[[100 Welcome to Outlines of Technique in Cardiac Surgery]] # A Coronary Anastomosis Primer ## Foreword The indications for and complications of surgical procedures are plentifully discussed in textbooks and journal articles, and details of procedural sequence may occasionally be mentioned in the written record. The surgical atlas will illustrate the exposure, show where the organs are excised and reconnected, but it is rare to find detailed descriptions of the exact methods of excision and reconnection, and almost unknown to read instruction as to technique and the handling of surgical instruments. It is assumed that this minutiae will be imparted personally, in the operating room, but the student of surgery will consult his or her experience and quickly realize that very little was ever formally explained. A disorganized mimicry, of good and bad habit alike, interspersed with random admonishments to hold one instrument this way or another are all the education that is usually encountered. The fortunate surgeon finds a skilled mentor to mimic, and if observant and possessed of an instinctive knack copies these methods without ever being aware or being made aware of the reasons behind their evolution. One need consider nothing more common than the widespread ignorance of how a right-handed surgeon should hold scissors to cut with the left hand to realize how little thought goes into technical education. For that matter, the infinite and barbaric variety of table manners when cutlery is handled at meals is all that is needed to triply underscore this point. But while a piece of meat may be pinned with one clenched fist and inelegantly sawn at with the other, and eventually be scooped into the maw of the unwashed in a variety of ways with little consequence except the disgust of a neighbor, one hopes that there is one superior, optimal way to skin a cat. How will we recognize it? Compared to others, the true way should be easy, quick, and efficient to perform. It should be the most respectful of tissue, and guarantee the greatest uniformity of excellent result. It should be easy to teach, and should be within the reach of the common man, erasing as much as possible the natural difference between the ordinary and truly gifted. Because it is all these things, it should have a reason why each motion is performed in a particular way and a particular order, and these reasons should fit together in an over-arching philosophy with the simplest system of principles. ### Exposure #### How wide the sternotomy Particularly when the heart is hypertrophied, exposing obtuse marginals close to the atrioventricular groove can be quite difficult. Some patients will have deep chests that accentuate this challenge. The heart holder will be crucial in this scenario. Another useful trick is to extend the midline incision along the linea alba, even up to midway between the xyphoid and the umbilicus. You will be able to spread the sternotomy very wide, and get maximum exposure. Another seldom utilized trick is to open up the right pleura, and drop the pericardial stays along the right side to allow the heart to rotate towards the left. Retracting the hypertrophied heart to the right to expose OM's can decrease venous drainage, and worsen collateral flow back bleeding through the arteriotomy, interfering with visualization. To improve exposure of the coronaries, particularly of left sided targets and the PD, the heart can be elevated from the pericardial well by tucking in at least one moist laparotomy sponge behind the left AV groove. This lap sponge usually only interferes with RCA exposure. #### The PLOM/LVbr One lap tucked behind the left AV groove, another moist lap pulls the apex of the heart to the right, and cephalad. #### The ramus One lap tucked behind the left AV groove, another is used to gently but firmly grasp the apex of the heart in an anterior-posterior fashion between the thumb (posterior) and the other fingers (anterior). The assistant uses his left hand to perform this maneuver, retracting the apex slightly towards the patient's right, while simultaneously gently twisting the heart counterclockwise. This rotation can make the ramus exposure as satisfying as that of the diagonal. #### The diagonal Usually the easiest target to expose, requiring only the moist lap sponge behind the LV AV groove. #### The LAD The LAD does not even require the whole LV AV groove lap sponge: instead of being scrunched up, only one fold is needed. #### The PD One LV AV groove lap sponge behind the LV, and the assistant uses another cold lap to retract the acute margin cephalad with his right hand. #### The RCA NO LV AV groove lap sponge behind the LV. The assistant uses a cold lap to retract the acute margin cephalad and slightly to the left. #### The heart holder, the heart jock and the hercules The heart is held in position by your assistant. If manpower is plentiful, the least skilled person can hold the heart, while the more skilled assistant provides retraction on the vein, or on the epicardium, to open up the target arteriotomy. Vein retraction is accomplished by pulling on adventitial scruff, at least 1 cm away from the anastomosis so as to prevent interference with the surgeon's instruments. In general, exposure is intermittent: retract, and when the surgeon's needle has entered the tissue, let go or relax the counter-traction. The skilled assistant also mans the sucker, which again, is used intermittently, and only just before the needle passes through the lip of the target arteriotomy. The assistant holding the heart uses a cold lap to aid in the traction: be wary of the friable epicardium that shears off the heart with this traction. This shear leaves a worrisome blood blister at best, and a bleeding raw surface at worst. An assistant pulling the heart towards his own side of the table uses the surfaces of fingers 2-5, often spreading them apart to help open the target arteriotomy. The assistant who pushes the heart away from his own side should use the thumb if possible, as this is usually more comfortable. The heart jock is an alternative method of exposure that requires no human assistant. The 'heart jock' is a sort of cargo netting made of cotton tapes. The tapes at the north and south ends of the net are substantially longer. The two ends of these long tapes on the patients right are passed, one through the sinus behind the aorta, and one through the completed space behind the IVC and caudad to the right inferior pulmonary vein, and therefore the net is layed out behind the heart. The left end of the long tapes are pulled up and over the heart, and secured to the drapes or the retractor (commonly an Ankenny retractor) on the patient's right, thereby lifting the heart. If a tape obscures the exposure of the target vessel, it is cut out. After the anastomoses to the targets are done, the conduit is passed through its hole in the net before the proximal is done. The hercules retractor is an articulated stabilizer arm that is attached to the sternal retractor, typically at the lower right hand corner of the toothed arm, closest to the winder. Wherever a human assistant would have applied a single thumb to provide appropriate exposure, pad with a folded lap or ray-tech, and push the hercules retractor aortic valve assistant “hand” against it. Arrange the links of the arm so that it is as out of the way as possible, and tighten the stabilizer arm handle to freeze it into position. With a good parsonnet retractor, no other assistant may be necessary. ![[Hercules3-Stabilizing-Arm3.png]] #### The Parsonnet retractor The Parsonnet self retaining retractor, also known as the eyelash retractor, is a useful adjunct to exposure of the target arteriotomy. It comes in a variety of sizes, only the smallest of which is typically practical. The 'eyelash'spikes on either side of this spring retractor are gently dug in to the epicardium on either side of the target arteriotomy, and upon release, the retractor opens the target up. If deployed too close to the arteriotomy, the retractor will interfere with suture placement, and will overly distort the arteriotomy, flattening it out in a way that promotes backwalling the anastomosis. The retractor is positioned to individual taste, but its open end is usually directed towards the apex on lateral and posterolateral targets, and towards the base for LAD targets. The retractor is removed after the anastomosis is tested for hemostasis, and it should be handed back and announced to the scrub nurse, as it is too easy to lose the device in the chest. ![[Parsonett Retractor.jpg]] #### The blower, the sucker Collateral bleeding through the target arteriotomy may obscure the anastomosis. Two schools of thought are applied to clearing this blood. The easiest to use is suction, through wall or cell-saver, using the metal Frazier suction tip. Do not occlude the side-hole of the frazier tip: the suction generated thereby is usually excessive, drawing the edges of the arteriotomy and conduit together, and obscuring the view. The vaccum source may be set too high, producing this effect regardless of occlusion of the side hole. The alternative is to use a blower, a wand that emits a focused jet of moisturized mist that blows away the collateralized blood and often flares out the tip of the conduit in a helpful fashion. This type of device gained popularity with the advent of off-pump bypass surgery. Initial versions had insufficiently moisturized air and were shown to desiccate and damage the endothelium of conduit and target. Even when appropriately moist, there is a tendency in diseased target vessels for the jet of mist to separate adventitial and intimal planes and unintentionally endarterectomize the target. The jet should be turned down to the lowest possible level, and only used when the needle is going through the target vessel wall. #### The coronary shunt/occluder Another option when collateral bleeding precludes optimal visualization is the use of a coronary shunt or occluder, another device popularized by the bloody fields typical of off-pump surgery. Upon arteriotomy, the target vessel is assessed with a coronary probe, allowing measurement of internal diameter, ensuring the absence of a distal stenosis, and confirming that arteriotomy has been made in the correct, previously-planned, segment of the target. A coronary shunt, a short, flexible, wire-bound tube with flared ends that correspond to the vessel internal diameter is selected according to the probe measurement. It is inserted, one end at a time, through the arteriotomy, bending like an inchworm before settling within the vessel, and preventing egress of collateral blood into the arteriotomy. Properly selected, the shunt can even be helpful with anastomosis by helping prevent back-walling at the heel and toe. Shunts will typically have a traction thread attached to their middle, attached at the other end to a light plastic torus. Gentle traction on this string away from the direction in which the needle is pointed on the target vessel can help in exposure. Excessive traction will dislodge one or both ends of the shunt. The traction string and its torus is flipped one way then the other as the anastomosis is completed. The last few bites of the anastomosis are left loose so that shunt can be pulled out of the vessel between the loops of suture. ![[Coronary Shunt.jpg]] ### Arteriotomy Once the target vessel has been identified and optimally exposed, the surgeon and assistant grab the epicardium on either side of the area with coronary debakeys and lift away from the heart. To prevent a backwall injury, it may be helpful to create the arteriotomy while the vessel is distended with antegrade or retrograde