Single-Bundle Versus Double-Bundle Acl Reconstructions in Isolation and in Conjunction with Extra-Articular Iliotibial Band Tenodesis

Abstract

Background

Intra-articular anterior cruciate ligament (ACL) reconstruction has been the primary treatment option for isolated ACL injuries for many years. An anatomic double-bundle reconstruction has been devised in an effort to improve rotational control. The role of the extra-articular iliotibial band tenodesis in ACL injuries has evolved from primary treatment, to an adjuvant secondary procedure, to being used more selectively in revision ACL reconstructions. Hypotheses: 1) Single-bundle and doublebundle intra-articular ACL reconstructions will both restore pre-injury laxity measurements in an isolated ACL injury cadaver model. 2) The deep iliotibial band structures contribute to rotational control and in a dual ACL + ITB injury cadaver model, ACL reconstruction alone cannot restore rotational control.

Study Design

Controlled Laboratory Design

Methods

17 fresh frozen cadavers received intra-articular reconstructions, seven single-bundle and ten double-bundle; laxity was measured with the ACL intact/ITB intact, ACL reconstructed/ITB intact, after cutting the ITB, and after an ITB tenodesis procedure; laxity measurements of anterior tibial translation(ATT) and internal rotation(IR) were measured following applications of an anterior shear force, an internal torque and a coupled anterior shear force-internal torque at 30 and 90 degrees of flexion.

Results

Single-bundle and double-bundle ACL reconstructions both restored IR to a native knee state under isolated internal torques and under coupled forces. Both reconstruction techniques also re-established anterior tibial translation to at least the pre-ACL injury level, with over-constraint in the double-bundle subgroup [5.00 (+2.11) to 3.50(+1.18), p-value 0.026] under coupled loads at 30 degrees of flexion. With the individual ACL reconstructions held constant, under coupled forces mean IR increased in the single-bundle subgroup [13.7(+1.1) to 17.6(+1.2), p-value 0.004] and the double-bundle subgroup [9.5(+1.0) to 12.4(+1.0), p-value 0.009] with the cutting of the ITB at 30 degrees. Under internal torque, mean IR increased in the single-bundle subgroup [14.0(+1.0) to 18.4(+1.6), p-value 0.016] with the cutting of the ITB at 30 degrees, while IR increased in the double-bundle subgroup [10.0(+1.3) to 13.4(+1.5), p-value 0.002] under the same internal torque at 90 degrees. With the ACL reconstruction held constant, ATT did not significantly change when the ITB was cut or when it was tenodesed under any specific loading condition.

Conclusion

Single-bundle and double-bundle intra-articular reconstructions were both able to restore internal rotation and anterior tibial translation to at least native knee laxity levels after an isolated laboratory ACL injury. When the ACL reconstructions were held constant, internal rotation statistically increased with the cutting of the ITB under multiple testing conditions in both the single-bundle and double-bundle subgroups.

Introduction

The treatment of ACL injuries has been researched and debated extensively in the literature amongst orthopedic surgeons since ACLs were first repaired in the 1950’s¹. As technology facilitated improvements in surgical technique, ACL treatment options focused on intra-articular variables of graft type, tunnel location, method for drilling tunnels, and number of bundles^2-5. The treatment goals of restored knee function and prevention of long-term consequences have remained constant. The focus of ACL reconstruction has become the restoration of native ACL anatomy.

The native ACL is comprised of three bundles, the anteromedial, the posterolateral, and the intermediate bundle⁶. Collectively, the bundles function to act as the primary static stabilizer of anterior tibial translation and internal rotation of the tibia throughout knee range of motion. However, each bundle restricts rotation and translation differently as the knee flexes and extends⁷. Historically, single bundle graft placement has more closely resembled the anatomical position of the an- teromedial bundle and has not reproduced the position and rotational control of the posterolateral bundle^8,9. Rotational stability has always been the most difficult motion to restrict with intra-articular reconstruction because of the intrinsically short biomechanical lever arm of the ACL graft.

Extra-articular reconstructions have the biomechanical advantage for rotational control and have been used in varying degrees to treat ACL injuries for decades^10-15. Terry et al. demonstrated that during an injury that resulted in a clinically deficient ACL, 93% of IT bands were also torn¹⁶. The magnitude of iliotibial band compromise is difficult to assess at the time of injury and its role in long-term success or failure of intra-articular reconstruction is even more unknown.

