Introduction

The Endobutton (EB) is the most widely used femoral fixation device worldwide that is designed specifically for soft tissue grafts. Pioneered by Dr. Thomas Rosenberg and introduced around 1990, it was the first device specifically designed to hold soft tissue grafts. As originally designed, the surgeon would tie a Dacron tape connecting the button to the tendon. In the past 5 years, this technique has been largely supplanted by use of the EB-CL (continuous loop), which obviates the need to tie knots. Due to the longevity of the device, there is a much greater literature concerning it than any of the other newer, soft tissue–specific devices.

Biomechanics

Numerous studies have analyzed the pullout strength and stiffness of this device.^1–7 Much was made at one point about a so-called “bungee effect,” in which the device’s fixation on the cortex and not in the tunnel supposedly resulted in lower stiffness. However, it has subsequently been shown that the greater stiffness resulting from cortical anchorage dwarfs the slightly reduced stiffness from the greater length of the construct.⁷ Also, as described elsewhere in the text, stiffness has no bearing on ultimate stability. This is because the stiffness of all grafts is independent of their mode of fixation once tunnel healing has occurred at about 2 months postoperatively. After this time, load is borne by the healed fibers that connect the graft to the tunnel, not by the device. Because the fixation device is not load bearing after this time, its stiffness is irrelevant. However, even before that time, stiffness is less important because the stiffness represents only elastic deformation of the graft. It is only plastic deformation, not elastic deformation, that will result in greater ultimate clinical laxity. In fact, reduced stiffness, or greater elasticity, in the graft-fixation construct will diminish the forces that tend to displace the fixation device in cyclical loading. This reduction in stiffness diminishes these forces by partially dissipating them in temporary elastic deformation of the graft. Therefore, ultimately, stability may actually be enhanced by protecting the construct from plastic deformation or fixation device slippage while tunnel healing is occurring.

Clinical Results

In the largest meta-analysis of anterior cruciate ligament reconstruction (ACLR) autografts, the EB-hamstring combination was found to have the highest stability rates of any graft-fixation construct when paired with modern tibial fixation.^8–14 Morbidity has been minimal.^15–18 In our experience,⁸ the EB has proven to be extremely reliable. After 13 years of continuous use, as well as follow-up of more than 80% of all implanted EBs, we have had no graft failures (see Chapter 69). Approximately 86% of grafts have had IKDC normal stability. We have had no hardware complications and no displaced EBs, and we have not had to remove an EB.

Surgical Technique

Femoral-Tunnel Formation

Principles of Femoral-Tunnel Drilling

Two femoral tunnels are necessary in sequence with each other, as will be described. They are drilled transtibially. A laser-marked, long transtibial guide pin with markings every 2 mm is used to drill the femoral tunnel at a 65-degree coronal angle. It is important to drill the tunnel such that the exit is in the femoral metaphysis rather than the femoral condyle if possible so that the tunnel will have adequate length. It has recently become apparent that a femoral-tunnel entry lower down at the 10 o’clock position rather than the 11 o’clock position (for a left knee) will improve rotational stability. This lower entry will, however, also result in a shorter tunnel because the exit will tend to be from the narrower condyle, which is lower down than the wider metaphysis. To compensate for this, the knee needs to be more flexed during drilling to 90 degrees or more in order to redirect the tunnel upward toward the metaphysis. The following is our technique for femoral-tunnel formation and EB fixation.

Notchplasty

We always perform a lateral notchplasty, but not roofplasty, of about 3 mm for visualization. All soft tissue should be removed from the tunnel so that the over-the-top position can be clearly seen. It should also be probed to make sure the surgeon knows where the back of the notch is.

Basic Technique

With the leg hanging free at its usual angle of about 75 degrees, a 1-mm deep indentation is made with the tip of the drill pin at the 10 o’clock position, and about 3 mm distal to (forward from) the over-the-top position. The tunnel entry point is found in this fashion because visualization is easiest at this degree of knee flexion. We then pull the pin back and make sure the location for the tunnel is satisfactory. The knee is then flexed to at least 90 degrees and usually to about 100 degrees. The pin is then reinserted into the indentation, and the tunnel is slowly drilled, taking care to stop when the resistance of cortical bone is reached. This is easily felt if the tunnel is drilled slowly and with a light touch. Cortex will usually be reached at 30 to 40 mm, as seen on the laser-marked pin.

