- Email: [email protected]
Semitendinosus muscle in anterior cruciate ligament surgery
Semitendinosus Muscle in Anterior Cruciate Ligament Surgery: Morphology and Function Karl Eriksson, M.D., Per Hamberg, M.D., Ph.D., Eva Jansson M.D., Ph.D., Hans Larsson, M.D., Adel Shalabi, M.D., and Torsten Wredmark, M.D., Ph.D.
To evaluate the fate of the hamstring muscles in general and the semitendinosus muscle in particular, after anterior cruciate ligament (ACL) reconstruction with an autologous semitendinosus tendon graft from the ipsilateral side. Type of Study: Prospective consecutive case series investigation. Methods: Included were I6 consecutive patients, I4 male and 2 female. with a mean age of 26 years. The inclusion criterion was chronic unilateral ACL insufficiency with no concomitant knee ligament injuries. ACL reconstruction was performed with a quadruple semitendinosus tendon graft using the EndoButton technique (Acufex. Mansfield. MA). Intraoperatively, muscle specimens were taken from the semitendinosus muscle on the harvested side. Follow-up at a minimum of 6 months included clinical examination, isokinetic strength performance, magnetic resonance imaging (MRI) of the thigh and knee. and ultrasound-guided muscle biopsy procurement from the semitendinosus muscle for histochemical and enzymatic analyses. Results: Of the patients. 75% showed regeneration of their semitendinosus tendons. The neotendons all inserted below the knee joint where they had fused with the gracilis tendon to a conjoined tendon inserting in the pes anserinus. The semitendinosus muscle had a smaller cross-sectional area on the operated side but none showed total atrophy. Less atrophy was present in the patients with a regenerated semitendinosus neotendon compared with those without regeneration (P = ,029). In the latter group the semimembranosus muscle seemed to compensate for this with hypertrophy (P = ,019). Cross-sectional muscle fiber areas, the relative number of each fiber type and oxidative potential as estimated by citrate synthase activity, showed no significant differences between the operated and nonoperated legs. The isokinetic strength of the hamstrings and quadriceps was significantly lower in the operated leg than in the nonoperated leg. Conclusions: With this surgical technique, the semitendinosus muscle can recover and the tendon has. according to the MRI images, a great potential to regenerate after its removal. Key Words: ACL reconstruction-Semitendinosus muscle-Regeneration-Histochemistry-MRI-Isokinetic strength.
Purpose:
I
n recent years, hamstring tendon graft has become an alternative to a bone-patellar tendon-bone (BPTB) graft in anterior cruciate ligament (ACL) re-
From the Departments of Orthopedics (K. E.. P. H.) and Radiology (H. L.). Stockholm Siider Hospital. Karolinska Institute; and the Departments of Clinical Physiology (E.J. j. Radiology (AS.). and Orthopedics (T. W.). Huddinge Universityv Hospital, Kamlinska Institute. Stockholm, Sweden. Supported by Grant No. 4494 from The Sb+*edish Medical Research Council. and grants from the Swedish Center for Reseurch in Sports and Stockholm Siider Hospital Research Foundation. Presented as part of Dr. Eriksson ‘s thesis, Karolinska Institute. Stockholm. Defended February 2. 2001. Address correspondence and reprint requests to Karl Eriksson. M.D., Department of Orthopedics, Stockholm Siider Hospital. Karolinska Institute. S-l 18 83 Stockholm, Sweden. E-mail: kalle. [email protected] 0 2001 by the Arthroscopy Association of North America 0749.8063/01/1708-2643$35.00/O doi:l0.1053/jnr.s.200/.20961
808
Arthroscopy:
The Journal
of Arthroscopic
and Related
construction surgery.i-x Many physicians criticize the high donor site morbidity when using a BPTB graft.“-ij The purported advantages of a semitendinosus graft material include lower donor-site morbidity and possibly less anterior knee pain,*,‘-5 a stronger graftih.17 because of the larger diameter than BPTB,3 and less quadriceps muscle strength deficit because the extensor mechanism is less affected.ia-ltJ Regardless of which autologous graft is chosen, donor-site morbidity will occur in some patients. Until now, only a few prospective randomized studies have focused on this issue. Yasuda et al.” reported no longterm differences in hamstring or quadriceps strength when comparing the autologous hamstring (semitendinosus and gracilis) tendons from the ipsilateral and contralateral sides. Likewise, neither Lipscomb et al.)” nor Maeda et al.’ found any reduction in hamstring strength in patients who underwent reconstruction
Surgetyv.
Vol 17. No 8 (October).
2001:
pp 80R-RI7
SEMITENDINOSUS
MUSCLE
with ipsilateral hamstring tendons. However. Marder et al.-’ reported that hamstring strength was lower with ipsilateral semitendinosus and gracilis grafts than with an ipsilateral BPTB graft. In these studies of isokinetic muscle strength of the hamstring muscles. both the semitendinosus and the gracilis tendons had been used for an autologous graft. In a recent study by Ohkoshi et al..” no differences in presurgical and postsurgical isokinetic peak torque values were seen in the hamstring muscles when the semitendinosus tendon was harvested from the contralateral leg. A few earlier magnetic resonance imaging (MRI) studies show that the semitendinosus tendon can regenerate. after being harvested for ligament reconstruction.‘J-‘5 The aims of the present study were to prospectively evaluate a patient group in which only the ipsilateral semitendinosus tendon had been harvested for an autologous graft. The data included results of clinical examination. isokinetic strength testing of the thigh. a biopsy examination of the semitendinosus muscle. and MRI. To our knowledge, no other study has described the histochemical. morphologic, and biochemical changes in hamstring muscles following ACL reconstruction. Indeed, little is known about the histochemistry, morphology and biochemistry of this muscle group.
METHODS Patients Sixteen patients. I4 male and 2 female, with a mean age of 26 years (range. I7 to 34 years) years were prospectively included in the study. The inclusion criterion was unilateral ACL injury with no concomitant ligament injuries. The median time from ACL trauma to reconstruction was 8 months (range, 3 to 24 months). No prior surgery except arthroscopic meniscus surgery had been performed on the involved knee. All patients underwent reconstruction with an ipsilatera1 quadruple semitendinosus tendon graft, using the EndoButton technique (Acufex, Mansfield, MA) described by Rosenberg and Graf.‘6 The right knee was involved in 9 cases and the left in 7. In 2 patients, because of chondral lesions, autologous osteochondral grafting was performed concomitantly with the ACL reconstruction. Both of these lesions were less than I cm’. Preoperative radiographs showed no arthritic changes in any of the knees. Nine patients (56%) had undergone partial meniscectomy (5 medial and 5 lateral) at their initial arthroscopy following ACL injury. Another 2 patients
IN ACL SURGERY
809
underwent partial medial meniscectomy at their reconstruction. leaving 5 patients (31%) without meniscus lesions at the time of reconstruction. The graft was harvested as described previously by Eriksson et al.‘j Before final preparation of the graft, muscle specimens attached to the musculotendinous junction were taken from the proximal part of the resected tendon at a distance of about 30 cm proximal to the tendon insertion on the tibia. A standardized rehabilitation protocol was used, including early rangeof-motion training. full weight bearing, and closedchain exercises. At a median of 7 months (range, 6 to I2 months), all patients underwent MRI scans of the knee and thigh. Ten months (range. 8 to I8 months) postoperatively. the evaluation was completed with a clinical examination, including an isokinetic muscle performance test. and percutaneous ultrasound-guided needle biopsy of the semitendinosus muscle. The nonoperated side was also examined as a reference. except for the control biopsies, which were taken only in 8 patients. One patient had an early complication with deep infection and was successfully reoperated on with an ipsilateral BPTB graft 5 months after the initial reconstruction. Therefore, his data from the isokinetic test as well as the thigh and quadriceps cross-sectional areas on MRI scans were excluded because they were also affected by the second operation. His results from the biopsy and MRI evaluation of the semitendinosus were included, however, because the primary failure did not affect tendon regeneration after his primary operation. The study was approved by the ethics committee of the Karolinska Institute.
MRI The MRI images were taken with a C-shaped 0.2Tesla resistive magnet (Magnetom Open; Siemens, Erlangen, Germany). T I - and T2-weighted transaxial MRI sequences of the thigh and knee joint were used and. in a few cases, additional sagittal sequences were made. In all images, the knee was held in full extension with an external rotation of 15” to 20”. T2weighted images were used to calculate semitendinosus, semimembranosus, and total muscle areas and pixel values over the semitendinosus, with TR (repetition time) of 4,914 milliseconds, TE (echo time) of 102 milliseconds, FOV (field-of-view) of 240 X 380 mm, matrix size of 126 X 256 pixels, and a slice thickness of 15 mm. Tl-weighted images of the knee joint were used for tendon evaluations with TR of 920 milliseconds, TE of 26 milliseconds, FOV of 150 X