cardioplegia, and after arteriotomy the cardioplegia is removed with root suction. Some may choose to denude the target vessel of epicardium on either side, to ensure midline arteriotomy, but this can involve a bloodier field, and deprive the anastomosis of some of its strongest tissue layers. Exposure of the vessel and arteriotomy can be accomplished with a specialized instrument, like a Wechsler, or "beaver" blade, or by an inexpensive #15. In either case, the vessel wall is stroked, in exactly the same spot with the 'belly' of the blade, and an arteriotomy created just large enough to admit the tip of the coronary scissors. Sometimes, the character or geometry or depth of the target vessel wall may make an #11 blade a better choice. The #11 is gently inserted for an extremely short distance at the shallowest angle of attack possible into the arterial target, with the edge of the blade away from the heart, and then is lifted up. The arteriotomy is extended just large enough to accomodate a coronary probe using forward or reverse (also known as circumflex) scissors. The probe will confirm absence of distal unexpected stenosis, and provide information as to the character of the vessel wall that may influence the direction in which the arteriotomy is extended. The size of the arteriotomy takes into consideration the size of the target, but is generally dictated by the size of the conduit. The conduit should be beveled no less than 45 degrees, and can be extended by a notch in the direction of the bevel that should be no greater than the length of the beveled cut, or the anastomosis will tend to have a "waist" in the conduit at the heel. The total opening of the conduit should be no less than 1.25 times the size of the target arteriotomy so that the anastomosis will have a pleasant and hemodynamically effective, rounded, 'cobra' head. The side-to-side, sequential anastomosis, which perforce is made small to avoid "seagulling" of the graft (see below), when manually flushed, is a reminder that small anastomoses can still provide excellent flow. #### Fixing the back wall Should the backwall be injured, typically during arteriotomy, to the extent that there is concern for dissection or of posterior bleeding, the arteriotomy is extended in either direction so that the injury is a the middle of the anastomosis, such that the patch angioplasty effect of anastomosis is at its greatest. The vessel is freed from surrounding tissues sufficiently to allow a stitch to be passed outside-in, on the far side of the injury, and then inside-out, emerging from the same side of the vessel, and is tied down. The needle and suture should be the finest possible to accomplish this, and multiple interrupted stitches should be used if necessary. ![[Coronary Back Wall Injury Repair .png]] ### THE ANASTOMOSIS Sewing one vessel to another is one of the most spiritually exalting, life-bestowing activities a human can perform. If simple commandments are not followed, however, it can be one of the most vexatious and injurious pursuits one person can inflict on another. "The most important 15 minutes between birth and death are when the left internal mammary is sewn to the left anterior descending". If the following rules are incorporated into the anastomosis, it will make the easy anastomosis swift, and the difficult anastomosis possible. #### The Rules ##### Two bites make up a stitch A rule regarding nomenclature: passing the needle through the wall of either the conduit or the target is a 'bite'. The combination of a bite through conduit and a bite through the corresponding vessel wall is called a stitch. By convention, imagining that the beveled graft looks like a leg with a shoe at the end of it, the furthest extent of the anastomosis, where the angle subtended by the anterior edge of the graft and the surface of the epicardium is obtuse, is known as the 'toe'. The opposite extent, where the angle between graft and epicardium is acute, is known as the 'heel'. The heel and the toe are not always oriented in the same direction as the target vessel, and are often at 90 degrees to it, in which case we refer to the 'anatomic' heel and toe (proximal and distal extents of the target arteriotomy), and the 'functional' heel and toe. ##### If you can draw it, you can sew it. The anastomosis exists first in your mind, and it should be the rare exception, such as when dictated by calcification of the vessel, where the stitches do not have the same ideal symmetry, spacing, and depth, as would be the case if the stitches were drawn on a sketch of an anastomosis. ##### Don't let the needle control you. The undisciplined or inexperienced surgeon inserts a needle into tissue, and eagerly awaits its unexpected exit somewhere removed from the entry. The disciplined surgeon lines his needle up like a billiard shot, and even though the cue is curved, knows exactly when the needle is driven in a specific spot, with a specific pitch, where the needle will emerge, and it therefore emerges exactly where desired. ##### The most difficult part of the anastomosis needs the best exposure The principle of exposure is one of the pillars that separates the novice from the adept. The two major techniques of coronary anastomosis, the tie-down and the parachute are designed, each in their own way, to provide maximum exposure to the most critical and difficult parts of the anastomosis: the portion that does not face you, and will not be stitched with forehand bites that are sewn towards you, and the heel of the anastomosis, particularly when target and/or conduit are small. ##### Do the most difficult part first A corollary to the previous rule. Once again, the difficult parts of the anastomosis; the portion that does not face you, and will not be stitched with forehand bites that are sewn towards you, and the heel of the anastomosis, particularly when target and/or conduit are small. Whenever possible, these are done first, when the unconstrained conduit can be freely positioned to allow optimal exposure. ##### See every bite of every stitch One should live in horror of the stitch that is not full thickness, or even worse, that accidentally incorporates the back wall, narrowing or closing the artery. For these to be avoided, exposure should allow visualization of each bite as it passes through target or conduit wall. This does not preclude taking both walls in the same 'bite', but this maneuver requires using the needle tip to 'pick up' the conduit vessel wall, then go through the wall for a short distance, then insert the needle tip back into the target arteriotomy to take the target wall, all without regrabbing the needle with the needle driver. This is obviously best practiced at the sides of the anastomosis, where visualization is easiest. With extremely rare exception, it should not be done at the heel or the toe. These should be 'taken in two'. ##### Don't pick up the needle by the tip Doing so will blunt the tip of the needle, or worse, turn it into a sort of crochet hook that snags every vessel wall where you don't want it to. For reasons best shrouded to preserve anonymity, this was known in my training as 'murphing' the needle. ##### Set up the anastomosis in such a way that most of it is sewing forehand towards yourself The easiest stitching is done sewing forehand towards oneself. While it is an obligation to achieve smoothness with backhand or forehand in either direction and with either hand, one still achieves the greatest speed and comfort when sewing forehand towards oneself. ##### Symmetry is hemostatic, or Pretty don’t bleed It is possible to sew a perfectly hemostatic proximal or distal anastomosis with six perfectly placed stitches. Regardless of the number of stitches, symmetry is the hallmark of hemostasis. Depth and spacing of the stitches should match up perfectly across the axis of the anastomosis. ##### Space more closely at toe and heel, more widely at sides Depth and spacing of the first two stitches on either side of the stitch exactly at the heel of toe of the anastomosis should be finer and closer in spacing than the stitches at the side. Too deep a stitch gathers in tissue at the heel and toe, and potentially narrows the target vessel, reducing the outflow and graft patency. Spacing these more closely is an attempt to prevent the need for 'fixit' stitches in these sensitive areas. Wider quicker bites at the sides that do not provide perfect hemostasis can be repaired with fixits where the patch angioplasty effect of the anastomosis makes it unlikely that the vessel will be narrowed. Fixits at the heel or toe should optimally be adventitial-epicardial, unless the leak is quite impressive. The stitch at the very heel and toe can be of greater depth than the stitches on either side, if greater strength is desired, or tissues are weak. Within reason, this does not typically narrow the target vessel as it only 'foreshortens' the vessel along its axis. ##### The "nailing schedule" of an anastomosis The preceding principle informs the concept of the "nailing schedule" of an anastomosis. In general, for the vast majority of appropriately sized anastomoses, there is a name, and appropriate spacing for each kind of symmetric stitch. These are: ![[Diagram of the Stitch Schedule.png]] - ###### The Wing, or the Half: - Halfway on each side from Toe to Heel. Called the Wing because it often coincides with the halfway point on the conduit where the beveled cut has been doubled by cutting towards the Heel to lengthen the anastomosis. This stitch can be deep in its bites into conduit and target as it is very difficult to narrow the anastomosis here. - ###### The Quarter, also known as Halfway to the Half: - Halfway from heel or toe to the Wing. Less deep in the bites than the Wing. - ###### The One-off: - The first stitch on either side of the Heel or the Toe. A third of the way from the Heel or Toe to the corresponding Quarter. These bites must be delicate on the target at least to avoid narrowing the anastomosis. - ###### The Toe: - Danger of backwalling is at a maximum. Delicate bite on conduit, bite on target can be deeper if desired since this bite usually doesn't tend to narrow the anastomosis. - ###### The Heel: - Danger of backwalling is at a maximum. Bite on target can be deeper if desired since this bite usually doesn't tend to narrow the anastomosis. The bite on the conduit can still be delicate, but it usually isn't as crucial as at the Toe. - ###### The Unnecessary Stitch: - The last bite, usually between the Quarter and the Wing. It really shouldn't need the extra bite, but it is usually easy to perform, you've got the needle right in your hands, and it is very annoying when you don't take it, and when you perfuse the graft, it leaks at precisely that point. The graft is not always lined up with the target. An orientation of a graft at 90 degrees (occasionally other angles) to the target is often required to set up a sequential, avoid an appendage, for a better graft lie, or to conserve conduit length. When this is the case, the target Heel and Toe are said to be the anatomic Heel and Toe, and the graft Heel and Toe are said to be the functional Heel and Toe. Delicate One-offs will still be necessary at the anatomic Heel and Toe. One-offs may or may not be necessary at the functional Heel and Toe, but they need not be delicate. ![[Conduit and Target at Ninety Degrees.png]] The small graft, even beveled, may not permit all these stitches. It will be wise to make a correspondingly small target arteriotomy. The same is the case for end-to-side sequential anastomoses, particularly when the graft is oriented at 90 degrees to the target. In these anastomoses, only the Toe, Heel, Wings and Quarters are usually