Chronic persistent cough is a common clinical symptom and sign encountered by physicians. A standardized diagnostic approach has been well established and confirmed in the literature. This approach emphasizes knowledge of the cough reflex arc and particularly examines the cough receptor and afferent pathways. Using this anatomic diagnostic approach, the etiology of cough can be determined and successful therapy initiated in nearly all patients. Familiarity with this approach appears to have a positive clinical outcome associated with it.
A 41 -year-old nonsmoking white male with a history of "asthma" as a child presented with a 6-week history of increased cough productive of yellow sputum starting with an illness manifested by fever, chills, and sweats. He also experienced dyspnea on exertion and chest tightness. At an outpatient center, he was given antibiotics on two separate occasions. Each time his cough and sputum production decreased.
Significant past medical history included "chest" problems starting during the first 2 months of life, "asthma" as a child, lingering colds with coughs each winter for years, infertility because of an absent vas deferens, recurrent bouts of pancreatitis, a negative family history for cystic fibrosis, and two sweat chloride determinations of 40 and 56 mEq/L, respectively. He denied sinusitis, postnasal drip, heartburn, dysphagia, and choking. He was taking no medications.
On physical examination, his blood pressure was 120/80 mm Hg; pulse rate, 80 beats/minute; and respiratory rate, 18 breaths/minute. Head, ears, eyes, nose and throat were normal. Examination of the chest revealed crackles at the left base and a regular rate and rhythm with no murmurs, gallops, or rubs. Abdominal examination showed a cholecystectomy scar while examination of his extremities revealed no clubbing, cyanosis, or edema. A chest radiograph showed nonspecific linear scars in the right upper lobe.
Pulmonary function studies revealed normal lung volumes with forced expiratory volume at 1 second (FEV1) 2.44 L; forced vital capacity (FVC), 3.75 L; and FEVj%/FVC, 65.
Based on this history, physical examination, and spirometry, a working diagnosis of asthma was made. lherapy with albuterol in an extended tablet form, 4 mg twice daily, and beclomethasone (via metered dose inhaler), 4 puffs twice daily, led to a significant decrease in his cough. Repeat spirometry 2 months later showed an FEV. of 3.51 L and an FVC of 5 L. Because the cough did not disappear and remained productive, a sputum culture was done and showed the presence ox Staphylococcus aureus. Because of this finding and a history of an absent vas deferens, a high resolution computed tomography scan was ordered and showed bronchiectasis. DNA analysis showed a positive delta F508 allele, but was negative for other gene mutations. Dicloxacillin was added to his medication regimen with near complete cessation of his cough and sputum production.
This case presents a rather unusual array of historic, physiologic, radiographic, and molecular findings suggesting a diagnosis of cystic fibrosis. It is presented to show just how complex the diagnostic workup of cough can become. The work-up for chronic cough is usually not as complicated.
Chronic cough is a common symptom that nearly all physicians encounter in daily practice. Chronic cough is defined as any cough lasting longer than 3 weeks and is the fifth most common symptom seen by physicians in their outpatient practices. It is responsible for approximately 30 million physician visits per year.1 In the United States, approximately $500 million are spent annually on prescription and over-thecounter antitussives.2
PHYSIOLOGY OF COUGH
Cough is a complex physiologic reflex that normally protects the lower respiratory tract from aspiration or helps rid the lower respiratory tract of both exogenous or endogenous materials. It also can be a pathologic reflex. It may represent the only sign of serious disease, may contribute to the spread of airborne infection, and occasionally may result in functional damage to patients.
The cough reflex has five components:
* cough receptors,
* afferent nerves (vague, trigeminal, glossopharyngeal, and phrenic),
* a cough center located in the medulla and separate from the respiratory center,
* efferent nerves (vague, phrenic, intercostal, lumbar, trigeminal, facial, and hypoglossal), and
* effector organs (diaphragm, intercostal and abdominal muscles, and muscles of the larynx, trachea, and bronchi as well as upper airway and accessory respiratory muscles) (Table 1 ).