810
K. ERIKSSON ET AL.
200 mm, matrix size of 144 X 256 pixels, and a slice thickness of 10 mm. In one patient, Tl-weighted images were used over the proximal thigh and knee joint. The semitendinosus tendon area and pixel values were calculated approximately 5 cm proximal to the joint line and the muscle area and muscle pixel determinations were made approximately 12 to 14 cm below the pubic bone, which corresponds to about 30 cm proximal to the pes anserinus. The corresponding values on the nonoperated leg were always determined at the same level. The MRI scans were evaluated by an experienced MRI radiologist. All images were analyzed twice at an interval of 3 months to ensure that the observer would not be influenced by the previous readings. The values given are the mean of these evaluations. The intraclass correlation (ICC) between the 2 readings was 0.7 to 0.9, which indicates a good correlation. Semitendinosus
Biopsy Analyses
The biopsy specimens of muscle taken intraoperatively from the leg were instantly frozen with Freon and stored at -70°C until all postoperative and control samples had been taken. An ultrasound-guided percutaneous needle biopsy technique was used for the postoperative and control specimens, using a 2. l-mm Bard Magnum needle (C.R. Bard, Covington, GA) that was fired 2 to 5 times, which procured approximately 1 to 5 mg wet weight of muscle per firing. In 2 cases, the needle position in the semitendinosus muscle belly was confirmed by simultaneous computed tomography scans. Before the biopsy, the patient was placed in the prone position, the semitendinosus muscle was identified with ultrasound and marked at the level corresponding to the site of the intraoperative sample, i.e., approximately 30 cm proximal to the pes anserinus. After local anesthetic infiltration of the skin and subcutaneous tissue, but not the muscle, a 3-mm skin incision was made, allowing the biopsy needle to approach the fascia easily. The needle was then fired under ultrasound guidance with free-hand technique and samples were collected and instantly frozen in isopentane precooled by liquid nitrogen and stored at -70°C pending analysis. Eight patients gave their approval to sampling from both the nonoperated and the operated legs. All samples (intraoperative, postoperative, and controls from the nonoperated leg) were analyzed at one time by the same examiner. Serial transverse sections (10 mm) were cut with a microtome at -20°C and stained for myofibrillar ATPase, after
preincubation at different pH levels in acid and alkaline buffers.“’ The fibers were classified on the basis of their staining intensity (myofbrillar ATPase) into fiber types I, 2A, 2B. and 2C. A computerized image analysis system (Scan-Beam; BioRad. Hadsund, Denmark) was used to calculate the relative distribution of fiber types in number (%) and cross-sectional fiber area for each fiber type. The average number of fibers per sampling occasion was 205 in the intraoperative specimen and 150 in the postoperative and control specimens. Fiber areas were measured on the stains preincubated at pH 4.6. The remaining portion of the biopsy samples was freeze-dried. dissected free from visible blood, extracellular fat, and connective tissue, under a dissection microscope. Citrate synthase activity was determined by a fluorometric method at 25°C using the method of Lowry and Passonneau after homogenization of the dried and weighed tissue in 0.1 mol/L of ice-cooled phosphate buffer (pH 7.7) with 0.5% bovine serum albumin (dilution 1 mg dry muscle in 400 mL buffer).Zx Because of freezing artifacts, only 12 of the 16 patients were represented with samples allowing fiber typing intraoperatively, 14 postoperatively, and, in the control group, all 8 patients had samples adequate for histochemical analysis. Enzyme determinations were made in all 16 patients intraoperatively, in all 15 patients postoperatively. and in all 8 controls. Clinical and Isokinetic
Strength
Evaluations
Fifteen patients underwent a standardized test protocol performed by an independent examiner, a physical therapist, including l-leg hop test, range-of-motion evaluation, and measurement of thigh circumference 15 cm proximal to the joint line and maximum calf muscle width. The Lysholm score was noted. Sagittal stability was determined with the Lachman test and the Stryker laxity tester (OSI; Stryker, Kalamazoo, MI), evaluating side-to-side difference with 89 N anterior force. Isokinetic concentric and eccentric hamstring muscle strength and concentric quadriceps strength were measured with a Kin-corn (Chattex, Chattanooga, TN). The test consisted of a lo-minute warm-up period on a cycle ergometer at 50 W, and 5 submaximal and 2 maximal isokinetic hamstring contractions on the Kin-corn. After a l-minute rest, the recording started with sitting thigh curls. Concentric and eccentric peak torque (Newton meters) between flexion angles of 45” to 90” were measured at an angle velocity of 90” per second. The test was performed
SEMITENDINOSUS
MUSCLE
IN ACL SURGERY
811
ST Muscle Operated Leg Regenerated n = 12 Nonregenerated n = J All patients n = I6 NOTE.
Values
are mean
ST Muscle Nonoperated
I l.S3 (7.65) 7.96 (2.47)
12.x9 (25X) 12.0 (3.601
IO.61
12.67 (2.76)
(2.98)
and standard
Leg
Ratio ST Operated/ Nonoperated Leg
SM Muscle Operated Leg
0.89
(0.08)
13.03
(4.9s)
0.66 0.84
(0.2 I ) (0.11)
14.33 13.37
(4.26) (4.66)
SM Muscle Nonoperated Leg 13.35 I I.57 12.88
(4.34) (3.76) (4.14)
Ratio SM Operated/ Nonoperated Leg 1.03 (0.14) I .26 (0.13) I .05 (0.23)
deviation.
with consecutive maximal torques (usually 3 to 5) and ended when the subject failed to reach the previous maximal value. An isokinetic concentric quadriceps contraction from 90” to 5” of knee flexion at 90” per second was also performed. One patient did his tests on a Biodex isokinetic machine (Biodex Corp. New York, NY) instead of Kin-corn because of a technical failure of the Kin-corn. and the same test protocol was used. In all patients, the nonoperated leg was examined for comparison. In addition to peak values in Newton meters, total work in Joules (area under the curve) was calculated in each test. Statistics All variables were subject to conventional descriptive statistics (mean, standard deviation for normally distributed variables, median and range for ordinal or skewed variables). The problem of multiplicity was discussed. Because of the exploratory nature of this study. any statistical significance was followed by a discussion of the clinical significance. No method for dealing with missing values was applied. If the variables were normally distributed, a parametric test was used, i.e.. all variables except MRI pixel values and Lysholm scores where nonparametric tests were used. The difference between preoperative and postoperative values, as well as operated and nonoperated leg, were analyzed by using the Student t test for dependent samples. The difference between patients with a regenerated semitendinosus tendon and those with no regeneration after surgery was analyzed by analysis of variance (ANOVA) with repeated measures of the preoperative and postoperative difference. This ANOVA describes the interaction between the preoperative and postoperative values and regeneration and no regeneration. The relationship between operated leg and nonoperated leg was in some variables described as a quotient and analyzed with the Student t test for independent samples. For preoperative postoperative difference and operated leg versus nonoper-
ated leg, the Wilcoxon rank sum test for dependent samples was used. The Mann-Whitney U test for independent samples was used for analysis of difference between the patients with regenerated and nonregenerated semitendinosus tendon. Correlations were calculated using the Pearson 2-tailed test. ICC with a 2-way mixed model with average measure was used for a measure of agreement between the 2 readings of the MRI examinations. The overall level of significance was set to P < .05. RESULTS MRI Evaluation Anatomy: Twelve of the 16 patients (11 males and 1 female) showed regeneration of the semitendinosus tendon (i.e., neotendon) with normal topography to and below the level of the joint line. At various distances 1 to 3 cm below the joint line, the neotendons had fused with that of the gracilis to form a conjoined tendon structure, which inserted in the pes anserinus. In 1 female patient, the neotendon had fused with the semimembranosus tendon proximal to the joint line. In the 3 remaining patients, no obvious tendon structure was seen distal to the semitendinosus muscle belly. However, no signs of retraction of the muscle bellies were noted. Of the 2 patients with concomitant osteochondral grafting at the time of reconstruction, 1 had a regenerated neotendon and the other did not. Area Determinations: The regenerated semitendinosus neotendon cross-sectional area was 0.27 2 0.08 cm’, and the area on the nonoperated leg was 0.18 + 0.03 cm’. There was no significant difference between the 2 sides (P = .14). The gracilis tendon areas were the same on both sides. The total thigh muscle area of the operated leg was 4% less than in the nonoperated leg (177 + 28 cm* v 185 + 31 cm*, P = .006). No significant difference in area ratio (operated/nonoperated leg) was found between regenerated and nonregenerated cases. The
812
K. ERIKSSON ET AL.
quadriceps muscle cross-sectional area on the operated side was 10% smaller than on the nonoperated side (P = .OOl), but there was no significant difference (ratio, 0.99 -+ 0.04, P = Sl) between the total hamstring area on the operated and nonoperated side. The semitendinosus muscle cross-sectional area was 16% smaller on the operated than on the nonoperated side (Table 1, P = .OOl). There was no statistically significant difference between the semimembranosus muscle cross-sectional area in the operated and nonoperated side when evaluating all patients (regenerated and nonregenerated). The ratios of the semitendinosus and semimembranosus muscle crosssectional areas between operated and nonoperated sides in the regenerated and the nonregenerated tendon groups are shown in Fig 1. Patients with no regeneration had a relatively larger semimembranosus muscle on the operated side compared with the nonoperated side (P = ,019). Pixel Values (Signal Intensity): The pixel values of the regenerated neotendons did not differ significantly from those of the nonoperated legs. Nor was there any difference between the gracilis tendons in the operated and nonoperated legs. The pixel values of the semitendinosus muscle in both legs are shown in Table 2. The 1 patient examined with Tl-weighted sequences over the thighs had higher levels in both legs, which explains the wider range in the nonregenerated group. No significant difference in pixel values was seen in the semimembranosus muscle on the operated and nonoperated side. Muscle Biopsy Evaluation No significant differences were found in citrate synthase activity, fiber type composition, or fiber type areas before and after the surgery in the operated leg. Nor were there any differences in these variables between the operated and nonoperated legs after surgery (8 subjects) (Tables 3 and 4). No differences were noted between the regenerated and nonregenerated groups in the variables studied. Using the t test on an intraindividual basis (n = 7), the mean fiber area in the postoperative samples in the operated legs tended to be smaller than the postoperative controls in the nonoperated legs (6,304 v 7,021 pm*, P = .16). Clinical Examination
and Isokinetic
Strength
The Lysholm score was 71 (range, 40-86) before surgery and 91 (range, 70-100) after surgery, a significant improvement (P < .OOl). The Stryker laxity test with 89 N showed a mean side-to-side difference
FIGURE 1. (A) Ratio of cross-sectional area in semimembranosus muscle of operated and nonoperated side, and nonregenerated and regenerated tendons, respectively. The area is relatively greater on the operated side in the cases with no regeneration (P = ,019). The median is indicated in the boxes, containing 50% of the values, The range is seen outside the squares. (B) Ratio of cross-sectional area in semitendinosus muscle of operated and nonoperated leg, and nonregenerated and regenerated tendons. respectively. The area is greater in the regenerated cases (P = ,029). The median is shown in the boxes, containing 50% of the values. The range is seen outside the squares.
preoperatively of 4 + 1.68 mm. Postoperatively, the side-to-side difference was 1.27 + 1 .I mm and the preoperative to postoperative laxity difference had improved significantly (P < .OOl). The mean laxity difference between preoperative and postoperative values was 2.7 & 1.5 mm. There were no significant differences in Lysholm score between the regenerated and nonregenerated tendon groups. The 2 patients
SEMITENDINOSUS
MUSCLE
r = .86). Approximately the same relationships were seen for total work (Joules). There was no statistically significant difference in hamstring peak torque ratio (operated v nonoperated side) between the regenerated and nonregenerated tendon groups. No correlations were seen between the total thigh muscle area and the hamstring isokinetic peak torque in either the operated or nonoperated leg. However, correlations were found between concentric quadriceps peak torque and total thigh muscle area both in the operated (P < .05, r = .62) and in the nonoperated (P < .05, r = .77). The isokinetic curves had similar geometrical appearances in the operated and nonoperated leg, and no significant additional or missing peaks were seen between the 2 sides, although differences were found between maximal values.
TABLE 2. Pixel Values and Ratios in the Semitendinosus Muscle Postoperatively in the Operated and Nonoperated Legs for the Regenerated and Nonregenerated Semitendinosus Tendon Groups Operated Leg Regenerated n= 12 Nonregenerated n=4 P value NOTE. comparison groups.
