Breathlessness debilitates millions of people with lung disease, heart failure and cancer. In one South Australian household survey, nearly one in ten people self-reported limiting chronic breathlessness, mostly attributed to lung or cardiac disease . Symptoms correlate poorly with the objective measures of disease (e.g. spirometry) , in part because of altered brain processing of respiratory sensations, leading to a disproportionately increased perception of breathlessness [3▪▪]. Breathlessness may be refractory, that is, persistent despite the optimal management of underlying disease [4,5].
Research into the possible mechanisms for the perception of breathlessness has largely concentrated on the peripheral pathways and sensory afferent sources of respiratory sensation. These are summarized in the American Thoracic Society position statement , in which the growing interest in the role of central nervous system (CNS) mechanisms is also highlighted.
In humans, such mechanisms are best investigated using noninvasive imaging techniques, preferably those which do not use radiation. Although most neuroimaging studies to date have investigated induced acute breathlessness, there are limitations to their interpretation in the investigation of chronic refractory breathlessness, in which it should not be assumed that mechanisms and responses to breathlessness or interventions are the same. Advances in neuroimaging methodology and analysis may enable a more detailed understanding of the mechanisms of this challenging symptom.
WHAT IS FUNCTIONAL MRI, HOW DOES IT WORK AND WHAT ARE THE LIMITATIONS OF THIS APPROACH?
Functional MRI (FMRI) allows the identification of location, pattern and time course of brain activity, in vivo. Contrast agents, radiation and radioactive tracers are not required. Brain activation causes localized increases in metabolic activity, resulting in increased cerebral blood flow, cerebral blood volume and oxygen saturations (Figs 1 and 2), thus decreasing localized deoxyhaemoglobin concentration. As deoxyhaemoglobin is more disruptive to a magnetic field than oxyhaemoglobin, there is a localized increase in magnetic resonance signal in the region of neural activation known as the blood oxygenation level dependent (BOLD) response.
FMRI studies are typically conducted in groups of 12–20 participants and involve repeatedly acquiring MRI scans optimized for FMRI, alternating ‘on’ periods of stimulation followed by ‘off’ periods of rest or a control (e.g. Figure 2). Such an approach has often been used in the FMRI studies of breathlessness, in which 15–30 s of stimulation are followed by rest periods. As a result of the signal drift, stimulus block durations of longer than 1 min can become ineffective.
FMRI analysis uses multiple linear regression to detect correlations in location, pattern and time course between the changes in a task or stimulus of interest and changes in BOLD response in each image element (voxel) of the brain. Usually, FMRI studies are presented as the results of findings in a group of participants merging their brain scans by morphing each individual's scan to a common template.
Neurovascular coupling is the cascade of events that increases capillary blood flow in concert with neuronal activity, and its potential for alteration must be considered when interpreting FMRI data. This is particularly relevant with breathlessness research and the following issues affecting the FMRI measurements should either be controlled or accounted for in the statistical analysis:
- Alterations in arterial carbon dioxide, pH, arterial oxygen tension and intrathoracic pressure have profound effects on vascular tone [8,9].
- Drugs [10,11] or disease states  may directly affect chemical signalling mechanisms to the cerebral vessels, vascular reactivity, cerebral metabolic rate or physiological fluctuations.
- Correction for physiological noise [13,14] (e.g. respiratory and cardiac motion) and head movement is required, especially in the studies of breathlessness in which stimulus-correlated artefact can be problematic.
Strategies to account for the above include meticulous control of end-tidal gases  or inclusion of control tasks . Combination of FMRI with electrophysiological measurement can provide a more direct signal of the neuronal activity, allowing the investigation of altered neurovascular coupling .