An in vivo study designed to assess the prevalence and impact of ITB injuries in clinically deficient ACL knees would be difficult to perform. Thus, a laboratory cadaver-based model was made. Within this model, two hypotheses focused on ACL reconstruction were tested. 1) Single-bundle and double-bundle intra-articular reconstructions will both restore pre-injury laxity measurements in an isolated ACL injury cadaver model. 2) The deep iliotibial band structures contribute to rotational control in a dual ACL + ITB injury cadaver model and thus, ACL reconstruction alone cannot restore rotational control.

Methods

Twenty paired fresh-frozen cadaveric knees (63-87 years old) were thawed in a lukewarm water bath 12-18 hours prior to use. Three knees were removed from the testing protocol. Two were removed secondary to computer error resulting in data loss and one for intra-articular tunnel misplacement (Figure 1). All specimens contained at minimum 18 cm of bone and soft tissue proximal and distal to the tibio-femoral joint line. Testing apparatus (Figure 2) rigidly secured cadavers proximally with a clamp around the femur. Poly (methyl methacralate) (PMMA) and a wooden dowel were placed within the intramedullary space of the femur to prevent femur fracture during testing. Flexion and extension was controlled distally by placing PMMA and a wooden dowel with a perpendicularly oriented K-wire within the tibial intramedullary space. Ropes that could be adjusted were secured to the K-wire and anchored to an external structure to allow testing at variable angles of flexion.

Figure 1. –Flow Diagram. Sequence of procedures and kinematic testing.

Figure 2. –A) Lateral and B) AP drawing of biomechanical apparatus. c) oblique photograph gurof specimen during testing in lab.

Soft tissues covering the tibial shaft were stripped to eight cm distal to the joint line. Exposure of the tibia was performed to allow a specially designed metal ring to be anchored to the tibia ten cm distal to the joint line. The center of the ring was aligned with the longitudinal axis of the tibia and the three fixation screws were equally spaced on the ring and tightened to the bone. The 4.15 cm fixed external radius of the ring provided a consistent moment arm for torque application. Tibial torsion was applied via two ropes, which were secured to the outside of the ring. The first rope rotated from the ring anteriorly, while the second rope rotated from the ring posteriorly. Each rope was run over freely rotating metal bars, medially and laterally, and connected to 6.1 kg weights respectively (5Nm torque). The heights of the metal bars were adjusted based upon the specimen and the flexion angle to maintain perpendicular alignment of the rotational moments to the longitudinal axis of the tibia. During torque application, both 6.1 kg weights were released simultaneously to equally load the tibial ring.

The anteriorly directed load on the tibia was produced via a 2.5 cm wide cloth strap that was wrapped superficially around the skin and soft tissues of the proximal tibia. The center of the strap was aligned parallel with the joint line three to four cm distal to the tibial plateau. A rope was tied to the front of the strap directly anterior to the tibial tubercle. The rope was strung over a freely rotating metal bar whose height was adjusted to maintain an anteriorly directed moment that was perpendicular to the longitudinal axis of the tibia. The distal end of the rope was tied to a 0.2 kg metal hook that held a 9.1 kg sandbag. The 5 Nm torque and 91.2 N anterior shear force were oriented perpendicular to the tibia and were designed to coincide with previously performed ACL biomechanical literature^{2,5,8,10,15,17,18}.

Prior to kinematic testing all knees were examined and determined to be stable to varus/valgus stress with no corresponding joint space opening. Lachman and pivot shift tests were also negative. OrthoPilot 2.0 (B/ Braun AESCULAP, Tuttlingen, Germany) computer- assisted-surgery system was used for obtainment of kinematic data, confirmation of ACL tunnel location, and verification of knee flexion angles. The accuracy of the OrthoPilot system has been documented as 0-1 degree and 0-1 mm¹⁹. The use of the OrthoPilot system has been documented as able to reproducibly measure laxity and position of ACL graft tunnels^19-21.

OrthoPilot references the knee through rigidly fixed femoral and tibial markers. The femoral marker was placed just proximal to the medial femoral epicondyle and the tibial marker was fixed to the anterior medial surface of the tibia 15-20 cm distal to the joint line. The joint line is assessed by repeatedly cycling the knee through flexion and extension. Additional required intra-articular anatomic reference points were tagged by the laboratory investigator using the OrthoPilot stylus.

Testing Protocol

Four different testing conditions were assessed for each specimen (Figure 1).

