Patients
Patients for this study were recruited between August 2001 and August 2004 in Orton Orthopaedic Hospital Helsinki, Finland. Every patient with an ACL injury was chronologically recruited, there were no refusals. We did not collect or record the numbers nor reasons of the excluded cases.
120 patients with fresh or chronic ACL injury were selected and they were randomized into 4 treatment groups, with different fixation methods. Every patient was operated on with autograft hamstring tendon reconstruction. The median age in different groups was; group I 31, group II 29, Group III 35, and group IV 32 years, respectively. The ratio of female/male patients in groups was 11/19, 11/19, 12/18, and 10/20, respectively. The median interval from injury to surgery was from 3 to 4 months. (Table 1)
TABLE 1 Demographic Variables in ACL Reconstruction in 4 Randomized Groups [5].
|
Rigidfix/
Intrafix
|
Rigidfix/
BioScrew
|
BioScrew/
Intrafix
|
BioScrew/
BioScrew
|
P value
|
Median age, y (range)
|
31
(18-50)
|
29
(18-50)
|
35
(20-48)
|
32
(18-49)
|
NS
|
Gender, F/M
|
11/19
|
11/19
|
12/18
|
10/20
|
NS
|
Median interval from injury to surgery (range)
|
4 mo
(1 wk – 10 mo)
|
4 mo
(1 wk – 2 y, 8 mo)
|
3.5 mo
(1 wk – 2 y, 11 mo)
|
3 mo
(1 wk – 8 y, 3 mo)
|
NS
|
Medial/lateral menisci resection
|
5/2
|
8/3
|
7/4
|
7/3
|
NS
|
Medial/lateral menisci repair
|
|
|
|
/1*
|
NS
|
NS not significant, F female, M male.
* Protocol deviation.
The results of 2 years of follow-up were published in American Journal of Sports Medicine in 2009. No statistically significant differences between the groups were seen 2 years postoperatively [5].
As the basis for the randomization and fixation of the graft, three devices were chosen. Femoral fixation was performed by one of the two devices, either 2 Rigidfix bioabsorbable rods (PLLA) or a BioScrew, a L-lactic acid, bioabsorbable interference screw. The length of the PLLA rod was 42 mm and the diameter was 3.3 mm. Fixation to the tibia was done with a nonabsorbable Intrafix device, which consists of an expansion sheath made from high-density polyethylene and a tapered expansion screw molded from Delrin (DePuy, Mitek), or a BioScrew. In group I, femoral Rigidfix cross-pins and a tibial expansion sheath with a tapered expansion screw (Intrafix, n=30) were used. In group II, femoral Rigidfix and tibial interference screw fixation (BioScrew, n=30) were used. In group III, femoral BioScrew and tibial Intrafix fixation (n=30) was done. In group IV, BioScrew fixation into both tunnels (n=30) was employed.
According to the study protocol, the inclusion criteria were fresh or chronic (less than 5 years old) unilateral ACL tears in female or male patients, with an age range of 18 to 50 years. Additional injuries could be an uncomplicated meniscal lesion, small tears (flaps) which were easily dealt with and minor chondral damage, at the most grade 2 Outerbridge small lesions. These injuries were thought not to have any impact on the clinical outcome.
Randomization was done by sealed, numbered envelopes containing information of the treatment group, and the patients as well as physiotherapists were blinded to the method used. Numbers were generated by a computer program using block randomization to ensure equal distribution of patients into each group. The diagnosis of an ACL tear was made clinically, by magnetic resonance imaging (MRI) examination when needed, and confirmed at arthrocsopy. The envelope was not opened until arthroscopy had confirmed the diagnosis after graft harvesting was completed. The two surgeons (AH, JS) were obviously not blinded to the method used.
To exclude a case the interval between the injury and surgery needed to exceed 5 years or be a revision procedure. Concomitant grade II to III collateral or posterior cruciate ligament tears and peripherally detached meniscal tears to be repaired as well as Outerbridge 3 to 4 chondral damage and arthrosis of the knee also excluded the case. Detached meniscal lesions need to be refixed and they were excluded as it would be difficult to differentiate symptoms related to the refixation from those of ACL reconstruction.
Surgery
A diagnostic arthroscopy and minor menisci procedures were performed; 40 meniscal resections and one meniscal repair. After this, the two hamstring tendons (STG) were harvested through a vertical incision situated medial to the tibial tuberosity. Harvesting was made with a Linvatec graft harvester. The grafts were of a diameter from 7 mm to 10 mm. One graft diameter in the Rigidfix / Intrafix group was not recorded. (Table 2)
TABLE 2 The Graft Diameter Between the Four Groups.
Group
|
10 mm
|
9 mm
|
8 mm
|
7 mm
|
|
Rigidfix/Intrafix
|
0
|
6
|
17
|
6
|
29
|
Rigidfix/Bioscrew
|
0
|
9
|
16
|
5
|
30
|
Bioscrew/Intrafix
|
1
|
6
|
19
|
4
|
30
|
Bioscrew/Bioscrew
|
0
|
4
|
25
|
1
|
30
|
After opening the treatment allocation envelope, the grafts were individually prepared for the randomized, specific fixation methods.
The drill tunnels with femoral Rigidfix fixation were made transtibially into the posterior fourth quadrant of the femoral condyle in the sagittal plane and into 10:30-o’clock position in the right and 1:30-o’clock position in the left knee in the frontal plane. In the tibia, sagittally, the drill guide was positioned intra-articularly at the ACL footprint with the drill guide targeted in the second quadrant from anterior to posterior. No C-arm was used.
