A 66-Year-Old Man with Intrapleural Leakage of Parenteral Nutrition Fluid Due to a Malpositioned Central Venous Catheter

Introduction

Pleural effusion is a common finding in patients admitted to the Intensive Care Unit (ICU) [1]. Although the differential diagnosis of pleural effusion is broad, less common causes should be considered in the appropriate clinical context. When a milky fluid is drained during thoracocentesis, the differential diagnosis typically includes chylothorax, pseudochylothorax, and empyema [2].

Chylothorax is defined as the presence of chyle in the pleural cavity [3]. Chyle is a milky-white fluid composed of lymph and fats produced in the small intestine during digestion and transported via lymph vessels [3]. Chylothorax is characterized by a pleural fluid triglyceride level greater than 110 mg/dL (1.24 mmol/L) and a cholesterol level below 200 mg/dL (5.18 mmol/L) [3]. Non-traumatic causes of chylothorax include congenital abnormalities, malignancy, such as lymphoma, infection, such as tuberculosis, and systemic conditions, such as systemic lupus erythematosus [4].

In contrast, pseudochylothorax is a cholesterol-rich pleural effusion with pleural cholesterol levels above 200 mg/dL (5.18 mmol/L) and triglyceride levels below 110 mg/dL (1.24 mmol/L) [4]. It is most frequently associated with chronic inflammatory disorders, such as chronic tuberculous pleuritis and rheumatoid arthritis [4].

In patients receiving total parenteral nutrition (TPN), another potential cause of a milky pleural effusion is TPN leakage through a malfunctioning or malpositioned central venous catheter (CVC) [5]. Among patients in the ICU, TPN is frequently administered when oral or enteral feeding is not feasible [6]. The essential components of TPN are carbohydrates, lipids, amino acids, vitamins, trace elements, electrolytes, and water. Because of its hypertonicity, it is typically delivered through a central line to prevent vascular irritation [6]. However, if a CVC is malpositioned, particularly with the catheter tip abutting or eroding the vessel wall, the hyperosmolar TPN solution can lead to vascular erosion and extravasation [5]. This rare complication is one among several potential complications associated with TPN administration and CVC use. TPN complications can be classified as metabolic, such as hyperglycemia, hypertriglyceridemia, electrolyte imbalance, refeeding syndrome, and hepatic dysfunction, infectious complications, and mechanical complications related to the venous access, such as arterial puncture, pneumothorax, and thrombosis [6–8].

Here, we describe the case of a 66-year-old man with bilateral intrapleural leak of TPN fluid, due to a malpositioned CVC.

Case Report

A 66-year-old man with a past medical history of chronic obstructive pulmonary disease and cholecystectomy with choledochotomy for choledocholithiasis, was admitted to the ICU with a diagnosis of septic shock due to cholangitis. Broad spectrum antibiotics and fluid resuscitation were initiated in the Emergency Department. Despite these measures, the patient’s condition progressed to septic shock, and he required vasopressor support and mechanical ventilation. Therefore, a CVC was placed into the left internal jugular vein under ultrasound guidance. Correct positioning of the catheter tip was confirmed with a plain film chest X-ray (Figure 1). The left internal jugular vein was chosen to preserve the right internal jugular vein for potential placement of a dialysis catheter, given the patient’s acute kidney injury on admission (serum creatinine 4.58 mg/dL, with an estimated glomerular filtration rate of 14 mL/min/1.73 m2), which, in case of progression, might have resulted in the need for renal replacement therapy.

Blood cultures grew Escherichia coli and Enterobacter cloacae. An endoscopic retrograde cholangiopancreatography with stent-placement in the common bile duct was performed. Because of gastrointestinal intolerance to enteral feeding, TPN was initiated on day 4 after admission. Routine chest X-ray on day 5 showed bilateral pleural fluid accumulation. At that time, the patient was afebrile and no longer required vasopressor support, although he remained intubated. Physical examination revealed bilateral diminished breath sounds at the lung bases, without evidence of peripheral edema. Laboratory results showed a C-reactive protein level of 49 mg/L (255 mg/L on admission; reference range <5 mg/L), total bilirubin level of 5.9 mg/dL (6.3 mg/dL on admission; reference range 0.67–1.17 mg/dL), total protein level of 62 g/L (reference range 60–80 g/L), and serum creatinine level of 3.08 mg/dL, with an estimated glomerular filtration rate of 20 mL/min/1.73 m2. Echocardiography showed a normal left ventricular ejection fraction of 60%, a non-dilated inferior vena cava, and a pericardial effusion measuring up to 1 cm adjacent to the right ventricle, without clinical or echocardiographic signs of cardiac tamponade. An 8.3 French pigtail catheter was placed in the left pleural space, immediately draining 2.5 L of a white, milky fluid (Figure 2). Analysis of the fluid revealed a transudate according to Light’s criteria (effusion protein 18 g/L, ratio to serum protein <0.5; effusion LDH 62 U/L, ratio to serum LDH <0.6) with triglyceride levels of 789 mg/dL, cholesterol <7 mg/dL, and glucose level of 772 mg/dL (serum glucose 119 mg/dL). Subsequent drainage of the right-sided pleural effusion yielded similar findings.

