- Email: [email protected]
Tension Pyopneumothorax
Tension Pyopneumothorax* Rare Presentation of Ruptured Barrett's Esophagus Maria A. Matsumo to, M.D. ; S. David Rockoff, M.D ., EG .G.P.; and Benjamin L. Aaron, M.D. , EG .G.P.
Tension pneumothorax following esophageal rupture is very rare. We report a case in which a perforated Barrett's esophagus rapidly developed a spontaneous tension pyopneumothorax. The mechanism for the tension remains obscure. Knowing that a tension pneumothorax can occur with esophageal rupture can help prevent misdiagnosis. (Chest 1993; 103:1604-06)
E
sophageal rupture is common and numerous causes have been reported.':' The classic radiographic findings, regardless of cause, include pleural effusion, pneumothorax , pneumomediastinum , and subcutaneous emphysema .'"·"·l7·'· However, review of the English language literature reveals only one prior case of a tension pneumothorax due to rupture of the esophagus." We report what is, to our knowledge, the first case of a tension pyopneumothorax resulting from a ruptured Barrett's esophageal ulcer and the second known case of a tension pneumothorax due to esophageal perforation. CASE REPORT
A 77-year-old white man presented to the emergency room with a 3-day history of cough, right pleuritic chest pain and increasing shortness of breath and a I-day history of severe right costovertebral pain . He had undergone two hiatal hernia operations and an abdominal aortic aneurysm repair and was receiving medication for peptic reflux . He was in moderate respiratory distress, and scattered rhonchi were heard in his lungs bilaterally. He had right eostovertebral angle tenderness and a soft, nontender abdomen. The WBC count was 12,600/mm 3 • The admission chest radiograph (Fig I) was initially interpreted as showing right pleural effusion, possible right lower lobe consolidation, and a midline mediastinum, leading to the clinical imp ression of right- sided pneumonia with a parapneumonic effusion. An abd ominal computed tomographic (Cf) scan was ordered to evaluate his upper abdomen and aorta. A regimen of ceftriaxone was started for possible pneumonia. However, while awaiting the CT scan, he rapidly developed a marked increase in shortness of breath and a decrease in oxygen saturation. He was given furosemide and a bronchodilator and became sufficiently stable to undergo the abdominal CT scan 5 h after admi ssion. On the CT scan, an unexpected right hydropneumothorax was seen, and the stud y was therefore extended to include more of the chest . The mediastinum was shifted toward the left (Fig 2), indicating the presence of a right tension hydropneumothorax. No mediastinal air was see n. A chest tube was placed on the right with an immediate rush of air and rapid drainage of approx imately 1,000 ml of light brown fluid . A chest radiograph obtained after insertion of the chest tube showed the med iastinum to have returned to the midline . The initial analysis of the pleural fluid showed a pH of6.98 and a glucose level of 18 mg/dl. *From the Department of Radiology (Drs. Matsumoto and Rockoff) and the Division of Thoracic and Cardiovascular Surgery (Dr. Aaron), George WashinbrtonUniversity Hospital, Washington, DC . Reprint requests: Dr. Rockoff, George Washington University Medical Genter; 901 23rd Street. NW, Washington, DG 20037
1604
FIGURE 1. Admission chest radiog raph shows right pleural effusion and right lower lobe consolid ation . Mediastinum is midlin e . Left eighth rib is surgically absent. Left lower lobe is hypoaerat ed . In the ICU , the prelimin ary diagnoses were pneumonia and empye ma, with a bronchopleural fistula as the caus e for the tension pneumothorax . The patien t required suppl em ental oxygen ; his tachypnea gradually decreased , and he remained afebrile . In the first 48 h the chest tub e drain ed over 2,500 ml of serosanguineous fluid, without an air leak. In the absence of a definitive diagnos is, anothe r chest CT scan was obtained , which showed a communication between the esophagus and the right pleural space (Fig 3). Complete analysis of the pleural fluid showed the following values: pl l . 6.9; glucose , IS mg/
FIGURf: 2. Ch est C'T scan (suboptimal du e to tachypnea) shows large tension hydropneumothorax on the right with air-fluid level (arrows) and marked med iastinal shift to the I('fl. l'I:o definit e medi astinal air was ide ntified , but a small amount could not be excluded bec ause of streak artifacts. TensionPyopneumothorax (Matsumoto, Rockoff. Aaron)
FICITIlE 3. A CT image of Iower thorax shows communication (thick, straight black arrlllcs) between the esophagus (E ) and the right pleural space (curved arnnc). Tiny pneumome-diastinum (thin arrow) is identified . Hight chest tube (rchitl' arrow) and bilateral pleural
effusion s ar e seen.
