Pneumothorax, Chylothorax, Hemothorax, and Fibrothorax


A pneumothorax is present when there is air in the pleural space. Pneumothoraces are classified as spontaneous pneumothoraces, which develop without preceding trauma or other obvious cause, and traumatic pneumothoraces, which develop as a result of direct or indirect trauma to the chest, including diagnostic or therapeutic maneuvers ( iatrogenic pneumothoraces ) . Spontaneous pneumothoraces are subclassified as primary or secondary. A primary spontaneous pneumothorax (PSP) presents in an otherwise healthy person without underlying lung disease. A secondary spontaneous pneumothorax (SSP) complicates an underlying lung disease, most commonly chronic obstructive pulmonary disease (COPD).

Most pleural effusions prove to be either an exudate or a transudate according to the criteria provided in Chapter 79 . Occasionally the liquid contents turn out to be chyle, pseudochyle, or blood. This chapter describes the pathogenesis and clinical manifestations of chylothorax, pseudochylothorax, and hemothorax. Fibrothorax, the sequela of chronic organizing pleural disease of any origin, is also considered.

Pathophysiology of Pneumothorax

In normal subjects the pressure in the pleural space is negative with respect to the alveolar pressure during the entire respiratory cycle. The pressure gradient between the alveoli and the pleural space—the transpulmonary pressure—is the result of the inherent elastic recoil of the lung. During spontaneous breathing, the pleural pressure is also negative with respect to the atmospheric pressure. The functional residual capacity, or resting end-expiratory volume of the lung, is the volume at which the inherent outward pull of the chest wall is equal to, but opposite in direction to, the inward pull (recoil) of the lung.

When a communication develops between an alveolus or other intrapulmonary air space and the pleural space, air will flow from the alveolus into the pleural space until there is no longer a pressure difference or until the communication is sealed. Similarly, when a communication develops through the chest wall between the atmosphere and the pleural space, air will enter the pleural space until the pressure gradient is eliminated or the communication is closed. The influence of a pneumothorax on the volume of the hemithorax and the lung is illustrated in Figure 81-1 . In this example, sufficient air has entered the pleural space to elevate the pleural pressure from −5 to −2.5 cm H 2 O, so that the transpulmonary or recoil pressure has decreased from 5 to 2.5 cm H 2 O. The amount of air necessary to effect this change in the pleural pressure can be seen to be equal to 33% of the patient’s vital capacity: most of this pleural air (25% of the vital capacity) is accounted for by air displaced from the lung, and the rest is accounted for by the expansion of the thoracic cavity along its pressure-volume curve (by 8% of the vital capacity). The rise in the pleural pressure also causes a shift of the mediastinum to the contralateral side, an enlarged hemithorax, and a depressed hemidiaphragm. These findings are expected and do not necessarily indicate that a tension pneumothorax is present.

Figure 81-1

Influence of a pneumothorax on the volumes of the lung and hemithorax.

In this illustration the pneumothorax has caused the intrapleural pressure to rise from −5 to −2.5 cm H 2 O. The lung and thoracic cavity move on their respective pressure-volume curves to new volumes for the lung (A) and thoracic cavity (B). Note that the volume of the hemi­thorax becomes larger as the volume of the lung becomes smaller. The changes in volumes of the hemithorax and lung are unequal because of the differences in the slopes of two pressure-volume curves. VC, vital capacity.

(Redrawn from Light RW: Pleural diseases , ed 4, Philadelphia, 2001, Lippincott Williams & Wilkins, p 286.)

The main physiologic consequences of a pneumothorax are a decrease in the vital capacity (as illustrated in Fig. 81-1 ) and the arterial P o 2 . In patients with PSP, the decrease in the vital capacity is usually well tolerated. If the lung function of the patient is abnormal before the development of the pneumothorax, however, the decrease in vital capacity may lead to respiratory insufficiency with alveolar hypoventilation and respiratory acidosis.

Most patients with a pneumothorax have a reduced arterial P o 2 and an increase in the alveolar-arterial oxygen tension difference. In a series of 12 patients with spontaneous pneumothorax, the arterial P o 2 was below 80 mm Hg in 9 (75%) and was below 55 mm Hg in 2 patients, both of whom had SSP.

The reduction in arterial P o 2 appears to be due to the creation of regions of the lung both with low ventilation-perfusion ratios and with absent ventilation (shunt), and occasionally due to alveolar hypoventilation. Norris and coworkers reported that the average right-to-left shunt in their 12 patients with spontaneous pneumothorax was more than 10%. Larger pneumothoraces were associated with greater shunts. When the pneumothorax occupied less than 25% of the hemithorax, the shunt was not increased.

After air is evacuated from the pleural space, the arterial P o 2 usually improves, but the improvement may take several hours. Norris and colleagues evacuated the pleural air from three patients with an initial shunt above 20%; within 90 minutes, the shunt had decreased below 10%, but it nonetheless remained above 5% in all patients. The delay in improvement may be related to the duration of the pneumothorax and the time necessary to expand collapsed alveoli.

Primary Spontaneous Pneumothorax


A study from Britain reported an incidence of spontaneous pneumothorax as 24.0 and 9.8 per 100,000 per year for men and women, respectively, and about one half of the pneumothoraces were PSPs. This extrapolates to an annual incidence of 22,500 in the United States.

Etiologic Factors

PSP, which develops in an otherwise healthy person without known lung disease, is traditionally believed to result from rupture of a subpleural emphysematous bleb often located in the apex of the lung. Blebs can be found in more than 75% of patients undergoing thoracoscopy for treatment of PSP. The pathogenesis of these subpleural blebs and the trigger(s) of their rupture are unclear. Such blebs have been attributed to congenital abnormalities, inflammation of the bronchioles, and disturbances of the collateral ventilation. There is a strong association between smoking and the development of a PSP. When four separate series of patients with PSP are combined, 461 of 505 patients (91%) with PSP were smokers or ex-smokers. The risk for a spontaneous pneumothorax is related to the level of cigarette smoking. In men the relative risk for a pneumothorax is 7 times higher in light smokers (1 to 12 cigarettes per day), 21 times higher in moderate smokers (13 to 22 cigarettes per day), and 102 times higher in heavy smokers (>22 cigarettes per day) than in nonsmokers. It is probable that smoking-induced disease of small airways contributes to the development of the subpleural blebs.

Patients with PSP tend to be taller and thinner than control persons. In one study, military recruits with a pneumothorax were on average 2 inches taller and 25 pounds lighter than the typical military recruit. An increased length of the chest may contribute to the formation of the subpleural blebs. Because pleural pressure falls about 0.20 cm H 2 O per centimeter of vertical height, pleural pressure will be more negative at the apex of the lung in taller persons; accordingly, the alveoli at their lung apex are subjected to a greater mean distending pressure. Over an extended period, this could lead to the formation of subpleural blebs in subjects genetically predisposed to bleb formation.

The risks for PSP can be inherited. The Birt-Hogg-Dubé syndrome is an autosomal dominant condition characterized by an increased incidence of spontaneous pneumothorax (see Fig. 69-10 ), benign skin tumors, and renal tumors. The genetic abnormality resides on chromosome 17p11.2 and involves mutations in the folliculin gene. Approximately 40% of patients with the mutation will have a pneumothorax. Pneumothorax is also more frequent in those with the Marfan syndrome and homocystinuria.

The traditional concept was that air leaked from one bleb supplied by a single airway. More recent data have challenged this concept; a study using inhaled fluorescein suggested that air may leak from more areas than just the blebs, raising the possibility of “pleural porosity,” in which the air leaks from multiple pores on the visceral pleura. Endobronchial one-way valve studies also found that occlusion of multiple segmental bronchi are necessary to stop air leaks. This suggests that air is leaking either via multiple sites, via collateral ventilation, or both.

Clinical Manifestations

PSP most commonly happens to those in their early 20s, and rarely after age 40. PSP usually develops while the patient is at rest, and rarely during heavy exercise.

