(Video 9.1). Dual injection showing a long RCA CTO, with the distal cap at a major bifurcation, and good septal collateral channels from the left system. A 6F XB 3.5 90 cm in used from the right radial into the left main, and an 8F JR 4 from the femoral approach into the RCA. A long 45 cm introducer is used from the femoral
(Video 9.2). Dual injection using a bilateral transradial approach. An 8Fr AL1 is inserted from the left radial, using a sheathless technique, and an XB 3.5 90 cm from the right radial
Pathways to the Distal Cap (Table 9.1)
Retrograde pathways to the distal cap of the CTO
Epicardial or intra-myocardial non-septal CC
Skill required to cross with wire
Risk as a conduit
Large, not too tortuous, previous cardiac surgery
Severely diseased SVG, several in-stent restenosis
Surgical clips to delineate trajectory, aortic stump
Able to dilate?
Higher risk if
Single tortuous CC
In the atrio-ventricular groove, intact pericardium
Internal mammary artery, especially if tortuous
Stump less occlusion, absence of clips
There are two types of conduits that can be used to deliver retrograde wires and microcatheters to the distal cap: collateral channels (CC) and surgical grafts. While CCs can be divided into septal CC, epicardial or intra-myocardial non-septal CC, surgical grafts can be divided into patent or occluded. The following section provides general principles with both conduits.
Septal channels are clearly the most commonly used CCs for the retrograde approach. They are safer than epicardial CCs, as a perforation usually fenestrates into a ventricular chamber or lead to benign septal hematomas, or arteriovenous fistulas. An exception to this general rule is that occasionally very proximal septal CC will communicate with one right postero-lateral (RPL) branch, or distal septal CC can emerge at the free wall of the RV and thus become by definition epicardial at some point during their course; perforation of such channels at these areas may lead to tamponade. The septal channels are best evaluated in the straight right anterior oblique (RAO) view, or the RAO cranial view (typically 30° right, 30° cranial) (Figs. 9.3 and 9.4, Videos 9.3, 9.4, and 9.5). Occasionally an RAO caudal view is useful to assess the collateral channel anatomy at the base of the heart, as they enter, or exit the PDA. The operator should look for the presence of connections from the donor vessel to the receiving vessel. Typically, septal CCs will collateralize a right coronary artery (RCA) CTO to the posterior descending artery (PDA) from the left anterior descending (LAD) artery, or an LAD CTO from the PDA. In the left dominant system, the septal CCs will collateralize an LAD CTO from a left PDA, or vice versa.
(Videos 9.3 and 9.4). Crossing septals using the surfing technique. (a) Presence of several septal collateral channels (CC), most CC1 and CC0, with possible connection to the recipient vessel indicated by the arrow (Video 9.3). (b) Finecross in proximal septal trunk. Connection with a Sion wire into the PLV branch after surfing (Video 9.4)
(Video 9.5). Presence of several septal collaterals and one epicardial collateral. Dual injection showing several CC0 and CC1 connections to the PDA. However, there is also a non-interventional collateral from the distal OM branch to the tip of the PDA. This epicardial and very tortuous CC is dominant and non-interventional (arrow)
The Rentropp score has been used to evaluate the efficiency of collateral flow to the occluded vessel, but is less useful to identify collaterals that can be used for the retrograde approach . There are methods to analyse which CC is most effective in providing flow to the receiving vessel. When looking at the angiography and analyzing it frame-by-frame, it is often clear to see where the receiving vessel first lights up with contrast. With this information, by playing the cine-angiogram frame-by-frame, forwards and backwards, we can identify the channel that provides this preferential flow to the vessel. It is usually the largest CC, but not necessarily the one that is the best to be crossed. Indeed, we usually prefer to attempt crossing other septal CCs to allow for the largest to provide flow to the target vessel during the intervention, when possible.