Misinterpretation (or overinterpretation) because of misleading statistical analysis
As FMRI datasets comprise the time course of hundreds of thousands of individual voxels each correlated with the stimulus of interest, appropriate statistical correction for multiple comparisons is very important. Cluster-based thresholding is commonly used, which takes into account correlated signal change in neighbouring voxels when determining significance. Appropriate correction for multiple comparisons is essential [7▪▪], and very small uncorrected P-values may appear convincing (e.g. P < 0.0001). However, because of the number of comparisons, such approaches are usually insufficiently robust and may confuse an unwary reader. Analyses restricted to defined brain areas (e.g. ‘region of interest’ or ‘small volume correction’ approaches) should be clearly explained to prevent misinterpretation as ‘identification of the brain area involved with a particular function’; such techniques restrict the statistical analysis to limited areas of the brain, thus require a priori hypotheses about that area's function. Another statistical technique requiring care is indirect comparisons. For example, if breathlessness is induced in two populations (e.g. men and women) and a particular brain area is seen to be active in women but not men, formal statistical comparison should inform conclusions rather than visual comparison of the two group maps [7▪▪].
This is challenging for the study of breathlessness. Although FMRI is best suited to stimuli of short duration, (e.g. <1 min), longer stimulus durations are usually required for a breathing challenge to be fully effective (unlike a painful stimulus elicited by a laser), and recovery may take several minutes or longer, especially in patients with chronic breathlessness.
Lying flat in a tunnel scanner may be challenging. With careful selection of people with lung and heart disease, FMRI should be possible [17,18]. However, it is debateable whether those with more severe breathlessness, particularly in those with high anxiety levels, would tolerate this procedure easily.
Although the most common trigger for disease-related breathlessness is exercise, head motion associated with exercise would render MRI images uninterpretable; therefore, to date, breathlessness has been induced ‘artificially’ either by manipulating blood gases, resistive load or by breathholding to produce ‘urge to breathe’ in healthy volunteers. In the future, upright MRI scanners may be suitable for FMRI in patients who suffer claustrophobia or orthopnoea, but current maximum field strengths are inadequate for FMRI. A more detailed explanation of FMRI and its limitations may be found elsewhere [16,19].
WHAT IS MAGNETOENCEPHALOGRAPHY, HOW DOES IT WORK AND WHAT ARE THE LIMITATIONS OF THIS APPROACH?
Magnetoencephalography (MEG) measures brain activity by detecting tiny magnetic fields generated by the electrical currents associated with neuronal firing [20▪▪,21]. These magnetic field changes are detected by superconducting quantum interference devices (SQUIDs) within a MEG headpiece (Fig. 3).
MEG can directly measure brain activity at millisecond resolution, which contrasts to the indirect, secondary blood-flow derived slower resolution (seconds) measures obtained with FMRI. It can detect and identify intrinsic oscillatory activity, such as alpha, beta and gamma rhythms. This type of temporal information can be used to investigate how activity within different nodes of a brain network relates to each other in both task-related experiments and in the resting state.
The ability to measure millisecond resolution brain activity gives MEG the potential advantage of directly measuring neuronal activity rather than the consequential changes in cerebral blood flow in investigations that may alter neurovascular coupling (see above), for example, opioids and hypercapnia. Signal drift is not a problem as it is with FMRI. The main advantages of MEG in the study of breathlessness are that it:
- can be used to test hypotheses relating to how brain areas interact with each other on a sub-millisecond scale;
- is less demanding than FMRI and may be tolerated by patients, even with severe breathlessness;
- there is potential to explore seated exercise to induce breathlessness, although head motion would need to be carefully controlled.
Although the statistical maps of neural electrical activity require a structural MRI scan (for registration of brain activity to a common template), this takes only a few minutes.
Environmental magnetic noise is a major consideration with MEG, hence the need for specialist shielding around MEG scanners. Magnetic interference can also be generated by the participants themselves; metal implants, skeletal muscle activity (e.g. eye movements) and electrocardiac activity. As with FMRI, head movement degrades image quality. Source localization with MEG is most efficient with clear a priori hypotheses about the location of brain activity to include in data analysis. Therefore, MEG may be less suited to the exploratory brain mapping approaches of early FMRI studies but much better at exploring how activity in different brain areas relate to each other.
HOW HAS NEUROIMAGING HELPED UNDERSTAND BREATHLESSNESS SO FAR?
The neuroimaging literature directly investigating breathlessness comprises the results of seven different studies (Table 1) . Most studies induced acute breathlessness in healthy volunteers, apart from one study of symptom perception in mild asthmatic patients . The studies are heterogeneous with different stimuli (resistive loading for varying durations, mechanical ventilation or hypercapnia); imaging methods (PET and FMRI at 2 and 3 T); statistical approaches; numbers of participants (n = 6–14) and approach to collection of behavioural data. Therefore, it is uncertain whether differences in the study findings represent biological or methodological differences. The small number of participants means it is likely that only the strongest effects would have been detected. It remains unclear how the neurophysiology of experimental breathlessness compares with that experienced by patients.