necessary, all of which must be symmetric and delicate. This type of anastomosis is more prone to "purse-string" narrowing, the prevention of which is discussed below. ##### Handle the conduit and target as you would want your own urethra handled No part of the conduit, excepting those portions that are to be trimmed away, or will lie outside the suture line of the anastomosis, and therefore will not contact blood, should ever be crushed or handled with forceps. The endothelium is very sensitive, and its long term responses to injury will do your anastomosis no good. The conduit should be handled by the adventitia. Expediency may tempt you to grasp the edge rather than the adventitia with your forceps to facilitate passage of the needle through the wall nearby. If this temptation cannot be resisted, only the tiniest edge should be grasped and the suture line should exclude this grasped edge. Optimally, the conduit need not be handled to assist passage of the needle. If the needle is oriented so it enters the tissue at ninety degrees, the tissue will not move. The needle tip passes through the tissue, and immediately, the fingers in the needle driver are rolled to allow the curve of the needle to go away from the heart, away from tissues that interfere with the anastomosis, so that it can be re-oriented to enter the target artery and confront the endothelium at 90 degrees, or to be regrasped by forceps or needle driver (away from the tip, of course). If this 90 degree transfixion technique does not work, you probably murphed the needle. ##### Let the curve of the needle keep you away from trouble Following the curve of the needle is crucial to the atraumatic creation of a hemostatic anastomosis. One marvels that Carrel and Guthrie perfected their techniques with straight needles and unswedged suture. The needle is passed through tissue at 90 degrees to it, not only for the precision of suture placement described above, but so it will gather the least tissue, and so that when it emerges from tissue, it will head away from the heart and away from other structures one does not wish to encounter. Barbarous and brutal neophytes will accomplish this elevation of the tip of the needle, not by rolling their fingers and following the curve, but by CRUELLY LIFTING THEIR ENTIRE HAND, THE NEEDLE, AND THE TRANSFIXED TISSUES. This senseless treatment of the arterial wall can rip the needle through it, requiring major alterations and endangering patency and hemostasis. Even when the arterial wall is not entirely ripped, there will be a slit in the wall that increases needle hole bleeding as the tip of an injury iceberg. The experienced heart surgeon, condemned to hell for unrepented pride and lust, will be forced by Beelzebub to watch an endless blooper reel of poor needle handling. ##### Lay out your fucking suture, away from the direction you are sewing The profanity provides necessary emphasis to this rule. The surgeon who has not learned to lay out the suture away from the direction of the suture line construction will tangle the suture, unintentionally lock it, promote the formation of knots that will be dragged through delicate vessel walls, increasing the time of anastomosis and consequently ischemia. An individual that has not learned suture handling when closing hernias and incisions has no right to join bile duct to bowel, let alone join the company of those who sew blood vessels together, and should confine their attempts to serve mankind to the removal of skin tags, debridement of decubiti, and toe amputation. ##### Orient the needle so that when you roll your fingers/wrist, the needle follows the spoke, goes in and exits at 90 degrees to the tissue When the conduit has been brought to the target by either parachute or tie-down technique, the last few rules can be consolidated as follows. The needle emerges from the last bite, is grasped by the forceps away from the tip the microsecond after it is released from the driver. Following the curve of the needle, it is pulled through the tissues, directed away from the heart, and is lifted in the air in such a way that conduit and target are approximated, but not enough to limit exposure of both walls for the next bite. This takes the slack out of the suture so it will also not interfere with the next bite. The needle is held up in the air with the forceps, away from towels and sponges, and the hand with the needle driver is oriented in whatever fashion necessary to perfectly align the needle for the next bite, without stammer. Note: the needle is NOT repositioned within the needle driver by opening it a little and pushing with the forceps, or some other inanimate object like a towel. This positioning can never hope to be as exact as can be accomplished using both of ones wonderfully articulated hands, and will only lead to further stammer and loss of time. The length of suture is then layed out away from the direction in which the stitch is traveling by pushing the forceps against it. The needle is than brought to conduit to enter at ninety degrees, aligned like a spoke on a wheel so that it will enter target artery at 90 degrees, and emerge from target adventitia, all with only a slightly altered roll of ones fingers around the needle drivers. This rolling motion of the fingers, and slightly of the wrist, can be applied to any needle driver. The Castroviejo needle driver designed specifically for the distal anastomosis, where precision is valued over strength, is a half-round on each side to allow for this roll, but even the flat handled Castroviejo made for the proximal can be rolled in this way. For that matter, an ordinary standard needle driver can be rolled, though with slightly more wrist rotation, and the roll is assisted by the added control surfaces of the finger tips inserted INTO THE FINGER HOLES. The only reason to palm the grip of ordinary needle drivers is when strength is valued over precision, and wrist and forearm rotation predominate, as when driving wires through the sternum. Any other approach has abandoned reason in favor of dogma, or worse, showmanship. ##### Always go inside out on target artery. This is a good rule of thumb for any vascular anastomosis, whether proximal or distal. The target is being bypassed because it is diseased, and its different layers will be disparately affected. It is too easy when passing the needle outside to in on a diseased vessel to take a skiving bite, that may not be full thickness, bouncing off the atherosclerotic intima. Such a bite sets up the target for a dissection that will not be detected until its catastrophic effects reveal themselves in the ICU. Moreover, at the toes and heels, an outside-in course would greatly risk a backwall bite. ##### Sew with your fingers more than your wrists, your wrists more than your elbows, your elbows more than your shoulders, your waist not at all. The efficient, smooth surgeon selects the body position that best suits the intended forehand or backhand bite, braces forearms or wrists on the sternal retractor to prevent tremor, and executes the motions described above by motion of fingers and wrists alone. It is possible to move crudely, with greater motion of the arms and shoulders when the vessels are large and easily exposed, but the smaller the vessel, and the deeper it lies within the chest, the less room there will be for this kind of sloppy technique. The surgeon that learns to use dexterous and precise movements when the anastomosis is easy, will have the advantage when the anastomosis is hard. ##### Shift your body position to allow the easiest backhand or forehand. Imagine an LAD target, exposed with a laparotomy sponge so that it is perfectly oriented along the cephalad/caudad axis, and superimpose a polar coordinate system over it, such that the feet are in the direction of zero degrees, and the head is at 180. Clearly, bites passed through target walls at the portion of the anastomosis at 90 degrees will be backhand, and those at 270 degrees will be forehand. As one sews approaching the vicinity of 0 and 180, the stitch can be transitioned to forehand or backhand as a matter of individual comfort. Between 165 and 180 degrees, the surgeon can continue taking backhand stitches, standing facing this part of the anastomosis, the surgeon’s body at 300 degrees, or the surgeon can stand at 200 degrees and take the stitches forehand. Likewise, when taking target bites in the vicinity of 345 degrees to 360 degrees, the surgeon can stand at the same compass point, and continue to take forehand stitches, or can shift to stand at 200 degrees and take backhand stitches. ##### Stop sewing with one arm before you start limiting the exposure of the other arm With both tie-down and parachute techniques, there will come a point during an anastomosis where it will be tempting to continue to sew with one arm of the suture. This will usually happen when a portion of the anastomosis when one has transitioned from an uncomfortable backhand to a comfortable forehand, or vice versa, but it will be time to set this arm aside, and start sewing with the other, or else the partially completed anastomosis will prevent mobilization of the remainder’s vessel walls, obscuring visualization of the bites that will be taken with the other suture arm. One example is the SVG to diagonal tie down anastomosis that will be described in greater detail momentarily. Let us denote end of the target arteriotomy closest to the base of the heart as six o’clock, and the opposite end as 12 o’clock. The anastomosis will be tied down at 12 o’clock, and continue with one arm counterclockwise until just past 6:30 to 7 o’clock, then the other arm will be used to go from 12 clockwise to meet at 7 o’clock. To take the first arm past 7 is to limit the exposure of the portion that remains. On a proximal going to a right sided target, where the anterior extent of the aortotomy is 12, and the most posterior is 6, the anastomosis is performed by a right handed surgeon with tie-down at six, then one arm clockwise up to around 10:30 to eleven, then the other arm from six counterclockwise to meet at eleven. If the first arm were continued to 12 or past, it would interfere with the bites thrown at 5:30 with the other arm. ##### Avoid the back wall Traction on the sides of the target arteriotomy opens the vessel, and allows the easiest visualization of the anterior and posterior walls of the artery. At the toe and heel, the view is necessarily more difficult, and traction can often flatten the vessel, drawing the anterior and posterior walls together, increasing the risk of inadvertently allowing the needle to gather the back wall into the bite. A backwall bite at the very heel and toe, or even on the bites on either side, would seriously limit outflow of the graft. A virtually fool-proof way of avoiding the backwall is to insert the needle into the target, parallel to its course, equidistant from anterior and posterior walls, a little bit farther than is required for the planned depth of the bite. The fingers are then rolled slightly, to allow the curved needle tip to gently touch the anterior wall of the target artery. This pushes the belly of the curved needle against the posterior wall. The needle is then drawn back very slightly, the tip barely grazing the anterior wall, until the tip is at the desired depth of the planned bite. The fingers are then rotated further to drive the needle through the tissue of the anterior wall. Had any posterior wall tissue been grabbed on the way in, it would have slipped off the tip of the needle with the backwards motion. The same technique can be used for the bites on either side of the heel and toe, except that after the backward stroke, the needle is tilted