Evaluation of complications of totally implantable central venous port system insertion.

The aim of the present study was to describe the procedure of totally implantable central venous port system (TICVPS) insertion performed at our center and investigate associated complications. The study retrospectively evaluated 827 patients who underwent a single-type TICVPS insertion from January 2013 to July 2015. The length of the procedure, long-term device function, angle (chamber-to-tip) and complications of TICVPS, including infection, skin erosion, occlusion, malposition and thrombosis, were analyzed from the patients' medical records. A total of 843 TICVPS insertions were performed in 827 patients. The TICVPS implantation was successful in all cases (100%). A total of 34 cases (4.0%) with complications were recorded. Complications at the chamber insertion site occurred in 11 patients (1.3%), including 5 infection (0.6%) and 6 erosion cases (0.7%). All patients with chamber insertion site infection were treated by administration of antibiotics and dressing. Of the patients in which chamber insertion site erosion occurred, 2 were subjected to TICVPS removal and reinsertion and 4 were treated with debridement, irrigation and resuture. The most common type of complication was catheter-associated (2.3%; n=19). Among these cases, 7 had catheter-associated infection (0.8%), 8 had catheter migration (1.0%) confirmed by chest radiography, 4 had catheter-associated thrombosis (0.5%) and 2 had chamber malposition (0.3%). The present retrospective study on TICVPS, which used a relatively large cohort, demonstrated a low complication rate (4.0%) compared with that reported in previous studies (5-20%). A well-designed procedure, experienced vascular surgeons, an aseptic operating room environment, ultrasound-guided puncture, a wide angle (chamber-to-tip) and the use of fluoroscopy with contrast agent may reduce the complication rate of TICVPS insertion.

Introduction

Niederhuber et al (1) reported on the first use of the totally implantable central venous port system (TICVPS) in 1982. Subsequently, the use of the TICVPS in patients undergoing chemotherapy, parenteral nutrition, intravenous injection, transfusion or repetitive laboratory analysis has increased (2,3). The TICVPS may reduce infection rates and thrombosis arising from recurrent puncture of the veins in patients with cancer (4,5). In addition, they also have no or only a minor impact on patients' daily activities, and cosmetic results after implantation are usually satisfactory (6–10). The TICVPS provides safe access to the central vein and long-term comfort and aesthetic satisfaction for patients who require long-term venous access.

The TICVPS is being implemented in >15 million patients per annum in the US, with associated complication rates ranging from 5–19% (11,12). Various types of TICVPS, which all provide venous access, are inserted by vascular surgeons, interventional radiologist and oncologists using a number of methods (13–19). Reported complications are mechanical complications during or directly after the insertion, including arterial puncture, nerve injury, hematoma and pneumothorax, and long-term complications, including infection and thrombosis (11). As the insertion techniques, management of the central venous port and catheter material are improved, the associated complications reduced, however when complications occur, the length of hospitalization and cost of medical care increase, and the rates of patient morbidity and mortality increase (20,21). Thus, it is important to reduce complications during or after TICVPS insertion.

The purpose of the present study was to describe the procedures for TICVPS insertion and to assess various postoperative complications associated with the TICVPS at our center. The present study also sought to determine how to reduce TICVPS-associated complications.

Patients and methods

Patients and data

The present study retrospectively reviewed data from patients who underwent TICVPS insertion between January 2013 and July 2015 at Pusan National University Yangsan Hospital (Yangsan, Republic of Korea). Exclusion criteria were as follows: Patients without fluoroscopy image results or follow-up not completed at Pusan National University Yangsan Hospital after TICVPS insertion. A single-type TICVPS (Districath; Districlass Medical SA, Chaponnay, France) was used. All procedures were performed by 2 vascular surgeons at a single center. Patient data and surgical radiographic imaging were collected from electronic medical records and a picture archiving and communication system tool (Marosis 5.4.10.71 PACS viewer; Marotech, Inc., Seoul, Korea), including operative and progress notes, as well as nursing records, in order to identify and record complications.