In spite of these considerations, the insular cortex, anterior cingulate cortex and the amygdala are the most commonly activated in response to experimental breathlessness. Activation in these brain areas is commonly observed in the studies of pain  (described by some as ‘affective pain areas’) supporting the popular comparison of pain with breathlessness. However, these brain areas subserve multiple different functions, for example, anxiety (relevant to breathlessness) and general bodily awareness (interoception). Therefore, although caution is advised when interpreting the findings in breathlessness based on similar activation patterns seen in different conditions (i.e. accepting reverse inferences).
The multidimensional model of breathlessness proposed by Lansing et al. describes breathlessness as having physical and emotional components comprising an immediate (unpleasantness) component and a longer term emotional impact. This model forms the basis for the recently developed Multidimensional Dyspnoea Profile , an assessment tool designed to behaviourally unpick separate components of clinical breathlessness.
The current neuroimaging literature supports this model of a primary sensory and a primary affective component. For example, Von Leupold et al. manipulated breathlessness unpleasantness with emotional picture viewing. They examined brain activity in the amygdala and anterior insula. Activity in these brain areas was enhanced in the more unpleasant state. Future studies would take this work forward by dissociating the brain areas responsible for the physical intensity of breathlessness from its unpleasantness, possibly using a whole-brain approach to discover which other areas in the brain contribute to breathlessness unpleasantness. Another future aim would be to determine how the amygdala and anterior insula interact with the rest of the brain's respiratory control network. Although methodologically challenging, disentangling unpleasantness and intensity is an important goal for FMRI research. Breathlessness unpleasantness relies more on the CNS processing and is an obvious target for CNS-focused treatments, whereas changing the intensity of breathlessness might rely more directly on the peripheral input and might, therefore, be expected to be changed more by treatments to the lungs or muscular system. There are no FMRI studies that have examined the neural underpinnings of longer term emotional impact on breathlessness processing (e.g. conditioned fear ). Such studies might better be performed in a clinical population, although fear conditioning paradigms in healthy volunteers may yield useful insights .
OPIOIDS AND BREATHLESSNESS
There is increasing clinical interest in the use of opioids for the treatment of refractory breathlessness [38▪▪,39], although their adverse effect profile and safety concerns necessitate their careful use [40▪▪,41]. Clinically, acutely administered opioids may reduce the physical intensity of breathlessness (by reducing the work of breathing through brainstem respiratory centre depression ) or by reducing the immediate unpleasantness of breathlessness. As there are few neuroimaging studies of opioid mechanisms on breathlessness and respiratory control, the following paragraph discusses the studies of breathlessness alongside appropriate studies of opioid action in pain to suggest potential avenues for future research.
Beta-endorphin is an endogenous ligand at the mu opioid receptor released during exercise . Using naloxone to block opioid receptors, studies performed in healthy volunteers  and in chronic obstructive pulmonary disease patients  demonstrate that endorphins dampen effort perception and breathlessness during exercise. Analogous to the findings in chronic pain [46,47], these studies suggest that endogenous opioids play an important role in breathlessness perception.