Chest pain and dyspnea are the main symptoms. The chest pain, often acute, is usually localized to the side of the pneumothorax. Rarely, Horner syndrome can develop, probably from traction on the sympathetic ganglion associated with the mediastinal shift.

Vital signs are usually normal, with the exception of a moderate tachycardia. A tension pneumothorax should be suspected if tachycardia (>140 beats/min), hypotension, cyanosis, or electromechanical dissociation is present. The side with pneumothorax may be larger than the contralateral side and move less during the respiratory cycle. Tactile fremitus is absent, the percussion note is hyperresonant, and the breath sounds are absent or reduced on the affected side. In those with a right-sided pneumothorax, the lower edge of the liver may be shifted inferiorly. With a large pneumothorax, the trachea may be shifted toward the contralateral side.


The diagnosis is usually suggested by the clinical history and physical examination and established by demonstrating a pleural line on the chest radiograph ( Fig. 81-2 ). A visceral pleural line can be distinguished from other lines such as skinfolds by the following criteria. This line is defined by air density on both sides of the line, whereas a skinfold is really an edge without air density on either side. A pleural line should generally be sharp and well defined, whereas a skinfold is often poorly defined, at least on one side. A pleural line can be followed continuously, roughly paralleling the chest wall inner contour so that, in the upright patient, it caps the apex of the lung (assuming no adhesions) and often tapers toward the lung base; a skinfold usually is not continuous, often fades at both ends, and often does not follow an anatomically sensible configuration. Finally, a pleural line should not be crossed by lung vessels, whereas a skinfold can show lung vessels beyond its edge ( eFig. 81-1 ).

Figure 81-2

Distinguishing a pneumothorax from a skinfold.

A, Chest radiograph of a pneumothorax. The pleural line has lucency on either side, representing air in the pleural space on one side of the line and air in the lung on the other. The line is sharply demarcated and can be traced along its course ( lower arrow ) . No blood vessels can be seen beyond the superior ( upper arrow ) and lateral ( middle arrow ) extent of the line. B, Chest radiograph of a skinfold ( arrows ) that could be mistaken for a pneumothorax. The border is more of an edge, with lucency on only one side. The edge is poorly defined and cannot be followed continuously ( lower arrow ) . Blood vessels can be traced beyond the border of the fold ( arrowhead ) .

In doubtful cases, lateral decubitus films (with the affected side up, eFig. 81-2 ), ultrasonography, or computed tomography (CT) (see Fig. 76-7 ) may facilitate the diagnosis. Expiratory films (see Fig. 18-5 ) are only slightly more sensitive than inspiratory films in detecting pneumothoraces and are not routinely recommended. A small pleural effusion is associated with a PSP in approximately 15% of cases and is manifested radiographically as an air-fluid level ( eFig. 81-3 ; see eFig. 81-2 ). The pleural fluid frequently contains a large percentage of eosinophils. On rare occasions, spontaneous pneumothorax is complicated by brisk pleural bleed, producing a hemopneumothorax. Emergency surgery is indicated if the patient is hemodynamically compromised.

When one is managing a patient with a pneumothorax, the amount of lung collapse should be estimated. One can first measure the lung and hemithorax “diameter”: the distance from the lung root to the visceral pleural line (lung diameter) or to the chest wall (hemithorax diameter). Because the volumes of the lung and the hemithorax are roughly proportional to the cube of their diameters, one can estimate the degree of collapse (percent pneumothorax, or PTX%) by measuring average diameters of the lung and the hemithorax, cubing these diameters, and using the following equation, known as the Light index :

<SPAN role=presentation tabIndex=0 id=MathJax-Element-1-Frame class=MathJax style="POSITION: relative" data-mathml='PTX%=100%×[1−(diameter of the lung/diameter of the hemithorax)3]’>PTX%=100%×[1(diameter of the lung/diameter of the hemithorax)3]PTX%=100%×[1−(diameter of the lung/diameter of the hemithorax)3]
PTX % = 100 % × [ 1 − ( diameter of the lung / diameter of the hemithorax ) 3 ]
There is an excellent correlation ( r = 0.84) between the Light index and the amount of air that can be aspirated from a pneumothorax.

Recurrence Rates

Following a PSP, a patient is at risk for recurrence particularly in the months immediately after the first episode. One study followed 153 PSP patients for a mean of 54 months and reported that 39% had a recurrent ipsilateral pneumothorax, most within the first year. Interestingly, 15% also developed a pneumothorax on the contralateral side.

There have been several attempts to predict who will develop recurrent pneumothorax. Patients who are tall, patients with a low body mass index, and those who continue to smoke are more likely to have a recurrence. Patients who have blebs or bullae or both on high-resolution CT scan are also more likely to have a recurrence. Once a patient has had one recurrence, the risk for another recurrence increases to more than 50% if no measure is taken to prevent the recurrence.


There are two goals to managing a patient with a PSP: to rid the pleural space of air and to prevent recurrence. Treatment options include observation, supplemental oxygen, simple aspiration, simple tube thoracostomy, tube thoracostomy with instillation of a pleurodesing agent, thoracoscopy with oversewing of the blebs and pleurodesis, and open thoracotomy. When selecting treatment for any given patient, it should be remembered that PSP is mainly a nuisance and rarely life-threatening. Published guidelines, such as from the American College of Chest Physicians and the British Thoracic Society, have emphasized the lack of controlled studies on the treatment of pneumothorax.


Once the communication between the alveoli and the pleural space is eliminated, the residual air in the pleural space will be gradually reabsorbed, albeit slowly. Kircher and Swartzel estimated that 1.25% of the volume of the hemithorax is absorbed each 24 hours. For a patient with a 20% pneumothorax, it will take 16 days for pleural air to be absorbed spontaneously. Supplemental oxygen may increase the rate of pleural air absorption (see the next section).

Supplemental Oxygen

In pneumothorax, gases move in and out of the pleural space from the capillaries in the visceral and parietal pleura. The movement of each gas depends on the gradient between its partial pressure in the capillaries and in the pleural space, the blood flow per unit surface area available for gas exchange, and the solubility of each gas in the surrounding tissues. Normally the sum of all the partial pressures in the capillary blood with a patient breathing room air is about 706 mm Hg (P n 2 , 573; P h 2 o , 47; P co 2 , 46; and P o 2 , 40 mm Hg). If it is assumed that the pleural pressure is approximately 0 when there is a pneumothorax, then the net gradient for gas absorption is only 54 mm Hg (760 − 706). If the patient is placed on 100% oxygen, however, the sum of all the partial pressures in the capillary blood will probably fall below 200 mm Hg (the P n 2 will approach 0, whereas the P o 2 will remain <100 mm Hg). The net gradient for gas absorption will exceed 550 mm Hg, or be 10 times greater than it was with the patient breathing room air.

Administration of humidified 100% oxygen to rabbits with experimentally induced pneumothoraces increased the rate of absorption by about sixfold. In subsequent studies in patients with a spontaneous pneumothorax, administration of high concentrations of supplemental oxygen increased the rate of absorption by fourfold. It is recommended that hospitalized patients with any type of pneumothorax who are not subjected to aspiration or tube thoracostomy be treated with high-flow supplemental oxygen.

Simple Aspiration

The initial treatment for most patients with PSP greater than 15% of the volume of the hemithorax should be simple aspiration. This procedure is successful in about 60% of patients with PSP. If successful, simple aspiration avoids hospitalization, and there is less pain from the smaller tube. The recurrence rates appear to be similar after simple aspiration and after tube thoracostomy.