Only the cough receptor component of this reflex has been studied extensively, but knowledge of the receptors is still incomplete. Animal and human experiments as well as disease states in humans suggest that cough receptors are located primarily in the larynx, trachea, and large bronchi. There are more receptors in the large proximal airways than in smaller peripheral airways - an observation emphasized by those who do fiberoptic bronchoscopy where coughing is significant as the bronchoscope stimulates the larynx, trachea, and proximal large airways; but when small airways are "tugged at" during transbronchoscopic lung biopsies, there is virtually no cough.
Cough receptors are found circumferentially in all areas, but are increased at the carini. These receptors are particularly suited to be stimulated by inhaled irritants, gases, and airborne particles. Removal of the mucosa of large airways in experimental animals does not abolish the cough reflex, suggesting at least two sets of cough receptors, one at or just below the epithelial surface and another site in deeper structures of the airways. Superficial receptors can be stimulated by substances having direct airway contact while those deeper receptors are more likely to be stimulated by mechanical deformation of the airways. Cough has been associated with diseases of the sinuses, pharynx, diaphragm, pericardium, and pleura, suggesting that there are cough receptors in these areas also.
There appear to be three types of afferent nerve endings in the respiratory tract: 1 ) myelinated slowly adapting pulmonary stretch receptors, 2) myelinated rapidly adapting stretch or irritant receptors, and 3) nonmyelinated C-fiber receptors, both bronchial and pulmonary.3 Slowly adapting pulmonary stretch receptors are not involved directly in initiating cough, but modify the cough reflex by affecting expiratory muscles. The larynx is felt to be an important site for the initiation of both cough and bronchoconstrictive reflexes following inhalation of irritants or particulate matter. Experimental evidence suggests that myelinated irregularly firing irritant receptors are responsible for laryngeal cough. C-fìber receptors also may be important. Capsaicin, an extract of red pepper, causes immediate and severe coughing and has been used in human studies to study the mechanisms of cough. Capsaicin- induced laryngeal cough can be decreased by anesthetizing the larynx with Udocaine. Airway rapidly adapting stretch or irritant receptors, the most important receptors, can be stimulated by tactile stimulation, distention and collapse of airways, chemical irritants, gases, and mediators. Pulmonary and bronchial C-fìbers inhibit the cough reflex.
Anatomy of a Cough Reflex
Afferent input from the tracheal and bronchial receptors is mediated through branches of the vagus nerve while the laryngeal receptors are handled by the superior laryngeal nerve. The pharynx is supplied by the glossopharyngeal nerve and pharyngeal branches of the vagus. The nerves themselves can be stimulated electrically as demonstrated in animal studies. Compression of these nerves by tumor can lead to cough. Transection of the nerve eliminates cough as demonstrated by ipsilateral vagotomy, which can eliminate cough in patients with lung cancer. Similarly, in patients with heart-lung transplant where there is no innervation below the tracheal anastomoses, it has been observed that there is a significant decrease in the ability of cough-provoking solutions to elicit a cough response.4 The cough center is located diffusely in the medulla and is separate from the respiratory center. There are higher cough centers as cough can be voluntarily initiated or suppressed. The efferent loop of the reflex arc includes the vagus nerve, which innervates the larynx and tracheobronchial tree, and the phrenic and spinal motor nerves, which innervate the effector organs.
An effective cough is achieved by high gas flows and velocities through the airways. There are three phases of a normal cough. The first phase is the inhalation phase during which a large volume of air is inhaled. The second compressive phase begins with glottic closure and compression of the gas within the thorax by muscles of expiration working against a closed glottis and a maximally contracted diaphragm. The third expiratory phase begins with a sudden opening of the glottis and expulsion of this trapped intrathoracic air at high velocities. During the expiratory phase, oscillations of tissue and gas cause the characteristic sound of cough.
EVALUATION OF CHRONIC COUGH
A systemic approach to the evaluation of chronic cough was presented by Irwin et al5 in 1981. It was designed as an anatomic diagnostic protocol as it emphasized evaluation of the afferent limb of the reflex arc and showed that the cause of cough could be reliably determined in nearly all patients. Three conditions - postnasal drip syndromes, bronchial asthma, and gastroesophageal reflux disease - accounted for the greatest number of etiologies. A comprehensive history and physical examination emphasizing the anatomic locations of the afferent loop of the reflex arc were found to be extremely important in establishing the cough-associated diagnosis. The history and physical examination, together with bronchial inhalation challenge when appropriate, helped diagnose 86% of cases. Once the diagnosis was established, successful outcome (ie, elimination of the cough or significant reduction of the cough) was obtained in nearly all patients.