TICVP insertion

The TICVP insertion procedures were performed in the operating room, and the surgeons constantly monitored the electrocardiogram (ECG), oxygen saturation and blood pressure of patients during the procedure. Antibiotics (2 g flomoxef sodium) were intravenously administered as a prophylaxis prior to the procedure. The patient was placed in the supine position, and the neck was slightly turned to the side opposite to that of the procedure. Betadine was applied around the procedure site, which was aseptically draped. The right internal jugular vein was primarily selected as the access vein. The left internal jugular vein was considered the second access vein if the right internal jugular vein had an anatomical abnormality or in cases of right breast cancer. The subclavian vein was used if the two internal jugular veins could not be accessed. Once the access vein was determined, venous puncture was performed under ultrasonography and the guide wire was placed in the needle. Although the needle was removed and the wire fixed into the vein with mosquito forceps, a skin incision of ~2 cm was created for the pocket of the chamber at the deltopectoral region, below the clavicle. After making the pocket wide enough to insert the chamber (semicircle, ~2 cm in diameter), a subcutaneous tunnel was created between the puncture and pocket sites with a tunneler, which was connected at the catheter end. The chamber was then placed at the pocket site, and the catheter was cut to place the tip of the catheter at the cavoatrial junction and into the puncture site. Finally, the function of the TICVPS was confirmed by aspirating a small amount of blood from the chamber with a non-coring needle. The blood flow through the catheter, the catheter angle and the catheter tip position were checked by injecting a small amount of contrast media under fluoroscopy with a mobile C-arm (OEC9900 Elite; GE Healthcare, Little Chalfont, UK; Figs. 1 and 2). The chamber was sutured on the fascia of the pectoralis major muscle for fixation of the chamber in the pocket after each checkpoint was confirmed. The skin incision was sutured subcutaneously with monocryl 5-0 and reinforced with Steri strips, and dressing was applied. The success of the procedure was due to the following features showcasing good clinical practice: i) All procedures were performed in a clean operating room with monitoring, ii) ultrasonography and fluoroscopy were used for guidance, iii) the procedure was performed by experienced surgeons, iv) selection of a wide catheter angle (>60°) and v) subcutaneous suture with Steri Strips.

Figure 1.

Caterer insertion. The insertion of the catheter tip using fluoroscopy. The blue arrow demonstrated the inspection of the caterer tip position with the injection of a small amount of contrast media.

Figure 2.

Intraoperative fluoroscopy. The catheter angle was measured using fluoroscopy with a mobile C-arm.

Two surgeons had extensive experience with central venous catheterization (CVC) under ultrasonography (>100 cases annually during 4 years) and also performed endovascular surgery (>200 cases annually) under ultrasonography and fluoroscopy, and a similar operative technique was used for TICVPS placement. A vascular surgeon experienced in TICVPS was defined as having performed >400 CVCs and 200 TICVPs.

Post-insertion exam and follow-up

After the insertion, the patient had a 1-day post insertion check-up by the Surgeon who had performed the TICVPS insertion and the wound site was examined for any immediate complications. The team of surgeons initiated the use of the TICVPS to check its patency. Subsequently, the patients were followed up over 30 days post-insertion and any observations were added to the patients' medical records.

Analyses of the present study were performed by reviewing the electronic medical records of the patients. The definition of periprocedural complications was classified into immediate, early and late complications. Immediate complications are intra-procedural. Early complications were defined as complications that arise within 24 h, which are mostly procedure-associated, and also complications that occur within 30 days after the procedure. Late complications are those that are detected beyond 30 days of insertion. The complication rates published in previous studies vary in their type. The present study focused on various important periprocedural complications.

