Technical Aspects

Initially, surgically placed subcutaneous ports were inserted via a blind percutaneous subclavian vein approach³ or by cutdown.¹⁵ Although central lines placed by interventional radiologists were also initially placed via the subclavian vein under fluoroscopic guidance,¹⁶ insertion of ports soon evolved from an ultrasound-guided subclavian vein approach⁴ to the preferred ultrasound-guided IJ vein approach.^5,6,17 Ultrasound-guided IJ access has become the preferred approach because it allows an easily compressible access (in the rare case of inadvertent arterial puncture), avoids the higher risk for pneumothorax associated with axillary/subclavian vein access,¹⁸ and avoids the “pinch-off syndrome” associated with medially placed subclavian catheters.¹⁹

Review of the patient’s history and any available relevant previous imaging studies is paramount. The goal is to uncover any evidence of central venous stenosis or occlusion. A history of multiple central venous lines or end-stage renal disease should be queried. Access via the subclavian vein should be avoided in dialysis patients in an attempt to preserve patency of the axillary and subclavian veins for any current or future surgical dialysis access (fistulas or shunts). Although it has traditionally been recommended to avoid the ipsilateral approach in patients who have undergone axillary lymph node dissection, it has been shown that the incidence of lymphedema after ipsilateral port placement is no higher than that reported after axillary lymph node dissection alone.²⁰ In addition, any thoracic malignancy requiring external beam irradiation should be considered and the planned radiation field avoided.

The technique described here is a minor modification of that previously described by Mauro, Jaques, and Morris.4,5 A routine procedure tray is augmented with a cutdown tray (see Table 120-1 and Fig. e120-1). The neck should be interrogated with ultrasound to confirm patency of the IJ vein, which is supported by easy compressibility of the vein and lack of adjacent collaterals. If ultrasound or previous imaging indicates that the IJ veins are occluded, the subclavian veins are a reasonable alternative. If there is complete central venous occlusion involving the brachiocephalic veins or superior vena cava (SVC), a femoral, translumbar, or transhepatic approach can be considered. Although these approaches are rarely required, they should be part of an interventionalist’s armamentarium.

The upper portion of the chest and neck are then prepared and draped in sterile fashion. Povidone-iodine solution has traditionally been used, but many institutions have switched to 2% chlorhexidine gluconate in 70% isopropyl alchohol (Chloraprep).²¹ The ultrasound probe is then also draped so it can be used for direct imaging during venous access. The procedure is typically performed with the patient under conscious sedation and continual monitoring of the electrocardiogram and pulse oximetry by a member of the radiology nursing staff. Local anesthesia is achieved at the entry site with 1% lidocaine with epinephrine, and this can be done under ultrasound guidance to plan the path of the access needle and avoid intravascular injection of lidocaine (Fig. 120-1). A small skin incision is made with a #11 blade. Through this small incision, a 21-gauge, 4-cm needle (some obese patients may require a 21-gauge, 7-cm needle) connected to a 20-mL syringe by short connector tubing is advanced in a medial and caudal direction under direct ultrasound guidance (Fig. 120-2). The operator should keep the needle tip visualized throughout the access process. Some operators prefer an 18-gauge needle to eliminate the need for a transition dilator. The operator can often optimize the angle of approach to avoid placing the carotid artery in the path of the trajectory of the needle. Gently “bobbing” the needle may improve localization of the needle tip. Once the tip of the needle “tents” the vein, a short, quick thrust will advance the needle into the vein. A tactile “pop” is often felt as the needle enters the vein. Again, with the aid of ultrasound guidance, care can be taken to avoid going completely through the vein into the carotid artery. Insertion of the tip of the needle within the vein is confirmed by observing the bright tip of the needle within the vein (Fig. 120-3) and by aspiration of blood through the 20-mL syringe. At times, the patient’s respirations will intermittently collapse the vein. In such cases, puncturing the vein during expiration will usually allow successful access. A 0.018-inch mandril wire is then advanced into the vein under fluoroscopic guidance so the access needle can be exchanged for a Cope transition dilator (Cook Medical, Bloomington, Ind.). Some operators prefer a micropuncture set (Cook Medical) to make the transition to a 0.035-inch system. The mandril wire is used to measure the distance from the access to the proximal aspect to midportion of the right atrium, and this access is then temporarily capped. The intravascular distance can be measured by placing the tip of the mandril wire in the desired location between the cavoatrial junction and the cephalad aspect of the right atrium, bending the exposed wire as it exits the transition dilator, and then pulling the wire back to the point of access in the vein, which can be visualized under fluoroscopy with a radiopaque marker at the skin. The wire should then be stored in a safe place on the table so it can be used for measurement after the catheter is tunneled to the right neck dermatotomy site.

