Irina Petrache
Steve N. Georas
Patients who develop acute respiratory problems usually present with symptoms that result in the rapid diagnosis and treatment of the underlying disorder. On the other hand, chronic diseases of the lung that cause slowly progressive symptoms may go undetected unless incidentally discovered as part of a general medical evaluation. This chapter discusses common pulmonary problems with which the general practitioner is often confronted: cough, hemoptysis, dyspnea, noncardiac chest pain, and the abnormal chest x-ray.
Cough
Cough is an important defense mechanism that clears the airways of both secretions and inhaled particles (1). Although it often is associated with other respiratory symptoms, cough may be the symptom that prompts a patient to seek medical advice, especially if the cough is associated with complications (e.g., fear of serious disease, exhaustion, insomnia, lifestyle change, pain, hoarseness, urinary incontinence). A cough is composed of three phases: a deep inspiration, closure of the glottis accompanied by a rapid increase in intrathoracic pressure, and a final opening of the glottis with an explosive release of pressure.
Mucosal neural receptors that initiate a cough reflex are located throughout the nasopharynx, ears, larynx, trachea, and bronchi down to the level of the terminal bronchioles. They are rapidly adapting receptors with thin myelinated nerve fibers and show varied sensitivities to different stimuli. Stimulation of cough receptors in the nasopharynx also may cause sneezing. In contrast, stimulation of laryngeal receptors may initiate cardiovascular, bronchoconstrictor, and laryngoconstrictor reflexes, whereas stimulation of tracheal and bronchial receptors may also cause bronchospasm and airway mucus secretion. After activation of the receptors, impulses are conducted along afferent pathways in the ninth and tenth cranial nerves to the cough center located diffusely in the medulla. The reflex is completed through efferent pathways that cause forceful contraction of the diaphragm and other expiratory muscles. Although many different stimuli activate these receptors, all initiate cough by some form of mechanical or chemical irritation. The expression of some irritant receptors appears to be increased in the airways of people with chronic cough. Additional factors, such as acute inflammation of the airways,
P.870
may disrupt the bronchial mucosa, increase its permeability, and expose the receptors. The accompanying increases in respiratory secretions will lead to cough. Environmental pollutants, such as cigarette smoke, can directly stimulate the receptors without necessarily provoking an inflammatory reaction. Finally, although stimulation of irritant receptors may cause reflex bronchoconstriction, the bronchospasm itself, through reflex pathways, induces cough.
|
TABLE 59.1 Causes of Cough |
||||||||||||||||||||||||||||
|
||||||||||||||||||||||||||||
Acute Cough Syndromes
Table 59.1 lists the causes of cough. Generally, acute coughs are self-limited (<3 weeks) and are caused by viral upper respiratory tract infections (2). In contrast, cough that is triggered by mild bronchospasm may persist for weeks to months after a viral upper respiratory tract infection. Usually, viral infections, atypical pneumonias, and Pneumocystis carinii pneumonia are associated with nonproductive coughs, and bacterial infections are associated with significant sputum production. Younger patients tend to have a more productive cough associated with pneumonia, whereas older individuals, especially those with chronic obstructive pulmonary disease (COPD), may retain secretions because of impaired ability to clear them. A productive cough that follows a typical viral syndrome may signal the development of a superimposed bacterial bronchitis or pneumonia. High concentrations of air pollutants, such as insoluble gases (e.g., ozone, SO3, NO2), which are not irritating to the upper airway, can cause either a dry or a productive cough secondary to chemical irritation.
Chronic Cough Syndromes
A persistent cough (generally lasting >3 weeks) often is more bothersome than the acute cough syndrome. The most common cause of chronic coughing is cigarette smoking (2). The so-called smokers’ cough, a manifestation of chronic bronchitis, is generally described as hacking, worse in the morning, and productive or dry, as sputum often is ignored by cigarette smokers. The number of cigarettes smoked bears little relationship to the development of cough. Perhaps because they inhale more deeply, smokers of marijuana may complain of a persistent cough after smoking only one to two cigarettes daily. Patients with central bronchogenic and mediastinal tumors often present with cough, whereas patients with metastatic tumors or peripheral lung cancers that arise outside the airways or beyond irritant receptors seldom do.
In nonsmokers, three entities account for most cases of chronic cough: postnasal drip, asthma, and gastroesophageal reflux disease (GERD). The most common cause of chronic cough is postnasal drip, resulting from chronic sinusitis or allergic rhinitis (1,2). It is important to recognize that bronchospasm in both smokers and nonsmokers can be associated with a chronic dry cough. Cough may be the only manifestation of mild asthma (cough variant asthma) and need not be associated with dyspnea, wheezing, or changes in baseline pulmonary function (3). GERD may present with only minimal gastrointestinal symptoms, significant nagging cough, and, occasionally, hoarseness (seeChapter 42). A nocturnal cough that is precipitated or increased by lying flat makes this diagnosis more likely.
A dry hacking cough associated with dyspnea is common in patients in heart failure (see Chapter 66). Cough due to congestive heart failure also is often initiated by lying down. Similarly, cough may precede the complaint of dyspnea in patients with pulmonary emboli or bronchiolitis obliterans organizing pneumonia (a patchy pneumonia, probably immunologic, that often responds to treatment with corticosteroids). Bronchiectasis and chronic pulmonary infections, such as tuberculosis or nontuberculous mycobacterial pneumonia in immunocompetent patients, and P. carinii pneumonia in patients with acquired immunodeficiency syndrome (AIDS) commonly cause coughing. A chronic nonproductive cough occurs in up to 10% of patients taking an angiotensin-converting enzyme inhibitor and remits shortly (within 4 weeks) after the drug is discontinued. Because angiotensin-converting enzyme inhibitors are the treatment of choice for many conditions (e.g., congestive heart failure), it may be worth trying to “treat through” the cough in some patients.
P.871
Pharmacologic approaches can be used to try to suppress cough in patients who require angiotensin-converting enzyme inhibitors, including cromolyn sodium (e.g., via metered-dose inhalers, two puffs four times per day), baclofen (5 mg three times per day for 1 week, 10 mg three times per day for 3 weeks), low-dose theophylline, or sulindac. However, these regimens have not been evaluated in controlled studies or with large numbers of patients (4). Because the incidence of cough is much lower with angiotensin II receptor antagonists, a trial of these agents also may be reasonable.
There are numerous less common causes of chronic cough. For example, paroxysmal coughing often followed by an inspiratory whoop may indicate infection with Bordetella pertussis. A chronic cough may be caused by a process that stimulates the neural receptors in the pleura and pericardium. Even impacted cerumen in the external auditory canal can elicit a chronic cough. If the history and physical examination are unrevealing, it often is tempting to attribute chronic cough to a psychogenic cause; however, this is a rare cause of coughing, most often reported in children (1).
Evaluation
The acute and chronic cough syndromes are evaluated in similar ways. Usually, a history and physical examination yield a presumptive diagnosis. Information should be obtained about the development, duration, character, and precipitants of the cough; environmental or occupational exposure; smoking history; and any history of asthma or COPD. A history of constant swallowing or of throat clearing is associated with postnasal drip, even though the patient may deny many other symptoms associated with rhinitis or sinusitis.
Although the physical examination seldom provides a specific diagnosis, it may provide important clues. Careful examination of the ears, nose, throat, and lungs may yield relevant clues to a diagnosis. Cobblestoning in the posterior oropharynx represents lymphoid hyperplasia and is commonly seen in patients with chronic sinusitis. Examination of the chest may reveal rhonchi caused by the loose secretions that result from acute or chronic infection. A localized wheeze suggests a bronchogenic tumor, whereas wheezing at end expiration suggests active bronchospasm. Finally, the physical examination allows observation of the quality and severity of the cough. A harsh cough associated with loose secretions is characteristic of tracheobronchitis resulting from viral upper respiratory tract infection. When little or no coughing occurs in the course of the visit, the patient should be asked to cough to determine whether the cough is productive or is associated with wheezing. This procedure is useful because some patients refuse to admit to expectoration of sputum and often unconsciously swallow their secretions.
If a diagnosis is not obvious after a history and physical examination, a chest x-ray is indicated. It may reveal a tumor, pneumonia, or another chronic inflammatory process involving the lung parenchyma. The x-ray also may demonstrate atelectasis associated with a bronchogenic tumor, aspirated foreign body, or bilateral hilar adenopathy suggesting sarcoidosis. In patients with a normal x-ray, spirometry can be used to look for obstructive airway disease. However, a normal spirogram does not necessarily exclude the diagnosis (seeChapter 60). When the chest x-ray is normal, bronchoscopy seldom provides additional useful information (2). Although a proximal bronchogenic tumor can be hidden on a chest x-ray by the mediastinal shadows, patients with these tumors often have associated hemoptysis. If the history, physical examination, chest x-ray, and spirogram are unrevealing and if the patient's cough persists after stopping new medicines, including angiotensin-converting enzyme inhibitors and β-blockers (including eye drops), referral to a subspecialist may be appropriate. Additional tests might include methacholine challenge (asthma), high-resolution chest computed tomography (CT) (bronchiectasis, interstitial lung disease), sinus x-ray or CT (chronic sinusitis), 24-hour pH probe (GERD), and, rarely, bronchoscopy (endobronchial tumor or aspirated foreign body) or cardiac evaluation (heart failure). Use of an algorithm-based approach that incorporates the clinical assessment of disease probability in evaluating patients with chronic cough has been validated, identifying a cause for cough in 93% of patients (2). Figure 59.1 shows such an algorithm, which was developed by a consensus panel.