The Werner classification identifies collaterals as either invisible (CC0), as thread-like tinny vessels (CC1) or as small vessel-like channels (CC2) (Fig. 9.5, Videos 9.6, 9.7, 9.8, 9.9, and 9.10) . This classification has been shown to be helpful to predict success in crossing a septal CC when using a Japanese-style technique with selective tip injection . However, when using a different technique such as the septal surfing, this classification is much less useful. We will discuss this technique and how to cross collaterals in the next section. It is also important to look if there is a large septal vessel trunk leading to the CCs. This is often the case with LAD to PDA CCs. However, as the branching of collaterals increases with distance from the LAD towards the PDA, the septal CCs are often very small and invisible close to the PDA (Figs. 9.4 and 9.5, Videos 9.5, 9.6, 9.7, 9.8, 9.9, and 9.10). This doesn’t mean that they don’t exist or connect however. As a general principle, CCs are most likely to be crossed easily if larger and straighter. Tortuosity of a septal CC is an issue and may preclude crossing with the wire. However, invisible collaterals, if straight, can be crossed with a surfing technique as explained later. Sometimes, challenges may arise at the entry of the septal CC, because of severe angulation or calcification. Despite connecting, such collateral may be very difficult to engage.
(Videos 9.6, 9.7, 9.8, 9.9, and 9.10). CC0, CC1 and CC0 channels. (a) Black arrow indicates a large but tortuous CC2 septal channel. However, the first septal (S1) provides small CC1 channels and even CC0 channels (white arrow) (Video 9.6). (b) Selective injection with FineCross shows how tortuous the largest collateral is (Videos 9.7, 9.8, and 9.9). (c) After a failed crossing attempt through the CC2, successful crossing was achieved using a surfing technique from S1, though an invisible CC0 connection (Video 9.10)
(Video 9.11). Interventional vs. non-interventional epicardial collateral channels. A post CABG RCA CTO receives very large but tortuous epicardial collateral (back arrow) from the native distal LCX to the PLV and also a straighter interventional epicardial CC from the OM branch (grafted with a LIMA) to the PDA (white arrow)
The epicardial collaterals are a separate subset of connections and pose a completely different set of challenges and considerations (Fig. 9.6). They typically arise from the postero-lateral (PL) to the obtuse marginal (OM) or diagonal branches, from the distal LCX to the PLVs, from diagonals to OM or vice-versa. They can also arise from the apical LAD or posterior descending artery (PDA) and feed any CTO vessel with a course on the surface of the heart. Also, epicardial CC are common from the RCA on the surface of the right ventricle, connecting or arising from the conal branches (with the LAD) or the acute marginal branches. They tend to be friable and by their location, they are at a higher risk of rupture following wire or microcatheter crossing. Larger collateral are the ones to prefer. Tortuosity is less significant issue with those collaterals. Patients who underwent prior cardiac surgery are at a much lower risk of complication in the case of an epicardial CC perforation, especially if the CC course is not basal, but rather run on the mid ventricular or apical portion of the left or right ventricles. Epicardial CC that are present in the atrio-ventricular (AV) groove of the heart are not trivial to cross, even in post-CABG patients; in case of perforation, the blood can accumulate posterior to the atria, leading to a contained hematoma and focal tamponade (Fig. 9.7). However, if a more apical collateral is perforated in a patient who underwent cardiac surgery in the past, such an event usually does not lead to tamponade, as the pericardium is adherent to the heart (Fig. 9.6, white arrow Video 9.11). In summary, epicardial channels can be considered good interventional routes if large, especially if not too tortuous, and if used in patients with prior cardiac surgery. An exception to this rule is the presence of AV groove collaterals that can lead to tamponade if perforated, both in patients with or without an intact pericardium. An additional point requiring special attention is the possibility that by occluding a dominant large epicardial collateral, the sole blood supply to the target vessel may be severely compromised and thus can lead to severe chest discomfort, hemodynamic and electrical instability. In such cases this collateral should be abandoned and alternative collateral channels should be explored. Only experienced operators should attempt crossing of tortuous and small epicardial collaterals, especially in patients with an intact pericardium.