Ratio Operated LegMonoperated Leg
Nonoperated Leg
105 (88-163)
91 (87-130)
I54 ( 125-799) ,015
1.17 (1.02-1.43)
108 (92-665) ,067
I .35 ( I .20-l .02
813
IN ACL SURGERY
.52)
Values are median and range. The P values refer to the between the regenerated and nonregenerated tendon
with osteochondral grafting had Lysholm scores of 89 (nonregenerated) and 99 (regenerated). The regenerated tendon group improved their Stryker side-to-side laxity significantly more than the nonregenerated group (ANOVA, P = .03). The difference in thigh circumference, operated versus nonoperated, was - 1.4 + 1.21 cm (P = .OOl) and the corresponding value for the calf was -0.48 + 0.93 cm (P = .08). No difference was seen between the regenerated and nonregenerated tendon groups. The l-leg hop test ratio (operated v nonoperated leg) was 0.94 (range, 0.771.02; P = .002) but no difference was found between the regenerated and nonregenerated tendon groups. A significant correlation was found between the l-leg hop and isokinetic concentric quadriceps strength ratios (operated v nonoperated side; P < .05, r = .60). However, no significant correlations were seen regarding the hop test and isokinetic concentric or eccentric hamstring strength. Isokinetic concentric and eccentric hamstring peak torque and concentric quadriceps peak torque are shown in Table 5. In the operated leg, we found significantly lower values in all tests (P < .05). There was also a significant correlation between the concentric and eccentric hamstring peak torques (P < .05,
DISCUSSION The main value of this study was the multidisciplinary approach to the donor site after ACL reconstruction with semitendinosus tendon. It included macromorphometry with MRI, micromorphometry with histochemical examination of the muscle biopsy specimen, and recordings of strength in the operated and nonoperated leg. The weakness in the study was the lack of preoperative MRI and isokinetic strength values. Regeneration of the resected semitendinosus tendon to a nearby anatomic distal insertion in 75% of the cases was previously reported by Eriksson et a1.23 The significance of pixel values in the regenerated tendon is hard to evaluate because there is no correlation between pixel values and histologic findings described in tendons. However, the MRI findings in the present study do correspond with a previous report of regenerated semitendinosus neotendons evaluated with MRI.23 The cross-sectional area of the semitendinosus tendon (10 to 20 mm2) in the present study does not entirely accord with the data of Hamada et al.19 who found it to be between 6.3 and 15 mm2. This could be
TABLE 3. Fiber Type Distribution and Citrate Synthase (CS) ( mmol/kg dry weight muscle X minutes) Activity in the Semitendinosus Muscle Intraoperatively and Postoperatively in Operated Legs and Postoperatively in Nonoperated Semitendinosus Muscle Type Preop operated leg n = I2 Postop operated leg n = I4 Postop nonoperated leg n = 8 NOTE.
Values
are mean
and standard
I %
50 (14) 54(17) 48(11) deviation.
Type
2A %
29 (IO)
27 (8) 34 (3)
Type
2B %
20 (16) I8 (13) I8 (IO)
Type
2C % l(1) 1 (2) 0 (0)
CS Activity 18 (8) n = 16 18 (6) n = 15 22 (6) n = 8
814 TABLE
K. ERIKSSON ET AL. 4.
Muscle
Fiber Area Postoperatively
(kn?)
Preop operated leg (n = 12) Postop operated leg (n = 14) Postop nonoperated leg (n = 8) NOTE.
Values
are mean
and standard
in Vurious Types in the Operuted Leg
of md
Fibers From the Semitendinosus Muscle IIltrcrc)pemti,,el~ Postoperatively in the Nompertrted Leg
Type
I Area
Type
2A Area
Type
5.250 5,264 5,799
(876) (2.062) (I .432)
6,878 6,770 7.902
( I .444) (2.4 19) (2.999)
6,687 6.781 6,977
5.
Isokinetic
Strength Hamstrings Concentric Peak Torque (NW
Postoperative operated leg Postoperative nonoperated leg Percent lower in operated leg Difference operated Y nonoperated P value NOTE.
Values
are mean
leg
and standard
Type
( I .785) (2,105) ( 1.955)
2C Area
5,142(1.371) 3,777 (456) No fibers present
Mean
Area
5.938 (1.173) 6.5 18 ( I .740) 6.828 ( 1.884)
deviation.
attributable to differences between Japanese and Caucasian anthropometry, because the tendon cross-sectional areas in our study resemble those reported by Tohyama et al., who examined a Caucasian population.30 When harvesting the semitendinosus tendon for ACL reconstruction, all muscles in the thigh were affected by the operation and the preoperative injury, since the total muscle area was about 5% less than that of the healthy side. The main cause of this atrophy was the quadriceps muscle (1 O%), not the hamstrings (1%). The semitendinosus muscle on the operated side was affected more by a reduction of 16% than the healthy side. When the MRI showed no regeneration of the tendon, the reduction in cross-sectional area of the semitendinosus muscle was significantly greater (34%). These patients also had a relative hypertrophy of the ipsilateral semimembranosus muscle belly, expressed as a significant difference in the muscle crosssectional area ratio (operated v nonoperated). The findings regarding greater hypotrophy of the semitendinosus muscle in patients with no regeneration have previously been reported by Eriksson et al.,23 but in that study the semimembranosus muscle was not analyzed for comparison. This agonist (semimembranosus) seems to compensate by hypertrophy for the loss in mass of the semitendinosus muscle if the tendon does not regenerate. Our findings regarding isokinetic hamstring strength support this view because no sig-
TABLE
2B Area
md
94 (25) I04 (25) IO -lO(l5) ,022 deviation.
(9O”/s)
nificant difference was seen between the tendon groups with or without regeneration. Simonian et al.?’ found no compensatory hypertrophy of the agonist muscles in their study of 9 patients. However they did not analyze their subgroups, regenerated (n = 6) and nonregenerated (n = 3), separately. Another difference between their study and ours is that they did not examine the thigh more proximally than 10 cm above the knee joint. The semitendinosus muscle acts as a part of the hamstring complex but does not play a dominant role in the knee flexors except during sitting thigh curls (supine position) when the semitendinosus, gracilis, and sartorius muscles are all strongly activated. In knee flexion exercises in a prone position, the semitendinosus and other hamstring muscles are only moderately activated.31 This was evident when we developed the test protocol for the isokinetic muscle test. Kramer et al.” performed isokinetic flexion tests 21 months after surgery with patients in the prone position and also found that the peak torques were lower in the operated than in the nonoperated leg. In their opinion, there was a distinct trend toward muscle imbalance, particularly in the muscle group affected by the semitendinosus tendon harvest. However, this conclusion could be questioned based on the relatively low specificity of their prone test situation. Most other studies have not stated whether hamstring torque was
in Operated
Hamstrings Concentric Total Work
and
Nonoperated
Hamstrings Eccentric Peak Torque
(J)
WV
53 (15) 62 (17) I5 -9 (I I) ,009
I I2 (27) I39 (30) I9 -28 (18) c.001
Legs Hamstrings Eccentric Total Work
(J) 68 (18) 88 (21) 23 -20 (IO) c.001
(n =
15)
Quadriceps Concentric Peak Torque
(Nm) 181 (34) 226 (55) 20 -45 (38) <.OOl
Quadriceps Concentric Total Work
(J) I82 (32) 234 (53) 22 -52 (45) <.OOl
SEMITENDINOSUS
MUSCLE
calculated with patients in the sitting or the prone position. 4. IY-2 I.25.33-35 Muscle imbalance and quadriceps and hamstring deficits in ACL-deficient knees have been found in previous studiesJh-“x and, when evaluating postoperative values, this must be taken into consideration. Other studies of the thigh cross-sectional area of ACLdeficient knees show deficits of 5% to 10% in the injured compared with the healthy leg.‘7.3Y-Ai It is not yet proven that harvest of the semitendinosus tendon is the main reason for knee flexion strength deficit after ACL reconstruction. It could be the injury itself. In fact, Yasuda et al.” found no difference in isokinetic strength before and 3 months after surgery if only a graft had been harvested from the leg and the ACL reconstruction had been performed on the contralateral side. The isokinetic concentric and eccentric hamstring strengths were significantly lower in the operated than in the nonoperated leg. The same relation was observed in the concentric quadriceps test. These results resemble those in a recent isokinetic follow-up study by Carter and Edinger.JZ who found 25% and 20% deficits in extension and flexion. respectively, in the operated leg 6 months after surgery. However, the results are not entirely comparable because their protocol differed from ours. Lipscomb et a1.,i9 in a retrospective study of 51 patients, found no significant side-to-side difference in hamstring strength (60” v 240” per second) 26 months after surgery, but did find a slight decrease in quadriceps strength at the same follow-up. Sekiya et al.,?’ in a 2-year follow-up study of quadriceps and hamstring isokinetic strength indices (operated v nonoperated leg), reported results similar to ours, with 10% and 20% strength deficit in the hamstrings and quadriceps, respectively. In that study, they also noted a correlation between quadriceps and hamstring isokinetic peak torque indices and the l-leg hop test index. We also found a correlation between the quadriceps isokinetic index and hop test, but no correlation with the hamstring index. Laxity was not correlated to the hop test, quadriceps, or hamstring index in either their study or in ours. Ohkoshi et al.Zz reported 99% hamstring peak torque strength in the operated leg but simultaneously found that peaks were missing in the latter half of the torque curves. They concluded that this was attributable to lack of semitendinosus action and compensation by other muscles. In our study, no obvious differences in torque curves were seen between operated and nonoperated legs. Comparison of the controls with the preoperative
IN ACL SURGERY
815
and postoperative citrate synthase enzyme activity levels, reflecting oxidative capacity, showed no abnormalities in semitendinosus oxidative muscle metabolism. Nor was the fiber type distribution altered in the operated versus nonoperated legs. or in the regenerated versus nonregenerated subgroups. When comparing the fiber type distribution in the semitendinosus muscle with the commonly examined vastus lateralis, the type 1 fiber proportion in our material seemed to be similar to previous findings in ACL-deficient patients.JJ.W The type 1 fiber proportion was similar to that in uninjured male patients of similar age, but type 2B fiber proportion in our study was slightly higher.-1s-J7As for fiber area determinations, no sign of denervation or fiber atrophy was seen. The area values for all fiber types agree with results of investigators concerning the quadriceps muscle in a trained population.“’ These findings support previous electromyographic results reported by Cross et al.“’ who found a normal pattern in the semitendinosus muscle in patients where the gracilis and semitendinosus tendons had previously been harvested for ACL reconstruction. CONCLUSIONS Harvest of the semitendinosus tendon appears to be a well-tolerated procedure in ACL reconstruction surgery, after which a structure that has the appearance of a tendon regenerates in 75% of the patients, but so far there is no histologic confirmation that it is in fact a tendon. In the remaining 25% of patients, the muscle crosssectional area was smaller. This was compensated for by a concomitant increase in the semimembranosus muscle cross-sectional area. In no case were signs of total atrophy or retraction of the semitendinosus muscle seen, and the oxidative muscle potential and fiber type distribution were the same as in the control leg. Nor were there any differences when comparing patients with regenerated and nonregenerated tendons. This was also true when comparing clinical outcome and isokinetic strength, although there was a strength deficit in the operated versus the nonoperated leg. In the present study, the functional and structural morbidity of the semitendinosus graft harvest seemed to be small and the muscle seemed to recover following harvest of its tendon. REFERENCES I. Aglietti P, Buzzi R, Zaccherotti G, De Biase P. Patellar tendon versus doubled semitendinosus and gracilis tendons for ante-
816
2.
