Effects of bone block position and orientation within the tibial tunnel for posterior cruciate ligament graft reconstructions: a cyclic loading study of bone-patellar tendon-bone allografts

Keith Markolf

Although good subjective results have been obtained after posterior cruciate ligament (PCL) reconstruction, objective improvement (as measured by restoration of posterior laxity) has been less predictable. Mariani et al. (4) reported on 24 patients with chronic isolated PCL tears who were treated with PCL reconstruction with bone-patellar tendon-bone autografts. Preoperative side-to-side differences in posterior laxity as measured with a KT-2000 arthrometer (Medmetric Corp., San Diego, California/ averaged 8.38 mm, and postoperative laxity measurements at 26.5 months averaged 4.08 mm. Normal laxity (0 to 2 mm) was restored in 25% of patients and near-normal laxity (3 to 5 mm) was achieved in 54%. In 13% of patients, the amount of laxity was abnormal (6 to 10 mm), and in 8% it was severely abnormal (>10 mm).

Noyes and Barber-Westin (8) reported on 25 patients with isolated PCL tears. They performed PCL reconstruction with allograft alone in 10 patients (bone-patellar tendon-bone in 6 patients and Achilles tendon in 4 patients) or with a ligament augmentation device in 15 patients (13 of the 15 had supplementation of the device with a bone-patellar tendon-bone allograft). At an average 45 months postoperatively, there were no significant objective or subjective differences between groups. However, collective analysis of the two groups showed that near-normal side-to-side difference in laxity, as measured with the KT-1000 arthrometer, was not restored in the majority of patients. In patients with chronic injuries, the side-to-side difference in posterior laxity was less than 2.5 mm in 25%, 3 to 5.5 mm in 12%, and more than 6 mm in 63%. In patients with acute injuries, the side-to-side difference in posterior laxity was 3 to 5.5 mm in 60% and more than 6 mm in 40%.

Abnormal posterior laxity after PCL graft reconstruction could be related to thinning and permanent elongation of the graft as it is cyclically loaded in vivo. (1, 7) A common method of PCL replacement involves use of a transtibial tunnel in which the graft must pass around an acute angle at the posterior tibia. The high local tissue stresses experienced by the graft as it passes around the posterior edge of the tibial tunnel could produce graft thinning and lead to permanent graft elongation, which in turn would be manifested as increased posterior knee laxity.

At the present time, there are no clear surgical guidelines regarding the position and orientation of the bone block within the tibial tunnel. It is typically placed somewhere near the posterior opening of the tibial tunnel, and it may be oriented with the bone facing anteriorly or posteriorly. With the bone block facing posteriorly, the graft tissue has a direct, uninterrupted path around the acute bend at the posterior tunnel edge. However, if the bone block is oriented anteriorly, the course of the graft tissue in the vicinity of the posterior tunnel edge is altered, because the graft must first pass around the end of the bone block and then around the posterior tunnel edge. This could change the stress distribution within the graft tissue where it makes the bend. Recessing the bone block at a distance from the tunnel edge could also produce changes, because this exposes a different portion of the graft tissue to the acute bend. Both of these factors could affect the cyclic performance of the graft construct and thereby affect postoperative knee laxity.

The purpose of this study was to measure the mechanical responses of bone-patellar tendon-bone allografts to cyclic loading tests for different bone block orientations within the tibial tunnel (anterior versus posterior) and positions within the tibial tunnel (flush with the posterior tunnel opening versus recessed 1 cm from the opening).

METHODS

Test Protocol

Fresh-frozen patella-patellar tendon-tibia specimens were divided into medial and lateral halves and 11-mm wide grafts were prepared. Graft pairs were assigned to both tibias from 64-year-old and 43-year-old male specimens. Tibial tunnels, 11 mm in diameter, which were drilled at an angle of 60[degrees] to horizontal, exited the tibia posteriorly at the center of the PCL attachment.

