Published ahead of print on June 15, 2007, doi:10.1164/rccm.200612-1856OC
© 2007 American Thoracic Society doi: 10.1164/rccm.200612-1856OC
Representation of Capsaicin-evoked Urge-to-Cough in the Human Brain Using Functional Magnetic Resonance Imaging1 The Howard Florey Institute, University of Melbourne, Parkville, Victoria, Australia Correspondence and requests for reprints should be addressed to Stuart B. Mazzone, Ph.D., The Howard Florey Institute, University of Melbourne, Parkville, Victoria, Australia 3010. E-mail: smazzone{at}florey.edu.au
Rationale: Coughing in humans is typically preceded by a desire (or urge) to cough. The neural circuitry involved in sensing airway irritation and generating the urge-to-cough in humans is essentially unknown. Objectives: The aim of the present study was to use functional brain imaging to describe the supramedullary regions that are activated in humans during capsaicin inhalation. Methods: Experiments were performed on 10 healthy subjects (5 males, 5 females). Capsaicin doses were individually tailored to evoke a transient and reversible urge-to-cough. Blood oxygen level–dependent (BOLD) functional magnetic resonance measures were collected during repeated 24-second challenges with capsaicin or saline inhalation and subjects were asked to rate the urge-to-cough intensity of each challenge. Measurements and Main Results: Capsaicin inhalation reliably evoked an urge-to-cough, which was associated with activations in a variety of brain regions, including the insula cortex, anterior midcingulate cortex, primary sensory cortex, orbitofrontal cortex, supplementary motor area, and cerebellum. Conclusions: These data provide the first insights into the cortical neuronal network involved in sensing airway irritation and modulating coughing in humans.
Key Words: cough reflex • supramedullary • placebo • interoception • BOLD signal • cortex
Coughing is typically thought of as a protective reflex response to airway irritation that is integrated predominately at the level of the brainstem. Studies in cats and guinea pigs show that cough and coughlike responses can be evoked in decerebrated or deeply anesthetized animals, suggesting that supramedullary influence may not be essential for producing the basic cough motor pattern (1–4). However, it is also clear that coughing can be voluntarily initiated and inhibited (5, 6), is highly susceptible to placebo-induced suppression (7), is lost or severely diminished (for most stimuli) during general anesthesia or sleep (3, 8), and is typically preceded by an awareness of the irritating stimulus and perceived need to cough (termed the urge-to-cough) (4, 9). This strongly suggests that inputs from higher brain regions are likely involved in sensing airway irritations and provide important descending regulatory control over the basic cough reflex motor event. Despite the circumstantial evidence for an involvement of higher brain centers in coughing, few studies have attempted to experimentally explore the nature of this control. In anesthetized cats, coughlike efforts can be evoked after electrical stimulation of the suprasylvian gyrus or amygdala, and reflex cough (evoked by electrical stimulation of sensory nerves in the superior laryngeal nerves) is inhibited by concomitant stimulation of the cingulate gyrus or orbital gyrus (10). Although these findings provide a conceptual framework for the existence of descending regulatory cough pathways, it is unclear how these data translate to human cough responses because there have been no reports of evoked coughing after stimulation of brain regions in humans (reviewed in Reference 11). Recently, there has been an interest in the conscious perception of airway irritation that humans experience with coughing. Davenport and colleagues showed with capsaicin inhalation in conscious human subjects that the urge-to-cough always precedes a cough motor event and, importantly, is evoked at a lower threshold than cough itself (9). Similar results were reported after activation of pulmonary afferents with lobeline in humans (4), which has led to the suggestion that cough may consist of a behavioral element in addition to the reflex response initiated from the airways (9, 11). Nevertheless, essentially nothing is known about the neural substrates that may be involved in this behavioral element. In the present study, we used functional brain imaging techniques to explore the brain regions activated in humans after capsaicin inhalation–evoked urge-to-cough.
Subject Recruitment and Cough Threshold Testing Recruitment and experimental procedures were approved by the Howard Florey Institute Human Research Ethics Committee. Ten (five females, five males) nonsmoking, healthy adult subjects, with no history of chronic respiratory or neurological disease, participated in the study after providing informed consent (Table 1). No subjects were taking medications likely to influence blood oxygen level–dependent (BOLD) responses. Each subject underwent an initial capsaicin cough threshold test. Subjects were fitted with a modified face mask that was attached via tubing to two jet nebulizers (RapidFlo; Allersearch, Scoresby, Victoria, Australia) in series with medical air (flow rate = 5 L/min). Thirty-second blocks of normal tidal breathing with doubling doses of capsaicin (starting at 0.49 µM), with 5 minutes between challenges, were used to determine the capsaicin concentration needed to evoke two or more cough efforts (denoted C2). The capsaicin dose immediately before the C2 dose was used in all subsequent challenges for imaging the urge-to-cough.