Results

A total of 843 TICVPS were inserted in 827 patients between January 2013 and July 2015. The procedure was successful in all cases. TICVPS insertion was performed twice in 16 patients in this study period. The demographic and clinical characteristics of the patients are listed in Table I. A total of 351 patients (42.4%) were males. The mean age of the patients was 58.2 years (range, 18–86 years), and 448 (54.2%) were below the age of 60 years. The average time the indwelling catheter was worn was 275.41 days (range, 1–782 days), and 325 patients (38.6%) had the indwelling catheter for >300 days. TICVPS insertion was performed under local and general anesthesia in 715 (84.8%) and 128 cases (15.2%), respectively. Patients under general anesthesia were simultaneously subjected to cancer surgery and TICVPS insertion. The most common TICVPS insertion sites were the right internal jugular, left internal jugular, right subclavian and left subclavian veins in 724 (85.9%), 113 (13.4%), 4 (0.5%) and 2 cases (0.2%), respectively. The mean catheter angle was 72.5° (range, 48.7–99.8°) and the mean body mass index of the patients was 23.0 kg/m2 (range, 13.2–44.4).

Table I.

Demographics and clinical characteristics of the patients (n=827).

Variable	Value
Sex
Male	351 (42.4)
Female	476 (57.6)
Age (years)	58.2±11.6 (18–86)
<60	448 (54.2)
≥60	379 (45.8)
Port implantation period (days)	275.4±146.5 (1–782)
<300	518 (61.4)
≥300	325 (38.6)
Anesthesia
Local	715 (84.8)
General	128 (15.2)
Port implantation site
Right internal jugular vein	724 (85.9)
Left internal jugular vein	113 (13.4)
Right subclavian vein	4 (0.5)
Left subclavian vein	2 (0.2)
Catheter angle (°)	72.5±20.0 (48.7–99.8)
Body mass index (kg/m2)	23.0±3.6 (13.2–44.4)
<17	24 (2.9)
≥17, <25	619 (74.4)
≥25	200 (23.7)
Underlying disease
Malignant solid tumor	766 (90.9)
Breast cancer	182 (21.6)
Gastrointestinal cancer	470 (55.8)
Gynecological cancer	44 (5.2)
Lung cancer	32 (3.8)
Other	38 (4.5)
Hematologic malignancy	71 (8.4)
Benign disease	6 (0.7)
Metastatic cancer	343 (40.7)

Values are expressed as n (%) or the mean ± standard deviation (range).

Solid tumors, hematologic cancers and benign tumors were present in 766 (90.9%), 71 (8.4%), and 6 patients (0.7%), respectively, wherein TICVPS was inserted for fluid resuscitation, parental nutrition or transfusion.

TICVPS-associated complications are described in Table II. A total of 34 (4.0%) complications were recorded. Catheter-associated complications were the most common type of complication, occurring in 19 cases (2.3%). Among these patients, 7 patients with catheter-associated infection underwent TICVPS removal and received antibiotics based on the results of the catheter tip culture. Catheter migration occurred in 8 patients, which was confirmed by chest radiography. Catheter tip malposition occurred in the neck vein in 6 cases, and extravenous catheter migration in 2 cases. Of these patients, 3 underwent removal and re-insertion of the catheter and 5 were subjected to re-positioning, including the 2 patients with extravenous catheter migration. The other 4 patients with catheter thrombosis underwent catheter removal (n=3) and anticoagulation therapy (n=1), including 1 patient whose catheter function was preserved.

Table II.

Complications of totally implantable central venous port system.

Type of complication	n (%)
Chamber site-associated	11 (1.3)
Infection	5 (0.6)
Erosion	6 (0.7)
Catheter-associated	19 (2.3)
Infection	7 (0.8)
Migration	8 (1.0)
Thrombosis	4 (0.5)
Other	4 (0.5)
Chamber malposition	2 (0.3)
Discomfort	1 (0.1)
Malfunction	1 (0.1)
Total	34 (4.0)

Complications of the chamber insertion site occurred in 11 cases (1.3%; 5 infections and 6 erosions). All infections were treated by administration of antibiotics and dressing. Of the 6 erosion cases, 2 underwent TICVPS removal and re-insertion, and 4 were treated with debridement, irrigation and restoration.