dl; ' ,-lOO WBC /mm'; 39 ,800 HBC/mm'; and amylase , 2,580 UlL. Based on these C1' and pleural fluid findings , the definitive diagnosis of an esophagopleural fistula was made . At sur ge r y a 3-cm-long perforation of the right posterior esophagus was identified . A distal esophagectomy with a thoracic esophagogastrostmny was performed . Histopathologic examination of the esophagus showed a Barrett's esophagus with extensive ulceration and perforation of a Barrett 's ulcer. The patient slowly recovered and was eventually dis charged . DISCUSSIO:-J
Iatrogenic causes are responsible for 48 to 68 percent of esophageal perforations, with endoscopy and dilation the most common .!"!"'!" Other causes include external penetrating or blunt trauma, 15.1" pressure necrosis or direct laceration by a foreign hody, II .'" severe emesis.":!" and , rarely, perforation of a Barretts ulcer or esophageal diverticulum . IH Barrett's esophagus is ~enerally an acquired lesion Sl'Condary to chronic gastroesophageal reflux with columnar epithelium found within the e s o p h a ~ u s .h .1 7 Ulceration and strictures han' been observed with Barrett's esophagus. 1. 11·1 1 TIJt> incidence of adenocarcinoma is said to lx- 10 to 41 percent ."!' Of the hundreds of cases of Barrett's esophagus literature, we found only 12 cases in the English lan~ua~e of rupture. 1·11 Perforations into the pleural spact'l.1.1 and into adjacent mediastinal structures'>" have been reported . The site of perforution was not specified in two caSI'S. I" The common clinical findings with esophageal perforation from any cause are nonspecific and include chest and/or hack pain , subcutaneous emphysema, fever, and dyspnea ."·17'-' TIlt' plain chest radiographs oftcn show pneurnomediastinum, pneumothorax. and pleural effusion, 15 I" hut were normal in 3.'3 percent of patients with iatrogenic perforations" and revealed " nond iagnostic ahnormalities" in oth er s.17 Esophageal rupture is associated with a mortality of 18 to 44 perc ent.":":" According to Ri~~ and Walker; IH regardless of cause, delayed recognition of esophageal perforation is the rule, tlu- major reasons being wrong diagnosis, failure to of the brown pleural fluid, and recognize the si~nificance
lack of awareness of the problem . Diagnosis is delayed for more than 24 h in 50 to 77 percent of cases," .15.'7 and the "classic" history and symptoms are often absent.15.17 The most important study to obtain is an esophagram, This has been shown to be diagnostic in 86 to 93 percent of cases.!"!"!" If the diagnosis is still not clear, a chest CT scan may demonstrate pneumomediastinum, abscess cavities adjacent to the esophagus, or the actual perforation site .20,21 The pleural fluid of esophageal rupture is characteristically purulent with an elevated amylase level." The treatment of esophageal perforation can be nonoperative or surgical depending on the location and nature of the perforation and the clinical setting.,us.l" Pertinent to our case, three of the four prior cases of Barrett's ulcer rupture with an esophagopleural fistula had chest radiographs with hydropneumothoraces.' None had a tension pneumothorax . The other reported case of a tension pneumothorax from an esophageal perforation was secondary to a fistula from an esophageal diverticulum to the pleural space.'" The presence of tension associated with the hydropneumothorax in our case remains unexplained . \Ve know of no discussions in the literature that elucidate a mechanism for its development. Perhaps our patient's previous hiatal hernia surgery and associated changes served to create a checkvalve type of obstruction at the site of rupture and thus permitted the tension pyopneumothorax to develop. In summary, esophageal perforation is very difficult to diagnose, and delay in treatment is common. The rare presence of a tension hydropneumothorax in esophageal perforation is especially misleading and is potentially a further caus e for delayed diagnosis and treatment. In the case we present, the chest CT findings and increased amylase in the pleural fluid were diagnostic. Knowledge that an esophageal rupture can cause a tension hydropneumothorax may prevent delay in the diagnosis of such cases. REFERENCES
