Introduction

Anterior cruciate ligament reconstruction (ACLR) autograft options include bone-patellar tendon-bone (BTB), quadriceps tendon (QT), and hamstrings tendon (HT). The gold standard autograft for ACLR is widely debated. In 1992, over 90% of ACLR autografts were BTB but by 2021, HT became the autograft of choice and was used in 50% of cases [1]. Despite the changing landscape, the clinical outcomes after ACLR with BTB or HT autograft are equivalent [2,3,4]. Benefits of HT include a minimally invasive harvest, reduced risk of knee stiffness, lower post-operative pain, applicability to paediatric ACLR, and avoidance of anterior knee pain and kneeling pain associated with donor-site morbidity [5].

The anterior cruciate ligament (ACL) is made up of the anteromedial and posterolateral bundles, which both contribute to rotational stability [6]. An anatomic double bundle (DB) reconstruction technique was previously popularized to recreate the native anatomy. However, multiple studies have failed to show a clinical improvement or decrease in re-injury rate after DB compared to a single bundle (SB) ACLR [6].

There is huge variation in the graft configuration, tunnel formation, and fixation techniques for SB ACLR. Anatomic tunnel positions are favoured to minimize the rate of re-rupture [7]. A more anatomic, oval femoral tunnel may improve clinical outcomes compared to a round femoral tunnel [8]. Fixation technique is intimately related to the graft configuration. The decision to use aperture or suspensory fixation for tibial and femoral fixation respectively should complement the graft configuration to maximize graft diameter while maintaining an adequate bone-graft interface and minimizing potential complications. The graft configuration chosen depends on the number of tendons harvested, the tendon length, the number of graft strands, and has a direct effect on the subsequent graft diameter and the graft length.

A common pitfall of HT is a small graft diameter, often less than 8 mm (mm), which is associated with a higher re-rupture rate and worse functional outcomes compared to a graft diameter greater than 8 mm [9]. Several studies have attempted to predict HT graft size pre-operatively based on anthropometric data and have found a consistent correlation between height, weight, sex, and graft size [10]. In general, taller, heavier males have larger, longer hamstrings tendons [10]. However, anthropometric measures only provide an estimate, may not be universal across ethnicities and do not account for for single bundle hamstrings autograft preparation techniques in ACLR. Variability in graft harvest technique. The semitendinosus (ST) tendon is always used in a conventional HT autograft, and the Gracilis (GC) tendon is also often harvested with the benefit of adding strands and increasing graft diameter. However, an advantage of an ST-only autograft may be greater knee flexion strength, specifically at greater than 70° of flexion compared to a combined ST and GC autograft ACLR [11]. Accordingly, the ideal HT technique should minimize donor site morbidity while achieving a graft diameter greater than 8 mm. The objective of this current concepts review is to describe the indications, surgical anatomy, technique, intraoperative tips, clinical outcomes, and complications

Indications and contraindications

General indications for ACLR include ACL insufficiency and the desire to return to pivoting activities. HT autograft may be favoured in those who commonly kneel as a part of their occupation, cultural practice, or hobbies such as gardening. Contraindications to an ACLR include active infection, advanced arthritis, and partial ACL tears with no instability or clinical laxity. A contraindication to using HT for an ACLR is HT deficiency due to previous harvest or injury. A BTB or QT autograft may be favoured in context of a combined ACL and medial collateral ligament (MCL) knee injury due to the dynamic valgus stability that the HT provide [12].

With respect to graft configuration, intra-operative indications for a 4-strand ST-only configuration include a tendon harvest of 26 cm (cm) and graft diameter of 8 mm or more. Indications for an 8-strand combined ST and GC configuration include tendon harvests of 26 cm and 4-strand ST-only graft diameter of less than 8 mm. Additionally, adjustable-loop suture implants must be available for both femoral and tibial suspensory fixation. Necessary tendon harvest lengths for a 5-strand combined ST and GC configuration include an ST harvest of at least 24 cm and a GC harvest of at least 16 cm. Either aperture or suspensory fixation may be utilized for a 5-strand combined ST and GC graft. Ultimately, surgeon preference and comfort with different configurations and fixation techniques are also important and play a large part in ensuring a successful procedure and outcome.