3.
4.
5. 6.
7. 8. 9.
IO.
1 I. 12.
13. 14. 15.
16.
17.
18.
K. ERIKSSON ET AL. rior cruciate ligament reconstruction. Am J Sports Med 1994; 22:21 I-218. Aglietti P, Buzzi R, Menchetti PM, Giron F. Arthroscopically assisted semitendinosus and gracilis tendon graft in reconstruction for acute anterior cruciate ligament injuries in athletes. Am J Sports Med 1996;24:726-73 I. Maeda A, Shino K, Horibe S, Nakata K, Buccafusca Cl. Anterior cruciate ligament reconstruction with multistranded autogenous semitendinosus tendon. Am J Sports Med 1996;24: 504-509. Marder RA, Raskind JR, Carroll M. Prospective evaluation of arthroscopically assisted anterior cruciate ligament reconstruction. Patellar tendon versus semitendinosus and gracilis tendons. Am J Sports Med 1991; 19:478-484. Brown CH Jr. Steiner ME. Carson EW. The use of hamstrine tendons for anterior c&ate ligament reconstruction. Tech: nique and results. C/in Sports Med 1993;12:723-756. Ferretti A, Conteduca F, De Carli A, Fontana M, Mariani PP. Results of reconstruction of the anterior cruciate ligament with the tendons of semitendinosus and gracilis in acute capsuloligamentous lesions of the knee. Iral J Orthop Traumaroj 1990; 16:452-458. Meystre JL, Vallotton J, Benvenuti JF. Double semitendinosus anterior cruciate ligament reconstruction: I O-year results. Knee Surg Sports Traumatol Arthrosc 1998;6:76-81. Rosenberg TD. Deffner KT. ACL reconstruction: Semitendinosus tendon is the graft of choice. Orthopedics 1997;20:396398. Kartus J, Stener S, Lindahl S, Engstrom B, Eriksson BI, Karlsson J. Factors affecting donor-site morbidity after anterior cruciate ligament reconstruction using bone-patellar tendon-bone autografts. Knee Surg Sports Traumatol Arthrosc 1997;5:222-228. Kartus J, Magnusson L, Stener S, Brandsson S, Eriksson BI, Karlsson J. Complications following arthroscopic anterior cruciate ligament reconstruction. A 2-5-year follow-up of 604 patients with special emphasis on anterior knee pain. Knee Surg Sports Traumarol Arrhrosc 1999;7:2-8. Kohn D, Sander-Beuermann A. Donor-site morbidity after harvest of a bone-tendon-bone patellar tendon autograft. Knee Surg Sports Traumarol Arrhrok 1994;2:2 19-223. Miller MD. Nichols T. Butler CA. Patella fracture and oroximal patellar tendon rupture following arthroscopic anterior cruciate ligament reconstruction. Arthroscopy 1999; I5:640643. Aglietti P, Buzzi R, D’ Andria S, Zaccherotti G. Patellofemoral problems after intraarticular anterior cruciate ligament reconstruction. C/in Orrhop 1993: 195-204. Sachs RA, Daniel DM, Stone ML, Garfein RF. Patellofemoral problems after anterior cruciate ligament reconstruction. Am J Sports Med 1989; 17:760-765. Shino K, Nakagawa S, Inoue M, Horibe S, Yoneda M. Deterioration of patellofemoral articular surfaces after anterior CNciate ligament reconstruction. Am J Sports Med 1993;2 I :206211. Zarzycki W, Mazurkiewicz S, Wisniewski P. [Research on strength of the grafts that are used in anterior cruciate ligament reconstruction]. Chir Narzadow Ruchu Orrop Pal 1999:64: 293-302. Hamner DL, Brown CH, Jr., Steiner ME, Hecker AT, Hayes WC. Hamstring tendon grafts for reconstruction of the anterior cruciate ligament: biomechanical evaluation of the use of multiple strands and tensioning techniques. J Bone Joinr Surg Am 1999;81:549-557. Rosenberg TD, Franklin JL, Baldwin GN, Nelson KA. Extensor mechanism function after patellar tendon graft harvest for anterior cruciate ligament reconstruction. Am J Sporrs Med 1992;20:5 19-526.
19. Lipscomb AB. Johnston RK, Snyder RB, Warburton MJ. Gilbert PP. Evaluation of hamstring strength following use of semitendinosus and gracilis tendons to reconstruct the anterior cruciate ligament. Am J Sports Med 1982; l&340-342. 20. Harter RA. Ostemig LR, Standifer LW. Isokinetic evaluation of quadriceps and hamstrings symmetry following anterior cruciate ligament reconstruction. Arch Phys Med Rehabil l990:7 I :465-468. 21. Yasuda K, Tsujino J. Ohkoshi Y. Tanabe Y, Kaneda K. Graft site morbidity with autogenous semitendinosus and gracilis tendons. Am J Sporrs Med 1995;23:706-714. 22. Ohkoshi Y. moue C. Yamane S. Hashimoto T, Ishida R. Changes in muscle strength properties caused by harvesting of autogenous semitendinosus tendon for reconstruction of contralateral anterior cruciate ligament. Arrhroscopy 1998; 14:580584. 23. Eriksson K, Larsson H, Wredmark T, Hamberg P. Semitendinosus tendon regeneration after harvesting for ACL reconstruction. A prospective MRI study. Knee Surg Sports Traumarol Arrhrosc 1999;7:220-225. 24. Cross MJ, Roger G. Kujawa P. Anderson IF. Regeneration of the semitendinosus and gracilis tendons following their transection for repair of the anterior cruciate ligament. Am J Sporrs Med 1992;20:22 l-223. 25. Simonian PT. Harrison SD. Cooley VJ. Escabedo EM, Deneka DA, Larson RV. Assessment of morbidity of semitendinosus and gracilis tendon harvest for ACL reconstruction. Am J Knee Surg 1997; 10:54-59. 26. Rosenberg TP. Graf B. Endoscopic technique for ACL reconstruction with Protrac tibia1 guide: EndoButton fixation. Mansfield, MA: Acufex Microsurgical. 1992. 27. Schantz P, Billeter R, Henriksson J, Jansson E. Traininginduced increase in myofibrillar ATPase intermediate fibers in human skeletal muscle. Muscle Nerve 1982;5:628-636. 28. Lin L, Sotonyi P, Somogyi E, et al. Coenzyme QIO content in different parts of the normal human heart. Clin Physiol 1988; 8:39 l-398. 29. Hamada M, Shino K, Mitsuoka T, Abe N, Horibe S. Crosssectional area measurement of the semitendinosus tendon for anterior cruciate ligament reconstruction. Arrhrascopy 1998; 14:696-70 I. 30. Tohyama H, Beynnon BD, Johnson RJ, Nichols CE. Renstrom PA. Morphometry of the semitendinosus and gracilis tendons with application to anterior cruciate ligament reconstruction. Knee Surg Sporrs Traumarol Arrhrosc 1993;l: 143-147. 3 I. Tesch PA. Target bodybuilding. Champaign, IL: Human Kinetics, 1999. 32. Kramer J, Nusca D, Fowler P, Webster-Bogaert S. Knee flexor and extensor strength during concentric and eccentric muscle actions after anterior cruciate ligament reconstruction using the semitendinosus tendon and ligament augmentation device. Am J Sports Med 1993;2 I :285-29 I. 33. Sekiya I, Muneta T, Ogiuchi T, Yagishita K, Yamamoto H. Significance of the single-legged hop test to the anterior cruciate ligament-reconstructed knee in relation to muscle strength and anterior laxity. Am J Sports Med 1998;26:384388. 34. Karlson JA, Steiner ME, Brown CH, Johnston J. Anterior cruciate ligament reconstruction using gracilis and semitendinosus tendons. Comparison of through-the-condyle and over-the-top graft placements. Am J Sports Med 1994;22: 659-666. 35. Li RCT. Isokinetic strength of the quadriceps and hamstrings and functional ability of anterior cruciate deficient knees in recreational athletes. Br J Sports Med I996;30: 16 I - 164. 36. Shirakura K, Kato K, Udagawa E. Characteristics of the isokinetic performance of patients with injured cruciate ligaments. Am J Sporrs Med 1992;20:754-760.
SEMITENDINOSUS 37.
38.
39.
40.
41.
42.
MUSCLE
Lorentzon R, Elmqvist LG. Sjostrom M, Fagerlund M, Fuglmeyer AR. Thigh musculature in relation to chronic anterior cruciate ligament tear: muscle size, morphology. and mechanical output before reconstruction. Am J Sports Med 1989; 17: 423-429. ltoh H. lchihashi N, Maruyama T, Kurosaka M, Hirohata K. Weakness of thigh muscles in individuals sustaining anterior cruciate lieament iniurv. Kobe J Med Sci 1992:38:93-107. Gerber C,-Hoppele; H, Claassen H, Robotti G, Zehnder R, Jakob RP. The lower-extremity musculature in chronic symptomatic instability of the anterior cruciate ligament. J Bone Joint Surg Am 1985;67: 1034-1043. Arangio GA, Chen C. Kalady M. Reed JF 3rd. Thigh muscle size and strength after anterior cruciate ligament reconstruction and rehabilitation. J Orthop Sports Phys Ther 1997;26: 238-243. Kariya Y, Itoh M. Nakamura T, Yagi K, Kurosawa H. Magnetic resonance imagine and soectroscouv of thigh muscles in cruciate ligament insufficiency. Acra &hop S&d 1989;60: 322-325. Carter TR. Edinger S. Isokinetic evaluation of anterior cruciate
43.
44.
45.
46.
47.
IN ACL SURGERY
817
ligament reconstruction: Hamstring versus pateliar tendon. Arlhroscopy 1999; 15: l69- 172. Arvidsson I. Arvidsson H, Eriksson E, Jansson E. Prevention of quadriceps wasting after immobilization: An evaluation of the effect of electrical stimulation. Orrhopedics 1986;9: I5 l91528. Haggmark T, Eriksson E. Jansson E. Muscle fiber type changes in human skeletal muscle after injuries and immobilization. Orthopedics 1986;9: I8 I -I 85. Berg HE, Dudley GA. Hather B, Tesch PA. Work capacity and metabolic and morphologic characteristics of the human quadriceps muscle in response to unloading. Clin Physiol 1993;13: 337-347. Maughan RJ. Nimmo MA. The influence of variations in muscle fibre composition on muscle strength and cross-sectional area in untrained males. J Physiol (land) 1984;35 I :29931 I. Esbiomsson M. Jansson E, Sundberg CJ. Svlven C. Eiken 0. Nygren A, Kaijser L. Muscle fibre tyjes andenzyme activities after training with local leg ischemia in man. Acta Physiol Scmd 1993;148:233-241.