I Starnes VA , Adkins RB, Ballinger JF, Sawyers JL. Barrett's esophagus: a surgical entity. Arch Surg 1984; 119:563-67 2 Cappell MS, Sciales C, Biempica L. Esophageal perforation at a Barrett's ulcer. J Clin G;lstroenteroll989; 11:663-66 3 Limburg AJ, lIesselink EJ, Kleibeuker JB . Barrett's ulcer: cause of spo ntaneous oesophageal perforation. Gut 1989; 30:404-05 4 Andersson R, Nilsson S. Perforated Barrett's ulcer with esophago-pleural fistula. Acta Chir Scand 1985; 151:495-96 5 Borne J. Goldwater L. Columnar cell-lined esophagus: assessment of etiology and treatment-a 22 year experience. J Thorac Cardiovasc Surg 1976; 71:825-34 6 Cerstenberger PO, Pellegrini CA, TIerney LM . Barrett's ulcer of the esophagus: previously unrecognized cause of acquired esophagoresplratory fistula . Am J Med 1986; 81 :713-17 7 Diehl JT, Thomas L, Bloom MB , Dresdale AR, Harasimowicz P, Daly BOT, et al. Tracheoesophageal fistula associ ated with Barrett's ulcer: the importance of reflux control , Ann Thorne Surg 1988; 45:449-50 8 Lamhert DR , Llaneza PP, Caglani RD , Lach RD , Beaver WL. Esophageal-atrial fistula. J Clin Gastroenteml 198.; 9:345-49 9 ltahashi 1111, Granada LO. Cerebral food embolism secondary to esophageal-cardiac perforation, JAMA 1972; 219:373-75 10 Larsen K, Jensen BS, Axelsen F. Perforation and rupture of the esophagus. Scand J Thorne Cardiovase Surg 1983; 17:311-16 11 Cooper BT, Barbe...at GO. Barrett's oesophagus: a clinical stud y CHEST I 103 I 5 I MAY, 1993
1605
of 52 patients. Q J Med 1987; 62:97-108 12 Kerlin P, D'Mellow G, Van Deth A. Barrett's esophagus: clinical , endoscopic, and histologic spectrum in fifty patients. Aust NZ J Med 1986; 16:198-205 13 Herlihy KJ, Orlando RC, Bryson JC, Bozyrnski EM, Carney CN, Powell D\V. Barrett's esophagus: clinical, endoscopic. histologic, manometric, and electrical potential difference characteristics. Gastroenterology 1984; 86 :436-43 14 Bladergroen MR, Lowe JE, Postlethwait RW Diagnosis and recommended management of esophageal perforation and rupture. Ann Thorac Surg 1986; 42;235-39 15 Flynn AE, Verrier ED, Way L'N, Thomas AN, Pellegrini CA. Esophageal perforation. Arch Surg 1989; 124:1211-15 16 Goldstein LA, Thompson WR. Esophageal perforations: a 15 year experience. Am J Surg 1982; 143:495-503 17 Pate J\v' Walker WA, Cole FH , Owen E\V, Johnson WHo Spontaneous rupture of the esophagus: a 30 year experience. Ann Thorac Surg 1989; 47:689-92 18 Rigg KM, Walker RW, Tension pneumothorax secondary to ruptured oesophageal diverticulum. Br J Clin Pract 1990; 44 : 528-29 19 Michel L, Grillo HC, Malt RA. Operative and nonoperative management of esophageal perforations. Ann Surg 1981; 194:5763 20 Faling LJ, Pugatch RD, Bobbins AH . The diagnosis of unsuspected esophageal perforation hy computed tomography, Am J Med Sci 1981; 281:31-4 21 Backer CL, LoCicero J, Hartz RS, Donaldson JS, Shields T. Computed tomography in patients with esophageal perforation. Chest 1990; 98:1078-80 22 Sherr Hp, Light RW, Merson Mil , WolfRO, Taylor LL, Hendrix TR . Origin of pleural fluid amylase in esophageal rupture , Ann Intern Med 1972; 76:985-86
Pneumomediastinum Caused by Subcutaneous Emphysema in the Shoulder* A Rare Complication of Arthroscopy Kam-Yung
[AU,
M.D . , EC.C.P.