Subsequently, three additional studies have confirmed the findings of Irwin's original study. Using a similar anatomic diagnostic protocol, the three studies have confirmed that in nonsmoking patients, three conditions account for the majority of the associated diagnoses in patients with chronic cough.6"8 These include postnasal drip syndromes, cough variant asthma, and gastroesophageal reflux disease (Table 2). A total of 335 prospectively studied patients have been reported. Thirty-three percent had postnasal drip, 26% bronchial asthma, 10% bronchial asthma and postnasal drip, and 11% gastroesophageal reflux disease as the cause of cough.
Common Causes of Chronic Cough
The following aspects of the history and physical examination have been found to be useful: cough following an upper respiratory tract infection or allergy exposure and a history of throat clearing or a sensation of secretions in the posterior nasopharynx (postnasal drip) are typical historical features in patients with postnasal drip syndromes. The finding of a granular appearance of the posterior nasopharyngeal mucosa and mucopurulent secretions adhering to the posterior nasopharynx suggests the diagnosis. Abnormal sinus radiographs or computed tomography scans are also helpful. Asthma is suggested by a history of paroxysmal dry hacking cough aggravated with exercise, cold air, sleep or allergy exposure or irritants. The diagnosis can be confirmed by demonstrating reversible airflow obstruction by spirometry. If spirometry is normal, bronchial inhalation challenge may be necessary. Gastroesophageal reflux disease is suggested by the typical heartburn or acid indigestion history. In elderly patients, symptoms of dysphagia may be the only clue. Documentation by abnormal barium swallow, esophageal manometry, or esophageal pH probing may be necessary to confirm the diagnosis. In most patients, the specific etiology of cough can be found using this approach.9
Pratter et al8 offer an algorithmic approach to the diagnosis of chronic cough. Because asthma and postnasal drip syndromes are so common, some authors have advocated streamlining the work-up by empirically treating patients with a history suggesting postnasal drip with an antihistamine/decongestant preparation for 7 days and patients with a history suggesting bronchial asthma with an inhaled beta-agonist for up to 7 days. If the patient responds to either of these therapies, it is concluded that postnasal drip or asthma is the specific etiology. If the cough does not respond to this specific therapy, then a more comprehensive evaluation should be pursued.
COUGH VARIANT ASTHMA
Cough is a common symptom noted in nearly all patients with asthma in association with wheezing, shortness of breath, and chest tightness. McFadden10 described known asthmatic patients who presented with only cough as a symptom of an exacerbation . Corrao et al11 presented six patients with chronic cough of unknown etiology. All had normal spirograms, but positive bronchial inhalation challenges to methacholine and all responded to bronchodilators, suggesting that these patients had cough variant asthma.
Johnson and Osborn12 reviewed the literature of cough variant asthma in 1991. Fifteen clinically oriented articles were summarized. These included both pediatric and adult patients. The frequency of cough ranged from 39% in children to 43% in adults with cough being the only symptom of asthma. Adults up to 65 years of age have been described, as well as children ages to 3 to 1 7 years. Fifty-three percent of subjects were male in the pediatric group and 46% were male in the adult population. Cough symptoms were described as nocturnal in 72%, exercise-induced in 78%, cold air stimulated in 44%, and upper respiratory infection triggered in nearly 100%. The cough was usually of long-standing duration. Patients with cough variant asthma typically had normal spirometry and peak expiratory flows without change postbronchodilator. Bronchial hyperreactivity could be demonstrated in nearly all by bronchial provocation challenges. The natural history of cough variant asthma revealed that a number of patients eventually developed wheezing and shortness of breath requiring chronic therapy. It appears that more children develop typical asthma symptoms than adults (9% to 75% of children and 50% of adults). The effectiveness of therapy has been based on self-reporting, ie, the results have relied on the patients reporting the effect of therapy on cough. No objective cough counting techniques have been used in these studies.