• 1.ACL intact/ITB intact- represents the native knee before injury and reconstructive surgery.
• 2.ACL reconstruction/ITB intact – represents an ACL reconstructed knee (either single or double bundle) with an uninjured ITB.
• 3.ACL reconstruction/ITB deficient – represents a combination ACL+ITB injury where only the ACL is reconstructed.
• 4.ACL reconstruction/ITB tenodesis – represents a combination ACL+ITB injury where both the ACL and ITB were reconstructed.

Assessment of Laxity

Each specimen was evaluated at 30 and 90 degrees of flexion. The OrthoPilot 2.0 CAS system measured anterior tibial translation (ATT) and rotational change of the tibia in reference to the rigidly fixed femur at each testing point. Measurements were based on a zero point made with the femur rigidly fixed proximally and the tibia secured in the desired angle of flexion by distal rope attachment to the tibial intramedullary dowel.

• 1.Neutral drawer - anterior 91.2 N shear force was applied to the proximal tibia strap.
• 2.Internal torque - 5 Nm torque was applied to the tibial ring.
• 3.Coupled anterior drawer-internal torque – anterior 91.2 N shear force was applied in combination with the 5 Nm internal torque.

Intra-articular ACL Reconstruction

Single bundle

The ACL was cut and debrided. A centrally located guide pin in the tibial footprint was used to coincide with the suggested seven to eight mm of distance between the PCL and the guide pin, one knee had six mm of distance²². A ten mm drill was used to create the tunnel. The femoral guide pin was placed trans-tibially by manipulating the flexed knee to accommodate a corresponding ten o’clock or two o’clock position. Tunnel position was confirmed by arthroscopic visualization and supported by the CAS system. A ten mm drill was also used for this tunnel. After drilling, a posterior wall of one to four mm was visualized in all specimens.

In order to eliminate cadaveric tendon inconsistencies as a possible source of error, synthetic rope was used. Elongation of several samples of ten mm diameter synthetic rope were tested. These ropes were secured at one end and weight was hung from the other end. No time related graft elongation was measured. A 15 cm polypropylene rope with a diameter of ten mm was chosen based upon consistent size with biologic grafts and its ability to resist deformation. Once the graft was placed, the knee was flexed and extended in external rotation three times to assess graft isometry before securing. The graft was tightened and secured at 30 degrees of flexion using two perpendicularly oriented hemostats clamped directly against both the tibia and the femur. Graft tension was arthroscopically visualized and palpated with a curved probe after fixation and periodically during kinematic testing to monitor for fixation slippage.

Double bundle

The ACL was cut and debrided. The anteromedial bundle was first made by drilling a guide pin to exit the tibial footprint centered at 45-52% of the medial width of the tibial plateau and ten mm anterior to the PCL. The femoral tunnel was drilled through a medial joint line incision in an effort to reproduce the native AM and PL footprints and avoid potential complications noted by Giron et al⁴. The femoral guide pin was positioned with the use of OrthoPilot between the 1:30 and 2:30 or 9:30 and 10:30 position and was visualized to be within the native footprint. The guide pin was located approximately six to seven mm from the “over-the-top” position.

Both tibial and femoral tunnels were drilled with a five mm drill bit. (Figure 3)

Figure 3. –Arthroscopy screen captures. Image 1: PcL aligned diagonally with AM and PL tibial tunnels anterior to it; Image 2: AM and PL femoral tunnels with boney bridge visible; Image 3: synthetic AM and PL bundles twisting around each other at 90 degrees of flexion.

The posterolateral bundle was made by positioning a guide pin within the tibial footprint centered on the tibia at 50-55% of the medial width of the tibial plateau and three to four mm anterior to the PCL. This tunnel was kept lateral to ensure that the tunnels did not converge on one another. A visible bony bridge between the tunnels was always intact before graft placement. The tunnel was always re-visualized to confirm placement within the ACL footprint. The femoral guide pin was placed between the 2:30 and 3:30 or 8:30 and 9:30 position and was approximately eight to nine mm from the “over-the-top” position. A visible bony bridge between the tunnels was always confirmed to be intact before graft placement.

Five mm grafts made of polyester that demonstrated no time dependent elongation were used as the double bundle grafts. These grafts were also secured with two perpendicular hemostats clamped directly against the bone and were tightened in the same dynamic fashion as the single bundle graft. The grafts were tightened in a sequence consistent with Cuomo et al, the posterolateral graft tightened first at 30 degrees and the anteromedial graft tightened second at 75 degrees².