An area micrometer with a test section 11 mm wide by 5 mm deep was used to measure thickness of the graft before and after cyclic testing; linear displacement of the dial indicator was recorded under a constant pressure of 0.12 MPa for 2 minutes, as described by Butler et al. (2) Measurements were taken at three different sites on the patellar tendon grafts. Since the graft lengths varied, the measurement sites were scaled in proportion to the free length of the graft. Site 1 was approximately 8 to 10 mm from the tibial bone block, site 3 was located at the mid-point of the graft (approximately 20 to 25 mm from the tibial bone block), and site 2 was halfway between sites 1 and 3 (approximately 14 to ]8 mm from the tibial bone block) (Fig. 1).

[FIGURE 1 OMITTED]

The tibial specimen (with distal end of the graft secured by clamped cables) was mounted onto an MTS model 812 servohydraulic materials test machine (MTS Systems Corp., Minneapolis, Minnesota) for cyclic loading. The tibial fixture was adjusted such that the angle of pull of the graft relative to the tibial plateau was 45[degrees]. Each graft construct was subjected to 2000 cycles of tensile loading at 0.5 Hz; the load applied to the graft varied from 20 to 220 N. During testing, the graft was continuously moistened with physiologic saline solution applied by using a drip system. All testing was performed at room temperature. Force versus elongation curves were monitored throughout testing. Because the target maximum load of 220 N varied slightly during the cyclic testing (because of periodic drift within the load-control electronics of the MTS machine), elongation of the graft during a loading cycle between 20 and 200 N of applied force (as measured by actuator displacement) was tabulated for cycles 1, 6, and 2000. The lower maximum load used for calculation of graft elongations assured that values would be available for all specimens. Further details of specimen preparation, graft thickness measurement, and the test protocols can be found in a prior publication from our laboratory. (7)

Experimental Design

Two variants of bone block placement were studied in two separate test series. In series 1 (27 graft pairs), all bone blocks were placed flush with the posterior opening of the tibial tunnel; one graft was oriented with the bone block facing posteriorly in the tunnel (Fig. 1A), whereas its pair was oriented with the bone block facing anteriorly in the tunnel (Fig. 1B). The mean age of the graft specimens in series 1 was 42.5 years (range, 18 to 65).

In series 2 (20 graft pairs), all grafts were oriented with the bone block facing posteriorly; one graft had the bone block flush with the posterior tunnel opening (Fig. 1A), whereas its pair had the bone block recessed into the tunnel, 1.0 cm away from the posterior opening (Fig. 2). Specimen ages were not available for grafts in series 2.

[FIGURE 2 OMITTED]

Statistical Analysis

Variables analyzed included measurements of graft thickness (at three graft sites), elongation of the graft during a loading cycle (between 20 and 200 N of applied force), and length of the graft (as recorded at 200 N of applied force). A two-way repeated-measures analysis of variance (ANOVA) model was used to determine the significance of differences between mean values of graft thickness at each measurement site (before and after cyclic testing) between tunnel position variants; the variants were anterior versus posterior bone block orientations and flush versus recessed bone block placements. Similar models were used to compare thickness measurements between sites for a given variant before and after cyclic testing. A two-way repeated-measures ANOVA model was also used to determine the significance of differences between mean values of graft elongation during a loading cycle and permanent increases in graft length with both graft variants between cycles 1, 6, and 2000. The level of significance was P < 0.05.

RESULTS

Series 1 (Bone Block Orientation)

In series 1, 3 of the 27 grafts (11.1%) with posterior-facing bone blocks failed at the acute corner before 2000 cycles of testing could be completed; cycles to failure for the failed grafts were 1300, 1450, and 1890, respectively. All specimens with anterior bone blocks survived the testing intact.

There was no significant difference in mean values of any measured parameter (graft thickness, graft elongation during a loading cycle, or change in graft length) between those posterior-facing block grafts in which the graft pair survived the testing and those posterior block grafts in which the graft pair did not (unpaired analysis). The donor ages of the three posterior bone block specimens that failed were 16, 35, and 50 years.

For the 24 graft pairs that survived testing, there was no significant difference in mean initial graft thickness between bone block orientations (for a given site) or between sites (for a given bone block orientation) (Table 1). Mean graft thickness was significantly reduced at all sites after cyclic loading with both bone block orientations (Table 1).