Perforation of an epicardial collateral in the atrio-ventricular groove in a post-CABG patient leading to a contained hematoma. Panels (a) and (b). The arrow indicated the perforation with contrast spilling (type 3 perforation). Panel (c). Echocardiogram showing a contained hematoma being to the left atrum (arrow) (From Spratt  with permission)
A sub-category of non-septal collaterals are the intra-myocardial collaterals (Fig. 9.8, Video 9.12). They tend to be present more often when there is a connection between the RPL and the OM branches, of from the OM to a distal PDA. They sometimes have the angiographic appearance of a network, they are occasionally compressed during systole; they pose unique challenges as they are often difficult to wire because of acute angulations as they dive into the myocardium and exit back to the epicardial surface. Advancing microcatheters can be also challenging and applying more forward pressure should be avoided since they may be rended along with the near-by myocardium leading to catastrophic tamponade. Instead very meticulous spinning of the microcatheter should be applied. These collaterals are best used in patients with prior cardiac surgery. Extreme care should be applied in patients with an intact pericardium (Fig. 9.8, Video 9.12). If still difficulties are encountered, then these collaterals should be abandoned and a different approach should be undertaken.
(Video 9.12). Intra-myocardial collateral channel from the OM to the PDA. During systole, the collateral disappears
With all collaterals, it is also important to analyse the relationship between the connection to the distal vessel and the distal cap of the CTO. If the CC connects too close to the distal cap, entry into the occlusion can sometimes become very difficult or even impossible. Regarding this point, the best collaterals are the ones that connect at a distance from the distal cap to allow for the microcatheter to engage the CTO without a steep angle.
Typical collateral patterns are depicted in Figs. 9.9, 9.10 and 9.11. For RCA CTOs, the most common dominant collaterals are septals (Fig. 9.9). For LCx CTOs, a frequent pattern of collateralization is from the diagonal branches (Fig. 9.10). Finally, for LAD CTOs, septals from the PDA are most often providing the collateral flow (Fig. 9.11).
Surgical grafts can serve as effective routes to the distal CTO vessel. Arterial grafts can sometimes be used to deliver gear to the distal cap. This is the case, for example, of a severe ostial diagonal disease filled with retrograde flow from the left internal mammary artery (LIMA) to the LAD, in the presence of a proximal LAD CTO. This common situation, which often arises after the occlusion of a saphenous vein graft (SVG) to the diagonal, can be dealt with opening the LAD CTO towards the diagonal, restoring antegrade flow to the anterolateral wall. In such a case, the LIMA can serve as a route to the distal LAD CTO cap. Another situation is the presence of a RIMA to the PDA, an RCA CTO, and severe disease in a large PL branch. Treating the PL branch from the right internal mammary artery (RIMA) is sometimes impossible because of the presence of two acute angulations. However, a retrograde recanalization of the RCA followed by stenting towards the PL will be most effective and efficient. Finally, the LIMA can be used to access the LAD which provides septal collaterals to the PDA, in the presence of an occluded native LAD. When considering the use of an arterial graft as a retrograde donor vessel, the following principles should be considered: (1) caution should be applied in the presence of severe loops. After advancing a microcatheter to the distal vessel and prior to attempting the retrograde recanalization, the operator should assess flow into the arterial graft. If flow is impaired with the retrograde microcatheter in place into the LAD, the procedure should be aborted and an alternative retrograde route or an antegrade approach selected instead. (2) Caution should be applied to arterial grafts that serve as the last conduit of blood supply to the heart; a donor artery complication would be disastrous. Caution should be applied while manipulating the guide and equipment in the LIMA since ostial guide dissection can occur with any movement of the gear. 6F guides are recommended to minimize this risk. Moreover, the selective cannulation of the LIMA with a 6F or 5.5F MAC catheter can really help to stabilize the catheter in place, while reducing the risk of ostial dissection. The ostial LIMA guide should be maintained in the field of view whenever possible. In general, we would lean towards having a higher threshold in using the LIMA as a retrograde conduit, compared to other grafts, or alternative routes.
SVGs can also be used. They can either be patent or occluded. Patent SVGs can be used for the same reasons as the ones cited in the previous section. In addition, the presence of severe disease in a degenerated SVG can raise the option of opening the native instead of treating the SVG, since long-term SVG patency may be limited despite the use of several DES (Fig. 9.12). Moreover, stenting degenerated grafts can lead to acute complications, such as no-reflow, especially if distal protection cannot be used. Another situation is the presence of recurrent restenotic lesions in an SVG, which is a prelude to occlusion of the graft in a near future. It is therefore wise to consider opening the native artery instead of keeping treating the diseased SVG.