With this procedure, a relatively small needle (≈16 gauge) with an internal polyethylene catheter is inserted into the second anterior intercostal space at the mid-clavicular line after local anesthesia. An alternative site is selected if the pneumothorax is loculated or if adhesions are present. After the needle is inserted, it is extracted, leaving the catheter tip in the pleural space. A three-way stopcock and a 60-mL syringe are attached to the catheter. Air is manually withdrawn until no more can be aspirated. The catheter is then occluded for several hours. If the chest radiograph confirms that there has been no recurrence, the catheter is removed and the patient is discharged. Alternatively, the patient can be observed overnight or can be discharged with a Heimlich one-way valve attached to the catheter. If the total volume of air aspirated exceeds 4 L and no resistance has been felt, it is assumed that there has been no reexpansion, and alternative procedures are initiated.

Patients with their first PSP should be managed initially with simple aspiration on an outpatient basis. Consideration can be given to overnight observation in hospital for those who reside a long distance from the hospital. Patients should return in 24 to 72 hours for a follow-up chest radiograph. If aspiration is unsuccessful, then either thoracoscopy or tube thoracostomy should be undertaken. In one study the intrapleural administration of 300 mg of minocycline after a successful pneumothorax aspiration decreased the incidence of subsequent PSP recurrence from 49% to 29%. Simple aspiration is not recommended for patients with secondary spontaneous pneumothoraces or with recurrent PSPs.

Tube Thoracostomy

For the past several decades, most patients with PSP have been managed initially with tube thoracostomy. It is recommended if simple aspiration proves ineffective and thoracoscopy is not readily available. It rapidly results in the reexpansion of the underlying lung and does not require prolonged hospitalization. In one series of 81 patients treated with tube thoracostomy, the average duration of hospitalization was only 4 days (range, 3 to 6 days). Only 3 patients (4%) had persistent air leaks after several days of chest tube drainage.

Tube thoracostomy with relatively small tubes (8 to 16 French) or pigtail catheters (8 to 10 French) appear to be as effective as larger tubes. It is advisable to use a water-seal chamber and to avoid suction for the first 24 hours of tube thoracostomy to reduce the risks of reexpansion pulmonary edema.

After the lung has reexpanded and the air leak has ceased for 24 hours, the chest tube can be removed. Air leaks are present when there is bubbling through the water-seal chamber of the drainage system. If there is no bubbling on quiet respiration, the patient should be asked to cough. The absence of bubbling indicates there is no air leak. Whether clamping the tube to observe for the recurrence of pneumothorax before its removal is beneficial remains controversial. If there is a persistent air leak after 72 hours, consideration should be given to application of a blood patch (see later). If air leak persists over 72 hours after tube thoracostomy, thoracoscopy should be considered.

Tube Thoracostomy with Instillation of a Pleurodesing Agent

Efforts have been made to diminish the recurrence rates of PSP by injecting pleurodesing agents into the pleural space at the time of the initial episode. Thoracoscopy with stapling of blebs and pleural abrasion reduces the recurrence rate to less than 5% and is the preferred option. Otherwise, pleurodesis with talc slurry or doxycycline can reduce the recurrence rates from approximately 40% to 25% ( Fig. 81-3 ). Large-particle-size talc should be used to minimize the risks for acute lung inflammation. A tetracycline derivative (e.g., doxycycline or minocycline) is preferred by some (see Chapter 82 ). Bleomycin is not recommended because it does not produce a pleurodesis in animals with a normal pleura and has not been studied with pneumothorax in humans.

Figure 81-3

Pneumothorax recurrence.

The probability of recurrent pneumothorax after assignment to the tetracycline group was lower than in the control group in the Veterans Administration cooperative study on spontaneous pneumothorax.

(Redrawn from Light RW, O’Hara VS, Moritz TE, et al: Intrapleural tetracycline for the prevention of recurrent spontaneous pneumothorax. JAMA 264:2224–2230, 1990.)

Autologous Blood Patch for Persistent Air Leak

The presence of a persistent air leak leads to prolongation of the hospitalization of patients with spontaneous pneumothorax. One inexpensive, noninvasive treatment for prolonged air leaks is the application of an autologous blood patch. Venous blood (50 to 100 mL) is drawn and promptly instilled intrapleurally, without anticoagulation, through a chest tube. The chest tube should not be clamped because, with an ongoing air leak, there would be risk for tension pneumothorax. Elevating the tube (about 60 cm) keeps the blood within the pleural cavity without clamping. According to a review of the literature, the blood patch technique stopped the air leaks in 91.7% of 107 patients. Whether blood patches prevent recurrence is unknown, although, in a small study ( n = 32), no recurrences was found after a follow-up period of 12 to 48 months.


Video-assisted thoracic surgery (VATS) is the treatment of choice of PSP if aspiration fails or if pneumothorax recurs (see Chapter 24 ). In a meta-analysis of 27 studies, the recurrence rate was only 5.4%. During VATS, attempts are made to eliminate the blebs (e.g., by endo-stapling or suturing ) and to create a pleurodesis. Pleurodesis is probably best induced by pleural abrasion, but some recommend apical pleurectomy. Bleb stapling without pleurodesis has a higher recurrence rate than bleb stapling with pleurodesis and is not recommended.

The most common complication after VATS for the treatment of PSP is a persistent air leak, seen in less than 5% of cases. VATS offers shorter hospitalization stays (median, 3 days) and less morbidity when compared with thoracotomy.

Medical thoracoscopy with talc insufflations (without stapling of blebs and pleural abrasion) has been tried. In one study of 59 patients, this approach had a recurrence rate of 5% during a follow-up period of 5 years. Controlled studies are needed to determine whether thoracoscopy is as effective as VATS.

Open Thoracotomy

When VATS is not available, open thoracotomy with oversewing of the blebs and pleural abrasion is a reasonable alternative. A transaxillary mini-thoracotomy can minimize the trauma and the length of the scar. In a meta-analysis the recurrence rate after open thoracotomy was only 1.1% and significantly better than that for VATS. Various methods have been used for scarification of the pleural surfaces, ranging from visceral and parietal pleurectomy to pleural abrasion with dry sponges. Because all appear to be similarly effective, pleural abrasion with dry gauze is recommended because it is less traumatic than pleurectomy and does not appear to interfere with a subsequent thoracotomy.

Summary of Treatment

Most patients with their first PSP should be managed initially with simple aspiration. Patients who fail aspiration should be managed with tube thoracostomy. If the air leak persists, VATS should be performed promptly, with endo-stapling of blebs and pleural abrasion to create a pleurodesis. When thoracoscopy is unavailable, a blood patch or an attempt at pleurodesis with a tetracycline derivative or talc can be tried. Thoracotomy is effective but involves longer hospitalizations and more postoperative morbidity. Patients with recurrent PSP should undergo VATS or thoracotomy.

Secondary Spontaneous Pneumothorax

SSP, which develops in a patient with an underlying lung disease, is thus more serious than a primary pneumothorax because it further decreases the pulmonary function of a patient whose reserve is already diminished. In addition, the presence of the underlying lung disease makes diagnosis and management of the pneumothorax more difficult.


The incidence of SSP is similar to that of PSP. There are an estimated 15,000 new cases of SSP annually in the United States. Men older than age 75 have the highest per capita rate of pneumothorax, 60 per 100,000 per year.

Etiologic Factors

COPD is the most common underlying disease in patients with SSP, although almost every lung disease has been associated with SSP. In one series of 505 patients with SSP, 348 patients had COPD, 93 had tumors, 26 had sarcoidosis, 9 had tuberculosis, 16 had other pulmonary infections, and 13 had miscellaneous diseases. Among patients with COPD, the incidence of SSP increases with progressive severity of the COPD. In the Veterans Administration cooperative study on pneumothorax, 27% of the 229 participants had an FEV 1 /FVC ratio below 0.40. One of the more common causes of SSP is Pneumocystis jirovecii infection in patients with acquired immunodeficiency syndrome (AIDS) (see eFig. 90-15 ). There is also a high incidence in cystic fibrosis. Of the more than 28,000 patients on the Cystic Fibrosis Foundation registry, the incidence of pneumothorax was 3.4%. Patients with lymphangioleiomyomatosis ( Fig. 81-4 ; see Chapter 69 and eFig. 69-7 ) and Langerhans cell histiocytosis (formerly histiocytosis X ) (see Chapters 54 and 63 ) also have a high incidence of spontaneous pneumothorax.