The absolute change and percentage change in graft thickness were significantly less with anterior bone blocks than with posterior bone blocks at sites 1 and 2 (Table 1). The mean reductions in graft thickness at site 1 were 21.4% with the posterior block and 10.5% with the anterior block; corresponding reductions at site 2 were 10.6% and 5.7%, respectively (Table 1). For each bone block orientation, mean reductions in graft thickness at all three sites were significantly different from each other (Table 1).

There were no significant differences in mean graft elongation during a loading cycle (between 20 and 200 N of applied tensile force) between the two bone block orientations for cycles 1, 6, or 2000 (Table 2). Mean graft elongations during a loading cycle with both bone block orientations decreased approximately 50% alter 2000 cycles of loading (Table 2).

Grafts with both bone block orientations significantly increased in length after cyclic loading (Table 3). Mean increases in graft length with anterior and posterior bone block orientations between cycles 6 and 2000 and between cycles 1 and 2000 were significantly different from each other (Table 3). The mean increase in graft length after 2000 cycles was 6.67 mm with the posterior block and 5.50 mm with the anterior block (Table 3). A substantial portion of these graft length increases (40.0% with the anterior block and 42.9% with the posterior block) occurred between cycle 1 and cycle 6 (Table 3).

Series 2 (Bone Block Position)

In series 2, 3 of 20 grafts (15.0%) with recessed bone blocks failed at the acute corner before 2000 cycles of testing could be completed; cycles to failure for failed grafts were 1652, 1850, and 1942, respectively. All specimens with flush bone blocks survived the testing intact.

There was no significant difference in mean values of any measured parameter (graft thickness, graft elongation during a loading cycle, or change in graft length) between those recessed block grafts in which the graft pair survived the testing and those recessed block grafts in which the pair did not (unpaired analysis).

For the 17 graft pairs that survived testing, mean initial graft thicknesses for flush specimens at all three sites were significantly different from each other (Table 4); mean initial graft thicknesses for sites 2 and 3 were significantly different from site 1 with recessed grafts (Table 4). There were no significant differences in mean initial graft thicknesses between bone block positions for any given site (Table 4).

Mean graft thickness was significantly reduced at all sites after cyclic loading with both bone block positions (Table 4). The absolute change and percentage change in graft thicknesses were significantly less with flush bone blocks than with recessed bone blocks at all three sites (Table 4). The mean reductions in graft thickness at site 1 were 23.9% with the flush block and 26.0% with the recessed block; corresponding reductions at site 2 were 12.9% and 23.9% and, at site 3, 6.8% and 13.3% (Table 4). The mean reduction in graft thickness at site 1 with the flush bone block (where the graft tissue passed around the acute bend) was not significantly different from that at site 2 with the recessed graft (where a different portion of the graft tissue passed around the acute bend) (Table 4). For both bone block positions, mean reductions in graft thickness at site 3 were significantly less than the corresponding changes at sites 1 and 2 (Table 4).

There were no significant differences in mean graft elongation during a loading cycle (between 20 and 200 N of applied tensile force) between the two bone block orientations for cycles 1, 6, or 2000 (Table 5). Mean graft elongations during a loading cycle with both bone block orientations decreased approximately 40% after 2000 cycles of loading (Table 5).

Grafts with both bone block orientations significantly increased in length after cyclic loading (Table 6). Mean increases in graft length for anterior and posterior bone block orientations between cycles 1 and 6, cycles 6 and 2000, and cycles 1 and 2000 were significantly different from each other (Table 6). The mean increase in graft length after 2000 cycles was 6.68 mm with the flush block and 9.39 mm with the recessed bone block; these were significantly different from each other (Table 6). A substantial portion of these graft length increases (19.1% with the flush bone block and 18.3% with the recessed block) occurred between cycle 1 and cycle 6 (Table 6).