Behavioural studies in pain demonstrate that low doses of acutely administered opioids may have more effect on the affective component of pain than its sensory intensity . In a PET study, experimental heat pain increased endogenous opioidergic activation in areas of the ‘affective’ pain system (anterior cingulate and anterior insula) . In an FMRI study of the opioid alfentanil during painful stimulation, differential effects on ‘sensory’ (somatosensory cortices and posterior insula) and ‘affective’ (parahippocampal gyrus, amygdala and anterior insula) brain regions  in the processing of pain were demonstrated. Finally, an FMRI study  showed that the opioid remifentanil dampened brain responses to breath holding in the anterior cingulate, prefrontal and insular cortices. The study demonstrated that opioid action on respiratory control extends beyond the brainstem [14,51], and opioids dampen similar (affective) brain areas observed in the above pain studies in the context of respiration. Although these neuroimaging studies should be interpreted with care because of lack of direct behavioural correlation with imaging findings, taken together, converging lines of evidence suggest opioid action on brain processing of the emotional (unpleasant) aspect of breathlessness. Future studies focused on combining the behavioural and neuroimaging evidence will give a clearer understanding of how opioidergic mechanisms play a role in the genesis of breathlessness. Unanswered questions include whether it is possible to dissociate the neurophysiological processes underlying opioid effects on respiratory drive from their effects on perception of unpleasantness, and to what extent the breathlessness relieving effects of opioids are because of each of these effects. Selective antagonism of the respiratory depressive effects of opioids  whilst maintaining the reduction of unpleasantness perception remains an important goal and lead to a safer approach to use opioids in the clinical setting. Finally, a better understanding of the opioidergic brain pathways may lead to trials of other drugs already known to act on those pathways.
Despite these advances in understanding, challenges remain in the investigation of breathlessness (see Table 2). Although early studies using FMRI and PET were limited by technical challenges, many of these have either been overcome or are now much better understood. Hypotheses generated from FMRI can be tested at much higher temporal resolution with MEG, which can investigate brain network activity at millisecond resolution and as consequence of the seated position, tolerated by more severely affected patients.
A greater understanding of opioidergic mechanisms may help the development of novel interventions which reduce perception of breathlessness without clinically significant respiratory depression. In addition, testing of other interventions such as facial airflow and pulmonary rehabilitation in people with chronic refractory breathlessness may help identify distinct central processes involved with perception of and emotional response to breathlessness to directly inform use in clinical practice.
Neuroimaging of breathlessness remains in its infancy. However, advances in the understanding of central perception, combined with novel neuroimaging techniques, means that we are poised to increase our understanding of the brain processes of breathlessness and their modulation. This knowledge will inform use of clinical interventions in everyday practice.
The authors thank Anja Hayen for her critical appraisal of a previous version of this article and Michael Simpson for his critical review of the MEG section. K.P. is supported by the National Institute for Health Research Oxford Biomedical Research Centre based at Oxford University Hospitals NHS Trust and University of Oxford.
Conflicts of interest
There are no conflicts of interest.
REFERENCES AND RECOMMENDED READING
Papers of particular interest, published within the annual period of review, have been highlighted as:
- ▪ of special interest
- ▪▪ of outstanding interest
1. Johnson MJ, Bowden JA, Abernethy AP, Currow DC. To what causes do people attribute their chronic breathlessness? A population survey. J Palliat Med 2012; 15:744–750.
2. Jones PW. Health status measurement in chronic obstructive pulmonary disease. Thorax 2001; 56:880–887.
3▪▪. Hayen A, Herigstad M, Pattinson KT. Understanding dyspnea as a complex individual experience. Maturitas 2013; 76:45–50.
This article explains how interindividual differences in the psychological makeup may impact the cognitive processes underlying breathlessness.
4. Pratter MR, Abouzgheib W, Akers S, et al. An algorithmic approach to chronic dyspnea. Respir Med 2011; 105:1014–1021.
5. Booth S, Bausewein C, Higginson I, Moosavi SH. Pharmacological treatment of refractory breathlessness. Expert Rev Respir Med 2009; 3:21–36.
6. Parshall MB, Schwartzstein RM, Adams L, et al. An official American Thoracic Society statement: update on the mechanisms, assessment, and management of dyspnea. Am J Respir Crit Care Med 2012; 185:435–452.
7▪▪. Pattinson KT. Functional magnetic resonance imaging in anaesthesia research. Br J Anaesth 2013; 111:872–876.
This article gives a basic introduction into FMRI and its limitations, aimed at a clinical audience.
8. Hayen A, Herigstad M, Kelly M, et al. The effects of altered intrathoracic pressure on resting cerebral blood flow and its response to visual stimulation. NeuroImage 2012; 66C:479–488.
9. Wise RG, Pattinson KT, Bulte DP, et al. Dynamic forcing of end-tidal carbon dioxide and oxygen applied to functional magnetic resonance imaging. J Cereb Blood Flow Metab 2007; 27:1521–1532.