Treatment of a native artery CTO instead of a diseased graft. (a) Severely degenerated SVG to PDA with several stenoses (black arrows), and an RCA CTO (white arrows); (b) Post retrograde CTO PCI of the RCA via the SVG graft
SVGs can also be used even if occluded. This is best performed if the occlusion is recent, as crossing a thrombosed graft is relatively easy (Fig. 9.13). However, instead of treating the graft, the native artery CTO will be addressed with a retrograde approach. This can also be performed in SVGs that have been occluded for several weeks, even years. The ostial graft occlusion morphology should be taken into consideration since blunt/stumpless occlusions are harder to cross compared to tapered ones. We will discuss the techniques on how to perform these approaches in the next section.
Using an occluded graft as a route to the distal cap. (a) Sub-occluded severely degenerated old SVG to OM. (b) Access to the distal cap of the OM occlusion from the old graft. (c) Final result after rotational atherectomy and stenting
In summary, there are multiple ways to get to the distal cap (Table 9.1). If the algorithm favors the use of the retrograde approach, it should be attempted in the presence of collateral channels or grafts that are suitable for the approach. Even tiny or even invisible septal CC can be crossed with a surfing technique; therefore, in the case of an RCA or an LAD CTO with septal CCs, an attempt to cross the CC should be made before concluding that the collaterals are not suitable or a retrograde approach impossible.
Crossing the Collateral Channel or the Graft with a Wire and a Microcatheter
Crossing a Septal CC with the Wire
Initially, the technique involves the use of a microcatheter advanced on a workhorse wire. In the case of LAD to PDA CCs, the operator should target the collateral that is most likely to connect. Additional anatomic considerations when choosing the CC to cross include the tortuosity of the donor vessel, the angle of the take-off of the septal (a retroflex septal is obviously challenging to engage), and the point of insertion into the target vessel on relation to the distal cap, (insertion of the collateral very close, or at the distal cap is suboptimal). Usually, we will attempt to cross the largest and straightest one. However, other septal channels can be crossed also, even if there is no visible connection (CC0) (Fig. 9.14). If a large septal trunk is present, the Corsair catheter should be delivered on a workhorse wire shaped properly to engage into the proximal portion of the septal, followed by advancement of the catheter. The wire should be then be exchanged for a Sion wire (Asahi Intecc, Japan), with a small 30–45° bend at 1 mm from the tip. Alternatively, a Fielder FC wire can be used, although the Sion wire has clearly become the dominant wire for that task. This wire will be advanced in the CC in the search of a low resistance connection. We call this technique the surfing technique (Fig. 9.3, Videos 9.3 and 9.4). The wire is quickly advanced and pulled back when feeling a resistance, and redirected in a different channel. Once feeling some buckling, drilling of the wire can lead to passage into the recipient vessel. This technique offers the advantage to test several options of crossing from one proximal septal trunk. If the operator fails to connect from a given septal, the Corsair should be pulled, a different one is then engaged with the workhorse wire shaped appropriately for the proximal septal angulation. After advancement of the Corsair, the new septal branch with its CCs are “surfed” again with the Sion. It is common to see the wire entering into a ventricular cavity. The wire will then loop and will seem to float freely into a large cavity. It will then be simply pulled and re-advanced into a different direction. Entering a cavity is benign as long as the microcatheter is not advanced. A good septal track to the recipient vessel can be identified when a repetitive course of the wire is noticed with sequential pulling and pushing on the wire; if the wire keeps on making the same turns, it likely tracks a vessel structure that may finally connect. Therefore, it is advanced to the point of resistance, and drilled at this level, seeking a release into the recipient vessel. The feeling is often of a free wire that tracks nicely the PDA or LAD, and engage branches at the distal CTO cap. A wire that successfully crossed a septal will be free; with the heart beats, a too-an-fro movement of the wire can be seen and the wire often track side branches. Otherwise, the wire is probably not into the recipient vessel. A retrograde angiography should be performed to confirm that the distal wire tip is at the distal cap before advancing the Corsair through the CC.