slightly to the desired side before advancing. ![[How to Avoid the Back Wall.png]] ##### The worse the artery, the bigger the bite Somewhat paradoxically, the artery that is hardened is also weakened. The vessel that is calcified may resist penetration with the needle, and yet will be absurdly prone to rip through if excessive torque is applied to the needle, or too much tension is applied to the suture. Likewise, the asymmetrically calcified vessel will have a thick, rock-hard posterior wall, while the anterior wall is paper thin and also susceptible to ripping. Such a rip, at best is the source of nuisance needle hole bleeding, at worst, it will loosen a running stitch, and require closely spaced bites on either side of the rip, through equally delicate tissue. Whenever possible, such targets are to be avoided, but sometimes we have no choice. Delicate tissue handling, following the curve of the needle is necessary, as well as gently probing the artery wall with the needle to find a soft spot, and in general, taking a larger deeper bite of the target vessel wall. ##### The fewer the stitches, the greater the tendency to purse-string Economics tells us there are no solutions, only trade-offs. An anastomosis with a smaller number of bites is established more quickly, and with less ischemic time than one that is more closely stitched. As discussed above, there is a trade-off in that more fix-it stitches may be required. Another trade-off is that traction on the suture, or tightening with a nerve hook, is more evenly and easily distributed along an anastomosis with a smaller number of bites. As these anastomoses are typically also smaller, there is a danger of constricting the anastomosis by gathering the edges of the conduit and target artery in the ‘purse-string’ effect. When a six to eight bite anastomosis has been fashioned, the tendency to purse-string can be fought by lightly taking up the slack from either arm with a nerve hook, then tying the arms of the suture while the assistant flushes the conduit, or releases the bulldog clamp that controls flow to the anastomosis. The knot is tied down with only enough tension to stop bleeding from the anastomosis and no more. #### Anastomotic technique Two techniques exist for vascular anastomosis, each with its own advantages and disadvantages. The adept will utilize both where they are best suited. ##### Parachute The parachute technique nearly, but not completely, approximates the graft and target, and allows placement of sutures in each vessel with maximum degrees of freedom for each so that each bite is easily seen. The first few bites are traditionally placed where exposure is most difficult, and where backwalling the conduit would be most catastrophic. I don't understand people who use it anywhere else, or who don't take advantage of the technique to completely circumnavigate the back half of the anastomosis. When the backl half is completed, the conduit and target are connected by a continuous strand of suture going back and forth, resembling the suspension lines of paracord in a parachute. Gentle back-and-forth traction, and/or careful application of a nerve hook, is used to slide conduit and target together. Individual tastes may vary: some people will continue the anastomosis with the assistant providing continuous tension so that nerve-hooking is not necessary upon completion. I prefer to leave things a little loose to maximize visibility, and nerve hook at the end. ###### Advantages: Excellent visualization ###### Disadvantages: The "parachuting" traction does cause a little more injury to the vessel wall. It requires a skilled and attentive assistant, which may not always be available at all hours of the clock. ###### When to use: All arterial distals, all sequential end-to-sides, most arterial proximals, very small conduits. ###### The Role of the Assistant Secures the conduit by adventitia at the toe, so that the opening faces the target. Spreads open the conduit with a forceps inserted into it to facilitate first conduit bite. Thereafter, alternates holding the growing number of suture strands close to the conduit, close to the target. By doing so, the assistant keeps the strands out of the way, and their traction opens up the conduit or the target. Each time a bite is placed in either, re-grab, and retract. A forceps with a non-serrated edge is used to avoid weakening the suture. When the surgeon has reached 180 degrees from the starting point, finishing with a bite on the target, the assistant hands the other end of the suture to the surgeon, and then a back-and-forth traction is coordinated between the two. The surgeon pulls on the needle-driver end and the rubber-shod end simultaneously, and takes up as much of the slack as possible. Friction will prevent a complete approximation in one go, and there will be some uneven loops. Then the assistant pulls on the conduit, and the uneven loops will be taken up. The surgeon alternates with traction on both instruments again, and eventually the graft and target are in close contact. If this process is unsuccesful, or the tissues involved seem too delicate, the same can be accomplished with a nerve hook. Rules of the nerve hook: 1: only hook suture loops that originate and terminate OUTSIDE the anastomosis, 2: hooking the stitches is like the perfect call system for residents- every other one. In the absence of a skilled assistant I will sometimes get by using a 6-0 prolene passed through the toe adventia of a graft, secured to a neighboring pericardium in the vicinity of the target using a rubbershod to hold the conduit, and perform every other role myself. ##### Tie Down The Tie Down technique avoids the use of an assistant to hold the conduit close to the target. Instead, the first stitch secures the two structures together, and is tied down. Each arm of the suture is then brought up either side of the anastomosis until completion. The nerve hook is used to take up slack, and the suture arms are tied together. ###### Advantages: Minimal need for a skilled assistant. Speedy anastomosis. ###### The Role of the Assistant The assistant has a minimal role in this technique. The surgeon that lays out the suture appropriately will have little need for someone to follow. Occasional traction on epicardium or on the graft a distance away from the anastomosis will provide optimal visualization, but this can often be done by the surgeon. Application of the suction when required is helpful. When performing proximals on both sides of the aortic root, the technique allows simultaneous anastomosis of two grafts. ###### Disadvantages: Can limit visualization in the smaller target. It can be difficult to take the smallest possible bite of the conduit when circumstances require, as in the example of the very small conduit to aorta proximal. ###### When to use: Anytime exposure and conduit size permit. #### Creation of the Ideal Tie Down Coronary Anastomosis A description of ideal coronary anastomoses utilizing the Tie Down technique follows. This will illustrate the application of the rules and guidelines delineated above, which are as true of this technique as they are for the parachute. The motions described are for the right handed surgeon, but can be altered as necessary to facilitate the techniques for the left handed surgeon, though the sequences will need to change, in accordance with the rules and guidelines. ##### Nomenclature for orientation. For ease of our description, let us establish an orientation based on broadly applicable compass points. - Toe- self explanatory - Heel- self explanatory - Far- the side of the anastomosis further away from the surgeon, used when the toe or heel are oriented in the general direction of the ends of the table. - Near- the side of the anastomosis closest to the surgeon, used when the toe or heel are oriented in the general direction of the ends of the table. - Left- the side of the anastomosis on the left hand of the surgeon, used when the toe or heel are oriented in the general direction of the sides of the table. - Right- the side of the anastomosis on the right hand of the surgeon, used when the toe or heel are oriented in the general direction of the sides of the table. We can therefore refer to quadrants of the anastomosis in the following fashion: Toe-Right, Heel-Left, Toe-Far, as appropriate for the specific anastomosis. ##### Orienting the anastomosis As touched upon above, the anatomy of a specific vessel may vary, and the degree of manipulation applied to the heart can combine with that variation to make certain vessels capable of different styles of orientation. An LV branch parallel to, and close enough to the PD, with sufficient traction on the acute margin, can be treated like the PD. Likewise an extremely distal PLOM that is close to and parallel to the PD. The far-away LV Branch might be more fruitfully treated like the Lateral OM/Standard PLOM, with the apex in the air, pointing away from the spine. The Ramus might be rotated to be like a diagonal, or treated like an LOM. It is the surgeon's decision how to manage this exposure and orientation, but in practice, it is easy to determine what will work best. #### the PLOM/LVbr The apex of the heart is oriented in the air, away from the spine, pointing, if possible to the surgeon. The Toe and Heel are oriented in the axis of the sides of the table. Upon arteriotomy, the heel is tied down first, passing needles inside out on conduit and target, tied on the same side of the anastomosis, outside of the vessels. The conduit is flipped to lie inside the "V" formed by the two even arms of the suture. The left hand arm of the suture is laid to the left, and secured with the weight of a moist lap. The right hand suture is loaded forehand to begin the Heel/Right quadrant, which is deep in the chest in this orientation. - ##### Heel: Already tied. - ##### Heel/Right One-off: Forehand, outside in on conduit, backhand, inside out on target. - ##### Heel/Right Quarter: Forehand, outside in on conduit, backhand, inside out on target. - ##### Heel/Right Wing: Forehand, outside in on conduit, backhand, inside out on target. Lay out the right arm suture to the right, shift the moist lap from the left arm to secure it with its weight, load the left arm on the needle driver for forehand. We now begin the Heel/Left quadrant, continuing to Toe/Left, and concluding with Toe/Right. - ##### Heel/Left One-off Forehand, outside in on conduit, backhand, inside out on target. - ##### Heel/Left Quarter Forehand, outside in on conduit, forehand, inside out on target. This, and the other stitches up to the toe can often be "taken in one", which can save considerable time. - ##### Heel/Left Wing Forehand, outside in on conduit, forehand, inside out on target. I know, I know. Does the Wing belong to the Heel or Toe quadrant because its exactly halfway? Write your own fucking textbook. - ##### Toe/Left Quarter Forehand, outside in on conduit, forehand, inside out on target. - ##### Toe/Left One-off Forehand, outside in on conduit, forehand, inside out on target If there is any doubt, take this one and the next two stitches "in two bites." - ##### Toe Forehand, outside in on conduit, forehand, inside out on target. - ##### Toe/Right One-off Forehand, outside in on conduit, forehand, inside out on target As can be seen, this has maximized the amount of forehand utilization - ##### Toe/Right Quarter Consider doing this one, and the next, backhand, one bite. Outside to outside, of course. - ##### Toe/Right Unnecessary Stitch Nerve hook up the right hand arm, put in on tension. Nerve hook up the left hand arm, across the toe, snap off a needle, tie it down. #### the diagonal The apex of the heart is oriented towards