Figure 81-4

Chest CT scan of a patient with lymphangioleiomyomatosis.

Note the numerous thin-walled cysts, rounded and relatively uniform in shape, and a pneumothorax on the left.

(Courtesy Dr. Lisete Teixeira, University of Sao Paulo, Brazil.)

Clinical Manifestations

In general, the clinical features (dyspnea, chest pain, cyanosis, and hypotension) associated with SSP are more severe than those associated with PSP. In the Veterans Administration cooperative study, the mortality rate from pneumothorax was 1%.


The possibility of a pneumothorax should be considered in every patient with COPD who has a sudden increase in breathlessness and/or chest pain. Physical examination is not always helpful because patients with underlying disease may already have hyperexpanded lungs, decreased tactile fremitus, a hyperresonant percussion note, and distant breath sounds over both lung fields.

The demonstration of a pleural line on the chest radiograph can be difficult in radiographs of COPD patients because of their emphysematous lungs. Pneumothorax is therefore easily overlooked, particularly when the radiograph is overexposed. In patients with lung disease, the radiographic appearance of the pneumothorax can be altered by the underlying abnormalities. Areas of normal lung collapse more evenly and completely than do diseased areas with large bullae or severe emphysema, which have decreased elastic recoil and may trap gas.

It is important to distinguish a SSP from a thin-walled bulla in patients with COPD. The apparent pleural line with a large bulla is usually concave toward the lateral chest wall because it represents the medial border of the bulla, whereas the pleural line with a pneumothorax is usually convex toward the lateral chest wall. CT scanning may be necessary to diagnose pneumothorax in a patient with COPD and to discriminate between pneumothorax and bullae ( Fig. 81-5 ).

Figure 81-5

Bullae resembling pneumothorax.

A, Frontal chest radiograph of a patient with cystic fibrosis and extensive bullous disease of the right apex that was initially mistaken for a pneumothorax. Note lucent area ( arrows ) within the right upper lung field. B, CT scan confirms the presence of bullae ( * ) and excludes pneumothorax.

Recurrence Rates

The recurrence rates for SSPs are higher than those for PSP s whether or not measures are taken to prevent a recurrence. In observation studies over a 3- to 5-year period, the recurrence rates were shown to be higher in those with SSP (≈45%) than in patients with PSP (≈30%).


Urgent evaluation is indicated for any patient suspected of having a SSP, because death has been reported before a chest tube can be inserted. Such deaths were reported in 3 of 57 patients (5%) with pneumothorax secondary to COPD and in 3 of 15 patients (20%) with cystic fibrosis. The high immediate mortality rate emphasizes the need for prevention of recurrences.

If one has time, the initial treatment for nearly every patient with a SSP should be tube thoracostomy. Simple aspiration is not recommended except in an emergency (see later) because it frequently is ineffective and cannot prevent recurrence. The evacuation of even a small pneumothorax can rapidly improve symptoms. In patients receiving mechanical ventilation, immediate chest tube drainage is needed because the pneumothorax is likely to enlarge.

In SSP the lung is more difficult to expand and air leaks persist longer when compared with PSP. In SSP caused by COPD, the median time for lung expansion is 5 days (versus 1 day for PSP), and more patients require prolonged chest tube drainage. After 7 days of tube thoracostomy drainage, the lung remains unexpanded or the air leak persists in about 20% of patients with SSP.

Following thoracostomy, most patients with SSP should be considered for thoracoscopy. Thoracoscopy should certainly be performed in patients with a persistent air leak or an unexpanded lung after 72 hours of tube thoracostomy. If the lung expands and the air leak ceases within the first 72 hours, an attempt should be made to prevent a recurrence. For preventing recurrences, thoracoscopy is superior to chemical pleurodesis (recurrence rates of about 5% versus 20%, respectively). If thoracoscopy is unavailable/inappropriate, chemical pleurodesis should be performed (as discussed earlier) to prevent a recurrent pneumothorax. A blood patch can be considered if thoracoscopy is unavailable and the patient has a persistent air leak. Thoracotomy is another alternative.

One consideration for patients with SSP is the effect that a pleurodesing agent might have on future lung transplantation. In 1998, however, a consensus conference statement on lung transplantation in cystic fibrosis stated that pleurodesis is not a contraindication to lung transplantation. However, pleurodesis is likely to make future lung transplantation more difficult, and consultation with a transplant surgeon is advisable.

Pneumothorax Secondary to Pneumocystis in Patients with AIDS

Patients with AIDS and P. jirovecii infection have a relatively high incidence of spontaneous pneumothorax. Approximately 5% of patients with AIDS who receive prophylactic pentamidine will have a spontaneous pneumothorax. Most patients with AIDS who have a spontaneous pneumothorax have a history of P. jirovecii infection, many are taking prophylactic pentamidine, and most have a recurrence of P. jirovecii infection. The presence of multiple subpleural lung cysts or cavities, often associated with subpleural necrosis, may explain the high incidence of spontaneous pneumothorax (see eFig. 90-15 ). After having one pneumothorax, the patient is likely to experience a contralateral pneumothorax. It should be noted that iatrogenic pneumothoraces, particularly those related to mechanical ventilation or pulmonary procedures, are also common in patients with AIDS.

Perhaps due to the necrotic lung surrounding the ruptured cavity, the spontaneous pneumothorax associated with AIDS and P. jirovecii infection is notoriously difficult to treat. In one report of 20 patients, the median duration of tube thoracostomy was 20 days; 11 patients underwent pleurodesis, whereas 5 patients had thoracotomy.

Patients with AIDS and a spontaneous pneumothorax should be treated with tube thoracostomy. If the air leak persists for more than a few days, there are two options: a Heimlich valve or VATS. There is no literature on use of the blood patch technique in managing these patients. In general, the Heimlich valve is preferred because the patient can be managed as an outpatient.

If the air flow is too high for the Heimlich valve to maintain lung inflation, VATS can be performed. Wait reported that 30 of 32 patients with AIDS and spontaneous pneumothorax were managed successfully with talc insufflation without endo-stapling. If operative intervention is planned, it should be carried out early to avoid prolonged hospitalization and morbidity.

Pneumothorax Secondary to Tuberculosis

Between 1% and 3% of patients hospitalized for pulmonary tuberculosis will have a pneumothorax ( eFig. 81-4 ). All such patients should be treated initially with tube thoracostomy. In one series of 28 patients, 7 of the 11 patients treated by observation or repeated pleural aspiration died, compared with 1 of the 17 patients treated with tube thoracostomy (64% versus 6%, respectively). Antituberculous chemotherapy should be given concomitantly with the tube thoracostomy. Thoracoscopy or thoracotomy should be considered if the lung remains unexpanded or if there is an air leak after 7 days.

Iatrogenic Pneumothorax

Iatrogenic pneumothoraces are probably more common than PSPs and SSPs combined. Currently, the leading cause of iatrogenic pneumothorax is transthoracic needle aspiration (see Fig. 19-10 ). The incidence of iatrogenic pneumothorax with this procedure is about 25%, and about 10% of the patients with pneumothorax receive tube thoracostomy. This procedure is more likely to result in a pneumothorax if the patient has COPD, if the lesion is deep within the lung, or if the angle of the needle route is wide (which may correlate with an increased number of passes). Various maneuvers such as positioning the patient with the biopsied side down or using the blood patch technique have not consistently proved useful in diminishing the incidence of pneumothorax. One paper did demonstrate a significant reduction in pneumothoraces if the puncture access was sealed by instillation of NaCl 0.9% solution during extraction of the guide needle.

