Total parenteral nutrition (TPN) via central venous catheter (CVC) is widely used for nutritional support in clinical practice. Vascular erosion by CVCs used for TPN is reported in both neonate and adult population [1-4]. The incidence of this complication is reported as 0.17 to 0.4% per catheterization [1-3]. Previous studies and case series suggest that almost all erosions are associated with the left-sided catheter placement [1-3]. Because of rare occurrence, systematic review on this topic is not suitable, this review will focus on reported cohort study, case series, case reports and autopsy studies in adult patients [1-3,5-7].

Using search terms “central venous catheterization”, “pleural effusion”, “total parenteral nutrition”, Pubmed was interrogated to October, 30, 2024 to identify relevant papers on this topic. In addition, the reference lists of review papers and case reports were searched for relevant articles that may have been missed during the electronic search.

1. Incidence and demographic risk factors

To quantify the incidence of vascular erosion is difficult because vast majority of information is based on case reports and series. One study which involved 1,058 catheters, had an incidence of 0.4% [2]. The largest cohort study evaluating records of 1,499 patients (2,992 catheters) over the 14 year period showed incidence of 0.17% per catheter or 0.28 per 1,000 catheter days [3].

Due to smaller vessel size, female gender may be a risk factor for erosion [2], four of the five vascular erosion occurred in female patients in the largest cohort which were not statistically significant (p=0.18) [3]. However, it found that older age was a statistically significant risk factor (p=0.009) [3].

2. Details of CVC placement

Most of reported vascular erosion were associated with conventional CVC insertion [1-3,5-7]. Recently, cases involving patients who received peripherally inserted central catheter (PICC) were also reported in both neonate and adults [7-10].

Almost all vascular erosion occurred in patients with left-side approach [1-3,5,6,8,10,11]. Relative risk of vascular erosion occurring in left-side approach was 2.9 (95% CI 2.76–3.0; p=0.009) [3]. One reported right-side vascular erosion which occurred in patient with a history of implantable cardioverter-defibrillator (ICD) placement causing difficult insertion of PICC [8].

Catheters made of polyethylene are known to be stiffer than that made of polyurethane [12]. Three of five vascular erosions that occurred in one study were by catheters made by polyethylene [3]. Despite improvement in catheter materials vascular erosions continue to be reported [5,9,11].

All catheters that caused vascular erosion were 15 cm long and might pose a risk of lying at an acute angle to the superior vena cava (SVC) wall, typically during left-side approach [3]. In order to prevent catheter-related complications, one suggest that 20 cm long catheters should be used for left-sided insertion [13].

Studies also found that 16 G or smaller catheter size may be associated with lower incidence of vascular erosion [1,3,12].

3. Mechanism of vascular erosion by CVCs used for total parenteral nutrition

The exact pathogenesis of vascular erosion by CVCs are unclear. Previous studies speculated that left-sided catheters is liable to abut the right wall of the SVC at a sharp angle [1]. Impaction against blood vessel wall is postulated to result in endothelial damage and subsequent vascular erosion which can lead to leakage of infused fluids into mediastinum and/or pleural cavities [2,3,14-16]. Migration of catheters from patient movement, neck flexion with improper securement of catheters and positional change may cause more acute angle abutting SVC wall [3,9,11,17]. One autopsy study revealed a possible mark from the catheter tip in the left brachiocephalic vein (BCV) with a surrounding hematoma in the vessel wall and no evident external soft tissue hematoma in the area of the left BCV [9]. Given vascular erosion causing micro-vascular injury rather than macro-vascular damage, in particular, hyperosmolar fluids, such as TPN can cause osmotic injury, leading to vascular leakage despite the intact venous placement of the catheter [5,17,18]. This may partially explain the reason why right-sided extravasation and TPN leakage from non-sharp angled CVCs occurred in previous studies [1,8]. The sharp angle of the CVC to the SVC wall, the use of a hyperosmolar solution, and endothelial damage with thrombus formation have been suggested as a mechanism of TPN pleural effusion without hemothorax occurrence following CVC removal [19].

4. Symptoms and signs of vascular erosion

Dyspnea, chest pain and palpitations were reported symptoms [2,3,5,9,11]. Two case reported patients who suffered from severe respiratory distress requiring endotracheal intubation and mechanical ventilation [18,19]. Most of patients showed signs of pleural effusion [2,3,5,11,18,19]. Although ipsilateral pleural effusions were more frequently reported, bilateral pleural effusion was also developed [6,9,10,20]. In cases of bilateral pleural effusions, computerized tomography of the chest showed pneumomediastinum and increased density within mediastinal fat tissue in the anterior mediastinum suggesting chemically induced mediastinitis [18-20].

The onset of symptoms by extravasation appears 8 hours to 3.6 days after CVC placement [1-3,5,6,8]. However, delayed presentation after 6 to 10 days after insertion did occur as shown in previous reports [1,19,20]. Patient characteristics, type of catheter, time to onset of symptom, site of extravasation of total parenteral nutrition and mortality in the selected references are shown in Table 1.

Table 1 . Patient characteristics, type of catheter, time to onset of symptom, site of extravasation of total parenteral nutrition and mortality in the references.

Reference number	Number and gender of patients (male/female)	Type of catheter	Time to symptom after TPN infusion (days)	Right/left sided pleural effusion	Mortality (%)
Mukau [1]	4 (1/3)	CVC	1.6 (mean)	0/4	0
Duntley [2]	8	CVC	3.0±1.5 (mean)	1/7	12.5
Walshe [3]	5 (1/4)	CVC	3.6 (mean)	0/5	20
Paw [6]	1 (1/0)	CVC	2	Bilateral PE	0
Yang [8]	1 (0/1)	PICC	4	Bilateral PE	0
Siddiqui [10]	1 (0/1)	PICC	2	Bilateral PE	0
Inaba [19]	1 (0/1)	CVC	8	1/0	0
Byeon [20]	1 (0/1)	CVC	6	Bilateral PE	0

CVC: central venous catheter, TPN: total parenteral nutrition, PICC: peripherally inserted central catheter, PE: pleural effusion..

5. Diagnosis of vascular erosion

Since symptoms and signs of extravasation are non-specific, the diagnosis often delayed, which may contribute increased morbidity and mortality. Plural fluid analysis typically showed ‘milky white appearance’ and increased level of glucose, triglyceride, and potassium [3,6,8,9,11,17]. Biochemical analysis of the plural effusion and the composition of TPN revealed the presence of chylomicron in both fluids [5]. As high triglyceride level can be seen in both chylothorax and TPN extravasation, high glucose and potassium level led to the diagnosis of TPN leakage [11].

6. Treatment of vascular erosion by CVCs used for total parenteral nutrition

Once diagnosis is made, TPN should be discontinued immediately and affected CVCs removed. Pleural drainage effectively relieve symptoms and operative repair is seldom needed in case of penetration of the left subclavian vein [10].

Extravasation of TPN from vascular erosion of central vein is rare. Due to non-specific symptoms, a high index of suspicion for this possibility is warranted for the timely diagnosis, especially if new respiratory symptoms are developed following commencement of TPN via CVCs. Elevated levels of glucose and potassium in the pleural fluid analysis is helpful for the diagnosis. It is treated by discontinuation of TPN, removal of CVCs, and pleural drainage.

The author declares no conflicts of interest.