Image Acquisition and Experimental Protocol Brain images were acquired with a Siemens Trio 3T scanner (Siemens AG, Erlangen, Germany) at the Murdoch Children's Research Institute, Melbourne, Australia. Anatomical images were obtained from all subjects (echo time [TE] = 2.6 ms, repeat time [TR] = 1,900 ms; flip angle = 9°; 176 sagittal slices, 1 x 0.5 x 0.5 mm3). Functional brain imaging was performed using BOLD contrast. Subjects were fitted with a respiratory monitor that provided online respiratory rate measurements throughout the experiments. Echo planar images (TE = 40 ms, TR = 2,000 ms; flip angle = 90°; 29 transaxial slices, 3.3 x 3.3 x 4.0 mm3) were acquired during tidal inhalation of the subject's individualized concentration of either capsaicin or saline. Subjects were told to breathe through their mouths but were not instructed on how to respond during the challenges (i.e., they were not told to actively suppress cough). Three functional runs of 568-second duration were performed (284 vol/run) during which capsaicin or saline was delivered, blinded, four times each to the subject (2 x 4 stimulus blocks), for 24 seconds on each occasion, in a pseudorandom order (Figure 1). Visual cues were used to time the subjects' initial tidal inhalations with the onset of each challenge (breathing was then self-paced for the duration of the challenge). At the end of each challenge, subjects estimated their urge-to-cough using a 5-point scale (using their right-hand fingers and thumb), which ranged from "no urge" (rated 0) and "actually coughed" (rated 5). The capsaicin concentration was increased midexperiment in one subject because she completely habituated to the challenge (i.e., they experienced no urge-to-cough with their predetermined capsaicin concentration). At the conclusion of the experiment, subjects were asked to list any additional sensations that they experienced during the capsaicin challenges.
Data Analysis Sex effects on the cough thresholds and urge ratings were examined with independent t tests. Repeated-measures analysis of variance was used to establish any effects of order on the ratings of the urge-to-cough. Functional brain images were analyzed with the Expert Analysis Tool (www.fmrib.ox.ac.uk/fsl/) of the Centre for Functional Magnetic Resonance Imaging of the Brain (FMRIB) (Oxford, UK) (12). Preprocessing included brain segmentation, motion correction, high-pass filtering, and spatial smoothing with a Gaussian kernel of 6 mm3 full width at half maximum. Statistical analysis was performed with the linear model of FMRIB (13). Regressors convolved with the hemodynamic response function were generated to model each of the events incorporated in the experimental protocol, including the ready cue, capsaicin inhalation, saline inhalation, and ratings of the urge-to-cough. Evaluation of the motion correction parameters revealed an unacceptable level of head movement (> 3 mm) in one subject (no. 9). Data from subject 9 were excluded from group analyses. Motion correction parameters were incorporated as orthogonalized regressors in the analyses of the remaining nine subjects. Parametric maps were coregistered onto the subjects' high-resolution anatomical images, which were then registered to a standard brain (Montreal Neurological Institute 152 brain) (14) to permit group analysis. A mixed-effects analysis was used to identify capsaicin activation, incorporating variance within session and across time (fixed effects) and cross-session variance (random effects). Random-effects analysis was used to test for sex differences in capsaicin activation. Cluster thresholding was performed with a z-threshold of 2.3 and corrected p value of less than 0.01 (15). Regions of interest (ROIs) were generated for clusters of capsaicin activation. Mean capsaicin-related signal changes from ROIs were correlated with mean urge ratings for the 12 capsaicin events using a Pearson's r coefficient and a Bonferroni-corrected significance level of p < 0.009, having taken account of the mean correlation between measures of r = 0.47 (16).