Severe subcutaneous emphysema following arthroscopy of the shoulder developed in a 62-year-old man. It subsequently evolved into pneumomediastinum with respiratory distress, but the patient recovered spontaneously. To our knowledge, this constitutes the first report of pneumomediastinum caused by subcutaneous emphysema in the shoulder. (Chest 1993; 103:1606-07)
shoulder joint was distended with saline solution, of which the continuous inflow and outflow were regulated hy au arthroscopy infusion pump. Surgery and anesthesia were uneventful. hut toward the end of the 2-hour surgery, swelling and crepitation of the right ann and shoulder were noted. These rapidly spread to the right side of the chest, the neck, and the face and later also involved the left side as well . A chest radiograph (Fig I) demonstrated subcutaneous and mediastinal emphysema. Two hours after completion of surgery, the patient developed a fever of 38.4°(; . lie felt slightly short of breath and complained of sore throat and suhsternal chest pain . Arterial blood gas measurements, while the patient was breathing oxygen 2 Umin by nasal cannula, revealed that the PaO, was 57 mm IIg, the PaCO, was 38 mm IIg , and the I'll was 7.43 . With increase of Flo, to 40 percent (hy a Venti-mask), the PaO, increased to 81 mm IIg. Fiberoptic bronchoscopy disclosed no evidence of pharyngeal , laryngeal, or tracheal injury. Imipenem-cilastatln therapy W;lS started intravenously lie con tinued to run a fever fi.r the next 3 days and then defervesced . Blood and sputum cultures grew no microorganisms. Subsequent chest radiographs showed no development of pneumothorax. Both subcutaneous emphysema ami pneumomediastinum started to decrease on the third postoperative day and completely resolved on day 7. DISCUSSION
Air in the mediastinum can originate from fivl' sites : the neck , the alveoli , the airways, the esophagus, and the abdominal cavity. I In the case presented, there was no evidence of airway, esophageal, or abdominal injuries. The patient was not under prolonged or high positive pressure ventilation and therefore unlikely to have development of alveolar rupture . Had alveolar rupture occurred, giving rise to pneumomediastinum and subsequently subcutaneous emphysema, the latter would have been more symmetrical rather than commencing and remaining predominant around the right shoulder as occurred in this case . Thus, it appeared that the subcutaneous air in the neck, which originated from the right shoulder, was the most likely cause of this patient's pneumomediastin um. Subcutaneous emphysema in tilt' shoulder causing pneumomediastinum is, however, rather unusual. Deep cervical fascia divides the soft-tissue structures of the neck into three distinct compartments: the previsceral space, the visceral
S
ubcutaneous emphysema in the shoulder causing pneumomediastinum has not been described . I wish to report such a case, which occurred as a complication of arthroscopy of the shoulder. CASE REPORT
A 62-year-old man underwent arthroscopy of the right shoulder for debridement of the glenoid lahral tear. General anesthesia was induced, and the trachea was intuhated easily on the first attempt with an 8.0-mm oral endotracheal tube, It was then connected to a volume-cycled ventilator with a tidal volume of 600 ml and a rate of Hlmin. Peak inspiratory pressure was around 15 to 16 cm 11,0. The *From the Riverside Medical Clinic, Riverside, Calif. Reprint requests: Dr. [AU, .3660 Arlington Avenue, Riverside, CA 92506
1606
Fu.umc 1. Anteroposterior radiograph of the chest revealing the pneumomediastinum (lIrmws) and rndiolucent streaks of air in the subcutaneous tissue. Pneumomediastinum Caused by Subcutaneous Emphysema (K8n-Yung Lau)
Tension pneumothorax following esophageal rupture is very rare. We report a case in which a perforated Barrett's esophagus rapidly developed a spontaneous tension pyopneumothorax. The mechanism for the tension remains obscure. Knowing that a tension pneumothorax can occur with esophageal rupture can help prevent misdiagnosis. (Chest 1993; 103:1604-06)
E