To evaluate the fate of the hamstring muscles in general and the semitendinosus muscle in particular, after anterior cruciate ligament (ACL) reconstruction with an autologous semitendinosus tendon graft from the ipsilateral side. Type of Study: Prospective consecutive case series investigation. Methods: Included were I6 consecutive patients, I4 male and 2 female. with a mean age of 26 years. The inclusion criterion was chronic unilateral ACL insufficiency with no concomitant knee ligament injuries. ACL reconstruction was performed with a quadruple semitendinosus tendon graft using the EndoButton technique (Acufex. Mansfield. MA). Intraoperatively, muscle specimens were taken from the semitendinosus muscle on the harvested side. Follow-up at a minimum of 6 months included clinical examination, isokinetic strength performance, magnetic resonance imaging (MRI) of the thigh and knee. and ultrasound-guided muscle biopsy procurement from the semitendinosus muscle for histochemical and enzymatic analyses. Results: Of the patients. 75% showed regeneration of their semitendinosus tendons. The neotendons all inserted below the knee joint where they had fused with the gracilis tendon to a conjoined tendon inserting in the pes anserinus. The semitendinosus muscle had a smaller cross-sectional area on the operated side but none showed total atrophy. Less atrophy was present in the patients with a regenerated semitendinosus neotendon compared with those without regeneration (P = ,029). In the latter group the semimembranosus muscle seemed to compensate for this with hypertrophy (P = ,019). Cross-sectional muscle fiber areas, the relative number of each fiber type and oxidative potential as estimated by citrate synthase activity, showed no significant differences between the operated and nonoperated legs. The isokinetic strength of the hamstrings and quadriceps was significantly lower in the operated leg than in the nonoperated leg. Conclusions: With this surgical technique, the semitendinosus muscle can recover and the tendon has. according to the MRI images, a great potential to regenerate after its removal. Key Words: ACL reconstruction-Semitendinosus muscle-Regeneration-Histochemistry-MRI-Isokinetic strength.
Purpose:
I
n recent years, hamstring tendon graft has become an alternative to a bone-patellar tendon-bone (BPTB) graft in anterior cruciate ligament (ACL) re-
From the Departments of Orthopedics (K. E.. P. H.) and Radiology (H. L.). Stockholm Siider Hospital. Karolinska Institute; and the Departments of Clinical Physiology (E.J. j. Radiology (AS.). and Orthopedics (T. W.). Huddinge Universityv Hospital, Kamlinska Institute. Stockholm, Sweden. Supported by Grant No. 4494 from The Sb+*edish Medical Research Council. and grants from the Swedish Center for Reseurch in Sports and Stockholm Siider Hospital Research Foundation. Presented as part of Dr. Eriksson ‘s thesis, Karolinska Institute. Stockholm. Defended February 2. 2001. Address correspondence and reprint requests to Karl Eriksson. M.D., Department of Orthopedics, Stockholm Siider Hospital. Karolinska Institute. S-l 18 83 Stockholm, Sweden. E-mail: kalle. [email protected] 0 2001 by the Arthroscopy Association of North America 0749.8063/01/1708-2643$35.00/O doi:l0.1053/jnr.s.200/.20961
808
Arthroscopy:
The Journal
of Arthroscopic
and Related
construction surgery.i-x Many physicians criticize the high donor site morbidity when using a BPTB graft.“-ij The purported advantages of a semitendinosus graft material include lower donor-site morbidity and possibly less anterior knee pain,*,‘-5 a stronger graftih.17 because of the larger diameter than BPTB,3 and less quadriceps muscle strength deficit because the extensor mechanism is less affected.ia-ltJ Regardless of which autologous graft is chosen, donor-site morbidity will occur in some patients. Until now, only a few prospective randomized studies have focused on this issue. Yasuda et al.” reported no longterm differences in hamstring or quadriceps strength when comparing the autologous hamstring (semitendinosus and gracilis) tendons from the ipsilateral and contralateral sides. Likewise, neither Lipscomb et al.)” nor Maeda et al.’ found any reduction in hamstring strength in patients who underwent reconstruction
Surgetyv.
Vol 17. No 8 (October).
2001:
pp 80R-RI7
SEMITENDINOSUS
MUSCLE
with ipsilateral hamstring tendons. However. Marder et al.-’ reported that hamstring strength was lower with ipsilateral semitendinosus and gracilis grafts than with an ipsilateral BPTB graft. In these studies of isokinetic muscle strength of the hamstring muscles. both the semitendinosus and the gracilis tendons had been used for an autologous graft. In a recent study by Ohkoshi et al..” no differences in presurgical and postsurgical isokinetic peak torque values were seen in the hamstring muscles when the semitendinosus tendon was harvested from the contralateral leg. A few earlier magnetic resonance imaging (MRI) studies show that the semitendinosus tendon can regenerate. after being harvested for ligament reconstruction.‘J-‘5 The aims of the present study were to prospectively evaluate a patient group in which only the ipsilateral semitendinosus tendon had been harvested for an autologous graft. The data included results of clinical examination. isokinetic strength testing of the thigh. a biopsy examination of the semitendinosus muscle. and MRI. To our knowledge, no other study has described the histochemical. morphologic, and biochemical changes in hamstring muscles following ACL reconstruction. Indeed, little is known about the histochemistry, morphology and biochemistry of this muscle group.
METHODS Patients Sixteen patients. I4 male and 2 female, with a mean age of 26 years (range. I7 to 34 years) years were prospectively included in the study. The inclusion criterion was unilateral ACL injury with no concomitant ligament injuries. The median time from ACL trauma to reconstruction was 8 months (range, 3 to 24 months). No prior surgery except arthroscopic meniscus surgery had been performed on the involved knee. All patients underwent reconstruction with an ipsilatera1 quadruple semitendinosus tendon graft, using the EndoButton technique (Acufex, Mansfield, MA) described by Rosenberg and Graf.‘6 The right knee was involved in 9 cases and the left in 7. In 2 patients, because of chondral lesions, autologous osteochondral grafting was performed concomitantly with the ACL reconstruction. Both of these lesions were less than I cm’. Preoperative radiographs showed no arthritic changes in any of the knees. Nine patients (56%) had undergone partial meniscectomy (5 medial and 5 lateral) at their initial arthroscopy following ACL injury. Another 2 patients
IN ACL SURGERY
809
underwent partial medial meniscectomy at their reconstruction. leaving 5 patients (31%) without meniscus lesions at the time of reconstruction. The graft was harvested as described previously by Eriksson et al.‘j Before final preparation of the graft, muscle specimens attached to the musculotendinous junction were taken from the proximal part of the resected tendon at a distance of about 30 cm proximal to the tendon insertion on the tibia. A standardized rehabilitation protocol was used, including early rangeof-motion training. full weight bearing, and closedchain exercises. At a median of 7 months (range, 6 to I2 months), all patients underwent MRI scans of the knee and thigh. Ten months (range. 8 to I8 months) postoperatively. the evaluation was completed with a clinical examination, including an isokinetic muscle performance test. and percutaneous ultrasound-guided needle biopsy of the semitendinosus muscle. The nonoperated side was also examined as a reference. except for the control biopsies, which were taken only in 8 patients. One patient had an early complication with deep infection and was successfully reoperated on with an ipsilateral BPTB graft 5 months after the initial reconstruction. Therefore, his data from the isokinetic test as well as the thigh and quadriceps cross-sectional areas on MRI scans were excluded because they were also affected by the second operation. His results from the biopsy and MRI evaluation of the semitendinosus were included, however, because the primary failure did not affect tendon regeneration after his primary operation. The study was approved by the ethics committee of the Karolinska Institute.
MRI The MRI images were taken with a C-shaped 0.2Tesla resistive magnet (Magnetom Open; Siemens, Erlangen, Germany). T I - and T2-weighted transaxial MRI sequences of the thigh and knee joint were used and. in a few cases, additional sagittal sequences were made. In all images, the knee was held in full extension with an external rotation of 15” to 20”. T2weighted images were used to calculate semitendinosus, semimembranosus, and total muscle areas and pixel values over the semitendinosus, with TR (repetition time) of 4,914 milliseconds, TE (echo time) of 102 milliseconds, FOV (field-of-view) of 240 X 380 mm, matrix size of 126 X 256 pixels, and a slice thickness of 15 mm. Tl-weighted images of the knee joint were used for tendon evaluations with TR of 920 milliseconds, TE of 26 milliseconds, FOV of 150 X