the midline, having been rolled in that direction by a lap sponge tucked in behind the obtuse margin in the pericardial well. The Toe and Heel are oriented in the axis of the head and foot of the table. Upon arteriotomy, the toe is tied down first, passing needles inside out on conduit and target, tied on the same side of the anastomosis, outside of the vessels.. The conduit is laid out towards the surgeons left hand. The right hand arm of the suture is laid to the right, and secured with the weight of a moist lap. The left hand suture is loaded forehand to begin the Toe/Far quadrant. The suturing will be going from right to left. ![[Diagonal or LAD Vein Anastomosis.png]] - ##### Toe: Already tied. - ##### Toe/Far One-off: Forehand, outside in on conduit, backhand, inside out on target. - ##### Toe/Far Quarter: Forehand, outside in on conduit, backhand, inside out on target. If the angles are favorable, the surgeon can turn his body to take this in one, going backhand outside in on conduit, and continuing backhand inside out on target. - ##### Far Wing: Take this in one, going backhand outside in on conduit, and continuing backhand inside out on target. - ##### Heel/Far Quarter: Take this in one, going backhand outside in on conduit, and continuing backhand inside out on target. The conduit is now flipped over to the far side of the anastomosis. Gravity therefore opens up the conduit, and allows easy visualization of the inside of the conduit and target. - ##### Heel/Far One-off: Properly done, with enough degrees of freedom, this can be taken in one as well. Forehand outside in through conduit, inside out through target. - ##### Heel: Properly done, with enough degrees of freedom, this can be taken in one as well. Forehand outside in through conduit, inside out through target. Use the needle in your hand to help you "nerve-hook" the last few suture loops, and pull up some of the slack. You don't want it so tight that you can't separate conduit from target a little and get in the way of your own visualization, but you don't want enough interior loop from the far stitches to accidentally snag one of these when you are doing the near side. You are about to sew forehand towards yourself for the next two stitches, which is the most natural direction possible. - ##### Heel/Near One-off: Take this in one, forehand outside in through conduit, inside out through target. - ##### Heel/Near Quarter: Take this in one, forehand outside in through conduit, inside out through target. Because you have been sewing forehand towards yourself, you will be sorely tempted to take the next stitch the same way, but it must be resisted, or it will limit your visualization of the next quadrant. Lay down the left hand arm of the suture to your left, weigh it down with the moist lap. Load the right hand arm of the suture for forehand. - ##### Toe/Near One-off: Forehand outside in through conduit, and usually a forehand inside out through the target. A diagonal that angles out very sharply towards the obtuse margin, the second bite of the stitch will sometimes be more favorably done backhand inside out through target. - ##### Toe/Near Quarter: Take this in one, forehand outside in through conduit, inside out through target. - ##### Near Wing: Take this in one, forehand outside in through conduit, inside out through target. - ##### Unnecessary Stitch: Take this in one, forehand outside in through conduit, inside out through target. Nerve hook up the right hand arm, put in on tension. Nerve hook up the left hand arm, across the heel, snap off a needle, tie it down. #### the ramus The Toe-Heel axis is oriented along the foot and head of the table, so we use the far-near, toe-heel orientation terms, as in the diagonal. If the heart can be rotated sufficiently counterclockwise, the sequence of bites/stitches can be the same as that of a diagonal. If not, the sequence can be similar to that of a PLOM, except one ties down the far wing, lays down the conduit between the two arms of the suture, starts with the right arm of the suture to peform the Toe/Far Quarter, the Toe/Far One-off, and the Toe. I find it easiest to continue this arm through to the Toe/Near One-off and Toe/Near Quarter before laying it to the right, and switching to the left arm of the suture. #### The PD The Toe-Heel axis is oriented from the left to the right side of the table, so we use the left-right, toe-heel naming conventions. Here it is better to perform the right hand side of the anastomosis first, as gravity on the conduit will open up the left hand side and facilitate exposure. Upon arteriotomy, the toe is tied down first, passing needles inside out on conduit and target, tied on the same side of the anastomosis, outside of the vessels.. The conduit is laid out towards the surgeons left hand. The left hand arm of the suture is laid to the left, and secured with the weight of a moist lap. The right hand suture is loaded forehand to begin the Toe/Right quadrant. - ##### Toe/Right One-off: Forehand outside in on conduit, backhand inside out on target. - ##### Toe/Right Quarter: Forehand outside in on conduit, backhand inside out on target. - ##### Right Wing: Take this in one, forehand outside in through conduit, inside out through target. - ##### Heel/Right Quarter: Take this in one, forehand outside in through conduit, inside out through target. Flip the vein over the axis of the anastomosis, towards the right hand, and gravity will open up both conduit and target. - ##### Heel/Right One-off: Take this in two, forehand outside in through conduit, forehand inside out through target. - ##### Heel Take this in two, forehand outside in through conduit, forehand inside out through target. Use the needle in your hand to help you "nerve-hook" the last few suture loops, and pull up some of the slack. You don't want it so tight that you can't separate conduit from target a little and get in the way of your own visualization, but you don't want enough interior loop from the right side stitches to accidentally snag one of these when you are doing the left side. - ##### Heel/Left One-off: Take this in two, forehand outside in through conduit, forehand inside out through target. - ##### Heel/Left Quarter: Take this in two, forehand outside in through conduit, forehand inside out through target. Lay the right arm of the suture out to the left, weigh it down with a moist lap. Load the left arm of the suture forehand. - ##### Toe/Left One off: Take this in two, forehand outside in through conduit, backhand inside out through target. - ##### Toe/Left Quarter: Take this in one, forehand outside in through conduit, forehand inside out through target. - ##### Left Wing: Take this in one, forehand outside in through conduit, forehand inside out through target. - ##### Unnecessary Stitch: Take this in one, forehand outside in through conduit, forehand inside out through target. Nerve hook up the left hand arm, put in on tension. Nerve hook up the right hand arm, across the heel, snap off a needle, tie it down. #### The RCA There are two schools of thought with the RCA. Many excellent surgeons orient the anastomosis in the direction of the vessel, so the orientation will be Toe to the surgeon's right, Heel to the left, so we would use the Far-Near convention. Taking the action of the gravity into consideration, I would recommend tying the Toe down, doing the near side first, around the Heel to the Heel/Far Quarter, and coming up the Toe/Far quadrant with the other arm of the suture, finishing with the vein draped towards the surgeon. The advantage of this orientation is that it uses less conduit. The graft will tend to lie in the AV groove, alongside the course of the native RCA. In this anterior position, it may be vulnerable to an overlying chest tube. On a re-operative sternotomy, it will be susceptible to injury. In my opinion, the greatest danger will be in judging the length of the graft and the angle of the anastomosis taking into consideration the volume status of the right ventricle. I prefer to orient the graft at ninety degrees to the RCA, so that the toe is oriented towards the apex, the heel towards the right atrium. The Left/Right naming convention would be in use. The heel is tied down first, the conduit is lain between the "V" of the left and right arm of the sutures. Once again, planning for the effects of gravity on the conduit, the right side is done first, taking the suture arm up to the toe, then continuing with the left from Heel to Toe. Gravity will open the anastomosis to facilitate the second arm. #### The LCX The circumflex proper is rarely grafted, as it is challenging to access beneath the epicardium of the AV groove in the obtuse margin, where other fragile structures, prone to hemmorhage may lie. Sometimes there is no other choice. It can be approached at the base of the left atrial appendage, very very carefully. If the vessel can be safely found, and it is not too calcified, I would suggest painstaking, each stitch in two bites anastomosis. The toe should be tied down first, obviously in the direction of the back of the heart, the heel anterior, orienting the graft towards the left aspect of the aorta. I would tie the knot at the heel. #### The LAD On the rare occasion that a free conduit is grafted to the LAD, the stitching sequence should be the same as a diagonal. Usually, the surgeon is using a conduit that has limited mobility at the proximal end. A tie down technique for this anastomosis does exist, but the best visualization of the crucial heel stitches is obtained with a version of the parachute technique. The axis of the anastomosis is along the table, so we use the Far/Near nomenclature. The arterial graft is laid alongside the arteriotomy, a half centimeter to one centimeter apart, like an open face sandwich. The assistant holds what will be the near wing of the arterial conduit with a Mills forceps in the left hand, and with the right hand uses a coronary Debakey on the adventitia of the conduit, a half centimeter or so away from the anastomosis. The right hand can tilt the angle of the conduit in whichever way optimizes the view of the interior of the conduit for the surgeon. Usually, not much manipulation is required. Gentle traction on the Near Wing of the conduit as the anastomosis progresses puts the loops of suture on a little tension, which opens up both conduit and target. A 7-0 prolene is secured and weighted on one arm by a rubbershod mosquito clamp positioned as close as possible to the anastomosis. The surgeon holds the far wing of the conduit with a Mills forceps, and the anastomosis begins here. - ##### Far Wing: Forehand, outside in through conduit, backhand inside out through target. The needle is picked up with the forceps, ALWAYS to the left of the previous loops, in the direction of progression of the suture line. - ##### Heel/Far Quarter: Forehand, outside in through conduit, backhand inside out through target. - ##### Heel/Far One-off: Forehand, outside in through conduit, forehand inside out through target. - ##### Heel With a smaller conduit, this can occasionally be tricky. The tension of the suture loops to this point should open up the conduit, but if additional help is needed, either assistant or surgeon can gently insert the rounded disk tips of the Mills forcep into the heel of the conduit and allow the tips to expand slightly, opening up the conduit. If the assistant does so, the coronary Debakey is shifted to the Near wing, and the Mills are inserted using the left hand. The surgeon should tilt his or her body slightly towards the direction of the feet, so that an easy forehand outside-in stitch is performed along the axis of the conduit. If the surgeon is inserting Mills forceps