Interventional Techniques for Device Implantation

Published on 21/06/2015 by admin

Last modified 22/04/2025

Print this page

This article have been viewed 2414 times

23 Interventional Techniques for Device Implantation

Initially, device implantation was the province of surgeons, and thus implant problems were solved with surgical techniques, such as tunneling of a new lead for a patient with an occluded subclavian vein. As devices became smaller and almost exclusively transvenous, implantation moved to the province of the cardiologist and then subsequently to the electrophysiologist (EP). Until recently, the mechanics of transvenous device implantation were generally straightforward, adopting surgical solutions and applying an electrophysiology skill set and tools. Over the last several years, however, with the marked increase in device upgrades and cardiac resynchronization therapy (CRT), the complexity of device implantation has increased greatly, becoming similar to an interventional procedure rather than the traditional device implant. Lacking interventional training, EP physicians have struggled to solve difficult implant problems with their electrophysiology skill set and tools. As a result, even now, the CRT implant failure rate remains at least 7.5% in experienced hands.¹

Interventional versus Electrophysiologic Approach

It is preferable to learn a new skill set better suited to the procedure than to continue with familiar methods. How would the current implant procedure change if device implantation was assumed by an interventional cardiology skill set rather than electrophysiology?

Consider the occluded subclavian vein. With expanded indications for implantable cardioverter-defibrillators (ICDs),² CRT,^3,⁴ and increased survival of patients with cardiovascular disease,⁵ the implanting physician encounters a higher percentage of patients who have preexisting leads in whom subclavian vein obstruction or occlusion can be as high as 45%.⁶ Application of an interventional approach to the occluded vessel offers the implanting physician less invasive solutions, such as venoplasty versus lead extraction or tunneling.⁷^–¹⁰

Left ventricular (LV) lead placement for CRT requires the successful completion of multiple steps, many originating in the interventional arena. Learning to implant an LV lead by application of an existing electrophysiology (EP) skill set and tools can be a long, painful, and humiliating process. Once learned, the EP approach has limited success, even in experienced hands. Because an interventional approach is better suited to the procedure, learning to implant an LV lead is usually faster, less painful, and more satisfying despite the requirement to first learn a new skill set and tools. Subsequently, application of the interventional approach can increase implant success from 88% to 92% with the EP approach to 98% with the interventional approach.¹¹ This chapter takes the position that it is better to learn a new, unfamiliar skill set than to continue with the familiar that may not be best suited to the procedure (Box 23-1).

Box 23-1
Interventional Mind-Set Approach to Problems Encountered During Device Implantation

1 Good work cannot be done if you are in an uncomfortable position; you owe it to your patients to insist that the room be set up properly.

2 It is important to be alert to new tools that can make difficult cases easier.

3 Contrast is essential to success.

4 Risk of contrast nephropathy is patient specific.

5 Contrast can be used safely in most patients if pretreated for contrast nephropathy.

6 There is (or should be) a preshaped catheter to locate every vascular structure.

7 If the first catheter does not work, try another shape or create a shape.

8 Catheter manipulation requires both hands on the catheter.

9 The assistant should inject contrast so the operator can focus on catheter control.

10 Catheters must not be bent before they enter the patient because every bend in a catheter reduces control.

11 Almost “all” vascular occlusions can be crossed one way or another.

12 Almost “all” vascular occlusions can be dilated with a balloon.

13 All surgical procedures can be done better with wires, catheters, and balloons.

14 Extra-support wires are essential for advancing catheters through difficult anatomy; the stiffer the better.

15 Guide support is essential for successful balloon dilation.

Morbidity and Mortality

Does the application of an interventional skill set specific to device implantation really reduce morbidity and mortality? Many examples show this may be the case, but the best example is sending a patient for an epicardial lead because the implanting physician does not use guide support or perform venoplasty. Even in the most experienced hands, mortality is doubled (4.8% surgical vs. 2.5% transvenous; P = 0.46), and acute renal injury and infection are quintupled (26.2% vs. 4.9%: P = .0004; and11.9% vs. 2.4%; P = .03, respectively).¹² In practice, results may be worse. Surgical LV lead mortality from the REPLACE registry¹³ was greater than 8%. There were four deaths from among the 48 patients (11% of 434 attempted LV lead placements) who went for an epicardial lead following a failed transvenous approach. The 8% mortality assumes all 48 patients went to surgery, although it is not unusual for patients either to decline a surgical lead or to be rejected as too risky; thus the mortality may be even higher. Interventional techniques safely increase implant success, reducing the need for surgical epicardial leads.

Device Implantation–Specific Interventional Skill Set

Application of an interventional skill set specific to device implantation can improve patient satisfaction (one incision not two), reduce complications (avoid extraction or surgery), reduce implant time, and reduce overall mortality. With the assumption that the above is true, physicians implanting devices need to have or develop these skills to provide the best possible patient care. EP training programs should include opportunities to observe and learn device-specific interventional techniques as part of the curriculum. Opportunities for education in interventional device implantation techniques can be developed by attending educational programs or working with interventional cardiologists or interventional radiologists. Once it becomes clear that the application of an interventional skill set specific to device implantation really does reduce morbidity, professional societies will promote the importance of the “interventional skills specific to device implantation” (knowledge base and technical skills) and support education and training in the area.

We define a skill set as the combination of knowledge base and technical skills. Part of the interventional device implantation skill set involves basic interventional knowledge and technical skills unique to device implantation. A physician with formal interventional training in interventional cardiology or radiology may help provide the knowledge and technical skills necessary for device implantation. Because the terms, tools, and techniques used in the interventional approach to device implantation may be unfamiliar, a glossary box is provided for reference. See also the Interventional Implant Equipment List and manufacturers at the end of Chapter 22 and online.

Glossary and Definition of Terms

LV lead delivery system is defined as the set of catheters used to place an LV lead in a coronary vein. The most effective delivery system consists of three telecopying components: 1, the coronary sinus (CS) access catheter; 2, the delivery guide; and 3, the vein selector.

Coronary sinus (CS) access catheter is defined as a removable catheter designed to cannulate the CS and provide access and support for delivery of a left ventricular (LV) lead to the target vein. Delivery guides are intended to be advanced through a CS access catheter. The lumen size of the CS access catheter determines subsequent implant options. The lumen of a device company CS access catheters is 7 French (7F) or smaller, which restricts the ability to add a CS support wire while working close the CS os, and other advanced techniques. There are no device company delivery guides for 6F leads because they require at least an 8F lumen in the CS. The SafeSheath CSG Worley provides a 9F lumen in the CS, expanding options, including a delivery guide for leads a large as 7F.

Delivery guide is defined as a sliceable catheter designed to fit into the target vein and provide direction and support for advancing the LV lead. Delivery guides are not intended for CS cannulation. In the opinion of the authors, the device company delivery guides have two important limitations: (1) they are too small to deliver the many 6F leads currently available, and (2) by providing several shapes to try and match the venous anatomy, device company delivery guides are intended to locate and cannulate the target vein and deliver the lead. In the authors’ experience, it can be difficult if not impossible to locate, insert a wire, and/or cannulate the target vein without the assistance of a vein selector.

Vein selector is defined as a small (5 French, 5F), soft-tip catheter that telescopes inside a delivery guide to extend beyond the delivery guide and locate the target vein using puffs of contrast, provide direction and support for advancing wires into the vein, then serve as a rail over which to advance the delivery guide. None of the device companies offers vein selectors, relying on their delivery guides to act as both vein selector and delivery guides. In the authors’ experience, there are many situations where the delivery guide alone is not sufficient, and the only way to insert the delivery guide into the vein is to use a vein selector inside the guide.

Extra-support wires. Advancing a balloon, catheter, or sheath over a stiffer wire may be necessary to prevent buckling. We find the .035-inch Amplatz Extra Stiff (Cook) suitable in many cases, but on occasion the .035-inch Amplatz Ultra Stiff (Cook) is required. Currently, the Ultra is only available in the cumbersome length of 260 cm. The Amplatz wires are also available in .032-inch size, which may be required for some sheaths. The Amplatz wires are stiff and do not advance easily through tortuous or narrow venous structures; thus they are inserted via an exchange procedure using a small catheter.

Exchanging of wires to augment support. When a more supportive wire (e.g., Amplatz) is required to advance the guide/sheath, a small catheter (4F-5F) is advanced over the existing wire into the pulmonary artery or inferior vena cava, the existing wire removed, the Amplatz inserted, and the exchange catheter removed. A hydrophilic 4F to 5F catheter will advance where a catheter with a standard surface will not, so it is important to have hydrophilic catheters on hand for exchange. To avoid the need for exchange-length wires, the catheter should be 60 to 70 cm in length, but this length may not be readily available. Thus, we often cut the hub and 30 to 40 cm off a 100- to 130-cm catheter once the less supportive wire is removed.

Focused-force venoplasty. The addition of a stiff wire beside the inflated balloon focuses the force of the balloon against the occlusion. Focused force is used for both focal and diffuse occlusions.

Hydrophilic wires/catheters and balloons. Materials that are intrinsically hydrophilic or have a hydrophilic coating are slippery to the touch when wet and offer much less resistance to passage through vascular structures than standard materials of the same size and configuration. Their disadvantage is that hydrophilic wires/catheters can be difficult to handle and direct.

“Push-pull” technique is used to advance the tip of a delivery guide into the ostium of the target vein. With the tail of the balloon at the tip of the guide, advance the guide (push), holding traction on the balloon as it is deflated (pull).

Knowledge Base

The knowledge base of an implantation-specific interventional skill set includes equipment required for the procedure, the importance of contrast, patients at risk for contrast nephropathy, use of preprocedure hydration for safe use of IV contrast, and the importance of table position for maintaining focus on the procedure. Other areas include open-lumen catheter manipulation (keeping both hands on catheter) and wire exchange, preshaped guiding catheters used to cannulate specific venous structures, and the importance of having the assistant inject contrast while the operator keeps both hands on the catheter. Interventional implanters should know type and size of balloons to use for which veins, how to prevent hemothorax during subclavian venoplasty, which tools to use for crossing a venous obstruction, which implant problems respond to the balloon anchoring technique and which to the snare technique, and how sacrificing a lead, using a wire under the insulation, and snaring can be used to regain venous access.

Technical Skills

Technical skills in an interventional skill set include torque control using both hands on the catheter, catheter manipulation while contrast is injected by the assistant, vein-specific catheter techniques (specific preshaped catheter used a specific way for a specific vein), and techniques to gain and optimize and maintain guide support. Along with deploying balloons to serve as anchors, implanters must know techniques for crossing venous obstruction associated with device leads (wire manipulation veins and venoplasty of stenotic veins) as well as venoplasty of stenotic or occluded cardiac vein structures in patients with and without prior open-heart surgery.

Training

Aggarwal and Darzi ¹⁴ discuss how traditional EP and implanting physicians can become trained in the device implantation–specific interventional skill set. Their editorial “Technical skills training in the 21st century” addresses this issue as it relates to such areas as laparoscopic surgery, bronchoscopy, and colonoscopy. They describe proficiency-based training in which the environment can be modeled to simulate the experiences encountered and to generate feedback about both technical skills and knowledge base. Together, physicians who have adopted the interventional approach, device manufacturers, and professional societies can create a program to help the traditional EP physician develop the knowledge base and skill set specific to device implantation.

Prevention of Contrast Nephropathy

Use of Contrast Material

The interventional mind-set approaches implant problems with the following attitude about contrast:

1 Contrast is essential.

2 Safety of contrast is patient specific.

3 Safe use of contrast requires advanced patient preparation.

4 Most patients with a history of heart failure can be hydrated safely.

5 Renal failure is more difficult to treat than volume overload.

When considering the use of contrast for LV lead placement, keep in mind that a successful implant not only improves symptoms, but also reduces morbidity and mortality.¹⁵ Although uncommon in most situations,^16,¹⁷ contrast nephropathy can be an important complication of CRT associated with increased morbidity and mortality and extended hospital stay.¹⁸ Patients with congestive heart failure and preexisting renal insufficiency are at increased risk for contrast nephropathy,¹⁹ but ultimately, CRT improves renal function by improving cardiac function.²⁰ In fact, patients with moderate renal insufficiency may derive the greatest benefit from successful CRT.²¹

Recognizing the importance of contrast to a successful procedure, the interventional mind is proactive, always looking for ways to reduce risk,²² whereas the EP mind-set focuses on reducing the use of contrast. Despite every effort to limit its use, the volume of contrast frequently exceeds expectation. Further, the nephrotoxicity of even a small volume of contrast (10-30 mL) in the unprepared patient can be significant, whereas nephropathy can be reduced or eliminated if preventive steps are taken before the implant. Thus, it is important for implanting physicians to adopt the interventional mind-set and take a proactive approach to contrast nephropathy rather than simply trying to limit its use.