Extra-articular Release and ITB Tenodesis

The deep layers and capsule-osseous fibers of the ITB were released with a sharp knife. These layers were targeted for release because of their high co-injury rate in ACL deficient knees¹⁶. First, the skin was incised longitudinally over the ITB at the level of the joint line. The anterior border of the ITB was visualized and palpated. A blunt hemostat was used to dissect a plane between the superficial ITB fibers and the deep structures. A scalpel released deep ITB layers parallel to the joint line and capsule-osseous fibers from the femur without compromising posterolateral capsule, anterolateral capsule, lateral collateral ligament or popliteus tendon. Distally, the deep ITB fibers were released from the tibia down to, but not including Gerdy’s tubercle. Tubercle insertion remained intact.

The superficial ITB that had been left intact until this point was used as the graft material for the extra- articular tenodesis, originally described by Losee et al²³. A 15 mm wide by 120 mm long ITB graft was incised from the most anterior distal aspect of the ITB, with its insertion on Gerdy’s tubercle left intact. Fiberwire was tied with a Krachow stitch to the free proximal end of the dissected graft. A Steinman pin was used to identify the proper anchor point of the tenodesis ²³. This pin was located at the junction of the anterior border of the LCL and the superior edge of the popliteal tendon. The pin location was tested by placing the leg in external rotation, wrapping the graft around the pin, and then ranging the knee from 0 to 90 degrees of flexion. Acceptable pin placement was defined as no shortening of the graft greater than 2 mm during sagittal plane motion. The graft was anchored with a unicortical bone screw to the lateral femoral condyle under maximum tension.

Maximum tension during fixation of the extra-articular reconstruction was used to avoid variations between specimens related to graft tension. Overconstraint was not a concern in this study because the hypothesis was that the; increased laxity observed with ITB compromise would be corrected with a lateral extra-articular procedure. Although clinically, overconstraint is problematic, other studies have shown overconstraint can be avoided by decreasing tension on the graft during in vivo fixation²⁴.

Statistical Analysis

Multivariate linear mixed model analysis was used to test the effect on displacement and rotation of varying ITB states in an ACL reconstructed knee. From the fitted model, test of mean contrast was performed to test for mean change in displacement and rotation on ACL reconstructed specimens with ITB intact, ITB deficient, and ITB tenodesis. To account for the multiple tests performed (i.e. four pairwise comparisons), p-value was adjusted using Bonferroni’s method. These comparisons were done separately for single bundle and double bundle ACL reconstructions. The results of the statistical analyses are summarized in Tables 1-3.

Table 1.

Demonstrates the comparison between ACL intact, deficient and reconstructed states. (ACL = Anterior cruciate ligament, SD = Standard deviation, ATT = Anterior tibial translation, IR = Internal rotation, ITB = Iliotibial band, statistically significant p-value <0.05)

Effect of Single Bundle and Double Bundle ACL Reconstruction (with ITB Intact) on Displacement and Rotation
			Single Bundle					Double Bundle
Testing Condition	Knee Flexion	Displacement Measurement	ACLDeficient	ACL Intact	ACL Graft	Diff (ACL Intact - ACL Deficient)	Diff (ACL Intact - ACL Graft)	ACLDeficient	ACL Intact	ACL Graft	Diff (ACL Intact - ACL Deficient)	Diff (ACL Intact - ACL Graft)
			Mean (SD)	Mean (SD)	Mean (SD)	p-value	p-value	Mean (SD)	Mean (SD)	Mean (SD)	p-value	p-value
Neutral Drawer	30 degrees	ATT	10.7 (5.3)			0.013		9.2(2.2)			<0.001
							4.1 (1.1)			0.078		3.9 (1.9)			0.057
								6.7 (3.0)					2.8(1.5)
		90 degrees	ATT	4.7 (1.8)			0.103		5.6 (1.6)			0.004
					3.0 (1.0)			0.280		2.7 (1.1)			0.357
						3.6(1.1)					2.1 (1.6)
PNm Internal Torque	30 degrees	IR	29.6 (5.8)			<0.001		31.8 (4.1)			<0.001
				13.4 (4.7)			0.547		10.2 (1.6)			0.885
					14.0 (3.4)					10.6 (1.8)
	90 degrees	IR	21.4 (9.7)			0.021		26.8 (5.3)			<0.001
				12.9 (4.5)			0.700		9.1 (1.2)			0.116
					13.6 (3.2)					10.0 (3.1)
Coupled Anterior Drawer - Internal Torque	30 degrees	ATT	8.1 (3.1)			0.020		8.5 (2.5)			0.001
				5.3 (2.9)			0.361		5.0(2.1)			0.026
					6.7 (2.1)					3.5 (1.2)
		IR	29.0 (5.4)			<0.001		31.3 (4.1)			<0.001
				14.1 (4.3)			0.870		9.5 (2.2)			0.844
					13.7 (4.9)					9.5 (2.1)
	90 degrees	ATT	6.3 (2.9)			0.222		5.9 (2.2)			0.006
				5.0 (1.5)			1.000		3.8 (1.5)			0.081
					5.0(3.2)					2.9 (0.6)
		IR	23.9 (10.5)			0.018		25.6 (5.1)			<.001
				13.6 (2.9)			0.760		8.2 (2.4)			0.374
					13.3 (3.2)					9.0 (2.7)

