Cough in Health and Disease

Cough is an important defense mechanism that clears the airways of secretions and prevents entry of foreign bodies and irritants to the lower respiratory tract. It is a universal experience in health but also a nonspecific presenting feature of most respiratory conditions and a number of nonrespiratory conditions. Acute cough is one of the most common presenting symptoms in the patient population encountered by the general practitioner. In most cases, cough results from viral and bacterial upper respiratory tract infection, is a self-limiting problem, and does not require further evaluation, but a small proportion of patients will have persistent cough that necessitates specialist opinion.

Chronic cough is arbitrarily defined as presence of cough for longer than 8 weeks. It affects 3% to 10% of the general population and is responsible for between 10% and 20% of respiratory outpatient referrals. Chronic cough with significant sputum production (i.e., more than a tablespoonful per day) is likely to be due to intrapulmonary disease such as chronic bronchitis or bronchiectasis. A chronic dry or minimally productive cough may be related to extrapulmonary factors; the cough is likely to be the result of abnormal sensitization of the cough reflex secondary to the effects of local inflammation on sensory nerve endings or an intrinsic abnormality of airway nerves. Most patients complain of an abnormal sensation in the laryngeal area such as a tickle in the throat (laryngeal paresthesia). Chronic cough often is perceived as a trivial annoyance but can be a disabling problem responsible for impairment of quality of life and associated with distressing symptoms such as musculoskeletal chest pains, syncope, incontinence, disturbed sleep, and social embarrassment.

The key to successful management is establishing a clear diagnosis and applying effective treatment for long enough to reset the activity of cough receptors at a more physiologic level. Important pitfalls include atypical presentations, the presence of multiple pathologic conditions, and inadequate therapy of the underlying disorder. Further difficulty arises from the fact that evidence for the efficacy of specific therapies in chronic cough is largely based on expert opinion or uncontrolled trials, as well as the paucity of randomized controlled trials with well-validated outcome measures to guide the clinician. Nevertheless, a systematic approach based on the so-called anatomic diagnostic protocol, which focuses on disease processes within the anatomic distribution of vagal afferent nerves, seems to be successful, and various studies have reported a high rate of treatment success even in tertiary referral populations. The general consensus is that in most cases of chronic cough in patients with no other respiratory symptoms or signs and normal findings on spirometry and chest radiography, the underlying cause is asthma, eosinophilic bronchitis, gastroesophageal reflux, or rhinitis, or a combination of these. Many of these conditions can be recognized clinically, and successful diagnosis and management often are possible without recourse to expensive or invasive investigations.

This chapter focuses primarily on isolated chronic cough, because it is a common condition not dealt with elsewhere in this book. In addition, an isolated chronic cough often is a difficult diagnostic problem for both primary and secondary care physicians. This category of cough is thought to involve a primary abnormality of the cough reflex leading to a heightened response to known tussive stimuli.

The Cough Reflex

Cough is a reflex that occurs when afferent nerve receptors are stimulated by inhaled, aspirated, or endogenous substances. The most sensitive sites for initiating cough are the larynx, the carina, and the points of bronchial branching. Cough receptors also are present in extrapulmonary structures, including the esophagus, diaphragm, and stomach. A broad group of rapidly adapting “irritant” receptors (RARs) found in the larynx and tracheobronchial tree can be stimulated by a wide range of stimuli, including cigarette smoke, ammonia, ether vapor, acid and alkaline solutions, hypotonic and hypertonic saline, and mechanical stimulation by direct contact, mucus, or dust; all such stimuli can provoke cough. Another closely related fiber is the slowly adapting stretch receptor (SAR), which terminates inspiration and initiates expiration when the lungs are at an adequate level of inflation. SARs also may influence cough. C-fiber receptors, which have thin, nonmyelinated vagal afferent fibers, are found in the laryngeal, bronchial, and alveolar walls. They are relatively insensitive to mechanical stimulation and lung inflation but are exquisitely sensitive to chemicals such as bradykinin, capsaicin, prostaglandins, and acid pH. Stimuli that are known to cause cough in human subjects such as capsaicin, bradykinin, and citric acid activate C-fiber afferents, particularly those located in the bronchi. Afferent nerve fibers pass to a central cough receptor in the medulla, triggering a forced expiratory maneuver against a closed glottis, followed by glottal opening and high-velocity expiration (Figure 18-1).