Next, the tunnel and port pocket are planned. The port pocket should be positioned in the infraclavicular region a few centimeters below the clavicle and medial to the deltopectoral groove. The region of the port pocket and planned tunnel is then anesthetized with 1% lidocaine with epinephrine. The port pocket should allow firm bony support by the underlying ribs and yet avoid mammary tissue. In addition, the tunnel should have a gentle curve so the catheter does not kink. To allow a few minutes for the lidocaine to take effect, the catheter and port can be assembled and secured at this time. The junction of the port and catheter should be tested by accessing the port with a Huber needle, flushing the catheter, and pinching off the distal end of the catheter to challenge the catheter/port junction and confirm the absence of a leak.

An incision is then made with a #15 blade, just long enough to allow insertion of the port. Though originally described as an incision caudal to the port pocket,⁴ we now make the insertion incision just cephalad to the pocket. This approach allows easy access to the catheter/port junction and gives some flexibility if last-minute tailoring of catheter length is necessary. However, some practitioners still prefer to use the inferior incision approach. Usually the port pocket can be easily created by blunt dissection with a Kelly hemostat and small tracheal retractors or simply with the operator’s finger (Fig. 120-4). This allows any fibrous bands encountered to be freed and gross estimation of the pocket size required (the length of one’s distal phalanx is approximately the depth required for many ports). The pocket should be large enough to allow the port to be inserted easily, the incision to be closed without tension on the skin, and the incision to not overlie the diaphragm of the port. If a snug fit is achieved, anchoring sutures are not usually needed.²² However, if the port pocket is too large or if the patient is obese, it is prudent to place anchoring sutures to avoid the complication of the port flipping within the pocket and thereby prohibiting access (Fig. 120-5). Usually, two nonabsorbable sutures adequately anchor the port.

Once the port pocket is created, the catheter is tunneled from the infraclavicular incision to the right neck dermatotomy. The metal tunneling device is then shaped with a gentle curve (Fig. e120-2) to facilitate following a cephalad then medial path. The operator should initially tunnel cephalad, then use the curve of the tunneler to turn medially, all the while keeping a finger on the leading point of the tunneler to avoid inadvertent injury to surrounding structures. Attention to the tip of the tunneler is of particular importance when it is passing over the clavicle, to ensure it does in fact pass superficial to the clavicle and avoids an intrathoracic trajectory. Once the catheter is brought through the tunnel (Fig. 120-6), the port is positioned within the pocket. If anchoring sutures are going to be used, they can be placed at this time. The port may be held in place by accessing it with a noncoring Huber needle. The catheter is then cut to length with the previously marked mandril wire used for measuring the distance. The Cope transition dilator is then exchanged over a 0.035-inch wire for a peel-away sheath.