Table 3.

Double bundle reconstruction was held constant, while the ITB state was varied. (ATT = Anterior tibial translation (mm), IR = internal rotation (degrees) , ACL = anterior cruciate ligament , ITB = iliotibial band, SD =standard deviation, CI = confidence interval, Bon adj = Bonferroni’s adjustment, statistically significant p-value <0.05)

Kinematic Effects of Varying the ITB State in Double Bundle ACL Reconstruction Knees
Testing Condition	Knee Flexion	Displacement Measurement	ITB State	Change due to ITB state
			ACL Graft/ ITB Intact	ACL Graft/ ITB Deficient	ACL Graft/ ITB Tenodesis	ITB Def - ITB Intact	ITB Tenodesis - ITB Def	ITB Tenodesis - ITB Intact
			Mean (SD)	Mean (SD)	Mean (SD)	Mean Diff (95% CI)	Bon adj p-value	Mean Diff (95% CI)	Bon adj p-value	Mean Diff (95% CI)	Bon adj p-value
Neutral Drawer	30 degrees	ATT	2.6 (0.4)	2.8 (0.7)	2.0 (0.5)	-0.2 (-1.6, 1.2)	=0.99	-0.6 (-1.9, 0.7)	0.972	-0.8 (-1.7, 0.1)	0.266
	90 degrees	ATT	2.1 (0.4)	2.1 (0.5)	1.7 (0.2)	0.0 (-1.2, 1.2)	=0.99	-0.4 (-1.4, 0.6)	=0.99	-0.4 (-1.0, 0.2)	0.621
SNm Internal Torque	30 degrees	IR	10.6 (0.8)	12.2 (1.3)	10.2 (1.1)	1.6 (-0.8, 4.0)	0.546	-2.0 (-5.4, 1.4)	0.67	-0.4 (-2.4, 1.6)	=0.99
	90 degrees	IR	10.0 (1.3)	13.4 (1.5)	7.5 (0.8)	3.4 (1.7, 5.1)	0.002	-5.9 (-8.9, -2.9)	0.002	-2.5 (-5.2, 0.2)	0.2
coupled Anterior Drawer - Internal Torqeo	30 degrees	ATT	3.5 (0.6)	4.6 (0.6)	3.0 (0.6)	1.1 (-0.1, 2.3)	0.207	-1.6 (-3.3, 0.1)	0.186	-0.5 (-2.3, 1.3)	=0.99
		IR	9.5 (1.0)	12.4 (1.0)	8.9 (0.9)	2.9 (1.1, 4.7)	0.009	-3.5 (-6.2, -0.8)	0.044	-0.6 (-3.5, 2.3)	=0.99
	90 degrees	ATT	2.9 (0.6)	3.3 (0.6)	2.7 (0.4)	0.4 (-0.6, 1.5)	=0.99	-0.6 (-2.0, 0.7)	=0.99	-0.2 (-1.5, 1.1)	=0.99
		IR	9.0 (1.1)	10.4 (1.2)	7.7 (0.7)	1.4 (-0.6, 3.5)	0.496	-2.7 (-5.5, 0.02)	0.154	-1.3 (-3.9, 1.3)	0.905

Results

Isolated ACL Reconstruction (Table 1)

Single Bundle

Under all of the variable loading conditions at 30 and 90 degrees of flexion, no anterior translational or internal rotational laxity measurement was statistically different when comparing the native ACL with single-bundle ACL reconstruction results. When assessing for laxity changes from the native ACL state to the ACL deficient state, all laxity measurements statistically increased except for the anterior tibial translation data at 90 degrees of flexion. These results demonstrated that knee laxity measurements increased with the cutting of the ACL in isolation and single-bundle reconstruction restored knee laxity to the level of the native ACL.

