Reconstruction of the anterior cruciate ligament (ACL) is a procedure that has increased in frequency in the United States, with more than 200,000 procedures performed annually. Despite the frequency with which it is performed, it is a procedure fraught with technical challenges. Graft failure and resultant recurrent instability can occur with all types of grafts. A review of board-collection data from the American Board of Orthopaedic Surgery demonstrated an overall complication rate of 9.0%. These data are reported from the first 2 years postoperatively, and early failures are usually associated with technical mistakes or failure of secure graft fixation. The incidence of long-term graft failure and recurrent instability has been reported to range from 5% to 40%. , Although the primary goal of ACL surgery is to restore anteroposterior knee stability to allow restoration of a preinjury level of activity, the graft source employed for ACL reconstruction (ACLR) can impact this goal. Irrespective of the choice of graft material, an adequate “time zero” graft strength, proper anatomical tunnels, and secure fixation within the femur and tibia are paramount to allow early mobilization and rehabilitation. Subsequently, biological incorporation of the graft with a gradual “ligamentization” of the graft tissue to allow restoration of normal knee function is necessary for clinical success. Another factor that can negatively impact clinical success is donor site morbidity, the result of harvesting an autograft, and is a complication seen with greater frequency with some graft choices.
All grafts can fail, and all can result in successful restoration of stability. The optimal graft material remains controversial. The choice of ACL graft material falls into two broad categories: autograft, which is a patient’s own tissue, and allograft, which is donated cadaveric tissue. Although other graft sources have been employed (specifically xenograft, which is a donated tissue from a different species, and man-made or manufactured graft material) these last two represent either a historical footnote or a very small percentage of ACL reconstructions, so a detailed discussion of their use is outside the scope of this chapter. Autograft options include the patellar tendon, hamstring tendon (HT), and the quadriceps tendon (QT). Allograft options are even more variable and include numerous cadaveric tissue sources and graft preparation techniques. Allografts can be subdivided into those with bone and those with only soft tissue. Allograft tissue sources include numerous tendons and ligaments typically harvested from the lower extremity. Given the plethora of graft choices, a surgeon’s knowledge of their advantages and disadvantages and of the management of graft-related complications is paramount to maximizing successful patient outcomes. These factors should be discussed with all patients before an ACL reconstruction so an educated and informed choice can be made.
Considered the “gold standard” by many surgeons, the bone-patellar tendon-bone (BPTB) autograft is the most commonly employed graft for ACL reconstruction. This autograft is harvested from the middle third of the patient’s patellar tendon and typically includes small bone blocks from the patella and the proximal tibia. Although many alternatives have been proposed, the BPTB graft is typically secured in the tibia and femur with metal or bioabsorbable interference screws that securely fixate the bone plugs within the tunnels. In addition to its high ultimate load-to-failure ratio in comparison with the native ACL (2977 N vs. 2100 N), the BPTB autograft has the advantage of osseous integration because of the presence of the bone plugs secured within the femoral and tibial tunnels, which allows for rigid fixation. , The BPTB autograft can be harvested from either the injured knee (the more common source) or, in some cases, from the contralateral knee. Freedman’s metaanalysis comparing BPTB and HT autografts demonstrated less residual anterior knee laxity in the reconstructions performed using the BPTB graft, as shown by the percentage of patients with less than 3 mm side-to-side differences in KT-1000 arthrometry testing. As such, the biomechanical stiffness of this graft makes it the most commonly used option for athletes.