Chamber malposition occurred in 2 patients and they underwent chamber repositioning. One patient had discomfort at the insertion site and requested TICVPS removal. TICVPS malfunction occurred in 1 case. Repositioning was performed first, but malfunction was not resolved; hence, the TICVPS was removed and another one was inserted.

Discussion

Insertion of the TICVPS is predominantly carried out by surgeons, who perform venous cut down or use anatomic landmarks to identify a suitable entry site. However, certain interventional radiologists perform image-assisted percutaneous TICVPS insertion using ultrasound guidance with the Seldinger technique at the access site and fluoroscopy to check the catheter placement with good success rates comparable to those of the surgeons (11). No significant differences in the rate of infection between the angiographic suite and the operation theater, were reported, with P<0.743. When intervention radiologists performed TICVPS, the infection rate was relatively high (>5%) (11). In the present study, the overall complication rate was 4.0%, with no mortality recorded in 843 cases of TICVPS insertion. In terms of the complication rate, the present results are superior to those reported in other studies (5–19%). A comparison of complications reported in various studies is presented in Table III (12,17,20,22). All patients of the present study underwent the insertion procedure performed according to good clinical practice, as mentioned above. Proper positioning of the catheter tip is important, as shortcomings thereof are associated with TICVPS malfunction and catheter-associated complications. In cases of catheter migration or catheter tip malposition, catheter malfunction, pain or swelling may occur. Son et al (23) reported that the risk for catheter thrombosis was high when the catheter tip was above the superior vena cava.

Table III.

Comparison of the frequency of complications (%) between different studies.

Complication	Teichgräber et al (17) (n=3,160)	Babu et al (22) (n=180)	Kim et al (20) (n=179)	McGee and Gould (12) (review)	Present study (n=827)
Pneumothorax	0	3.7		<0.1–0.2	0
Bleeding Hematoma	0.2	7.4		<0.1–2.2	0.2
Malposition	0	5.6			0.1
Deep vein thrombosis	0.5	4.6	4.5	4.3	0.4
Pain	0.2				0.1
Allergic reaction	0.02				0
Catheter associated bloodstream infection	5.1	18.8	12.8	13.6	0.8
Pocket infection	0.3				0.6
Migration	0.6	10	10		1
Skin erosion	0.2	0	0	9.3	0.7
Accidental dislodgement	0.4	10.2	4.5	7.5	0.2
Malfunction	0.1			4.6	0.01

Schenck et al (24) recommended using intra-atrial ECG techniques to determine the catheter tip position. In this study, the electrical current transducer was connected to the catheter and the cable attached to lead II of a standard ECG monitor. The catheter was slowly advanced by monitoring the morphological changes of the P wave until the tip reached the desired position. The catheter tip is close to the sinoatrial node (i.e., in the upper part of the right atrium) when the P wave reaches its maximum height. One centimeter above this point, when the P wave is at half of its maximal height, the tip is close to the atriocaval junction. It may be justified to perform delayed postoperative chest radiography to confirm central venous catheter line tip placement.

The acceptable complication rate observed in the present study (4.0%) may be due to several factors. The position of the catheter tip was determined using fluoroscopy with contrast media during the procedure. The blood flow in the catheter, as well as the catheter angle and catheter tip position were checked using fluoroscopy with contrast media. Prior to the end of the procedure, the catheter tip position and angle were adjusted if it was not appropriately placed.

Complications at the chamber insertion site included infection and erosion. Infection of the chamber insertion site comprised erythema, tenderness and occasional discharge (25). In the present study, 5 patients with chamber insertion site infection presented with redness, swelling and pain, and they were administered antibiotics.

