Introduction

Anterior cruciate ligament tears, which are a common musculoskeletal injury, are being treated with the ligament reconstruction in increasing numbers. It is clear that a perfect graft for ACL reconstruction does not exist. Because of this, the surgeon must be familiar with all varieties of possible graft choices. Graft selection is an important decision for optimal outcomes following reconstruction, and there are several graft options, such as autografts, allografts, and synthetic grafts, available for the ACL reconstruction (West and Harner 2005).

Most of the primary ACL reconstructions are performed with using either bone–patellar–bone (BPTB) or hamstring autografts, but there is still a search for the ideal graft option (Duquin et al. 2009). The most important reason for this is donor (harvest) site morbidities. These morbidities can be grouped into anterior knee pain and discomfort resulting from decreased function, including range of motion and muscular strength; local discomfort caused by numbness, tenderness, or an inability to kneel; and late tissue reaction at the donor site (Franceschi et al. 2008).

One of the alternative techniques to the most common two grafts is the QT autograft, which has long been reported as an ACL graft option (Santori et al. 2004). Quadriceps tendon grafts have been utilized less frequently than other types of autografts, but because of low donor-site morbidity reported, there has been recent popularization of this graft (De Angelis and Fulkerson 2007).

History of the QT

Blauth reported his technique for harvesting the central third of the QT with patellar bone block at one end in 1984 (Blauth 1984). Then Staubli published his series about QT reconstruction in 1997 (Staubli 1997). In the original technique, the QT graft was composed of both tendon and a patellar bone block. In 1999, Fulkerson modified this technique by only using the QT without a bone block to decrease the risk of the patellar fracture (Fulkerson 1999).

Properties of the QT Autograft

Anatomically the quadriceps tendon consisted of three layers. The rectus femoris (RF) tendon is present at the most surface layer. The middle layer consists of the vastus lateralis (VL) and medialis (VM) tendons, and the deep layer is the vastus intermedius (VI) tendon. The strength and size of the quadriceps come from the VL and VI tendons, which were overlapped and were firmly connected. The narrowest width of the RF was reported to be 15.3 mm, and the narrowest point is 4.8 mm proximal to the upper end of the patella. The average length of the RF is 27.3 cm (Iriuchishima et al. 2012). The entire quadriceps tendon may be used as an ACL graft, but VL and VI tendons show different fiber directions so that the strength of the quadriceps tendon is questionable (Iriuchishima et al. 2012).

The quadriceps tendon attaches to a broad and profound area of the proximal pole of the patella, whereas the PT originates superficially from the anterior aspect of the patellar apex (Staubli et al. 1999). The amount of fibrocartilage is significantly greater in the attachment zone of the QT compared to the PT (Evans et al. 1990). This may suggest that QT graft may be a more physiologic graft choice because of the angulation between the ligament and the bone during knee movements.

The sagittal angle between the axis of the QT and the patella changes by approximately 30° during flexion–extension of the knee, which may result in a better range of motion after an ACL reconstruction. This angle remains unchanged for the PT origin (Eijden et al. 1987). These anatomical and biomechanical properties of the QT tendon reveal that QT tendon is a good alternative to the other autograft sources.

The quadriceps tendon has four layers that were oriented longitudinally and obliquely. The biomechanical strength, large cross-sectional area, appropriate length, and the advantage of having a bone plug on one end are the advantages of the QT (Brand et al. 2000). Anterior knee pain has been reported to be less with this graft compared with a BPTB graft (Theut et al. 2003).

In Staubli et al. study, QT graft and BPTB grafts were compared. The mean QT graft lengths averaged 87 ± 9.7 mm for the right knee and 85.2 ± 8.4 mm for the left knee without the bone block. The tendinous part of the patellar tendon was on average 51.6 ± 6.9 mm in the right knee and 52.2 ± 4.8 mm in the left knee. The cross-sectional area of a 10 mm wide QT graft averaged 64.4 ± 8.4 mm2. This value was greater when compared with the patellar tendon (36.8 ± 5.7 mm2). These anatomical studies reveal that the quadriceps tendon is thicker, longer, and wider than the patellar tendon (Staubli et al. 1999).