Mechanical ventilation is another risk factor. In one early series of 553 patients undergoing mechanical ventilation, 4% developed a pneumothorax. The frequency of the pneumothorax was increased if the patient had aspiration pneumonia (37%), was treated with positive end-expiratory pressure (15%), had intubation of the right main-stem bronchus (13%), or had COPD (8%). In another study, the incidence of pneumothorax in 725 patients with acute respiratory distress syndrome was 6.9%. In patients with acute respiratory distress syndrome, the incidence of barotrauma is higher if the ventilator plateau pressure exceeds 35 cm H 2 O or if the lung compliance is below 30 mL/cm H 2 O. In patients with asthma, barotrauma rates were high and mortality was similarly high; now both have been reduced by low-pressure permissive hypercapnic ventilatory strategies.

Other common causes of iatrogenic pneumothorax (with approximate incidences) are thoracentesis (2.5%), pleural biopsy (8%), and transbronchial lung biopsy (6%) ( eFig. 81-5 ). Iatrogenic pneumothorax also frequently complicates cardiopulmonary resuscitation (CPR). Other procedures associated with the development of an iatrogenic pneumothorax include subclavian ( eFig. 81-6 ) or internal jugular vein catheterization, tracheostomy, intercostal nerve block, mediastinoscopy, liver biopsy, and the insertion of small nasogastric tubes. Radiofrequency ablation is increasingly used to treat lung tumors and carries a high incidence of pneumothorax (≈30%). In a series of 137 procedures, 27 resulted in symptomatic pneumothoraces requiring chest tube drainage. In another series, the risk for pneumothorax was associated with the number of tumors ablated, electrode positions, and the electrode trajectory through aerated lung.


Iatrogenic pneumothorax should be suspected in any patient treated by mechanical ventilation whose clinical condition suddenly deteriorates. A sensitive indicator of the development of a pneumothorax in such patients is increased peak and plateau pressures if the patient is on a volume-cycled ventilator or a decreased tidal volume if the patient is on a pressure-cycled ventilator.

The diagnosis should also be suspected in any patient who becomes more dyspneic after an intervention procedure that has been associated with development of a pneumothorax. Importantly symptoms from the pneumothorax may not be evident for 24 hours or longer after the procedure. The diagnosis needs confirmation by ultrasound or radiographic imaging. In patients with extensive pulmonary opacities, there may be little evidence of lung collapse, but the air in the pleural space may instead be indicated by the “deep sulcus” sign ( Fig. 81-6 ).

Figure 81-6

Appearance of pneumothorax in a supine patient.

Detailed views of the right costophrenic angle from successive supine chest radiographs taken before ( A ) and after ( B ) the development of a pneumothorax. Both illustrations show diffuse infiltration in the right lower lung field. Note change in costophrenic angles ( arrows ) caused by air in the pleural space between the diaphragm and chest wall: the “deep sulcus” sign. Air in the anterior pleural space can also be recognized by an air-tissue interface ( arrowheads ) .


The treatment of iatrogenic pneumothorax differs from that of spontaneous pneumothorax in that preventing recurrence is not an issue. If the patient has minimal/no symptoms and the pneumothorax is small (e.g., <15% of the volume of the hemithorax), the patient can be observed. The administration of supplemental oxygen increases resolution rate (see earlier section). Otherwise, the procedure of choice is simple aspiration with a plastic catheter, as described for PSP. If simple aspiration fails, tube thoracostomy should be performed. Small-bore chest tubes with Heimlich valves are quite effective in this situation.

The management of patients with an iatrogenic pneumothorax secondary to positive-pressure mechanical ventilation should include urgent tube thoracostomy because tension pneumothorax can easily develop when the positive pressure forces air into the pleural space under high pressure until it exceeds atmospheric pressure. If the patient continues to receive mechanical ventilation, the chest tube should be left in place for at least 48 hours after the air leak stops. At times, a bronchopleural fistula is so large that a high percentage of the total inspired volume exits through the chest tube. The air exiting through the chest tube still provides effective ventilation, however, because it contains levels of carbon dioxide similar to those of exhaled gas. Two studies in adults have demonstrated that high-frequency ventilation does not consistently improve gas exchange or decrease air leak through the fistula.

Traumatic (Noniatrogenic) Pneumothorax

The incidence of pneumothorax after blunt trauma depends on the severity of the trauma. The incidence of pneumothorax exceeds 35% in some series.


A traumatic pneumothorax can result from either penetrating or nonpenetrating chest trauma (see Fig. 76-4 , Fig. 76-7 , Fig. 76-8 , Fig. 76-10 and eFigs. 76-10 and 84-2 ). With penetrating chest trauma, the wound allows air to enter the pleural space via the chest wall or via the visceral pleura from the tracheobronchial tree. With nonpenetrating trauma, a pneumothorax may develop if the visceral pleura is lacerated secondary to a rib fracture or dislocation. In the majority of patients with pneumothorax secondary to nonpenetrating trauma, however, there are no associated rib fractures. It is thought that the sudden chest compression abruptly increases the alveolar pressure, which may cause alveolar rupture. Air then enters the interstitial space and dissects toward either the visceral pleura or the mediastinum. A pneumothorax develops when either the visceral or the mediastinal pleura ruptures.

Diagnosis and Treatment

The diagnosis of pneumothorax is made by ultrasonography, chest radiography, or CT. A pneumothorax detectable on CT, but not on the chest radiograph, is called an occult pneumothorax and accounts for about 40% of traumatic pneumothoraces (see Fig. 76-7 ). In recent years, ultrasonography (see Chapter 20 ) performed by emergency department physicians or surgeons has been used increasingly to determine whether a pneumothorax is present. Ultrasonography is more sensitive than supine chest radiographs in identifying pneumothoraces, but there are also false positives, particularly in patients with underlying lung diseases, especially COPD.

Most traumatic pneumothoraces should be treated initially with tube thoracostomy. If the patient has an occult pneumothorax or if the distance between the lung and chest wall does not exceed 1.5 cm, however, tube thoracostomy is probably not indicated unless the patient is receiving mechanical ventilation. In one series, 333 patients with pneumothoraces less than 1.5 cm were initially managed without chest tubes, and only 33 (10%) subsequently required tube thoracostomy. When traumatic pneumothoraces are treated with tube thoracostomy, the lung usually expands and the air leak ceases within 24 hours. If the leak persists for more than a few days, consideration should be given to performing thoracoscopy to identify and repair the site of the air leak.

Immediate thoracotomy is indicated for traumatic pneumothorax in two uncommon scenarios. The first is fracture of the trachea or a major bronchus (see eFig. 76-10 and 84-2 ), which usually presents together with an anterior/lateral fracture of one or more of the first three ribs and is associated with at least some hemoptysis. Fiberoptic bronchoscopy to search for a bronchial tear followed by its surgical repair usually restores full function of the distal lung.

The second diagnosis to consider is traumatic rupture of the esophagus, which is almost always accompanied by a hydropneumothorax (see Fig. 84-7 ). A reliable screening test for esophageal rupture is measurement of the pleural fluid amylase concentration. If this level is elevated, contrast radiographic studies of the esophagus should be performed. It is important to establish the diagnosis of esophageal rupture expeditiously because the mortality approaches 100% if surgical treatment is not performed promptly.

Air Travel and Pneumothorax

For patients who have suffered a pneumothorax, the Aerospace Medical Association suggests that flying be allowed 2 to 3 weeks after radiologic resolution of the pneumothorax. One study of 12 patients with traumatic pneumothorax reported that the 10 patients who waited at least 14 days before travel were all asymptomatic in flight, whereas 1 of 2 patients who flew earlier than 14 days developed respiratory distress in flight with symptoms suggesting a recurrent pneumothorax. However, a recent article suggested that patients who do not have a pneumothorax 48 hours after discharge can travel safely.

