Arthroscopy: The Journal of Arthroscopic and Related Surgery 8(3):375-379Published by Raven Press, Ltd. © 1992 Arthroscopy Association of North America

Case Report

Arthroscopic Roofplasty: A Method for Correcting anExtension Deficit Caused by Roof Impingement of an

Anterior Cruciate Ligament Graft

Stephen M. Howell, M.D., Major, USAFR, MC

Summary: Intercondylar roof impingement should be suspected in any patient having dif-ficulty regaining knee extension following an anterior cruciate ligament (ACL) reconstruc-tion. An arthroscopically performed roofplasty can eliminate the clinical complaints. The extent of bone removal can be planned from studying a lateral radiograph taken with the knee in terminal extension. Extension exercises or passive extension devices should not be used in patients with extension deficits caused by roof impingement because they may further damage the graft. Key Words: Roofplasty—Roof impingement—Graft, anterior cruciate ligament—Extension deficit.

Intercondylar notch impingement of an anterior cru-ciate ligament (ACL) graft is a perplexing problem con-fronting the reconstructive knee surgeon. Skilled knee surgeons, despite their keen awareness of this problem, continue to report late complications of pain, limited knee extension, synovitis, graft abrasion, and late graft rupture attributed to graft impingement (1-4).

At least three anatomic causes of ACL graft impinge-ment have been documented: (a) The volume of the intercondylar notch may be congenitally small, resulting in stenosis of the ACL by a narrow notch (5). (b) Chroni-cally unstable knees acquire osteophytes that stenose the aperture of the intercondylar notch. These osteophytes, if not removed during the ACL reconstruction, can result in late graft abrasion (6,7). (c) Tibial tunnel placement

anterior to the slope of the intercondylar roof can result in graft impingement as the knee is extended (2, 8-10).

This report describes the clinical course of a pa-tient with roof impingement of an ACL graft caused by an anteriorly placed tibial tunnel. Serial magnetic reso-nance (MR) findings of the graft are presented and suc-cessful arthroscopic correction of the impingement by a delayed roofplasty is described.

CASE REPORTA 30-year-old man who is a recreational basket-ball

player suffered an acute ACL rupture during a noncon-tact deceleration maneuver while playing basketball. Six weeks following the injury he reported complaints of pain, effusion, and limited knee extension to the ortho-paedic clinic. Physical examination revealed a moderate knee effusion, range of motion from 15 to 115˚, a posi-tive Lachman test, no joint line pain, and no varus/valgus or posterior instability. Pivot shift testing was prevented by pain. KT-1000 ligament laxity testing (MedMetric, San Diego, CA, U.S.A.) revealed an increase in anterior tibial translation of 2.5 mm at 89 newton (N) (20#) and 5 mm during a maximal manual drawer test compared

From the SGHT/David Grant Medical Center, Travis AFB, California, and the Department of Orthopedics: University of California, Davis Medical Center, Sacramento, California.

Address correspondence and reprint requests to Dr. Stephen M. Howell, 7601 Timberlake, Suite 103, Sacramento, CA 95823, U.S.A.

The views expressed herein are those of the author and do not reflect the official policy or position of the United States Department of Defense or the United States Government.

to the normal side.An arthroscopically assisted patellar tendon bone

reconstruction was performed 8 weeks postinjury. Knee motion under anesthesia, before reconstruction, was from 10 to 135˚. Pivot shift was grade 2. There were no meniscal lesions blocking extension. Femoral and tibial pilot pins were inserted by using a front entry guide system (Acufex Microsurgical, Norwood, MA, U.S.A.). The tibial guide pin was placed 5 mm anterior and medial to the center of the ACL insertion. A tensi-ometer (Acufex Microsurgical, Norwood* MA, U.S.A.) confirmed that the tibial-femoral separation distance decreased 1 mm during knee flexion from 10 to 90˚. A 10 mm wide patellar tendon graft was harvested and secured in the femoral and tibial tunnels with the medial-lateral dimensions of the tendon rotated 90˚ and ori-ented in the sagittal plane. Side wall impingement was corrected by a wallplasty that removed -5 mm of the medial aspect of the lateral femoral condyle. Limited knee extension, caused by preoperative soft tissue con-tracture, prevented complete assessment of the potential for intercondylar roof impingement.

A roofplasty did not appear to be needed. An extraar-ticular augmentation was performed by securing the posterior 12 mm of the iliotibial band to the posterior femur with a screw and washer.

Postoperatively, the knee was placed in a contin-uous passive motion machine for 4 days until 90˚ of flexion was regained. A long leg hinged brace with an extension stop at 35˚ of flexion was worn full time for the first 5 weeks. Full weight bearing was begun at 7 weeks. A derotation brace with an extension stop at 35˚ of flexion was used only during strengthening exercises (.5 h/day) for the next 4 1/2 months. A brace was not worn for walking and activities of daily living. At 6 months the extension stop was removed from the dero-tation brace and progressive return to athletics was per-mitted. The patient was enrolled in a daily supervised physical therapy program for the first 6 months.