DISCUSSION

The work presented here represents an extension of a prior study performed in this laboratory, (7) In that study, tibial tunnel versus tibial inlay grafts were tested by using essentially the same methods presented here. However, there is an important difference between the two studies. In the prior study, specimens were cycled between 50 and 300 N and, in the present study, grafts were cycled between 20 and 220 N. In the prior study, 32.2% of the grafts failed before 2000 cycles of testing could be completed. In hopes of reducing the failure rate, we decided to reduce the upper and lower load limits in the present study and thereby obtain more quantitative data related to graft thickness and graft elongation. The grafts used in the prior study were positioned in the tunnel with the bone block posterior and flush with the posterior opening of the tunnel. This configuration was identical to the present series 1 tests with a posterior bone block and series 2 tests with a flush bone block.

The failure rate for series 1 grafts with posterior bone blocks (11%) and series 2 grafts with flush bone blocks (0%) is lower than the failure rate of 32.2% for the grafts of the prior study. The percent changes in graft thickness after cyclic loading were approximately 47% to 74% less than corresponding values from the prior study, whereas changes in graft elongation were approximately 28% to 34% lower. These lower values for the present study are consistent with the 80 N (26.6%) reduction in maximum tensile force applied to the grafts.

In series 1 specimens, there was no significant difference in mean reduction in graft thickness between anterior and posterior bone block orientations at site 3. This would be expected because this site is farthest from the acute corner of the tibia. However, mean reductions in graft thickness at sites 1 and 2 with the anterior bone block were significantly less than changes at these sites with the posterior bone block. We believe this difference is related to the altered course of the graft tissue with an anterior bone block. With a posterior bone block, the graft passes directly around the tunnel edge, making an acute bend. With the anterior bone block, the graft fibers must first pass around the end of the bone block and then around the tunnel edge. This effectively increases the overall radius of curvature of the graft bend, which would theoretically reduce the stresses within the graft tissue. This could also explain why the graft elongation after 2000 loading cycles with the anterior bone block was 1.17 mm (17.5%) less than with the posterior bone block.

The results for bone block placement within the tibial tunnel require a more careful analysis. Graft elongations after 2000 cycles of loading with flush grafts were 28.8% less than with recessed grafts. This result is more difficult to explain and could be related to variations in fiber strength along the length of the graft. We hypothesize that the graft fibers near the region where they insert into the bone block may be locally stronger” than those removed from this location. Thus, when the bone block is recessed into the tunnel 1 cm away from the acute bone edge, the graft fibers passing around the corner could be mechanically inferior and more prone to thinning and elongation. When the bone block is flush with the posterior tunnel opening, the fibers passing over the tunnel edge are closer to the bone block, and perhaps stronger, it is possible that the effects of recessing the graft would be different with the bone block facing anteriorly; this was not studied in our test protocol.

The comparison of graft thickness measurement sites for series 2 specimens was not direct. For example, with flush grafts, tissue at site 1 passed around the posterior tunnel edge. With recessed grafts, tissue at site 2 passed around the tunnel edge. The mean reduction in graft thickness at site 1 with flush bone blocks was 23.9%; the mean reduction at site 2 with recessed grafts was also 23.9%. This was a consistent finding because these two measurement sites were both at the acute graft bend. The mean reduction of 12.9% at site 2 with flush grafts is consistent with the 13.3% for site 3 with recessed grafts, because these measurement sites are roughly the same distance from the tunnel edge. The one finding that is somewhat puzzling is the mean graft thickness reduction of 26.01% at site 1 for recessed grafts, because, theoretically, graft tissue at this location (which is well within the tunnel) should experience somewhat less tensile force because of frictional losses around the acute graft bend. The greater reductions in graft thickness at site 3 for recessed grafts (13.3%) compared with flush grafts (6.77%) appears reasonable because site 3 for recessed grafts is closer to the tunnel edge.

The findings related to graft elongation (between 20 and 200 N) during a loading cycle were very similar to those of the prior study. (7) This quantity is a rough measure of specimen stiffness. In the prior study, we found that this measurement was reduced approximately 50% between cycle 1 and cycle 2000, indicating that graft stiffness had approximately doubled. In the present study, the corresponding reductions in graft elongation during a loading cycle were approximately 38% to 49%. Stiffening of the graft after cyclic loading could be the result of permanent changes in collagen fiber alignment in the central portion of the graft or in the vicinity of the acute bend. A more parallel fiber alignment could theoretically stiffen the graft tissue.