Neonatal Pneumothorax

A spontaneous pneumothorax is present shortly after birth in 1% to 2% of all infants, and the pneumothorax is symptomatic in approximately half of these. It is twice as common in male infants. Affected infants are usually full- or post-term and have a history of fetal distress requiring resuscitation or a difficult delivery with evidence of aspiration of meconium, blood, or mucus.

There is a high incidence (19% in a series of 295 infants) of pneumothorax in infants with neonatal respiratory distress syndrome ( eFig. 81-7 ). Pneumothorax developed in 29% of those requiring intermittent positive-pressure ventilation with positive end-expiratory pressure, in 11% of those requiring continuous positive airway pressure, and in only 4% of those not requiring respiratory assistance. In a series of more than 20,000 babies with birth weights between 401 and 1500 g (0.9 to 3.3 pounds) born in 1999, the incidence of pneumothorax was 6.3%, but, in those babies weighing less than 750 g (1.7 pounds), the incidence was 15%.


The development of a spontaneous neonatal pneumothorax is related to the mechanical problems of expanding the lung for the first time. During the first few breaths of life, the transpulmonary pressures average 40 cm H 2 O, with occasional pressures as high as 100 cm H 2 O. At birth the alveoli usually open in rapid sequence, but, if bronchial obstruction is present from the aspiration of blood, meconium, or mucus, high transpulmonary pressures may lead to rupture of the lung. A transpulmonary pressure of 60 cm H 2 O ruptures an adult lung, whereas a transpulmonary pressure of 45 cm H 2 O ruptures a neonatal rabbit lung.

Clinical Manifestations

With neonatal spontaneous pneumothorax, the signs vary from none to severe acute respiratory distress, depending on the size of the pneumothorax. With a small pneumothorax, there may be mild apneic spells with some irritability or restlessness. With a large pneumothorax, there may be severe respiratory distress with marked tachypnea, grunting, retractions, and cyanosis. In newborn babies the detection of pneumothorax by physical examination is often difficult because breath sounds are widely transmitted in the small neonatal thorax from the contralateral lung. The most reliable sign is a shift of the apical heart impulse away from the side of the pneumothorax. The diagnosis requires radiographic confirmation (see eFig. 81-7 ).

The development of a pneumothorax in a patient with neonatal respiratory distress syndrome is usually heralded by a change in the vital signs. In one series, cardiac arrest marked the development of the pneumothorax in 12 of the 49 patients; most of the other patients had a decrease in the blood pressure, pulse, or respiratory rate. In another series, however, the earliest signs were an increase in the blood pressure, heart rate, or pulse pressure. A pneumothorax in an infant with the neonatal respiratory distress syndrome is associated with mortalities exceeding 60% in some studies.


An infant with a spontaneous pneumothorax who is asymptomatic or mildly symptomatic can be observed closely because the pneumothorax usually resolves within a few days. It is necessary to observe the patient closely in case the pneumothorax enlarges or a tension pneumothorax develops. Supplemental oxygen can hasten the resolution of the pneumothorax, but it should be administered with care because of the dangers of retrolental fibroplasia. Tube thoracostomy should be performed on any neonate who is more than mildly symptomatic. In one series of 76 infants with spontaneous pneumothorax, respiratory failure necessitated mechanical ventilation in 18, and pulmonary hypertension requiring either nitric oxide or extracorporeal membrane oxygenation developed in 7. However, all the patients had complete resolution of their pulmonary compromise. Tube thoracostomy should be performed in virtually all infants with the neonatal respiratory distress syndrome and pneumothorax, because the pneumothorax further compromises the ventilatory status and tends to increase in size.

Catamenial Pneumothorax

A catamenial pneumothorax is a pneumothorax that develops in conjunction with menstruation ( Fig. 81-7 ). Up until 2004, 229 cases had been reported, but it is probably underdiagnosed and underreported. With catamenial pneumothorax, respiratory symptoms usually develop within 24 to 48 hours of the onset of the menstrual flow. Most pneumothoraces are right sided, but left-sided and bilateral pneumothoraces have been reported. Catamenial pneumothoraces tend to be recurrent. On average, patients have approximately five pneumothoraces before the diagnosis is recognized.

Figure 81-7

Catamenial pneumothorax.

Frontal chest radiograph from a 32-year-old woman showing a small right-sided pneumothorax ( arrow ) . This was the fourth episode with onset during menses.

(Courtesy Dr. Lewis S. Lehman, San Francisco.)


The pathogenesis of catamenial pneumothorax is unclear. When the syndrome was initially described, it was hypothesized that air gained access to the peritoneal cavity during menstruation and then entered the pleural cavity through a diaphragmatic defect. In a subsequent review of 28 patients who had undergone thoracoscopy, endometriosis, primarily diaphragmatic, was present in 18 and diaphragmatic perforations or nodules were present in 21. It has been suggested that, with diaphragmatic endometriosis, the endometrial tissue undergoes cyclical necrosis leading to a diaphragmatic defect. These authors concluded that diaphragmatic abnormalities play a fundamental role in the pathogenesis of catamenial pneumothorax. Alternatively, endometriosis of the visceral pleura could lead to alveolar pleural air leaks during menstruation.

Diagnosis and Treatment

Any woman who has a spontaneous pneumothorax within the first 48 hours of her menstrual period should be suspected of having a catamenial pneumothorax. The treatment of catamenial pneumothorax is aimed at treating endometriosis, known or suspected, by suppressing the ectopic endometrium. This can be attempted by suppression of ovulation with oral contraceptives or by suppression of gonadotropins with danazol or gonadotropin-releasing hormone to produce a medical oophorectomy. Alternative treatments include thoracoscopy with stapling of blebs, closure of diaphragmatic defects and parietal abrasion or pleurectomy, or pleurodesis. In one series, 28 patients were treated with removal of the diaphragmatic perforation and pleurodesis at thoracoscopy plus suppression of gonadotropins, and the recurrence rate was still 32%.

Tension Pneumothorax

A tension pneumothorax is present when the intrapleural pressure exceeds the atmospheric pressure throughout expiration and often during inspiration as well. Most patients who develop a tension pneumothorax are receiving positive pressure to their airways, either during mechanical ventilation or during resuscitation. For a tension pneumothorax to develop in a spontaneously breathing person, some type of one-way valve mechanism must be present so more air enters the pleural space on inspiration than leaves the pleural space on expiration, and thus air accumulates in the pleural space under positive pressure.


The development of a tension pneumothorax is usually heralded by a sudden deterioration in the cardiopulmonary status of the patient. This is probably related to the combination of a decreased cardiac output due to impaired venous return and profound hypoxia due to ventilation-perfusion mismatches. In mechanically ventilated sheep, an induced tension pneumothorax (mean pleural pressure of +25 cm H 2 O) reduced the cardiac output from 3.5 to 1.1 L/min. The arterial P o 2 also fell from a baseline value of 159 to 59 mm Hg. Comparable reductions in cardiac output and oxygen saturation were seen in pigs and in dogs following induction of tension pneumothoraces. Similarly, in patients on mechanical ventilation who develop tension pneumothorax, there is a large drop in the cardiac output.

Clinical Manifestations

Tension pneumothorax most commonly develops in patients receiving positive-pressure mechanical ventilation or CPR. Occasionally a tension pneumothorax will evolve during the course of a spontaneous pneumothorax or during hyperbaric oxygen therapy. Tension pneumothorax can develop from improper connection of one-way flutter valves with small-caliber chest tubes.

The clinical picture of a tension pneumothorax is often characterized by respiratory distress, cyanosis, marked tachycardia, and profuse diaphoresis, marked hypoxemia and sometimes respiratory acidosis.