Air embolism can be avoided by pinching the sheath while introducing the catheter into the sheath or by covering the open end of the peel-away sheath (Fig. 120-7) until the catheter is quickly inserted into the sheath. The catheter can then be advanced through the peel-away sheath (Fig. 120-8) and positioned with its tip in the region between the cavoatrial junction and the cephalad aspect of the right atrium. Dialysis catheter manufacturers have developed valved peel-away sheaths to minimize the risk of air embolism and diminish blood loss (AirGuard Valved Introducer [Bard Access Systems, Salt Lake City, Utah]). This technology is now available for the smaller-caliber sheaths used for port insertion. Other maneuvers may be used, such as having patients hold their breath or hum; however, following such simple commands may prove difficult, depending on the degree of sedation at that time. Attention to the patient’s respiratory cycle is critical at this step so that occlusion of the sheath is not released at the point of inspiration. The catheter should not be advanced through the peel-away sheath if the patient is agitated, coughing, or crying, because an air embolus could easily result. The length of the catheter should be checked before the entire peel-away sheath is removed. If the catheter is too long or short, this should quickly be remedied at this time. Placement of a new peel-away sheath, which could be temporarily capped while the appropriate alterations are made, may be required.

Once the catheter is in appropriate position, aspiration of blood followed by flushing with normal saline confirms patency. This is followed by 5 mL of dilute heparin flush (100 units/mL) unless the port is a valved port, in which case it can simply be flushed with 20 mL of saline. The Huber needle anchors the port within the pocket while the pocket incision is closed in two layers, usually with subcutaneous 3-0 Vicryl followed by a 4-0 Vicryl running subcuticular suture. Other combinations of two-layer closure may be preferred by individual operators, and some incorporate the use of adhesive strips (Steri-Strips [3M, St. Paul, Minn.]) or liquid skin adhesive (Dermabond [Ethicon, Somerville, N.J.]). Regardless of the precise suturing technique used, two-layer closure is preferred to avoid wound dehiscence. If the patient is to receive chemotherapy within the next 3 days, the port is left accessed. At completion of the procedure, the area is cleaned, and sterile gauze with an occlusive dressing is applied.

Alternative Access

In the event of central venous occlusion, unconventional venous access must be considered.4,5,7,23 Potential alternative routes include well-developed collaterals, recanalized occluded veins, femoral, translumbar, and transhepatic access.²³ A femoral approach is rarely considered because of the increased risk of infection associated with this site; however, if it is used, maximal sterile barrier technique should be rigorously employed, and the port should be positioned over the ribs to allow a firm backing for access (Fig. 120-9).

Though more frequently used for dialysis or tunneled catheter access, a translumbar approach should be considered for port insertion if the SVC is occluded. A patent IVC must be confirmed by either venography at the time of the procedure or by review of recent cross-sectional imaging. In addition, the patient must be cooperative and be able to be positioned in a semiprone or decubitus position. The skin should be inspected for any lesions that would prohibit this type of access. After placing a guidewire, snare, or catheter in the IVC from a femoral approach to serve as a target, the catheter or guidewire is secured and the patient placed in the prone or semiprone (right side up) position. After the right flank is prepared and draped, an entry site approximately 7 to 10 cm lateral to the midline and just superior to the right iliac crest is identified, and a local anesthetic is applied. A 21-gauge needle is then advanced medially and superiorly under fluoroscopic guidance by targeting the guidewire or catheter at the L3 level. Angled fluoroscopy and parallax are key components of this procedure. Alternatively, computed tomography (CT) guidance may also be used to confirm access to the IVC.

Once access to the IVC is established, a 0.018-inch mandril wire is advanced into the right atrium. With the use of a transition dilator such as the Cope, the system is upsized to a 0.035-inch system. Oftentimes there is difficulty advancing the catheter through the peel-away sheath at the point of entry into the IVC because of angulation; in such cases, the catheter can be advanced over a stiff hydrophilic guidewire and through the peel-away sheath.7,23 For port access, the pocket should be created over the lower lateral ribs to provide firm support. A two-stage tunnel may be necessary to allow a gentle curve of the catheter to the entry site.