As part of a separate ongoing study, serial sagittal MR scans were obtained at weeks 1, 6, 11, 24, and 48 postoperatively to follow any time-related changes in the ACL graft appearance. The MR scans were per-formed using a 0.35 Tesla superconducting magnet with a dedicated quadrature detection knee coil (Diasonics, San Francisco, CA, U.S.A.). Imaging was confined to ten 2.5 mm thick sagittal sections (0.625 mm2 pixels) centered about the intercondylar region of the knee. The knee was externally rotated 10 to 15˚ to align the graft

optimally in the sagittal plane. Image acquisition was performed with the standard spin-echo technique using a 1500-msec TR and 50-msec TE. Encoding and recon-struction was performed with the standard two-dimen-sional Fourier transformation technique using 256 phase encoding steps and two excitations (15 min acquisition time).

The patient’s progress in physical therapy plateaued at 12 weeks postoperatively. He complained of pain, an effusion, and a persistent 10˚ flexion contracture. Physi-cal exam findings supported these complaints. A review of the first three MR scans revealed a step-wise deterio-ration of the appearance of the graft. At 1 week the graft was of low signal intensity, black and homogeneous in appearance from origin to insertion, and indistinguish-able from the MR appearance of the posterior cruciate ligament and the quadriceps tendon (Fig. 1). The knee was in -20˚ of flexion during the imaging procedure, with the graft and the intercondylar roof in direct con-tact. By week 6 the graft began to thin in its sagittal dimension due to the impingement of the intercondylar roof as the knee regained extension (Fig. 2). The MR scan at 11 weeks revealed regionalized edema confined to the distal one-third of the graft. The graft almost

FIG. 1. One-week graft of low signal intensity, and homogeneous in appearance from origin to insertion, and indistinguishable from the magnetic resonance (MR) appearance of the posterior cruciate ligament and the quadriceps tendon. The graft was aligned adjacent to the intercondylar roof in the partially flexed knee (20˚) (arrow).

S.M. HOWELL376

Arthroscopy, Vol. 8, No. 3, 1992

Because of the MR relationship of the graft to the intercondylar roof, a lateral radiograph was obtained with the knee in maximal extension (Fig. 4). The tibial tunnel was noted to be located partially anterior to the intercondylar roof verifying the impingement noticed on MR scanning (2,8).

A repeat arthroscopy was performed at 14 weeks. Adhesions in the suprapatellar pouch were minimal and did not require removal. The fat pad was firmer in con-sistency than normal, but not fibrosed. Limited removal of the fat pad was required to view the graft. A roof-plasty was performed. The inclination of the intercondy-lar roof was made more horizontal by removing 8 mm of bone. The postroofplasty notch view displays the loca-tion and extent of bone removal (Fig. 5). A manipulation using anesthesia was not performed.

The patient’s complaints of pain, effusion, and exten-sion deficit disappeared by 3 weeks after the roofplasty. The knee regained 10˚ of extension with a motion arc of 0 to 135˚. An MR scan of the ACL graft at 24 weeks (10 weeks after the delayed roofplasty) revealed restoration of a normal MR signal.

FIG. 2. Week 6. Graft beginning to thin in its sagittal dimension and bow posteri-orly due to the direct impingement by the intercondylar roof as the knee regained extension.

FIG. 3. The MR image at 11 weeks revealed regionalized edema confined to the distal one-third of the graft. Note the direct contact between the roof and graft without any intervening soft tissue (arrow). The graft appeared discontinuous, however; the knee remained stable.

appeared discontinuous (Fig. 3). There was no inter-posed soft tissue between the graft and the intercondylar roof on any of the MR scans. The impingement of the graft was caused by the intercondylar roof.

FIG. 4. A lateral radiograph with the knee in maximal extension (-10˚ of flexion) revealed that the tibial tunnel was located partially anterior to the slope of the intercondylar roof, verifying the impingement noticed on MR scanning.

ROOFPLASTY FOR EXTENSION DEFICIT 377

Arthroscopy, Vol. 8, No. 3, 1992

The increased MR signal under the distal edge of the intercondylar roof regained the low signal intensity identical to the 1 week appearance of the graft (Fig. 6).

The clinical result of the reconstruction was excel-lent at 25 months. The range of motion in the operated knee equaled that of the normal knee (0 to 135˚). KT-1000 ligament laxity testing revealed equal anterior translation during an 89 N and manual maximum test (MMT) test. The involved knee achieved 100% of the distance jumped by the normal knee in the single-leg hop test (11). The Lysholm score was 98 (12). An MR scan at 48 weeks confirmed that the ACL graft main-tained a normal MR signal.