810
K. ERIKSSON ET AL.
200 mm, matrix size of 144 X 256 pixels, and a slice thickness of 10 mm. In one patient, Tl-weighted images were used over the proximal thigh and knee joint. The semitendinosus tendon area and pixel values were calculated approximately 5 cm proximal to the joint line and the muscle area and muscle pixel determinations were made approximately 12 to 14 cm below the pubic bone, which corresponds to about 30 cm proximal to the pes anserinus. The corresponding values on the nonoperated leg were always determined at the same level. The MRI scans were evaluated by an experienced MRI radiologist. All images were analyzed twice at an interval of 3 months to ensure that the observer would not be influenced by the previous readings. The values given are the mean of these evaluations. The intraclass correlation (ICC) between the 2 readings was 0.7 to 0.9, which indicates a good correlation. Semitendinosus
Biopsy Analyses
The biopsy specimens of muscle taken intraoperatively from the leg were instantly frozen with Freon and stored at -70°C until all postoperative and control samples had been taken. An ultrasound-guided percutaneous needle biopsy technique was used for the postoperative and control specimens, using a 2. l-mm Bard Magnum needle (C.R. Bard, Covington, GA) that was fired 2 to 5 times, which procured approximately 1 to 5 mg wet weight of muscle per firing. In 2 cases, the needle position in the semitendinosus muscle belly was confirmed by simultaneous computed tomography scans. Before the biopsy, the patient was placed in the prone position, the semitendinosus muscle was identified with ultrasound and marked at the level corresponding to the site of the intraoperative sample, i.e., approximately 30 cm proximal to the pes anserinus. After local anesthetic infiltration of the skin and subcutaneous tissue, but not the muscle, a 3-mm skin incision was made, allowing the biopsy needle to approach the fascia easily. The needle was then fired under ultrasound guidance with free-hand technique and samples were collected and instantly frozen in isopentane precooled by liquid nitrogen and stored at -70°C pending analysis. Eight patients gave their approval to sampling from both the nonoperated and the operated legs. All samples (intraoperative, postoperative, and controls from the nonoperated leg) were analyzed at one time by the same examiner. Serial transverse sections (10 mm) were cut with a microtome at -20°C and stained for myofibrillar ATPase, after
preincubation at different pH levels in acid and alkaline buffers.“’ The fibers were classified on the basis of their staining intensity (myofbrillar ATPase) into fiber types I, 2A, 2B. and 2C. A computerized image analysis system (Scan-Beam; BioRad. Hadsund, Denmark) was used to calculate the relative distribution of fiber types in number (%) and cross-sectional fiber area for each fiber type. The average number of fibers per sampling occasion was 205 in the intraoperative specimen and 150 in the postoperative and control specimens. Fiber areas were measured on the stains preincubated at pH 4.6. The remaining portion of the biopsy samples was freeze-dried. dissected free from visible blood, extracellular fat, and connective tissue, under a dissection microscope. Citrate synthase activity was determined by a fluorometric method at 25°C using the method of Lowry and Passonneau after homogenization of the dried and weighed tissue in 0.1 mol/L of ice-cooled phosphate buffer (pH 7.7) with 0.5% bovine serum albumin (dilution 1 mg dry muscle in 400 mL buffer).Zx Because of freezing artifacts, only 12 of the 16 patients were represented with samples allowing fiber typing intraoperatively, 14 postoperatively, and, in the control group, all 8 patients had samples adequate for histochemical analysis. Enzyme determinations were made in all 16 patients intraoperatively, in all 15 patients postoperatively. and in all 8 controls. Clinical and Isokinetic
Strength
Evaluations
Fifteen patients underwent a standardized test protocol performed by an independent examiner, a physical therapist, including l-leg hop test, range-of-motion evaluation, and measurement of thigh circumference 15 cm proximal to the joint line and maximum calf muscle width. The Lysholm score was noted. Sagittal stability was determined with the Lachman test and the Stryker laxity tester (OSI; Stryker, Kalamazoo, MI), evaluating side-to-side difference with 89 N anterior force. Isokinetic concentric and eccentric hamstring muscle strength and concentric quadriceps strength were measured with a Kin-corn (Chattex, Chattanooga, TN). The test consisted of a lo-minute warm-up period on a cycle ergometer at 50 W, and 5 submaximal and 2 maximal isokinetic hamstring contractions on the Kin-corn. After a l-minute rest, the recording started with sitting thigh curls. Concentric and eccentric peak torque (Newton meters) between flexion angles of 45” to 90” were measured at an angle velocity of 90” per second. The test was performed
SEMITENDINOSUS
MUSCLE
IN ACL SURGERY
811
ST Muscle Operated Leg Regenerated n = 12 Nonregenerated n = J All patients n = I6 NOTE.
Values
are mean
ST Muscle Nonoperated
I l.S3 (7.65) 7.96 (2.47)
12.x9 (25X) 12.0 (3.601
IO.61
12.67 (2.76)
(2.98)
and standard
Leg
Ratio ST Operated/ Nonoperated Leg
SM Muscle Operated Leg
0.89
(0.08)
13.03
(4.9s)
0.66 0.84
(0.2 I ) (0.11)
14.33 13.37
(4.26) (4.66)
SM Muscle Nonoperated Leg 13.35 I I.57 12.88
(4.34) (3.76) (4.14)
Ratio SM Operated/ Nonoperated Leg 1.03 (0.14) I .26 (0.13) I .05 (0.23)
deviation.
with consecutive maximal torques (usually 3 to 5) and ended when the subject failed to reach the previous maximal value. An isokinetic concentric quadriceps contraction from 90” to 5” of knee flexion at 90” per second was also performed. One patient did his tests on a Biodex isokinetic machine (Biodex Corp. New York, NY) instead of Kin-corn because of a technical failure of the Kin-corn. and the same test protocol was used. In all patients, the nonoperated leg was examined for comparison. In addition to peak values in Newton meters, total work in Joules (area under the curve) was calculated in each test. Statistics All variables were subject to conventional descriptive statistics (mean, standard deviation for normally distributed variables, median and range for ordinal or skewed variables). The problem of multiplicity was discussed. Because of the exploratory nature of this study. any statistical significance was followed by a discussion of the clinical significance. No method for dealing with missing values was applied. If the variables were normally distributed, a parametric test was used, i.e.. all variables except MRI pixel values and Lysholm scores where nonparametric tests were used. The difference between preoperative and postoperative values, as well as operated and nonoperated leg, were analyzed by using the Student t test for dependent samples. The difference between patients with a regenerated semitendinosus tendon and those with no regeneration after surgery was analyzed by analysis of variance (ANOVA) with repeated measures of the preoperative and postoperative difference. This ANOVA describes the interaction between the preoperative and postoperative values and regeneration and no regeneration. The relationship between operated leg and nonoperated leg was in some variables described as a quotient and analyzed with the Student t test for independent samples. For preoperative postoperative difference and operated leg versus nonoper-
ated leg, the Wilcoxon rank sum test for dependent samples was used. The Mann-Whitney U test for independent samples was used for analysis of difference between the patients with regenerated and nonregenerated semitendinosus tendon. Correlations were calculated using the Pearson 2-tailed test. ICC with a 2-way mixed model with average measure was used for a measure of agreement between the 2 readings of the MRI examinations. The overall level of significance was set to P < .05. RESULTS MRI Evaluation Anatomy: Twelve of the 16 patients (11 males and 1 female) showed regeneration of the semitendinosus tendon (i.e., neotendon) with normal topography to and below the level of the joint line. At various distances 1 to 3 cm below the joint line, the neotendons had fused with that of the gracilis to form a conjoined tendon structure, which inserted in the pes anserinus. In 1 female patient, the neotendon had fused with the semimembranosus tendon proximal to the joint line. In the 3 remaining patients, no obvious tendon structure was seen distal to the semitendinosus muscle belly. However, no signs of retraction of the muscle bellies were noted. Of the 2 patients with concomitant osteochondral grafting at the time of reconstruction, 1 had a regenerated neotendon and the other did not. Area Determinations: The regenerated semitendinosus neotendon cross-sectional area was 0.27 2 0.08 cm’, and the area on the nonoperated leg was 0.18 + 0.03 cm’. There was no significant difference between the 2 sides (P = .14). The gracilis tendon areas were the same on both sides. The total thigh muscle area of the operated leg was 4% less than in the nonoperated leg (177 + 28 cm* v 185 + 31 cm*, P = .006). No significant difference in area ratio (operated/nonoperated leg) was found between regenerated and nonregenerated cases. The
812
K. ERIKSSON ET AL.
quadriceps muscle cross-sectional area on the operated side was 10% smaller than on the nonoperated side (P = .OOl), but there was no significant difference (ratio, 0.99 -+ 0.04, P = Sl) between the total hamstring area on the operated and nonoperated side. The semitendinosus muscle cross-sectional area was 16% smaller on the operated than on the nonoperated side (Table 1, P = .OOl). There was no statistically significant difference between the semimembranosus muscle cross-sectional area in the operated and nonoperated side when evaluating all patients (regenerated and nonregenerated). The ratios of the semitendinosus and semimembranosus muscle crosssectional areas between operated and nonoperated sides in the regenerated and the nonregenerated tendon groups are shown in Fig 1. Patients with no regeneration had a relatively larger semimembranosus muscle on the operated side compared with the nonoperated side (P = ,019). Pixel Values (Signal Intensity): The pixel values of the regenerated neotendons did not differ significantly from those of the nonoperated legs. Nor was there any difference between the gracilis tendons in the operated and nonoperated legs. The pixel values of the semitendinosus muscle in both legs are shown in Table 2. The 1 patient examined with Tl-weighted sequences over the thighs had higher levels in both legs, which explains the wider range in the nonregenerated group. No significant difference in pixel values was seen in the semimembranosus muscle on the operated and nonoperated side. Muscle Biopsy Evaluation No significant differences were found in citrate synthase activity, fiber type composition, or fiber type areas before and after the surgery in the operated leg. Nor were there any differences in these variables between the operated and nonoperated legs after surgery (8 subjects) (Tables 3 and 4). No differences were noted between the regenerated and nonregenerated groups in the variables studied. Using the t test on an intraindividual basis (n = 7), the mean fiber area in the postoperative samples in the operated legs tended to be smaller than the postoperative controls in the nonoperated legs (6,304 v 7,021 pm*, P = .16). Clinical Examination
and Isokinetic
Strength
The Lysholm score was 71 (range, 40-86) before surgery and 91 (range, 70-100) after surgery, a significant improvement (P < .OOl). The Stryker laxity test with 89 N showed a mean side-to-side difference
FIGURE 1. (A) Ratio of cross-sectional area in semimembranosus muscle of operated and nonoperated side, and nonregenerated and regenerated tendons, respectively. The area is relatively greater on the operated side in the cases with no regeneration (P = ,019). The median is indicated in the boxes, containing 50% of the values, The range is seen outside the squares. (B) Ratio of cross-sectional area in semitendinosus muscle of operated and nonoperated leg, and nonregenerated and regenerated tendons. respectively. The area is greater in the regenerated cases (P = ,029). The median is shown in the boxes, containing 50% of the values. The range is seen outside the squares.
preoperatively of 4 + 1.68 mm. Postoperatively, the side-to-side difference was 1.27 + 1 .I mm and the preoperative to postoperative laxity difference had improved significantly (P < .OOl). The mean laxity difference between preoperative and postoperative values was 2.7 & 1.5 mm. There were no significant differences in Lysholm score between the regenerated and nonregenerated tendon groups. The 2 patients
SEMITENDINOSUS
MUSCLE
r = .86). Approximately the same relationships were seen for total work (Joules). There was no statistically significant difference in hamstring peak torque ratio (operated v nonoperated side) between the regenerated and nonregenerated tendon groups. No correlations were seen between the total thigh muscle area and the hamstring isokinetic peak torque in either the operated or nonoperated leg. However, correlations were found between concentric quadriceps peak torque and total thigh muscle area both in the operated (P < .05, r = .62) and in the nonoperated (P < .05, r = .77). The isokinetic curves had similar geometrical appearances in the operated and nonoperated leg, and no significant additional or missing peaks were seen between the 2 sides, although differences were found between maximal values.
TABLE 2. Pixel Values and Ratios in the Semitendinosus Muscle Postoperatively in the Operated and Nonoperated Legs for the Regenerated and Nonregenerated Semitendinosus Tendon Groups Operated Leg Regenerated n= 12 Nonregenerated n=4 P value NOTE. comparison groups.