tips into the conduit, it is done with the left hand in an overhand position. Then the surgeon tilts his or her body back to face the table, and an easy inside-out forehand is taken through the target. - ##### Heel/Near One-off: Still grasping the near wing with a Mills, the surgeon tilts body slightly towards the head, outside-in through conduit, inside out through target. The conduit is now starting to “flip” over into its final configuration. Forehand outside in through conduit, forehand inside out through target. - ##### Near Wing: The assistant releases the Near Wing, and the surgeon picks it up with a Mills forceps. As always, adventitia can be grasped, but if required, the rim of arterial tissue grasped full thickness by the forceps can be excluded from the anastomosis by the suture line. Forehand outside in through conduit, forehand inside out through target. The delicacy of arterial tissue does not permit the see-saw method of bringing down the parachute. The loops must be nerve-hooked. Only hook the loops exterior to the anastomosis. Don’t try to take up all of the slack in one loop, go back and forth until the conduit is against the target. Try to take up most of the slack in the direction of the sewing arm. Take the rubbershod and shift its position towards the patient’s left shoulder so that a slight tension tends to open up the remainder of the anastomosis. - ##### Toe/Near Quarter: Forehand outside in through conduit, forehand inside out through target, with the body of the surgeon tilted slightly towards the head of the bed. - ##### Toe/Near One-off: Forehand outside in through conduit, forehand inside out through target, with the body of the surgeon tilted slightly towards the head of the bed. - ##### Toe: Forehand outside in through conduit, surgeon faces the bed, backhand inside out of the target. - ##### Toe/Far One-off: Backhand outside in through conduit. Assistant retracts on target epicardium to open up the anastomosis. Backhand inside out through target. This stitch can often be taken in one. - ##### Toe/Far Quarter: Backhand outside in through conduit. Assistant retracts on target epicardium to open up the anastomosis. Backhand inside out through target. This stitch can often be taken in one. - ##### Unnecessary Stitch: Backhand outside in through conduit. Assistant retracts on target epicardium to open up the anastomosis. Backhand inside out through target. This stitch can often be taken in one. Give the anastomosis a final nerve hook, snap off the needle, tie it down. I prefer to perform this anastomosis last, after proximals are taken off the aorta. Whatever air has been introduced to the aortic root during the proximals will have dissolved at the conclusion of this anastomosis. The spring clip occluder is removed from the anastomosis. The surgeon watches the reperfusion attentively for any compromise in hemostasis, and for the reassuring flush of blood distending the conduit and visibly filling the distal LAD. It is time to place pacing wires and reperfuse the rest of the heart through the root with warm blood through the root vent catheter, while putting the head down. ### Proximals As the saying goes, proximals are one of the few things in life that *can* be perfect. The creation of the aortotomy is covered in the section on determining graft length. It will be obvious that grafts coming up the right side of the heart should come off the right aspect of the aorta, typically at around 90 degrees to the aorta, and vice versa for the left sided grafts. The angle of the right sided vein graft will vary only slightly, as the area of the aorta suitable for the proximal is relatively circumscribed: caudal to the pericardial reflection over the SVC, cephalad to the right atrial appendage, anterior to the SVC itself. The only other factor, which is very rarely limiting, is the course of the aortotomy for a simultaneous Aortic Valve Replacement, which is usually well below the level of the ideal RCA proximal. On the left, the angle of the proximals to the Aorta often takes a slight angle towards the left shoulder to avoid the infundibulum or a larger than normal Pulmonary Artery. To assess this angle, or as discussed below, to properly determine length of grafts, it will usually be a good idea to distend the right heart by temporarily cutting down on venous drainage. As far as the angle of a SVG to the Aorta itself, regarded from the cross-sectional plane of the Aorta, the "Proud" graft, that is to say, the graft that tumescently protrudes outwards from the Aorta at 90 degrees, should be avoided. It is likely to kink when the mediastinal tissues cover the heart once more. This is done by a combination of a bevel in the anastomosis, and an aortotomy on the side, rather than on the anterior surface. If a larger than normal conduit is utilized, and not much bevel is desired, than the aortotomies for right and left targets must come directly from the lateral surfaces of the aorta, carefully planned so that the graft "rests" on the SVC or the PA respectively. As the grafts on either side emerge as described above, the axis of the anastomosis determines a Left/Right nomenclature. A tie-down technique is preferred except when the conduit is very small, as in a free arterial graft, or a similarly sized vein graft. The Heel is always tied down first. The surgeon side determines Left/Right as usual. To protect against torsion of the graft, the stitches for the Heel tie-down are ritually placed in the same way. The arm passed inside out through conduit is brought to the head of the table, weighed down with a moist lap. The other arm is loaded forehand weighted with the needle driver, which is rested below the sternal retractor, towards the foot of the table until the hole-punch aortotomy is performed. Then it is passed inside out through the Aorta in such a way that the two suture arms, evened out, are laid out towards the head of the table. The knot is tied down at the Heel on the "North" side of the anastomosis, i.e. towards the head. This is known as "Tieing North." One suture arm is passed underneath the vein, and weighted down with a moist lap. The RCA begins with a series of backhand outside-in through conduit, forehand inside out through Aorta. At the Left Wing, you can transition to a backhand outside in through conduit, backhand inside out through Aorta that can be taken in one. Stop at the Left/Toe Quarter so as to avoid limitation of the exposure of the other side of the proximal, weight that arm of the suture down, and load the other forehand. Forehand up, outside in through conduit, forehand down, inside out through Aorta. At the Right Wing, the stitch can be taken in one, surgeon's body tilted towards the foot, forehand outside-in through conduit, inside-out through Aorta, continuing in this fashion until the Toe/Right One-off, when it will become more comfortable to take these in one, backhand all the way until the Unnecessary stitch closes the gap. Left sided grafts, performed from the surgeon's side of the table begin with the Heel/Right One-off, forehand up through the conduit, backhand inside out through the Aorta. One is permitted perhaps a single "In One" stitch, all forehand at the Toe/Right Quarter or One-off before switching to the other arm of the suture. Then it is forehand up through conduit, backhand inside out through the aorta for the Heel/Left One-off, before transitioning to "In One" all forehand, maybe at Heel/Left Quarter, definitely at Left Wing, and then all the way through around the Toe to the Unnecessary bite. When operating with a skilled assistant, if the principles described above are followed, it is easy to sew simultaneous proximals, a right for the surgeon, a left for the assistant, and thereby cut down on ischemic time. Small conduit, free arterial conduit, when taken off the Aorta is best anastomosed using a parachute technique. The aortotomy must be carefully created, preferably with a single fire of the hole punch, or the use of the smaller punch. Too large an aortotomy will stretch out and flatten the anastomosis. The bites on the conduit must be as delicate as possible, particularly from heel to wings, or it will distort the inflow to the graft. If these precautions are followed, the anastomosis can be very satisfying. Alternatives with many advocates are to take the proximals of these grafts off of other *in situ* grafts, or off the hoods of vein grafts, or even off pericardial patches sewn to the Aorta. ### Order of bypasses Priority is given in the emergency situation, where ischemia is already in progress, to the culprit lesion. As soon as the distal is performed, blood cardioplegia is infused down such a graft, mitigating the ischemia as quickly is possible. Otherwise, consider the territory provided by the target, and the size of the target to assign a priority that allocates the best quality segment of conduit to the most important target. Whenever I can, I like to perform right sided anastomoses first, as this permits me to perfuse these grafts with cardioplegia. This can be done by direct cannulation of the graft and instillation of cardioplegia, and/or via the root after completion of the proximal. Antegrade cardioplegia may reach these targets imperfectly because of the coronary disease, and retrograde cardioplegia wont reach them because of anatomy. This strategy can help avoid the necessity for retrograde cannulation in isolated bypass grafting surgery. Carefully inspect the conduits before going on cardiopulmonary bypass. Is there an inopportune varicosity? Does an irritating bifurcation give you a suboptimal length in the middle of the harvested saphenous vein? Plan accordingly. If necessary to assist in planning, a method is described in the next section that can help predict what length of vein is required to reach a specific target. The harvested vein can be laid out, and individual segments can be measured and trimmed for specific targets using this method. My preference, when circumstances permit, is to work with the entire length of vein, trimming it as you go from one graft to another. ### Lengths of grafts It is axiomatic with grafts to obtuse marginals, "A thousand too long before one too short." To unpack the axiom, the consequences of a graft that is too short are obvious, but left sided grafts that are too long are usually well tolerated (to a point). There is plenty of space in the pericardium behind the AV groove to accomodate the extra length or a course behind the left atrial appendage can be selected. The excess length can be laid out over the infundibulum, or over the aortopulmonary trunk where a little wad of surgicel can be tucked into the excess loop to maintain a graceful curve. On the right side of the heart, there is less margin to accommodate a graft of excessive length. All this said, I prefer just a little generosity on the length when it is possible, as I believe there is a tendency to underestimate the effect of the appropriate volume load on the expansion of the ventricles. Also, on catheterizations of patients previously grafted (by other surgeons, of course), I have noticed grafts that seem to tent up their targets. To me this suggests that the partially devascularized saphenous vein graft contracts a little as it heals. Wherever you have determined is the exact place to trim the graft to create the proximal, there will invariably be swelling of the vein that signals a valve will be found there. Avoid it if you can. Whenever you transect a vein to create an anastomosis, look inside of it for valves. If you have no other choice, trim the valves out with a tonotomy scissors. It is preferable to work with endothelium that has not been provoked. A warning about the