After exposure to contrast, the kidneys concentrate contrast in the urinary space within the renal tubules, releasing nitric oxide from endothelial cells, triggering a transient vasodilation, followed by more prolonged vasoconstriction, and causing tubular cell damage, extravasation into the peritubular space, and constriction of the lattice-like peritubular blood vessels in the outer medulla. Sustained vasoconstriction can last for hours to days, resulting in ischemic injury to the outer medulla, particularly in those with chronic kidney disease and diabetes mellitus ²³ (Box 23-2).

Contrast-induced nephropathy (CIN) is typically defined as a 25% or greater increase in baseline serum creatinine, or a 0.5-mg/dL increase without another identifiable cause. CIN occurs 24 to 48 hours after exposure, with serum creatinine peaking at 3 to 5 days, returning to normal within 2 weeks, although usually by 7 to 10 days. The approach to the prevention includes (1) hydration to dilute and flush contrast from the kidney, (2) selection of less nephrotoxic contrast agent, (3) drugs to protect the kidney and prevent vasoconstriction, and (4) hemofiltration or hemodialysis.

A bedside risk score stratifies patients into low (0-7 points), medium (8-14), and high-risk (≥15). The score is calculated as follows: age 80 or older (2 points), female (1.5), diabetes (3), urgent (PCI) (2.5), emergent PCI (3.5), congestive heart failure (4.5), serum creatinine of 1.3 to 1.9 mg/dL (5) or 2 mg/dL or greater (10), and pre-PCI use of intra-aortic balloon pump (13). Based on the score, high-risk patients can be identified for early administration of more aggressive and intensive prophylaxis.^24,²⁵

Withdrawal of Nephrotoxic Drugs

Common nephrotoxic drugs include nonsteroidal anti-inflammatory drugs (NSAIDs), metformin, diuretics (e.g., furosemide), angiotensin-converting enzyme (ACE) inhibitors, aminoglycosides, sulfonamides and penicillin, amphotericin, cyclosporin A, vancomycin, and cisplatin. These agents should be withdrawn before contrast media administration in all patients at risk. Metformin is the most important because it can produce intramedullar metabolic acidosis, which promotes nephropathy. Almost all patients can tolerate suspension of diuretics on the day of the implant procedure.

Periprocedural Hydration

Hydration to dilute and flush contrast from the kidney is the most important step in the prevention of CIN. Hydration has two parts: (1) the fluid used: half-normal saline, normal saline, or bicarbonate, and (2) the protocol: (a) “overnight” (12 hours before and 12 hours after) or (b) “bolus” (3 mL/kg 1 hour before and 1-1.5 mL/kg 4-6 hours after procedure).

Hydration Fluids

Hypotonic Saline

A 0.45 normal saline (NS) solution (in 5% glucose) should not be used; it is less effective than isotonic saline. In a study of 1383 patients, CIN was 2.0% with 0.45 NS versus 0.7% with isotonic saline.²⁶

Isotonic Saline

Multiple clinical trials confirm that NS definitively reduces CIN. For example, Solomon et al.²⁷ reported that with saline hydration, “There is little risk of clinically important nephrotoxicity attributable to contrast material (122 ± 90 mL) for patients (220) with diabetes and normal renal function or for nondiabetic patients with preexisting renal insufficiency. The risk for those with both diabetes and preexisting renal insufficiency is about 9%, which is lower than previously reported.” None of their patients required dialysis.

Sodium Bicarbonate

A 2004 study suggested that sodium bicarbonate was superior to saline hydration,²⁸ although the hydration protocols were not the same. Subsequently, five published randomized clinical trials and three meta-analyses have provided conflicting results.²⁹^–³¹ Interpreting the data requires recognizing that in three of the six published studies (and several abstracts), the hydration protocol for saline and bicarbonate were not the same; thus the difference seen may reflect the effect of the infusion protocol rather than the infused solution. Only three published studies used the same infusion protocol.³²^–³⁴ The smaller studies found bicarbonate superior to saline,^33,³⁴ whereas the larger study found no difference.³²

Hydration Protocol

There is no direct comparison of overnight versus bolus infusion. However, analysis of the bicarbonate versus saline trials suggests bolus hydration is superior to overnight hydration, particularly in patients with congestive heart failure (CHF). When overnight saline was compared to bolus bicarbonate, saline was reduced for CHF patients in two studies ^35,³⁶ (not defined by Merten et al.²⁸). In the bicarbonate arm, the loading dose (3 mL/kg/hr for 1 hour) was not adjusted, but the infusion (1 mL/kg/hr during and 6 hours after procedure) was reduced to 0.5 mL/kg in one trial.³² Bicarbonate was superior to saline in the two studies where bicarbonate was not reduced, but equal to saline when the bicarbonate infusion was reduced for CHF.

These results suggest that reducing the dose of overnight saline results in less effective hydration than bolus hydration, provided it is not adjusted for CHF. For a 70-kg person, the volume of the unadjusted bolus infusion protocol is 620 mL, whereas the overnight infusion protocol (reduced for CHF) is 840 mL; thus the bolus protocol is superior to overnight hydration, particularly in patients with CHF. Our recommendation therefore is to avoid overnight hydration and use acute hydration (3 mL/kg) starting 1 hour before the procedure, then either 1 mL/kg during and 6 hours after the procedure or 1.5 mL/kg for 4 hours. Another advantage of bolus hydration is that it can be started shortly before the procedure.

Volume of Contrast Media

Although studies show that the risk of CIN is dose dependent, there does not appear to be a threshold value below which CIN does not occur. Patients with preexisting renal dysfunction and diabetes appear to be at highest risk for CIN. Retrospective studies of cardiac catheterization suggest a lower risk of CIN for contrast loads less than 100 mL for interventional procedures.

Choice of Contrast Agent

Types of Iodinated Contrast Agents

All contrast agents consist of a benzoic acid molecule with three atoms of iodine replacing hydrogen atoms on the benzene ring. Classification is based on (1) the charge on the iodinate molecule (ionic vs. nonionic), (2) osmolarity, and (3) molecular structure (monomer vs. dimer). The osmolality of the contrast agents is a measure of the number of particles in solution. The original contrast agents were ionic monomers and very hypertonic (osmolality of metrizoate = 2100), whereas newer contrast agents were nonionic and less hypertonic (osmolality of iopamidol = 796). Thus, ionic metrizoate was called “high osmolar,” referring to its osmolality; the terms high osmolar and ionic were used interchangeably. Currently, the term “high osmolar,” referring to the high osmolality of ionic agents, has no meaning because the osmolality of ionic agents such as ioxaglate (580) are the same or lower than for nonionic agents. More important (and confusing), the term “low osmolar” now refers to an agent with osmolality higher than 290. At present, the higher the osmolality, the more “low osmolar” is the agent, which is the opposite of the original description. By modern convention, iodixanol (osmolality of 290) is “iso-osmolar” and all agents greater than 290 are “low osmolar,” regardless of whether they are ionic or nonionic. The higher the osmolality, the more “low osmolar” it becomes. For example, ionic ioxaglate (osmolality of 580) is “low osmolar” compared to nonionic iodixanol (290). If one used the previous nomenclature, ioxaglate would be both “high osmolar” (because it is ionic) and “low osmolar” because it has a higher osmolality than iodixanol. Do not use the term “high osmolar” in relation to contrast.

The osmolality of nonionic agents follows: iodixanol 290, ioxilan 695, iopromide 774, iopamidol 796, and iohexol 884 mOsm/kg. Iodixanol is iso-osmotic, and all the rest are “low osmolar.”

Comparison of Contrast Agent Nephrotoxicity

Ionic vs. Nonionic

One randomized prospective evaluation of ionic diatrizoate (measured osmolality = 1500) vs. nonionic iopamidol (osmolality 796) found no difference in the incidence of nephrotoxicity.³⁷ However the subsequent RECOVER study compared ionic, “low-osmolar” ioxaglate (osmolality 580) to nonionic, iso-osmolar iodixanol (osmolality 290) in patients with renal insufficiency undergoing coronary angiography with or without PCI.³⁸ In addition to ionic versus nonionic, a second variable was “low osmolar” (580) versus iso-osmolar (291). The incidence of nephropathy was significantly lower with iso-osmolar nonionic iodixanol (7.9%) than with “low osmolar” ionic ioxaglate (17.0%) (P = .021). Unfortunately, some patients received hydration with 0.45 NS (ineffective; see earlier), and thus the difference may be caused by inadequate hydration. Further, the findings may also reflect the difference in osmolality.

Low-Osmolar Nonionic vs. Iso-Osmolar Nonionic

The NEPHRIC Study compared low-osmolar iohexol (osmolality = 884) with an iso-osmolar iodixanol (osmolality 290) and found nephropathy less likely to develop in high-risk patients with iso-osmolar iodixanol.³⁹ Because NEPHRIC was small and hydration not uniform, and because several subsequent studies were flawed and produced conflicting results, the CARE trial was conducted to settle the issue. The Cardiac Angiography in Renally Impaired Patients study was a prospective, multicenter, randomized, double-blind comparison of the “low osmolar” nonionic iopamidol (osmolality 796) to the iso-osmolar nonionic iodixanol (osmolality 290) in high-risk patients.⁴⁰ CIN, defined by multiple endpoints, was not statistically different in high-risk patients, with or without diabetes mellitus. All patients received prophylactic volume expansion with isotonic sodium bicarbonate solution, administered at 3 mL/kg/hr for 1 hour before angiography, and at 1 mL/kg/hr during and for 6 hours after angiography. Prophylactic N-acetylcysteine regimen of 1200 mg twice daily administered on the day before and on the day of the study procedure was used in some centers. The authors concluded that any true difference between the agents was small and unlikely to be clinically significant. Thus, if bolus hydration is used, either nonionic iso-osmolar iodixanol or nonionic low-osmolar iopamidol can be used. The conflicting results of NEPHRIC and CARE may reflect the approach to hydration.

In summary, adequate hydration with saline or bicarbonate (3 mL/kg load, then 1 mL/kg for 6 hours) apparently reduces the difference (if any) among the different agents. When hydration is suboptimal, the nonionic agents, particularly those with an osmolality close to isotonic (iodixanol osmolality = 290) may be superior.

Gadolinium

Although they may be less nephrotoxic than iodinated contrast agents, gadolinium-based contrast agents should never be used because of their association with nephrogenic systemic fibrosis, a relatively recently recognized,⁴¹ rapidly progressive fibrosing disorder of skin and internal organs. Nephrogenic systemic fibrosis is seen in patients with kidney disease (estimated glomerular filtration rate <30) who receive gadolinium-containing contrast agents.⁴²

Carbon Dioxide

For more than five decades, gaseous CO₂ has been recognized by interventionalists as an alternative to iodinated contrast for diagnostic as well as interventional procedures.^43,⁴⁴ Winters et al.⁴⁵ recently described their favorable experience with gaseous CO₂ venography rather than iodinated dye in patients undergoing pacemaker or ICD lead revisions.

Drugs to Prevent Renal Failure

Iloprost

A prostacyclin analogue, iloprost works “presumably as a renal vasodilator and cellular protective agent downstream in the prostaglandin cascade to prevent contrast-induced acute kidney injury.”⁵⁰ In a randomized, double-blind, placebo-controlled study of 208 patients with a serum creatinine of 1.4 mg/dL or greater, Spargias et al.⁵¹ reported CIN in 22% of controls and 8% of the iloprost group (1 ng/kg/min). All patients were hydrated with saline. Although dose-dependent hypotension and platelet inhibition are the principal side effects of iloprost, the incidence of hypotension and bleeding was not significantly different between groups.

Hemofiltration and Hemodialysis

In patients with a creatinine greater than 2 mg/dL, hemofiltration (4 hours before and continued for 24 hours after the procedure) compared to isotonic saline prevented the deterioration of renal function, decreased the need for renal replacement therapy, and improved in-hospital and long-term outcomes.⁵² Similarly, in patients with a creatinine more than 3.5 mg/dL, hemodialysis (a single, 4-hour treatment initiated immediately after procedure) reduced length of stay and the need for renal replacement therapy at hospital discharge.⁵³

Treatments to Avoid

Fenoldopam (systemic IV administration), dopamine, calcium channel blockers, atrial natriuretic peptide, and l-arginine are not effective, whereas furosemide, mannitol, and endothelin receptor antagonists are potentially detrimental.⁵⁴

We have successfully adopted the following reasonable approach. Because patients with CHF are at increased risk and the serum creatinine levels of patients with New York Heart Association (NYHA) Class III or IV CHF can vary from day to day, we pretreat all patients undergoing a CRT implant, as follows:

1 Withdraw nephrotoxic drugs.

2 Bolus hydration with either isotonic saline or bicarbonate, 3 mL/kg/hr for 1 hour before, then 1 mL/kg/hr (consider 0.5 mL/kg/hr for recent CHF decompensation) during, and for 6 hours (or 1.5 mg/kg/hr for 4 hours) after the procedure. Because the total volume of intravenous (IV) hydration is lower with bolus hydration and the results possibly better, we no longer recommend overnight hydration, particularly for those with CHF.

