Introduction

This chapter will describe a technique that provides the highest strength known graft for use in revision and high-risk primary anterior cruciate ligament reconstruction (ACLR), while allowing flexibility for femoral tunnel creation in the revision setting. Autograft has consistently shown better results than allograft and should be used whenever possible for ACL revision. Hamstring (HS), bone–patellar tendon–bone (BPTB), and quadriceps tendon (QT) are the only three autografts with proven reported efficacy. We prefer HS graft partially because the utilization of both the semitendinosus (ST) and gracilis (Gr) allows a much stronger graft to be fashioned than is possible with either the patellar tendon or QT grafts, which are limited in size to avoid rupture of the remaining patellar and QTs. The traditional 2ST/2Gr graft is already roughly 50% stronger than a 10-mm BPTB or a QT graft, and can be further strengthened by augmenting it with the extra harvested tendon that is almost always available but is usually discarded because of lack of need.

Graft Configurations

The ST length to the musculotendinous junction is generally 23–27 cm. The graft harvest extends to this length and usually far beyond. However, the amount of ST needed for graft use is only about 16 cm for a double-length graft, and much less can also be used. Thus the typical harvested graft can easily contribute a third strand—that is, a 24-cm triple ST, or even as little as a 21-cm triple strand. In many patients, the ST harvest is long enough to fashion a four-strand graft. Most surgeons, including the author, feel comfortable with a 30-cm graft for a quadruple ST graft. However, the primary inventor and proponent of quadruple ST used grafts that were significantly shorter. Regarding the Gr, the numbers are similar. The harvest is usually long enough to fashion a third strand. While length of the harvest may be enough for a fourth strand, the thinness of the Gr to this length renders a fourth strand that is difficult to suture and of minimal additional value in augmenting strength. However, a third strand is useful for revisions and for high-risk primaries. Possible graft configurations in ascending order of strength are a six-strand 3ST/3Gr, six-strand 4ST/2Gr, or even seven-strand (4ST/3Gr) graft. As the grafts get stronger, their length may decrease, with the limiting factor being having enough graft to fill the tendon tunnels. Thus the surgeon must be sure that he or she is comfortable with the amount of graft available for tunnels. A five-strand 3ST/2Gr graft is an excellent combination of graft strength and length, and has had good clinical success. However, if the harvest is sufficiently long, an even stronger 4ST can be safely used.

Graft Configurations

The ST length to the musculotendinous junction is generally 23–27 cm. The graft harvest extends to this length and usually far beyond. However, the amount of ST needed for graft use is only about 16 cm for a double-length graft, and much less can also be used. Thus the typical harvested graft can easily contribute a third strand—that is, a 24-cm triple ST, or even as little as a 21-cm triple strand. In many patients, the ST harvest is long enough to fashion a four-strand graft. Most surgeons, including the author, feel comfortable with a 30-cm graft for a quadruple ST graft. However, the primary inventor and proponent of quadruple ST used grafts that were significantly shorter. Regarding the Gr, the numbers are similar. The harvest is usually long enough to fashion a third strand. While length of the harvest may be enough for a fourth strand, the thinness of the Gr to this length renders a fourth strand that is difficult to suture and of minimal additional value in augmenting strength. However, a third strand is useful for revisions and for high-risk primaries. Possible graft configurations in ascending order of strength are a six-strand 3ST/3Gr, six-strand 4ST/2Gr, or even seven-strand (4ST/3Gr) graft. As the grafts get stronger, their length may decrease, with the limiting factor being having enough graft to fill the tendon tunnels. Thus the surgeon must be sure that he or she is comfortable with the amount of graft available for tunnels. A five-strand 3ST/2Gr graft is an excellent combination of graft strength and length, and has had good clinical success. However, if the harvest is sufficiently long, an even stronger 4ST can be safely used.