excessively large diameter saphenous vein graft (a.k.a the “snausage”). Think twice about using this conduit, and consider harvesting another length. These are very hard to work with, involving very large anastomoses on either end, with a tendency to stick out at right angles from the heart and the Aorta. Their length must be perfectly judged, or they will kink very easily. Blood does not flow very well in these grafts, and has a tendency to stagnate. Be suspicious of the possibility of a snausage if a venous ultrasound shows a 6mm-8mm diameter or greater: when the graft is pressurized, you may not like it. As a rule of thumb, a conduit as long as a standard pair of Metzenbaum scissors will span the distance from the Aorta to the back of most hearts. At the beginning of your career, or that of your harvesting assistant, request a full length more than what you need. As confidence grows, request a half length more than what you need. This will help cover the unexpected injury to the mammary during harvest, the surprise targets, and all manner of uninvited phenomena. To determine length after a distal anastomosis, lower the heart into position while infusing heparinized saline into the preferred vein cannula (described below). The vein cannula will slide into a non-torsed orientation as the vein is draped in the desired course. With the heart lowered into normal anatomic position, antegrade cardioplegia is administered through the root. After a desired dose is given, while continuing cardioplegia administration, the perfusionist is directed to distend the right heart. Heparinized saline flush is continued down the graft to determine length, and when satisfied, the heart is emptied while the graft is trimmed, carefully preserving orientation by grasping adventitia at the intended toe. The suture is passed and the orientation marked as described above. The Aortic root is distended temporarily with cardioplegia, the eleven blade is inserted just large enough to insert the hole punch, cardioplegia is stopped but not vented yet. It is easier to insert the hole punch with a little tone in the Aorta. The hole punch is nearly but not completely closed on the Aortic wall, while the root is vented. When it begins to inflate, the hole punch is fired. For some reason, this leaves a nicer, round hole. You will also have less blood emerge into the well. Turn down the root vent to avoid entraining too much air into the root. Begin your proximal. If you need to plan exactly how much vein is required, the following steps can be taken to estimate. Expose the desired target. Put a zero silk tie on a right angle, and put the tip of the right angle on the area of intended arteriotomy. The assistant uses a Debakey forceps to place the other end against the intended site of the proximal on the Aorta, and a second Debakey takes the length of silk and forces it into contact with the pericardium. A medium clip is placed on the silk where it touches the Aorta and the silk is cut. The process is repeated as necessary, marking each silk with a different number of clips to distinguish the grafts. To conserve vein on a left sided anastomosis, consider orientation of the distal anastomosis ninety degrees to the target. Instead of creating a course that goes backward to the AV groove before traveling over the PA to the Aorta, it will use a "Great Circle Navigation" route to get to the proximal. The former technique is less sensitive to changes in ventricular distension than the latter, but the latter uses less vein. #### When Grafts are Too Short Veins to left-sided targets can be taken off of other veins to left-sided targets as a Y graft. If the LIMA is generous enough, they can be taken off the LIMA. It is better to come to this decision before the LIMA is anastomosed. The vein anastomosis is performed with the LIMA reflected upwards so that the anastomosis is performed to the underside of the LIMA, and it should occur so the anastomosis is within the pericardium rather than originate from a segment within the left pleura. This technique is also useful when calcification of the aorta denies this as a source of the proximal. For right-sided vein grafts, and when all other techniques have failed with left-sided grafts, harvest an additional length of vein, and add to the length of the graft. Spatulate the ends of both segments opposite each other to expand the anastomosis circumference and counteract scarring and pures-stringing as causes of stenosis at that site. I will mention the following technique, not because I would recommend it, but only because I have seen it done with survival of the patient in someone elses hands. I shudder to mention the use of pledgeted stitches from the acute margin of the right ventricle to the anterior surface as a “scrunch stitch” to accomodate a short vein graft. The LIMA to LAD graft that is too short offers a few options. First, bear in mind that the open sternotomy distracts the source of the LIMA from its target. Relax the sternal retractor and assess the length again. An in-situ LIMA graft is too short if tension does not allow movement in any direction by one centimeter. Obviously, the graft can be taken free, detaching it from the subclavian and taking it off the hood of a vein graft, or directly off the aorta. If the in-situ graft is pedicled, multiple focal skeletonizations, as many as necessary can be performed. At as many places as necessary, cut through the fascia/internal thoracic muscle pedicle, free the accompanying veins from the mammary, clip the veins on either side and sever them. A surprising amount of length can be generated this way. At least, it will surprise those who do not entirely harvest their in-situ grafts in a skeletonized fashion as a matter of routine, as I do. A LIMA that has not been harvested fully to the subclavian because of technique or because of anatomic limitations (osteophytes, etc), or in the setting of chronically obstructed lungs may have tension introduced by medial intrusion of the left upper lobe. An extremely useful trick here is to carefully plan the course of the mammary THROUGH the portruding lobe, and create a channel with a lung-load stapler. Most of the time, only a single fire is necessary. A small wedge can be taken if necessary. ### When Bifurcated Veins Attack Sometimes, the greater saphenous vein will bifurcate into two smaller branches, often fusing back into one common vein some centimeters later. This is usually a disadvantage, and if the situation is encountered with ultrasound prior to incision, consideration should be given to chosing another segment for harvest. When no other choice exists, the lemons can be turned to lemonade. If the vein bifurcates and fuses again into a single conduit over one length, one option is to ignore the bifurcation, and perform anastomoses using those portions where the vein is single. The blood will travel down both limbs to perfuse the target. The bifurcation can sometimes be harnessed as a natural "Y", so that a single proximal can perfuse two targets. If both limbs are of adequate size, sometimes this can permit two lengths for one harvested length, as can be seen in the illustration. Most often the bifurcation will result in a length that, if reversed, would produce a graft with a very small proximal ending in a larger distal. A solution to this situation that should only be employed in desperation is to use a valvulotome to cut the valves in the vein and use it in a non-reversed fashion. The larger end will thereby form the proximal, and the small end the distal. The appropriately sized LeMaitre valvulotome is best when available, as it more thoroughly removes the valvular tissue circumferentially and with minimal endothelial disruption. The Mills valvulotome is more readily available. It cuts through the valve tissue in one place in the leaflet rather than excising the leaflets. Be careful not to let the Mills valvulotome accidentally enter the inevitably nearby side branch, as traction in this position will cut the wall of the vein. ### Sequential Grafts The trade-off inherent to sequential grafts is that the greater outflow of the sequential graft results in greater patency over time, but that there is greater technical challenge in the construction of these grafts. If the surgeon does not rise to the challenge, the eggs of all the distals on the graft are in the basket of the single proximal. A few guidelines will maximize the chances of success. There are two kinds of sequential: the serpentine and the diamond. The serpentine sequential has the anastomosis oriented along the length of the target, as well as that of the conduit, in such a way that the anastomosis will make both vessels run in the same direction. The anastomosis length can be larger than that of diamond, if desired as there is less risk of "seagulling" (see below). This kind of sequential will NOT conserve vein, and must be carefully planned, occasionally with tacking stitches so that resulting loops of graft will not flip over and occlude the graft. This technique can be useful when conduit is plentiful, but does not have sufficient diameter to anastomose in the diamond technique. The diamond technique makes a small incision in the conduit, along the conduit, and a small incision in the target, also along its length, but the graft and target will be at ninety degrees or so to each other. The stereotypical diamond sequential bypasses a LOM and an PLOM. The distal is constructed first with vein at ninety degrees to the PLOM so that the vein is directed to intersect at ninety degrees with the LOM on its way to the proximal off the left side of the Aorta. Artistry is involved in laying out the graft after the distal to plan how much length is desired between anastomoses to allow for expansion of the heart. Watch out for valves, and avoid them. Sometimes Janus, God of Thresholds is kind, and a vein branch will be positioned exactly where you want it. The branch can then be truncated to use graftotomy for the anastomosis. If not, mark the position of the intended vein incision with a marking pen. The vein incision can be made with a fifteen blade between two coronary Debakeys grasping the adventitia on either side. I find the greatest control by grasping the adventitia between two coronary Debakeys while distending the vein with flush and allowing it to kink slightly between the forceps. I then use the very tips of the tonotomy scissors to make a small snip that has never resulted in a backwall injury for me. Incisions in conduit and target must be kept small, and the bites must be delicate, or they will be a tendency to flatten the graft at the sequential in a way that limits flow to the distal limb. When seen from the side, such a graft looks like a child's illustration of a seagull, hence the term "seagulling." If the sequential target is to a deeply intramyocardial target, the danger of this is greater. This would be a good time to be blessed by a sidebranch on the vein conduit right at the site of the anastomosis. The small anastomosis should be performed in a parachute fashion, and a nine-stitch schedule (eight compass points and one unnecessary stitch if desired) is used. I find the Mills forceps very useful in this setting: their small rounded tips are easily inserted into the conduit opening and allowed to spread to facilitate the interior view as the needle is passed through the wall from outside to inside. The tips are in position to grab the needle as it emerges from the conduit opening. The back half from conduit toe to conduit heel is completed, the parachute is brought down with nerve hook applied to the exterior loops, then the front half. The Mills insertion in the nearly completed anastomosis