Therapy
Specific therapy for the various acute inflammatory and irritating processes likely to cause coughing is discussed in detail in individual chapters dealing with these topics.
In general, viral tracheobronchitis requires only symptomatic therapy because coughing usually subsides spontaneously in 2 to 4 weeks. Patients with persistent coughing and a history or physical examination compatible with bronchospasm may benefit from bronchodilators. Treatment should begin with an inhaled β2-sympathomimetic agonist. A detailed therapeutic approach to the pharmacologic treatment of bronchospasm is presented in Chapter 60.
Cessation of cigarette smoking and avoidance of a polluted environment may be the most important aspects of the therapy for both acute and chronic cough. It often is difficult to convey to a smoker that smoking as few as one or two cigarettes per day causes airway irritation and inflammation. Ipratropium bromide may improve cough and decrease sputum production in patients with chronic bronchitis (see Chapter 60) (1).
|
FIGURE 59.1. Guidelines for evaluating chronic cough in immunocompetent adults. ACEI, angiotensin-converting enzyme inhibitor; BaE, barium esophagography; GERD, gastroesophageal reflux disease; HRCT, high-resolution computed tomography; Hx, history; PE, physical examination; PNDS, postnasal drip syndrome. (From Irwin RS, Boulet LP, Cloutier MM, et al. Managing cough as a defense mechanism and as a symptom. A consensus panel report of the American College of Chest Physicians. Chest 1998;114:133S, with permission.) |
P.872
The treatment of postnasal drip and GERD is discussed in Chapters 33 and 42, respectively. In most patients, cough improves within 1 week of initiation of therapy for postnasal drip, but GERD may not resolve for months, even with optimal therapy (2). Removal of impacted cerumen in the auditory canal provides immediate relief. Approximately one fourth of patients with chronic cough referred for subspecialty evaluation had more than one cause (2). Thus, if specific therapy does not eliminate the cause, additional testing and treatment should be pursued.
After specific therapy has been initiated, the use of antitussives should be considered. Despite the enormous demands for antitussives, these preparations are absolutely necessary in only a few situations. Moreover, the expectoration of sputum is a major goal of therapy for chronic obstructive airway disease. Therefore, when antitussives are needed in patients with productive coughs, it usually is better to attempt cough reduction (not total suppression), primarily to allow patients to sleep and to avoid posttussive syncope, stress incontinence, or straining of the chest wall or abdominal muscles. In the United States, several hundred cough and decongestant preparations, usually sold as combination products, are available. Many of these preparations combine so-called expectorants with antitussives and should be avoided because, insofar as they
P.873
have an effect, they work at cross purposes. In prospective double-blinded studies of patients with cough associated with the common cold, the combination of dexbrompheniramine, an antihistamine (contained, for example, in Cheracol and Drixoral), and pseudoephedrine, a vasoconstrictor (6 mg/120 mg twice per day, orally for 1 week), reduced symptoms compared with placebo (5), as did naproxen (500 mg loading dose, then 200–500 mg three times per day orally for 5 days) (6). Another randomized study of 97 patients with cough secondary to upper respiratory tract infections found no difference between guaifenesin alone versus guaifenesin plus codeine or guaifenesin plus dextromethorphan in reducing coughing (7).
|
TABLE 59.2 Nonnarcotic Antitussives |
|||||||||||||||
|
Antitussives act on the cough reflex by either anesthetizing the peripheral irritant receptors or increasing the threshold of the cough center. The two most effective nonnarcotic antitussives are dextromethorphan and benzonatate, although the latter has not been rigorously studied in a placebo-controlled randomized trial (Table 59.2). Dextromethorphan is chemically derived from the opiates; however, it is classified as nonnarcotic because at prescribed dosages it has no sedative or analgesic effects and therefore has little potential for abuse. It is available over the counter in a variety of preparations (e.g., Dimetane DX). Dextromethorphan suppresses cough centrally. Occasionally, the drug causes nausea, dizziness, or vertigo, and overdosage of >200 mg may lead to central nervous system (CNS) depression. Benzonatate is a peripherally acting anesthetic similar to tetracaine. Rarely, it causes headaches, dizziness, and nausea or gastrointestinal upset. The drug should not be chewed or sucked because this action will result in an unpleasant taste and prolonged oral pharyngeal anesthesia. Overdosage has been associated with CNS stimulation and tremors, which may lead to seizures followed by profound CNS depression. It is reasonable to treat patients initially with dextromethorphan and, if intolerable cough persists, to substitute benzonatate.
If nonnarcotic antitussives are ineffective, codeine can be tried. Many clinicians prescribe codeine preferentially to patients with persistent cough because it is a more potent cough suppressant than the nonnarcotic agents. Codeine is effective in dosages of 15 to 30 mg administered every 3 to 6 hours. The common side effects—nausea, vomiting, constipation, dry mouth, and sedation—usually are not experienced at these lower dosages.
Hemoptysis
Hemoptysis is defined as the expectoration of blood from below the vocal cords. It can range from flecks of blood in sputum to the coughing of large amounts (>1 L) of blood. Distinguishing between hemoptysis and hematemesis occasionally is difficult. Blood from the lungs usually is bright red and frothy, has an alkaline pH, and usually is mixed with sputum containing macrophages and white blood cells. Often, patients with hemoptysis complain of a tickling or irritation in the chest. On the other hand, hematemesis is characterized by blood that is darker brown, has an acid pH, and is mixed with food particles. Sometimes blood from a lesion in the sinuses or in the upper airway is aspirated and later expectorated, giving the appearance that bleeding occurred in the lower respiratory tract. A careful history and physical examination must be performed to avoid inappropriate evaluation or treatment. The patient should be instructed to collect and save the bloody sputum so that the hemoptysis can be quantified. Nevertheless, a history of hemoptysis should not be ignored if a patient cannot produce a specimen on command, because the symptoms can be intermittent. Table 59.3 summarizes the various pulmonary causes of hemoptysis.
In the typical ambulatory practice, chronic bronchitis is by far the most common cause of blood streaking of the sputum, followed less commonly by lung cancer (8,9). The likelihood of a particular diagnosis depends on the patient population (e.g., smokers vs. nonsmokers) (8). Bronchiectasis is less common today in the industrial world because of mass screening for tuberculosis, childhood vaccinations for measles and whooping cough, and antibiotic treatment of serious respiratory infections. Active cavitary tuberculosis is also a less common cause of hemoptysis than it once was, but residual upper lobe bronchiectasis, the result of old tuberculosis infection, is still seen.Bronchogenic carcinoma (see Chapter 61) presents with hemoptysis at two stages. Blood-streaked sputum may be a brief manifestation of a small irritative mucosal lesion. This symptom may resolve, only to be replaced by major hemoptysis from a large endobronchial tumor that is friable or necrotic or is eroding central vessels. Usually blood from a necrotizing pneumonia or a lung abscess is mixed with pus, and the sputum appears red-brown or red-green. Hemoptysis from pulmonary emboli, a manifestation of pulmonary
P.874
infarction, is rare because of the lung's dual blood supply, unless patients have significant heart or lung disease. Even with the advent of fiberoptic bronchoscopy, the cause of hemoptysis remains undiagnosed 8% to 15% of the time (10,11). The 5-year survival rate for patients with cryptogenic hemoptysis (hemoptysis with normal chest x-ray and a negative bronchoscopy) is very good (85%–95%) (10).
|
TABLE 59.3 Pulmonary Causes of Hemoptysis |
||||||||||||||||||||||||||
|
||||||||||||||||||||||||||
Table 59.3 also lists less common causes of hemoptysis, but this ranking reflects to some extent the location of a practice. For example, mycetomas and parasitic infections that cause hemoptysis are much more common in areas of the country where those problems are endemic. Hemoptysis is common in bronchial carcinoids by virtue of their endobronchial location and marked vascularity. Hemoptysis caused by pulmonary metastasis from a solid tumor is rare. Its presence raises the possibility of endobronchial metastases, which are most common in patients with breast, colon, and kidney cancer and those with malignant melanoma. Patients with mitral stenosis and pulmonary vascular congestion are prone to hemoptysis with any source of lung irritation. Although certainly less common today, this valvular abnormality often is silent and the history of rheumatic fever forgotten. Patients taking the anticoagulants warfarin or heparin may develop hemoptysis, especially if there is an associated inflammation of the airways. Occasionally, blunt chest trauma produces hemoptysis in an otherwise healthy individual. Very rarely, hemoptysis is due to intrathoracic endometriosis, in which case it occurs at the time of menstruation.