Cough Thresholds and Capsaicin-evoked Urge-to-Cough Ratings As has been previously reported by others (17), capsaicin-evoked cough thresholds were on average lower in the female subjects compared with the males, albeit this was largely driven by the high thresholds displayed by subjects 1 and 2 (Table 1). Accordingly, the average capsaicin concentration used to evoke the urge-to-cough during the magnetic resonance imaging acquisitions was also lower for the females (7.03 ± 3.51 and 1.08 ± 0.24 for males and females, respectively; p < 0.05). Nevertheless, the mean urge-to-cough rating across all capsaicin challenges during the functional runs was not different between the male and female subjects (2.3 ± 0.4 and 2.5 ± 0.3 for males and females, respectively; t[8] = 0.3, p = not significant). Saline inhalation during the scanning sessions failed to evoke an urge-to-cough in any of the subjects tested (mean rating score, 0 ± 0; 12 saline challenges per subject). Conversely, capsaicin inhalation reliably produced an urge-to-cough (perceived as laryngeal irritation) in all subjects tested. The temporal nature of this urge was striking, developing and persisting during the course of a 24-second challenge but rapidly dissipating after the first exhalation after cessation of the challenge. Although most subjects also reported other sensations in addition to urge-to-cough (Table 1), there was no consistent additional sensation that could account for the group-evoked brain activations described below. For example, one subject described burning of the lips that began toward the end of a functional run and persisted continuously for the remainder of the experiment (i.e., the burning was present during saline challenges and in between challenges). Other capsaicin-evoked events included persistent running nose (one subject), swallowing at higher urge intensities (one subject), inconsistent tearing of the eyes that was secondary to the urge-to-cough and not due to capsaicin vapor entering the eyes (three subjects), occasional detection of a discernable odor (two subjects), and a difference in sound between challenges (one subject). No subjects reported any sensations from the face, tongue, or oral cavity. Capsaicin inhalation was also associated with a modest (statistically insignificant, p > 0.05) reduction in respiratory rate during the first functional run, but not during the subsequent runs (Figure 2).
Although all subjects reported a capsaicin-evoked urge-to-cough during the scanning runs, the magnitude of this urge was variable between challenges in any given subject. Analysis of the urge rating scores across all challenges and the three functional runs revealed within-run (F[1,9] = 8.1, p < 0.05) and between-run (F[1,9] = 5.8, p < 0.05) habituation of the capsaicin-evoked urge ratings. Indeed, in one subject, complete tachyphylaxis occurred to her predetermined individualized dose of capsaicin, necessitating an increase in the capsaicin challenge dose mid–scanning session to ensure adequate capsaicin-evoked urge-to-cough. There were no instances of any subject rating a capsaicin challenge as 5 (i.e., no subjects actually coughed in the scanner in response to capsaicin).
Functional Brain Imaging
Correlation between the Magnitude of Regional Brain Activations and the Urge-to-Cough Scores Given that each capsaicin challenge produced variable subjective reports of each subject's urge-to-cough experience (score ratings ranging from 1 to 4), we compared the magnitude of the activations in each brain region with the subjective rating provided by the subject (Table 2). A significant correlation between the magnitude of the signal change and the magnitude of the urge rating was found in the anterior midcingulate cortex (BA 24/32, r2 = 0.59), the right primary somatosensory cortex (BA 43, r2 = 0.59), and the supplementary motor area (SMA; BA 6, r2 = 0.69). There was a moderate correlation between urge ratings and signal changes in the left middle frontal gyrus and left cerebellum. All other brain regions that showed capsaicin-evoked activations were poorly correlated with the subject's urge score.
Circumstantial evidence suggests that cough in humans involves higher brain regulation over the basic reflex-evoked change in respiratory motor pattern that occurs at the brainstem level. In the present study, we show that capsaicin inhalation in healthy humans, at doses insufficient to cause coughing, evokes a transient urge-to-cough that is associated with activation of a variety of brain regions, including the primary sensory and motor cortex, insula, orbitofrontal cortex and discrete regions of the anterior cingulate cortex, and cerebellum. In some of these regions, the magnitude of the activation was correlated with the intensity of the perceived urge. These data provide the first insights into the supramedullary neuronal network that is likely involved in the conscious perception of airway irritation and/or the regulation of coughing in humans.
