In this chapter, the potential complications during the harvest of hamstring, bone-patellar tendon-bone (BPTB), and quadriceps tendons for anterior cruciate ligament reconstruction (ACLR) will be discussed. Regardless of which graft is harvested and its potential complication, it is essential that the surgeon discuss these issues preoperatively with the patient.
Hamstring Tendon Harvest
Hamstring tendon autograft is one of the most common grafts used around the world for ACLR. As with all grafts, there are some inherent risks and complications with harvest of the hamstring tendons.
Sensory Nerve Injury
The reported nerve injury rate during hamstring tendon harvest ranges from 22% to 74%. Although injury to the infrapatellar branch of the saphenous nerve (IPBSN) is more commonly discussed, injury to the sartorial branch of the saphenous nerve (SBSN) is also of concern, with the sensory distribution shown in Fig. 6.1 . This SBSN is the terminal branch of the saphenous nerve and exits the sartorius fascia at a mean of 7.2 cm from the distal gracilis insertion to become subcutaneous. This nerve is in close association with the gracilis for 4.6 cm before exiting layer 1 of the knee upon cadaveric dissection. Clinically, this translates to the SBSN being in close proximity to the gracilis tendon in the distal thigh and put this structure at risk during tendon stripper passage proximally. Sanders and colleagues revealed that of the 74% of instances in which patients experienced postoperative sensory nerve disturbance, 32% of these injured both the sartorial and infrapatellar branch, 23% injured only the sartorial branch, and 19% injured the infrapatellar branch.
Modifications to the surgical approach have been investigated to potentially decrease nerve injury. Incisions that are in line with the distal hamstring tendons have been investigated in an effort to potentially decrease an iatrogenic injury to the IPBSN. These modifications include anteromedial oblique and horizontally oriented incisions along the tendons. A metaanalysis was performed that evaluated studies that compared nerve injury between vertical, oblique, and horizontal incisions during hamstring tendon harvest. The risk of iatrogenic injury to the IPBSN was significantly greater when a vertical incision was used compared with an oblique or horizontal incision. Although modifying the orientation of the skin incision may decrease the incidence of IPBSN injury, it does not change the more significant injury to the sartorial branch that is injured proximally during the harvest. This is owing to the SBSN being intimately involved with the gracilis tendon for a portion of its course in the distal thigh and damage occurring from the passage of the tendon stripper proximally during harvest. Overall, it is difficult to avoid injury to the IPBSN during the initial approach to the hamstring tendons, regardless of incision orientation when approached from the traditional anteromedial approach.
Use of a posterior incision, as seen in Fig. 6.2 , has been described, and may decrease injury to both branches of the saphenous nerve. , First, it avoids the anteromedial incision which puts the IPBSN at risk. Second, it allows for the tendons to be harvested distally in line with the SBSN, which may protect the nerve. However, there is a learning curve for surgeons unfamiliar with this posterior technique. To help increase initial surgeon comfort with this approach, a second small anteromedial incision perpendicular to the tendons can be made by placing an index finger in the posterior incision, following the tendons distally, and tenting the skin to mark the incision site anteromedially. As for the orientation of the anatomy for surgeons unfamiliar with this technique, the semitendinosus is typically deeper and lateral to the gracilis. Also, one can trace these tendons distally, and the semitendinosus has a more distal insertion on the tibia compared with gracilis. Letartre and associates demonstrated no postoperative sensory deficits in 90 patients after using the posterior mini incision technique. In this same study, the average length of tendon obtained was 270 mm, which is more than adequate for the variety of anterior cruciate ligament (ACL) techniques and fixation options.
The authors of this chapter will use the posterior incision to help combat the issue of iatrogenic sensory nerve injury for the reasons described above. Ultimately, this was also one of the factors in deciding to regularly use an alternative graft choice.
Residual Hamstring Weakness
Hamstring weakness after harvest is a concern in all athletes. Furthermore, there is concern that further weakening of the hamstrings, especially in female athletes, could potentially lead to an increased ACL rerupture rate. Kriellaars et al. measured the knee flexion strength postoperatively in patients who had had hamstring and patellar tendon harvests compared with control groups. They demonstrated that the knee flexors in patients who underwent hamstring harvest for ACLR had significant deficits compared with the BPTB group and control subjects. This specific study found flexion strength deficits of up to 50% in the hamstring tendon group compared with control groups at over 1 year postoperatively. In addition to flexion deficits, significant tibial internal rotation deficits at the knee have been demonstrated when both hamstring tendons are harvested, compared with the semitendinosus only. Kaeding et al. found that hamstring strength asymmetry was common 3 years after ACLR with hamstring tendon autograft. Some 50% of their cohort had greater than 15% deficit compared with the unaffected limb nearly 3 years after surgery. They also demonstrated decreased tibial internal rotation during weight acceptance and increased tibial external rotation during jogging at initial contact and during weight acceptance compared with the control group. They ultimately demonstrated altered knee mechanics during normal gait and jogging compared with patients with symmetric hamstring strength at 3 years after surgery. In comparison, MacDonald et al. found knee flexion deficits at 3 months after hamstring tendon autograft, but this difference did not remain beyond longer-term follow-up. Overall, the literature is inconclusive with regard to the clinical effect of possible knee flexion deficits after hamstring tendon autograft, but it is still an important factor for the surgeon to consider when selecting appropriate graft choice for athletes, as well as their specific sport.
