doi:10.1378/chest.06-2843
(Chest. 2007; 131:1567-1569)
© 2007 American College of Chest Physicians
Elevated Glucose in Pleural Effusion*
An Early Clue to Esophageal Perforation
Khalid F. Almoosa, MD, FCCP;
Noel Wardell, MD and
Shahrokh Javaheri, MD, FCCP
* From the Department of Internal Medicine, Veterans’ Affairs Medical Center (Drs. Almoosa and Javaheri), Cincinnati, OH; and Kaiser Sunnyside Medical Center (Dr. Wardell), Cincinnati, OH.
Correspondence to: Shahrokh Javaheri, MD, FCCP, Pulmonary Section, 111F, VA Medical Center, 3200 Vine St, Cincinnati, OH 45220; e-mail: javaheri{at}snorenomore.com
 |
Abstract
|
|---|
Esophageal perforation is a rare cause of pleural effusions that carries with it a high mortality rate if diagnosis is delayed. The chemical characteristics of the pleural effusion include high amylase and low pH levels. However, these variables are nonspecific. We present a case of pleural effusion in a patient with markedly elevated glucose levels from recent cola ingestion as a specific clue to the diagnosis of esophageal perforation.
Key Words: esophageal perforation • glucose • lactobacillus • pleural effusion
|
Introduction
|
|---|
Chemical analysis of pleural fluid has proved to be useful in the diagnosis and management of a variety of disorders. The separation of fluid into an exudate or transudate can be accomplished with chemical analysis, including measurement of protein and lactate dehydrogenase (LDH) levels.1
A low pleural fluid glucose level has also been helpful in diagnosing certain effusions such as those due to rheumatoid arthritis and empyema.1 A high pleural fluid glucose level is a rare chemical finding in pleural fluid analysis. A few cases have been reported that were due to vascular erosion with leakage of parenteral nutrition fluid into the pleural space. We recently observed a patient with very high glucose levels in a pleural effusion, which led to our suspicion of esophageal rupture.
Esophageal rupture is a rare cause of pleural effusion that carries with it a high mortality rate if diagnosis is delayed.2 Clinically, diagnosis should be suspected if a pleural effusion develops soon after esophageal endoscopy, blunt chest or abdominal trauma, esophageal foreign-body removal, or severe vomiting. Chemical analysis of the effusion is helpful for diagnosis, showing an exudative process with high amylase levels and low pH levels.12 In the present case, a highly elevated pleural glucose level following cola ingestion increased our suspicion of esophageal perforation and served as an early clue to its diagnosis.
|
Case Reports
|
|---|
A 60-year-old man with unresectable esophageal cancer was admitted to the hospital for persistent fevers and pneumonia. He was undergoing radiation treatment and had recently completed his first cycle of cisplatin chemotherapy. At a routine follow-up clinic appointment, he was found to have a temperature of 39°C. He complained of general malaise and mild purulent sputum production. A chest roentgenogram showed a lobar infiltrate.
The patient was not in acute distress. Predominantly right bibasilar crackles were present with increased dullness to percussion and egophony changes. Heart sounds were normal without murmurs, rubs, or gallops. He was leukopenic with a WBC count of 1,700 cells/mL and an absolute neutrophil count of 1,386 cells/mL. Hemoglobin was 12 g/dL. A chest roentgenogram obtained at hospital admission revealed a right lower lobe infiltrate without pleural effusion, pneumomediastinum, or pneumothorax. A diagnosis of pneumonia was made; because of leukopenia he was treated with ceftazidime, gentamicin, and clindamycin.