3 Consider iloprost (±)

4 Use either iso-osmolar iodixanol or iopamidol.

5 Consider hemofiltration and hemodialysis in extremely high-risk patients (see Brown risk score earlier).

Additional Concerns

No data suggest that shellfish allergies predict increased risk of adverse reactions to contrast. It is currently thought that shellfish allergy is related to tropomyosin, and patients with seafood allergies do not have an increased risk of iodine allergy.

A variety of regimens are used to prevent allergic reactions to iodine. Specific regimens will vary from center to center. Preprocedure medications include diphenhydramine (Benadryl), a histamine receptor blocker, and steroids beginning 12 hours before the procedure. Typically, prednisone or methylprednisolone is administered 12 hours and immediately before the procedure.

Preimplant Hydration

Preimplant hydration is discussed earlier.

Conclusion

Implanting physicians who are not accustomed to preimplant hydration are extremely concerned that CRT patients will suffer CHF decompensation. Thus they avoid hydration, increasing the risk of nephrotoxicity that can occur even when small volumes of contrast are used. Our experience with routine hydration for device implantation does not support the level of concern frequently expressed. Further, many CRT patients are hydrated for their interventional procedures (e.g., cardiac catheterization and PCI) without acute decompensation. Finally, there is the risk/benefit ratio of hydration to prevent CIN versus the possibility of exacerbating CHF. Contrast nephropathy results in an increased length of stay, higher morbidity, and both short-term and long-term mortality. Contrast nephrotoxicity is reduced if not eliminated by hydration. With the new, acute bolus approach to hydration, the total volume of IV hydration is lower, and the patient is under observation during the daytime hours. CHF may or may not develop as a result of hydration and can be treated if it does; thus hydration should be employed (Box 23-3).

Box 23-3
steps to prevent contrast nephropathy

1 Hold diuretics on the day of the procedure.

2 Preimplant hydration to prevent contrast nephropathy: NaCl vs. NaHCO₃ (probably no advantage of one over the other)

Recommendations:

a 3 mg/kg for 1 hour before procedure

b 1 mg/kg during and for 6 hours after procedure, 0.5 mL/kg if recent CHF exacerbation

3 Contrast agents: in general, contrast agents with osmolality closer to isotonic (≈290) seem to be superior to agents with high osmolality (>500) in patients with diabetes mellitus and chronic renal failure. However, isotonic iodixanol and iopamidol (osmolality = 796), seem to be equivalent.

Preparations for Implantation

Room Setup and Table Position

The equipment required for the procedure needs to be on hand in the room, including the appropriate implant table (Fig. 23-1). The table and the assistant need to be in the appropriate position (Fig. 23-2). If the monitor is not in a comfortable position, mount a flat-screen monitor on an IV setup pole.

Figure 23-1 Implant table designed for cardiac resynchronization therapy.

A, Skewed view; B, side view. The table is designed to be in the T-bone position for dealing with subclavian obstructions and during left ventricular lead implantation. The legs are set back to allow the image intensifier rotate under the table when imaging in the right anterior oblique projection. In addition, there is space between the legs of the implant table and the base of the patient table for the fluoroscopy pedals.

(Courtesy Oscor, Palm Harbor, Fla.)

Figure 23-2 Importance of table position while working with long wires and catheters.

A, Implant equipment table is in its usual position behind the operator. Wires and catheters are bent 90 degrees as they exit the pocket, making them prone to fall off the table. The 90-degree bend makes open-lumen catheters prone to kink, resulting in collapse of the lumen, destroying the function of the catheter. The 90-degree bend also cuts torque control of the catheter by 50%, making it more difficult to cannulate the coronary sinus or target veins. Wire exchanges are difficult to perform. B, Table is perpendicular to the patient table. The catheters and wires fall naturally on the table. The operator and assistant are not distracted from the procedure by trying to keep the wires from falling off the table and becoming contaminated. Open-lumen catheters are less likely to kink, and torque control is optimal. Wire exchanges are easily performed by pinning the wire to the table and sliding the catheter back to the hand pinning the wire.

Catheter Manipulation and Contrast Injection

Electrophysiology physicians are used to working with electrode catheters that have almost infinite torque control and do not kink or twist as do open-lumen catheters. As a result, keeping the catheter straight before it enters the body and applying torque with both hands to distribute the force are not as essential as with open-lumen catheters. In addition, we become accustomed to working alone with our EP catheter, so when contrast needs to be injected, it is not the EP mind-set to do it alone. To work effectively with open-lumen catheters and inject contrast, the approach needs to change from an EP to interventional simply because the interventional approach to catheter manipulation and contrast injection is the way the tools were designed to be used (Fig. 23-3)

Figure 23-3 How to torque an open lumen catheter and inject contrast.

A, Operator is attempting to control the catheter and inject contrast simultaneously. B, Operator has both hands on the catheter. The assistant can inject contrast on request of the operator. Interventional principles dictate that both hands must be on the catheter for effective catheter manipulation. Using both hands spreads the torque over a larger area, reducing the chance the lumen will collapse, destroying the catheter. The addition of the Y adapter with hemostatic valve and rotating hub makes it much easier to hold the catheter in position and insert a wire when needed without loss of catheter position. When the operator attempts to inject contrast and manipulate the catheter, two events sabotage the result: the operator must redirect focus from the tip of the catheter to the syringe, and catheter position is compromised because the plunger hand no longer is in control of the catheter.

Balloon Basics

Type of Catheter for Balloon Mounting

Balloon catheters are either “over the wire” (OTW) or rapid exchange (Fig. 23-4). The terms rapid exchange and monorail are synonymous. OTW coronary balloons are about 150 cm and thus require a 300-cm wire, which is very difficult to handle. We always use a rapid-exchange coronary balloon that requires a standard-length wire. Peripheral balloons are 40 to 60 cm; thus an OTW peripheral balloon uses a standard-length wire, which makes rapid-exchange models unnecessary. When balloons are made specifically for coronary venoplasty, they will be short enough to be OTW and still use standard-length wires.

Figure 23-4 Comparison of over-the-wire balloons with monorail balloons.

A, The wire over which the balloon is tracked exits the catheter at the end of the catheter near the inflation port. B, The wire over which the balloon is tracked exits the side of the balloon 20 to 30 cm from the tip. The advantage of an over-the-wire balloon is better ability to push the balloon and change the wire, if needed, distal to the obstruction. The disadvantage is the wire over which the balloon tracks must be at least twice the length of the balloon. When working with short, peripheral balloons, the length of wire required is not problematic. Working with an over-the-wire coronary balloon (150 cm in length) requires a 300-cm wire, and thus the operator and assistant must cope with 150 cm outside the body. The disadvantage of the monorail (also referred to as rapid exchange) is that the balloon cannot be used to change wires distal to the obstruction, and there is less “pushability” because the wire does run the length of the balloon.

Choice of Angioplasty Balloon and Inflation Pressure for Venoplasty

Compliance of Balloon

The term compliance refers to how a material deforms, or stretches, under pressure. This is an important consideration when selecting a balloon angioplasty catheter. Balloons of the same dimension will produce considerably greater force at the same pressure if the balloon material does not stretch. When a material stretches or deforms under pressure, it prevents the concentration of force from being focused at the stenosis. The balloon material stretches around the lesion rather than exerting force on it (Fig. 23-5, A) The more noncompliant the balloon material, the more dilating force it transmits, the more uniformly it expands, and the less risk it may pose of dilating healthy tissue (Fig. 23-5, B)

Figure 23-5 Comparison of compliant to noncompliant balloons.

A, Compliant balloon is inflated against an obstruction. B, Noncompliant balloon is inflated against the obstruction.

A more compliant balloon material exerts less dilation force at the lesion and increases the risk of “dog boning” and potential vessel dissections. Balloon dilation force is potentially applied outside the point of resistance or lesion. A less compliant balloon provides more dilation force within a vessel or metal stent, potentially reducing the risk of trauma outside the edges of the lesion. Figure 23-6 illustrates the impact of compliance on nominal pressure, rated burst pressure, and balloon size between nominal and rated burst pressure.

Figure 23-6 Compliance tables from compliant (A) and noncompliant (B) balloon.

The material from which the compliant balloon is made is thinner than the noncompliant balloon; as a result the nominal atmospheric (ATM) pressure of the compliant balloon (A = 6 atmospheres) is lower than the noncompliant balloon (B = 12 atm). The rated burst pressure of the compliant balloon (A = 12 atm) is lower than the rated burst pressure of the noncompliant balloon (B = 20 atm). In addition, the size of the compliant balloon increases from 3.5 mm at nominal pressure to 3.81 mm at rated burst pressure (A), whereas the noncompliant balloon increases to only 3.66 mm at rated burst pressure. Thinner balloon material also makes the profile (outside diameter) of the compliant balloon smaller than that of the noncompliant balloon.

More Compliant Balloons

• Softer, more flexible, and more susceptible to growth outside a hard lesion or metal stent.

• Higher rate of balloon diameter growth and will typically grow in the areas of least resistance.

• Can grow in diameter outside the lesion and apply high pressure to healthy tissue.

• Pressure tends to migrate to that area of the balloon with the least resistance. This can cause a “dog boning” effect, which could result in damage of healthy tissue outside the intended treatment area.

Less Compliant Balloons

• Typically stiffer, thicker, and less susceptible to growth outside a lesion.

• Balloon holds its shape despite added pressure, exerting more force against a hard lesion, potentially increasing success.

Technique Considerations

For coronary vein venoplasty, a “noncompliant balloon” (midrange in compliance scale) is a good place to start. Such balloons track easily and relieve most stenotic areas. If the balloon will not advance across the stenosis, the first step is to reevaluate the guide support. After a guide with a size and shape suitable for the target vein is in place, the operator should consider downsizing the balloon, using one of the same composition (e.g., 3 mm to 2 mm), and changing to a more compliant balloon. Once the smaller balloon is inflated across the lesion, the balloon should be reinserted and inflated. If there is a residual “waist” (indentation in balloon created by obstruction) after full inflation of a noncompliant balloon, an ultra-noncompliant balloon (Kevlar) can be tried. The addition of a second (stiff) angioplasty wire beside the balloon before inflation will focus the force of the balloon on the narrowing. Because it can be difficult to place a second wire across a lesion after initial attempts at dilation, we always place two wires across the lesion initially, one stiff and one floppy, before starting. The balloon goes over the floppy wire, and the stiff wire is used to concentrate forces at the site of stenosis.

Profile of Balloon

The outer diameter (OD) of the balloon in its wrapped, never-inflated state is referred to as the profile; the lower the profile, the easier it is to pass the balloon through the obstruction. Noncompliant balloons are made of thicker material; thus the OD of a 3-mm compliant balloon is less than OD of a noncompliant 3-mm balloon. Once a balloon is inflated, the profile is greater because of “winging” (Fig. 23-7). The balloon material is wrapped around the shaft, and once deflated, it remains unwrapped, increasing its profile.

Figure 23-7 Balloon material unwraps after first inflation, increasing balloon profile.

After the first inflation, a balloon designed for dilation will have a larger outside diameter, making it more difficult to place across an obstruction. The first inflation should be central, at the superior vena cava (SVC)–innominate junction, for subclavian obstructions and distal in the coronary veins.

Diameter of Balloon

When dealing with coronary arteries, the initial size of the balloon is based on the estimated size of the normal coronary artery proximal and distal to the stenosis. However, in the coronary veins, the size of the pacing lead determines the size of the balloon. In the coronary branch veins, a 3.0-mm balloon inflated to at least nominal pressure, or until the waist is eliminated, is satisfactory for 6-French (6F) leads, 2.5 mm for 5F Leads, and 2.0 mm for 4F leads. In the main body of the proximal coronary sinus (CS), a 6.0-mm balloon is used, while a 4.0- to 6.0-mm balloon is used for the mid to distal CS.