is very handy in performing the last stitch. As usual with parachute anastomoses, the closer the rubber-shodded end of the suture is to the anastomosis, and the closer the conduit is held to the target, the easier it is complete the back half of the parachute with the available suture. Remember that the fewer the stitches, the greater the potential to "purse-string" the anastomosis when the knot is tied. It would be a good idea to flush the graft while tieing the knot, and only tieing with enough force to stop the leaking. The hemostasis and flow can be checked by placing a spring occluder (a.k.a. soft dog) on the limb to the distal target. It is very impressive to see how good the flow is through such small anastomoses. The most distal target should be the one with the best outflow. Otherwise there will be a tendency for the flow to go preferentially down the intermediate sequential target, and the distal limb may go down. #### Sequentials Involving Arterial Grafts The radial graft lends itself well to sequentials, but precision is even more of a necessity to avoid flattening the graft at the side-to-side anastomosis. Sometimes, a truly exceptional diagonal target is worthy of consideration for a sequential diagonal and LAD sequence of anastomoses. Some important issues need to be considered. The LIMA must be quite generous to allow this to occur. Is the diagonal branch truly worthy of polyamory in the traditional marriage of LIMA to LAD? What angle does the diagonal take off of the LAD? The smaller the angle, the more likely that the diagonal and LAD anastomoses will both be in alignment with the LIMA. The wider the angle, and the more proximal the take-off is along the LAD, the more likely that one anastomosis will be aligned with LIMA, and the other be at right angles. It is extremely important to take into account the rotation of the heart involved in exposing the diagonal and the LAD. Properly exposed, the diagonal and LAD will seem to take a straightforward course that lines up well with the LIMA. When the heart is rotated back into the normal position, however, there may be a tendency to kink the LIMA at the diagonal anastomosis. The greater the angle of the take-off, and the more proximal the take-off of the diagonal, the greater the tendency for this to happen. If rotation has been taken into account, the wider angle take-off has most often required the LIMA-diagonal anastomosis to be in-line, and the LIMA-LAD anastomosis to be at right angles, as the mammary approaches the LAD lateral to medial. It is always easiest to the middle anastomosis before the distal on this kind of in-situ graft. The sequential LIMA taken to two different targets on the LAD is, in my belief, the superior solution to the serially diseased, multi-segment LAD, a subject worthy of brief discussion. If an LAD has a set of lesions that divides into at least two segments, a few options exist. 1. Graft just one segment with the LIMA. Choose the one with the most septals, and rely on future native vessel PCI to open the other segment up to the LIMA should it ever become necessary. The future PCI can sometimes be performed via the LIMA itself. 2. Graft one segment with the LIMA, the other with a vein. 3. If the disease subdividing the LAD is focal enough, consider crossing the disease with the arteriotomy (if calcification permits) and performing a patch angioplasty with the longer than usual LIMA anastomosis (optimal) or with a vein patch and put the LIMA on the vein patch (least optimal) or put the LIMA on the best segment itself (in-between optimal). Bear in mind, the vessel in the vicinity of the calcification you cut across may be difficult to penetrate with the suture needle, but it will be difficult to narrow as you search for a soft spot because of the patch angioplasty. 4. Endarterectomize the subdividing disease, tack down the intima posteriorly above and below the anastomosis using a technique similar to that used to close backwall injuries (as described above). Complete with longer anastomosis as in option 3. 5. Sequential anastomosis of LIMA to LAD segments. Also gives you the opportunity, at the cost of some of your myocardial protection, of removing the soft dog before the distal anastomosis to appreciate the excellent flow through the LIMA distal to the previous anastomosis. You may possibly appreciate the degree to which this approach might have been unecessary by observing the rich flow through the distal LAD arteriotomy from the proximal anastomosis when the soft dog is temporarily removed. ## Notes on Instruments ### Castroviejo Needle Drivers Castroviejo needle drivers are designed to perform small anastomoses in cramped spaces using small curved needles passing through delicate tissues. The unifying design feature is a pointed and precise clamping surface that is held in the hand so that the entire driver can be rolled in your fingers in a fashion that employs the curve of the needles to maximum advantage. Some castroviejo drivers emphasize this motion using "half-round" handles with milled grips for extra precision in the fingers. These are intended for distals. Flat handle castroviejo's trade off some of the precision and delicacy for greater mechanical advantages and are more appropriate for proximals or larger anastomoses. Some surgeons are too brutal to ever know the difference. Another design element variation is in the presence of the "clicker" that keeps the tips together, needle firmly grasped, until the surgeon releases the clicker. The clicker is vulnerable to poor handling, and can fail at annoying times. Some surgeons eschew the clicker for this and another important reason: they feel that the clicker permits motions injurious to the vessel wall. The situation they fear is when the clicker is operated when the needle is through tissue: this motion, particularly when the needle is already being grasped in the other hand with a receiving forceps is admittedly dangerous. The adept surgeon practices this sequence of steps until they are fluid: unclick the castroviejo with the needle still grasped tight, roll the fingers to drive needle through tissue following the curve, release the needle from the castroviejo the microsecond before it is grasped by the forceps. The more fluid the steps, the more they will approach practical simultaneity. Do they reach simultaneity? Perhaps- but if so, with the practice there will be no dangerous uncontrolled motion. ### Forceps Three common varieties of forceps are used in coronary anastomosis. The Mills, the Coronary Debakey, and the Golds. They exist on two different spectra: tissue-handling and needle-handling. The Mills have rounded, small and precise, flat rings at the tips. They usually have half-round handles. The coronary Debakey is a small tip version of the familiar instrument, with the grasping surface irregular with very fine teeth. The Gold forceps have grasping surfaces that are smooth, and often coated with diamond-dust to maximize their needle grip. They tend to slip right off tissue. The forceps are organized along the spectra in the following way: Needle handling: The Golds are superior to the Coronary Debakey which are superior to the Mills. Tissue handling: The Mills are more precise and delicate in gripping tissue, the Debakey has the most grip, but less precision and delicacy than the Mills, the Golds. Want to grasp some epicardium on either side of a target while you clear it off and enter it with a fifteen blade? The Debakey is your friend. Want to reliably but delicately hold some adventitia, or maneuver within a tight conduit? Mills are the right choice, but prepare to have the needle take unpredictable angles when you grasp it with the forceps. Proximal vein grafts and most easy distal anastomoses with vein? You wont need much tissue grasping, but Golds will make needle positioning in your driver easy. The proper selection of forceps for the right purpose diminishes your "stammer", the term we use for repetitive motions, wasted attempts to accomplish the same goal. Stammer slows you down not just with the wasted motion but by distracting your attention from the main task. A big percentage of stammer comes from needle-handling. You will come to have an instinct about what needle posture is required for the next bite, backhand or forehand, and what angle. Even when you do, you may have difficulty achieving the desired needle position in your driver. The brute uses his non-dominant fingers to grasp the needle in position for the dominant hand castroviejo to achieve the right angle. It is more civilized to use an instrument, and builds your facility. It is a rookie mistake to unclick the castroviejo with the needle in its grasp and to try to achieve the right angle by twisting it inside the castroviejo with the forceps in your other hand. It is very difficult to get the precise angle this way because of inherent overcontrol. Speed comes from the blend of two options: either learn to use your castroviejo more skillfully to accomodate an awkward needle angle, or achieve the right needle angle efficiently. The way to do so is to grasp the needle firmly in whatever forceps you have chosen, however awkward the angle of the needle within the forceps in your non-dominant hand so the needle is stable. Then, keeping the finger grip of the forceps constant, rotate your hand and wrist to make the needle take the right angle in space relative to a convenient grasp with your dominant hand castroviejo. ### Vein Cannula Some people like to use a plastic or metal vein cannula inserted into the proximal (reversed if vein) conduit. These cannula are typically too large for arterial conduits. Once the conduit is in place, the cannula is secured with a silk tie. Since the conduit wont really be able to rotate over the cannula, it is a good idea to mark the vein with a dashed line to make sure it wont twist and to keep track of the orientation. This kind of cannula has excellent flow, and is very unlikely to kink. The preferred technique at my institution is the use a 20g short plastic IV catheter inserted into the proximal conduit, and temporarily secured with a soft spring clamp bulldog. Downsides: It is easy to carelessly pull the IV cannula out. The flow can be limited, particularly if the IV cannula becomes kinked. The advantage of the technique is that as the conduit is pressurized with saline through the IV cannula to lay it out on the way to its proximal, the conduit will rotate over the cannula as necessary for the optimum lie. You get to use all of the conduit this way, and sometimes you need to use every millimeter. This cannula can be used in all forms of conduit. The Nerve Hook Rubber Shod ## Afterword Much more can be said of the minutiae surrounding Coronary Artery Bypass Grafting: harvesting technique, skeletonization, conduit preparation, pericardial windows, other conduit options, conformations of multiple arterial grafts. It is hoped this will be a dynamic document, and perhaps there will be time to make these and other additions. The glamour of bypass grafting has faded with its ubiquity, so common and 'routine' a procedure has it become. It is disregarded as the "bread and butter" of our craft, while people seek specialization towards what are felt to be more creative aspects of cardiac surgery. Nevertheless, there is a great deal of artistry in the bypass graft, if you know where to find it. Nothing else requires this kind of precision. You will go a far way in medicine before you find a treatment with greater and more durable impact than bypass grafting. As the wags of my childhood used to say, "The most important ten minutes of your life between birth and death are when your mammary gets sewn to your LAD".