Cough is an inherent protective respiratory tract symptom. Duration of cough has been defined as acute (lasting up to 4 weeks), subacute (up to 8 weeks), and chronic (more than 8 weeks).3 In this document, CC will be called specific, if it is associated with a known cause, or non-specific, if not. Finally, we will examine current knowledge and the available guidelines on refractory CC (RCC), i.e., CC that persists despite treatment targeting known associated conditions.

CC is a very common symptom in clinical practice, and prevalence in the general population ranges from 12% to 3.3%.4,5 It is closely related with tobacco use, and the prevalence of CC in smokers is 3 times that of never-smokers or former smokers.6 A higher prevalence of CC has also been associated with environmental pollution.7

During the initial contact with the CC patient, the physician should determine the general causes of CC and the associated warning symptoms (Table 1). In all CC guidelines currently in use, if the patient has a normal chest X-ray, does not smoke and is not receiving angiotensin-converting enzyme inhibitors (ACEI), the causes associated with CC that will directly impact on the initial treatment are considered.

CC is reported by patients as either a single isolated symptom or one of several. Neurobiological mechanisms in chronic pain and in CC are similar,8 so the patient with chronic pain or CC responds more intensely to a painful stimulus or tussigenic stimulus of a specific magnitude than a healthy individual; this effect has been called hypertussia (similar to hyperalgesia). If the patient receives a stimulus that is not at all painful or tussigenic and responds excessively, the condition is called allotussia (or allodynia, in the case of pain).9 Hypertussia or allotussia are clinical conditions that are now grouped under the heading of “chronic cough hypersensitivity syndrome” (CCHS). The cough circuit is unquestionably complex, and may involve interaction between the different stimuli from the very start8 (Fig. 1). In short, the possible impact of CC on physiology means that the following must be taken into consideration when treating this condition: (a) cough neural pathways themselves may be affected, or (b) an aggravating factor that alters the cough reflex may need to be treated (Fig. 2).

Fig. 1.

Interactions among peripheral stimuli of cough reflex.

(0.17MB).

Fig. 2.

Two possible access pathways of stimuli of chronic cough.

(0.19MB).

The neurological mechanism of cough in humans originates from stimulation of 2 types of neuron terminals that converge in the cough center: unmyelinated C fibers and myelinated Aδ fibers.10,11 Most C fibers respond to a range of irritant stimuli of inflammatory origin, while Aδ fibers respond to mechanical and acid stimuli.12

Excitability of the CNS cough center is increased by 3 mechanisms13: peripheral, central and secondary hypersensitivities (Fig. 3). In the case of central sensitization, paresthesia is often observed in the area of the larynx, as well as hypertussia or allotussia, indicating a neuropathic response.14 In the second mechanism, connections with the cough center via the emotional brain lead to participation of consciousness and the emotional status in the control of cough.

Fig. 3.

Neurological feedback circuit of chronic cough.

(0.36MB).

Adapted from Pacheco.13

Furthermore, it is important to remember that the phenomenon of different converging peripheral stimuli can have a practical application, as suggested by the Australian cough guidelines of 2010, which recommend simultaneous triple therapy, proton pump inhibitors (PPIs), and speech pathology intervention for RCC15 (Strong recommendation/very low evidence). When central hypersensitivity has developed, a mechanism that is much more sensitive to mild peripheral stimuli is established, and this may be the cause of the phenomenon known as visceral hypersensitivity, also called secondary hypersensitivity (Fig. 3). The problem is that central hypersensitivity may decrease clinically, without changes being observed in the CC capsaicin challenge test due to peripheral hypersensitivity.16

Patients’ perception of cough may be expressed differently, depending on the causative pathology. In general, cough accompanying upper airway diseases or laryngopharyngeal reflux (LPR) manifests as irritation in the throat. The “urge to cough” has been the subject of recent research which has located the origin of this impulse within the brain.17,18 Several standardized questionnaires have recently been developed to determine the characteristics of cough. These need further studies before they can be validated,19 and their utility in practice is still limited (Weak recommendation/low evidence). The intensity and frequency of cough have little diagnostic value, and these parameters are of most use when studying the therapeutic effect of antitussive agents. The intensity of cough can be determined with symptom questionnaires or quantified scales, mainly of the visual analog type (Weak recommendation/low evidence). The impact of cough on health-related quality of life is a useful parameter that can be measured objectively. To date, the most widely accepted tool is the Leicester Cough Questionnaire (LCQ).20 The importance of measuring the impact of cough on quality of life is high (Strong recommendation/moderate evidence).