Tension pneumothorax should be suspected in patients receiving mechanical ventilation who suddenly deteriorate. In this situation the peak pressures on the ventilator usually increase markedly if the patient is on volume-type ventilation, whereas the tidal volumes decrease markedly if the patient is on pressure-support ventilation. Tension pneumothorax should also be suspected in any patient undergoing CPR in whom ventilation becomes difficult. In one series of 3500 autopsies, an unsuspected tension pneumothorax was found in 12 cadavers; 10 had received mechanical ventilation, and 9 CPR. Tension pneumothorax should also be suspected in patients with a known pneumothorax who deteriorate suddenly or in patients who have undergone a procedure known to cause a pneumothorax.

Diagnosis and Treatment

A tension pneumothorax is a medical emergency. One should not waste time to establish the diagnosis radiologically because the clinical situation and the physical findings usually strongly suggest the diagnosis ( see Fig. 76-4 and eFig. 81-7 ). The patient should immediately be given high-flow supplemental oxygen. Once the abnormal hemithorax is identified, a small (14- to 16-gauge) catheter with needle should be immediately inserted into the pleural space through the second anterior intercostal space. The catheter should be left in place and in communication with the atmosphere until air ceases to exit through the syringe. Additional air can be withdrawn from the pleural space with the syringe and the three-way stopcock. The patient should be prepared for immediate tube thoracostomy.

Reexpansion Pulmonary Edema

Unilateral pulmonary edema ( reexpansion pulmonary edema [RPE]) may develop in certain patients whose lung has been rapidly reinflated after a period of collapse secondary to a pneumothorax or a pleural effusion. Patients with RPE have various degrees of hypoxia and hypotension. On occasion the pulmonary edema becomes bilateral, and the patient requires intubation and mechanical ventilation. On rare occasions the syndrome has been fatal, including in otherwise healthy, young persons. The incidence of RPE is probably relatively low because there were no instances of RPE in the Veterans Administration cooperative study of more than 200 patients with spontaneous pneumothorax. In a retrospective study of 320 episodes of pneumothorax, RPE developed in 3 patients (1.0%). A literature review in 1988 revealed 53 cases of RPE edema, which was fatal in 11 (21%) cases. This is likely an overestimation, because nonfatal cases are less likely to be reported and many patients are asymptomatic.


RPE appears to be due to increased permeability of the pulmonary vasculature. In both experimental animals and humans, the edema fluid has a high protein content, suggesting that edema forms because of increased capillary leak rather than increased hydrostatic pressure. It has been hypothesized that the mechanical stresses applied to the lung during reexpansion damage the capillaries and lead to pulmonary edema. Reperfusion injury due to reactive oxygen species is another possibility. However, several animal studies demonstrated that the administration of reactive oxygen species–scavenging compounds such as dimethylthiourea, catalase, or superoxide dismutase all partially inhibit the neutrophilic infiltration associated with the development of RPE but do not impressively decrease the amount of edema in the experimental situation. Moreover, neutrophil depletion does not affect the amount of edema. For these reasons, mechanical stress on the lung is currently considered to be the most likely cause of RPE.

In humans, most cases of RPE develop when the pneumothorax or pleural effusion has been present for at least 3 days and when negative pressure has been applied to the pleural space. Similar findings have been confirmed in experimental animal studies.

Clinical Manifestations

Patients with RPE typically have pernicious coughing or chest tightness during or immediately after tube thoracostomy or large-volume thoracentesis. The symptoms usually progress for 12 to 24 hours, and serial chest radiographs reveal progressive ipsilateral pulmonary edema (see Fig. 76-6 ), which may progress to involve the contralateral lung. Treatment is primarily supportive, with the administration of supplemental oxygen and diuretics, and intubation and mechanical ventilation when necessary. One report suggested that the syndrome could be aborted if the patient is treated with continuous positive airway pressure within the first hour of the development of the syndrome. Chest tubes should be placed to underwater-seal drainage if the syndrome develops to avoid further aggravation with suction.


Because RPE can be fatal, it is important to prevent it when possible. The possibility of its development should be considered in any patient with a large pneumothorax or pleural effusion subjected to tube thoracostomy or thoracentesis. When tube thoracostomy is performed for spontaneous pneumothorax, the tubes should initially be connected to an underwater-seal drainage apparatus rather than to negative pleural pressure. If underwater-seal drainage does not cause reexpansion of the underlying lung within 24 to 48 hours, then negative pressure can be applied to the pleural space.

When a thoracentesis is performed, the procedure should be terminated if the patient develops tightness of the chest or experiences coughing. In a series of 941 thoracenteses, including 119 that drained more than 1500 mL, opacities consistent with RPE were noted in only two patients (0.2%) from whom 1000 and 1200 mL of pleural fluid had been removed: neither was symptomatic or required treatment. Thus, RPE is rare and rarely clinically significant. Patients should be monitored during thoracentesis for symptoms, and, although pleural manometry has been recommended to follow pleural pressures, such approaches have not yet been shown to alter the incidence of this rare complication.


Pleural fluid can be milky or turbid. When this cloudiness persists after centrifugation, it is almost always due to high lipid content in the pleural fluid. High levels of lipid can accumulate in pleural liquid through two separate mechanisms. In one, chyle enters the pleural space following disruption of the thoracic duct, producing a chylothorax (chylous pleural effusion). In the second, large amounts of cholesterol or lecithin-globulin complexes accumulate in a pleural effusion to produce a pseudochylothorax (chyliform or cholesterol pleural effusion). It is important to recognize and differentiate these two conditions because their etiology and management are completely different.


A chylothorax forms when the thoracic duct, which carries dietary fat in the form of chylomicrons, becomes disrupted. Chylomicrons are formed in the intestine, after which they enter the intestinal lacteal vessels and are then transported to the cisterna chyli. The thoracic duct, a 2- to 3-mm-wide thin-walled conduit, leaves the cisterna chyli and passes through the aortic hiatus of the diaphragm on the anterior surface of the vertebral body between the aorta and the azygos vein into the posterior mediastinum. The thoracic duct then ascends extrapleurally in the posterior mediastinum along the right side of the anterior surface of the vertebral column. Between the level of the fourth and sixth thoracic vertebrae, the duct crosses to the left side of the vertebral column and continues cranially to enter the superior mediastinum between the aortic arch and the subclavian artery and the left side of the esophagus. As a result, rupture of the thoracic duct in its caudal portion tends to produce right-sided chylothoraces, and rupture of the cranial portion of the duct tends to produce left-sided chylothoraces. Once the thoracic duct passes the thoracic inlet, it arches above the clavicle and passes anterior to the left subclavian artery, vertebral artery, and thyrocervical trunk to terminate in the region of the left jugular and subclavian veins.

Although the route just described is the typical one, there are wide anatomic variations throughout the course of the duct. For example, duplication or even triplication of the thoracic duct is present in 40% of the population. Also, many collateral vessels and lymphaticovenous anastomoses are known to exist. Presumably these channels transport the chyle to the blood following therapeutic ligation of the thoracic duct. The wide range of anatomic variation and the multiple collateral channels render the thoracic duct at risk for injury, especially during thoracic surgery.

Chyle, the liquid drained from the thoracic duct, is a milky, opalescent fluid that usually separates into three layers on standing: a creamy uppermost layer containing chylomicrons, a milky intermediate layer, and a dependent layer containing cellular elements. Chyle is bacteriostatic, not irritating, and not known to induce pleural thickening.

The thoracic duct normally conveys between 1500 and 2500 mL of chyle daily. Ingestion of fat can increase the flow of lymph in the thoracic duct by two to tenfold for several hours. Ingestion of liquid, but not protein or carbohydrate, also increases the flow, whereas starvation decreases chyle flow. The primary cellular component of chyle is the T lymphocyte. Prolonged loss of chyle can result in severe nutritional depletion, dehydration, electrolyte loss, and hypolipidemia, as well as depletion of T- and B-cell lymphocytes and immunodeficiency.


Chylothorax can result from traumatic (including surgical) and nontraumatic causes: their relative frequencies vary among series. In a summary of five separate series totaling 143 patients, greater than 50% of chylothoraces were caused by tumors, especially lymphoma ( ), followed by trauma. However, a single institute series of 203 patients revealed trauma/surgery as the most common (50%) cause compared to nontraumatic causes (44%).