Evaluation
The diagnostic evaluation of hemoptysis is aimed at determining the cause, localizing the site, and quantifying the amount of bleeding. The history and physical examination are directed at uncovering clues to the causes outlined in Table 59.3. An attempt should be made to quantitate the amount of hemoptysis by history and by collection of expectorated blood. Massive hemoptysis, generally defined as more than a few hundred milliliters of blood during a 24-hour period, is a medical emergency, and survival of the patient depends on rapid diagnosis and treatment (12). The principal risk of massive hemoptysis is asphyxiation, not exsanguination, with the rate of bleeding as the most important prognostic factor. Patients with underlying lung disease are less able to tolerate spillage of blood into other portions of the lung before acute respiratory failure develops.
During the physical examination, extrathoracic sources of bleeding from the nasal passages, sinuses, and pharynx should be sought. Physical findings may be helpful. Digital clubbing may be seen in non–small cell lung cancer, lung abscess, or bronchiectasis. Scattered ecchymoses, multiple petechiae, or gastrointestinal bleeding suggests a defect in hemostasis, and telangiectasia of the skin, lips, or buccal mucosa is consistent with hereditary hemorrhagic telangiectasia (or Osler–Weber–Rendu syndrome). Ulceration and crusting of the nasal septum may represent upper airway involvement of Wegener granulomatosis. The significance of unilateral wheezing or crackles must be interpreted with caution because these sounds may be produced by aspirated blood or secretions rather than by endobronchial tumor.
A chest x-ray is essential because acute inflammatory diseases, such as active tuberculosis, pneumonia, and lung abscess, will produce obvious radiographic abnormalities. Typically, lung cancers associated with hemoptysis are centrally located squamous cell carcinomas, and approximately half are cavitary. However, localization of the bleeding source often is precluded by bilateral aspiration of blood or by the presence of bilateral pulmonary disease. Patients with bronchitis often have normal chest x-rays, and the findings on plain film of focal bronchiectasis may be nonspecific. If the bronchiectasis is a result of old tuberculosis, however, apical scarring may suggest the diagnosis; otherwise, there may be increased or crowded lung
P.875
markings, thickened dilated bronchi, multiple cystic cavities (1–3 mm in diameter), or infiltrates because of recurrent infection. The chest CT is more sensitive than the chest x-ray in detecting bronchiectasis and generally is sufficient to make the diagnosis (13). Differentiating bronchitis from bronchiectasis by history and chest x-ray sometimes is difficult. This distinction may not be critical, however, because the acute medical management of bronchitis and bronchiectasis in patients with hemoptysis is the same.
When the chest x-ray is normal or nonlocalizing, endobronchial malignancy is the principal diagnosis to exclude, although bronchitis is the most likely diagnosis. Individuals younger than 40 years, including smokers, with hemoptysis that has lasted <1 week are unlikely to have cancer. In such patients, observation is a reasonable initial approach (14).
Persistent or recurrent hemoptysis mandates a thorough evaluation that includes bronchoscopy. Patients with normal chest x-rays who are at increased risk for lung cancer (age >40 years, >20 pack per year cigarette smoker) should undergo bronchoscopy. Still, lung cancer will be discovered at bronchoscopy in only approximately 5% of these patients (see Chapter 61) (14,15). Sputum cytology may provide the diagnosis in as many as half of these patients, but in general bronchoscopy still is required to locate the site of malignancy (lung vs. upper aerodigestive tract) and to plan for therapy. In the evaluation of recurrent or persistent hemoptysis, most clinicians view bronchoscopy and chest CT as complementary, with CT being helpful in guiding the bronchoscopy and/or angiography to the regions of highest yield (13,16).
Therapy
Blood irritates the tracheobronchial tree and triggers constant coughing, which by itself is traumatic. Mild cough suppression may help, but the patient must be able to expectorate blood as it accumulates. Specific treatment depends on the underlying cause of hemoptysis. Chronic bronchitis, with intercurrent hemoptysis, usually is treated on an ambulatory basis with antimicrobial drugs (see Chapter 60) for 10 to 14 days. In such circumstances, blood streaking of the sputum usually stops in 2 to 3 days.
No clinical criteria or radiographic signs predict massive hemoptysis, and the quantity of hemoptysis does not necessarily indicate the seriousness of the patient's underlying disease. Thus, given the tendency for rebleeding and the often unpredictable clinical course, a low threshold for hospitalization is warranted. If massive hemoptysis is present, consideration should be given to early bronchoscopy and interventional angiography with bronchial artery embolization (12,17) or to surgical resection.
Dyspnea
Breathing is an unconscious act that usually occurs effortlessly, yet even normal people become aware of their breathing during deep sighs or during moderate to severe exercise. Dyspnea, the abnormal uncomfortable sensation of breathlessness, is difficult to define because patients often cannot accurately perceive or quantitate the feeling. Similar to an individual's threshold for pain recognition, the complaint of dyspnea depends on both the individual's limit for discomfort and the specific circumstances that provoke shortness of breath. Thus, dyspnea must be defined in terms of what is abnormal for a particular individual in the context of his or her fitness level and of the amount of activity that is associated with breathlessness. Some patients become dyspneic with relatively small measurable alterations in ventilation, whereas others (e.g., patients who are hyperventilating with Kussmaul breathing) may not complain of dyspnea. Fortunately, a reasonable correlation exists between the degree of dyspnea and objective measurements of physiologic dysfunction.
Often, the actual complaint of dyspnea may not be expressed as such. It may vary depending on the type of precipitating illness and on whether it developed abruptly or over a longer period. Thus, asthmatic patients may complain of tightness in the chest and acute shortness of breath, whereas patients with acute pulmonary embolism (PE) may state that their breath has suddenly “been taken away,” and they cannot get enough air even though they ventilate easily. A sensation of air hunger or suffocating is typical for patients with congestive heart failure. In contrast, patients with emphysema or neuromuscular diseases may note an increased effort or work of breathing and may modify their lifestyles and dismiss the sensation of breathlessness as part of their advancing age.
Normal Ventilation
No single mechanism is responsible for dyspnea. Because dyspnea is the result of a variety of diverse influences acting alone or together, a brief discussion of the control of ventilation may help the practitioner understand the complexity of dyspnea and the reason why this sensation often does not immediately respond to correction of obvious physiologic abnormalities. Normally, ventilation is coupled to the individual's metabolic demands as reflected in the oxygen consumption and carbon dioxide elimination necessary to meet a given level of activity. These needs are sensed by peripheral (carotid and aortic bodies) and central (medullary) chemical chemoreceptors that respond to the O2, CO2, and pH of blood and cerebrospinal fluid. The acute stimulation of these receptors provokes changes in minute ventilation. In addition, the control and regulation of the rate and pattern of breathing are influenced
P.876
by the reflex effects of activation of neural receptors that lie in the lung parenchyma, airways, blood vessels, respiratory muscles, and chest wall. For example, receptors in the chest wall and diaphragm respond to increased stiffness (decreased compliance) in the lung that occurs with fluid accumulation or with interstitial fibrosis. In addition, interstitial edema may activate nerve fibers located in the alveolar interstitium and may cause reflex dyspnea in patients with pulmonary edema. Other receptors located in the airway epithelium cause rapid shallow breathing, coughing, and bronchospasm when irritating substances are inhaled. Finally, the CNS alone can cause large alterations in breathing that lead to hyperventilation in association with anxiety attacks. This discussion should help in understanding, for example, why correction of arterial hypoxemia alone in a patient with an asthmatic attack usually does not relieve the sensation of breathlessness. In this situation, dyspnea results from the complex interaction of both chemical and neural stimuli to breathe, coupled with an individual's response to these signals. Therefore, correction of only one of these problems is not sufficient to abolish dyspnea. A detailed consensus panel report on the pathophysiology and management of dyspnea has been published (18).
|
TABLE 59.4 Causes of Dyspnea |
|||||||||||||||||||||||||||||||||||||||||||||||||||
|
|||||||||||||||||||||||||||||||||||||||||||||||||||
Evaluation
The causes of dyspnea are diverse and include essentially all diseases that result in significant functional impairment of either the respiratory system (gas exchange and pulmonary mechanics) or the cardiovascular system (circulatory and cardiac function) and any hematologic abnormality that impairs oxygen delivery. Table 59.4 summarizes the general disease categories likely to cause abnormal breathlessness.