Surgical technique

Number of tendons

A single 26 cm ST tendon harvest is necessary to procure an 8 mm diameter, 6.5 cm long graft utilizing an adjustable suspensory fixation technique. A minimum graft length of 6 cm achieves 2 cm of bony contact in both the femoral and tibial tunnels given an intra-articular graft length of 2 cm. If the ST tendon harvest is less than 26 cm, the GC tendon should be harvested in preparation for a two-tendon graft (Fig. 1) (Supplementary Digital Content 1).

Fig. 1
figure 1

Graft configuration decision-making algorithm to maximize graft diameter. ST, semitendinosus; cm, centimeters; GC, gracilis; mm, millimeters

Full size image

Number of strands

In the last 5 years, 20 techniques describing different graft configurations ranging from 3 to 8 strands have been published. Single tendon grafts produce up to 4 strands while combined ST and GC grafts produce up to 8 strands. The number of strands invariably depends on the number of tendons harvested, the harvested tendon length and the tendon quality. Additionally, the femoral and tibial fixation techniques must be considered simultaneously as they impact the graft length needed. Figure 1 provides an algorithm for graft configuration based on the number of tendons harvested and the length of the harvested tendons (Fig. 1).

Suturing technique

The suturing technique used to secure the graft configuration plays a small, yet potentially important role in overall graft strength. While no study has assessed for a difference in clinical outcomes based on graft suturing technique alone, biomechanical studies provide some insight. The technique chosen plays a larger role in graft configurations that require strand connection by suturing at both ends compared to one end of the graft [13]. A conventional, buried Figure 4 knot has inferior strength compared to a ripstop, modified Mason-Allen knot in a bovine extensor tendon biomechanical model [13]. A supplementary uniform graft whipstitch with absorbable suture increases immediate graft load to failure, but this finding is limited to an in vitro setting and may not translate to long-term clinical outcomes [14]. Last, in the setting of suspensory fixation, a Polyethylene suture is superior to braided polyester or Mersilene tape to connect the graft and button due to its superior load to failure, elongation, and stiffness properties [15].

Fixation techniques

Suspensory and aperture fixation are both acceptable options during ACLR. Aperture fixation achieved with an interference screw is technically simpler to carry out due to the constant diameter of the tunnel and provides a longer bone-graft interface for graft incorporation. However, this technique requires a longer graft to obtain press-fit fixation at the outer cortex of the tibial tunnel. Complications are uncommon and include intra-articular screw penetration, tunnel osteolysis, tunnel enlargement, late screw fracture and screw migration [16]. In contrast, suspensory fixation, such as a button, requires a shorter tibial tunnel socket but is more challenging secondary to the maintenance of a closed tunnel end to facilitate button fixation on the outer cortex. A tibial socket is beneficial because it allows for a shorter graft with more strands and a greater diameter. However, a drawback is that there is also less bone-graft interface for graft incorporation. Complications are uncommon, apply to non-adjustable and adjustable devices and include tunnel widening due to side-to-side or longitudinal graft motion termed windshield-wiper and bungee cord effects respectively, and intra- and extra-articular button migration [16]. A pitfall specific to some adjustable suspensory devices is lateral femoral outer cortical breach during graft passage if the femoral tunnel is drilled distally in metaphyseal bone due to significant force being placed on the cortex during tensioning [17].

Configuration descriptions

  • A. Long ST tendon harvest, all-suspensory fixation

A 26 cm ST tendon harvest or longer is necessary to produce a 4-strand ST-only graft. An 8-strand combined ST and GC graft uses the same technique after the ST and GC tendons have been sutured together at both ends and is detailed in Fig. 2A. A graft preparation board is used to prepare an adjustable button-loop suspensory system on both the femoral and tibial sides. The harvested tendon is doubled over the adjustable loop on the tibial side (Fig. 2B), then each half is folded over the adjustable loop on the femoral side to produce 4 strands (Fig. 2C). The preliminary graft diameter is sized, and if it is less than 8 mm, the GC tendon is also harvested (Fig. 1). If the 4-strand ST graft is greater than 8 mm alone, the ends of the tendon are buried within the initial tendon fold and sutured into place. If the GC tendon harvested is 24 cm or longer, it is quadrupled in the same fashion as the ST, with cerclage sutures combining both tendons to create a two-tendon, 8-strand graft (Fig. 2D). If the GC tendon is 20–24 cm long, it is tripled to produce a 7-strand graft. If the GC tendon is 16–20 cm long, it is doubled to produce a 6-strand graft.