Length of Balloon

The length of noncompliant balloon suitable for coronary venoplasty is 15 to 21 mm. For enlargement of a straight, diffusely narrow vein segment, a longer balloon may be chosen. Longer balloons tend to be more difficult to advance because of their greater surface area of resistance in the stenotic vein segment. If a very tight, discrete stenosis is encountered, a shorter balloon with less resistance may be a better choice. However, longer balloons tend to be more stable in the vein as well. The composition of the balloon material also influences the length. When an ultra-noncompliant or a cutting balloon is required, 9 to 10 mm is usually the best starting length. The material required to construct the ultra-noncompliant balloon is relatively thick and stiff, and thus it is difficult to track through any curved vein segments, particularly if the balloon is long.

Inflation Pressure

Reviewing the packing material included with the balloon gives the best indicator of the appropriate pressure, with the following definitions kept in mind:

Nominal pressure: Pressure required for that balloon to achieve the rated size in an open system. If the balloon is not inflated to the nominal pressure, it will not reach the rated size.

Overinflated: Balloon size when the nominal pressure is exceeded. Compliant balloons increase beyond nominal more than noncompliant. The size of the balloon at pressures that exceed nominal is printed on the packaging along with the nominal pressure (see Fig. 23-6).

Rated burst pressure: Pressure above which the balloon will burst (usually). Compliant balloons tend to have lower rated burst pressure (see Fig. 23-6).

Effective pressure: Pressure required to relieve the obstruction. The effective pressure varies according to the nature of the stenosis. In patients with prior open-heart surgery, significant fibrosis or even suture material may be contributing to the obstruction. The effective pressure is judged from observation of the balloon. The pressure is effective when the balloon achieves a uniform size and the waist is eliminated.

Box 23-4 summarizes major considerations in the preparation of a venoplasty balloon.

Box 23-4
Preparation of Venoplasty Balloon

1 Partially fill the syringe of the inflation device with half-strength contrast.

2 Connect the inflation device to the balloon with a three-way stopcock.

3 Turn stopcock off to the balloon.

4 Clear air from the line with contrast from the inflation syringe.

5 Turn the stopcock off to the inflation device.

6 Connect a 30-mL (cc) syringe filled with half-strength contrast to the stopcock.

7 Draw back on the 30-mL syringe to remove air from the balloon.

8 With the 30-mL syringe vertical, tap bubbles to the top of contrast.

9 Release to fill the balloon with contrast from the 30-cc syringe.

10 Repeat the sequence: drawback, tap bubbles to the top, release.

11 Turn the stopcock off to the syringe and remove the syringe.

12 Draw back on the inflation device to clear residual bubbles from the line.

13 Holding the syringe of the inflation vertical, tap bubbles to the top.

14 Remove the protective cover from the balloon.

15 Proceed as ready for procedure.

Important: Do not test-inflate the balloon before insertion.

Crossing Obstructed or Occluded Vein in Patients with Preexisting Leads

It is important to remember that “subclavian” is commonly used shorthand for the axillary/subclavian/innominate vein. Some degree of subclavian vein occlusion is common in patients with existing leads. The severity of the obstruction is often defined based on a venogram performed by contrast injection through an arm vein (peripheral venogram). When the peripheral venogram reveals an occlusion, it is best to enter the vein as far peripheral to the occlusion as possible to provide a stable access for approaching the occlusion (Fig. 23-8). A standard short J wire is then advanced to the occlusion and the needle exchanged for the dilator from a 5F sheath. The dilator is then connected to the Y adapter of the injection system (Fig. 23-9), and contrast is injected at the site of occlusion.

Figure 23-8 Venous access peripheral to occlusion on brachial venogram.

A, Thin-walled 18-gauge needle is positioned above the vein peripheral to the occlusion as 20 mL of contrast is injected in a brachial vein. B, As the needle is advanced into the vein, the column of contrast is seen to indent, confirming location. To ensure stable access from which to work, the needle should enter the vein as far peripheral to the occlusion as possible.

Figure 23-9 Injection system used to define the occlusion.

The dilator is from a 5-French (5F) sheath. When the J wire provided with the access sheath will not advance, the dilator from a 5F sheath is advanced over the J wire and connected to the Y adapter. Contrast injection through the dilator at the site of occlusion will reveal openings in the occlusion. The glide wire can then be directed through the occlusion using a steering handle.

The length and severity of the occlusion may be either overestimated or underestimated by the peripheral venogram. In Figure 23-10 the peripheral venogram reveals extensive collaterals, and the occlusion appears complete in A, whereas in B, contrast injected at the site of occlusion demonstrates an opening that was easily crossed with an angled glide wire and torque device. In Figure 23-11, despite collaterals, the peripheral venogram in A suggests an opening that will be easy to cross, whereas in B, with injection of contrast at the occlusion, there is a total occlusion over 2 cm. Although it is important to perform contrast injection at the site of obstruction, the true, clinically relevant severity of obstruction is ultimately determined by whether a wire will cross the obstruction. For example, the venogram may demonstrate a hopelessly long, total occlusion, but the wire may cross easily (Fig. 23-12). Conversely, the stenosis may be short but impossible to cross with a wire. Accordingly, we define the obstruction based on how difficult it is to deliver a wire across the occlusion, as follows:

1 Moderate: Glide wire is easily advanced, but catheters do not advance easily.

2 Severe: Glide wire does not easily cross.

3 Total: Despite aggressive manipulation with multiple wires and assist catheters, a wire will not cross.

Figure 23-10 Contrast injected at site of occlusion reveals opening.

A, When contrast is injected through a vein in the left arm, the subclavian appears to be totally occluded with extensive collaterals on brachial venogram. B, When contrast is injected at the site of occlusion, it is apparent that the vein is stenotic, not occluded, and the opening to facilitate passing the wire is demonstrated.

Figure 23-11 Contrast injected at site of occlusion reveals total occlusion.

A, Although there are collaterals, it appears that the vein is patent on brachial venogram, as indicated by black arrowheads. B, Contrast injection through the dilator reveals that the vein is occluded, as indicated by the arrows marking where the occlusion starts and ends.

Figure 23-12 Total occlusion demonstrated by contrast injection at site of occlusion.

Despite the extensive collaterals, this total occlusion was crossed with a .035-inch angled hydrophilic guidewire directed with a torque device in less than 1 minute. Currently, most implanting physicians do not attempt to cross the occlusion and resort to extraction or move to the other side, an unfortunate outcome for the patient.

Crossing an Occluded “Subclavian” vein

Based on the venogram, enter the vein peripheral to the obstruction, and advance the standard short, .035-inch J wire to the occlusion. Remove the needle and advance the dilator from a 4F to 5F sheath to the obstruction, and remove the J wire. Attach the injection system to the dilator, and inject contrast to define the extent of the occlusion. Often, an opening is apparent despite an apparent total occlusion on the peripheral venogram. The severity of the occlusion should never be based on the peripheral venogram. An angled .035-inch glide wire is then inserted through the Y adapter and a torque device attached. The wire is then directed into the opening defined by contrast injection. Additional puffs of contrast can be injected to define the opening, as needed. As the wire is advanced, the dilator is advanced to provide support for the wire.

In some cases the axis of the dilator directs the wire away from the opening (Fig. 23-13; see also Fig. 23-18, A and B). Faced with this situation, we have evaluated a variety of catheters and catheter shapes and found a hydrophilic, left vertebral diagnostic catheter (vert slip-cath, vert) (Fig. 23-14), in conjunction with a torque device and angled glide wire (Fig. 23-15), to be most useful. The short tip of the “vert” can be turned in the vein peripheral to the occlusion, directing the glide wire toward the opening (Fig. 23-16). The hydrophilic coating on the catheter is important to ensure the catheter will advance as the glide wire is advanced; however, it can slip through the operator’s fingers and out of the vein. To avoid loss of venous access while using the vert, it is useful to exchange the dilator of the 5F sheath for the 5F sheath (Fig. 23-17). The sheath is more secure in the vein than the vert, even if only a few millimeters are in the vein. Once in place, the vert is connected to Y adapter of the injection system and advanced to the occlusion with puffs of contrast. Once at the occlusion, the vert is turned in the direction of the opening and the glide wire advanced; the vert is useful for peripheral occlusions (Fig. 23-18) as well as central occlusions (Fig. 23-19). The support and direction of the vert can also be used to direct the wire through what appears to be a total occlusion (Fig. 23-20). In addition, the vert can be advanced without a guidewire using contrast and orthogonal views to confirm position (Fig 23-21). Figure 23-22 illustrates the synergism of the glide wire and the left vertebral catheter.

Figure 23-13 Dilator directs glide wire away from opening in occlusion.

A, Glide wire follows the direction of the dilator. B, Working with the torque device, the tip of the wire is directed toward the opening but will not cross.

Figure 23-14 Two angled hydrophilic catheters that are useful for crossing an occluded vein.

The short tip allows the catheter to be manipulated within the vein to direct and support the wire. Glide-cath (Angled Taper Radiofocus Glidecath, 5F 100 cm, Terumo Medical) is glide wire catheter 4 French (4F) in diameter with .038-inch lumen and 10-mm angled tip. Vert slip-cath (Vert Slip-Cath Beacon Tip; Cook) is left vertebral diagnostic catheter 5F in diameter with .038-inch lumen and 12-mm angled tip. The “vert” provides more support for the glide wire but may not advance as easily over the wire (5F) as the glide-cath (4F). Both catheters are available in 4F and 5F sizes.

Figure 23-15 System used to cross difficult subclavian/innominate occlusions.

A, Glide wire is advanced through Y adapter, and the torque device is attached to the distal end of the glide wire. B, Direction of the glide wire can be determined by the left vertebral diagnostic catheter (vert). Contrast can be injected through the .038-inch lumen of the vert with the .035-inch glide wire in place.

Figure 23-16 Vert directing glide wire.

A, Tip of the left vertebral diagnostic catheter (vert) is directed inferiorly toward the opening in the occlusion. B, Glide wire advances through the obstruction.

Figure 23-17 Dilator exchanged for sheath to prevent loss of venous access.

A, Glide wire is advanced deep into the collaterals before the dilator is removed. B, Contrast is injected in the sheath to ensure the tip of the sheath is in the vein before removing the glide wire.

Figure 23-18 Vert used to cross occlusion.

A, Contrast is injected through the superiorly directed dilator of a 5F sheath, demonstrating an inferior opening. B, Dilator directs the glide wire away from the opening. C, Dilator is exchanged for a left vertebral diagnostic catheter (vert), which is directed toward the opening confirmed using contrast. D, Glide wire is easily advanced through the opening for subsequent venoplasty.

Figure 23-19 Vert crossing central obstruction at peripheral end of proximal coil.

A, Venography near the site of occlusion when a glide wire would not advance reveals the opening is located superiorly. B, Left vertebral diagnostic catheter (vert) is directed to the site of occlusion. C, Glide wire is advanced across the occlusion, using the vert for direction and support.

Figure 23-20 Vert directs glide wire through total occlusion despite no visible opening.

A, Contrast injection at the site of occlusion fails to reveal an opening through which to advance the wire. B, Left vertebral diagnostic catheter (vert) was used to provide direction and support for the glide wire, which was advanced through the occlusion.

Figure 23-21 Vert can be used to cross occlusions without glide wire.

A, Peripheral venogram suggests total occlusion with extensive collaterals. B, Contrast injection at the site of obstruction reveals an opening. C, 5-French left vertebral diagnostic catheter (5F vert) is directed along the leads with puffs of contrast to confirm position. D, Tip of the vert is along the leads. Contrast injected at this position refluxes back along the leads into the collaterals, confirming it is still intravascular. With long occlusions, it may be useful to advance the vert alone as far as possible before utilizing the glide wire.

Figure 23-22 Vert and glide wire used together to cross a total occlusion.

A, Angled glide wire is advanced beyond the tip of the left vertebral diagnostic catheter (vert). The tip of the glide wire is off the axis of the leads. Once orthogonal views confirm the wire is along the lead, as it appears in the anteroposterior (AP) projection, the vert is advanced over the glide wire until the tip reaches the point where glide wire is off axis (stop vert advance). The glide wire is retracted and the tip of the vert then directed to the leads, and the glide wire advanced further along the leads. B, Tip of the vert is advanced to where the glide wire goes off axis near the junction of the innominate vein and the superior vena cava (SVC). C, Glide wire was withdrawn into the vert, the vert then directed toward the SVC, and the glide wire advanced. D, Glide wire is in the SVC. To ensure proper position, the wire should be advanced to the pulmonary artery and its position confirmed in orthogonal views.