Minimum Tunnel Length

Our minimum acceptable tunnel length is 30 mm, which, when the 15-mm continuous loop length is subtracted, leaves 15 mm of graft in the tunnel. The cortex is 3 to 5 mm thick. Thus we make sure that the laser pin shows at least 27 mm inserted into the condyle before reaching the femoral cortex to ensure that the total channel length will be at least 30 mm after drilling through the cortex.

Redrilling if the First Tunnel Is too Short

If the tunnel is less than 27 mm we withdraw the pin, relax the knee to 75 degrees or so, and make a new pilot indentation at about the 10:30 position and about 5 mm distal to over-the-top position (i.e., slightly higher and more distal or forward than the original hole). We then flex the knee to 5 or 10 degrees more than on the first pass, at least 95 degrees, and redrill the hole. The new tunnel should be significantly longer. The changes in tunnel location, along with the greater knee flexion, yield longer tunnels and will still result in a tunnel in a very acceptable location. If the knee is adequately flexed the first time, however, a second pass will rarely be necessary.

Finishing the Femoral Tunnel

Once a satisfactory tunnel has been found, the laser pin is further drilled through the cortex. We observe the laser markings when giving way is achieved, which indicates the pin has burst through the cortex, so we have a good idea of the total channel length. Next, the appropriate acorn reamer is inserted over the pin. This is drilled nearly to the cortex, as measured from the laser pin. If high resistance is felt before the anticipated point is reached, drilling should be stopped to avoid breaking through the opposite cortex. Generally the acorn reamer will be drilled 1 to 2 mm shorter than the laser pin length indication of where the cortex was reached. The socket will usually be about 6 mm shorter than the total channel length. It must be within 9 mm of the length of the total tunnel to allow a 15-mm EB-CL to be used. This is because at least 6 mm of extra length is needed to allow a turning radius for the EB as it sits outside the femoral cortex (i.e., 6 + 9 = 15 mm). Once the socket is drilled, the acorn drill bit is withdrawn and a 4.5-mm bit is inserted and drilled through the outer cortex with the guide pin still in place (Fig. 31-1). After the 4.5-mm tunnel is drilled, the guide pin is removed with the 4.5-mm drill bit. Next, the long-depth gauge is used to measure the total channel length (Fig. 31-2). If it exceeds the socket length by more than 9 mm, the guide pin and then the acorn drill bit should be reinserted over the guide pin, and the socket should be drilled a little farther so that it is within 6 to 9 mm of the total length. If the far wall of the cortex is accidentally violated, this can be easily salvaged with an Xtendobutton, as will be described.

Fig. 31-1 The 4.5-mm drill bit is drilled through the cortex over the previously drilled, laser-marked, long guide pin after the socket has been drilled.

Fig. 31-2 The total channel length is measured with the long-depth gauge while viewing the femoral-tunnel entry point arthroscopically.

Calculating Endobutton–Continuous Loop Length

The socket length is subtracted from the total channel length and 6 or 7 mm are added for a turning radius. If that number is a multiple of 5, then that is the EB-CL used. If it is not a multiple of 5, then the next-largest number that is a multiple of 5 is selected. For example, if the total channel is 34 mm and the socket is 27 mm, the difference would be 7 mm. 7 + 6 = 13. The next-largest number that is a multiple of 5 is 15, so a 15-mm EB-CL is selected. In this case, 34 – 15 = 19, so 19 mm of graft would be in the femoral tunnel.

Preparing the Endobutton/Graft Construct

The EB-CL is positioned in the holder on the Graftmaster board. The graft is then passed through the fabric loop attached to the EB (Fig. 31-3). A violet #3–0 or 4–0 monofilament absorbable suture is then sewn across the graft (Fig. 31-4) at a distance from the EB that is 2 mm greater than the total channel length, as identified on the Graftmaster board on which the construct is positioned. In the earlier example, the total channel length was 34 mm, so the suture would be placed at 36 mm from the EB. This will provide arthroscopic evidence that the graft has been passed to the proper depth later in the procedure. This area should also be marked with a marking pin. This absorbable colored suture also locks the graft in place, preventing it from sliding on the fabric loop. A #5 suture is then passed through one eyehole and a #2 suture through the other eyehole of the EB. All four suture ends are then passed through the eye of the long passing pin (Fig. 31-5), and a hemostat is applied to the four suture ends near their tips to lock the graft onto the pin.

Fig. 31-3 The prepared graft is passed through the fabric loop until each arm of the graft is the same length.

Fig. 31-4 The colored absorbable monofilament suture locks the graft to prevent sliding and also marks the entry point for the femoral tunnel.