With respect to harvesting complications, a possible remedy for this issue is to harvest one tendon as opposed to both. Possible graft configurations with one hamstring tendon are discussed in the small graft diameter section later. Newer anatomic reconstruction techniques such as the all-inside ACL, which requires less graft length, have allowed surgeons to harvest only one tendon to help attempt to decrease these postoperative strength deficits. If residual hamstring weakness is of particular concern in an athlete, alternative grafts can be used, such as BPTB or quadriceps tendon grafts.
Premature Graft Transection
Premature graft transection can be a serious complication. The first step to safe harvest is clear identification of the semitendinosus and gracilis. Typically the sartorial fascia can be incised and reflected. It is easiest to identify the tendons on the backside of the sartorial fascia when reflected. The tendons also typically blend together distally near the tibial attachment, so it is easier to identify separate tendons more proximally. A right-angle clamp or equivalent can be used to separate and isolate the tendons once identified after blunt dissection proximally between the tendons and tibia. It is important to note that the medial collateral ligament inserts under the hamstring tendons on the tibia, so care must be taken to not damage this structure. The right-angle clamp or equivalent can also be used to free up the tendons from the overlying sartorial fascia, once identified.
The most common reason for premature transection of hamstring tendons is failure to release the fascial band attachments of the semitendinosus. These fascial bands have also been referred to as an accessory insertion of the semitendinosus tendon. Warren et al. found this accessory insertion present in 77% of the cadaver knees they dissected. This accessory tendon attaches approximately 8 to 10 cm proximal to the insertion site on the tibia. It is important to free up this amount of tendon proximal to this accessory tendon’s insertion to help avoid premature transection. A helpful technique to make sure that the fascial bands are released is to pull on the hamstring tendons once freed and whipstitched. If the posterior calf muscles do not move or “bounce” while pulling on the hamstrings and the hamstring can be pulled laterally over the tibial tubercle, then the fascial bands have been adequately released.
If the gracilis tendon is cut prematurely but still has approximately 10 cm of length intact, this single strand could be sutured to the doubled semitendinosus graft. If the semitendinosus, the more robust tendon, is cut prematurely, an alternative graft is typically needed, because obtaining a graft diameter of at least 8 mm is unlikely.
Using the posterior approach described previously is also advantageous with regard to premature graft amputation. Prodromos reported no occurrences of premature amputation of the hamstring tendons in a series of 203 patients. As for the semitendinosus reflection to the gastrocnemius with this posterior approach, this typically can be directly identified through the posterior approach because the incision is typically centered over this anatomy. If using two small incisions, it is recommended that the tendon stripper be started at the tendon insertions, then brought to the level of the posterior incision, and that the semitendinosus reflection be clearly identified at the end of the stripper where it falls into the “axilla” of the accessory tendon and can be easily transected with a 15 blade.
Small Diameter Graft
The quadrupled hamstring graft has been used historically for ACLR. If these grafts have a diameter of less than 8 mm, increased rates of graft failure are seen compared with larger diameter grafts. A major downside of using hamstring grafts is that there is no easy and reliable method to preoperatively determine potential graft diameter. Thus multiple techniques have been described to help ensure adequate graft diameter is obtained.
Vinagre and associates described various techniques to prepare hamstring tendon autografts. This ranged from a 2-strand construct with one tendon up to an 8-strand graft with two tendons. These techniques can be complicated, and it is often difficult to obtain uniform tension in all of the limbs. In contrast, the “graft-link” construct is a reproducible graft preparation technique where the semitendinosus is quadrupled. This technique has the advantage of enabling uniform loading of the construct, as well as using only the semitendinosus tendon. Leaving the gracilis tendon intact can help with postoperative hamstring weakness. However, if the graft diameter is less than 8 mm, the remaining gracilis tendon can be used to augment the graft. Because of the shorter graft length with some of these techniques, they are better used with all-inside ACL techniques with suspensory fixation.
Alternatively, the hamstring autograft can be augmented with allograft tissue. Johnson and colleagues evaluated allograft augmentation in younger patients. They revealed that the graft diameter was 7.8 mm in the nonaugmented group versus 9.9 mm in the augmented group. The failure rate in the nonaugmented group was 28.3% compared with only 11.9% in the allograft augmented group. If this technique is used, the authors recommended placing the autograft tissue around the allograft tissue so that the autograft tissue is preferentially exposed to the joint and bony tunnels.