DISCUSSIONThe MR appearance of impinged and unimpinged

ACL grafts has been studied over time and is markedly different (2,13). Grafts placed in tibial tunnels anterior to the slope of the intercondylar roof suffer from roof impingement (2,8-10). Characteristically, roof impinge-ment results in an increase in signal intensity in the distal portion of the graft by 3 months postimplanta-tion. An impinged graft directly contacts the intercondy-lar roof and may be seen to bow posteriorly if the scan

is obtained with the knee maximally extended (2,13). This regionalized signal increase does not spontaneously revert to a low signal intensity (13). The signal increase persists during the first 3 years of implantation (2).

In contrast, grafts placed posterior to the slope of the intercondylar roof are not impinged (2.8,10). Unimpinged grafts maintain a low signal intensity during the entire first year of implantation (2). They do not acquire a regionalized MR signal increase. Clinically, knees with unimpinged grafts are more likely to regain full extension than impinged grafts (2).

Clearly, the graft in this patient was subjected to roof impingement. The lateral radiograph confirmed that the tibial tunnel was anterior to the slope of the inter-condylar roof. The signal increase of the distal one-third of the graft on the 11-week MR scan was diagnostic. Direct contact between the intercondylar roof and the graft was seen on all the MR scans.

There was no interposed soft tissue on the MR scan to suggest that a “cyclops syndrome” was the cause of the impingement (3). Therefore, the extension deficit was caused by the graft impinging against the bone of the intercondylar roof.

The delayed roofplasty eliminated the flexion con-tracture within 3 weeks of raising the intercondylar roof.

FIG. 5. A delayed roofplasty was performed that elevated the slope of the inter-condylar roof. Eight millimeter of bone was removed from the roof arthroscopi-cally. The postroofplasty notch view displays the location and extent of bone removal.

FIG. 6. An MR scan of the anterior cruciate ligament (ACL) graft at 24 weeks (10 weeks after the delayed roofplasty) revealed restoration of a normal MR signal. The edematous portion under the distal edge of the intercondylar roof had regained the low signal intensity present at 1 week. The homogeneous, low signal intensity of the graft is characteristic of an unimpinged reconstruction.

S.M. HOWELL378

Arthroscopy, Vol. 8, No. 3, 1992

The MR signal of the graft returned to the original low signal intensity within 10 weeks after eliminating the roof impingement. The MR signal of impinged grafts has been shown to remain elevated for 1 to 3 years after implantation (2,13). Therefore, the roofplasty was responsible for returning the MR appearance of the graft to the low signal intensity characteristic of an unimpinged graft.

REFERENCES 1. Fullerton LR, Andrews JR. Mechanical block to extension following

augmentation of the anterior cruciate ligament: A case report. Am J Sports Med 1984;12:166-8.

2. Howell SM. Berns GS, Farley TE. Signal intensity measurements of unimpinged and impinged anterior cruciate ligament grafts. Radiology 1991;179:639-43.

3. Jackson DW, Shaefer RK. Cyclops syndrome: loss of extension follow-ing intra-articular anterior cruciate ligament reconstruction. Arthros-copy 1990;6:171-8.

4. Lukianov AV, Richmond JC, Barrett GR, et al. A multi-center study on the results of anterior cruciate ligament reconstruction using a dacron ligament prosthesis in salvage cases. Am J Sports Med 1989;17:380-6.

5. Anderson AF, Lipscomb BA, Addlestone RB. Analysis of the intercondylar notch by computed tomography. Am J Sports Med 1987;15:547-52.

6. Gillquist J, Odensten M. Arthroscopic reconstruction of the anterior cruciate ligament. Arthroscopy 1988~45-9.

7. Odensten M, Gillquist J. Functional anatomy of the anterior cruciate ligament and a rationale for reconstruction. J Bone Joint Surg 1985: 67A:257-62.

8. Howell SM, Clark JA, Farley TE. A rationale for predicting anterior cruciate graft impingement by the intercondylar roof an MRI study. Am J Sports Med 1991: 19:276-81.

9. Sidles J, Larson R, Garbini J, et al. Ligament length relationships in the moving knee. J Orthop Res 1988; 6:593-610.

10. Tanzer M, Lenczer E. The relationship of intercondylar notch size and content to notchplasty requirement in anterior cruciate ligament sur-gery. Arthroscopy 1990;4:89-93.

11. Sachs R, Daniel D, Stone M, et al. Patellofemoral problems after anterior cruciate ligament reconstruction. Am J Sports Med 1989;17:76&5.

12. Lysholm J, Gillquist J. Evaluation of knee ligament surgery results with special emphasis on use of a scoring scale. Am J Sports Med 1982;10:150-4.

13. Howell SM, Clark JA, Blasier RD. Serial magnetic resonance imaging of hamstring ACL autografts during the first year of implantation: a preliminary study. Am J Sports Med 1991; 19:42-7.

ROOFPLASTY FOR EXTENSION DEFICIT 379

Arthroscopy, Vol. 8, No. 3, 1992