Ratio Operated LegMonoperated Leg
Nonoperated Leg
105 (88-163)
91 (87-130)
I54 ( 125-799) ,015
1.17 (1.02-1.43)
108 (92-665) ,067
I .35 ( I .20-l .02
813
IN ACL SURGERY
.52)
Values are median and range. The P values refer to the between the regenerated and nonregenerated tendon
with osteochondral grafting had Lysholm scores of 89 (nonregenerated) and 99 (regenerated). The regenerated tendon group improved their Stryker side-to-side laxity significantly more than the nonregenerated group (ANOVA, P = .03). The difference in thigh circumference, operated versus nonoperated, was - 1.4 + 1.21 cm (P = .OOl) and the corresponding value for the calf was -0.48 + 0.93 cm (P = .08). No difference was seen between the regenerated and nonregenerated tendon groups. The l-leg hop test ratio (operated v nonoperated leg) was 0.94 (range, 0.771.02; P = .002) but no difference was found between the regenerated and nonregenerated tendon groups. A significant correlation was found between the l-leg hop and isokinetic concentric quadriceps strength ratios (operated v nonoperated side; P < .05, r = .60). However, no significant correlations were seen regarding the hop test and isokinetic concentric or eccentric hamstring strength. Isokinetic concentric and eccentric hamstring peak torque and concentric quadriceps peak torque are shown in Table 5. In the operated leg, we found significantly lower values in all tests (P < .05). There was also a significant correlation between the concentric and eccentric hamstring peak torques (P < .05,
DISCUSSION The main value of this study was the multidisciplinary approach to the donor site after ACL reconstruction with semitendinosus tendon. It included macromorphometry with MRI, micromorphometry with histochemical examination of the muscle biopsy specimen, and recordings of strength in the operated and nonoperated leg. The weakness in the study was the lack of preoperative MRI and isokinetic strength values. Regeneration of the resected semitendinosus tendon to a nearby anatomic distal insertion in 75% of the cases was previously reported by Eriksson et a1.23 The significance of pixel values in the regenerated tendon is hard to evaluate because there is no correlation between pixel values and histologic findings described in tendons. However, the MRI findings in the present study do correspond with a previous report of regenerated semitendinosus neotendons evaluated with MRI.23 The cross-sectional area of the semitendinosus tendon (10 to 20 mm2) in the present study does not entirely accord with the data of Hamada et al.19 who found it to be between 6.3 and 15 mm2. This could be
TABLE 3. Fiber Type Distribution and Citrate Synthase (CS) ( mmol/kg dry weight muscle X minutes) Activity in the Semitendinosus Muscle Intraoperatively and Postoperatively in Operated Legs and Postoperatively in Nonoperated Semitendinosus Muscle Type Preop operated leg n = I2 Postop operated leg n = I4 Postop nonoperated leg n = 8 NOTE.
Values
are mean
and standard
I %
50 (14) 54(17) 48(11) deviation.
Type
2A %
29 (IO)
27 (8) 34 (3)
Type
2B %
20 (16) I8 (13) I8 (IO)
Type
2C % l(1) 1 (2) 0 (0)
CS Activity 18 (8) n = 16 18 (6) n = 15 22 (6) n = 8
814 TABLE
K. ERIKSSON ET AL. 4.
Muscle
Fiber Area Postoperatively
(kn?)
Preop operated leg (n = 12) Postop operated leg (n = 14) Postop nonoperated leg (n = 8) NOTE.
Values
are mean
and standard
in Vurious Types in the Operuted Leg
of md
Fibers From the Semitendinosus Muscle IIltrcrc)pemti,,el~ Postoperatively in the Nompertrted Leg
Type
I Area
Type
2A Area
Type
5.250 5,264 5,799
(876) (2.062) (I .432)
6,878 6,770 7.902
( I .444) (2.4 19) (2.999)
6,687 6.781 6,977
5.
Isokinetic
Strength Hamstrings Concentric Peak Torque (NW
Postoperative operated leg Postoperative nonoperated leg Percent lower in operated leg Difference operated Y nonoperated P value NOTE.
Values
are mean
leg
and standard
Type
( I .785) (2,105) ( 1.955)
2C Area
5,142(1.371) 3,777 (456) No fibers present
Mean
Area
5.938 (1.173) 6.5 18 ( I .740) 6.828 ( 1.884)
deviation.
attributable to differences between Japanese and Caucasian anthropometry, because the tendon cross-sectional areas in our study resemble those reported by Tohyama et al., who examined a Caucasian population.30 When harvesting the semitendinosus tendon for ACL reconstruction, all muscles in the thigh were affected by the operation and the preoperative injury, since the total muscle area was about 5% less than that of the healthy side. The main cause of this atrophy was the quadriceps muscle (1 O%), not the hamstrings (1%). The semitendinosus muscle on the operated side was affected more by a reduction of 16% than the healthy side. When the MRI showed no regeneration of the tendon, the reduction in cross-sectional area of the semitendinosus muscle was significantly greater (34%). These patients also had a relative hypertrophy of the ipsilateral semimembranosus muscle belly, expressed as a significant difference in the muscle crosssectional area ratio (operated v nonoperated). The findings regarding greater hypotrophy of the semitendinosus muscle in patients with no regeneration have previously been reported by Eriksson et al.,23 but in that study the semimembranosus muscle was not analyzed for comparison. This agonist (semimembranosus) seems to compensate by hypertrophy for the loss in mass of the semitendinosus muscle if the tendon does not regenerate. Our findings regarding isokinetic hamstring strength support this view because no sig-
TABLE
2B Area
md
94 (25) I04 (25) IO -lO(l5) ,022 deviation.
(9O”/s)
nificant difference was seen between the tendon groups with or without regeneration. Simonian et al.?’ found no compensatory hypertrophy of the agonist muscles in their study of 9 patients. However they did not analyze their subgroups, regenerated (n = 6) and nonregenerated (n = 3), separately. Another difference between their study and ours is that they did not examine the thigh more proximally than 10 cm above the knee joint. The semitendinosus muscle acts as a part of the hamstring complex but does not play a dominant role in the knee flexors except during sitting thigh curls (supine position) when the semitendinosus, gracilis, and sartorius muscles are all strongly activated. In knee flexion exercises in a prone position, the semitendinosus and other hamstring muscles are only moderately activated.31 This was evident when we developed the test protocol for the isokinetic muscle test. Kramer et al.” performed isokinetic flexion tests 21 months after surgery with patients in the prone position and also found that the peak torques were lower in the operated than in the nonoperated leg. In their opinion, there was a distinct trend toward muscle imbalance, particularly in the muscle group affected by the semitendinosus tendon harvest. However, this conclusion could be questioned based on the relatively low specificity of their prone test situation. Most other studies have not stated whether hamstring torque was
in Operated
Hamstrings Concentric Total Work
and
Nonoperated
Hamstrings Eccentric Peak Torque
(J)
WV
53 (15) 62 (17) I5 -9 (I I) ,009
I I2 (27) I39 (30) I9 -28 (18) c.001
Legs Hamstrings Eccentric Total Work
(J) 68 (18) 88 (21) 23 -20 (IO) c.001
(n =
15)
Quadriceps Concentric Peak Torque
(Nm) 181 (34) 226 (55) 20 -45 (38) <.OOl
Quadriceps Concentric Total Work
(J) I82 (32) 234 (53) 22 -52 (45) <.OOl
SEMITENDINOSUS
MUSCLE
calculated with patients in the sitting or the prone position. 4. IY-2 I.25.33-35 Muscle imbalance and quadriceps and hamstring deficits in ACL-deficient knees have been found in previous studiesJh-“x and, when evaluating postoperative values, this must be taken into consideration. Other studies of the thigh cross-sectional area of ACLdeficient knees show deficits of 5% to 10% in the injured compared with the healthy leg.‘7.3Y-Ai It is not yet proven that harvest of the semitendinosus tendon is the main reason for knee flexion strength deficit after ACL reconstruction. It could be the injury itself. In fact, Yasuda et al.” found no difference in isokinetic strength before and 3 months after surgery if only a graft had been harvested from the leg and the ACL reconstruction had been performed on the contralateral side. The isokinetic concentric and eccentric hamstring strengths were significantly lower in the operated than in the nonoperated leg. The same relation was observed in the concentric quadriceps test. These results resemble those in a recent isokinetic follow-up study by Carter and Edinger.JZ who found 25% and 20% deficits in extension and flexion. respectively, in the operated leg 6 months after surgery. However, the results are not entirely comparable because their protocol differed from ours. Lipscomb et a1.,i9 in a retrospective study of 51 patients, found no significant side-to-side difference in hamstring strength (60” v 240” per second) 26 months after surgery, but did find a slight decrease in quadriceps strength at the same follow-up. Sekiya et al.,?’ in a 2-year follow-up study of quadriceps and hamstring isokinetic strength indices (operated v nonoperated leg), reported results similar to ours, with 10% and 20% strength deficit in the hamstrings and quadriceps, respectively. In that study, they also noted a correlation between quadriceps and hamstring isokinetic peak torque indices and the l-leg hop test index. We also found a correlation between the quadriceps isokinetic index and hop test, but no correlation with the hamstring index. Laxity was not correlated to the hop test, quadriceps, or hamstring index in either their study or in ours. Ohkoshi et al.Zz reported 99% hamstring peak torque strength in the operated leg but simultaneously found that peaks were missing in the latter half of the torque curves. They concluded that this was attributable to lack of semitendinosus action and compensation by other muscles. In our study, no obvious differences in torque curves were seen between operated and nonoperated legs. Comparison of the controls with the preoperative
IN ACL SURGERY
815
and postoperative citrate synthase enzyme activity levels, reflecting oxidative capacity, showed no abnormalities in semitendinosus oxidative muscle metabolism. Nor was the fiber type distribution altered in the operated versus nonoperated legs. or in the regenerated versus nonregenerated subgroups. When comparing the fiber type distribution in the semitendinosus muscle with the commonly examined vastus lateralis, the type 1 fiber proportion in our material seemed to be similar to previous findings in ACL-deficient patients.JJ.W The type 1 fiber proportion was similar to that in uninjured male patients of similar age, but type 2B fiber proportion in our study was slightly higher.-1s-J7As for fiber area determinations, no sign of denervation or fiber atrophy was seen. The area values for all fiber types agree with results of investigators concerning the quadriceps muscle in a trained population.“’ These findings support previous electromyographic results reported by Cross et al.“’ who found a normal pattern in the semitendinosus muscle in patients where the gracilis and semitendinosus tendons had previously been harvested for ACL reconstruction. CONCLUSIONS Harvest of the semitendinosus tendon appears to be a well-tolerated procedure in ACL reconstruction surgery, after which a structure that has the appearance of a tendon regenerates in 75% of the patients, but so far there is no histologic confirmation that it is in fact a tendon. In the remaining 25% of patients, the muscle crosssectional area was smaller. This was compensated for by a concomitant increase in the semimembranosus muscle cross-sectional area. In no case were signs of total atrophy or retraction of the semitendinosus muscle seen, and the oxidative muscle potential and fiber type distribution were the same as in the control leg. Nor were there any differences when comparing patients with regenerated and nonregenerated tendons. This was also true when comparing clinical outcome and isokinetic strength, although there was a strength deficit in the operated versus the nonoperated leg. In the present study, the functional and structural morbidity of the semitendinosus graft harvest seemed to be small and the muscle seemed to recover following harvest of its tendon. REFERENCES I. Aglietti P, Buzzi R, Zaccherotti G, De Biase P. Patellar tendon versus doubled semitendinosus and gracilis tendons for ante-
816
2.