Crossing invisible septal collateral channels. (a) Retrograde injection showing no clear evidence of connection from the septal branches, especially from the first septal. (b) Connection through an invisible septal channel into the PDA with the surfing technique. (c) Another case showing several CC1 connections from the LAD to the PDA. (d) Connection to the PDA through an invisible septal channel
If the septal surfing technique fails, it may be necessary to perform a tip injection of contrast from the Corsair. This technique involves pulling out the wire, aspirating back blood from the Corsair, and then injecting 2–3 cc of contrast using a small syringe, to assess the course of the collateral channel. Such a technique is useful especially in the case of large but tortuous CC2 collaterals that offers resistance to passage of the guidewire (Fig. 9.5, Videos 9.6, 9.7, 9.8, 9.9, and 9.10). A distal tip injection can help to identify bends in the CC that was not identified with non-selective injections. It is often very useful to use a straight left anterior oblique (LAO) projection to better assess the collateral with the tip injection. This second injection can be performed with only saline into the Corsair, flushing the remaining contrast out of the Corsair and providing information from another view before reinserting the wire. The tip injection technique carries the risk of septal dissection, or perforation, which will preclude subsequent use of this route. However, septal surfing can also lead to septal perforation. They are usually smaller and appear as small stains in the septum (Fig. 9.15). They are for the vast majority benign.
Failed attempt to connect retrograde from the LAD with some benign septal stains of contrast (arrow)
If the septal branches are tiny, it is often helpful to start surfing with a microcatheter that has a smaller tip. The Finecross (Terumo, Japan) can be very helpful to follow a workhorse wire that was advanced into a small septal CC at its origin; then, the wire can be exchanged for the Sion and septal surfing performed (Fig. 9.3, Videos 9.3 and 9.4). If the wire connects into the recipient vessel, an attempt to cross the collateral with the Finecross can be done. The Finecross can be manipulated with the same rotations as for the Corsair. If it crosses, this catheter will be adequate for less complex CTOs that are likely to be traversed with a retrograde true-to-true (TTT) crossing strategy. However, if a long CTO is attempted, switching for a Corsair will tremendously improve the support from the retrograde side for subsequent wire manoeuvers, especially involving knuckled wires. This is because of the screwing pattern of the Corsair that stabilize the catheter in its position, providing additional back-up support that will be lacking with the Finecross. Exchanging the Finecross for a Corsair or a Turnpike (Vascular Solutions, US) can be done with a trapping technique, even in a 6F guide, using a 2.0 mm balloon. The trapping technique is explained in Chap. 4.
When performing septal surfing in the case of a PDA providing septal collaterals to the LAD, the same principles apply. One major nuance if that the septal CCs are often very tiny or even invisible close to the PDA, as branching increases towards the inferior septum. However, if the wire successfully engages a connecting collateral, its advancement becomes easier as it moves towards the LAD, with the wire navigating progressively into larger vessels. It is therefore unusual to be able to deliver a Corsair in a septal branch to start surfing (Fig. 9.16, Video 9.13). The best technique is the leave the Corsair in the PDA, and to start surfing from the PDA, leaving the wire tip as free as possible as it engages several little septal branches. If one offers minimal resistance, the wire can be pushed further and directed into the LAD. If a larger CC is present and seems to connect, it should be first targeted. Once the wire advances freely, engages branches in the recipient vessel, and when a to-and-fro movement of the wire is visualized, it is safe to follow with the microcatheter. A retrograde injection should however be performed to confirm the location before advancing the microcatheter in case of doubt.