Most patients have had cough disappear or significantly decrease with bronchodilators. In those patients not responding to beta-agonists and theophylline, systemic steroids work in nearly all patients. Beneficial effects have been seen with cromolyn sodium as well as inhaled corticosteroids. It must be emphasized that a patient with a cough and a positive bronchial inhalation challenge who does not respond to bronchodilators and corticosteroids should be evaluated for another cause of cough as a number of patients may have more than one etiology.7 Why patients with cough variant asthma do not manifest more typical features of asthma is not well understood. Asthmatics do not appear to have heightened cough receptor thresholds.13 Cough receptors in patients with cough variant asthma may be oversensitized.14
An association between cough and bronchoconstriction had been made by Salem and Aviado15 when they postulated that cough receptors were stimulated by local bronchoconstriction, ie, mechanical deformation. They based this postulate on observations that ephedrine and isoproterenol, primary bronchodilators, were also effective antitussive agents. While histamine and methacholine in man causes cough that may parallel the degree of induced bronchoconstriction, cough receptors and irritant receptors appears to be functionally different. Cough and bronchoconstriction can be stimulated independently of each other and while closely related, neither is interdependent. Because of the different receptors, it has been further postulated that bronchodilators may be effective antitussives by "resetting" airway receptors, thereby making them less sensitive to inhaled stimuli.
Airway inflammation has become the hallmark of the current understanding of the pathophysiology of bronchial asthma. In patients with cough variant asthma, bronchoalveolar ravage studies as well as endobronchial biopsies have shown inflammation similar to that found in asthmatics with the more typical symptoms of cough, wheezing, and shortness of breath. These data suggest that the same pathophysiologoc mechanisms are playing a role in cough variant asthma. Recently, Boulet et al16 described a group of adults with chronic cough and no evidence of asthma, but with diagnoses that included postnasal drip syndrome, gastroesophageal reflux disease, both, or none of the above. In each patient, bronchoalveolar ravage and endobronchial biopsies showed a significant amount of inflammation similar to that descabed in asthmatics with the exception of no basement membrane thickening. This study suggested airway inflammation may have a role in the etiology of cough, even in those patients with other well-defined syndromes that do not appear to directly effect the lower respiratory tract. This inflammation may be increasing the sensitivity of the cough receptor.
POSTNASAL DRIP SYNDROMES
There has been a great deal of controversy around whether nasopharyngeal processes can induce cough. Both animal and human experimentation has been equivocal, and while cough receptors are found in large numbers in the larynx, there is often no connection between the postnasal drip syndrome and the larynx directly. The typical postnasal drip syndromes include allergic rhinitis, sinusitis (acute or chronic), vasomotor rhinitis, primary nasal polyposis, chronic rhinitis with non-IgE associated bronchial asthma, endocrine related rhinitis (pregnancy or hypothyroidism), and rhinitis medicamentosa. All of these have typical symptoms of sneezing, cough, nasal congestion, and posterior nasopharyngeal drainage. In addition to a boggy erythematous friable mucosa in the anterior nates, there is often a granular appearance to the mucosa of the posterior nasopharynx or the gross appearance of mucoid secretions. In this clinical setting, the cough may be directly related to these inflammatory processes of the upper airway. These findings are so commonly encountered in patients with chronic cough that a therapeutic trial of an antihistamine-decongestant prior to formal diagnostic work-up has been recommended.8
GASTROESOPHAGEAL REFLUX DISEASE
A number of respiratory symptoms and diseases have been associated with gastroesophageal reflux disease. These include chronic cough, asthma, bronchitis, bronchiectasis, hemoptysis, aspiration syndromes, adult respiratory distress syndrome, atelectasis, and pulmonary fibrosis.17 In many patients, typical symptoms of substernal burning, indigestion, regurgitation of digested material, dysphagia, hoarseness, or choking can be elicited by history. In some patients, overt symptoms may be lacking and therefore gastroesophageal reflux disease as an etiology may be overlooked. In situations in which the diagnostic work-up for cough has been unrevealing, diagnostic testing of the esophagus should be considered. The tests include barium swallow, esophagoscopy, esophageal motility, and 24-hour esophageal pH probing. A therapeutic trial of H2 blockers or sucralfate might be tried prior to invasive testing; however, failure of these agents does not exclude the diagnoses as long-term therapy of up to 6 months may be necessary.