sophageal rupture is common and numerous causes have been reported.':' The classic radiographic findings, regardless of cause, include pleural effusion, pneumothorax , pneumomediastinum , and subcutaneous emphysema .'"·"·l7·'· However, review of the English language literature reveals only one prior case of a tension pneumothorax due to rupture of the esophagus." We report what is, to our knowledge, the first case of a tension pyopneumothorax resulting from a ruptured Barrett's esophageal ulcer and the second known case of a tension pneumothorax due to esophageal perforation. CASE REPORT
A 77-year-old white man presented to the emergency room with a 3-day history of cough, right pleuritic chest pain and increasing shortness of breath and a I-day history of severe right costovertebral pain . He had undergone two hiatal hernia operations and an abdominal aortic aneurysm repair and was receiving medication for peptic reflux . He was in moderate respiratory distress, and scattered rhonchi were heard in his lungs bilaterally. He had right eostovertebral angle tenderness and a soft, nontender abdomen. The WBC count was 12,600/mm 3 • The admission chest radiograph (Fig I) was initially interpreted as showing right pleural effusion, possible right lower lobe consolidation, and a midline mediastinum, leading to the clinical imp ression of right- sided pneumonia with a parapneumonic effusion. An abd ominal computed tomographic (Cf) scan was ordered to evaluate his upper abdomen and aorta. A regimen of ceftriaxone was started for possible pneumonia. However, while awaiting the CT scan, he rapidly developed a marked increase in shortness of breath and a decrease in oxygen saturation. He was given furosemide and a bronchodilator and became sufficiently stable to undergo the abdominal CT scan 5 h after admi ssion. On the CT scan, an unexpected right hydropneumothorax was seen, and the stud y was therefore extended to include more of the chest . The mediastinum was shifted toward the left (Fig 2), indicating the presence of a right tension hydropneumothorax. No mediastinal air was see n. A chest tube was placed on the right with an immediate rush of air and rapid drainage of approx imately 1,000 ml of light brown fluid . A chest radiograph obtained after insertion of the chest tube showed the med iastinum to have returned to the midline . The initial analysis of the pleural fluid showed a pH of6.98 and a glucose level of 18 mg/dl. *From the Department of Radiology (Drs. Matsumoto and Rockoff) and the Division of Thoracic and Cardiovascular Surgery (Dr. Aaron), George WashinbrtonUniversity Hospital, Washington, DC . Reprint requests: Dr. Rockoff, George Washington University Medical Genter; 901 23rd Street. NW, Washington, DG 20037
1604
FIGURE 1. Admission chest radiog raph shows right pleural effusion and right lower lobe consolid ation . Mediastinum is midlin e . Left eighth rib is surgically absent. Left lower lobe is hypoaerat ed . In the ICU , the prelimin ary diagnoses were pneumonia and empye ma, with a bronchopleural fistula as the caus e for the tension pneumothorax . The patien t required suppl em ental oxygen ; his tachypnea gradually decreased , and he remained afebrile . In the first 48 h the chest tub e drain ed over 2,500 ml of serosanguineous fluid, without an air leak. In the absence of a definitive diagnos is, anothe r chest CT scan was obtained , which showed a communication between the esophagus and the right pleural space (Fig 3). Complete analysis of the pleural fluid showed the following values: pl l . 6.9; glucose , IS mg/
FIGURf: 2. Ch est C'T scan (suboptimal du e to tachypnea) shows large tension hydropneumothorax on the right with air-fluid level (arrows) and marked med iastinal shift to the I('fl. l'I:o definit e medi astinal air was ide ntified , but a small amount could not be excluded bec ause of streak artifacts. TensionPyopneumothorax (Matsumoto, Rockoff. Aaron)
FICITIlE 3. A CT image of Iower thorax shows communication (thick, straight black arrlllcs) between the esophagus (E ) and the right pleural space (curved arnnc). Tiny pneumome-diastinum (thin arrow) is identified . Hight chest tube (rchitl' arrow) and bilateral pleural
effusion s ar e seen.