Figure 18-1 Representative scheme of afferent and efferent pathways that regulate cough, and of the pathophysiology of the enhanced cough reflex. Laryngeal and pulmonary receptors, such as rapidly adapting receptors (RARs), C-fibers, and slowly adapting fibers (SARs), and putative cough receptors provide input to the brain stem medullary central cough generator through the intermediary of the relay neurons in the nucleus tractus solitarius (nTS). The central cough generator then decides and coordinates the efferent output to the muscles that causes the cough effort. Also present is an efferent output to airway smooth muscle and mucosal glands (not shown). The cerebral cortex may control the motor output of cough volitionally or influence the “urge-to-cough” sensation. Potential factors involved in the enhancement of the cough reflex both in the upper airways and at the brain stem level are illustrated. Inflammatory responses with expression of mediators, neuropeptides, and vanilloid receptors (e.g., TRPV1) may influence the cough response and brain stem modulation of the afferent pathways. CGRP, calcitonin gene–related peptide; LTD₄, leukotriene D₄; PGE₂, prostaglandin E₂; TNF-α, tumor necrosis factor-α; TRPV1, transient receptor potential cation channel subfamily V member 1 [i.e., vanilloid receptor 1 or capsaicin receptor].

Differential Diagnosis of Cough

The causes of cough can be conveniently divided into acute and chronic (Box 18-1 and Table 18-1). An acute cough is arbitrarily defined as a cough of less than 3 weeks’ duration. Infectious and allergic conditions are by far the most common etiologic disorders. Most acute coughs related to viral upper respiratory tract infection resolve by 3 weeks, but a small proportion become persistent and require further evaluation.

Most pulmonary conditions implicated in causing chronic cough, such as chronic obstructive pulmonary disease, lung cancer, an inhaled foreign body, pulmonary tuberculosis, sarcoidosis, idiopathic pulmonary fibrosis, and heart failure, will be obvious on clinical assessment, spirometry, and chest radiography. Assessment and management of these conditions are dealt with elsewhere in this book. Thus, a majority of patients referred for investigation of chronic cough are nonsmokers with normal findings on physical examination and chest radiography. Most present with a nonproductive or minimally productive cough, and 60% to 75% are female. A recognized tendency is for cough to manifest initially around the time of menopause. The most common conditions implicated in aggravating or causing chronic cough in these patients are listed in Table 18-1.

Table 18-1 Common Conditions Implicated in Causing Chronic Cough

ACE, angiotensin-converting enzyme.

Clinical Assessment

An initial assessment of a patient with chronic cough is directed at finding a specific cause, assessing severity, and initiating trials of treatment. A careful history and physical examination are paramount in the evaluation of a patient with chronic cough (Table 18-2). Details of the factors surrounding the onset of cough and associated symptoms and a careful assessment of the upper airways and the respiratory system are particularly important. Basic initial investigations should include up-to-date chest radiography, spirometry, and tests of bronchodilator reversibility, if appropriate. An abrupt onset of coughing while eating or chewing should raise the possibility of an inhaled foreign body, and the onset of cough shortly after introduction of angiotensin-converting enzyme (ACE) inhibitor therapy suggests ACE inhibitor associated cough. The presence of significant quantities of sputum, hemoptysis, systemic symptoms, prominent breathlessness, wheeze, or abnormal physical signs increases the probability of intrinsic lung disease and should trigger appropriate investigations, which may include a CT scan of the chest and bronchoscopy even in the absence of suggestive findings with more simple investigations. The onset of cough with symptoms suggesting an upper or lower respiratory tract infection raises the possibility of a postinfectious cough; prominent whoops, a very troublesome nocturnal cough, and cough associated with vomiting all are associated with pertussis, a condition that is increasingly recognized in both school-age children and adults. Otherwise, little evidence is available to suggest that information on the timing, nature, complications, and potential aggravating factors is predictive of the underlying cause of the cough.

Table 18-2 Initial Evaluation of the Patient with Chronic Cough

ACE, angiotensin-converting enzyme; CT, computed tomography.