  • B. Short ST tendon harvest, mixed suspensory and aperture fixation

Fig. 2
figure 2

a When both the ST and GC tendons measure 26 cm or longer, they are cut to equal lengths and sutured together at the ends in preparation for an 8-strand graft. b The combined GC, and ST are halved lengthwise by passing them through the femoral adjustable-loop implant. c Both ends of the combined ST and GC are then passed through the tibial adjustable-loop implant far enough to produce 4 equal strands, each with two limbs of tendon. d The tendon ends are secured at the femoral end and the graft is reinforced with non-absorbable, circumferential cerclage sutures to solidify the 8-strand graft

Full size image

If the initial ST harvest is less than 26 cm, the GC is immediately harvested. The ST and GC tendon harvest lengths dictate the resulting graft configuration, number of strands, and femoral and tibial fixation strategies (Fig. 1).

To triple a tendon, fold it over an adjustable loop at one third of its total length (Fig. 3A), then fold the long end of the tendon over a second adjustable loop at half of the distance of the remaining length (Fig. 3B). If using aperture fixation at one end, instead of folding the tendon over an adjustable loop initially, fold it over a non-absorbable suture, which will function as the distal end to provide graft control while securing the graft with an interference screw. If doubling a tendon, fold it over an adjustable loop, or non-absorbable suture, depending on the chosen tibial fixation technique, at half of the tendon length and suture the tendon ends together (Fig. 3C). Circumferential cerclage sutures may be added at both graft ends regardless of graft configuration and fixation techniques to consolidate the two-tendon graft (Fig. 2D).

Fig. 3
figure 3

a Two thirds of the ST tendon is passed through the femoral adjustable-loop implant. b The long end of the ST tendon is halved lengthwise by passing it through the tibial adjustable-loop implant. c The end of the ST tendon is sutured to the femoral loop. The GC tendon is the halved lengthwise by passing it through the femoral adjustable-loop implant, superficial to the ST tendon loop. d The suture limbs from the GC ends are added to the tibial adjustable-loop implant to create a 5-strand graft

Full size image

Graft fixation

If a suspensory fixation strategy is chosen, femoral and tibial sockets are drilled. The femoral tunnel length is marked of the femoral side of the graft to confirm seating of the graft in the femoral socket. The femoral shuttle suture is looped through the femoral passing suture to introduce the adjustable button-loop system into the knee through the anteromedial portal. The button is flipped on the lateral femoral cortex and the loop is shortened to bring the graft into the knee joint and seat it in the femoral socket. The tibial adjustable button-loop system is passed through the tibial tunnel using a tibial passing suture. A stand-alone button is used to fix the adjustable button-loop system on the tibial side by passing the button through the loop and shortening the loop in knee extension with re-tensioning after cycling the knee through a flexion–extension arc.

If an aperture fixation strategy is chosen, a femoral socket and open tibial tunnel may be drilled. The graft is passed into the knee from the tibial tunnel into the femoral socket using the femoral shuttle suture. Under arthroscopic visualization an interference screw is placed at the femoral aperture after the graft is seated. The tibial fixation is completed with an interference screw placed at the outer tibial cortex under direct open visualization with the graft under tension in knee extension.

Configuration technique review

Methods

A review of the most recent SB HT ACLR autograft techniques was completed to identify configurations the options available based on the number and length of hamstrings tendons harvested. Three online databases (Embase, PubMed, and MEDLINE) were searched from database inception until April 10, 2021, updated September 24, 2022, and May 2, 2024, for literature that investigated SB HT ACLR autograft configuration techniques by two orthopaedic surgeon reviewers independently (X.X and X.X). The search included broad terms such as “anterior cruciate ligament reconstruction”, “autograft”, and “hamstrings” (Supplementary Material: Appendix 1) and was completed using the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. Any disagreements were deliberated between the two reviewers, and a senior author (X.X) was consulted if a consensus could not be obtained. The research question and study eligibility criteria were established a priori. The inclusion criteria were English language studies that detailed the autograft configuration technique of a SB HT ACLR in humans. Exclusion criteria included multi-ligament knee reconstruction, quadriceps tendon autograft, patellar tendon autograft, allograft reconstruction, double or triple bundle reconstruction, ACL augmentation, non-human studies, and no explicit graft preparation technique. The 20 most recent SB HT ACLR graft configuration surgical technique publications were included (Fig. 4). Data were collected and recorded in an Excel spreadsheet (Version 16.84, Microsoft Corporation). Abstracted data included the manuscript title, author(s), year of publication, study design, sample size (if any), minimum necessary tendon length, average graft length, and graft diameter, and is summarized in Table 1 [9, 18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36].