Total Occlusion: Unable to Manipulate Wire

In most cases, a wire can be manipulated across the obstruction and venoplasty safely performed.³ However, some are wire refractory. Most of the approaches to a wire-refractory total occlusion have limitations. A very proximal subclavian stick increases the risk of hemothorax, pneumothorax, and, if successful, the risk of clavicular crush. The supraclavicular approach carries the risk of insulation damage or lead fracture from contact with the clavicle.⁵⁵ Tunneling requires a second incision (with associated risk of infection) and sacrifice of the other subclavian vein.⁵⁶ In addition, the tunneled lead is prone to erosion and fracture. When a wire cannot be advanced across the occlusion, there are two options: sacrificing one of the leads or using specialized crossing tools (microdissection, laser wire, radiofrequency guidewire).

Establishing Venous Access By Sacrificing an Existing Lead

Prepectoral Sheath Extraction (Powered or Mechanical)

Gula et al.⁵⁷ report that “central venous occlusion is not an obstacle to device upgrade with the assistance of laser extraction.” Although extraction of a functional lead can be performed safely in some centers by experienced operators, it still carries a small risk of serious complications (including death), loss of the functional lead, and dislodgement of adjacent lead(s) tethered to the lead being extracted. In addition, the lead may be extracted without regaining venous access. When the lead dislodges prematurely during prepectoral sheath extraction, the lead can be snared and held in the circulation while the sheath is forced through the fibrous tissue over the lead. However, it might be easier and safer to cut off the connector, insert a wire in the lumen, and extract the lead using the femoral approach, followed by venoplasty (see later). If the lead is not functional, extraction is a more appealing option, but careful assessment of the risk/benefit of the extraction procedure and the experience of the operator must be considered.⁵⁸ Before the development of the techniques described later, extraction was the only option for implantation of a new lead (see Chapter 26).

Prepectoral Wire under Insulation Technique Followed by Venoplasty

The lead to be sacrificed is partially extracted and a wire inserted under the insulation (Figs. 23-23 and 23-24). Using a stiff stylet, the lead is advanced back into the circulation, carrying along the wire. The wire and lead are separated by advancing the lead and fixing the wire. The lead is then extracted, retaining the wire. The ability to readvance the lead into the circulation with a stiff stylet depends on (1) implant duration, which determines the extent of fibrous tissue encasing the lead; (2) size of the wire under the insulation; and (3) lubricity of the wire under the insulation. This method is usually reserved for leads in place less than 1 month but can be successful up to 2 years if a hydrophilic .014-inch angioplasty wire is placed under the insulation. Once the .014-inch angioplasty wire is free in the circulation, it is exchanged for a .035-inch extra-stiff J wire using a 4F to 5F hydrophilic catheter. When used for leads implanted for more than 1 month, venoplasty is usually required because of the fibrous sheath that develops around the lead. When the lead cannot be advanced back into the circulation, femoral extraction can be used.

Figure 23-23 Angioplasty wire inserted under insulation of 7F pacing lead.

A, Thin-walled 21-gauge needle is inserted under the insulation (larger needles tear the insulation of 7-French leads) and the dry, extra-stiff, .014-inch hydrophilic angioplasty wire is inserted through the needle under the insulation then advanced beyond the tip of the needle as far as possible. The extra-stiff wire will advance further than a floppy wire. B, Needle is removed, retaining the wire under the insulation. It is important that the hydrophilic wire under the insulation is dry to maintain a high coefficient of friction.

Figure 23-24 Snare and wire under insulation of sacrificed lead to gain venous access.

A, Defective ICD lead in place for 18 months was unscrewed and withdrawn using traction alone, until the proximal coil reaches the pocket and the tip was in the right atrium (RA). B, A .014-inch extra-support hydrophilic wire was inserted 5 mm under the insulation starting 1 cm proximal to the proximal coil. C, Lead could not be advanced back into the circulation and thus was snared from femoral vein. D, Lead is pulled back into the circulation until the wire reaches the right atrium where the lead and wire are separated by holding the wire in place and continuing to pull the lead with the snare. Once the lead and wire are separated, the lead is extracted into the pocket. In retrospect, it makes more sense to remove the IS-1connector and extract via the femoral vein separating the lead from the wire. Further, once it becomes clear that the lead will not advance back into the circulation, a wire can be inserted into the stylet lumen rather than under the insulation.

Femoral Extraction with Wire in Stylet Lumen Followed by Venoplasty

When advancing the lead back into the circulation is not possible, the lead can be snared and extracted using the femoral route. With femoral extraction, the wire can be placed in the stylet lumen after the connector is removed, rather than under the insulation. Once the .014-inch angioplasty wire is free in the circulation, it is exchanged for a .035-inch extra-stiff J wire using a 4F to 5F hydrophilic catheter, followed by venoplasty.

Establishing Venous Access with Specialized Tools Using Leads for Direction

Various mechanical devices have been utilized to cross wire-refractory total occlusions, including a flexible hollow tube with screwlike surface contour (Tornus), microdissection, and laser. We evaluated the Tornus, but after our first case, had very limited success.⁵⁹

Microdissection

The Frontrunner XP CTO Catheter (Cordis Endovascular, Miami) is a device designed for microdissection through total occlusions in interventional radiology (Fig. 23-25). For short, total venous occlusions, the Frontrunner is often very successful^60,⁶¹ (Fig. 23-26), but in some cases the jaws will not engage the occlusion. Microdissection is a slow process; thus, although it is possible to open long, total occlusions, it requires a long procedure (Fig. 23-27). In addition to being a slow process, microdissection can result in the jaws of the device stuck in the fibrous tissue surrounding the lead beyond the total occlusion, unable to reenter the true lumen (Fig. 23-28). This is addressed with a directed needle reentry catheter⁶² (OutBack LTD Re-Entry Catheter, Cordis), a companion device of the Frontrunner (Fig. 23-29). The OutBack is an open-lumen catheter with a curved, extendable, retractable 21-gauge needle on the distal end. The OutBack with needle retracted is advanced through the channel created by the Frontrunner catheter, until the tip is in the fibrous tissue outside the distal true lumen. The needle is advanced with the tip angled toward the true lumen. A .014-inch guidewire is then advanced into the distal true lumen, the OutBack catheter is withdrawn over the wire, and venoplasty is performed.

Figure 23-25 Blunt microdissection tool.

A, The Frontrunner with jaws closed (top) and open (bottom). The ruler provides a reference for the size of the device (Frontrunner XP CTO Catheter, Cordis Endovascular, Miami). B, The Micro Guide Catheter through which the Frontrunner is passed. The lumen of the Micro Guide easily tracks over a 0.014-inch guidewire, but the lumen is large enough to accommodate two 0.014-inch guidewires.

Figure 23-26 Blunt microdissection tool successfully crosses total occlusion.

A, The length of the total occlusion (between two arrowheads) is short (3-5 mm). B, When the Frontrunner is advanced through the 6F sheath in the proximal axillary vein, it tends to be directed superior to the interface between the pacing leads and the occlusion. C, A short (65-cm) 6F multipurpose guide is advanced through the 6F sheath to direct the Frontrunner to the occlusion adjacent to the pacing leads. D, The Frontrunner (arrow) is across the lesion in the vein which is filled with contrast (C, arrow). The Micro Guide catheter (LuMend) is then advanced over the Frontrunner into the venous lumen, the Frontrunner is removed, a wire is advanced through the Frontrunner into the vein, and the Micro Guide is removed. Venoplasty is then performed.

Figure 23-27 Retrograde microdissection of a long total occlusion.

A, Both retrograde and antegrade catheters define the occlusion, which extends from the subclavian vein to the superior vena cava. B, Attempts to engage the occlusion are unsuccessful. Despite the use of various guide shapes, the microdissection catheter (Frontrunner XP CTO Catheter) repeatedly went into the collateral vein. C, Microdissection is successful in reaching the level of the open vein. However, on the first attempt, the microdissection catheter will not reenter the lumen. On the second try, starting at the inferior vena cava, the lumen of the vein is entered. A 4.5F straight braided guide with radiopaque tip (Micro Guide Catheter) through which the Frontrunner passes is advanced into the true lumen proximally. D, A microsnare is advanced through the sheath in the proximal vein. The wire originating in the femoral vein is snared into the sheath, providing access to the heart for lead placement.

Figure 23-28 Failure of microdissection catheter to enter true lumen despite successful microdissection beyond the occlusion.

A, The long total occlusion from the innominate vein to the superior vena cava (SVC) with extensive collateral vessels is shown. B, The long total occlusion is successfully crossed via antegrade microdissection along the inner curve of the leads. However, the Frontrunner does not cross into the true lumen, but remains in the fibrous tissue surrounding the pacing leads. C, The microdissection catheter and guide (Frontrunner XP CTO Catheter) are again advanced beyond the total occlusion, this time along the outer curve of the lead, but remain in the fibrous tissue surrounding the pacing leads. D, Contrast material injected into the guide catheter shows that it is in the tissues outside the SVC. Attempts to direct the microdissection catheter into the true lumen of the SVC are unsuccessful. The contrast staining had no clinical impact.

Figure 23-29 Directed needle catheter used to enter distal true lumen.

A, The short total occlusion with proximal and distal true lumen (arrows). B, The jaws of the microdissection catheter are open in the proximal portion of the occlusion. C, A nylon catheter with stainless steel braid and radiopaque tip (Micro Guide Catheter) is advanced over the Frontrunner through the track created in the fibrous tissue by microdissection. Despite being beyond the true lumen, the catheter is not in the true lumen. D, The OutBack Re-Entry catheter with needle retracted is advanced to the proximal portion of the distal true lumen. E, The curved hollow needle is advanced into the true lumen, and a .014-inch guidewire into the distal true lumen. F, Final venoplasty is performed with an interventional radiology, 6-mm-diameter, 4-cm-long balloon over a .035-inch Amplatz wire; predilation was performed with a 3-mm-diameter, 23-mm-long coronary balloon over the 0.014-inch guidewire, which was then changed for the .035-inch wire using the Micro Guide catheter.

Despite some success with microdissection, we found it time-consuming. Further, in some cases where the Frontrunner was in the fibrous tissue beyond the occlusion, we were not able to redirect the catheter back into the lumen with the OutBack. Ultimately, we were successful in only 60% of patients, and despite experience, the procedure time did not diminish. Finally, there is no central lumen; thus, to determine if it has reached the true lumen, a 4.5F guide catheter (Micro Guide Catheter, Cordis) must be advanced and the Frontrunner removed, creating a 4.5F hole.

Radiofrequency Guidewire

Several groups report successful recanalization of a long-standing, complete left subclavian vein occlusion by radiofrequency (RF) perforation with use of an RF guidewire in dialysis patients, some of whom have had a pacemaker on the same side.^62,⁶³ Radiofrequency is also used as an alternative to needle puncture in cases of complete superior vena cava (SVC) occlusion after Mustard repair.⁶⁴ Based on this experience, radiofrequency could be used to regain central venous access by adding a lead to an existing device. Although possible, it seems unlikely that the RF energy would be sufficient to damage the adjacent leads. As with microdissection, however, the RF wire does not have a central lumen; thus, to determine if it has reached the true lumen, a guide catheter must be advanced and RF wire removed. It is more comforting to have a device with a central lumen, where a wire can be advanced to confirm arrival in the central lumen.

Direct Needle Puncture

Honnef et al.⁶⁵ report using a transjugular intrahepatic portosystemic shunt set (TIPS needle) followed by venoplasty and stent to recanalize the chronically occluded left brachiocephalic of a hemodialysis patient. Sadarmin et al.⁶⁶ implanted an ICD despite chronic total occlusion of the SVC by directing a 21F Colapinto needle (Cook Medical) from the right internal jugular vein through the occlusion, followed by venoplasty and stent. The report was followed by an editor’s warning that insertion of a needle into a total occlusion is fraught with failure or complication, even with fluoroscopy. Concerns include perforating the vessel, then possibly reentering the true lumen distal to the occlusion.

Laser Crossing of Total Occlusions

The open-lumen excimer laser can be used to cross total occlusions in the arterial system.⁶⁷ The laser directs high-energy photons to cause molecular bond disruption at the cellular level (photoablation) without thermal damage to the surrounding tissue. The excimer laser transmits ultraviolet energy from the source (CVX-300TM Excimer Laser System, Spectranetics, Colorado Springs) to photoablate fibrous, calcific, and atheromatous material. Laser photons cause vaporization of contrast; thus residual contrast must be flushed from all catheters, the lasing site, and vascular structures adjacent to the lasing site prior to lasing. Laser catheters are indicated for treatment of both lower-extremity and coronary arteries when a total obstruction cannot be crossed with standard guidewires.¹⁶ The risk of perforation is greatest when the diameter of the laser catheter is close to the diameter of the vessel. In the subclavian, the 0.9-mm diameter laser is far smaller than the vessel. The laser must be advanced at less than 1 mm per second to allow time for the high-energy photons to disrupt the molecular bonds at the cellular level without thermal damage to the surrounding tissue.