Findings on history and physical examination often are unremarkable, in which case the patient evaluation should focus on the recognition of corticosteroid-responsive conditions (i.e., asthma and eosinophilic bronchitis) and extrapulmonary factors that may be aggravating the cough, such as rhinitis and gastroesophageal reflux. One approach to the assessment of patients with chronic cough is outlined in Figure 18-2. The emphasis is on early recognition of corticosteroid-responsive cough, which can be detected easily in most cases with appropriate investigations or treatment trials. By contrast, the management of patients with nonasthmatic cough can be complex, time-consuming, and expensive, often with disappointing response to specific therapy. Far from clear, however, is whether extrapulmonary factors implicated in the pathogenesis of cough are aggravating a preexisting tendency or represent the underlying cause. Several factors point to the former, including the tendency for nonasthmatic chronic cough to affect middle-aged women and the frequent clinical observation that interventions against potential causes of chronic cough often help, but rarely cure the cough. It is best to have no preconceptions about the underlying causative factors in nonasthmatic cough and to view extrapulmonary factors such as rhinitis and gastroesophageal reflux as potential aggravating factors rather than as causes of the problem. This model has the advantage of providing a basis for the incomplete response to the treatment of these conditions seen in many patients; it also should stimulate research into the cause and treatment of the underlying heightened cough reflex sensitivity.

Figure 18-2 Diagnostic algorithm for investigation of chronic cough. ACE, angiotensin-converting enzyme; EB, eosinophilic bronchitis; GERD, gastroesophageal reflux disease.

A further difficulty in evaluating patients with nonasthmatic chronic cough is the poor correlation between the presence of symptoms or abnormalities on investigation of the potential aggravating factor and the success of treatment directed against that factor. Thus, the diagnosis is secured largely by demonstrating clinical improvement as indicated by decrease in frequency or severity of cough after a suitable trial of treatment. Spontaneous improvement is common, and multiple potential aggravating factors typically are present; these factors add another layer of complexity to the clinical encounter. Results of treatment trials are more easily interpreted when combined with attempts to assess cough severity objectively before and after treatment. Suitable methods for such assessment include use of a simple cough visual analogue (with scores of 0 to 100 mm) (Figure 18-3), cough-specific health-related quality of life questionnaires, and evaluation of cough reflex sensitivity. A 15-mm change in visual analogue score, a 2-point change in Leicester Cough Questionnaire quality of life score, and a 2–doubling dose change in C2 (concentration of capsaicin that causes 2 coughs) can be regarded as evidence of a significant response to treatment. The remainder of this section focuses on the evaluation of the commoner conditions implicated in causing chronic cough.

Figure 18-3 Cough severity assessment tools: A, cough-specific health-related quality of life questionnaires, such as the Leicester Cough Questionnaire; B, cough reflex sensitivity measurement; C, 24-hour ambulatory automated cough frequency monitoring; and D, 100-mm cough visual analogue scale.

Cough Variant Asthma/Eosinophilic Bronchitis

Asthma is a condition characterized by airway hyperresponsiveness and inflammation that manifests with variable symptoms of cough, dyspnea, and wheeze. A subgroup of patients can present with an isolated chronic cough known as cough variant asthma. Heightened cough reflex sensitivity is a common finding in cough variant asthma but not in non–cough-predominant asthma. The airway inflammation in cough variant asthma is essentially similar to that seen in classic asthma. The cough typically is dry or minimally productive; it may occur nocturnally, or after exercise or allergen and occupational exposure, although clinical clues often are lacking. The key to diagnosing asthma is demonstrating variable airflow obstruction. Serial peak flow recordings and spirometry with bronchodilator response are routine first-line investigations, but findings often are normal in cough variant asthma. Demonstration of airway hyperresponsiveness by bronchoprovocation testing is a more sensitive and specific index of variable airflow obstruction, and the hyperresponsiveness may be the only abnormality found. Blood or sputum eosinophilia, a raised exhaled nitric oxide concentration, the presence of atopy, and a positive result on allergen skin prick testing or allergen-specific IgE assay provide supportive evidence for the presence of asthma.

The diagnosis of so-called cough variant asthma usually is confirmed by observation of clinical improvement with therapy. Although originally described as manifesting with a bronchodilator-responsive cough, cough variant asthma is much more commonly associated with a corticosteroid-responsive cough, and inhaled corticosteroids are the mainstay of treatment. A blood eosinophil count above 0.4 × 10⁹/L, a sputum eosinophil count more than 3%, and an exhaled nitric oxide concentration above 50 parts per billion (ppb) at an exhalation flow of 50 L/minute are strongly associated with a positive response to corticosteroid therapy in patients with cough. These tests are arguably more fruitful and clinically informative than tests of airway dysfunction. A 2-week trial of oral corticosteroids may be a useful alternative if these tests are unavailable.