3.
4.
5. 6.
7. 8. 9.
IO.
1 I. 12.
13. 14. 15.
16.
17.
18.
K. ERIKSSON ET AL. rior cruciate ligament reconstruction. Am J Sports Med 1994; 22:21 I-218. Aglietti P, Buzzi R, Menchetti PM, Giron F. Arthroscopically assisted semitendinosus and gracilis tendon graft in reconstruction for acute anterior cruciate ligament injuries in athletes. Am J Sports Med 1996;24:726-73 I. Maeda A, Shino K, Horibe S, Nakata K, Buccafusca Cl. Anterior cruciate ligament reconstruction with multistranded autogenous semitendinosus tendon. Am J Sports Med 1996;24: 504-509. Marder RA, Raskind JR, Carroll M. Prospective evaluation of arthroscopically assisted anterior cruciate ligament reconstruction. Patellar tendon versus semitendinosus and gracilis tendons. Am J Sports Med 1991; 19:478-484. Brown CH Jr. Steiner ME. Carson EW. The use of hamstrine tendons for anterior c&ate ligament reconstruction. Tech: nique and results. C/in Sports Med 1993;12:723-756. Ferretti A, Conteduca F, De Carli A, Fontana M, Mariani PP. Results of reconstruction of the anterior cruciate ligament with the tendons of semitendinosus and gracilis in acute capsuloligamentous lesions of the knee. Iral J Orthop Traumaroj 1990; 16:452-458. Meystre JL, Vallotton J, Benvenuti JF. Double semitendinosus anterior cruciate ligament reconstruction: I O-year results. Knee Surg Sports Traumatol Arthrosc 1998;6:76-81. Rosenberg TD. Deffner KT. ACL reconstruction: Semitendinosus tendon is the graft of choice. Orthopedics 1997;20:396398. Kartus J, Stener S, Lindahl S, Engstrom B, Eriksson BI, Karlsson J. Factors affecting donor-site morbidity after anterior cruciate ligament reconstruction using bone-patellar tendon-bone autografts. Knee Surg Sports Traumatol Arthrosc 1997;5:222-228. Kartus J, Magnusson L, Stener S, Brandsson S, Eriksson BI, Karlsson J. Complications following arthroscopic anterior cruciate ligament reconstruction. A 2-5-year follow-up of 604 patients with special emphasis on anterior knee pain. Knee Surg Sports Traumarol Arrhrosc 1999;7:2-8. Kohn D, Sander-Beuermann A. Donor-site morbidity after harvest of a bone-tendon-bone patellar tendon autograft. Knee Surg Sports Traumarol Arrhrok 1994;2:2 19-223. Miller MD. Nichols T. Butler CA. Patella fracture and oroximal patellar tendon rupture following arthroscopic anterior cruciate ligament reconstruction. Arthroscopy 1999; I5:640643. Aglietti P, Buzzi R, D’ Andria S, Zaccherotti G. Patellofemoral problems after intraarticular anterior cruciate ligament reconstruction. C/in Orrhop 1993: 195-204. Sachs RA, Daniel DM, Stone ML, Garfein RF. Patellofemoral problems after anterior cruciate ligament reconstruction. Am J Sports Med 1989; 17:760-765. Shino K, Nakagawa S, Inoue M, Horibe S, Yoneda M. Deterioration of patellofemoral articular surfaces after anterior CNciate ligament reconstruction. Am J Sports Med 1993;2 I :206211. Zarzycki W, Mazurkiewicz S, Wisniewski P. [Research on strength of the grafts that are used in anterior cruciate ligament reconstruction]. Chir Narzadow Ruchu Orrop Pal 1999:64: 293-302. Hamner DL, Brown CH, Jr., Steiner ME, Hecker AT, Hayes WC. Hamstring tendon grafts for reconstruction of the anterior cruciate ligament: biomechanical evaluation of the use of multiple strands and tensioning techniques. J Bone Joinr Surg Am 1999;81:549-557. Rosenberg TD, Franklin JL, Baldwin GN, Nelson KA. Extensor mechanism function after patellar tendon graft harvest for anterior cruciate ligament reconstruction. Am J Sporrs Med 1992;20:5 19-526.
19. Lipscomb AB. Johnston RK, Snyder RB, Warburton MJ. Gilbert PP. Evaluation of hamstring strength following use of semitendinosus and gracilis tendons to reconstruct the anterior cruciate ligament. Am J Sports Med 1982; l&340-342. 20. Harter RA. Ostemig LR, Standifer LW. Isokinetic evaluation of quadriceps and hamstrings symmetry following anterior cruciate ligament reconstruction. Arch Phys Med Rehabil l990:7 I :465-468. 21. Yasuda K, Tsujino J. Ohkoshi Y. Tanabe Y, Kaneda K. Graft site morbidity with autogenous semitendinosus and gracilis tendons. Am J Sporrs Med 1995;23:706-714. 22. Ohkoshi Y. moue C. Yamane S. Hashimoto T, Ishida R. Changes in muscle strength properties caused by harvesting of autogenous semitendinosus tendon for reconstruction of contralateral anterior cruciate ligament. Arrhroscopy 1998; 14:580584. 23. Eriksson K, Larsson H, Wredmark T, Hamberg P. Semitendinosus tendon regeneration after harvesting for ACL reconstruction. A prospective MRI study. Knee Surg Sports Traumarol Arrhrosc 1999;7:220-225. 24. Cross MJ, Roger G. Kujawa P. Anderson IF. Regeneration of the semitendinosus and gracilis tendons following their transection for repair of the anterior cruciate ligament. Am J Sporrs Med 1992;20:22 l-223. 25. Simonian PT. Harrison SD. Cooley VJ. Escabedo EM, Deneka DA, Larson RV. Assessment of morbidity of semitendinosus and gracilis tendon harvest for ACL reconstruction. Am J Knee Surg 1997; 10:54-59. 26. Rosenberg TP. Graf B. Endoscopic technique for ACL reconstruction with Protrac tibia1 guide: EndoButton fixation. Mansfield, MA: Acufex Microsurgical. 1992. 27. Schantz P, Billeter R, Henriksson J, Jansson E. Traininginduced increase in myofibrillar ATPase intermediate fibers in human skeletal muscle. Muscle Nerve 1982;5:628-636. 28. Lin L, Sotonyi P, Somogyi E, et al. Coenzyme QIO content in different parts of the normal human heart. Clin Physiol 1988; 8:39 l-398. 29. Hamada M, Shino K, Mitsuoka T, Abe N, Horibe S. Crosssectional area measurement of the semitendinosus tendon for anterior cruciate ligament reconstruction. Arrhrascopy 1998; 14:696-70 I. 30. Tohyama H, Beynnon BD, Johnson RJ, Nichols CE. Renstrom PA. Morphometry of the semitendinosus and gracilis tendons with application to anterior cruciate ligament reconstruction. Knee Surg Sporrs Traumarol Arrhrosc 1993;l: 143-147. 3 I. Tesch PA. Target bodybuilding. Champaign, IL: Human Kinetics, 1999. 32. Kramer J, Nusca D, Fowler P, Webster-Bogaert S. Knee flexor and extensor strength during concentric and eccentric muscle actions after anterior cruciate ligament reconstruction using the semitendinosus tendon and ligament augmentation device. Am J Sports Med 1993;2 I :285-29 I. 33. Sekiya I, Muneta T, Ogiuchi T, Yagishita K, Yamamoto H. Significance of the single-legged hop test to the anterior cruciate ligament-reconstructed knee in relation to muscle strength and anterior laxity. Am J Sports Med 1998;26:384388. 34. Karlson JA, Steiner ME, Brown CH, Johnston J. Anterior cruciate ligament reconstruction using gracilis and semitendinosus tendons. Comparison of through-the-condyle and over-the-top graft placements. Am J Sports Med 1994;22: 659-666. 35. Li RCT. Isokinetic strength of the quadriceps and hamstrings and functional ability of anterior cruciate deficient knees in recreational athletes. Br J Sports Med I996;30: 16 I - 164. 36. Shirakura K, Kato K, Udagawa E. Characteristics of the isokinetic performance of patients with injured cruciate ligaments. Am J Sporrs Med 1992;20:754-760.
SEMITENDINOSUS 37.
38.
39.
40.
41.
42.
MUSCLE
Lorentzon R, Elmqvist LG. Sjostrom M, Fagerlund M, Fuglmeyer AR. Thigh musculature in relation to chronic anterior cruciate ligament tear: muscle size, morphology. and mechanical output before reconstruction. Am J Sports Med 1989; 17: 423-429. ltoh H. lchihashi N, Maruyama T, Kurosaka M, Hirohata K. Weakness of thigh muscles in individuals sustaining anterior cruciate lieament iniurv. Kobe J Med Sci 1992:38:93-107. Gerber C,-Hoppele; H, Claassen H, Robotti G, Zehnder R, Jakob RP. The lower-extremity musculature in chronic symptomatic instability of the anterior cruciate ligament. J Bone Joint Surg Am 1985;67: 1034-1043. Arangio GA, Chen C. Kalady M. Reed JF 3rd. Thigh muscle size and strength after anterior cruciate ligament reconstruction and rehabilitation. J Orthop Sports Phys Ther 1997;26: 238-243. Kariya Y, Itoh M. Nakamura T, Yagi K, Kurosawa H. Magnetic resonance imagine and soectroscouv of thigh muscles in cruciate ligament insufficiency. Acra &hop S&d 1989;60: 322-325. Carter TR. Edinger S. Isokinetic evaluation of anterior cruciate
43.
44.
45.
46.
47.
IN ACL SURGERY
817
ligament reconstruction: Hamstring versus pateliar tendon. Arlhroscopy 1999; 15: l69- 172. Arvidsson I. Arvidsson H, Eriksson E, Jansson E. Prevention of quadriceps wasting after immobilization: An evaluation of the effect of electrical stimulation. Orrhopedics 1986;9: I5 l91528. Haggmark T, Eriksson E. Jansson E. Muscle fiber type changes in human skeletal muscle after injuries and immobilization. Orthopedics 1986;9: I8 I -I 85. Berg HE, Dudley GA. Hather B, Tesch PA. Work capacity and metabolic and morphologic characteristics of the human quadriceps muscle in response to unloading. Clin Physiol 1993;13: 337-347. Maughan RJ. Nimmo MA. The influence of variations in muscle fibre composition on muscle strength and cross-sectional area in untrained males. J Physiol (land) 1984;35 I :29931 I. Esbiomsson M. Jansson E, Sundberg CJ. Svlven C. Eiken 0. Nygren A, Kaijser L. Muscle fibre tyjes andenzyme activities after training with local leg ischemia in man. Acta Physiol Scmd 1993;148:233-241.