The ultimate tensile loads of the intact ACL, 10 mm QT graft, and BPTB graft are 2,160 ± 157 N, 2,173 ± 618 N, and 1,953 ± 325 N, respectively, which means that the QT is biomechanically an option to the other choices (Woo et al. 1991). The quadriceps tendon is thicker and wider than the patella tendon. One can obtain a quadriceps tendon graft for ACL reconstruction with a volume, which is 50 % larger than a bone–patellar tendon–bone graft of similar width (Fulkerson and Langeland 1995).

The early physical findings of hamstring, patellar tendon, and quadriceps tendon ACL reconstructions were prospectively compared by Joseph and colleagues, and they reported that the quadriceps free tendon group achieved knee extension earlier than the patellar tendon group, and QT group required less pain medication postoperatively than the other two groups (Joseph et al. 2006). The central quadriceps tendon was reported to be as strong after graft harvest as the patellar tendon when testing to failure in cadaveric knees (Adams et al. 2006). Also in another study, there was a significant influence of bone plug length on the stability of the press-fit fixation technique. If graft harvest generates bone plugs of 15 mm length, knee flexion should be limited to 60° during the first 3 weeks of postperative period. Biomechanically loading characteristics when tested to ultimate failure with a rising load angle using a 25 mm length patellar bone plug were comparable for the quadriceps tendon with bone plug and a traditional bone–patellar tendon–bone construct. According to that study, the biomechanical characteristics of QTPB grafts are comparable to BPTB grafts in femoral press-fit fixation technique (Dargel et al. 2006).

Biomechanical and cadaveric studies showed that the normal ACL’s cross-sectional area was 44 mm2 and an ultimate load to failure was 1,725–2,160 N (Noyes et al. 1984). This load to failure was highest when the ACL was intact in its anatomical position (Woo et al. 1991). The QT provides a thicker graft than other two common autograft options with an adequate load to failure in cadaveric models (West and Harner 2005). In comparison, the BPTB autograft, four-strand hamstring graft, and QT graft have different cross-sectional areas (35, 53, and 62 mm2, respectively), while the greatest is the QT. All autografts (hamstring 4,090 N, patella 2977 N, quadriceps 2,352 N) have a greater load to failure than the native ACL (Table 1).

Table 1 Comparison of tensile loads and load to failure of grafts according to Woo’s, Noyes’, and West’s studies
Full size table

Mulford et al. (2012) have reported that 81–95 % of patients had a normal Lachman test and 80–95 % a normal pivot shift test compared with the 76–100 % and 81–100 % following a PT graft and 64–100 % and 72–100 % following a hamstring graft, respectively. They also reported that the IKDC scores were A and B (normal and nearly normal) in 88 % compared with the A and B in 91.6 % of reconstructions using PT and in 90.7 % of hamstring grafts. The mean Lysholm score was 91 points for QT reconstruction, whereas it was 91–93 and 80–94, respectively, following PT and hamstring ACL reconstructions (Mulford et al. 2012) (Table 2).

Table 2 Lachman, pivot-shift, IKDC, and Lysholm scores comparison of three graft options according to Milford
Full size table

Evidence from anatomic, cryosectional, and structural properties analyses has reported that the QT graft can be a valuable and versatile adjunct for the ACL reconstruction (Staubli et al. 1996). The multilayer structure of the QT allows this graft to be split, on one end into two separate tails, which makes it a good option for a double-bundle repair. It may be used with one side tunnel with bone block and the other side for double-bundle tunnels for split tails of the QT (Caborn and Chang 2005).

The properties of QT autograft are as follows:

  • Harvesting is easy after one has an initial learning curve.

  • The graft can be obtained with or without a patellar bone block.

  • In revision operations, the diameter can often be sufficient to accommodate an expanded tibial tunnel.

  • Donor-site problems have been reported to be than the patellar tendon.

  • Biomechanical characteristics are good.

  • Cross-sectional area is larger when compared to the PT.

  • Quadriceps inhibition after the harvest is usually less.