Double Bundle

The laxity measurements under the ACL deficient state were all statistically larger than the native ACL measurements. Following double-bundle reconstruction, all laxity measurements, under all loading conditions, were restored to non-statistically different levels as the native ACL. The only significant finding was overconstraint under the coupled anterior shear force-internal torque at 30 and 90 degrees of flexion. The double-bundle reconstruction was ableto restore anterior tibitd franslation and internal rotation regardless of simulated load to at least the level of laxity of the native ACL.

ACL Reconstructed/ITB state variable

Single Bundle Reconstructions Constant (Table 2)

Table 2.

Single Bundle ACL reconstruction was held constant and the ITB state was varied. (ATT = Anterior tibial translation (mm), IR = internal rotation (degrees), ACL = anterior cruciate ligament , ITB = iliotibial band, SD =stan- dard deviation, CI = confidence interval, Bon adj = Bonferroni’s adjustment, statistically significant p-value <0.05)

Kinematic Effects of Varying the ITB State in Single Bundle ACL Reconstruction Knees
Testing Condition	Knee Flexion	Displacement Measurement	ITB State			Change due to ITB state
			ACL Graft/ ITB Intact	ACL Graft/ ITB Deficient	ACL Graft/ ITB Tenodesis	ITB Def - ITB Intact	ITB Tenodesis - ITB Def	ITB Tenodesis - ITB Intact
			Mean (SD)	Mean (SD)	Mean (SD)	Mean Diff (95% CI)	Bon adj p-value	Mean Diff (95% CI)	Bon adj p-value	Mean Diff (95% CI)	Bon adj p-value
Neutral Drawer	30 degrees	ATT	6.7 (0.8)	7.1 (0.5)	6.4 (0.6)	0.4 (-1.3, 2.1)	=0.99	-0.7 (-2.2, 0.8)	0.972	-0.3 (-1.4, 0.8)	=0.99
	90 degrees	ATT	3.6 (0.5)	3.3 (0.6)	3.3 (0.3)	-0.3 (-1.7, 1.1)	=0.99	0.0 (-1.2, 1.2)	=0.99	-0.3 (-1.0, 0.5)	=0.99
5Nm Internal Torque	30 degrees	IR	14.0 (1.0)	18.4 (1.6)	10.9 (1.3)	4.4 (1.5,7.3)	0.016	-7.6(-11.6, -3.6)	0.003	-3.1 (-5.5, -0.8)	0.036
	90 degrees	IR	13.6 (1.5)	14.9 (1.8)	8.1 (0.9)	1.3 (-0.7, 3.2)	0.558	-6.7(-10.3, -3.2)	0.007	-5.4 (-8.6, -2.2)	0.002
Coupled Anterior Drawer - Internal Torque	30 degrees	ATT	6.7 (0.7)	7.4 (0.8)	6.0 (0.7)	0.7 (-0.7, 2.1)	0.917	-1.4 (-3.4, 0.6)	0.471	-0.7 (-2.9, 1.5)	=0.99
		IR	13.7 (1.1)	17.6 (1.2)	10.1 (1.1)	3.9 (1.7, 6.0)	0.004	-7.4(-10.7, -4.2)	0.001	-3.6 (-7.1, -0.1)	0.137
	90 degrees	ATT	5.0 (0.7)	4.6 (0.7)	3.4 (0.5)	-0.4 (-1.6, 0.8)	=0.99	-1.1 (-2.8, 0.5)	0.465	-1.6 (-3.1, 0.0)	0.142
		IR	13.3 (1.3)	14.7 (1.4)	9.6 (0.9)	1.4 (-0.9, 3.8)	0.666	-5.1 (-8.3, -1.9)	0.01	-3.7 (-6.7, -0.7)	0.058

Internal Torque

Under isolated internal torque, internal rotational laxity increased within the single bundle reconstruction subgroup from 14.0 degrees to 18.4 degrees (P=0.016) at 30 degrees of flexion when the ITB was released. Upon ITB tenodesis, internal rotation decreased to 10.9 degrees. The decreased rotation with tenodesis was significant when compared to when the ITB was intact (P=0.036) and when the ITB was deficient (P=0.003). At 90 degrees of flexion, ITB tenodesis demonstrated less internal rotation, 8.1 degrees, than when the ITB was deficient, 14.9 degrees, or when the ITB was intact 13.6 degrees (P=0.007 and P=0.002 respectively).