The second aspect to consider is the study of the tussigenic reflex and its sensitivity with objective challenge techniques using inhaled substances. These methods have been described in recent guidelines published by the European Respiratory Society.21 Capsaicin acts directly on specific TRPV1 receptors. New ways of expressing results have been recently proposed, including the complete analysis of dose–response curve, measuring ED50 and Emax.22 Sensitivity and specificity of the capsaicin test in the differential diagnosis of CC and in healthy individuals are low.23,24 It is currently recommended for use in epidemiological studies or for determining the effect of drugs (Strong recommendation/moderate evidence). Inhalation of mannitol in dry powder correlates well with the capsaicin test, but the sensitivity and specificity of the technique are similarly low (Strong recommendation/low evidence). Quantification of the fraction of nitric oxide in exhaled air (FeNO), when high (>30 or 38 parts per billion [ppb], depending on the author), predicts a favorable response to corticosteroid treatment of CC.25 Likewise, the quantification of eosinophils in sputum and peripheral blood can characterize a group of CC patients with eosinophilic inflammation of the airways who are potential responders to corticosteroid treatment26 (Strong recommendation, low quality of evidence).

The physiopathological interpretation of CC has changed recently, and it is now understood as a unitary neurological response to stimuli received from distinct but interactive anatomical origins27,28 (Strong recommendation/moderate evidence) (Fig. 1). Current thinking is that (a) CC is partially or totally resistant to specific treatment in up to 2/3 of patients,29 and (b) most individual carriers of the anatomical-diagnostic triad of conditions do not present CC.

New perspectives in the mechanism of the cough reflex suggest that in these patients, most symptoms are centered in the larynx.30 This has produced the concept of the “laryngeal hypersensitivity syndrome” (LHS) as a fundamental clinical basis for CC, particularly in its refractory form.31

In prolonged or persistent bacterial bronchitis, extended treatment of over 2 weeks with antibiotics targeting the causative pathogen leads to complete resolution of CC32 (Strong recommendation/moderate evidence). It is now agreed that there are 2 asthma phenotypes: eosinophilic and neutrophilic asthma. It was reported recently that 75% of patients with neutrophilic asthma have moderate to severe CC33 (Weak recommendation/low evidence). Studies of eosinophilic airway inflammation measured by FeNO and eosinophils in sputum are highly specific and sensitive and predict good response to corticosteroids34 (Strong recommendation/moderate evidence). The diagnosis of asthma is defined as reversible bronchial obstruction, variable maximum expiratory flow and bronchial hyperreactivity (BHR), as determined by spirometry with bronchodilator testing, monitoring of maximum expiratory flow, or bronchial challenge testing (BCT) (Strong recommendation/high evidence). Asthma can be ruled out by a negative BCT, but if it is positive, the positive predictive value ranges between 60% and 80%35 (Strong recommendation/high evidence). BHR with no evidence of variable flow obstruction associated with CC suggests a diagnosis of cough-equivalent asthma or cough-variant asthma36; in this entity, antileukotrienes appear to be more effective than in conventional asthma37 (Strong recommendation/low evidence). A form of CC called eosinophilic bronchitis (EB), characterized by eosinophils in induced sputum, absence of BHR and good response to corticosteroids, was recently defined: prevalence is 7%–33%.38 Similarities and differences between these entities are summarized in Table 2. With respect to the recently described entity of neutrophilic asthma associated with CC and gastroesophageal reflux (GER), treatment with macrolides has been proposed, but the effect on CC has not been specified39 (Weak recommendation/low evidence). Similarly, treatment with the monoclonal antibody mepolizumab has shown success in eosinophilic asthma, although it failed to resolve the accompanying CC,40 suggesting that the inflammation may be mast cell-mediated.

Laryngeal conditions have usually been considered in the diagnostic triad as chronic upper airway cough syndrome (CUACS), formerly known as post-nasal drip; nevertheless, in view of the high rate of laryngeal symptoms in CC, LHS has recently been associated with laryngeal neuropathy.

There are 5 upper airway conditions that may present with CC: allergic rhinitis, chronic rhinosinusitis in the adult, obstructive sleep apnea, vocal cord dysfunction (VCD), and extra-esophageal manifestation of ERG or LPR.

Allergic rhinitis: this is diagnosed from signs and symptoms of nasal inflammation and the identification of specific allergens in skin prick testing or RAST. This condition should be managed according to the recommendations of the recent guidelines.41Chronic rhinosinusitis: treatment consists of irrigation of the nostrils with saline solution, oral corticosteroids for at least 1 month, and oral antibiotics for up to 3 months, in the case of purulent sinusitis (Strong recommendation/moderate evidence).

Obstructive sleep apnea (see section on “Chronic cough and sleep apnea–hypopnea syndrome”).