Skin erosion at the chamber insertion site is a rare long-term complication. The skin overlying the chamber generally breaks down, exposing the device in the subcutaneous space (26,27). Skin erosion is a gradual process, which results in infection. This may manifest systemically as a fever with chills and/or locally with discharge or abscess (28). However, erosion without infection has also been documented (29). Incision site tension, repeated abrasion or repeated needle puncture may also result in skin erosion.

A recent study suggested that the incidence of skin erosion was 1% (30). The incidence of skin erosion in the present study was 0.7% (6/843). The pocket was bigger than the chamber (semicircle, ~2 cm in diameter) to reduce the tension of the skin incision after chamber insertion. A thick skin flap was also created to withstand repeated puncture and weight loss. Patients undergoing chemotherapy are more likely to lose weight due to chemotherapy-associated side effects, and therefore, the use of a thick skin flap is appropriate in these patients. The subcutaneous skin incision site was sutured with monocryl by a well-trained surgeon and reinforced with an aseptic Steri strip to reduce skin infections and dehiscence. A previous study reported that subcutaneous suture closure of the incision site reduced wound disruption (31). The methods mentioned above may have resulted in a lower incidence of skin erosion and lower complication rates in the present study.

In the present study, all patients underwent TICVPS in a hybrid operation room, which fulfilled aseptic criteria for a standard surgical room and imaging equipment from a mobile C arm or angio suite (32), which may have contributed to reducing sources of infection.

Catheter-associated thrombosis may occur spontaneously or from a prothrombotic state associated with an underlying malignancy or treatment (25). The association between cancer and thrombosis arises as a consequence of cancer treatment and direct vessel trauma, which is a result of long-term central venous catheter placement (30). Thrombosis may cause several symptoms associated with loss of catheter function, including an increased risk of infection, pulmonary embolism and postphlebitic syndrome; it is also associated with greater cost (33). Catheter-associated thrombosis has a reported incidence of 0.3–28.3% (17,34,35). In the present study, the incidence of catheter-associated thrombosis was 0.5% (4/843). At our center, the catheter angle was constantly checked using fluoroscopy and it was attempted to adjust the catheter angle to >60° during the procedure. A sharp catheter angle causes poor blood flow in the catheter, and thrombosis may easily occur. At our center, the right internal jugular vein was primarily selected as the access vein, as the jugular vein has a lower risk for catheter-associated thrombosis than the subclavian vein (36), and the right internal jugular vein provides direct access to the superior vena cava (37). The left internal jugular vein was considered as the secondary access vein if the right internal jugular vein had an anatomical abnormality or in patients with right breast cancer, due to radical axillary lymph node dissection and postoperative radiotherapy (38). The creation of a catheter angle of >60° and selection of the appropriate access vein during the procedure may have reduced the occurrence of catheter-associated thrombosis in our center.

Patients were diagnosed with catheter-associated infections if they had at least 2 positive blood culture results, obtained from at least 2 separate sites at different times, with evidence of colonization of the catheter with the same organism. The fulfillment of the latter part of the definition may only be determined by removing the catheter (25). In the present study, 7 (0.8%) cases of catheter-associated infection occurred. In a previous study, the overall incidence of catheter-associated infection was reported as 0–6.8% (39). To reduce infection-associated complications, TICVPS insertion was performed in the operating room under aseptic conditions. All healthcare professionals who participated in the procedure wore surgical gowns and it was attempted to minimize the length of the surgery.

In the present study, no periprocedural complication, e.g., pneumothorax, was recorded. Ultrasound-guided puncture reduces the rate of these complications. Port site discomfort was recorded in 1 patient, which may have been caused by nerve injury. Potential early complications, including pneumothorax, air embolism or arterial puncture may be fatal, but these did not occur in the present study (40,41).

In conclusion, low complication rates of TICVPS insertion were observed in the present, large, retrospective study. Complication rates may be reduced by using a well-designed procedure, experienced vascular surgeons, an aseptic environment, ultrasound-guided puncture and fluoroscopy with contrast media.