dl; ' ,-lOO WBC /mm'; 39 ,800 HBC/mm'; and amylase , 2,580 UlL. Based on these C1' and pleural fluid findings , the definitive diagnosis of an esophagopleural fistula was made . At sur ge r y a 3-cm-long perforation of the right posterior esophagus was identified . A distal esophagectomy with a thoracic esophagogastrostmny was performed . Histopathologic examination of the esophagus showed a Barrett's esophagus with extensive ulceration and perforation of a Barrett 's ulcer. The patient slowly recovered and was eventually dis charged . DISCUSSIO:-J
Iatrogenic causes are responsible for 48 to 68 percent of esophageal perforations, with endoscopy and dilation the most common .!"!"'!" Other causes include external penetrating or blunt trauma, 15.1" pressure necrosis or direct laceration by a foreign hody, II .'" severe emesis.":!" and , rarely, perforation of a Barretts ulcer or esophageal diverticulum . IH Barrett's esophagus is ~enerally an acquired lesion Sl'Condary to chronic gastroesophageal reflux with columnar epithelium found within the e s o p h a ~ u s .h .1 7 Ulceration and strictures han' been observed with Barrett's esophagus. 1. 11·1 1 TIJt> incidence of adenocarcinoma is said to lx- 10 to 41 percent ."!' Of the hundreds of cases of Barrett's esophagus literature, we found only 12 cases in the English lan~ua~e of rupture. 1·11 Perforations into the pleural spact'l.1.1 and into adjacent mediastinal structures'>" have been reported . The site of perforution was not specified in two caSI'S. I" The common clinical findings with esophageal perforation from any cause are nonspecific and include chest and/or hack pain , subcutaneous emphysema, fever, and dyspnea ."·17'-' TIlt' plain chest radiographs oftcn show pneurnomediastinum, pneumothorax. and pleural effusion, 15 I" hut were normal in 3.'3 percent of patients with iatrogenic perforations" and revealed " nond iagnostic ahnormalities" in oth er s.17 Esophageal rupture is associated with a mortality of 18 to 44 perc ent.":":" According to Ri~~ and Walker; IH regardless of cause, delayed recognition of esophageal perforation is the rule, tlu- major reasons being wrong diagnosis, failure to of the brown pleural fluid, and recognize the si~nificance
lack of awareness of the problem . Diagnosis is delayed for more than 24 h in 50 to 77 percent of cases," .15.'7 and the "classic" history and symptoms are often absent.15.17 The most important study to obtain is an esophagram, This has been shown to be diagnostic in 86 to 93 percent of cases.!"!"!" If the diagnosis is still not clear, a chest CT scan may demonstrate pneumomediastinum, abscess cavities adjacent to the esophagus, or the actual perforation site .20,21 The pleural fluid of esophageal rupture is characteristically purulent with an elevated amylase level." The treatment of esophageal perforation can be nonoperative or surgical depending on the location and nature of the perforation and the clinical setting.,us.l" Pertinent to our case, three of the four prior cases of Barrett's ulcer rupture with an esophagopleural fistula had chest radiographs with hydropneumothoraces.' None had a tension pneumothorax . The other reported case of a tension pneumothorax from an esophageal perforation was secondary to a fistula from an esophageal diverticulum to the pleural space.'" The presence of tension associated with the hydropneumothorax in our case remains unexplained . \Ve know of no discussions in the literature that elucidate a mechanism for its development. Perhaps our patient's previous hiatal hernia surgery and associated changes served to create a checkvalve type of obstruction at the site of rupture and thus permitted the tension pyopneumothorax to develop. In summary, esophageal perforation is very difficult to diagnose, and delay in treatment is common. The rare presence of a tension hydropneumothorax in esophageal perforation is especially misleading and is potentially a further caus e for delayed diagnosis and treatment. In the case we present, the chest CT findings and increased amylase in the pleural fluid were diagnostic. Knowledge that an esophageal rupture can cause a tension hydropneumothorax may prevent delay in the diagnosis of such cases. REFERENCES