However, several reported postoperative complications are important to consider in counseling patients on the selection of a BPTB graft. Kneeling pain is defined as pain produced by direct pressure on the patellar tendon. This symptom has been studied independently of generalized anterior knee pain and is detected by having the patient walk on his or her knees. Ejerhed reported kneeling pain in nearly twice as many patients who underwent ACL reconstruction with BPTB graft compared with HT autograft (53% vs. 23%). Anterior knee pain following patellar tendon harvesting is the most common postoperative complication, and occurs in up to 46% of patients. Interestingly, Shelton’s comparative study of autograft BPTB versus allograft BPTB ACL reconstructions demonstrated no difference in the incidence of anterior knee pain between the two cohorts at 15-month follow-up. Geib found the incidence of anterior knee pain to be more than 5-fold higher with BPTB autograft versus QT autograft with a single bone plug (27% vs. 5%). Most studies show a tendency for decreased knee pain with the use of HT autograft compared with BPTB as well. ,
Because the knee extensor mechanism is disrupted with BPTB autograft harvest, loss of knee extension strength has been shown to be significantly greater in patients following use of a BPTB graft, with reported deficits of 20% at 1-year follow-up and 15% at 3 years. Patella fracture and anterior knee paresthesias are other potential donor-site complications, and are discussed in a later section.
The development of osteoarthritis following ACL injury and subsequent reconstruction, although multifactorial in nature, is a known sequela with a prevalence of 21% to 48% in patients with concomitant meniscal injury. A long-term prospective follow-up study by Pinczewski et al. demonstrated a significantly higher rate of radiographically evident osteoarthritis following ACL reconstruction in patients who had a BPTB autograft in comparison with HT autograft (39% at 10 years for BPTB compared with 18% at 10 years for HT). The reason for this increased rate of arthritis is unknown.
Despite these complications and its relative contraindications in patients with preexisting anterior knee pain or work-related kneeling requirements (such as priests, builders, plumbers, mechanics, and combat soldiers), the BPTB autograft remains the second most popular graft option for ACL reconstructions worldwide.
The use of the patient’s tendons for a graft source, specifically the semitendinosus alone or in combination with the gracilis, has been coined the HT autograft . Paradoxically, the gracilis is not a hamstring but an adductor. Like the BPTB, the ultimate tensile load of a quadrupled HT autograft (2352 N) is greater than that of the native ACL. The autograft can be harvested from the injured knee or from the contralateral limb. The tendons are typically identified at their insertion of the proximal tibia and then “stripped” from their muscle attachment. They are then typically prepared by folding the long tendons in half, resulting in a four-stranded graft of 7 to 10 mm in diameter. A major advantage of the HT autograft is the preservation of the extensor mechanism, the primary focus of postoperative rehabilitation following ACL reconstruction. When compared with the nonoperated leg, HT autograft has been shown in a level 1 study to provide higher maintained knee extension strength than BPTB (92% vs. 85%). As alluded to previously, this graft option is preferable when anterior knee pain is preexisting or mitigation of the risk is desired.
As expected, multiple studies have demonstrated that, following an ACL reconstruction using a HT autograft, patients have significantly lower knee flexion strength in the operated compared with the nonoperated leg in the immediate postoperative period than those who received a BPTB autograft (90% in HT autograft vs. 102% in BPTB autograft). , Level 1 evidence has shown however, that flexion strength deficits may be transient, and may not affect an individual’s level of participation in sporting activities at long-term follow-up. , An additional theoretical disadvantage of HT graft is the longer graft integration and healing times within bone tunnels owing to the absence of bone plugs because it has been shown in a canine model to take 12 weeks for collagen fibers to form an attachment to bone that resembles Sharpey’s fibers. The clinical significance of this prolonged ligamentization process remains uncertain. Another complication associated with HT autograft is the inconsistency of HT size. Smaller tendons will result in a smaller autograft diameter and may lead to increased failure. A case-control study of 491 primary HT autograft ACL reconstructions over a 6-year period with 1.9-year follow-up found a 0.82 times lower likelihood of a patient undergoing revision surgery with every 0.5-mm incremental increase in graft diameter within the range of 7.0 to 9.0 mm. In this cohort, a patient with a graft 9 mm in diameter was 55% less likely to have undergone a revision for graft failure than a patient with a graft 7 mm diameter. Although researchers have tried to predict tendon size with preoperative imaging, it remains a challenge to accurately identify patients with smaller tendons.