Our experience with the laser began when microdissection failed,⁶⁸ but the laser was successful in crossing the total occlusion (Fig. 23-30). Subsequently, we have used the laser successfully,⁶⁹ now in 15 of 17 consecutive cases; 16 on the left and one right. The acute angle where the right innominate enters the SVC could make it more difficult; however, we used a 5F internal mammary artery (IMA) guide to successfully direct the laser. The clinical characteristics for our patients are shown in Tables 23-1. Sixteen of the 17 had existing leads; one patient presented with an infected system on the right and an occlusion on the left related to leads extracted 27 months previously for an infected system. Most of the upgrades were for an LV lead. Fifteen of the 17 cases were successful. The two unsuccessful laser-crossing cases both had long total occlusion where the lumen distal to the occlusion was stenotic, making it difficult to define as a target for the laser. There were no clinical complications, although contrast stains were observed in 11 of 17 patients. In six patients the stains were mild (<1 cm in diameter), in four moderate (1-2 cm), and one extensive (3.5 cm). Tables 23-1 and 23-2 present clinical and procedural details of the laser-crossing cases.

Figure 23-30 Peripheral and central extent of 10-mm occlusion well defined, and single application of laser energy successful.

A, Antegrade contrast injection at the site of occlusion defines both the peripheral and central extent of the occlusion. B, Tip of the 0.9-mm open-lumen laser (Turbo Elite OTW laser ablation catheter, Spectranetics, Colorado Springs) is at the occlusion. C, After one application of laser energy, the tip is in the lumen central to the occlusion. A wire was passed through the laser into the pulmonary artery and venoplasty performed.

TABLE 23-1 Clinical and Patient Characteristics of Laser-Crossing Cases

TABLE 23-2 Procedural Characteristics of Laser-Crossing Cases

As illustrated in Figure 23-30, when the peripheral and central extent of the occlusion is well defined by antegrade contrast injection and the occlusion is less than 10 mm, crossing the occlusion may require only a few applications of laser energy. However, peripheral contrast injection even at the site of occlusion may not define the central extent of the occlusion as well as it first appears (Fig. 23-31, A). When the laser did not reach the lumen central to the occlusion after four applications of energy, a femoral catheter was advanced to define this lumen (Fig. 23-31, B). In cases with known long total occlusions we start with both femoral and pectoral access (Fig. 23-32). The precise direction of the laser may not be apparent from the anteroposterior (AP) projection (Fig. 23-33). Thus, the tip position is always confirmed in orthogonal views before lasing. Depending on the length of the occlusion and position of the laser tip, forward lasing may be continued, or the laser withdrawn and redirected with the guide. Long total occlusions can be time-consuming and require multiple applications of laser energy. Before each subsequent application of laser energy, the tip is positioned firmly against the occlusion and its position confirmed in orthogonal views. When the lumen central to the occlusion is not well defined, it may be necessary to place a guide from the femoral in the central lumen and direct the laser into the lumen of the femoral guide (Fig. 23-34). When resistance decreases or the tip appears to be in the vein or guide central to the occlusion, the preloaded 300-cm, .014-inch angioplasty wire is advanced into the pulmonary artery (PA) or inferior vena cava (IVC) and venoplasty performed.⁷ If the wire does not advance freely to the PA, it must be assumed that the laser is not in the lumen; it is withdrawn and redirected with the guide.

Figure 23-31 Antegrade and retrograde venogram of a wire refractory chronic total occlusion.

A, Contrast injection at the site of occlusion clearly demonstrates the distal (peripheral) extent of the occlusion. From this venogram, it is not clear whether site 1 or site 2 is the proximal (central) extent of the occlusion. B, Central extent of the occlusion is defined via retrograde contrast injection. When the tip of the laser did not reach the central lumen after five laser applications, a 5F left vertebral catheter (Left Vertebral Slip-Cath with Beacon Tip, Cook, Bloomington, Ind) was advanced through an 8F braided sheath (Channel FX Transseptal Sheath, Bard Electrophysiology Division, Lowell, Mass) to site 2, the proximal (central) end of the occlusion. Site 1 is an out-of-plane collateral. Contrast injection defines the target lumen for the laser.

Figure 23-32 Venograms of a wire refractory chronic total occlusion.

A, Central extent of the occlusion is not defined by contrast injection at the site of occlusion through the dilator. B, Central extent of the occlusion is defined by a retrograde venogram. A 5F left vertebral catheter (Left Vertebral Slip-Cath with Beacon Tip, Cook) is advanced through an 8F guide (Channel FX Transseptal Sheath, Bard Electrophysiology) to probe for the innominate vein where the leads enter the superior vena cava. The tip of the left vertebral catheter is advanced and contrast injected to define the central extent of the occlusion.

Figure 23-33 Demonstration of importance of orthogonal views for laser crossing of total occlusions.

A, Tip of the 1.3-mm open-lumen laser appears to be in the lumen of the vein distal to the occlusion. The tip of the braided 8F femoral guide is at the superior vena cava (SVC)–innominate junction, supported by the .035-inch J wire curled in the lumen of the vein central to the occlusion. The 7F axillary guide adjacent to the leads directs the 1.3-mm open-lumen laser. In this projection the tip of the laser appears to be in the lumen of the vein central to the occlusion, as defined by the curled J wire. B, When the camera is rotated, the tip of the laser appears to be in the SVC, however, the wire does not advance freely.

Figure 23-34 When orthogonal views confirm tip positions, laser enters femoral guide.

A, Using the 7F guide, the tip of the laser is positioned at the tip of the 8F femoral guide. The femoral guide was advanced into the vein to provide a target for the laser. The .035-inch J wire is in position in the central portion of the innominate vein to keep the femoral guide in the vein. B, With application of laser energy, the tip of the laser entered the femoral guide.

Figures 23-35 and 23-36 illustrate a right-sided implant with total occlusion of the SVC above the left innominate vein. The patient has an arteriovenous fistula for dialysis in the left arm. The existing leads were implanted 5 years previously via the internal jugular and tunneled to the left side after extraction of an infected system on the right. Although the failed right ventricular lead could have been extracted for access, the lead would need to be tunneled from the right internal jugular to the left side over the clavicle and across the sternum, resulting in the same stress factors that caused the lead to fail initially. One patient with a chronic total occlusion without leads to follow required a laser to cross the obstruction (see Figs. 23-39 to 23-41). This case is discussed in detail in the section on total occlusions without implanted leads to follow or extract.

Figure 23-35 Superior vena cava (SVC) occlusion with patent left innominate vein.

A, Contrast injection in the right axillary vein reveals total SVC occlusion related to pacing leads that had been inserted in the right internal jugular vein and tunneled to the left pectoral area 5 years previously. B, Contrast injection at the site of occlusion through a catheter inserted in the leg demonstrates a patent left innominate vein. The patient has a left arteriovenous fistula for hemodialysis. C and D, In preparation for laser crossing, the extent of the occlusion is documented in the AP and RAO views via simultaneous contrast injection.

Figure 23-36 Laser opening of occluded superior vena cava (SVC).

A, The 0.9-mm laser is advanced through a 5F internal mammary artery (IMA) guide catheter (Sherpa NX Active IMA Medtronic Vascular, Danvers Mass). The IMA guide provides direction and support for the laser. The central extent of the occlusion is defined by the left vertebral catheter (“vert”) introduced through a braided 8F sheath (Preface Biosense Webster, Cordis, Miami). B, After three applications of laser energy, the tip is central to the occlusion. The 1.47-mm internal diameter of the IMA makes it difficult or impossible to track through dense fibrous tissue over the 0.9-mm laser. C, A 300-cm, .014-inch angioplasty wire is advanced through the central lumen and the laser removed. D, After venoplasty with a 3 × 15–mm coronary balloon, a vert is used to upsize to a .035-inch extra-stiff wire. A 6-mm × 4-cm Kevlar balloon is then advanced through the obstruction and inflated to 30 atm until the waist is eliminated and the pressure stable.

Limitations of Current Equipment and Procedure for Laser Crossing of Wire-Refractory Total Occlusions

For the purposes of crossing an occlusion, then advancing a wire, the over-the-wire (not rapid-exchange) laser catheter is required. The current 120-cm over-the-wire laser catheter is far too long for this application; the ideal length is 45 cm. Another limitation of the current equipment is the lack of an ideal companion guide. Keeping the laser advancing in the correct direction requires a guide catheter for support. To ensure that the guide can follow the laser, the lumen of the guide must match the outside diameter of the laser, and the guide should have a hydrophilic coating. On several occasions, the hole created in the fibrous tissue by the 0.9-mm laser was be too small to allow the guide to follow the laser, and we changed to the 1.3-mm-diameter laser. The ideal guide for this application would be 60 cm long, have a left vertebral shape, a hydrophilic coating, and an internal diameter (ID) of 1.0 mm. Finally, a means to better define and enlarge the lumen of the vein proximal (central) to the occlusion will make it easier to direct the laser.

The requirements for operation of the laser include (1) training in laser safety and physics, (2) review of laser operation, (3) demonstration of laser operation, (4). hands-on training, and (5) a company representative to assist for a minimum of three cases. For physicians already trained in laser lead extraction, minimal additional training is required.

Because a laser easily advances through tissue, it is possible to enter an artery, causing an arteriovenous fistula. Neurovascular structures, the chest cavity, or any other adjacent structure in the mediastinum could also be penetrated unless the tip location is confirmed in orthogonal views before and during lasing. A 1-mm opening created by laser perforation through a vein surrounded by fibrous tissue is very unlikely to have serious consequences, unless the perforation is not recognized and venoplasty is performed.

Preventing Complications

The laser catheter is used safely in the arterial system, where high pressure makes the consequence of perforation much worse than in the venous system. In addition, lasing is not really being performed in a vein, but rather into the extensive, dense fibrotic tissue surrounding the preexisting leads. To prevent barotrauma from laser vaporization of contrast, residual contrast must be flushed with saline from all catheters, the lasing site, and vascular structures adjacent to the lasing site before lasing. To prevent thermal damage to the surrounding tissue, the laser is advanced at less than 1 mm per second. To prevent perforation, the laser tip position is confirmed in orthogonal views before and during lasing, ensuring that the laser catheter remains in the fibrous tissue until it reaches the lumen proximal (central) to the occlusion.

Although laser perforation is the complication most often considered, the real danger results from posterior deviation of the wire into a small branch vein or collateral that is not apparent on the AP projection. To ensure that the angioplasty wire is in the true lumen as it is advanced, the tip must move freely, remaining in proximity and parallel to the leads until it reaches the PA or IVC. Unrecognized wire deviation seems to occur most often at the innominate-SVC junction. Fortunately, wire deviation is easily recognized when the wire fails to advance into the PA or femoral vein. Proper wire position is confirmed by panning along the length of the wire in the left anterior oblique (LAO) and right anterior oblique (RAO), as well as AP projection. The balloon should never be inflated (or progressively larger dilators be advanced) until the wire is in the femoral vein or PA and its position adjacent to the leads is confirmed by orthogonal projections. Unlike laser lead extraction, where catastrophic complications can occur despite experience and optimal technique, venoplasty following lasing should not result in a catastrophic complication, unless wire deviation/perforation is unrecognized and followed by venoplasty. Thus, we do not require operating room backup or arterial pressure monitoring, as recommended for lead extraction.

Summary

We find laser crossing of wire-resistant total occlusions to be successful in a high percentage (15 of 17) of patients who needed device upgrades. The safety of the procedure remains to be established in a multisite evaluation. If directions are followed and precautions heeded, this technique for crossing wire-refractory total occlusion can be learned by physicians with training and experience in venoplasty and laser lead extraction and deserves further evaluation. Availability of a short (45-cm), over-the-wire laser and a 60-cm-long, left vertebral–shaped guide with hydrophilic coating and ID of 1.0 mm will facilitate the procedure.

Total Occlusions in Patients Without Implanted Leads

Access to the central vasculature can be regained using interventional techniques. However, most EP physicians send patients with bilaterally occluded subclavian/innominate vein for an epicardial or a femoral system. A variety of occlusions can be successfully opened, although the success rate is lower than in patients with retained pacing or defibrillation leads. The occlusion may be short (Fig. 23-37) or long (Fig. 23-38). There may be flush occlusion at the innominate-SVC junction (see Fig. 23-38), or the innominate may be patent and well reconstituted central to the occlusion (see Fig. 23-37). It may be possible to cross the obstruction with a wire, or it may require a laser (Figs. 23-39 to 23-41).