Evidence for Cortical Involvement in Cough
Activated Brain Regions and Possible Implications A dense cluster of activity was concentrated bilaterally in the primary sensory cortex, at the caudal limits of the central sulcus (near the junction with the lateral sulcus). This region likely corresponds to BA 43, and was shown to receive sensory inputs from the throat in the classic studies by Penfield and Boldrey (28). We also noted activity in the orbitofrontal cortex, which receives extensive sensory inputs from somatic and visceral sources and may integrate visceral sensory and motor information to modulate behavior (29). Capsaicin-evoked activations were concentrated in the lateral regions of the orbitofrontal cortex, which may be involved in the evaluation of the experiences of unpleasantness (30). The insula cortex is also believed to play a prominent role in visceral sensory processing, and a variety of cardiorespiratory and gastrointestinal afferent nerves provide input to insula neurons (31, 32). Activations in the insula cortex have been reported for pain (22–24), thirst (25), bladder distension (26), dyspnea (27), esophageal distension (33), and itch (34). We noted bilateral activations that were mostly confined to the anterior insula. This region is commonly activated in studies of both somatic and visceral pain (22–24), and may be a key site for the perception of dyspnea in humans (27). It was interesting that the magnitude of the BOLD signal change in most of the brain regions activated during capsaicin inhalation showed little or no correlation with the urge-to-cough rating score provided by the subject. This was a surprising observation as we expected that the level of sensory input to the brain dictates the level of urge. However, in addition to S1 in the right hemisphere, the anterior midcingulate cortex and the SMA did show urge-dependent activation patterns. A previous study in cats reported that coughing evoked by stimulation of the superior laryngeal nerve could be suppressed with concurrent electrical stimulation of the anterior cingulate gyrus (10), suggesting that this region is involved in higher level inhibition of coughing. The SMA has also been shown to play a prominent role in sensory suppression (35). The anterior cingulate cortex and the SMA may represent components of an inhibitory control mechanism of the cough reflex and perhaps the correlation between the BOLD signal magnitude and the urge-to-cough rating for these regions reflects a greater need to actively suppress cough as urge intensity increases. This hypothesis is supported by other functional imaging studies in which activity in the anterior cingulate and/or the SMA shows a high correlation with sensory-driven urges when subjects are instructed to consciously inhibit the impending motor event. For example, the urge to void when the bladder is filled (36), the urge to breathe when patients are on mechanical ventilation at low tidal volumes (27), the urge to scratch after cutaneous injections of histamine or antigen (34), and the urge to drink in alcoholics (37) are all associated with urge-dependent increases in anterior cingulate/SMA activity. Fentanyl analgesia during painful thermal stimuli also increases activity in the SMA and this is associated with a decreased pain perception (38). Nevertheless, additional studies will be necessary to confirm an inhibitory role of the anterior cingulate and/or SMA on the cough reflex. Some of the brain regions activated in response to capsaicin inhalation may not contribute to generating the urge-to-cough. Capsaicin was delivered via a face mask (oral breathing) and may be in contact with sensory nerves associated with the external nares, lips, tongue, and oral cavity. Most subjects reported additional sensations, which included persistent burning of the nose, sporadic detection of an odor, or sound. However there was no sensation (other than urge-to-cough) that was commonly experienced by a significant percentage of the subjects, making it unlikely other sensory nerve populations contributed significantly to the group BOLD signal changes. Capsaicin-evoked changes in motor activities could also contribute to the brain regions activated in the present study. Indeed, we noted activity in the primary motor cortex and cerebellum after capsaicin inhalation and some subjects reported tearing of the eyes (a motor response to strong urge-to-cough intensities) and frequent swallowing during capsaicin challenges. In addition, there was a modest (but insignificant) reduction in breathing frequency in most subjects during the initial (but not the latter) capsaicin challenges. Again, these motor events were not consistent across the group, and given that many of the brain regions activated after capsaicin challenge have also been implicated in the processing of other sensory modalities (21–27), it seems reasonable to speculate that much of the overall pattern of brain activity observed represents elements involved in perceiving and responding to airway irritation. In conclusion, the results of the present study suggest that a complex network of cortical and subcortical brain regions are activated in association with capsaicin inhalation and the urge-to-cough. The precise role of individual regions in sensing airway irritation, generating urge, or in the behavioral modification of cough is speculative. Nevertheless, the pattern of activation is consistent with what has been reported in humans after other types of sensory stimuli and therefore supports the suggestion that a common cortical network is likely involved in shaping responses to a variety of cutaneous and visceral stimuli.
The authors thank Dr. Michael Ditchfield and Mr. Michael Kean of the Children's MRI Centre (Melbourne, Australia) for the assistance they provided during the execution of this study.
Supported by National Health and Medical Research Council of Australia (nos. 350333, 454776 [S.B.M.], and 400317 [G.F.E.]), the Robert J. Jr. and Helen C. Kleberg Foundation, and the G. Harold and Leila Y. Mathers Charitable Foundation (M.J.F., G.F.E.). Originally Published in Press as DOI: 10.1164/rccm.200612-1856OC on June 15, 2007 Conflict of Interest Statement: None of the authors has a financial relationship with a commercial entity that has an interest in the subject of this manuscript. Received in original form December 21, 2006; accepted in final form May 29, 2007
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
TOP
ABSTRACT
AT A GLANCE COMMENTARY