A strip of QT with or without a bone block from the patella is a less commonly employed autograft source. Like BPTB autografts, QT autografts offer a high ultimate tensile load (2,352 N), as well as bone-to-bone healing when a patellar plug is harvested. Other added advantages over BPTB include reduced anterior knee pain (8.3% vs. 39%) and reduced incidence of paresthesia from cutting the infrapatellar branch of the saphenous nerve. The ability to customize graft length may be another benefit in patients where a short or long BPTB graft presents a length-tunnel mismatch. Lee reviewed 247 patients and found similar range of motion and International Knee Documentation Committee (IKDC) scores at 2-year follow-up in comparison with BPTB autografts.
However, similar to BPTB, loss of knee extension strength was reported to be 20% at 1 year. Additional disadvantages include cosmetic concerns and a more technically demanding harvest. A cadaveric study demonstrated an average postharvest QT strength of 2430 N following removal of a 10-mm wide strip of autograft tendon, which remains higher than that of a native patellar tendon. This may explain why residual quadriceps rupture is an uncommon cause of extensor mechanism disruption following ACL reconstruction. However, a single report of late donor site rupture at the vastus medialis has been reported in the literature in a patient whose knee buckled after missing three steps descending from a train ladder.
The use of any cadaveric tissue, whether harvested aseptically or processed in any way, is broadly classified as an allograft. A myriad of allograft options exists for ACL reconstruction. Some of the options with allograft bone attached include the QT, patellar tendon, and Achilles tendon. Those without allograft bone include anterior and posterior tibialis, peroneal tendon, HT, and, less commonly, the tensor fascia lata. Use of allograft tissue for ACL reconstruction confers several advantages, including shorter surgical time, a wide selection of tissue options, lower risk of postoperative pain, and most notably, no donor site morbidity as allograft use obviates the need to take the tissue from the patient undergoing the ACL reconstruction. Cost of allograft tissue may be an important consideration in certain practice settings. A retrospective review at a large academic medical center estimated that the mean total hospital cost for allograft ACL reconstruction between 2004-2005 was $5195, or 28% higher than using autografts.
Major disadvantages, as discussed in later sections, include an increased rate of graft failure, especially in the younger patients, and risk of disease transmission. Other preoperative considerations should include the potential for immune rejection and delay in graft incorporation. Jackson et al. demonstrated in a goat model that, at 6 months, allografts had persistently smaller cross-sectional area with greater numbers of large collagen fibers, indicative of remodeling delay, compared with BPTB autografts.
Several graft sterilization techniques have been proposed, ranging from antibiotic soaks to chemical sterilization and gamma-irradiation. Ethylene oxide, an industrial fumigant commonly used to sterilize medical equipment, does not alter the mechanical properties of graft tissue but may cause host tissue reactions upon transplantation. In a retrospective review of 109 patients who underwent ACL reconstruction with freeze-dried, ethylene oxide-sterilized BPTB allografts, 6.4% developed a persistent intraarticular inflammatory reaction that resolved upon removal of the allograft. The authors therefore recommended against usage of this agent. Peracetic acid, another agent which has been used to successfully sterilize bone allografts without compromising mechanical properties, has been shown to reduce graft stiffness and failure load, and is not recommended. Gamma-irradiation induces excitation of molecules and ions, resulting in crosslinking reactions that subsequently lead to pathogen destruction. High-dose irradiation (3 Mrad or more) is an unacceptable technique that alters tissue mechanics; however, lower-dose irradiation, although controversial, is sometimes used. Rappe et al. noted higher early failure rates in a cohort receiving irradiated Achilles allograft at 2.0 to 2.5 Mrad in comparison with a cohort that received nonirradiated allografts (33% vs. 2.4%) at 6-month follow-up. However, a recent level 1 randomized controlled trial (RCT) comparing irradiated with nonirradiated HT allograft found no differences in IKDC, Tegner, and Lysholm scores at final follow-up. The authors did note a significant difference in knee stability, however, citing a lower proportion of patients with side-to-side KT-2000 measurements less than 3 mm in the irradiated group (32% vs. 84%). As such, although surgeons must have a general understanding of available sterilization techniques, there remains a lack of consensus on optimal graft processing.