I Starnes VA , Adkins RB, Ballinger JF, Sawyers JL. Barrett's esophagus: a surgical entity. Arch Surg 1984; 119:563-67 2 Cappell MS, Sciales C, Biempica L. Esophageal perforation at a Barrett's ulcer. J Clin G;lstroenteroll989; 11:663-66 3 Limburg AJ, lIesselink EJ, Kleibeuker JB . Barrett's ulcer: cause of spo ntaneous oesophageal perforation. Gut 1989; 30:404-05 4 Andersson R, Nilsson S. Perforated Barrett's ulcer with esophago-pleural fistula. Acta Chir Scand 1985; 151:495-96 5 Borne J. Goldwater L. Columnar cell-lined esophagus: assessment of etiology and treatment-a 22 year experience. J Thorac Cardiovasc Surg 1976; 71:825-34 6 Cerstenberger PO, Pellegrini CA, TIerney LM . Barrett's ulcer of the esophagus: previously unrecognized cause of acquired esophagoresplratory fistula . Am J Med 1986; 81 :713-17 7 Diehl JT, Thomas L, Bloom MB , Dresdale AR, Harasimowicz P, Daly BOT, et al. Tracheoesophageal fistula associ ated with Barrett's ulcer: the importance of reflux control , Ann Thorne Surg 1988; 45:449-50 8 Lamhert DR , Llaneza PP, Caglani RD , Lach RD , Beaver WL. Esophageal-atrial fistula. J Clin Gastroenteml 198.; 9:345-49 9 ltahashi 1111, Granada LO. Cerebral food embolism secondary to esophageal-cardiac perforation, JAMA 1972; 219:373-75 10 Larsen K, Jensen BS, Axelsen F. Perforation and rupture of the esophagus. Scand J Thorne Cardiovase Surg 1983; 17:311-16 11 Cooper BT, Barbe...at GO. Barrett's oesophagus: a clinical stud y CHEST I 103 I 5 I MAY, 1993
1605
of 52 patients. Q J Med 1987; 62:97-108 12 Kerlin P, D'Mellow G, Van Deth A. Barrett's esophagus: clinical , endoscopic, and histologic spectrum in fifty patients. Aust NZ J Med 1986; 16:198-205 13 Herlihy KJ, Orlando RC, Bryson JC, Bozyrnski EM, Carney CN, Powell D\V. Barrett's esophagus: clinical, endoscopic. histologic, manometric, and electrical potential difference characteristics. Gastroenterology 1984; 86 :436-43 14 Bladergroen MR, Lowe JE, Postlethwait RW Diagnosis and recommended management of esophageal perforation and rupture. Ann Thorac Surg 1986; 42;235-39 15 Flynn AE, Verrier ED, Way L'N, Thomas AN, Pellegrini CA. Esophageal perforation. Arch Surg 1989; 124:1211-15 16 Goldstein LA, Thompson WR. Esophageal perforations: a 15 year experience. Am J Surg 1982; 143:495-503 17 Pate J\v' Walker WA, Cole FH , Owen E\V, Johnson WHo Spontaneous rupture of the esophagus: a 30 year experience. Ann Thorac Surg 1989; 47:689-92 18 Rigg KM, Walker RW, Tension pneumothorax secondary to ruptured oesophageal diverticulum. Br J Clin Pract 1990; 44 : 528-29 19 Michel L, Grillo HC, Malt RA. Operative and nonoperative management of esophageal perforations. Ann Surg 1981; 194:5763 20 Faling LJ, Pugatch RD, Bobbins AH . The diagnosis of unsuspected esophageal perforation hy computed tomography, Am J Med Sci 1981; 281:31-4 21 Backer CL, LoCicero J, Hartz RS, Donaldson JS, Shields T. Computed tomography in patients with esophageal perforation. Chest 1990; 98:1078-80 22 Sherr Hp, Light RW, Merson Mil , WolfRO, Taylor LL, Hendrix TR . Origin of pleural fluid amylase in esophageal rupture , Ann Intern Med 1972; 76:985-86
Pneumomediastinum Caused by Subcutaneous Emphysema in the Shoulder* A Rare Complication of Arthroscopy Kam-Yung
[AU,
M.D . , EC.C.P.