Bone-Patellar Tendon-Bone Harvest
The BPTB was the most commonly used graft before 2000. Its primary advantage is that bone-to-bone healing of the graft with rigid fixation can be achieved in each of the bone tunnels. Secondarily, its graft length can be easily determined preoperatively using both x-ray and magnetic resonance imaging. However, there are complications and disadvantages innate to this graft that will be discussed later.
Sensory Nerve Injury
Postoperative numbness over the anterior knee is a common complaint, with an incidence ranging from 20% to 60% caused by injury of the IPBSN. , Numbness was found following the use of both vertical and horizontal harvest incisions. Some 59% of patients who had a vertical incision were found to have postoperative numbness, compared with 43% of patients with a horizontal incision. In this particular study, the horizontal incision was made at the junction of the middle and distal thirds of the patellar tendon. Tsuda found that only 17% of 75 patients who had two transverse incisions centered over the inferior pole of the patella and superior border of the tibial tubercle had postoperative numbness. Lastly, a cadaver study was performed to investigate if the position of the knee influenced nerve injury during exposure of the patellar tendon. It was found that that the distance of the IPBSN from the typical skin incision was greater with the knee in flexion, which may help to decrease nerve injury.
Anterior Knee Pain
Postoperative anterior knee pain is a significant issue and can limit full recovery after ACL surgery. This can be defined as pain over the front of the knee at rest or activity, as well as when kneeling or squatting. The incidence of anterior knee pain after BPTB harvest is as high as 4% to 60% postoperatively. Corry et al. reported that at 2 years after surgery 31% of BPTB autograft patients had pain when kneeling compared with 6% of those who were reconstructed with hamstring tendon autografts.
The cause of the frontal knee pain is multifactorial, and includes bone harvest, injury to the IPBSN, reduced strength, and loss of motion. The most obvious potential cause is operative site morbidity involved with removing bone plugs from the tibia and femur. Surgical techniques such as bone grafting of the donor site and closure of the patellar tendon defect have been evaluated to potentially decrease local discomfort in the donor site region. In a prospective randomized study, Brandsson et al. revealed that suturing of the patellar tendon defect and bone grafting of the defect in the patella did not reduce anterior knee problems or donor site morbidity. A second study by Boszotta and Prunner reported similar results: bone grafting of the patellar defect did not reduce kneeling complaints or patellofemoral problems. To the contrary, Tsuda found that bone grafting of the harvest sites decreased the incidence of frontal knee pain to 17% in comparison to the previously quoted incidence of 40% to 60%. We recommend bone grafting of the defects with bone from the plugs to not only potentially help with frontal knee pain, but also with palpable defects, in addition to gentle reapproximation of the superficial patellar tendon, as discussed later.
Another intraoperative factor for anterior knee pain is damage to the IPBSN. Several studies have correlated sensory disturbance on the anterior knee to painful kneeling. The techniques described earlier can be used to help avoid injury to avoid anterior knee pain.
Patellar fracture is a serious complication of BPTB harvest and has a reported rate of 0.23% to 2.3%. Patellar fractures can occur both intraoperatively and postoperatively. Intraoperative fractures typically are oriented longitudinally and can be caused by imprecise bone cuts, dull saw blades, extension of the bone plug harvest proximal to the mid patella, or overly aggressive levering of the patellar bone plug out of the donor site. Postoperatively, patellar fractures occurred on average 57 days after surgery ( Fig. 6.3 ). Risk factors for postoperative patella fractures include squared patellar cuts and large bone plug harvests either proximal to the equator of the patella or deeper than one-third of the patellar thickness. Thus to avoid either intraoperative or postoperative patella fracture, the following tips should be followed:
Use a sharp saw blade.
Avoid harvesting a bone plug greater than one-half of the length of the patella and deeper than one-third of the thickness, which is typically 10 mm. Using a saw blade that is marked at 10 mm can help guide the depth of the cut. If a curved osteotome is used to free up the bone plug after the initial cuts are made, it is possible to fracture the patella or produce an undersized bone plug if the deep surface is not well defined. An alternative approach is to undercut the bone plug at the distal pole of the patella using the oscillating saw that will set the depth of the bone plug and avoid a near full -thickness bone plug. To do this, the tibial bone plug must be harvested first and then reflected superiorly after the initial patellar cuts to visualize this undercut.
Angle 30 degrees toward the center of the bone plug with the longitudinal cuts.
Predrill the superior horizontal border of the bone plug with multiple small unicortical holes. Then use a curved osteotome to complete the cut, as this avoids a stress riser being formed from excursion of the blade beyond the horizontal border of the bone plug.