In ambulatory practice, the major causes of dyspnea are obstructive airway disease and atherosclerotic and hypertensive heart disease, either alone or in combination. The prevalence of symptomatic lung disease in a specific geographic region or socioeconomic group is further modified by the prevalence of cigarette use, urban pollution, and occupational exposure to inhaled toxic substances. The clinical circumstances and sequence of events in which dyspnea occurs will aid in its evaluation.
One of the first steps in evaluating a patient who complains of dyspnea is deciding whether the symptoms reflect an acute or a chronic event, because the more serious causes of dyspnea tend to present abruptly. In general, dyspnea of sudden onset is easier to evaluate, but the workup must proceed quickly to determine whether the patient should be admitted to the hospital for more intensive evaluation and therapy. On the other hand, the evaluation of chronic dyspnea usually can be accomplished more slowly in an ambulatory setting.
Acute Dyspnea
The history, physical examination, and chest x-ray form the focal point of the evaluation of a patient with acute dyspnea. In a young patient, the medical history and physical examination alone often suggest the presumptive diagnosis. When necessary, additional distinction of primary
P.877
cardiac from pulmonary disorders will be aided by the chest x-ray, spirogram, and electrocardiogram.
Acute tracheobronchitis should be considered in the middle-aged smoker with cough, dyspnea, and purulent sputum in association with a clear chest x-ray. When wheezing and rhonchi are present, the term asthmatic bronchitis is often used. Spontaneous pneumothorax (see below) presents with sudden sharp chest pain and dyspnea. A small but significant pneumothorax on chest x-ray can easily be missed, and diagnostic accuracy will be improved with an expiratory film. Previously undiagnosed interstitial lung disease, bullous lung disease, andcystic fibrosis also may present with spontaneous pneumothoraces. In these cases, the chest film should demonstrate characteristic abnormalities.
Acute dyspnea in association with fever, cough, and purulent sputum with localized infiltrates suggests pneumonia, usually bacterial. Diffuse infiltrates and nonproductive cough suggest atypical pneumonia (see Chapter 33).
The patient with acute dyspnea and heart failure has usual cardiac symptoms and signs, including paroxysmal nocturnal dyspnea, crackles, cardiomegaly, and a symmetric interstitial pattern with or without pleural effusions (see Chapter 66). Psychogenic dyspnea, or the hyperventilation syndrome, has a rapid onset and usually is found in patients with anxiety disorders (see Chapter 22). This syndrome should be considered in young patients in whom dyspnea is unrelated to exertion and is associated with somatic complaints and excessive fearfulness (19).
Less common, but important, causes of acute dyspnea include acute foreign-body aspiration, usually evident from the history of aspiration and a physical examination that demonstrates decreased breath sounds over the part of the lung supplied by the occluded bronchus. During the heating season or in certain industrial settings, carbon monoxide intoxication should be considered as a cause of headaches and dyspnea. Diagnosis requires a high degree of suspicion and awareness of the problem. Confirmation requires measurement of carboxyhemoglobin with a co-oximeter. The partial pressure of oxygen measured in the arterial blood gas sample will remain normal.
Pulmonary Embolism
PE is a major life-threatening cause of acute dyspnea, but the diagnosis can be difficult. Its evaluation requires a systematic approach with a logical sequence of diagnostic testing. Approximately 75% of patients suspected of having deep venous thrombosis (DVT) or PE do not actually have these conditions (20).
The incidence of PE is high in patients with chronic obstructive lung disease or congestive heart failure and in those with risk factors for venous thromboembolism (e.g., cancer, prolonged immobilization, or a strong family history of DVT; see Chapter 57) (20). In a prospective study of older women (>60 years of age), obesity, heavy cigarette smoking, and high blood pressure were significant risk factors for PE (21). In younger women, use of oral contraceptives substantially increases the risk of PE (21). The physical examination usually is not helpful in the diagnosis, especially because many of the patients have underlying respiratory and cardiovascular diseases that may themselves produce abnormal physical findings, such as tachycardia, tachypnea, distended neck veins, and an accentuated pulmonic component of the second heart sound.
Most laboratory tests are not useful in the diagnosis of PE. Chest x-rays often are abnormal, but the findings are nonspecific (localized infiltrates or oligemia, atelectasis, elevated hemidiaphragm, pleural effusion). Arterial blood gas tensions often are abnormal (reduced PaO2and PaCo2) but are not helpful diagnostically, in part because of considerable variation and in part because of the high prevalence of cardiopulmonary diseases that alters both PaO2 and PaCo2. An exception is the D-dimer assay. D-Dimer is a breakdown product of fibrinolysis, and its concentration in the blood is increased when clotting occurs. When coupled with a low clinical suspicion of DVT or PE, a negative quantitative D-dimer assay essentially excludes the diagnosis (22, 23, 24). For example, if the pretest probability of DVT is <10%, then the posttest probability of DVT after a negative D-dimer rapid enzyme-linked immunosorbent assay (ELISA) is <1% (23). In a 2005 study, none of 232 patients with low or intermediate clinical probability of PE and negative D-dimer assay developed venous thromboembolism after 3 months of followup (24). The diagnostic utility of the D-dimer assays is strongly dependent on how it is measured. ELISAs and quantitative rapid ELISAs have the best sensitivity and negative likelihood ratios (23). Of note, a positive D-dimer assay is not useful in establishing the diagnosis of PE. Furthermore, if the pretest probability of PE is high, then a negative D-dimer assay should not be used to rule out the diagnosis, and additional imaging studies should be obtained. The best way to determine the pretest probability of PE currently not known. In general, the overall impression of an experienced clinician fares about as well as clinical prediction rules for PE, which may be especially valuable for less experienced clinicians (e.g., house staff in training) (25). This is an active area of research where additional prospective studies are needed (see Chapter 57).
If the pretest probability of PE is high (or if the D-dimer assay is positive), then additional imaging studies should be obtained. Three imaging studies of the chest can be performed: pulmonary angiography, ventilation–perfusion (V/Q) scanning, and contrast-enhanced CT. Pulmonary angiography has been considered the “gold standard” for diagnosing pulmonary emboli. However, angiography is uncomfortable for the patient, involves a significant dye load, and requires prolonged breath-holding (up to 30 seconds).
P.878
This can be challenging in acutely dyspneic patients and may result in angiograms that are difficult to interpret, especially in subsegmental vessels. V/Q scans or chest CT scans are much less invasive and are the preferred initial tests in the workup of suspected PE. Which of these two scans is the test of choice is a subject of debate. CT scans now are more commonly ordered than V/Q scans in hospitalized patients with suspected PE (26). The choice of V/Q versus CT scan likely will be determined by local institutional expertise in the acquisition and interpretation of these two studies.
Patient Experience
Little discomfort is associated with a lung scan. With a V/Q scan, the patient should be instructed that he or she will inhale a mixture of oxygen and xenon for 3 to 4 minutes, followed by a venous injection of radioactive-labeled technetium. Several different projections are recorded on a scanner while the patient is lying on a table and breathing spontaneously. During a spiral CT scan, as the patient moves through the scanner without stopping, the x-ray tube rotates continuously in the same direction. Acquisition of the scan is carefully timed with an intravenous injection of contrast dye. The patient will have to hold his or her breath, for as little as 10 seconds in newer CT scanners. Patients may experience some mild hot flashes as the dye is injected.
The V/Q scan is a highly sensitive test, but it can be nonspecific depending on the configuration, location, and number of perfusion defects seen. There are well-established criteria for interpreting the results of V/Q scans (27). In general, the greater the perfusion defects without corresponding ventilation defects, the higher the “probability” of the scan. If the V/Q scan is normal, the diagnosis of an acute PE is excluded. Conversely, a high-probability scan is associated with an 85% to 90% chance of PE. As with the D-dimer assay, the predictive value of V/Q scans is highly dependent on the pretest probability of PE (27). Unfortunately, only 10% to 15% of patients with suspected PE will have a high-probability scan, and fewer than 5% of scans will be normal (27). Most patients will have an intermediate-probability (or nondiagnostic) V/Q scan and require further testing to confirm or exclude the diagnosis. Remember that a “low-probability” scan does not exclude the diagnosis of PE. In particular, if there is a high clinical suspicion, up to 40% of patients with low-probability V/Q scans will have documented PE on pulmonary angiography (27). In patients with nondiagnostic V/Q scans, abnormal compression ultrasonography (or impedance plethysmography) may detect a proximal DVT and confirm the need for anticoagulation (22). DVT will be detected by initial testing in approximately 10% of these patients, who should be hospitalized for initiation of anticoagulant therapy (see Chapter 57). Because ultrasonography does not detect calf vein thromboses, some patients with an initially negative test are at risk for propagating a thrombus and having a PE. This occurs 2% to 15% of the time, depending on risk factors for DVT. If there is adequate cardiopulmonary reserve, serial noninvasive testing (e.g., at days 5 and 10) is a reasonable strategy. Alternatively, if clinical suspicion for PE is high or there is limited cardiorespiratory reserve, additional imaging studies should be obtained.