Severe subcutaneous emphysema following arthroscopy of the shoulder developed in a 62-year-old man. It subsequently evolved into pneumomediastinum with respiratory distress, but the patient recovered spontaneously. To our knowledge, this constitutes the first report of pneumomediastinum caused by subcutaneous emphysema in the shoulder. (Chest 1993; 103:1606-07)
shoulder joint was distended with saline solution, of which the continuous inflow and outflow were regulated hy au arthroscopy infusion pump. Surgery and anesthesia were uneventful. hut toward the end of the 2-hour surgery, swelling and crepitation of the right ann and shoulder were noted. These rapidly spread to the right side of the chest, the neck, and the face and later also involved the left side as well . A chest radiograph (Fig I) demonstrated subcutaneous and mediastinal emphysema. Two hours after completion of surgery, the patient developed a fever of 38.4°(; . lie felt slightly short of breath and complained of sore throat and suhsternal chest pain . Arterial blood gas measurements, while the patient was breathing oxygen 2 Umin by nasal cannula, revealed that the PaO, was 57 mm IIg, the PaCO, was 38 mm IIg , and the I'll was 7.43 . With increase of Flo, to 40 percent (hy a Venti-mask), the PaO, increased to 81 mm IIg. Fiberoptic bronchoscopy disclosed no evidence of pharyngeal , laryngeal, or tracheal injury. Imipenem-cilastatln therapy W;lS started intravenously lie con tinued to run a fever fi.r the next 3 days and then defervesced . Blood and sputum cultures grew no microorganisms. Subsequent chest radiographs showed no development of pneumothorax. Both subcutaneous emphysema ami pneumomediastinum started to decrease on the third postoperative day and completely resolved on day 7. DISCUSSION
Air in the mediastinum can originate from fivl' sites : the neck , the alveoli , the airways, the esophagus, and the abdominal cavity. I In the case presented, there was no evidence of airway, esophageal, or abdominal injuries. The patient was not under prolonged or high positive pressure ventilation and therefore unlikely to have development of alveolar rupture . Had alveolar rupture occurred, giving rise to pneumomediastinum and subsequently subcutaneous emphysema, the latter would have been more symmetrical rather than commencing and remaining predominant around the right shoulder as occurred in this case . Thus, it appeared that the subcutaneous air in the neck, which originated from the right shoulder, was the most likely cause of this patient's pneumomediastin um. Subcutaneous emphysema in tilt' shoulder causing pneumomediastinum is, however, rather unusual. Deep cervical fascia divides the soft-tissue structures of the neck into three distinct compartments: the previsceral space, the visceral
S
ubcutaneous emphysema in the shoulder causing pneumomediastinum has not been described . I wish to report such a case, which occurred as a complication of arthroscopy of the shoulder. CASE REPORT
A 62-year-old man underwent arthroscopy of the right shoulder for debridement of the glenoid lahral tear. General anesthesia was induced, and the trachea was intuhated easily on the first attempt with an 8.0-mm oral endotracheal tube, It was then connected to a volume-cycled ventilator with a tidal volume of 600 ml and a rate of Hlmin. Peak inspiratory pressure was around 15 to 16 cm 11,0. The *From the Riverside Medical Clinic, Riverside, Calif. Reprint requests: Dr. [AU, .3660 Arlington Avenue, Riverside, CA 92506
1606
Fu.umc 1. Anteroposterior radiograph of the chest revealing the pneumomediastinum (lIrmws) and rndiolucent streaks of air in the subcutaneous tissue. Pneumomediastinum Caused by Subcutaneous Emphysema (K8n-Yung Lau)