  • The residual strength of the extensor mechanism has been reported to be less impaired compared to PT graft (Fulkerson and Langeland 1995; Morgan et al. 1995; Staubli et al. 1996; Noronha 2002; Adams et al. 2006; Chen et al. 2006).

The QT autograft is an alternative to BPTB, especially in patients who need to kneel frequently or who require deep flexion of the knee. However, the best fixation method still has to be decided (Fulkerson and Langeland 1995; Morgan et al. 1995; Staubli et al. 1996; Theut et al. 2003).

Harvesting the central third of the QT does not affect the function of the extensor mechanism as much compared with the bone–tendon–bone autografts (Pigozzi et al. 2004). Biomechanically, Fulkerson tested both the native intact QT and the residual QT after harvest and had a statistically higher strength at failure than the corresponding patellar tendon construct (Fulkerson 1999).

Harvesting the quadriceps tendon autograft is not difficult and obtaining this graft can be safe, but surgeons should be aware of the anatomy of the proximal patella with a curved proximal surface, dense cortical bone, and closely adherent suprapatellar pouch. Donor-site morbidity has been reported to be minimal with QT graft. The quadriceps–patellar bone graft is sufficiently large and strong, and it can achieve good ligament function after reconstruction (Chen et al. 2006).

Patient Selection

Patients, who participate in pivoting and high valgus stress activities and sports, who have a concomitant MCL injury, whose jobs require kneeling or long periods of knee flexion, could be treated with QT graft. With using this graft, the medial stabilizers are spared. It may also be used for patient with patella baja, Osgood–Schlatter’s disease sequelae, and patellar tendinitis, and it can also be used for ACL revision surgeries and multiple knee ligament injuries. A poorly motivated female patient where postoperative quadriceps weakness is a potential problem may be a relative contraindication. When the QT is used for revision surgery, the results also have been acceptable (Garofalo et al. 2006).

In one study after follow-up for more than 36 months, simultaneous arthroscopically assisted reconstruction of both ACL and PCL using hamstring and quadriceps autografts for restoring knee stability were reported to be effective and safe. Harvesting the QT autograft was reportedly easy during arthroscopic surgery for PCL reconstruction, and the affixed sutures of the tendinous end of the quadriceps tendon–patella construct pass easily from the tibial tunnel through the joint to the femoral tunnel, while the bone plug remains within the tibial tunnel. In this study the ACL reconstruction portion of the procedure had a more predictable outcome than the PCL reconstruction (Lo et al. 2009).

Surgical Technique

A suprapatellar incision, about 5 cm long, is performed. Skin flaps are raised, the fascia over the tendon is incised, and the entire width of the QT is exposed. The central portion of the tendon with a patellar bone block is harvested with the careful dissection to the joint. The deep synovium of the suprapatellar pouch should not be violated during the tendon harvesting to prevent fluid leakage during the arthroscopy. If the pouch is opened, subcutaneous fluid distention, decrease in the operative vision, and adhesions between tendon and anterior femur may occur (Woo et al. 1991). Bending the knee over 120o may provide a satisfactory arthroscopic view. If any violation occurs, clamping or suturing the defect will decrease the leakage.

The midportion one cm width of the QT is harvested longitudinally until the superior end of the patella distally (Fig.1). After freeing the tendon, a 25 mm long, 10 mm wide, and 5 mm thick bone block is harvested from the proximal patella (Fig. 2). During the bony harvest, one should keep in mind that the proximal part of the patella is oblique and short, and oblique harvesting of the patella should be avoided. After the graft is harvested, the defect can be filled with bone chips from the tibial tunnel drilling if available. During the learning curve, the harvesting incision may be extended longer for less complications and operation time.