Newer-generation “multislice” CT scanners are becoming widely available and provide exceptional resolution of lung structures and vessels even to the subsegmental level. By injecting a carefully timed bolus of contrast medium, images can be obtained in a single breath-hold, thus allowing visualization of the main, lobar, and segmental pulmonary arteries in most patients. The diagnosis of PE usually is obvious as a low-density filling defect within the vessel lumen, but there is a lack of widely agreed upon standards for diagnosing PE by spiral CT. Therefore, a highly trained interpreter with a knowledge of bronchovascular anatomy (and its variants) is essential. The advantages of a spiral CT include its relatively quick acquisition time and its ability to diagnose abnormalities within the lung parenchyma and mediastinum. In several studies, unsuspected abnormalities (e.g., pneumothorax, cancer) were found in 10% to 30% of subjects undergoing spiral CT to “rule out PE” (28). Studies have shown that multislice CT scanners are very sensitive. For example, a clinical trial that used a D-dimer assay in conjunction with multislice CT scanning found that compression ultrasonography did not improve diagnostic yield in patients with a negative CT scan (24). Therefore, ultrasonography may be unnecessary in patients with negative multislice CT scans. Furthermore, the subsequent rate of venous thromboembolism in patients with negative CT scans is very low (e.g., approximately 1.5% at 3 months), comparable to the rate in patients with negative pulmonary angiography (24,29). This finding has fueled the argument that CT scans should become the new “gold standard.” The CT scan versus angiogram debate is unlikely to be resolved by a clinical trial. A definitive study would require randomly assigning patients to CT scan or angiography, withholding anticoagulation in those with negative studies, and observing patients for subsequent venous thromboembolism. Such a study would require thousands of patients in each arm and is unlikely to be performed soon.
The resolution of PE varies and can occur as early as 1 to 2 weeks in patients with small emboli. In patients with larger emboli and in patients with underlying cardiopulmonary disease, angiographic evidence of emboli may persists for 2 to 3 months (30). If chest pain occurs after discharge, a subsequent lung V/Q or chest CT scan is necessary to determine whether embolization has recurred.
Evaluation of Chronic or Progressive Dyspnea
In contrast to acute dyspnea, chronic dyspnea usually is more difficult to diagnose and often requires more extensive diagnostic procedures; therefore, the evaluation
P.879
should proceed in a logical sequence to avoid expensive and invasive laboratory testing. Because shortness of breath is appropriate to certain levels of activity depending on the fitness of the individual, the clinician must decide whether the patient's symptoms are abnormal and over what period they have developed. Many patients with chronic cardiopulmonary disease or chronic anemia adapt to the insidious onset of dyspnea by subconsciously changing daily habits and avoiding physical activity. The degree of dyspnea should be determined by comparing the patient's abilities to perform work with an appropriate peer group and with his or her baseline performance. Thus, the complaint of dyspnea in a 35-year-old patient who normally runs 5 miles and now becomes short of breath after running only 2 miles should not be ignored.
The most useful initial test is the chest x-ray, which often is abnormal and therefore directs subsequent evaluation. Patients with COPDassociated with emphysema have hyperinflation, decreased lung markings, and often evidence of bullous formation (see Chapter 60). Largepleural effusions, lung cancer, or heart disease associated with dyspnea results in obvious changes in the chest roentgenogram with evidence of fluid occupying at least half of one hemithorax, large mass lesions, or cardiomegaly, respectively. Interstitial lung disease that has led to fibrosis is revealed by chest x-ray, although early cases of interstitial lung disease will be detectable only by high-resolution CT scan. Unilateral hemidiaphragm paralysis results in obvious asymmetry in lung expansion. Patients with this condition often describe orthopnea secondary to difficulty with diaphragmatic excursion in the recumbent position.
Patients with dyspnea and a normal or nonspecific chest x-ray represent a challenging group to diagnose. Table 59.5 provides an approach to the evaluation of these patients. Most of these patients have obstructive lung disease. A spirogram is useful to screen for occult lung disease because a normal spirogram nearly excludes significant parenchymal disease. Although patients with exercise-induced asthma may have a normal spirogram during symptom-free periods, more commonly there is evidence of slight reduction in the baseline forced expiratory volume as a percentage of forced vital capacity. Home peak flow monitoring may confirm the diagnosis in these patients. Chapter 60 discusses additional specialized procedures that aid in the diagnosis of exercise-induced asthma. In one study of 72 patients referred for evaluation of chronic dyspnea not diagnosed by history, physical examination, chest x-ray, or spirometry, the two most common diagnoses were asthma/reactive airway diseases (approximately 17%) and hyperventilation syndrome (approximately 20%). There was a wide spectrum of underlying diseases in the remaining cases. Notably, 20% of patients remained undiagnosed despite extensive evaluation (31).
In general, obesity is not associated with dyspnea unless body weight is markedly increased (50%–100% or more, or ≥100 lb over ideal weight). Primary pulmonary hypertension may be associated with subtle dilation of the pulmonary arteries on chest x-ray and is most commonly seen in young women. Pulmonary hypertension also can occur in association with collagen vascular diseases and should be suspected especially in patients with dyspnea on exertion in whom the chest X-ray does not show parenchymal lung disease. Upper airway obstruction resulting from goiter, for example, often is not apparent on a routine posteroanterior and lateral chest x-ray. Anemia usually does not cause dyspnea unless it has developed acutely (blood loss or hemolysis) or is relatively severe (e.g., hematocrit values ≤20%). Some patients with advanced hepatic cirrhosis complain of severe dyspnea that is especially worse when they are upright (“platypnea”). These patients experience excess shunting of blood through abnormal vascular channels in the lung (the hepatopulmonary syndrome) (32).
|
TABLE 59.5 Workup of Chronic Dyspnea when the Initial Workup (e.g., Chest X-Ray, Spirometry) Is Unrevealing |
||||||||||||||||||||||||||||||||
|
||||||||||||||||||||||||||||||||
In patients in whom the diagnosis is uncertain, laboratory testing should include a hemoglobin determination or hematocrit value to determine whether the patient has severe anemia or erythrocytosis, and thyroid function
P.880
studies if symptoms of thyroid dysfunction (e.g., unexplained weight gain, fatigue) are noted.
Complete pulmonary function tests should be performed. In addition to baseline spirometry (see Chapter 60), other pulmonary function tests include measurements of total lung capacity and functional residual capacity, which quantitate the degree of hyperinflation or restriction. Categorization of a disorder as obstructive or restrictive will direct the clinician to a narrowed list of causes. Thediffusing capacity measures the amount of alveolar capillary surface area available for gas exchange. Thus, the diffusing capacity is reduced in patients with PE and other vascular occlusive diseases and in patients with emphysema. In contrast, an elevated diffusing capacity is found in conditions that increase the pulmonary blood volume, for example, erythrocytosis, early congestive heart failure, or obesity. Measurement of inspiratory and expiratory pressures helps characterize neuromuscular problems. Flow–volume loops help identify upper airway sources of obstruction.Chapter 60 describes the experience of the patient during the performance of these tests.
Cardiovascular testing should be performed. Chapters 65 and 66 discuss the use of specialized noninvasive cardiovascular evaluation, including echocardiography and nuclear scanning, to assess right and left ventricular function or the presence of valvular heart disease.
If a patient is dyspneic on exertion and baseline testing of cardiopulmonary function, as described earlier, is normal or only mildly abnormal, exercise testing should be considered. In general, two types of exercise tests are available. The first is a standard cardiac stress test, during which the patient exercises and is observed for the development of chest pain and for electrocardiographic or radionuclide ischemic changes (see Chapter 62). The second type of exercise test is a cardiopulmonary stress test in which cardiac function, pulmonary gas exchange, ventilation, and physical fitness are quantitated at specific workloads. The two types of tests are similar, but the patient should be told that the cardiopulmonary test requires continuous exercise while breathing into a mouthpiece, and measurement of oxygenation is made by either an oximeter or an indwelling arterial line. Such complicated cardiopulmonary stress testing is justified and useful to determine whether dyspnea is caused by cardiac disease; pulmonary disease, including exercise-induced asthma or occult pulmonary vascular disease; deconditioning; or combinations of these conditions. This type of testing is particularly useful in evaluating patients for disability compensation because static pulmonary function and noninvasive cardiac testing may not accurately predict the functional state of a given patient during actual working conditions. The referring clinician usually can determine presumptively the most likely cause of dyspnea and can make the appropriate referral for the specific exercise test. In large hospital centers having combined cardiopulmonary laboratories, simultaneous consultation and exercise testing by cardiologists and pulmonologists may be available. Stress testing also can be useful in deciding whether or not the patient needs supplemental oxygen (see Chapter 60).