There was a puzzling difference in graft failure rates between series I and series 2 specimens under identical test conditions. The failure rate for series 1 grafts with posterior bone blocks was 11%, whereas that for series 2 grafts with flush bone blocks was 0%. Series 1 tests were performed first and, at that time, we did not appreciate the possible importance of graft position in the tibial tunnel. Because the focus of series 1 tests was to study anterior versus posterior bone block orientations, it is possible that less attention was paid to placing the bone block exactly flush with the posterior tunnel opening. Some of the bone blocks in series 1 tests may have been recessed slightly more than others, thus leading to a higher failure rate for these grafts. In fact, this possibility led us to the series 2 tests, in which bone block position in the tunnel was selected as a variable for study.

Differences between our cyclic loading test configuration and the in vivo environment deserve comment. The exact tensile loading conditions for a PCL graft during in vivo activities are unknown; tensile force is most likely applied to the graft tissue in combination with some degree of knee flexion-extension. This would change the angle of the graft relative to the tibial plateau during the loading cycle. Our test configuration represented a simplified case; the tibia was stationary, and cyclic tensile force was applied to the proximal bone block with the graft at a constant angle relative to the tibial plateau. The purpose of our tests was to measure differences in graft survivability, thinning, and elongation resulting from changes in placement of the distal bone block within the tibial tunnel. Changing the angle of pull (relative to the tibia) during the test would have added an additional degree of complexity and, in our opinion, would not have substantially altered the comparative results.

The four tibias (from two pairs) selected for these studies had exceptional bone quality. The tunnel edges were repeatedly checked for mechanical degradation and none was observed. The graft pairs were randomly assigned between tibias in approximately equal numbers. This provided a powerful repeated-measures study design.

A surgeon typically has flexibility in positioning and fixing the bone blocks of the graft within the tibial and femoral tunnels. Prior studies from this laboratory have shown the advantages of first fixing the bone block of the graft in the tibial tunnel, and then pretensioning the femoral end of the graft before fixing it. (5, 6) Those studies demonstrated that when the femoral end of the graft was tensioned, all of the pretension force was applied directly to the intraarticular portion of the graft, avoiding the frictional losses that would occur at the tunnel edge if the tibial end of the graft were tensioned. The present study further suggests that before the bone block is fixed in the tibial tunnel, it should be positioned with its end flush to the posterior tibial opening, and with the bone block facing anteriorly.

One possible problem with orienting the bone block anteriorly is the possibility that it could effectively shorten the graft. This could compromise fixation of the bone block within the femoral tunnel if the allograft tissue was somewhat shorter than required. (3) If the allograft were longer than required, the effective shortening produced by anterior bone block orientation could prove to be an advantage, as this could result in enough graft length to allow the bone block to be fully recessed into the femoral tunnel.

The clinical significance of differences in permanent graft elongation between test configurations deserves special mention. On average, recessed grafts elongated 2.71 mm more than flush grafts, and grafts with anterior bone blocks elongated 1.17 mm more than those with posterior bone blocks. In our view, elongation differences on the order of 2 mm could be important clinically in terms of posterior knee stability. The fact that substantial portions of the permanent elongation occurred after only 6 cycles of loading emphasizes the importance of preconditioning the graft tissues before they are fixed in the knee. If not, unwanted posterior laxity could develop soon after the patient becomes ambulatory.

Our recommendations related to bone block position and orientation are based on results of in vitro testing of grafts that were fixed in the tunnel without the use of standard surgical hardware such as an interference screw, Endobutton (Smith & Nephew Endoscopy, Andover, Massachusetts), or screw and washer. Rigidity of fixation could also be a factor in the ultimate elongation of a cyclically loaded graft. Fixation issues related to bone quality have not been addressed in this study. Although we initially suspected that specimens from younger donors would perform better in our tests, this was not the case. Regression analyses revealed no significant correlations of any measured parameter with specimen age.