The unpredictable anatomy of the thoracic duct and associated accessory lymphatics makes them vulnerable to injury during cardiovascular or thoracic surgical procedures, especially those involving the posterior mediastinum. Esophagectomy, for example, is complicated by chylothorax in 1% to 4% of cases. Overall, the incidence of chylo­thorax after cardiothoracic operations is low (0.5% to 2.5%). Surgical procedures that involve mobilization of the left subclavian artery are particularly likely to be complicated by chylothorax. Transplantation of the lung and heart (or both) can sever the lymphatic drainage and result in chylothoraces. Chylothorax has been reported to complicate a wide range of other operations, including esophagoscopy, stellate ganglion blockade, thoracic sympathectomy, high translumbar aortography, lung resections, thyroid surgery, spinal surgery, and even gastric banding. Bilateral chylothoraces can develop following bilateral neck dissection. Chylothorax is common after surgical repair of congenital diaphragmatic hernia (incidence, 4.6% in a series of 1383 patients), especially if patch repair or extracorporeal membrane oxygenation is employed. Most (>80%) patients can be treated conservatively, and mortality is not increased. To aid in the identification of the thoracic duct during surgery, ingestion of cream (to increase the flow and size of the thoracic duct) before high-risk operations (e.g., esophagectomy) has been advocated.

Nonsurgical trauma can also lead to chylothorax. The thoracic duct may be disrupted with penetrating injuries involving the neck or thorax. Nonpenetrating trauma in which the spine is hyperextended or a vertebra is fractured can lead to chylothorax, particularly if the injury follows shortly after the ingestion of a fatty meal. Less impressive traumas such as weight lifting, straining, severe bouts of coughing or vomiting, childbirth, and vigorous stretching while yawning have been associated with chylothorax. Chylothoraces secondary to closed trauma are usually on the right side, with the site of rupture in the region of the 9th or 10th thoracic vertebra.

Chylothoraces can also arise from transdiaphragmatic movement of chylous ascites. Causes of chylous ascites include many of the causes of chylothorax. In addition, cirrhosis may be a cause of chyloascites and associated chylothorax. These chylothoraces are transudative, probably from dilution by low-protein cirrhotic ascitic fluid.

There are many other causes of chylothorax, but all together they account only for a small percentage of all chylothoraces. Lymphangioleiomyomatosis (see also Chapter 69 ) and other lymphatic abnormalities, such as pulmonary lymphangiectasis, yellow nail syndrome, lymph node enlargement, and lymphangitis of the thoracic duct can result in chylothorax, as can tuberous sclerosis, amyloidosis, mediastinal fibrosis, lupus, and Gorham syndrome. Elevated pressure of the venous system into which the thoracic duct drains (e.g., superior vena caval or subclavian vein thrombosis/obstruction) is another cause of chylothorax. In postsurgical patients, those with central venous thrombosis were five times more likely to develop chylothorax. Chylothorax accounts for about one fifth of pleural effusions in patients with superior vena cava syndrome.

Chylothorax with no identifiable cause (5% to 10% of all cases) is labeled idiopathic. Lymphoma, however, must be excluded.

Fetal and Neonatal Chylothorax

Fetal chylothorax is an uncommon but important condition that requires monitoring and management to avoid serious complications, including spontaneous abortion or death after birth. Fetal chylothoraces are often termed primary fetal pleural effusions, because no obvious cause can be identified in most cases. Congenital lymphangiectasis is a rare condition that can produce fetal chylothorax. Cytogenetic analysis of the cells (mainly lymphocytes) may help detect underlying chromosomal abnormalities.

Chylothorax is the most common type of neonatal pleural effusion, with an incidence estimated at 1 : 15,000. Of note, the pleural fluid remains clear rather than milky until milk feeding begins. It may be a result of persistent fetal chylothorax but can also be due to developmental abnormalities of the thoracic duct or its rupture from trauma during delivery. Increased venous pressure, especially from congenital heart diseases or from thrombosis of central venous catheters, is another recognized mechanism of neonatal chylothorax. In many cases the chylothorax is idiopathic.

Mutations of several candidate genes have recently been linked with congenital chylothorax, including integrin α 9 ( ITGA9 ) and β 1 ( ITGB1 ), vascular endothelial growth factor receptor 3 ( FLT4 , and FOXC2 ). The integrin α 9 β 1 is widely expressed in smooth muscles and is a receptor for extracellular matrix proteins and vascular cell adhesion molecule-1. Animal studies suggested that the α9 subunit is required for the normal development of the lymphatic system, including the thoracic duct. Mice with homozygous null mutation of the α9 subunit develop large bilateral congenital chylothoraces. An autosomal inheritance of a heterozygous missense mutation (c.1210G>A, p.G404S) of ITGA9 has been found in four of five fetuses with chylothoraces who did not respond to prenatal pleurodesis but not in those who responded.

The gene for vascular endothelial growth factor C was found to be important in tumor-related lymphangiogenesis and chyloascites formation in mice bearing ovarian carcinoma. Chy-3-mutant mice, which carry a chromosomal deletion that includes Vegfc , develop hypoplastic dermal lymphatic drainage and resultant chyloascites and lymphedema. Transgenic mice with mutations in the Pi3kca ( phosphoinositide 3-kinase ) gene also develop defective lymphatics and chyloascites. Rasa1 (also known as p120 RasGAP) is a Ras GTPase-activating protein that functions as a regulator of blood vessel growth in adult mice and humans. In mice, systemic loss of Rasa1 resulted in early lethality caused by chylothorax with underlying extensive lymphatic vessel hyperplasia and leakage.

Clinical Manifestations

The symptoms, physical findings, and radiographic features of chylothorax are the same as those encountered in patients with comparably sized pleural effusions of any cause. Pleuritic chest pain and fever are rare because chyle is not proinflammatory. Chylopericardium or chyloascites can be present concurrently.

With nontraumatic chylothorax, insidious onset of dyspnea on exertion is common. With traumatic chylothorax, there is usually a latent period of 2 to 10 days between the trauma and the clinical presentation of the pleural effusion. During this latent period, chyle may accumulate in the posterior mediastinum to form a chyloma—visible radiologically as a posterior mediastinal mass —which eventually ruptures into the pleural cavity, giving rise to a chylothorax.

Neonatal chylothorax may present with respiratory distress in the first few days of life. Fifty percent of the infants have symptoms within the first 24 hours, whereas 75% have symptoms by the end of the first week. Most neonatal chylothoraces are either right sided or bilateral, but rarely left sided. There is a high frequency of neonatal chylothoraces in infants with hydramnios. Fetal chylothorax is often diagnosed only on ultrasonography.

The main threat to life with chylothorax is from external drainage leading to inanition. The daily loss of 1.5 to 2.5 L of fluid rich in protein, fats, electrolytes, and lymphocytes will rapidly render patients malnourished and immunocompromised.


The distinctive white, odorless, milky appearance should suggest the diagnosis, though chylothorax must be differentiated from empyema and pseudochylothorax ( Fig. 81-8 ). In empyema, the milky appearance is caused by suspended leukocytes and debris, which will sediment upon centrifugation, leaving a clear supernatant. In both chylous and chyliform pleural effusions, the milky appearance is caused by high lipid levels, and the supernatant will remain cloudy after centrifugation. The lipids in chyliform effusion are cholesterol or lecithin-globulin complexes rather than chylomicrons (as in chylothoraces). It should be emphasized that the pleural fluid with chylothorax can occasionally be bloody or even clear yellow. In one series only 44% of chylothoraces were milky.

Jul 21, 2019 | Posted by in CARDIOLOGY | Comments Off on Pneumothorax, Chylothorax, Hemothorax, and Fibrothorax
Premium Wordpress Themes by UFO Themes