This approach to the evaluation of dyspnea almost always answers the questions necessary for diagnosis of the underlying condition and for establishment of a therapeutic regimen.
Therapy
Treatment of dyspnea is primarily aimed at therapy for the underlying cardiac, pulmonary, neuromuscular, or hematologic disorders causing the abnormal breathlessness. In certain patients with underlying irreversible lung disease, specific measures that improve respiratory muscle function may alleviate symptomatic dyspnea. Training programs that increase both muscle strength and endurance are available in selected pulmonary rehabilitation centers, even for patients with advanced lung disease. Pulmonary rehabilitation can result in decreased shortness of breath in many patients. Because anxiety and depression are commonly associated with the development of chronic cardiopulmonary disorders associated with dyspnea, appropriate anxiolytics or antidepressants may be considered. In patients with pulmonary disease, buspirone (20–30 mg/day) may be an effective anxiolytic and does not impair respiratory drive. Selective serotonin reuptake inhibitors may be particularly useful in patients with panic attacks (19). Patients with dyspnea and terminal lung diseases (e.g., cancer, COPD, interstitial pulmonary fibrosis) represent a particularly challenging group to manage. The patient and/or the primary caregiver should make plans early on (and not in the setting of acute decompensation) to deal with progressive dyspnea and end-of-life care. Low-dose narcotics (e.g., morphine sulfate) can be extremely useful as a palliative therapy for terminally ill dyspneic patients. Some subjects obtain marked relief from nebulized morphine, which may act via bronchial opioid receptors. The optimal dosing of nebulized morphine is unknown (typically initially 20 mg morphine sulfate in 5 mL normal saline) and requires further study.
Noncardiac Chest Pain
Chest pain is a particularly frightening symptom because of the widespread knowledge and concern about heart disease. However, nonspecific musculoskeletal pain is more common than angina, especially in younger patients (<40 years old). Most patients with chest pain can be evaluated and treated in an ambulatory setting; a few patients require referral to a specialist. The common noncardiac causes of chest pain are discussed in this section.
|
TABLE 59.6 Causes of Chest Pain |
||||||||||||||||||||||||
|
||||||||||||||||||||||||
P.881
Afferent neural impulses responsible for thoracic pain are carried by the sympathetic chain, vagus, and phrenic nerves. Visceral structures, which include the lung, diaphragm, heart, and esophagus, all lie within the thoracic cage and have overlapping innervation. Chest pain arising from these different organs often has similar referral patterns; thus, irritation of the diaphragmatic pleura, diaphragm, or pericardium from either thoracic or abdominal disease causes chest pain that radiates to the shoulder. In addition, patients may have difficulty localizing pain from the deeper anatomic structures within the chest, whereas diseases involving the superficial structures, muscles, and ribs are more easily localized. Because there is no sensory innervation of the lung parenchyma, alveolar or interstitial disease does not cause chest pain unless the pulmonary vasculature, bronchi, or pleura are involved. One study found that enhanced visceral (esophageal) pain perception underlies many cases of unexplained angina-like chest pain (33). Table 59.6 lists causes of chest pain.
Musculoskeletal pain is common in young individuals who increase their exercise abruptly (including patients who acutely hyperventilate as part of an anxiety state. A history of unusual exertion with increased breathing plus tenderness of intercostal muscles usually suffices to make this diagnosis.
Pain caused by tracheitis or tracheobronchitis is a distinctive substernal burning sensation that is precipitated by coughing and most often is associated with viral respiratory infections. This is in contrast to the sharp, stabbing, pleuritic chest pain experienced with pneumonia. The latter is clearly localized to the chest wall and arises from stretching the inflamed parietal pleura during breathing or coughing. Pleurodynia(or epidemic myalgia), characterized by fever, headache, and sudden onset of intense lower thoracic pleuritic pain, usually results from infection with Coxsackie B viruses.
Other causes of chest pain include costochondritis (Tietze syndrome), which is anterior localized pain associated with tenderness over one or more costochondral junctions; herpes zoster, which commonly causes unilateral aching or itching, limited to one dermatome, which may precede by several days the eruption of vesicles; rib fracture orbone metastases, which are more chronic and pleuritic in nature; andcervical spine disease with referred pain to the chest (see Chapter 70). Acute stabbing chest pain can occur with a spontaneous pneumothorax, which occurs primarily in young men or in older patients with obstructive pulmonary disease. Often, a small (<20%) pneumothorax is not accompanied by significant dyspnea in otherwise healthy individuals. Pleuritic chest pain that follows the abrupt onset of dyspnea should raise the suspicion of a PE. The pain associated with pulmonary hypertension is heavy and aching and often similar to that of cardiac ischemia. Gastrointestinal disorders, such as GERD or gastric or duodenal ulcer, usually are distinguished from cardiopulmonary chest pain by their association with eating and by their relief with use of antacids (see Chapters 42 and 43).
Evaluation
Many common causes of noncardiac chest pain can be diagnosed by a thorough history and physical examination. Because discrete anatomic structures must be involved to cause noncardiac chest pain, the physical examination is more useful in the diagnosis of noncardiac chest pain than in the diagnosis of dyspnea and hemoptysis. Inspection of the chest wall may reveal the characteristic unilateral eruption of herpes zoster along a dermatome. Light palpation over the chest wall elicits pain and crepitus from fractured ribs. Mild pressure over the costochondral junctions anteriorly reproduces the pain of Tietze syndrome. (In general, cardiac pain is not worsened by pressure over the chest wall.) In pneumonia or pulmonary infarction, a distinct friction rub is sometimes heard directly over the specific area of chest pain. With pericardial involvement, a friction rub that varies with respiration or with the cardiac cycle usually is present. A thorough abdominal examination is important because diseases involving the abdominal visceral organs can cause referred chest pain that is indistinguishable from that produced by involvement of the thoracic structures. Often, laboratory studies and a chest x-ray will not be necessary for the diagnosis of these common causes of chest pain.
Therapy
The treatment of chest pain requires therapy for the underlying disease process, as well as analgesic drugs for the
P.882
pain itself. Tracheal irritation is limited to the duration of the viral illness but can be treated by cough suppression and bronchodilators (seeCough). Tietze syndrome is treated with standard anti-inflammatory agents and heat. Although the pain of herpes zoster often is severe, it may be controlled with mild narcotics such as codeine 30 to 60 mg every 4 to 6 hours. The chest pain experienced in pulmonary hypertension often does not respond to treatment with nonnarcotic analgesics, and narcotics may be required if the hypertension does not improve with treatment.
Pleuritic pain in patients with pneumonia or PE responds to specific therapy for the inflammatory process. Nevertheless, narcotics may be needed to reduce splinting of the chest wall and thereby prevent atelectasis. Codeine (60 mg every 4 hours) usually is adequate therapy.
Abnormal Chest X-Ray
A chest x-ray must always be compared with prior films because an abnormality that has been stable in size for >2 years almost certainly is benign and may not require further evaluation. In addition, depending on the appearance of the abnormality and the patient's age, the most appropriate plan may be observation with serial chest x-rays.
Specific Patterns Indicative of an Abnormal Chest X-Ray
This section reviews common abnormalities that indicate the presence of pulmonary disease requiring further evaluation.
Air Bronchogram
Normally, bronchi beyond the mainstem division cannot be seen. However, when the alveoli surrounding a bronchus are devoid of air because of consolidation or, less commonly, collapse, an air bronchogram can be seen (Fig. 59.2). The presence of an air bronchogram indicates an alveolar filling process and is most commonly seen in pneumonia and pulmonary edema (cardiogenic and noncardiogenic). However, an air bronchogram is not present in every consolidated lung because bronchi may fill with secretions or exudate. Therefore, its absence is less significant than its presence.