The pathophysiology of gastroesophageal reflux disease- induced cough is unclear. Esophageal pH probing studies relate cough directly to the number of refluxes in die distal esophagus but not in the proximal esophagus, suggesting that a reflex mechanism is involved. Other investigators feel that direct aspiration of acid into the lower respiratory tract is necessary in order for cough to occur.
CHRONIC COUGH IN CHILDREN
The differential diagnoses of chronic cough in children is identical to that seen in adult populations. While no large prospective series studying the spectrum and frequency of the cause of chronic cough have been reported in the pediatric literature, Parks et al18 and Kamei19 have reviewed the common causes of a persistent cough in children. The list is summarized in Table 3. It includes the two most common etiologies, asthma and postnasal drip syndromes, but emphasizes that recurrent viral infections and atypical organisms can be associated with a chronic persistent cough. Foreign body aspiration, while less commonly associated with chronic cough in adults, must always be considered in children and must be evaluated using specific testing. Maximum inspiratory and expiratory posteroanterior chest radiographs (looking for unilateral air trapping) can be helpful but only bronchoscopy can definitely rule out a foreign body. Congenital causes are far more commonly diagnosed in children and should be considered whenever appropriate or when the common causes have been rule out. These etiologies include cystic fibrosis, dysmotile cilia syndrome, sequestration, bronchial cyst, aberrant blood vessels compressing large airways, and immunodeficiency syndromes.
HABIT OR PSYCHOGENIC COUGH
Psychogenic or habit cough is probably the most controversial of all etiologies because there is no diagnostic test or series of tests to define it. It must be a diagnosis of exclusion. It has been described more commonly in children but exists in adults. It is the rarest of causes of chronic cough. Lokshin and Weinberger20 recently reviewed the literature of psychogenic cough.
The earliest published report occurred in 1963, and 153 cases were published through 1991 . Only 4 of 153 patients were over 18 years. The characteristics include a loud "barking seal," paroxysmal, dry cough not occurring at night. It increases when attention is paid to it. It is seen most commonly in young females. There is no response to asthma medications or antitussives. Therapy includes psychological intervention and various types of suggestion therapy.
The aforementioned material suggests very strongly that the most successful therapy is that aimed at tiie specific etiology of the chronic cough. In the series cited in Table 2, successful outcome was obtained in more than 90% of patients, ie, the cough either disappeared or was decreased satisfactorily according to the patient. This success suggests that few patients will require antitussive therapy to control or modify the cough. Irwin and Curley21 reviewed cough modifiers in a recent publication. Therapy was divided into antitussive therapy, that which controls or eliminates cough, and protussive therapy, that which makes a cough more effective. Some antitussives alter mucociliary factors that stimulate cough receptors. They may act by:
Causes of Chronic Cough In Children
* increasing the volume of mucus (true expectorants),
* decreasing mucus production,
* changing the consistency of mucus (mucolytics), and
* increasing mucociliary clearance.
Irwin and Curley21 found in their review of the literature that only ipratropium bromide in chronic bronchitis, iodopropylidene glycerol in chronic bronchitis and asthma, guaimesal in acute and chronic bronchitis, and dexbrompheniramine malete plus pseudoephedrine sulfate in the common cold have been shown to be effective antitussive drugs in studies designed to decrease cough frequency and intensity in randomized double-blind placebo-controlled studies in human subjects. Other antitussives increase the threshold or latency period of the cough center. These include both narcotic and nonnarcotic agents. All narcotic agents of the phenanthrene alkaloid group (morphine, meperidine, and codeine) all appear to be effective antitussives. Effective nonnarcotic antitussives available in the United States include dextromethorphan, diphenhydramine, and caramiphen.
Protussive therapy is defined as treatment that increase the effectiveness of cough, namely, airway clearance. Hypertonic saline aerosol in patients with bronchitis and ameloride aerosol in patients with cystic fibrosis improved cough clearance, but studies have not correlated overall improvement in the patient's clinical condition.