Fig. 1
figure 1figure 1

Midportion of the QT of 1 cm width was dissected longitudinally till the superior end of the patella distally

Full size image
Fig. 2
figure 2figure 2

After freeing the tendon proximally, 25 mm long, 10 mm wide, and 5 mm thick bone block is freed from the patella

Full size image

The free end of the QT is stitched with #5 Ethibond (Ethicon, Somerville, NJ) sutures to facilitate graft passage (Fig. 3). Then, the diameter of the tendinous part of the graft is measured. The patellar bone plug should be 25 mm long, 10 mm wide, and 5 mm thick. In general, the tendon portion is 70–80 mm long, 10 mm wide, and 5–6 mm thick. The tendon of that size allows an easy passage of an 11 mm diameter template. After the graft preparation, the tibial footprint is prepared in the usual fashion, such as stump protection and limited notchplasty if needed, for a single-bundle reconstruction. The tibial tunnel is placed in slightly posteriorly to the central part of the residual ACL footprint.

Fig. 3
figure 3figure 3

Free end of the QT is stitched with the #5 Ethibond sutures and also another #5 Ethibond is passed through the bony end of the graft to facilitate graft passage

Full size image

The tibial drill guide is adjusted to a 55° angle, and the tibial tunnel is reamed starting from a 6 mm diameter and end up to a 10 mm reamer. Femoral and tibial tunnels are drilled 1 mm smaller than the desired size and are enlarged with appropriately sized dilators. After drilling the femoral and tibial tunnels, the bone plug is pulled with the sutures into the femoral socket through the anteromedial portal, and the free tendon end is pulled into the tibial tunnel with the grasping forceps holding the sutures at the end of the tendinous part. Both sides can be fixed with a 9 mm tricalcium phosphate interference screw (Mega fix, Karl Storz GmbH). The tension, stability, and impingement of the graft are checked after the fixation is completed (Fig. 4). Comparison of preoperative and postoperative X-rays from a revision case of a nonfunctioning ACL repair that was performed with a QT autograft is shown in (Fig. 5).

Fig. 4
figure 4figure 4

The tension of the graft is checked with the probe

Full size image
Fig. 5
figure 5figure 5

(a) Preoperative AP X-ray of a revision case of a nonfunctioning ACL repair. (b) Preoperative lateral X-ray of a revision case of a nonfunctioning ACL repair. (c) Postoperative AP X-ray. (d) Postoperative lateral X-ray

Full size image

Rehabilitation

The rehabilitation protocol is similar to the hamstring graft reconstruction rehabilitation. The rehabilitation program is started before surgery with patient education and achieving full ROM and if possible complete reduction of intra-articular inflammation and swelling before surgery to avoid arthrofibrosis.

On the first postoperative day, the swelling and pain are tried to be controlled to limit muscular inhibition and atrophy. Weight bearing is allowed the next day with crutches, and full passive extension, early initiation of quadriceps, and hamstring activity should be obtained at first week. Early open chain exercises that may theoretically shear or tear the weak immature ACL graft are avoided. Rehabilitation programs should involve functional testing and functional sport-specific training prior to return to sport.

Our rehabilitation program following ACL reconstruction with QT autograft is composed of five main phases:

  • Phase 1 (first 2 weeks) is the maximal protection phase. The active range of motion (AROM) 0–90° within 10 days, active quadriceps contraction, full weight bearing on crutches without knee brace, edema control, graft protection, and wound healing are the aims of this phase. Achieving full extension is important during phase 1. During the first 2–3 weeks, early rehabilitation exercises may appear difficult in some patients having suprapatellar pain and tenderness causing major discomfort.

  • Phase 2 (3–6 weeks) is the medium protection phase. The goals of this phase are full AROM equal to nonsurgical knee, normal gait without assistive device, and independent activities of daily living.

  • Phase 3 (7–12 weeks) is the minimum protection and strengthening phase. The goals are good strength of the quadriceps at the operative site with equal quadriceps muscle girth. Single-leg squats to 60° with good form should be performed by the end of this phase.

  • Phase 4 (12–20 weeks) is the progressive strengthening phase. The goals during this phase are pain-free running, landing, and jumping with the uninvolved limb dominating effort and gaining at least 75 % of the uninvolved knee in jump tests.

  • Phase 5 (20–26 weeks) is the return to sports phase and aims at least 85 % of the uninvolved knee in jumping tests. Single-leg squats, 20 repetitions to 60° of knee flexion, should be performed during this phase in addition to single-leg stance at least 60 s (Berker et al. 2009).