Silhouette Sign
Obliteration on a chest x-ray of the margin of a normally opaque structure in the chest by an abnormal pulmonary density is called thesilhouette sign. If the clinician has knowledge of thoracic anatomy and spatial relations, the silhouette sign can be used to localize abnormalities within the lung parenchyma. Edges of organs that are in contact with parenchymal infiltrates will be obliterated because the normal air interface is eliminated. On the other hand, intrathoracic lesions that are not anatomically contiguous will not interfere with the outlines of nearby structures. For example, obliteration of the right or left cardiac borders, which are anterior, localizes an abnormality to the right middle lobe or to the lingular segment of the left upper lobe, respectively (Fig. 59.3). In contrast, an infiltrate that overlaps, but does not obliterate, the cardiac border is posterior and represents a lower lobe lesion. Lower-lobe abnormalities obliterate diaphragmatic borders and on lateral chest x-ray are seen as an increased density over the vertebrae (spine sign). Obliteration of the left border of the aortic knob, a posterior structure, occurs with lesions in the apical posterior segment of the left upper lobe, whereas obliteration of the border of the ascending aorta, an anterior structure, occurs with lesions in the anterior segment of the right upper lobe.
|
FIGURE 59.2. Air bronchogram. The patient had fever and sputum production. The initial x-ray demonstrates a branching air bronchogram seen behind the heart on the left, which is consistent with a lower-lobe infiltrate. |
Collapse
The collapse (Fig. 59.4) or diminution in volume of the whole lung, a lobe, or a segment of one of the lobes can be an important clue to the presence of asymptomatic pulmonary disease, such as bronchogenic carcinoma, or it may be the cause of a symptom, such as dyspnea in an asthmatic patient with mucous plugging. The primary mechanisms that cause pulmonary collapse are bronchial obstruction from either an intrinsic bronchial mass or an extrinsic or intrinsic stenosis of the bronchus, compression of the lung from a large pleural effusion or from a
P.883
pneumothorax, peripheral bronchial plugging with subsequent pulmonary collapse, and contraction of the lung secondary to chronic inflammatory disease. The signs of collapse are related to anatomic landmarks within the lung and are manifested by displacement of the fissure in the lung, loss of aeration within the pulmonary parenchyma, and crowding of the vascular and bronchial lung markings. Other signs that suggest collapse reflect the secondary effects of loss of lung volume, such as elevation of the diaphragm, shift of the mediastinal structures toward the
P.884
collapsed area, diminution of the size of a hemithorax, compensatory hyperinflation, hilar displacement, and tracheal deviation. These latter signs are much more difficult to interpret in patients with underlying lung disease, in whom many of these signs may exist in the absence of collapse.
|
FIGURE 59.3. Silhouette sign. A: Middle-lobe infiltrate obscures the border of the heart. B: Previous x-ray shown for comparison. |
|
FIGURE 59.4. Collapse. A: Collapse of the right upper lobe and partial collapse of the left lower lobe in a patient complaining of cough and increased sputum. The middle lobe fissure is displaced upward, and there is blunting of the left hemidiaphragm. Note that there is no air bronchogram in either collapsed segment. B: Within 24 hours of initiating aggressive physical therapy, there is resolution of the collapse on the right and almost complete resolution on the left. |
|
FIGURE 59.5. Kerley B lines. Close view of the right lower lung in a patient with congestive heart failure demonstrates horizontal linear lines that run to the edge of the lung. |
Septal Markings
Normally, lung markings reflect vascular patterns within the pulmonary parenchyma and are rarely the result of the bronchi or the lymphatics. Three types of linear shadows represent septal markings within the lung: Kerley A lines, which are thin nonbranching lines several inches long radiating from the hilum that appear to cross blood vessels; Kerley B lines (Fig. 59.5), which are up to 1 inch long and are found at the lateral lung bases on the posteroanterior film or in the retrosternal clear space on the lateral film, radiating from the pleura; and Kerley C lines, which are fine interlacing structures throughout the lung parenchyma that produce a spider-web appearance. The fine linear or reticular pattern represented by Kerley lines are specific for interstitial lung disease. The most common cause of these lines is interstitial edema caused by congestive heart failure.
Common Problems in Patients with Abnormal Chest X-Rays
In this section, three general categories in which the chest x-ray provides the basis for diagnosis and further evaluation are discussed. A general approach is outlined, including the initial evaluation that should be completed by the clinician before referral is made to a pulmonary specialist or thoracic surgeon. Often, this diagnostic evaluation can be completed in an ambulatory setting.
Infiltrates
Infiltrates represent alveolar or interstitial lung disease and seldom show borders except at pleural surfaces. They may be radiographically separated into diffuse and focal infiltrates. Diffuse implies bilateral generalized involvement, if not of the entire lungs then at least of most of both lung fields. Focal implies discrete lesions that may or may not be bilateral but that have intervening normal lung tissue between the localized lesions.
Alveolar
Alveolar infiltrates (Fig. 59.6A) can be recognized by their fluffy margins, their coalescence into rosette formations, their occasional butterfly configuration involving hilar and central lung zones, and the presence of air bronchograms or air alveolograms.
Although there are numerous causes of diffuse interstitial pulmonary disease, there are few causes of diffuse alveolar lung disease. Therefore, the distinction between an alveolar and an interstitial process is important. Unfortunately, it is not always possible to distinguish between the two entities, and there may be a combination of both. Moreover, a disorder that begins as an interstitial process often can merge into an alveolar process, such as interstitial edema in early congestive heart failure progressing to florid pulmonary edema.
Table 59.7 lists the causes of diffuse alveolar pulmonary disease of the lung. The three most common causes—infection, edema, andhemorrhage—are characterized by rapid progression and regression. In contrast, diffuse interstitial disease develops more slowly. Therefore, the time course for the development of pulmonary symptoms and roentgenographic abnormalities is an important aspect in the differential diagnosis of diffuse lung disease.
Interstitial
An interstitial pattern may be primarily linear (reticular) or may consist of multiple, discrete, noncoalescent, round nodules, 1 to 5 mm in diameter (Fig. 59.6B). Although there
P.885
can be a summation effect, these small nodular densities retain a distinct identity in comparison with the larger fluffier infiltrates characteristic of alveolar disease. In certain disease processes, such as tuberculosis, histoplasmosis, or healed viral pneumonia, these nodules may calcify and thus appear more dense. Honeycombing is seen in advanced interstitial lung disease and represents end-stage irreversible scarring with thickened or dilated airways. It can be identified on a chest x-ray as round or oval irregular air spaces that have a reasonably uniform diameter of 1 to 10 mm and are arranged in weblike bunches, thus giving the impression of a honeycomb. Honeycombing is seen in a variety of fibrosing lung diseases, including idiopathic pulmonary fibrosis, sarcoidosis, asbestosis, chronic hypersensitivity pneumonitis, and eosinophilic granuloma. Bilateral interstitial infiltrates in the lower lung fields are common radiographic pattern. The most common diagnoses with this pattern are bronchiectasis, chronic aspiration, collagen vascular diseases, asbestosis, sarcoidosis, or idiopathic pulmonary fibrosis. Localized interstitial processes often represent residues of prior pulmonary infections. If new, they may represent acute processes such as mycoplasma pneumonia or lymphangitic spread of carcinoma.
|
FIGURE 59.6. A: Alveolar patterns. This patient has progressive dyspnea after inhaling fumes from an automobile accident. Compared with B, little air is visible in the infiltrate because fluid is filling the alveoli. B: Interstitial pattern. Bilateral interstitial infiltrates are seen in a patient with progressive dyspnea and fibrosis on biopsy. Compared with A, there is a lacy reticular appearance with accentuation of the air spaces by the fibrosis. |
|||||||||||||||||||||
|
TABLE 59.7 Causes of Diffuse Alveolar Pulmonary Disease |
|||||||||||||||||||||
|
Although interstitial lung disease may be “idiopathic,” this is a diagnosis of exclusion, and the primary goal in the evaluation is to determine whether a treatable disease is present (Table 59.8). The initial medical history, physical examination, and laboratory testing should be oriented toward evaluating the patient for the presence of sarcoidosis,
P.886
pulmonary involvement associated with collagen vascular disease, granulomatous lung infection (mycobacterium, fungus), or pneumoconiosis secondary to occupational exposure (see Chapter 8). Additional evaluation often includes complete pulmonary function testing (spirometry, lung volumes, diffusing capacity, room air arterial blood gas measurement), formal cardiopulmonary exercise testing, chest CT, and fiberoptic bronchoscopy with lavage and transbronchial lung biopsy. The transbronchial lung biopsy often is useful in diagnosing and guiding therapy in patients with sarcoidosis, lymphangitic spread of tumor, infectious disease, collagen vascular pulmonary diseases, and idiopathic pulmonary fibrosis. If the tissue obtained with transbronchial biopsy is inadequate for diagnosis, consultation with a thoracic surgeon for a surgical lung biopsy (video-assisted thoracoscopy or open lung biopsy) may be indicated.
|
TABLE 59.8 Causes of Diffuse Interstitial Pulmonary Disease |
||||||||||||||||||||||||||||||
|
||||||||||||||||||||||||||||||
Slowly Resolving or Recurrent Infiltrates
Endobronchial tumor must be considered in patients who have a slowly resolving pneumonia. Most patients with community-acquired pneumonia respond rapidly to antibiotic therapy, and their chest x-ray will return to baseline over 2 to 4 weeks (see Chapter 33). In one study, radiographic resolution of pneumonia was inversely related to age and was delayed in smokers, patients with multilobar infiltrates, and in hospitalized patients (34). Patients with chronic obstructive lung disease (see Chapter 60) who have difficulty mobilizing bronchial secretions and patients with superimposed congestive heart failure or with necrotizing or multilobar pneumonia may have pulmonary infiltrates that persist for 4 to 5 months. If a patient shows symptomatic improvement and slow but progressive roentgenographic clearing, observation is warranted. If roentgenographic abnormalities persist and the patient fails to show clinical improvement or if the pulmonary infiltrate is found in an asymptomatic individual, further evaluation to exclude lung cancer is warranted (see Chapter 61). In contrast to an older patient, a younger patient with recurrent sinopulmonary infections should be suspected of having an abnormal host defense such as cystic fibrosis or an immunoglobulin deficiency state, which may escape detection until adolescence or early adulthood. In addition, an undiagnosed immunodeficiency such as AIDS should be considered if a patient with a presumed community-acquired pneumonia fails to respond to appropriate antibiotics. Older patients with recurrent pneumonias should be evaluated for the possibility of lung cancer. In evaluating patients with recurrent pneumonia, it is essential to review previous chest x-rays to document the anatomic location and the characteristics of the recurrent pulmonary infiltrate. In general, if a pneumonia recurs in multiple lobes or in pulmonary segments that are unrelated anatomically, bronchogenic carcinoma is unlikely. For example, a recurrent pneumonia that initially involves the right upper lobe and subsequently the right lower lobe is unlikely to be the result of a single endobronchial tumor. In addition to the anatomic location, the time during which recurrent pneumonias occurred should be considered. If the interval is >2 years, malignancy is unlikely.