Coupled Anterior Drawer – Internal Torque

At 30 degrees of flexion, internal rotational laxity increased from 13.7 degrees to 17.6 degrees (P=0.004) with the release of the deep and capsule-osseous fibers of the ITB. The increase in internal rotation with the cutting of the ITB was not significant after the placement of the ITB tenodesis. Status-post ITB tenodesis, internal rotation decreased to 10.1 degrees, which was significant when compared to the ITB deficient state (P=0.001). At 90 degrees of flexion, an additional significant decrease from the ITB deficient state to the ITB tenodesis state was measured, 14.7 degrees to 9.6 degrees (P=0.01).

Double Bundle (Table 3)

Internal Torque

Internal rotation in the double bundle reconstruction model increased from 10.0 degrees to 13.4 degrees (P=0.002) at 90 degrees of flexion when the select fibers of the ITB were cut. This increase in rotation was significantly decreased to 7.5 degrees of internal rotation (P=0.002) with the ITB tenodesis.

Coupled Anterior Drawer – Internal Torque

The double bundle reconstruction subgroup demonstrated a similar laxity pattern under coupled anterior drawer-internal rotation as the single bundle subgroup under coupled anterior drawer-internal rotation at 30 degrees of flexion. The double bundle ACL reconstruction was not able to maintain internal rotational control when the ITB was compromised either. Internal rotation increased from 9.5 degrees to 12.4 degrees (P=0.009). The increase in rotation was decreased to 8.9 degrees (P=0.044) with ITB tenodesis.

Discussion

The role of the ACL as the primary rotational and anterior translational restraint of the knee is well proven^25-26. However, the ability to restore native stability to the knee following ACL injury has been more complicated. Continued instability, most specifically rotational laxity, and high long-term osteoarthritis rates in ACL reconstructed knees have led the push for more anatomic reconstruc- tions^12,27,28. Single-bundle reconstructions historically imitate the function of the anteromedial bundle, while intra-articular double-bundle reconstructions attempt to restore the static rotational control of the posterolateral bundle^6,7.

Double bundle reconsfruction hasbeen documented to decriose the eute ed the clinically impcbtndt pivot shift^29,30. While the pivot shift, a dynamic assessment of laxity that can be reproduced in the office, correlates more closely with a patient’s perceived stability³¹. Not all studies support the claim of double bundle superior- ity^32,33. Some biomechanical laboratory assessments have shown a difference in rotational control, but large clinical follow-ups have not shown a functional difference^27,34.

The results found within this cadaver model show that both single-bundle and double-bundle reconstruction techniques are able to restore static laxity under variable loading conditions to those of an uninjured native knee. The argument against single-bundle reconstruction is the inability to reconstruct the posterolateral bundle fibers. A single-bundle footprint of a ten mm diameter tunnel covers approximately 82% of the native femoral footprint, as observed by Harner et al³⁵. The femoral tunnel may be placed further lateral in an attempt to simulate the function of both native ACL bundles. Tsuda et al showed that a laterally based single tunnel could replicate both rotational and translational knee laxity measurements of double bundle reconstructions^27,36.

Single-bundle and double-bundle reconstructions are successful procedures with low rates of failure^12,37. The most common modes of failure, in order of occurrence, are traumatic re-injury, technical mal-position of the tunnels or fixation, and biologic failure of the graft³⁸. Misplaced tunnels during an ACL revision can be easily addressed, but fixation failure and failure of graft incorporation are more difficult to define as a specific cause. Although the ACL is the primary restraint to anterior tibial translation and internal rotation, secondary structures decrease the strain on the remodeling ACL, which is most important during the first six months after surgery when the graft is remodeling^39,40.

An extra-articular tenodesis functions as a secondary restraint. Engbretsen et al. showed that by adding an extra-articular ITB tenodesis, ACL forces could be reduced by up to 43%⁴¹. However, biomechanical data has been mixed about its role in restoring knee stability after ACL injury^25,42. Historical literature has shown that extra-articular tenodesis in addition to intra-articular ACL reconstructions for chronic ACL tears have improved return to level of activity, improved anterior laxity, and improved overall functional scores⁴³.