Fig. 31-5 All four sutures—two #5 and two #2 sutures—are passed through the eye of the long passing pin.

Passing the Graft

The long pin with the graft and sutures attached is then passed into the tibial tunnel into the knee, where it is visualized with the arthroscope. The knee should be flexed during this passage to the same degree as when the femoral tunnel was drilled. The pin is then inserted into the femoral tunnel and passed proximalward out through a puncture in the soft tissue, where it is pulled free of the passing sutures (Fig. 31-6). Sometimes the pin must be redirected subtly within the femoral socket before it finds the smaller 4.5-mm tunnel and makes its way into the soft tissue. The two #5 sutures are then grasped with a Kocher clamp. The sutures are wound onto the closed Kocher jaws like a spool. The Kocher with wound sutures is then strongly pulled until the graft can be felt to pass into the femoral socket and become fully seated. The previously inserted violet suture and ink markings on the graft are visualized arthroscopically just outside the femoral-tunnel opening, which confirms that the graft is in proper position. If the graft is not fully seated, the Kocher can be wound like a windlass to pull in the graft. Cycling of the knee may also help.

Fig. 31-6 A,B, The long guide pin is passed into the knee and out through the soft tissue with the graft construct attached.

Seating the Endobutton

The two #2 sutures are then pulled to flatten the EB on the external femoral surface. After slack is taken up, a slight toggling should be felt (Fig. 31-7). Strong retrograde tension should be applied to the whipstitches as they dangle out the tibial incision while the knee is flexed and extended to make sure the graft cannot be pulled back out (Fig. 31-8). We then wrap the sutures around the smooth shank of the tibial screw and hold very strong tension while fully flexing and extending the knee three times. This serves to eliminate slack from the graft. Equally importantly, this will serve to pull the graft out of the knee if the EB has not been seated. For this reason, this step is very important. We have had the graft withdraw in this fashion two or three times and have successfully reseated it each time (see “Troubleshooting”). We have never had a graft pull out later. However, if this step had been omitted, later loosening could have occurred in these cases. The tibial screw or other tibial fixation method is then tightened.

Fig. 31-7 The #2 sutures are used to flatten the Endobutton, seating it on the femoral cortex after the graft has been pulled into the socket with the #5 sutures.

Fig. 31-8 Strong retrograde tension fails to dislodge the graft when the knee is cycled once the Endobutton has been properly seated.

Removing the Passing Sutures

After the EB is flattened and tibial fixation has been applied, the passing sutures should be pulled out from the EB. The #2 suture should be tried first, making sure that it slides freely (if not, see problem 3 in “Troubleshooting”). If it does slide freely, cut one limb off at skin level, and pull the other end out and discard it. This is repeated with the #5 suture.

Troubleshooting

In this section, we list all the EB problems that we have either encountered or theorized but not encountered. In all cases the difficulty was dealt with, the surgery was finished with only a minor delay, and a good result was obtained. In 13 years of continuous use we have never had an EB-related complication, EB failure, or migration. We have never had to remove an EB nor had a graft fail when fixated with an EB.⁸ Rarely, however, one of the following problems may arise. If the surgeon is prepared, these problems should pose no difficulty. The EB has proven to be remarkably trouble free over the most extended use of any soft tissue femoral fixation device.

The Xtendobutton

Recently released, the Xtendobutton is a larger button that attaches to the standard EB (Fig. 31-9), effectively enlarging its profile as described earlier. It can be used routinely to eliminate the need to drill the narrower 4.5-mm tunnel. This larger profile can be used with larger tunnels up to at least a 10-mm tunnel. We still use the smaller 4.5-mm tunnel in conjunction with the larger socket because we believe the smaller bone removal may be beneficial and because we are comfortable with this technique after 13 years of continuous use. However, it is likely that the use of the Xtendobutton will prove to be just as reliable. Xtendobutton use eliminates the calculations involved with two tunnels, which may facilitate the procedure for the occasional user. As described earlier, the Xtendobutton is of great value if the surgeon penetrates out the far femoral cortex with the larger socket reamer such that the standard EB is too small to use. Before the introduction of the Xtendobutton a second incision would have had to be made over the femur for insertion of a femoral screw. The Xtendobutton allows the procedure to be completed in the usual fashion without a second femoral incision (Fig. 31-10).

Fig. 31-9 The Xtendobutton fits over the Endobutton so that it is large enough to hold outside a 10-mm femoral tunnel.

Fig. 31-10 In this revision case, the Xtendobutton was used to anchor the graft on the outside of a 9-mm femoral tunnel.

Conclusions