The approach to patients with chronic cough has been well defined and evaluated in the literature through a number of prospective studies. Meticulous attention to detail of the afferent loop of the cough reflex has helped identify the cause of cough in most patients. The most common causes appear to be similar in both children and adults and include asthma, postnasal drip syndromes, gastroesophageal reflux diseases, and aspiration. In children, recurrent viral infections and infections with atypical organisms also are very prevalent. Specific therapy directed at the cause alleviates the cough in most patients. In some patients, there may be more than one cause of cough. Invasive testing (eg, bronchoscopy and esophageal pH probing) is rarely necessary. In patients in whom a specific cause cannot be identified or in whom cough modifiers are necessary while specific therapy is taking hold, antitussives of both the narcotic and nonnarcotic variety are helpful.
1. Braman SS, Corrao WM. Chronic cough. Prim Care. 1985;12:217-225.
2. Corrao WM. Chronic cough: an approach to management. Compr Trier. 1986;12:14-19.
3. Korpas J, Widdicombe JG. Aspects of the cough reflex. Respfr Med. 1991;85(suppl A):3-5.
4. Higenbottam T1 Jackson M, Woolman P, Lowry R. Wallwork J. The cough response to ultrasonically nebulized distilled water in heart-lung transplantation patients. Am Rev ResprrDis. 1989;140:58-61.
5. Irwin RS, Corrao WM, Pratter MR. Chronic persistent cough in the adult: the spectrum and frequency of causes and successful outcome of specific therapy. Am Rev Respfr Dis. 1981;123:413-417.
6. Poe RH, Harder RV, Israel RH, Kallay MC. Chronic persistent cough. Chest. 1989;95:723-728.
7. Irwin RS, Curley FJ, French CL. Chronic cough. The spectrum and frequency of causes, key components of the diagnostic evaluation, and outcome of specific therapy. Am Rev Respfr Dis. 1990·,14?:640-647.
8. Pratter MR, Bam« T, Aken SM, Dubois ]. An algorithmic approach to chronic cough. Ann Intern Med. 1993;119:977-983.
9. Braman SS, Corrao WM. Cough: differential diagnosis and treatment. Clin Chest Med. 1987;8:177-188.
10. McFadden ER Jr. Exertional dyspnea and cough as preludes to acute attacks of bronchial asthma. N Engf ; Med. 1975;292:555-559.
11. Corrao WMi Braman SS, Irwin RS. Chronic cough as die »le presenting manifestation of bronchial asthma. N EnfdJMed. 1979;30Ck633-637.
12. Johnson D, Osborn LM. Cough variant asthma: a review of the clinical literature. J Asthma. 1991;28(2):85-90.
13. Pomerantz MA1 Braman SS. Cough reflexes in asthma. Am Rev Respir Dis. 1992;145(suppl A):12.
14. Choudry NB, Fuller RW. Sensitivity of the cough reflex in patients with chronic cough. Eur Respfr ). 1992;5:296-300.
15. Salem H, Aviado DM. Antitussive drugs. Am J Med Sri. 1964;247:585-600.
16. Boulet LP, Milot J, Boutet M, St Georges F, Laviolette M. Airway inflammation in nonasthmatic subjects with chronic cough. Am } Kespir Cri« Care Med. 1994;149:482-489.
17. Irwin RS, Zawacki JK, Curley FJ, French CL, Hoffman PJ. Chronic cough as the sole presenting manifestation of gastroesophageal reflux. Am Rev Respir Dis. 1989;140:1294-1300.
18. Parks DP, Ahrens RC Humphries CT, Weinberger MM. Chronic cough in childhood: approach to diagnosis and treatment. J Pedían. 1989;1 1 5:856-862.
19. Kamei RK. Chronic cough in children. PeaW CIm North Am. 1991;38:593-605.
20. Lokshin BM, Weinberg« M. The habit cough syndrome. The American Journal of Asthma 9 Allergy for Pediatricians. 1993;7:11-15.
21. Irwin RS, Curley FJ. The treatment of cough: a comprehensive review. Chest. 1991;99:1477-1484.'
Anatomy of a Cough Reflex
Common Causes of Chronic Cough
Causes of Chronic Cough In Children