Tibial Tunnel

Two potentially negative consequences of a larger area graft are a smaller relative surface area of graft contact in the bony tunnel for healing, and a wider tibial tunnel that may produce a width of graft that impinges in the femoral notch. Both of these consequences are related to the custom of creating a circular tibial tunnel. This is done for reasons of mechanical ease. However, the true anatomic tibial footprint is not a circle but rather an oval. The prevalent double-bundle techniques, which feature two tunnels, are really using two tunnels as the easiest way to approximate a single anatomic oval. In the femur, in particular, there is no easy way to create an oval tunnel. However, in the tibia, this can easily be done by downsizing the drill size 2 mm from the measured size and then toggling the drill tip after it emerges through the tibial eminence a distance of 50% of the drill bit width. This produces an oval inlet of roughly the same total area as a circular tunnel of the measured graft size would be. It is important to begin the tunnel with a larger drill bit, the same as the measured size of the graft, to allow the graft to enter the external side of the tunnel.

This oval tunnel has two advantages: First, the oval shape presents greater surface area for graft healing than a circular tunnel of the same area. Second, the graft is 2 mm, or more than 20%, narrower than a circular tunnel would be, preventing it from potentially impinging in the notch. It should be noted that an oval tunnel precludes use of interference-fit intratunnel fixation. Therefore external tibial cortical friction fixation, such as with a whipstitch-post construct, must be used.

Femoral Tunnel

Unfortunately it is far more difficult to create an oval femoral tunnel. To do so would necessitate drilling a tunnel from outside-in, and most femoral tunnel techniques and systems are designed to be drilled inside-out. Second, if an outside-in system were used, it would tend to compromise the button fixation on the femur, currently the most popular method of HS fixation, because of unpredictable enlargement of the external cortical border of the tunnel where button fixation takes place. If such a system were used, it would require a larger button than the ones most commonly used. Since an oval femoral tunnel is impractical, in general the best answer is two close or (ideally) overlapping femoral tunnels placed within the femoral anatomic footprint of the ACL. To keep these tunnels reliably within the footprint, any wall between them will need to be quite small, and ideally there will be no wall at all. Thus interference fixation should also not be used there.

Ideally, the tunnels should also not be parallel. If they were close together or overlapping in the notch and they were parallel, then the outlet on the femur would produce an opening that would be potentially too large to allow an adequate cortical buttress to provide reliable cortical button fixation. We avoid parallelism by drilling the anteromedial (AM) tunnel transtibially and the posterolateral (PL) tunnel outside-in. This double-bundle femoral technique achieves the twin goal of allowing a much greater surface area for healing than a single tunnel, especially important for a large graft, and also allows the graft to stay nestled proximally in the notch. A large circular graft will extend distally in the notch, potentially outside the footprint of the ACL femoral attachment. These more distal fibers would be subjected to excessive tension forces in high flexion and possibly stretch out, thus negating their desired beneficial effect of improving graft tensile strength.

Results

We used this technique on a consecutive series of 126 ACL-reconstructed patients, both primary and revision. We had excellent results with no graft failures. KT-1000 and rating results were indistinguishable from our previously reported results using single-bundle four-strand HS graft. However, given the excellent results using the single-bundle technique, any improvement would have been difficult to detect. We were happy to find no adverse effect of the stronger graft, and no significant complications of any kind.

Conclusion

The double-bundle technique, with a five-strand or greater HS graft and ovalized tibial tunnel described here, has proven to be a safe, highly effective technique. It requires 15–30 minutes more surgical time than a traditional single-bundle 4HS technique, with similar surgical outcomes. For this reason we do not think it is routinely warranted for primary ACLR. However, for revision ACLR; for patients such as ski racers or larger patients, who are at high risk of ACL rupture; and for patients with HS tendons of small girth, we think it is a safe, highly effective procedure.

Perhaps most importantly, there does not need to be any learning curve regarding patient outcome with this procedure, although there will be a modest one for length of procedure. Since failures due to technical difficulty have plagued double-bundle ACLR procedures, we feel this is a significant advantage. One reason for the lack of difficulty is that no hyperflexion of the knee is required. The only unfamiliar part of the technique for most surgeons who perform ACLR is the use of the outside-in guide for creation of the PL tunnel. However, this is a simple technique that should require no learning curve for any experienced ACL surgeon.

Another advantage of this technique is that it allows smaller individual femoral tunnels to be used rather than one large tunnel. Half of our revision cases require autologous bone grafting due to prior mildly misplaced femoral tunnels or because of tunnel widening. In some cases we are able to fit two smaller tunnels within the anatomic footprint and thus execute a revision procedure in one stage without bone grafting, which would have required two stages with bone grafting otherwise.