For the Rigidfix femoral fixation groups the graft was constructed, according to the manufacturer’s instructions, using whipstiches of No. 1 Vicryl (Ethicon Inc, Johnson&Johnson, Somerville, New Jersey) to join the doubled limbs of the semitendinosus and gracilis tendons together. The tunnels were made with the use of a drill guide, the depth of the transtibially drilled femoral tunnel was 30 mm to 40 mm. With the Rigidfix instrumentation, 2 transverse tunnels were drilled to receive the fixation devices. After the graft was passed into the drill tunnel by pulling it with No. 2 Vicryl loop left in the graft end, 2 Rigidfix implants were tapped through the drill guide sleeves transfixing the graft and advanced to the bone medial to the drill tunnel (Figure 1).
For the Intrafix tibial fixation, No. 1 absorbable whipstich in the graft ends was used. The graft was spread into 4 quadrants between the sleeve and the drill tunnel. After cycling the knee 10 to 15 times, the graft was tightened. The expansion sleeve and the screw were introduced concentrically by which the 4 limbs of the graft could be compressed between the bone tunnel and the device (Figure 1). The screws used were of three different sizes: if the graft was less than or equal to 8 mm, the 6 mm to 8 mm screw was used (7-9 mm screw if the bone quality was suboptimal), and if the graft was larger than 8 mm, the 8 mm to 10 mm screw was used.
In the BioScrew fixation, the femoral drill tunnel was performed with the “outside-in” technique using the rear-entry guide (Linvatec). A tight whipstich with No. 1 Vicryl was used to join the graft limbs together to form a single bundle. The bioabsorbable interference screw of 30 mm length and diameter equal to the graft and tunnel was inserted eccentrically, compressing the graft agaist the bony tunnel wall (Figure 1).
If the bone was soft and the surgeon felt that the screw did not tighten sufficiently, the fixation was reinforced with a resorbable button by which the whipstich sutures were secured outside the femoral and/or tibial tunnel. If the graft ends protruded from the tibial tunnel entrance, the screw fixation was supplemented with a staple (Richards; Smith&Nephew, Memphis, Tennessee). (Table 3)
TABLE 3 Back up Fixations in the Four Study Groups.
Back up fixation
Group# / Name
|
Button
Femur / Tibia
|
Staple
Femur / Tibia
|
I / Rigidfix/Intrafix
|
0 / 0
|
1 / 0
|
II / Rigidfix/Bioscrew
|
1 / 2
|
0 / 27
|
III / Bioscrew/Intrafix
|
14 / 0
|
0 / 0
|
IV / Bioscrew/Bioscrew
|
9 / 13
|
2 / 14
|
Before the tibial fixation the knee was cycled 10-15 times through the range of motion in all 4 groups. The graft was then tightened to 40-N force, as instructed by the manufacturer, with the Mitek tensioning device (Intrafix fixation), and in the BioScrew fixation a manual estimate of 40-N was done. The knee was held as commonly recommended in 30° flexion, and the distal ends of the graft were secured according to the group specified.
Post-operative care
Postoperatively no knee brace was used. Knees were mobilized immediately in the same way in all 4 groups. Crutches were in use for the first 4 weeks. Full weightbearing was allowed after 2 weeks. Active quadriceps activity, e.g. exercise cycle and exercises in water, was allowed from 3 to 4 weeks after the operation. Straigth leg raise was allowed at 6 weeks and jogging at 12 weeks after the operation. Return to sports was allowed after 6 months, contact sports and downhill skiing from 9 months postoperatively. After one year postoperatively there were no restrictions. The physiotherapist used closed kinetic chain muscle rehabilitation principles in patient guiding.
Evaluation methods
Clinical examination was performed preoperatively and at 1, 2, and a minimum of 5 years postoperatively by the not blinded two surgeons (AH, JS). The Lachman and pivot-shift tests were graded negative I (pivot-shift: glide), slightly positive II (pivot-shift: shift), and clearly positive III with no end point (pivot-shift: clunk). Anteroposterior knee laxity (side-to-side difference with 35-pound and manual maximum force; KT-2000 arthrometer, MEDmetric Corporation, San Diego, California) and isokinetic peak muscle torques of the knee extensors and flexors at 60deg/s and 180deg/s (Lido MultiJoint II, West Sacramento, California) were measured, also. The knee scores were completed by the patients: Tegner activity level was used to define the actual physical performance (1-10), for subjective evaluation Lysholm knee scores (0-100) was used. The subjective International Knee Documentation Committee (IKDC) score was calculated based on the information gathered by the 2000 IKDC KNEE EXAMINATION form. For another subjective evaluation the “patellofemoral scores” (KPS 0-100) was used, also.
Statistical Analysis
Statistical analysis was done using the BMDP statistical package (Statistical Solutions Ltd, Cork, Ireland). The parametric data between the 4 groups were evaluated with analysis of variance. The nonparametric data were evaluated with chi-square (between the groups) and with McNemar’s or sign test (comparison over time within a group). The minimum level of significance was P=.05. Lysholm score was one of the main outcome measurements used. With a difference of 10 points, a clinically significant difference between the treated groups would be acknowledged. When type I error (alfa) was set at 0.05 and power (beta error, power 1-beta) at 80%, sufficient total computerized sample size would be 54 for 2 treatment groups, as calculated by the computer program. Based on our power calculations (Lysholm, 10 points difference, SD 13) the planned minimum sample size was 4 x 27 = 108 patients. 102 patients participated in the 5-year follow-up and the SD of all patients combined in Lysholm score was 12 i.e. slightly smaller than in our power calculations. Therefore, the power of our study was on same level as planned and the risk of type II error was not evident.