Fever persisted, and 3 days later a repeat chest roentgenogram showed a new right pleural effusion with an air-fluid level (Fig 1
). It was thought that a parapneumonic effusion and an accompanying self-limited bronchopleural fistula had developed in the patient. The pleural fluid protein level was 0.7 g/dL (serum protein level, 4.5 g/dL; ratio, 0.2), and the LDH level was 65 U/L (serum LDH level, 106 U/L; ratio, 0.6). The pleural fluid glucose level, measured by spectrophotometry, was 843 mg/dL. The simultaneous serum glucose level was 115 mg/dL (ratio, 7.3). Measurements of the pleural fluid amylase level and pH were not obtained. Pleural fluid analysis also showed numerous WBCs and many budding yeast cells. Gram stain and culture findings showed the presence of Lactobacillus acidophilus and Candida albicans. Blood cultures grew L acidophilus. On further questioning by the pulmonary service, the patient stated that he had ingested a cola 2 h before thoracentesis. A repeat physical examination revealed no subcutaneous crepitus. Based on the clinical history of esophageal carcinoma, elevated pleural fluid glucose levels, the recent ingestion of a cola, and a new air-fluid level, esophageal perforation was suspected. A CT scan of the thorax confirmed distal esophageal perforation with fistula formation involving the right pleural space (Fig 2
). In order to confirm the oral source of organisms found in the pleural fluid, a mouth swab was obtained for a culture of L acidophilus and C albicans, which grew C albicans only. The patient refused surgical intervention and died several days later.
|
Discussion
|
|---|
Esophageal perforation is a rare cause of pleural effusions, and the resulting mortality correlates directly with the length of delay between the onset of symptoms and diagnosis and treatment.2 Early detection and intervention, therefore, are crucial to improving patient outcomes. Clinical signs and symptoms can be subtle and nonspecific. Chest pain is the most common presenting symptom, although dyspnea, cough, and hematemesis may also be present. Except for fever and tachycardia, the physical examination is usually unremarkable. Subcutaneous crepitation is characteristic but rarely present. Chest roentgenograms usually show a pleural effusion, but may also reveal mediastinal widening, subcutaneous and mediastinal air, or air-fluid levels. A CT scan often shows a pleuroesophageal tract, although contrast esophagography may become necessary.3
Analysis of pleural fluid, demonstrating high amylase and low pH levels, has traditionally been used to make the diagnosis in clinically suspected patients.1 Elevated pleural fluid amylase levels are of parotid gland origin4 and are highly suggestive of an esophageal perforation. Discriminating the amylase origin also requires the laboratory to measure isoamylase levels. Amylase levels, however, may be high in pleural effusions secondary to malignancy and pancreatitis.567 Similarly, a low pH is nonspecific, and may occur in a number of diseases causing pleural effusion, such as empyema, cancer, rheumatoid and lupus pleuritis, hemothorax, and urinothorax. Esophageal perforation was not initially suspected in our patient because the chest roentgenogram did not show any evidence of mediastinal or pleural pathology. The pleural effusion that appeared later was thought to be parapneumonic in nature. However, the finding of a highly elevated glucose level in the pleural fluid soon after oral ingestion of a high glucose load was a clue to the diagnosis, which was later confirmed by chest CT scan.
Pleural fluid glucose levels are generally not very helpful in identifying the cause of a pleural effusion, except for rheumatoid-associated effusions and empyemas in which the glucose level is characteristically very low.6 Despite Lactobacillus-infected pleural fluid and the presence of numerous WBCs, our patient had an extremely elevated pleural fluid glucose level of 843 mg/dL.
Glucose transport from blood to pleural fluid occurs mainly by simple diffusion across endothelial and mesothelial membranes, resulting in similar levels in both fluids.6 Mildly elevated pleural fluid glucose levels (approximately 200 mg/dL) have been observed in transudative effusions and are due to concomitant hyperglycemia.5 Our patient’s serum glucose level was 115 mg/dL. We therefore suspected leakage from an exogenous source.
Seven cases of elevated pleural fluid glucose levels (range, 174 to 916 mg/dL) have been reported,89 all in patients undergoing parenteral nutrition with vascular erosion. Since our patient did not have a central line and was normoglycemic, the high pleural fluid glucose level suggested a leakage from another source, and we suspected that esophageal perforation was present. On further questioning, the patient stated that he had ingested a can of cola shortly before thoracentesis, lending support to this theory. A 330-mL container of cola contains 96 g of glucose. Our patient had a glucose level of 8.43 g/L. We suspected that glucose from the cola leaked into the pleural space and created an osmotic gradient from the plasma to the pleural space. This gradient promoted the movement of water into the pleural space and consequently diluted the pleural fluid contents. In this context, we note that, despite the presence of lactobacilli together with many WBCs, concentrations of both pleural fluid protein and LDH were low. We propose that this was probably due to dilution.