Eosinophilic bronchitis is an increasingly recognized entity that presents with a corticosteroid responsive cough and is characterized by sputum eosinophilia (Figure 18-4), heightened cough reflex sensitivity, but no evidence of variable airflow obstruction or airway hyperresponsiveness. Most patients also have a raised exhaled nitric oxide concentration, and blood eosinophilia often is present. Studies suggest that eosinophilic bronchitis is responsible for 10% to 15% of cases of chronic cough. The airway inflammation is similar to that seen in asthma although available evidence indicates that differences in the pattern of airway dysfunction are due to differences in the site of mast cell localization within the airway, with infiltration of the epithelium occurring in eosinophilic bronchitis and infiltration of the airway wall smooth muscle occurring in asthma. Recognition of eosinophilic bronchitis is important, because like cough variant asthma, it responds well to inhaled corticosteroids. This is best achieved by assessing airway inflammation using induced sputum or exhaled nitric oxide; if these techniques are not available, a trial of corticosteroid therapy is indicated irrespective of the presence of airway hyperresponsiveness.

Figure 18-4 Sputum eosinophilia in eosinophilic bronchitis. Eosinophils are the cells with cytoplasmic granules stained red.

Further Investigations

The use of fiberoptic bronchoscopy and high-resolution CT scanning should be reserved for evaluation of patients suspected of inhaling a foreign body; for those with suggestive symptoms, signs, or chest radiograph findings; and for patients with no objective evidence of more common causes of cough, because the investigations are invasive and expensive, with low diagnostic yield. Cough reflex sensitivity measurement has limited value in the validation of the presence of chronic cough in clinical practice because of the wide overlap of cough sensitivity between health and respiratory disease causing cough. Ambulatory cough monitors have the advantage of providing objective evidence of the presence and frequency of cough, but their routine use is hampered by automation difficulty of analysis of the recordings. Recent advances in recording devices and improved battery life have led to a renewed interest in cough monitor development, and they should be available for clinical use in the near future.

Treatment

Treatment directed at the specific cause of chronic cough is summarized in Table 18-3. With use of the anatomic diagnostic protocol, success rates of up to 95% in the management of chronic cough have been reported. The success rate goes down to 60% to 80% in specialist cough clinics, possibly owing to the complexity of cases referred. Reassessment of the patient after treatment with exclusion of additional aggravating factors or causes forms an integral part of management for chronic cough. A common dilemma faced by physicians managing these patients is that the diagnosis of cough often depends on successful trials of treatment, which if unsuccessful leads to the difficult question of whether the underlying condition has not responded or is not responsible for the cough. In some situations, the use of objective tests to make a diagnosis and careful validation of the effect of therapy for the underlying condition should minimize this problem. Therapeutic interventions for common causes of chronic cough are discussed next.

Table 18-3 Specific Therapy for Chronic Cough

ACE, angiotensin-converting enzyme; GER, gastroesophageal reflux.

Unexplained Chronic Cough

Cough remains unexplained after extensive investigations and treatment trials in up to 40% of patients. These patients are predominantly middle-aged females with objective evidence of airway abnormalities, including increased cough reflex sensitivity and airway inflammation. Organ-specific autoimmune diseases are common, suggesting that the airway abnormalities may have an autoimmune basis.

Unexplained cough is responsible for considerable physical and psychological morbidity. Many patients with unexplained chronic cough are labeled with a diagnosis of psychogenic cough, although little evidence is available to support this view, and it is perhaps more likely that any abnormal illness behavior is secondary to the adverse impact of cough on psychosocial aspects of quality of life. In evaluating a patient with unexplained cough, it is important to recognize common pitfalls in managing chronic cough (Box 18-2). Therapy for idiopathic chronic cough is disappointing, and there is therefore a large unmet need for better antitussive treatment in these patients. Referral to a respiratory physiotherapist or speech therapist for cough management coaching may be of some help.

Conclusion

All pulmonary and many nonpulmonary conditions can manifest with cough. Most will be evident after a simple clinical assessment that includes a careful history, physical examination, plain chest radiography, and spirometry. Cough that remains unexplained after such an assessment is a common reason for referral to a specialist. Potential causes include asthma, eosinophilic bronchitis, rhinitis, and gastroesophageal reflux. Satisfactory outcomes can be achieved in a majority of patients with a management strategy that includes targeted investigations and carefully controlled treatment trials. Complete cure is not always possible, however, particularly in patients with cough thought to be due to extrapulmonary factors, and a significant minority of predominantly middle-aged women have unexplained chronic cough. Whether this reflects failure to identify important causes or represents inadequate treatment of established factors is unclear.