Figure 23-37 Venous access despite total subclavian occlusion without pacing leads to provide direction.

A, Brachial vein contrast injection demonstrates total occlusion of the subclavian vein. B, Central extent of the occlusion is defined by contrast injection from a catheter introduced from the left femoral vein. C, A venogram balloon is inflated to facilitate visualization of the central extent of the occlusion and the hydrophilic 5F left vertebral catheter (vert) is advanced across the occlusion. D, A .035-inch extra-support guidewire is advanced into the pulmonary artery through the vert, the vert is removed, the balloon is advanced across the occlusion and inflated to eliminate the obstruction.

Figure 23-38 Venous access despite long total subclavian occlusion without pacing leads to provide direction.

A, Innominate/subclavian vein is seen to be occluded along its entire length. SVC, Superior vena cava. B, Dilator from a 5F sheath is advanced over a glide wire into the collaterals. C, Glide wire enters the SVC. D, Venoplasty is performed to open the innominate for lead placement.

Figure 23-39 Total occlusion with no lead to follow or extract that required laser: part 1.

A, Venogram central to the occlusion reveals the innominate is reconstituted and widely patent beyond the occlusion. Vert, Left ventricular diagnostic catheter. B, Peripheral venogram reveals collaterals and a small axillary vein. C, On left anterior oblique (LAO) view, the vert introduced in the axillary vein is advanced to the tip of the femoral vert. An angioplasty wire would not advance. D, Anteroposterior (AP) view confirms the position of the two catheters. Despite the direction and support of the vert, the wire would not cross the occlusion. Unsuccessful attempts were also made to advance wires from the vert central to the occlusion.

Figure 23-40 Total occlusion with no lead to follow or extract that required laser: part 2.

A, The 6F Rapido advanced from the axillary vein seems to be well aligned with the femoral vert. The inner diameter (ID) of the vert is too small for the laser to pass. A 6F sheath was placed and the 6F outer-diameter (OD) Rapido advanced. B, After advancing the laser under power to the tip of the vert, an angioplasty wire was advanced but did not reach the lumen of the vein central to the occlusion. The Rapido and vert were realigned and position confirmed in LAO (C), AP, and RAO (D) projections.

Figure 23-41 Total occlusion with no lead to follow or extract that required laser: part 3.

A, After the tip of the Rapido from the axillary vein is realigned with the tip of the vert from the femoral vein, the laser is introduced and advanced toward the tip of the vert. As the laser is advanced, the tip of the vert is pushed back. B, A 300-cm, .018-inch extra-support angioplasty wire is advanced into the lumen of the vert. C, Angioplasty wire is clearly seen in the lumen of 8F femoral sheath when the vert is withdrawn. Using a 5F hydrophilic catheter (vert) the angioplasty wire is exchanged for a .035-inch Amplatz Extra Support Wire. D, A 6 mm × 4 cm, noncompliant peripheral balloon is advanced and inflated to eliminate the stenosis.

When approaching a case with known occlusion and no leads to follow we start by evaluating the central extent of the occlusion. First, a long, 8F braided sheath (Preface Guiding Sheath, Cordis) is advanced to provide a stable supportive platform in the SVC. A 5F left vertebral diagnostic catheter is then advanced into the SVC and puffs of contrast used to locate the innominate (Fig. 23-37, B). To make the central extent of the occlusion easier to visualize for the prepectoral approach to the occlusion, a venogram balloon is inflated (see Fig. 23-37, C). Moving from the femoral to the prepectoral area, venous access is obtained in the axillary vein peripheral to the occlusion, and a 5F sheath inserted. Using the venogram as a target, the obstruction is crossed with a glide wire and a 5F left vertebral catheter (see Fig. 23-37, C). Venoplasty is performed once the wire is across the occlusion and position is confirmed in orthogonal views (Fig. 23-37, D). Figure 23-38 is an example of a long total occlusion without wires to follow where there is flush occlusion of the innominate as it enters the SVC. When it is not possible to cross into the innominate vein central to the occlusion with a wire, a laser can be used to bridge the gap. See Figures 23-39 to 23-41 for details of the procedure. Box 23-4 lists venoplasty equipment. Box 23-5 lists the equipment required for crossing an obstructed axillary/subclavian vein and performing venoplasty. Subclavian venoplasty is covered in the next section.

Box 23-5
Equipment for Crossing and Dilating Stenotic/Occluded Subclavian Veins

Contrast injection system: injection syringe, stopcock, extension, reservoir syringe, Y adapter

.018-inch angled-tip hydrophilic guidewire, 150 to 180 cm in length (Angled Zipwire, Boston Scientific)

.035-inch angled-tip hydrophilic guidewire (Angled glide wire, Terumo, or Angled Zipwire, Boston Scientific)

Steering handle for glide wire: torque device for .025- to .038-inch wires (Mallinckrodt)

Extra-support wire = .035 inch, 145 cm, 3-mm tip curve (Amplatz Extra Stiff, Cook)

Ultra-support wire = .035 inch, 260 cm, 3-mm tip curve (Amplatz Ultra Stiff, Cook)

4- to 5-French hydrophilic catheter with 10- to 12-mm angled tip, 100 cm (Vert Slip-Cath Beacon Tip, Cook, or Angled Taper GlideCath, Terumo)

Noncompliant, 6 mm × 4 cm, subclavian venoplasty balloon

For wires ≤0.035 inch (PowerFlex-P3, Cordis)

For wires ≤0.018 inch (Sterling, Boston Scientific)

Ultra-noncompliant, 6 mm × 4 cm, subclavian balloon for refractory stenosis (Conquest, Kevlar); rated burst = 30 atm (Bard)

30-mL syringe (to pull negative pressure for balloon preparation)

Y adapter with Luer lock

Small bowl for mixing contrast, 50/50 (contrast/saline)

See Interventional Equipment List at the end of Chapter 22 and online for details and ordering information.

John Gurley and the group from the Gill Heart Institute at the University of Kentucky described inside-out central venous access (IOCVA) in eight patients.^69a IOCVA is an important, well-reasoned new approach for central venous access that can be applied to patients with chronic total SVC and subclavian vein occlusions that cannot be recanalized with currently available technology. IOCVA is a new concept that reverses the approach by directing a needle outward from within the body. The authors reasoned that IOCVA can be performed safely because there are consistent anatomic relationships adjacent to the thoracic central veins. The central veins are bounded anteriorly by the clavicle, soft tissues, and skin. A needle directed anteriorly will not encounter arteries, nerves, or pleura, which lie posterior and lateral to the central veins. IOCVA allows the placement of conventional leads and devices in optimum locations.

Subclavian Venoplasty: Response to Stenosis between Axillary Vein and Right Atrium

In this section we use the term subclavian venoplasty as shorthand to describe dilation of the axillary, subclavian, and/or innominate veins. Once the wire is successfully across the obstruction, the fibrous tissue surrounding the leads must be dilated to allow unfettered access to the cardiac chambers. At this point the options are venoplasty or progressively larger dilators. The use of progressively larger dilators is an anachronism reflecting that surgeons, not cardiologists, initially implanted devices. In the late 20th century, when EP physicians took over device implantation from the surgeons, we followed their lead in part because the EP skill set did not include balloon dilation. The modern approach to vascular obstruction should favor balloon dilation over using progressively larger dilators.

Venoplasty vs. Progressively Larger Dilators

Progressive larger dilators have been used for many years to enlarge the opening of a focal subclavian stenosis but were often unsuccessful when the obstruction extended from the subclavian vein to the SVC. When a focal peripheral stenosis is opened, approximately 30% have a stenosis at the innominate-SVC junction that is more difficult to address with dilators. Even when a “subclavian” stenosis can be opened with dilators, the result is always incomplete. The frequency of this problem is increasing as we perform more device upgrades. In most cases with progressively larger dilators, the stenotic area continues to restrict sheath and lead manipulation. Implantation of a right atrial (RA) or right ventricular (RV) lead after incomplete innominate/subclavian dilation is time-consuming and frustrating but often possible. However, it adds a significant degree of difficulty to a CRT implant that more frequently can result in failure. Restricted access makes it more difficult to cannulate the CS, and all subsequent manipulation of catheters and leads is impaired. Friction between the lead and catheters makes final sheath removal more likely to result in LV lead dislodgement.

Drawbacks to Progressively Larger Dilators

Prolonged Implant Time

Use of dilators prolongs the case because it takes longer to dilate the occlusion with dilators than balloons. Also, catheter and lead manipulation is restricted, and thus each step in the implant procedure is more difficult and takes longer.

Failed Implant

Use of dilators (not balloons) reduces the chance of a successful implant because many obstructions, particularly those with a distal stenosis, cannot be dilated successfully with dilators, resulting in a right-sided implant that is more difficult and less likely to be successful. Also, restricted catheter manipulation makes the implant more difficult, increasing the risk of failure.

Increased Risk of Infection

Use of dilators (not balloons) increases the risk of infection because of the longer procedure times. Also, when the dilators fail and the other side is used, there are two incisions, each with its own risk of infection. When laser extraction is used for access, the pocket must be opened twice instead of once, if the laser and surgical backup are not immediately available during the initial procedure.

Increased Cost

Use of dilators (not balloons) increases the cost of the procedure. Dilator failure may result in lead extraction for access, requiring a second procedure with general anesthesia, and a surgical consult. Prolonged implant time increases the staff and room costs and reduces the number of cases that can be performed.

Suboptimal Clinical Result

Use of dilators (not balloons) results in a suboptimal result because of increased risk of infection, loss of venous access from using the other side, increased risk of lead erosion if the lead is tunneled, and increased risk of death if extraction is employed for access. Increased mortality results if the patient has a failed implant and is sent for an epicardial lead, or if the patient has a failed lead implant and is not sent to surgery.

Increased Risk of Complications with dilators

Because venoplasty is new and many implanting physicians have no experience with balloons, there is a tendency to regard progressively larger dilators as safer than venoplasty. Paradoxically, the concern that venoplasty is dangerous probably makes it safer than dilators because more attention is paid to the wire position.

Figures 23-42 and 23-43 illustrate a catastrophic vascular event during the use of progressively larger dilators. It is important to note the glide wire did not perforate the dilator but simply exited through a branch feeding the innominate vein. The branch vein is a thin-walled structure not encased in fibrous tissue; thus, advancing the dilator lacerated the vein. Wire position in the AP projection can be deceptive. To be certain the wire has not exited the vein, progressively larger dilators (and balloons) should never be advanced until the wire is in the PA or IVC and confirmed to be adjacent to the leads in orthogonal views.

Figure 23-42 Hemodynamic collapse during use of progressively larger dilators to open subclavian occlusion.

A, Venogram produced by contrast injection in a peripheral vein reveals subclavian obstruction with collaterals. B, Venogram reveals obstruction at the superior vena cava (SVC)–innominate vein junction. C, Progressively larger dilators are advanced over the .035-inch angled-tip hydrophilic wire. The wire appears to be in the vein in the AP projection. D, With the development of hemodynamic collapse, contrast was injected through the sheath and flowed freely into the pleural space.

Figure 23-43 Left anterior oblique (LAO) projection reveals hydrophilic wire and contrast in pleural space.

A, Hydrophilic wire is seen to exit the innominate vein into a branch vein off the innominate and into the pleural space. B, Contrast injected through the sheath flows freely into the pleural space through the laceration in the venous branch produced by the dilator. The wire exited the innominate vein through a branch vein peripheral to the obstruction at the SVC-innominate junction. It did not perforate the vein.

Performing Subclavian/Innominate Venoplasty

As previously discussed, it is important to define clearly the situation with contrast injection at the site of occlusion (Fig. 23-44). Before starting venoplasty, the implanting physician must consider which wire is across the occlusion and what level of support is required to advance the balloon beyond the obstruction. The .035-inch glide wire that is most frequently used does not provide optimal support. Thus, upsizing to a .035-inch extra-stiff wire using a 4F to 5F hydrophilic catheter is a worthwhile first step rather than first trying to advance the balloon.

Figure 23-44 Arm and local venograms of a typical high-grade stenosis at axillary-subclavian junction.