10. Masamoto K, Kanno I. Anesthesia and the quantitative evaluation of neurovascular coupling. J Cereb Blood Flow Metab 2012; 32:1233–1247.
11. Pattinson KT, Rogers R, Mayhew SD, et al. Pharmacological FMRI: measuring opioid effects on the BOLD response to hypercapnia. J Cereb Blood Flow Metab 2007; 27:414–423.
12. Mogi M, Horiuchi M. Neurovascular coupling in cognitive impairment associated with diabetes mellitus. Circ J 2011; 75:1042–1048.
13. Brooks JC, Faull OK, Pattinson KT, Jenkinson M. Physiological noise in brainstem FMRI. Front Hum Neurosci 2013; 7:623.
14. Harvey AK, Pattinson KT, Brooks JC, et al. Brainstem functional magnetic resonance imaging: disentangling signal from physiological noise. J Magn Reson Imaging 2008; 28:1337–1344.
15. Pattinson KT, Governo RJ, MacIntosh BJ, et al. Opioids depress cortical centers responsible for the volitional control of respiration. J Neurosci 2009; 29:8177–8186.
16. Diukova A, Ware J, Smith JE, et al. Separating neural and vascular effects of caffeine using simultaneous EEG–FMRI: differential effects of caffeine on cognitive and sensorimotor brain responses. NeuroImage 2012; 62:239–249.
17. Hogg JC, McDonough JE, Suzuki M. Small airway obstruction in COPD: new insights based on micro-CT imaging and MRI imaging. Chest 2013; 143:1436–1443.
18. White JA, Patel MR. The role of cardiovascular MRI in heart failure and the cardiomyopathies. Cardiol Clin 2007; 25:71–95.vi.
19. Gore JC. Principles and practice of functional MRI of the human brain. J Clin Invest 2003; 112:4–9.
20▪▪. Proudfoot M, Woolrich MW, Nobre AC, Turner MR. Magnetoencephalography. Pract Neurol 2014; 0:1–8. doi:10.1136/practneurol-2013-000768.
This article gives a basic introduction into MEG and its limitations, aimed at a clinical audience.
21. Hari R, Parkkonen L, Nangini C. The brain in time: insights from neuromagnetic recordings. Ann N Y Acad Sci 2010; 1191:89–109.
22. Liotti M, Brannan S, Egan G, et al. Brain responses associated with consciousness of breathlessness (air hunger). Proc Natl Acad Sci USA 2001; 98:2035–2040.
23. Brannan S, Liotti M, Egan G, et al. Neuroimaging
of cerebral activations and deactivations associated with hypercapnia and hunger for air. Proc Natl Acad Sci USA 2001; 98:2029–2034.
24. Parsons LM, Egan G, Liotti M, et al. Neuroimaging
evidence implicating cerebellum in the experience of hypercapnia and hunger for air. Proc Natl Acad Sci USA 2001; 98:2041–2046.
25. Banzett RB, Mulnier HE, Murphy K, et al. Breathlessness in humans activates insular cortex. Neuroreport 2000; 11:2117–2120.
26. Peiffer C, Poline JB, Thivard L, et al. Neural substrates for the perception of acutely induced dyspnea. Am J Respir Crit Care Med 2001; 163:951–957.
27. Evans KC, Banzett RB, Adams L, et al. BOLD fMRI identifies limbic, paralimbic, and cerebellar activation during air hunger. J Neurophysiol 2002; 88:1500–1511.
28. Peiffer C, Costes N, Herve P, Garcia-Larrea L. Relief of dyspnea involves a characteristic brain activation and a specific quality of sensation. Am J Respir Crit Care Med 2008; 177:440–449.
29. Von Leupoldt A, Sommer T, Kegat S, et al. The unpleasantness of perceived dyspnea is processed in the anterior insula and amygdala. Am J Respir Crit Care Med 2008; 177:1026–1032.
30. Von Leupoldt A, Sommer T, Kegat S, et al. Down-regulation of insular cortex responses to dyspnea and pain in asthma. Am J Respir Crit Care Med 2009; 180:232–238.
31. Von Leupoldt A, Sommer T, Kegat S, et al. Dyspnea and pain share emotion-related brain network. NeuroImage 2009; 48:200–206.