Translumbar catheter complications are similar to those encountered with other access routes and include infection, catheter tip migration, kinking, fibrin sheath formation, and rarely, caval occlusion.²³ In the pediatric population, interval growth may result in migration of the catheter tip to an extravascular position, so plain radiographs should be obtained intermittently to ensure appropriate tip position.²³

If both the SVC and IVC are occluded, the transhepatic route is another alternative.5,7,23 Access to the middle hepatic vein can be achieved with a 21-gauge needle under ultrasound and fluoroscopic guidance. Once access to the hepatic vein is confirmed by injection of contrast material, a mandril guidewire can be advanced into the right atrium to allow the system to be upsized to a 0.035-inch system. The port pocket should be created over the bony support of the rib cage and at a point that is easily accessible. This access is subject to respiratory movement, which may result in migration of the catheter to an extravascular position. Complications of transhepatic catheter placement include intraabdominal hemorrhage (any coagulopathy should be corrected) and hepatic vein thrombosis, although this is uncommon. Again, interval growth is an issue with children, and the catheter tip should be monitored with intermittent plain radiographs.^7,23

Tip Position

Perhaps no other topic is more passionately debated in the central venous access arena than the distal catheter tip position. This contentious topic is impacted by operators’ experiences, local policies, and conflicting recommendations27–32 and is further complicated by anatomic misinterpretations.^33,34 At our institution, the catheter tip of a subcutaneous port is considered to be optimally placed if it is in the region between the cavoatrial junction and the cephalad portion of the right atrium (Fig. 120-10). This takes into consideration the expected cephalad migration with the patient in the upright position,^18,28,35,36 and the improved function with the catheter tip in this position.^28,36 However, other authors vehemently recommend that the catheter tip be placed in the SVC³¹ no deeper than the cavoatrial junction to avoid potential cardiac complications, including arrhythmias, tricuspid valve injury, and the most dreaded complication, cardiac perforation and tamponade.³⁷ This is further complicated by the fact that positioning the catheter outside the atrium does not guarantee there will be no vascular perforation.^29,38 Detailed discussion of this debate is beyond the scope of this chapter; the reader is directed to articles that discuss these points in detail.^28,29,32 In addition, review of the anatomy is in order before further contemplating this debate, because the true cavoatrial junction in adolescents and young adults has been identified as more caudally located than commonly believed, and is best estimated by a point approximately two vertebral bodies below the carina.³⁴

Port Selection: Material, Size, and Number of Lumina

Most ports are made of either titanium or plastic (Fig. e120-3). Plastic ports cause only minimal if any imaging artifacts, whereas titanium ports cause minimal artifact on CT (Fig. 120-11) and only local artifact on magnetic resonance imaging (MRI). Both materials are MRI compatible. Single- and double-lumen ports are available, with single-lumen ports being smaller and less expensive and catheters ranging from 6F to 10F. Dual-lumen ports are larger and more expensive, possibly migrate more,³⁵ and are attached to larger 10F to 13F catheters. Kowalski et al. showed a significant difference in migration between 10F single-lumen ports and 12F double-lumen ports, possibly because of the larger device settling within the soft tissues of the chest wall and thereby withdrawing the catheter.³⁵ Because double-lumen tunneled catheters have been noted to result in more total venous occlusion than is the case with single-lumen catheters,³⁹ and because catheters smaller than 2.8 mm in outer diameter are associated with few instances of thrombosis,⁴⁰ the larger-caliber catheters used in double-lumen ports may also be at increased risk for venous thrombosis. Moreover, one should use central venous devices with the fewest number of lumens necessary to reduce the number of portals for colonization.^41,42 However, dual-lumen ports offer the benefit of being able to administer incompatible medications or concomitantly administer fluids or blood products with chemotherapy. Therefore, double-lumen ports are usually reserved for situations in which incompatible chemotherapeutic agents or concomitant blood product and fluid administration are genuinely anticipated.