The advent of synthetic graft material dates back to the 1980s. Benefits of synthetic grafts include avoidance of donor morbidity, a readily available supply, and significant tensile strength. Although its introduction initially spawned enthusiasm, long-term outcome data on early synthetic designs have demonstrated high complication rates, and the practice of using synthetic grafts has largely fallen out of favor in the United States. In 1981, Dandy first implemented a carbon-fiber ACL graft, the sequelae of which included rupture and carbon deposition within the knee joint. A prospective study with 4-year follow-up using Dacron, a central core of tightly woven tapes encased by a velour sheath, demonstrated a 23% failure rate in primary ACL reconstructions and a 78% failure rate in salvage cases (revision or multiligamentous injury pattern). Similarly, Ventura found a 28% graft failure rate and a 100% rate of degenerative arthritic changes in 126 patients treated with polyethylene terephthalate (PET) synthetic ligaments at 19-year follow-up. Long-term data following usage of the Leeds-Keio ligament, which consists of tubularized polyester fibers, also show a high failure rate (28%) and increased degenerative changes compared with the unaffected knee. Other studies have demonstrated crossinfections from allogenic material, immunological response, tunnel osteolysis, and foreign body synovitis. ,
A newer generation of synthetic graft was proposed in the late 1990s, with the advent of the Ligament Advanced Reinforcement System. This graft, also made of PET fibers, contains an intraarticular component with longitudinal fibers which mimic the native ACL and also allow for tissue in-growth. A multicenter study of 159 patients with a minimum of 3 years’ follow-up demonstrated an overall complication rate of 5.7%, including three cases of graft failure and one case of knee synovitis, whereas a long-term study of 26 patients at 11.6 years demonstrated a 15.4% failure rate with no cases of infection or synovitis. , As such, it has been suggested that surgical indications do exist for synthetic grafts, although some authors advocate restricted use to patients over 40 years of age, or those needing quicker rehabilitation than is required for a biological graft or allograft, such as in the case of a once-in-a-life time sporting event. Future studies are needed to clarify treatment recommendations for usage of synthetic grafts in ACL reconstruction.
An overarching concern of surgeons and their assistants entrusted with preparing ACL grafts is contamination attributed to a dropped graft or via contact with nonsterile objects. Plante found that hamstring autografts left on the floor for 5 seconds had a contamination rate of 33% (10 of 30 specimens). According to a survey of sports medicine surgeons, 49 of 196 surgeons surveyed (25%) reported experience with at least one episode of intraoperative graft contamination. The two options that exist are to attempt to clean the graft or to choose another graft option. Of the 57 total reported events, the surgeon cleansed the graft and proceeded with the original procedure in 43 instances (75%), whereas an alternative autograft or allograft was used in the remainder of cases.
The three main cleansing agents that have been studied in graft decontamination are chlorhexidine, povidone-iodine, and antibiotic washes. Chlorhexidine is a bactericidal and fungicidal agent that damages bacterial membranes. Povidone-iodine uses iodine particles to kill microorganisms. Molina washed contaminated ACL grafts harvested from primary total knee arthroplasties using povidone-iodine, chlorhexidine, or a combination antibiotic wash (neomycin/polymyxin B) for 90 seconds and demonstrated the increased efficacy of chlorhexidine. For each group of 50 specimens, only one positive culture was reported in those treated with chlorhexidine, compared with three samples in the antibiotic wash group and 12 samples in the povidone-iodine wash group. Although case reports of arthroscopic lavage using low-dose chlorhexidine have documented damage to joint surfaces and reactive synovitis, no reports exist for grafts cleansed with chlorhexidine before implantation. Fig. 5.1 summarizes steps for graft sterilization in the event of contamination, should the surgeon choose to proceed with using the graft.