Vocal cord dysfunction: VCD is diagnosed from episodic narrowing of the vocal cords during inspiration producing dyspnea on inspiration and CC.42 CC occurs in more than 50% of adults with this condition. VCD is diagnosed by the observation of glottic narrowing on laryngoscopy or by a fall of more than 25% in inspiratory flow during the serum saline challenge maneuver.43 Management consists of the treatment of associated comorbidities such as asthma, rhinosinusitis, GER or the use of ACEI (Weak recommendation/low evidence), and the application of speech pathology intervention43 (Strong recommendation/moderate evidence).

Laryngopharyngeal reflux (LPR): see below.

Gastroesophageal reflux disease (GERD) has been associated with a variety of extra-esophageal manifestations, including CC.44

In both healthy individuals and patients with CC, cough is generally more frequent during the day. In some sleep-related diseases, such as sleep apnea–hypopnea syndrome (SAHS), the situation is different.58 The prevalence of CC in this population can be as high as 33%–39%.59 In our experience,60 42% of patients with SAHS reported CC, of whom 31% had high GER symptom scores. With regard to the effects of CPAP on cough, a significant improvement was found, with resolution of cough in 67%. Similarly, when the prevalence of SAHS was analyzed in patients with CC,61 44% were found to have criteria for SAHS. To summarize, CC is frequently associated with SAHS, particularly in patients with associated GER (Strong recommendation/moderate evidence). CPAP treatment may be indicated for the treatment of CC (Weak recommendation/moderate evidence), although more research is required.

Most patients with CC who present in primary care (PC) can be diagnosed by following the algorithm presented in Fig. 4 (Strong recommendation/moderate evidence). In 1 study, only 31% of the patients had any radiological changes that might guide diagnosis.67 Protocolized empirical and sequential treatment has been shown to be cost-effective.68,69 In a recent study of 112 patients with CC,69 sensitivity and specificity of symptoms for 3 of the diseases most frequently associated with CC were determined, as shown in Table 6.

Fig. 4.

Algorithm for the management of chronic cough in primary care.

(0.22MB).

Algorithms for the management of patients with CC referred from PC to a higher level of medical care are given in Figs. 5 and 6. After ruling out neutrophilic asthma in the CC patient, treatment alternatives are neuromodulators, basically gabapentin (300–1800mg/day) and amitriptyline (10–20mg/day), and speech pathology intervention. The recommendation for both options is strong, while evidence remains weak (Fig. 7).

Fig. 5.

Practical management of patients with chronic cough.

(0.2MB).

Fig. 6.

Algorithm for the management of refractory chronic cough.

(0.13MB).

Fig. 7.

Management of chronic cough in specialized units.

(0.33MB).

The American and Australian-New Zealand guidelines define CC as cough lasting more than 4 weeks,70 while the British guidelines71 set it at more than 8 weeks. According to several epidemiological studies, causes of CC in the child vary depending on age (Strong recommendation/high evidence). In schoolchildren, the most common causes are asthma (27%), cough-equivalent asthma (15.5%) and GER (10%). After adolescence, causes for CT are considered the same as in adults.

It is currently assumed that if partial control of cough can be achieved, this may be sufficient, since it is difficult to eradicate completely, except in some cases (Fig. 5). Unsatisfactory treatment in CC may be due to several causes74 (Table 8). In RCC, treatment possibilities must first be optimized (Fig. 6), and then 2 aspects should be given special attention: (a) LHS, and (b) central hypersensitivity75 (Strong recommendation/moderate evidence). Antitussive treatments should be used for “urge-to-cough”, but designed to avoid modification of the protective reflex cough. The available evidence suggests that nebulized lidocaine is an option for second-line treatment.76

In the second case, central hypersensitivity and treatment in CC, there is growing interest in the use of gabapentin in RCC, with positive responses in around 60% of patients (Strong recommendation/moderate evidence).77 Neuromodulator therapy is being proposed as the future in the treatment of central hypersensitivity.78 Other similar agents that have been studied in small, observational studies include pregabalin, amitriptyline, and baclofen. Amitriptyline or low-dose morphine have shown subjective improvement in the LCQ (Weak recommendation/moderate evidence).79

The efficacy of over-the-counter cough suppressants on CC has not been demonstrated.80 A recent review of dextromethorphan found a modest decrease in cough severity and frequency when compared to placebo.81 Codeine has shown no improvement in capsaicin-induced cough reflex testing.82 Its use in CC that does not respond to non-narcotic antitussives is currently recommended (Weak recommendation/low evidence).

Drugs for reducing peripheral hypersensitivity focusing on TRP receptor blockade have recently been evaluated.83 Inhaled tiotropium has recently been seen in laboratory animals to inhibit TRPV1 receptors.84 Another promising agent in the treatment of CC is AF-219, a receptor antagonist that acts against the purinergic receptor P2X3, and also against peripheral C fibers,85 and is effective in reducing CC frequency.