Banaszynski K, et al. Noemí Tomsen https://orcid.org/0000-0001 Age-related differences in pressures generated -7149-7948 during isometric presses and swallows by healthy adults. Dysphagia 2016; 31: 90–96. 13. Hamdy S, Aziz Q, Rothwell JC, et al. The cortical References topography of human swallowing musculature in 1. Clavé P and Shaker R. Dysphagia: current reality health and disease. Nat Med 1996; 2: 1217–1224. and scope of the problem. Nat Rev Gastroenterol 14. Hamdy S, Aziz Q, Rothwell JC, et al. Cranial nerve Hepatol 2015; 12: 259–270. modulation of human cortical swallowing motor 2. Ekberg O, Hamdy S, Woisard V, et al. Social and pathways. Am J Physiol 1997; 272: G802–G808. psychological burden of dysphagia: its impact on 15. Rofes L, Cola PC and Clave P. The effects of diagnosis and treatment. Dysphagia 2002; 17: sensory stimulation on neurogenic oropharyngeal 139–146. dysphagia. J Gastroenterol Hepatol Res 2014; 3: 3. Baijens LWJ, Clavé P, Cras P, et al. European 1066–1072. society for swallowing disorders – European 16. Cabib C, Ortega O, Vilardell N, et al. Chronic Union Geriatric Medicine Society white paper: post-stroke oropharyngeal dysphagia is associated oropharyngeal dysphagia as a geriatric syndrome. with impaired cortical activation to pharyngeal Clin Interv Aging 2016; 11: 1403–1428. sensory inputs. Eur J Neurol 2017; 24: 1355–1362. 4. Rofes L, Arreola V, Romea M, et al. 17. Rofes L, Ortega O, Vilardell N, et al. Pathophysiology of oropharyngeal dysphagia in the Spatiotemporal characteristics of the pharyngeal frail elderly. Neurogastroenterol Motil 2010; 22: 1–9. event-related potential in healthy subjects and 5. Ortega O, Rofes L, Martin A, et al. A older patients with oropharyngeal dysfunction. comparative study between two sensory Neurogastroenterol Motil 2016; 29: 1–11. 12 journals.sagepub.com/home/tag N Tomsen, O Ortega et al. 18. Rofes L, Arreola V, Martin A, et al. Natural 28. Baijens LW, Speyer R, Passos VL, et al. capsaicinoids improve swallow response in older Swallowing in Parkinson patients versus patients with oropharyngeal dysphagia. Gut 2013; healthy controls: reliability of measurements in 62: 1280–1287. videofluoroscopy. Gastroenterol Res Pract 2011; 2011: 380682. 19. Nakato R, Manabe N, Shimizu S, et al. Effects of capsaicin on older patients with oropharyngeal 29. Hind JA, Gensler G, Brandt DK, et al. dysphagia: a double-blind, placebo-controlled, Comparison of trained clinician ratings with crossover study. Digestion 2017; 95: 210–220. expert ratings of aspiration on videofluoroscopic images from a randomized clinical trial. 20. Ebihara T, Sekizawa K, Nakazawa H, et al. Dysphagia. 2009; 24: 211–217. Capsaicin and swallow reflex. Lancet 1993; 341: 432. 30. Kelly AM, Drinnan MJ and Leslie P. Assessing penetration and aspiration: how do 21. Ebihara T, Takahashi H, Ebihara S, et al. videofluoroscopy and fiberoptic endoscopic Capsaicin troche for swallowing dysfunction in evaluation of swallowing compare? Laryngoscope older people. J Am Geriatr Soc 2005; 53: 824–828. 2007; 117: 1723–1727. 22. Ebihara T, Ebihara S, Yamazaki M, et al. 31. Rosenbek JC, Robbins JA, Roecker PDEB, et al. Intensive stepwise method for oral intake using A penetration-aspiration scale. Dysphagia 1996; a combination of transient receptor potential 11: 93–98. stimulation and olfactory stimulation inhibits the incidence of pneumonia in dysphagic older 32. Klem GH, Lüders HO, Jasper HH, et al. The adults. J Am Geriatr Soc 2010; 58: 196–198. ten-twenty electrode system of the International Federation. The International Federation of 23. Alvarez-Berdugo D, Rofes L, Arreola V, et al. Clinical Neurophysiology. Electroencephalogr Clin A comparative study on the therapeutic effect Neurophysiol Suppl 1999; 52: 3–6. of TRPV1, TRPA1, and TRPM8 agonists on swallowing dysfunction associated with aging and 33. Ertekin C. Voluntary versus spontaneous neurological diseases. Neurogastroenterol Motil swallowing in man. Dysphagia 2011; 26: 183– 2018; 30: e13185. 192. 24. Clavé P, Arreola V, Romea M, et al. Accuracy 34. Toogood JA, Smith RC, Stevens TK, et al. of the volume-viscosity swallow test for clinical Swallowing preparation and execution: insights screening of oropharyngeal dysphagia and from a delayed-response functional magnetic aspiration. Clin Nutr 2008; 27: 806–815. resonance imaging (fMRI) study. Dysphagia 2017; 32: 526–541. 25. Mahoney FI and Barthel DW. Functional evaluation: the Barthel index. Md State Med J 35. Alvarez-Berdugo D, Rofes L, Farré R, et al. 1965; 14: 56–61. Localization and expression of TRPV1 and TRPA1 in the human oropharynx and larynx. 26. Kaiser MJ, Bauer JM, Ramsch C, et al. Validation Neurogastroenterol Motil 2016; 28: 91–100. of the Mini Nutritional Assessment short-form (MNA-SF): a practical tool for identification of 36. Teismann IK, Suntrup S, Warnecke T, et al. nutritional status. J Nutr Health Aging 2009; 13: Cortical swallowing processing in early subacute 782–788. stroke. BMC Neurol 2011; 11: 34. 27. Charlson ME, Pompei P, Ales KL, et al. A new 37. Suntrup-Krueger S, Bittner S, Recker S, et al. Visit SAGE journals online method of classifying prognostic comorbidity in Electrical pharyngeal stimulation increases journals.sagepub.com/ home/tag longitudinal studies: development and validation. substance P level in saliva. Neurogastroenterol J Chronic Dis 1987; 40: 373–383. Motil 2016; 28: 855–860. SAGE journals journals.sagepub.com/home/tag 13