TABLE 1

Series 1: Effects of Cyclic Loading on Graft Thickness for the 24 Pairs

of Grafts that Completed 2000 Cycles (Mean [+ or -] SD)

Graft thickness (mm)

Posterior bone block

Variable Site 1 Site 2

Before cycling 3.33 (0.60) 3.22 (0.49)

After 2000 cycles 2.62 (0.54) (a) 2.89 (0.51) (a)

Difference -0.71 (0.21) -0.33 (0.16) (b)

% change -21.4 (5.4) -10.6 (5.4) (b)

Graft thickness (mm)

Posterior bone Anterior bone

block block

Variable Site 3 Site 1

Before cycling 3.41 (0.54) 3.52 (0.53)

After 2000 cycles 3.28 (0.53) (a) 3.15 (0.53) (a)

Difference -0.13 (0.16) (b,c) -0.37 (0.23) (d)

% change -3.8 (4.8) (b,c) -10.5 (6.7) (d)

Graft thickness (mm)

Anterior bone block

Variable Site 2 Site 3

Before cycling 3.50 (0.48) 3.58 (0.53)

After 2000 cycles 3.29 (0.43) (a) 3.44 (0.47) (a)

Difference -0.21 (0.19) (b,d) -0.14 (0.12) (b,c)

% change -5.7 (5.1) (b,d) -3.6 (3.1) (b,c)

(a) Significantly different from before cycling.

(b) Significantly different from site 1.

(d) Significantly different from posterior.

TABLE 2

Series 1: Effects of Cyclic Loading on Graft Elongation in 24

Graft Pairs that Completed 2000 Cycles (Mean [+ or -] SD) (a)

Graft elongation (mm)

Cycle Posterior bone block Anterior bone block

1 5.60 (0.92) 5.57 (0.86)

6 3.29 (0.30) (b) 3.16 (0.37) (b)

2000 2.81 (0.38) (c) 2.71 (0.36) (c)

(a) Recorded between 20 and 200 N of tensile force.

(b) Significantly different from cycle 1.

TABLE 3

Series 1: Effects of Cyclic Loading on Graft Length for 21 Graft

Pairs (Mean [+ or -] SD) (a)

Graft, length increase (mm)

Difference Posterior bone block Anterior bone block

Cycle 6-cycle 1 2.67 (0.45) 2.36 (0.65)

Cycle 2000-cycle 6 4.00 (0.70) 3.15 (0.77) (b)

Cycle 2000-cycle 1 6.67 (0.66) 5.50 (0.78) (b)

(a) Recorded at 200 N of graft force.

(b) Significantly different from posterior.

TABLE 4

Series 2: Effects of Cyclic Loading on Graft Thickness for 17

Graft Pairs that Completed 2000 Cycles (Mean [+ or -] SD)

Graft thickness (mm)

Flush

Variable Site 1 Site 2

Before cycling 3.85 (0.56) 3.53 (0.47) (a)

After 2000 cycles 2.93 (0.63) (c) 3.08 (0.51) (a,c)

Difference -0.91 (0.53) -0.45 (0.18) (a)

% change -23.89 (13.24) -12.95 (5.45) (a)

Graft thickness (mm)

Flush Recessed

Variable Site 3 Site 1

Before cycling 3.38 (0.41) (a,b) 3.81 (0.60)

After 2000 cycles 3.15 (0.44) (c) 2.82 (0.40) (c)

Difference -0.23 (0.15) (a,b) -1.03 (0.51) (d)

% change -6.77 (4.56) (a,b) -26.01 (11.37) (d)

Graft thickness (mm)

Recessed

Variable Site 2 Site 3

Before cycling 3.32 (0.31) (a) 3.33 (0.39) (a)

After 2000 cycles 2.53 (0.41) (c,d) 2.90 (0.46) (c)

Difference -0.79 (0.27) (d) -0.44 (0.19) (a,b,d)

% change -23.86 (8.63) (d) -13.3 (5.76) (a,b,d)

(a) Significantly different from site 1.