A, From 15 to 20 mL of contrast material is injected into a vein in the distal left arm, revealing a stenosis at the junction of the axillary and subclavian veins. B, The contrast material is injected through a 5F sheath inserted in the axillary vein proximal to the stenosis. To create a flexible stable platform from which to approach the occlusion, one obtains venous access and inserts a 5F sheath percutaneously in the axillary vein as distal to the stenosis as possible. If venous access is too close to the occlusion, it will also be difficult to direct the wire to different potential routes through the stenosis.

In patients with previously implanted leads, venoplasty may be required anywhere between the right atrium and the axillary vein (Fig. 23-45). The most common site of obstruction is in the area of the clavicle (Fig. 23-46). A central occlusion is encountered in approximately 30% of patients with a peripheral occlusion. An isolated central occlusion accounts for approximately 10% of venoplasties. The lumen of the balloon must be large enough to accommodate the wire across the stenotic area, usually a .035-inch angled glide wire (Terumo Medical, Somerset, NJ). In addition, the inflated diameter of the balloon must be large enough for the 9F sheath (9F ID/11F OD) to pass freely after dilation (6 mm). Lastly, the balloon must be long enough to dilate from the right atrium to the axillary vein with three to five inflations (4 cm).

Figure 23-45 Dilation of a proximal subclavian vein occlusion.

A, Collateral vessels are seen (arrows) around the subclavian vein occlusion that developed in association with the leads of a previous implantable cardioverter-defibrillator (ICD). B, Despite what appeared to be a total occlusion, a 0.035-inch angled Glide wire (Terumo Medical, Somerset, NJ) crossed the occlusion and advanced through the heart into the pulmonary artery. A 6-mm-diameter, 4-cm-long balloon with a 0.038-inch wire lumen is across the occlusion and inflated to 15 atm, where two residual waists are seen. With inflation to 20 atm, both waists were eliminated.

Figure 23-46 Most common sites of occlusion associated with device leads.

Diagram of venous anatomy shows that the most common site of obstruction is in the area of the clavicle. A central occlusion is encountered in approximately 30% of patients with a peripheral occlusion. An isolated central occlusion accounts for approximately 10% of venoplasty procedures.

Initially, the balloon is inflated to the “nominal pressure” (listed on the package). At the nominal pressure, the balloon reaches the labeled diameter. The pressure is then increased until the “waist” opens or the balloon bursts. A balloon typically used by interventional radiologists is the PowerFlex P3, which is 6 mm × 4 cm with a .038-inch lumen (Cordis, Miami). If the stenosis does not submit to the initial balloon, an ultra-noncompliant (aramid fiber [Kevlar]), high-pressure balloon may be used (Conquest PTA Dilatation Catheter [Bard Peripheral Vascular, Tempe, Ariz.] or equivalent). Once the proximal stenosis is dilated, the balloon is advanced to the RA-SVC junction and inflated to nominal pressures from the right atrium back to the axillary vein in an overlapping manner. In this way, any occult distal stenotic areas are revealed and opened (Fig. 23-47).

Figure 23-47 Dilation of both proximal (peripheral) and distal (central) subtotal venous occlusions encountered in same patient.

A, The balloon is inflated at the site of the proximal occlusion to 15 atm with elimination of the waist and proximal stenosis. When the sheath was advanced, we encountered an obstruction at the SVC-innominate junction. B, The balloon is advanced to the junction of the superior vena cava and innominate vein and inflated, demonstrating a waist in the balloon in the second stenotic location. If the stenosis is severe, winging of the balloon after the initial inflation may make it impossible to advance to the SVC-innominate junction. Because 30% of patients with a peripheral stenosis also have a central stenosis, we now start our dilation at the SVC-innominate junction.

Balloons for Subclavian Vein Venoplasty

Why not use coronary balloons for subclavian venoplasty? We switched from coronary balloons to peripheral balloons for several reasons, including the diameter of the wire across the stenosis. When dealing with stenotic subclavian veins, we usually start with a .035-inch angled-tip glide wire. Coronary balloons track over wires .014 inch or less; thus, to use a coronary balloon, the wire must be downsized to .014 inch, where many peripheral balloons are designed to track over wires .038 inch or less. In addition, coronary balloons are less than 21 mm in length. The number of head-to-tail inflations required for the 21-mm balloon (Fig. 23-48, A) to cover the vein from the SVC-innominate junction to the pocket is prohibitive compared to the 40-mm peripheral balloon (Fig. 23-48, B). Finally the diameter of most coronary balloons is less than 4 mm, which is too small for passing a 9F sheath. Although we do not know the optimal size of balloon, we started with a peripheral balloon 6 mm × 4 cm in diameter and have not found the need to change after experience with more than 300 cases.

Figure 23-48 Comparison of “long” to “peripheral” balloon for dilation of occlusion at superior vena cava (SVC)–innominate vein junction.

A, Coronary balloon is inflated. B, Peripheral balloon is inflated. Note the difference in size. Consider the number of head to tail inflations required to go from the SVC-innominate junction to the pocket with the coronary balloon compared to the peripheral balloon.

A coronary balloon may be necessary to predilate the lesion when the only wire to cross the occlusion is .014-inch or .018-inch angioplasty wire (Fig. 23-49). The Hi-Torque Cross-It 300XT (Guidant, Boston Scientific, Natick, Mass.) and the .014-inch Crosswire (Terumo Medical), among others, are specifically designed to cross total coronary occlusions. They may be used to cross a subclavian occlusion when the .035-inch wire will not advance. When the only wire that will cross the lesion is a .014-inch, we start with a 2- to 3-mm coronary balloon. Once the stenosis is dilated with the coronary balloon, our usual .038-inch lumen peripheral balloon may track over the .014- or .018-inch wire.

Figure 23-49 Predilation of subclavian occlusion with coronary balloon.

A, The .035-inch angled Terumo Glide wire will not cross the total occlusion. B, The 0.014-inch Terumo Crosswire is successful in crossing the total occlusion. C, Coronary balloon is inflated at the site of occlusion. D, The 6-mm-diameter, 4-cm-long peripheral balloon with a .038-inch lumen is advanced over the .014-inch Terumo Crosswire and inflated. Peripheral balloons designed to track over .018-inch wires are now available if the .038-inch lumen peripheral balloon will not track over the .014-inch wire. Alternatively, the .014 wire can be upsized to a .035 extra-stiff wire using a 4F to 5F hydrophilic catheter (e.g., “vert”).

We now have two choices: (1) a .018-inch lumen peripheral balloon, or (2) upsize to a .035-inch extra-stiff wire (Amplatz Extra Stiff Wire Guide, Cook Vascular, Leechburg, Pa.) and dilate with a “standard” peripheral balloon.

Caution: When dealing with a peripheral stenosis, it is tempting to use a short dilator to upsize to a larger wire. Avoid this mistake at all cost. Over 30% of patients with a peripheral obstruction also have a central occlusion that the hydrophilic wire may cross without the obstruction becoming apparent. If the hydrophilic wire has crossed an unrecognized central stenosis and a short dilator is used, the extra-stiff J wire will not advance through the obstruction. Attempting to advance this J wire through an obstruction tends to disrupt the lumen, making it impossible to insert the hydrophilic wire back into the central circulation. When upsizing or exchanging a wire, always use a catheter long enough to reach the PA or IVC. A 4F to 5F hydrophilic catheter is ideal for this purpose. A nonhydrophilic catheter of the same size may not advance over the wire.

In the past, we usually upsized the wire rather than try to track the .038-inch lumen balloon over the .014- or .018-inch wire. Although the .018-inch lumen peripheral balloon is an option, our experience with these new balloons is minimal, and it is reassuring to have a .035-inch extra-stiff wire across the occlusion.

In summary, coronary balloons are too short, their diameters too small, and their lumens too narrow for the typical .035-inch access wire. Coronary balloons are used when a .014-inch wire is all that will cross the occlusion. As soon as the stenosis is partially dilated, the .014 wire is usually replaced with a .035 wire (see Fig. 23-49). Recent availability of .018-inch lumen peripheral balloons provides more options.

Short Balloons for Central Stenosis (Superior Vena Cava–Innominate Vein Junction)

As previously noted and shown in Figure 23-47, many patients with a peripheral stenosis also have a central stenosis. In these cases, progressive dilation of the proximal occlusion leaves the distal stenosis untouched, further restricting lead manipulation. In some patients, only a distal occlusion is encountered. Initially, we used short balloons (<20 mm) for the SVC-innominate junction occlusion (Fig. 23-50), thinking there was too sharp an angle to accommodate a longer balloon. However, on further reflection we realized that the wire is not constrained with the lead but enters the SVC; thus the central end of the balloon is free and is not forced to follow the bend of the leads. Figure 23-50 shows a distal-only occlusion that appears to be on a sharp bend, dilated with a short balloon at multiple sites.

Figure 23-50 Dilation of a total occlusion on a bend with short balloons.

A, The proximal subclavian vein is open, but the distal vein at the junction with the superior vena cava (SVC) is totally occluded. Collateral vessels have developed around the occlusion. B, Despite what appears to be chronic total occlusion, an angioplasty wire designed to cross total arterial occlusions is successfully advanced into the right atrium. A short (23-mm) balloon is inflated at the site of occlusion. C, The short balloon is again inflated, this time more distally. The short balloon is chosen because the occlusion occurs on a bend. Contrast material used to facilitate crossing of the total occlusion can be seen staining the tissues at the site of previous occlusion. No adverse clinical event occurred despite the contrast stain.

Kevlar Balloons and Focused-Force Venoplasty

In some patients, the stenosis is particularly resistant to dilation, requiring special balloons. In Figure 23-51, an angled .035-inch angled glide wire was manipulated across the stenosis, but the sheath would not advance. The stenosis was dilated with a 6 mm × 4 cm balloon delivered over the glide wire. Figure 23-51, A, demonstrates the balloon dilated at 16 atmospheres (atm); at 20 atm, the balloon ruptured without eliminating the stenosis. A Kevlar balloon (Conquest) was then advanced across the occlusion (Fig. 23-51, B). Once the second balloon was inflated to 22 atm, the stenosis was relieved, and the 9F sheath (9F ID/11F OD) passed easily. When the Kevlar balloon is not effective, a stiff wire is placed beside the balloon. This is a useful technique for a focal stenosis that does not respond, as seen in Figure 23-52. Focused-force venoplasty is also useful for an elastic obstruction,⁷⁰ where the balloon inflates completely but rebounds as soon as the balloon is deflated (Fig. 23-53).

Figure 23-51 Venoplasty of a stenotic subclavian vein that required an aramid fiber (Kevlar) balloon to eliminate the waist.

A, The 6-mm-diameter, 4-cm-long peripheral balloon is inflated at the site of stenosis to rated burst pressure (16 atm) but the waist remained. At 20 atm the balloon burst. B, Kevlar balloon replaces the standard balloon. The waist is eliminated at 22 atm.

Figure 23-52 Focused-force venoplasty to resolve a focal obstruction.

A, The waist remains despite inflating the PTA balloon to 16 atm. B, The waist remains despite inflating the Kevlar balloon to 30 atm. C, Kevlar balloon is inflated to 20 atm with an extra stiff .035-inch guidewire beside the balloon to focus the force of the inflated balloon on the residual waist (focused-force venoplasty). D, When Kevlar balloon is inflated to 30 atm against the wire, the obstruction opens.

Figure 23-53 Focused-force venoplasty for an elastic obstruction.

A, Central stenosis appears complete, but a trickle of contrast is seen to cross (black arrowheads). The hydrophilic 5F left vertebral diagnostic catheter (vert) is redirected, placing the tip at the opening, and an angioplasty wire is advanced across the occlusion. B, Coronary balloon is placed across the obstruction and inflated until the waist is eliminated. The .014-inch wire is exchanged for a .035-inch extra-support wire using the hydrophilic 5F vert. C, A 6-mm × 4-cm balloon was inflated to 16 atm and the waist eliminated; however, the sheath would not advance. A Kevlar balloon was inflated to 30 atm, and still the sheath would not advance. D, Balloon was exchanged for a long 5F sheath. Two .035-inch extra-stiff wires were advanced into the right ventricle and the sheath removed. The balloon was advanced over one wire beside the other and inflated to 30 atm, after which the sheath advanced.

“Complications” of Subclavian/Innominate Venoplasty

Despite our initial concerns, we find balloon dilation of venous structures remarkably safe as long as the balloon is never used until the wire is advanced successfully to the PA. Figure 23-54 illustrates the potentially catastrophic consequences of not heeding this axiom; in A the wire position appeared satisfactory, but it would not advance to the PA, the tip apparently stuck in the RA; the LAO view in B illustrates how deceptive wire position can be in the AP projection. Figures 23-42 and 23-43 illustrate the consequence of not abiding by the rule when using progressively larger dilators (not venoplasty).