The finding of Lactobacillus species in the pleural fluid is also consistent with an esophagopleural communication, as has been reported elsewhere.10 Lactobacillus species are nonsporulating, anaerobic, or facultative aerobic Gram-positive rods that frequently colonize the GI and urogenital tracts and seldom cause disease, even in the immunocompromised patient.11 Therefore, the detection of lactobacilli in an organ or body cavity outside of the GI or urogenital tract may suggest the presence of an abnormal communication between either of these cavities and the Lactobacillus-contaminated area. Since the esophagus is the only GI organ in the thoracic cavity, the presence of lactobacilli in the pleural space may indicate a pathologic communication with the esophageal lumen, most likely a perforation. If this occurs in the setting of high clinical suspicion, such as esophageal cancer in the case presented, then this finding may be diagnostic. Several previous case reports10 of lactobacillus empyema included patients with esophageal carcinoma and other esophageal pathology requiring endoscopic intervention.
The finding of Candida species in both the oral cavity and the pleural fluid of our patient also suggests a communication between the GI and pleural cavities. L acidophilus is a fastidious organism that is extremely difficult to grow from oral specimens. Therefore, it did not grow from the mouth specimen.
|
Conclusion
|
|---|
The finding of high glucose levels in the pleural fluid of our patient, coupled with the presence of GI tract-colonizing organisms, was the earliest clue to the diagnosis of esophageal perforation. Therefore, in the appropriate clinical setting, this finding is highly suggestive of esophageal rupture.
|
Footnotes
|
|---|
Abbreviation: LDH = lactate dehydrogenase
The authors have no conflicts of interest to disclose.
Received for publication November 22, 2006.
Accepted for publication January 9, 2007.
|
References
|
|---|
- Sahn, SA (1988) State of the art: the pleura. Am Rev Respir Dis 138,184-234[ISI][Medline]
- Attar, S, Hankins, JR, Suter, CM, et al Esophageal perforation: a therapeutic challenge. Ann Thorac Surg 1990;50,45-49[Abstract]
- White, CS, Templeton, PA, Attar, S Esophageal perforation: CT findings. AJR Am J Roentgenol 1993;160,767-770[Abstract/Free Full Text]
- Sherr, HP, Light, RW, Merson, MH, et al Origin of pleural fluid amylase in esophageal rupture. Ann Intern Med 1972;76,985-986[ISI][Medline]
- Light, RW, Ball, WC, Jr Glucose and amylase in pleural effusions. JAMA 1973;225,257-259[CrossRef][ISI][Medline]
- Light, RW Clinical manifestations and useful tests. Light, RW eds. Pleural diseases 2001,42-86 Lippincott Williams & Wilkins. Baltimore, MD:
- Light, RW Pleural effusion secondary to diseases of the gastrointestinal tract. Light, RW eds. Pleural diseases 2001,226-240 Lippincott Williams & Wilkins. Baltimore, MD:
- Bennett, MR, Chaudhry, RM, Owens, GR Elevated pleural fluid glucose: a risk for tension hydrothorax. South Med J 1986;79,1287-1289[CrossRef][ISI][Medline]
- Duntley, P, Siever, J, Korwes, ML, et al Vascular erosion by central venous catheters: clinical features and outcome. Chest 1992;101,1633-1638[Abstract/Free Full Text]
- Civen, R, Jousimies-Somer, H, Marina, M, et al A retrospective review of cases of anaerobic empyema and update of bacteriology. Clin Infect Dis 1995;20(suppl),S224-S229[ISI][Medline]
- Antony, SJ, Stratton, CW, Dummer, JS Lactobacillus bacteremia: description of the clinical course in adult patients without endocarditis. Clin Infect Dis 1996;23,773-778[ISI][Medline]
TOP
Abstract
Introduction