The composition of the pleural fluid, characterized by elevated triglyceride and glucose levels, matched the composition of the TPN the patient had been receiving. The diagnosis of TPN leakage was further supported by contrast injection through the CVC under X-ray guidance, which demonstrated contrast extravasation in the upper mediastinum, just caudal of the catheter tip (Figure 3). Subsequently, the CVC was removed. Both chest drains were successfully removed 2 days after their placement, as they had ceased to drain fluid. The patient could be extubated shortly thereafter.

A follow-up chest X-ray after removal of the drains confirmed resolution of the bilateral pleural effusions (Figure 4). On echocardiography, the pericardial effusion had decreased. Since our patient showed no signs of cardiac tamponade, pericardiocentesis was not performed, leaving the relationship between the pericardial effusion and parenteral nutrition leakage uncertain. Chest computed tomography, performed after drainage of the pleural effusions and after removal of the CVC, showed no evidence of mediastinitis or pneumomediastinum. The patient was discharged from the ICU. Unfortunately, he died 1 month later due to an exacerbation of chronic obstructive pulmonary disease.

Discussion

A pleural effusion with triglyceride levels >110 mg/dL in a patient receiving TPN can indicate intrapleural leakage of TPN. This possibility should prompt clinicians to discontinue TPN and perform additional analyses, including glucose and potassium level measurements.

Thoracocentesis in our patient revealed a white, milky fluid with triglyceride levels >110 mg/dL, which may have resulted in mistaking the TPN-related pleural effusion for chylothorax. The clue to diagnosing a TPN leakage-related pleural effusion is to add glucose and potassium measurement to the biochemical work-up [9]. Unfortunately, we did not measure pleural fluid potassium in our patient. In contrast to chylothorax, TPN leakage pleural effusion is characterized by elevated glucose and potassium levels, with a pleural fluid to serum glucose/potassium ratio >1 and with a pleural fluid composition similar to the administered TPN [10].

To support the diagnosis of leakage of TPN, a chest X-ray was performed, while injecting intravenous contrast in the CVC, which confirmed contrast extravasation in the mediastinum. To confirm the diagnosis of leakage into the pleural space, methylene blue could have been injected in the CVC to check for blue discoloration of the pleural fluid, whereby leakage not only into the mediastinum, but also in the pleural space could have been confirmed.

TPN-related pleural effusion can occur if a catheter perforates the vein wall directly by mechanical force (mostly seen in subclavian vein catheterization) or indirectly through hyperosmolality-induced endothelial damage and vascular erosion caused by the hyperosmotic parenteral nutrition [5]. The catheter can be correctly positioned initially, but perforate the vein wall after migration of the catheter [10].

The incidence of TPN-related pleural effusion is not well known. In a review of prospectively collected intravenous nutrition records in a tertiary hospital, the incidence of vascular erosion in patients receiving parenteral nutrition through a CVC was 0.17% per inserted catheter (2992 catheters over a 14-year period) [5]. The incidence was higher in left-sided catheters, with a relative risk of 2.9, as in our case. Left-sided catheters seem to have a higher risk of erosion due to their sharper angle entering the superior vena cava, which increases vessel wall injury [5]. Preferential use of right-sided central access can be considered where possible. Catheter migration from patient neck movement can worsen the risk of vascular erosion if the catheter is not securely positioned [5]. The average indwelling time of the CVC in parenteral nutrition leakage pleural effusion was 5 to 8 days, and onset of symptoms occurred at a mean of 3.6 days after catheter insertion [5]. TPN intrapleural leakage was also described with peripherally inserted central catheters [11]. Mortality rates of 12.5% to 20% have been reported in 2 small case series [5,12].

In the majority of previously described cases, the pleural effusion was bilateral [2,5,9,13–16], as was the case in our patient, which is remarkable, as it is generally believed that in humans there are no connections or fenestrations between both pleural cavities. This finding suggests the existence of interpleural communications, at least in some patients. However, the exact route of the TPN leakage remains unclear, whether from the mediastinum into both pleural spaces, from the left to the right pleural space via interpleural connections, or from other routes.

The treatment of TPN-related pleural effusion consists of immediate discontinuation of parenteral nutrition administration, removal of the catheter and pleural drainage [16].

Conclusions

In patients with a CVC receiving parenteral nutrition, a white, milky pleural effusion can be caused by TPN leakage in the pleural cavity. Elevated glucose and potassium levels in the fluid, with a comparable composition to the TPN, can be useful clues to differentiate from a chylous effusion.