Recent clinical five year follow-up has demonstrated faster return to sport, less kneeling pain, and high capacity to return to normal muscle trophism with coupled intra- and extra-articular procedures²⁹. While long term concerns of lateral compartment arthritis secondary to overload with overtightening of an extra- articular tenodesis exist, this was not observed in a 10-13 year follow-up by Marcucci et al.¹². Pernin et al., at mean 24.5 year follow-up of intra-articular ACL reconstruction and extra-articular tenodesis, noted that radiographic joint space narrowing was directly attributed to meniscal and chondral damage observed at the time of original surgery. If meniscus and cartilage surfaces appeared uninjured at the time of ACL injury, only 38% of patients demonstrated radiographic signs of osteoarthritis at long-term follow-up ³⁷.

Assuming the laboratory data showing that ITB deficiency leads to increased rotational laxity in the ACL reconstructed knee can be extrapolated to a clinical application, these results still do not mean that all ACL injuries need an adjunctive ITB reconstruction. Rather, it shows that the ITB may have a greater role in post-operative ACL reconstruction kinematics than previously thought. When these conclusions are applied to Terry et al.’s findings that 93% of clinically damaged ACLs also have ITB injury it presents a potential frequently overlooked point.

If ITB injuries occur at Terry et al.’s reported rate and the ITB plays a significant role in rotational control following ACL+ITB injuries, why are there not more obvious problems with instability after ACL reconstruction? The answer is more likely multi-factorial. First, the natural healing process of the ITB following injury is not completely understood. If the immobilization period following ACL reconstruction is enough time for the ITB to naturally heal, then increased rotational laxity would not be appreciated when the patient is allowed to return to activity. Second, the ITB injury in this laboratory study was extensive. The degree of ITB injury in most ACL patients might not be as severe, thus lessening the frequency of increased rotational laxity seen clinically. Third, the rotational changes measured in the lab, despite statistical significance, may not be clinically significant. The significant increases in internal rotation were only three to four degrees. This might be too small of an increase in laxity for a patient to note a problem. Additionally, the patient’s perception of instability does not always align with its mechanical function considering laxity is a quantitative outcome, while instability is a patient’s perceived qualitative outcome.

While anterolateral reconstructions are not performed on a regular basis in conjunction with an intra-articular ACL reconstruction, the procedure does have its place. McGuire et al. describes three specific indications: failed ACL reconstruction with proper technique and follow-up, persistent laxity in a knee with prior lateral reconstruction in conjunction with intra-articular ACL reconstruction, and knee dislocation¹³. The first indication is the one that pertains most to this discussion. ITB injury does not appear to have a large impact on the majority of ACL reconstruction patients, despite reported high co-injury rate. This is seen in the high success rate of primary ACL reconstruction. However, when patients have complications status post ACL reconstruction and the intra-articular procedure appears to be well performed, other sources of injury need to be evaluated. In these circumstances, it would be appropriate to evaluate the ITB more thoroughly and consider an extra-articular procedure.

The design of this cadaver-based laboratory model lent itself to numerous limitations. One issue being what is a significant ITB injury? The capsule-osseous fibers and the deep layer of the ITB were released in this study. These are the two most common layers injured per Terry et al., however, how often clinically are both of these layers completely torn¹⁶? If the ITB was over released in comparison to what occurs clinically, this study’s results would over estimate the potential laxity following ITB injury. Another limitation to this study was it being an in vitro cadaver model. Potential inconsistencies of cadaver tissue, lack of active knee restraints, and drying of tissues during the lengthy protocol present possible sources of error. The synthetic rope used for the ACL grafts, both single and double-bundle, provided consistency, yet synthetic grafts are not used clinically⁴⁴. Obtaining autograft for each specimen would have added additional variability secondary to tissue obtained, technique used for acquisition, and length of time added to procedure. Another potential issue with the chosen graft technique was that it could potentially slip since it was not rigidly fixed to the tibia or femur. However, double clamping with clamps perpendicular to each other, clamping directly against bone, as well as repeated graft tension evaluation decreased this risk.

Computer assisted surgery has been well documented as a reliable option for reproducible tunnel placement and laxity measurements^45-47. Also, in a prospective randomized control trial, there was no measurable functional difference between ACL reconstructions done with navigation or manual tunnel placement by the surgeon⁴⁸.

Conclusion

In a cadaver-based laboratory model, single-bundle and double-bundle intra-articular reconstructions were both able to restore internal rotation and anterior tibial translation to at least native knee laxity levels after an isolated ACL injury. When the ACL reconstructions were held constant and the ITB deep capsule-osseous fibers and deep fibers connecting to the proximal tibia were released, internal rotation significantly increased under multiple testing conditions in both the single-bundle and double-bundle subgroups. The increase in laxity demonstrated with ITB release was eliminated with ITB tenodesis.