(Video 9.8). Septal surfing from the PDA to the LAD. The Corsair is positioned into the PDA, and the wire manipulated to engage septals and complete the course up to the LAD
Advancing the Microcatheter to the Distal Cap Through the Septal CC
If a FineCross was used to surf the septal CCs, crossing of the collateral can be attempted with this catheter. However, the FineCross is not a septal dilator, and sometimes will not be as efficient in crossing a collateral. Alternative clockwise and counter clockwise rotations can be applied to help deliver the catheter to the distal cap. However, the FineCross provides much less support for retrograde crossing of the CTO. It should therefore be reserved for simpler and shorter CTOs where a true-to-true crossing is likely to be successful. In all the other situations, the Corsair is the preferred catheter. If a FineCross was first used, it can be exchanged for a Corsair using the trapping balloon technique as explained in Chap. 4. As described in Chap. 3, it is engineered as a septal dilator. It will be advanced with alternative clockwise and counter-clockwise rotations to the distal cap. It will often stop advancing at the inferior portion of the septum, where the septal CC is smaller and angulated. If is often necessary to apply 5–10 turns in one direction followed by a change of direction to help the catheter advance. If that step fails, the Corsair can be withdrawn and a 1.25 or 1.5 balloon used to dilate the septal at the point of resistance (Fig. 9.17, Videos 9.14, 9.15, 9.16, 9.17, 9.18, 9.19, 9.20, and 9.21). Then, crossing with the Corsair is attempted again. Occasionally using the 135 cm version of the Corsair to dilate a resistant part in the septum is more effective, followed by the exchanged for a long Corsair to complete the procedure. Alternatively, especially if already on the table, the reattempt can be done with the FineCross. Lack of support is also one of the common cause of absence of progression of the Corsair into the septal. It is possibly more frequent with 6F than with 8F guide catheters, although it can be encountered with the latter. The Corsair can be withdrawn, and a MAC technique used to increase the support. We like to use the 5.5F version of the GuideLiner that offers adequate inner lumen size to accommodate the Corsair, but also has a smaller outer diameter when delivered deep into the donor vessel, thus reducing the risk of donor artery dissection (Fig. 9.17d, Video 9.17). The GuideLiner can sometimes be pushed to the level of the septal and the Corsair re-advanced. However, once the Corsair is successfully delivered to the distal cap, the GuideLiner should be retrieved into the guide to minimize donor artery trauma. Alternatively, an anchoring balloon technique (see Chap. 12) can be used to improve support. Finally, the Corsair can sometimes dry through the septal, or even get “fatigued” after several clockwise and counter-clockwise rotations. The Corsair should be left alone for a few seconds, and movements re-applied after this short rest. Otherwise, it is wise to simply take it out, flush it and rewet the outer surface of the hydrophilic coating of the catheter as well as the wire. If this steps fails to help crossing the collateral with the Corsair, use of a new Corsair (or a TurnPike, Vascular Solutions, US) can be necessary. Flushing the lumen of the Corsair with Rotaglide (Boston Scientific, US) can help reduce the friction and minimize the occurrence of Corsair fatigue. This phenomenon should be recognized early on during the procedure to allow for a timely change for a new Corsair before the wire gets entrapped into the Corsair, in which case the whole system has to be removed as a unit. If all those steps failed, which is very unusual, the retrograde wire should be left in place and an antegrade approach attempted, using the retrograde wire as a target (kissing wire technique).
(Videos 9.14, 9.15, 9.16, 9.17, 9.18, 9.19, 9.20, and 9.21). Aorto-ostial RCA CTO treated with the retrograde approach. (a) Flush aorto-ostial RCA CTO, with septal collaterals, mandating a retrograde approach (Video 9.14). (b) Small septal channels connecting to PDA and PLV (Video 9.15). (c) One small invisible septal was crossed into a PLV with a surfing technique (Finecross and Sion wire), through a stent that jailed the first septal (Video 9.16). (d) After a failed attempt to deliver a Corsair through the septal CC (after dilating the stent struts at the ostium of the septal with a 1.5 mm balloon), a septal balloon dilation is performed at the point where the Corsair failed to progress with a 1.25 balloon at 12 atm (black arrow), with the support of a 5.5 F GuideLiner into the LAD (white arrow) (Video 9.17). (e) Corsair successfully delivered to the distal CTO cap (Video 9.18). (f) A true-to-true CTO crossing into the aorta was done (Video 9.19). (g) After snaring a long retrograde wire, balloon dilation and stenting with DES on the externalized guidewire (Video 9.21). (h) Final result (Video 9.21)
Crossing an Epicardial CC with the Wire and the Microcatheter
Surfing has no role in wiring the epicardial collaterals. Wiring epicardial collaterals is a very controlled and meticulous technique that relies on detailed definition of their course using frequent tip injections in different angiographic projections, as single projections frequently are inadequate to delineate their complex three-dimensional course. Furthermore, epicardial collaterals can change configuration as wires and microcatheters are introduced and advanced, a factor that should be taken into consideration while performing this most technically demanding maneuver.