Results

Patient satisfaction, knee joint function, and donor-site morbidity differ for QT graft from other autograft choices. The BPTB graft is better according to the IKDC scores, whereas functional parameters such as Lysholm and Noyes scores report comparable results. The donor-site morbidity was reported to be significantly less with QT grafts than the BPTB grafts. The QT may be a preferable graft source for patients who load the knee joint and who participate in activities with deep knee function. Also, the QT may be an alternative for revision ACL surgery when the patella tendon already was harvested (Gorschewsky et al. 2007).

The use of the QT as a graft choice for ACL reconstruction has been reported to provide acceptable clinical outcomes (Howe et al. 1991; Kaplan et al. 1991; Staubli 1997; Chen et al. 1999; Kim et al. 2001; Theut et al. 2003). From these outcome analyses, the average Lysholm knee scores were 93 points at the final assessment compared to preoperatively 61 points. During strenuous or moderate activities, minimal pain and swelling were reported for 91 % of the patients after reconstruction, and moreover, during that activities 85 % of the patients did not display partial giving way of their reconstructed knee, and 91 % of the patients reported no full giving way according to IKDC guideline. Less than 5 mm ligament laxity postoperatively with KT-1000 arthrometer tests was observed in 94 % of the patients. The average anterior displacement preoperatively and postoperatively was 11.88 ± 1.09 mm and 1.74 ± 1.80 mm, respectively (Chen et al. 2006).

Satisfactory results with QT reconstruction were reported with improved Lysholm scores in which 94 % of the patients were graded A or B with a median laxity of 2 mm postoperatively. Extension peak torque of the quadriceps muscle recovered to 82 % and 89 % of that of the contralateral knee at 180°/s at 1 year and 2 years after surgery, respectively. The patellar congruence angle and Insall–Salvati ratio did not show any significant change after QT usage (Lee et al. 2004).

In one study, most patients recovered to 80 % or more of extensor and flexor muscle strength (94 % and 91 %, respectively). In this study, only 56 % of the patients could recover over 90 % of extensor muscle strength and 50 % of the patients over 90 % of flexor muscle strength compared to the normal side, but the reason for it was explained with the sources of the patients, which included only 47 % (16 of 34 patients) of professional and recreational athletes (Chen et al. 2006).

Bone tunnel enlargement was observed in 37 % of patients in Segawa’s study. Enlargement occurred in 25 % of the femoral tunnels and 30 % of the tibial tunnels. Enlargement of both tunnels occurred in 18 % of the knees (Segawa et al. 2001). In Chen et al. study with bone–quadriceps tendon graft, tunnel expansion with more than 1 mm was only found in two (6 %) tibial tunnels (Chen et al. 2006).

Discussion

Patellar, hamstring, and quadriceps tendon are three autologous graft choices currently mainly used for ACL reconstruction. Clinical studies have not revealed any major differences in clinical outcomes among these grafts regardless of fixation technique yet. Mostly, other factors, such as graft harvest morbidity and complications, are more important when comparing different grafts.

The QT bone or QT graft is gaining popularity because of high reported rates of chronic patellofemoral problems with patellar tendon grafts. Any solutions such as suturing the tendon gap or bone grafting the patellar defect do not reduce anterior knee problems and kneeling complaints to treat this frustrating morbidity (Shino et al. 1993, Breitfuss et al. 1996).

In one study, no patients had signs or symptoms of patellofemoral pain at minimum 2-year follow-up evaluation, which is an important advantage in comparison to central third patellar tendon reconstructions. There was no case of quadriceps tendon rupture following harvest of the central quadriceps tendon (Theut et al. 2003).

Because a perfect graft for ACL reconstructions does not exist, the surgeon must be familiar with all varieties of possible graft choices. With understanding the anatomy of the proximal patella, which includes a curved proximal surface, dense cortical bone, and a closely adherent suprapatellar pouch, graft harvesting becomes easier. The quadriceps tendon autograft is usually not difficult to harvest and surgeons can obtain and use this graft safely.

Cross-References