32. Herigstad M, Hayen A, Wiech K, Pattinson KT. Dyspnoea and the brain. Respir Med 2011; 105:809–817.
33. Tracey I, Mantyh PW. The cerebral signature for pain perception and its modulation. Neuron 2007; 55:377–391.
34. Lansing RW, Gracely RH, Banzett RB. The multiple dimensions of dyspnea: review and hypotheses. Respir Physiol Neurobiol 2009; 167:53–60.
35. Meek PM, Banzett R, Parshall MB, et al. Reliability and validity of the multidimensional dyspnea profile. Chest 2012; 141:1546–1553.
36. Janssens T, De Peuter S, Stans L, et al. Dyspnea perception in COPD: association between anxiety, dyspnea-related fear, and dyspnea in a pulmonary rehabilitation program. Chest 2011; 140:618–625.
37. Pappens M, Smets E, Vansteenwegen D, et al. Learning to fear suffocation: a new paradigm for interoceptive fear conditioning. Psychophysiology 2012; 49:821–828.
38▪▪. Johnson MJ, Bland JM, Oxberry SG, et al. Opioids for chronic refractory breathlessness: patient predictors of beneficial response. Eur Respir J 2013; 42:758–766.
This pooled analysis of four clinical trials of opioids for refractory breathlessness did not demonstrate a relationship betwen the underlying condition causing the breathlessness and the response to opioids. This may indicate a final common pathway in the perception of breathlessness targeted by opioids.
39. Johnson MJ, Abernethy AP, Currow DC. Gaps in the evidence base of opioids for refractory breathlessness. A future work plan? J Pain Symptom Manage 2012; 43:614–624.
40▪▪. Baumblatt JA, Wiedeman C, Dunn JR, et al. High-risk use by patients prescribed opioids for pain and its role in overdose deaths. JAMA Intern Med 2014; 174:796–801. doi:10.1001/jamainternmed.2013.12711
This study raises legitimate concerns about the long-term use of opioids for chronic nonmalignant pain which are relevant for the use of opioids for breathlessness in people with stable long-term conditions such as COPD.
41. Society TBPOpioids for persistent pain: good practice. London:The British Pain Society; 2010.
42. Pattinson KT. Opioids and the control of respiration. Br J Anaesth 2008; 100:747–758.
43. Farrell PA, Kjaer M, Bach FW, Galbo H. Beta-endorphin and adrenocorticotropin response to supramaximal treadmill exercise in trained and untrained males. Acta Physiol Scand 1987; 130:619–625.
44. Sgherza AL, Axen K, Fain R, et al. Effect of naloxone on perceived exertion and exercise capacity during maximal cycle ergometry. J Appl Physiol 19852002; 93:2023–2028.
45. Mahler DA, Murray JA, Waterman LA, et al. Endogenous opioids modify dyspnoea during treadmill exercise in patients with COPD. Eur Respir J 2009; 33:771–777.
46. Jones AK, Kitchen ND, Watabe H, et al. Measurement of changes in opioid receptor binding in vivo during trigeminal neuralgic pain using [11C] diprenorphine and positron emission tomography. J Cereb Blood Flow Metab 1999; 19:803–808.
47. Jones AK, Cunningham VJ, Ha-Kawa S, et al. Changes in central opioid receptor binding in relation to inflammation and pain in patients with rheumatoid arthritis. Br J Rheumatol 1994; 33:909–916.
48. Price DD, Von der Gruen A, Miller J, et al. A psychophysical analysis of morphine analgesia. Pain 1985; 22:261–269.
49. Zubieta JK, Smith YR, Bueller JA, et al. Regional mu opioid receptor regulation of sensory and affective dimensions of pain. Science 2001; 293:311–315.
50. Oertel BG, Preibisch C, Wallenhorst T, et al. Differential opioid action on sensory and affective cerebral pain processing. Clin Pharmacol Ther 2008; 83:577–588.
51. Pattinson KT, Mitsis GD, Harvey AK, et al. Determination of the human brainstem respiratory control network and its cortical connections in vivo using functional and structural imaging. NeuroImage 2009; 44:295–305.
52. Dahan A, Aarts L, Smith TW. Incidence, reversal, and prevention of opioid-induced respiratory depression. Anesthesiology 2010; 112:226–238.