(b) Significantly different, from site 2.

(d) Significantly different from flush.

TABLE 5

Series 2: Effects of Cyclic Loading on Graft Elongation for 17

Specimens that Completed 2000 Cycles (Mean [+ or -] SD) (a)

Graft elongation (mm)

Cycle Flush Recessed

1 5.30 (1.02) 5.62 (1.21)

6 3.78 (0.38) (b) 3.83 (0.28) (b)

2000 3.26 (0.62) (c) 3.33 (0.49) (c)

(a) Recorded between 20 and 200 N of tensile force.

(b) Significantly different from cycle 1.

TABLE 6

Series 2: Effects of Cyclic Loading on Graft Length for 17 Graft

Pairs that Completed 2000 Cycles (Mean [+ or -] SD) (a)

Graft length increase (min)

Difference Flush Recessed

Cycle 6-cycle 1 1.28 (0.38) 1.72 (0.61) (b)

Cycle 2000-cycle 6 5.39 (2.47) 7.68 (5.36)

Cycle 2000-cycle 1 6.68 (2.65) 9.39 (5.77) (b)

(a) Recorded at 200 N of graft force.

(b) Significantly different from flush.

ACKNOWLEDGMENT

This study was supported by a grant from the Musculo-skeletal Transplant Foundation, Edison, New Jersey, which also provided the tissue specimens used for this study.

REFERENCES

(1.) Bergfeld JA, McAllister DR, Parker RD, et al: A biomechanical comparison of posterior cruciate ligament reconstruction techniques. Am J Sports Med 29: 129-136, 2001

(2.) Butler DL, Grood ES, Noyes FR, et al: Effects of structure and strain measurement technique on the material properties of young human tendons and fascia. J Biomech 17: 579-596, 1984

(3.) Clayer M, Atkinson R: A method of pre-operative assessment for posterior cruciate ligament reconstruction. Aust N Z J Surg 64: 319-321, 1994

(4.) Mariani PP, Adriani E, Santori N, et al: Arthroscopic posterior cruciate ligament reconstruction with bone-tendon-bone patellar graft. Knee Surg Sports Traumatol Arthrosc 5: 239-244, 1997

(5.) Markolf KL, Slauterbeck JR, Armstrong KL, et al: A biomechanical study of replacement of the posterior cruciate ligament with a graft. Part I. Isometry, pre-tension of the graft, and anterior-posterior laxity. J Bone Joint Surg 79A: 375-380, 1997

(6.) Markolf KL, Slauterbeck JR, Armstrong KL, et al: A biomechanical study of replacement of the posterior cruciate ligament with a graft. Part II. Forces in the graft compared with forces in the intact ligament. J Bone Joint Surg 79A: 381-386, 1997

(7.) Markolf KL, Zemanovic JR, McAllister DR: Cyclic loading of posterior cruciate ligament replacements fixed with tibial tunnel and tibial inlay methods. J Bone Joint Surg 84A: 518-524, 2002

(8.) Noyes FR, Barber-Westin SD: Posterior cruciate ligament allograft reconstruction with and without a ligament augmentation device. Arthroscopy 10: 371-382, 1994

Keith Markolf, * ([dagger]) PhD, Mark Davies, * MD, Bojan Zoric, * MD, and David McAllister, ([double dagger]) MD

From the * Biomechanics Research Section, and ([double dagger]) David Geffen School of Medicine at UCLA, Department of Orthopaedic Surgery, University of California, Los Angeles, Los Angeles, California

([dagger]) Address correspondence and reprint requests to Keith Markolf, PhD, Biomechanics Research Section, David Geffen School of Medicine at UCLA, Department of Orthopaedic Surgery, University of California, Los Angeles, 21-67 UCLA Rehabilitation Center, 1000 Veteran Avenue, Los Angeles, CA 90095-6902.

No author or related institution has received any financial benefit from research in this study. See “Acknowledgment” for funding information.

Soft tissue restraints to lateral patellar translation in the human knee

Quantification of Posterior Capsule Tightness and Motion Loss in Patients with Shoulder Impingement