Fig. 4
figure 4

Preferred reporting items for systematic reviews and meta-analyses flow diagram

Full size image
Table 1 Graft characteristics
Full size table

Results

Nine separate graft configurations have been reported from a three-strand, single tendon graft to an eight-strand, two-tendon graft using a variety of aperture or suspensory femoral and tibial fixation methods. Five decision-making algorithms for graft configuration have been described [19, 27, 31, 32, 36]. Thirteen studies compared graft configurations in addition to detailing the graft technique [9, 18, 20,21,22,23, 25, 26, 29, 30, 32, 33, 35]. Overall, 1299 patients were included in multiple different comparisons of graft configurations and fixation techniques.

Tendon and graft parameters

Of the 20 publications reviewed, 8 described graft configurations using a single tendon [21,22,23, 27, 30, 33, 35, 36] while 18 used two tendons [9, 18,19,20,21, 23,24,25,26,27,28,29,30,31,32,33,34, 36]. The conventional, and most described configuration was a 4-strand, doubled ST and GC graft (13 studies) [9, 18, 20, 21, 23,24,25,26, 29,30,31, 33, 34]. Other commonly described graft configurations included a 4-strand, quadrupled ST configuration (7 studies) [23, 27, 30, 32, 33, 35, 36], a 5-strand, tripled ST and doubled GC configuration (8 studies) [9, 18,19,20,21, 25, 26, 29], and an 8-strand, quadrupled ST and GC configuration (3 studies) [28, 32, 36]. Among all studies, the mean number of strands used in the final graft configuration was 4.9, the mean number of ST strands used was 3.0 and the mean number of GC strands used was 2.6 when the GC was utilized. The ST harvest length ranged from 20 to 32 cm. The GC harvest length ranged from 16 cm to 28.2 cm.

Fixation strategy

The femoral and tibial fixation strategies depended on the graft configuration and resultant graft length. Both aperture and suspensory fixation were utilized on both sides. For femoral fixation, 14 studies [18, 23,24,25, 27,28,29,30,31,32,33,34,35,36] used a suspensory strategy and 4 studies used an aperture strategy [9, 19, 20, 26]. On the tibial side, 6 studies [27, 28, 32, 33, 35, 36] used a suspensory strategy and 11 studies used an aperture strategy [9, 18,19,20, 23, 24, 26, 29, 31, 33, 34]. In general, a graft length of at least 9 cm was needed to use aperture fixation on both sides and a graft length of at least 6 cm, which ensures 2 cm of bone-graft contact within each tunnel was needed to use suspensory fixation on both sides.

Configuration decision-making

The most common graft configuration decision-making algorithm entailed a preliminary assessment of a 4-strand, quadrupled ST graft diameter and the addition of a GC if the initial diameter was less than 8 mm [27, 32, 36]. A second algorithm involved a preliminary assessment of the conventional 4-strand doubled ST and GC graft diameter and altering the configuration to a 5-strand, tripled ST, doubled GC graft if the initial diameter was less than 8 mm 19. A third algorithm described changing the tibial fixation strategy from aperture to suspensory in the presence of a short or small diameter graft [31].

Clinical outcomes

Six studies compared a conventional 4-strand, doubled ST and GC graft configuration to a 5-strand, tripled ST and doubled GC graft configuration [9, 18, 20, 21, 25, 26]. No differences were reported in the failure rate or included patient reported outcome measures in any study. However, all 6 studies reported a significantly larger graft diameter in the 5-strand graft group compared to 4-strand group.

Three studies compared a conventional 4-strand, doubled ST and GC graft configuration using either suspensory or aperture femoral fixation and aperture tibial fixation to a 4-strand, quadrupled ST graft configuration with suspensory fixation for both femoral and tibial sides [23, 30, 33]. No differences were reported in the failure rate or patient reported clinical outcomes in any study. Similarly, no report of differences in graft diameter were included in any study. Four studies compared other graft configurations from 3-strand, tripled ST grafts to 8-strand, quadrupled ST, and GC grafts [18, 22, 29, 32]. Regardless of the graft configurations compared, no study reported a difference in failure rates. In every study, the more strands incorporated, the larger the graft diameter.