Apical
Roentgenographic patterns that range from increased pulmonary markings or minor scarring to cystic or cavitary disease in the upper lobes often are interpreted by a radiologist as showing old granulomatous disease or active tuberculous infection. The evaluation of these patients includes questioning about previous tuberculous lung disease (including the type and duration of antituberculous therapy) and assessment of the reactivity of the tuberculin skin test. Comparison with prior chest x-rays is useful because the activity of an infiltrate cannot be determined
P.887
from an isolated chest x-ray. If old chest x-rays demonstrate that no change has occurred, further evaluation may be unnecessary. Chapter 34 discusses the evaluation and treatment of patients with tuberculosis.
|
FIGURE 59.7. Subpulmonic effusion. A: The diaphragm appears to be elevated on the right, representing subpulmonic fluid. B: When the patient is placed in the right lateral decubitus position, fluid layers on the right and tracks in the minor fissure and along the apex and the diaphragm. |
Superior sulcus tumors, usually squamous cell carcinomas, arise in the extreme apex of the lung and may be difficult to distinguish from pleural thickening or old granulomatous disease. Later in the disease course, x-rays may reveal erosion of adjacent ribs or vertebrae by the tumor (see Chapter 61).
Pleural Effusion
Small amounts of free fluid within the pleural space will obliterate the costophrenic or costocardiac angles. Because the density of pleural fluid is greater than the density of the lung, a subpulmonic collection will laterally displace the crest of the diaphragm (Fig. 59.7A). An increased density between the stomach gas bubble and pulmonary tissue may also indicate the presence of fluid within the pleural space. The diagnosis of a large pleural effusion is not difficult because fluid within the pleural space on an upright chest x-ray will form a concave density across the chest cavity; decubitus x-rays will demonstrate the free flowing pleural fluid in the dependent hemithorax (Fig. 59.7B). If the patient is recumbent, the pleural fluid will layer over the entire hemithorax, causing the lung to appear opaque.
In general, when pleural effusion is seen on a chest x-ray, a sample should be obtained for analysis. The obvious exception is a patient who develops acute pulmonary edema associated with a rapidly developing pleural effusion that resolves with therapy for congestive heart failure (see Chapter 66). Table 59.9 lists the causes of pleural effusion. A diagnostic thoracentesis to determine the cause of the effusion generally can be done when the thickness of the fluid between the inner border of the rib and lung is >1 cm on a lateral decubitus x-ray. The fluid should be sent for white blood cell and differential count, total protein concentration, lactate dehydrogenase (LDH) concentration, glucose concentration, Gram stain, cultures (aerobic and anaerobic bacteria, mycobacterium, fungus), pH (if parapneumonic effusion is present), and cytopathology (if malignancy is suspected). Specialized analysis of pleural fluid can help in cases where the diagnosis is not immediately obvious (e.g., amylase in cases of esophageal rupture or pancreatic disease, or adenosine deaminase in suspected tuberculous pleuritis). Most pleural effusions are clear and straw colored, and deviations from this norm may be diagnostically helpful. For example, a bloody effusion suggests tumor and, less commonly, PE with infarction, tuberculosis, or trauma; a lime-green effusion
P.888
suggests tuberculosis; a white milky effusion suggests a chylous exudate; and a viscous fluid with feculent odor strongly suggests an anaerobic empyema.
|
TABLE 59.9 Causes of Pleural Effusion |
|||||||||||||||||||||||||||||||||
|
|||||||||||||||||||||||||||||||||
It is important to classify pleural effusion as either transudative or exudative (35,36). Transudates caused by increased hydrostatic pressure or decreased plasma oncotic pressure have low protein and LDH concentrations. They generally are caused by congestive heart failure, cirrhosis with ascites, or nephrotic syndrome. They do not require further diagnostic evaluation (with respect to the effusion itself), and treatment is directed at the underlying cause. Exudates, as a result of increased protein permeability of the pleural blood vessels, generally are caused by inflammation or tumor infiltration. Exudates are defined by one of the following criteria: pleural fluid protein concentration that is >50% of the concentration of serum protein, pleural fluid LDH concentration that is >60% of the concentration of serum LDH, or pleural fluid LDH concentration that is >67% of the upper limit of normal serum LDH (35,36). Figure 59.8 shows a useful algorithm for working up a pleural effusion. Pleural effusions with protein concentrations >3 g/100 mL nearly always are exudates. If the pleural fluid is defined as exudative but the clinical picture is more consistent with a transudate, the serum–pleural fluid albumin gradient should be measured. If this gradient is >1.2 g/dL, the effusion probably is transudative (36). A cell count and pleural fluid cytologic study should be performed because the presence of polymorphonuclear leukocytes in pleural fluid suggests acute inflammation and infection, whereas >50% lymphocytes suggests tuberculosis or malignancy. Tuberculous pleuritis should be considered in any patient with an exudative pleural effusion and a history of tuberculosis exposure or a positive tuberculin skin reaction. The presence of >5% pleural mesothelial cells makes the diagnosis of tuberculosis unlikely. Low pleural fluid glucose concentration (<50 mg/dL) occurs in infections (parapneumonic effusions, tuberculosis), rheumatoid arthritis, and occasionally in malignancy. Pleural fluid pH determination is important not only in the differential diagnosis (acidic pH is seen with infection, malignancy, severe inflammation, esophageal rupture) but also in the management of parapneumonic pleural effusions. In this setting, pH <7.20 signals the possibility of a complicated parapneumonic effusion, which requires a more aggressive approach (e.g., a drainage procedure) (37). If a pleural effusion is bloody or appears infected, the patient should be hospitalized for further diagnostic studies and therapy (e.g., chest tube drainage, pleural biopsy).
Pneumothorax
The three major types of pneumothorax are spontaneous, iatrogenic, and traumatic. Of the three types, the general practitioner is most commonly faced with a spontaneous pneumothorax either in a young healthy individual or in an older patient with underlying pulmonary disease. In the former, a subpleural apical bleb ruptures into the pleural space, causing collection of varying amounts of air. Patients with spontaneous pneumothoraces tend to be tall, thin, young, male smokers (38). These patients generally are at rest when they first experience symptoms. In the older patient, emphysema with concomitant bullous disease is commonly associated with a pneumothorax. The abrupt onset of pleuritic chest pain or dyspnea is characteristic. Pneumothoraces are diagnosed by demonstrating a visceral pleural line on the chest x-ray.
After the diagnosis of pneumothorax is made, the patient may require observation (ambulatory or inpatient) or insertion of a chest tube. For patients with a pneumothorax >15% of the volume of the hemithorax, needle or catheter aspiration (e.g., using a thoracentesis kit) can obviate the need for hospitalization. This should be performed only by experienced personnel because of the risk for visceral pleural laceration. In general, patients with underlying lung disease should be hospitalized and may require a chest tube because of their limited pulmonary reserve. On the other hand, a healthy patient with a small pneumothorax (<15%) who is not in distress can remain at home and be followed by serial chest x-rays. The rate of reabsorption
P.889
is slow. Assuming approximately 1.25% of the volume is reabsorbed per day (38), a 15% pneumothorax will take approximately 12 days to reabsorb spontaneously. Supplemental oxygen can enhance the rate of reabsorption by increasing the transpulmonary nitrogen gradient.
|
FIGURE 59.8. Algorithm for working up undiagnosed pleural effusions. (From Light RW. Clinical practice. Pleural effusion. N Engl J Med 2002;346:1971, with permission.) |
Solitary Nodule
A solitary nodule is a radiologic finding that always requires evaluation. Chapter 61 provides a detailed discussion of the problem.
P.890
Specific References
For annotated General References and resources related to this chapter, visit http://www.hopkinsbayview.org/PAMreferences.