Discussion

There are multiple configuration variations available to achieve adequate autograft diameter when performing a SB HT ACLR. This review provides an algorithm to maximize autograft diameter for single-tendon or multi-tendon harvests of different lengths by optimizing the number of graft strands and femoral and tibial fixation techniques.

All femoral and tibial fixation techniques included in the review ensured a minimum of 2 cm of bone-graft overlap. Previous basic science research only weakly supports the general belief that more bone-graft overlap leads to improved graft healing [37]. One previous extra-articular animal model found improved pull-out strength with 2 cm of overlap compared to 1 cm [38], while a separate intra-articular animal model found no difference between 5 and 15 mm of overlap [39]. The different animal models may explain the difference in findings. The amount of bone-graft overlap and associated fixation technique may also play a role in tunnel widening and the need for staged procedures in the context of an ACLR failure. Both aperture and suspensory fixation may lead to tunnel widening and it is currently unclear whether one fixation strategy is superior to the other [41,42,43].

Two of the most common graft configurations for SB HT ACLR are 4-strand doubled ST and GC, and 5-strand tripled ST and doubled GC. The studies included in our review found no difference in failure rates or patient-reported outcomes between these two graft configurations. However, the 5-strand tripled ST and doubled GC graft resulted in a larger diameter compared to the 4-strand doubled ST and GC. These findings are in keeping with a recent subgroup analysis of 399 patients from a large randomized controlled trial comparing SB HT ACLR with and without lateral extra-articular tenodesis (LET). The subgroup analysis compared the clinical outcomes after 4-strand and 5-strand SB HT ACLR with or without concomitant LET and found no difference in failure rate or clinical outcomes after two years [40]. One possible explanation for an equivalent failure rate after 4- and 5-strand SB HT ACLR may be related to the definition of clinical failure and graft plastic deformation compared to complete rupture. It may be that while a larger diameter HT autograft provides greater resistance to complete rupture, there is little difference in the resistance to 3 mm of plastic deformation, which may be enough to result in clinical symptoms of recurrent instability [44].

Pearls and pitfalls

A successful ACLR is contingent on thoughtful preoperative planning and the ability to utilize multiple alternative strategies if unforeseen circumstances arise. An examination under anesthesia before proceeding with surgery should be completed and may provide information that changes the surgical plan. During the hamstrings harvest, ensure all fascial bands are completely released before the proximal harvest to avoid a short tendon harvest or tendon amputation. If a short tendon harvest occurs, a further 3–4 cm may be added to the total tendon length by harvesting the periosteal footprint on the proximal tibia.

The central theme of this review is to arm the orthopaedic surgeon with multiple graft configurations and accompanying fixation strategies to maximize graft diameter in the setting of a short hamstrings harvest. Preparation for different graft configurations begins with awareness of the operating room resources and communication with hospital administration to confirm the availability of preferred fixation strategies. In the setting of minimal available OR resources, a basic trauma small fragment set, and non-absorbable suture can be used to tension the graft around a screw at the cortical tunnel end. To avoid lateral femoral outer cortical breach during graft passage with an adjustable suspensory device, avoid drilling the femoral tunnel too distal in metaphyseal bone [17].

Limitations

Limitations of the current literature that compares SB HT ACLR configurations include a lack of level 1 studies comparing the graft diameter, re-injury rate and clinical outcomes between configurations, short follow-up times, a lack of consistent outcome measures and small sample sizes.

Future perspectives

Randomized controlled trials (RCTs) comparing SB HT ACLR graft configurations would help identify the ideal graft configuration and fixation strategy for each population subgroup. The ability to convert the graft configuration intra-operatively remains useful in the scenario of a short graft harvest or when limited fixation methods are available.

Conclusion

Multiple configurations exist to obtain an appropriate graft diameter during a SB HT ACLR. Pre-operative planning should identify what implants are available and which femoral and tibial fixation techniques will be employed. The 5-strand, tripled ST and doubled GC provides a simple alternative to the conventional 4-strand, doubled ST and GC graft configuration using the same femoral and tibial fixation methods. An 8-strand, quadrupled ST and GC graft configuration provides a simple alternative to a 4-strand, quadrupled ST graft configuration using the same femoral and tibial fixation methods. An intra-operative assessment of harvested tendon length and preliminary graft diameter facilitates changes in the graft configuration to obtain an appropriate diameter graft to minimize the risk of re-injury post-operatively.