Causes of graft failure can be divided into three categories:  technical, biologic, and external (5).  Technical causes include nonanatomic tunnel placement, hardware failure, bone plug fracture, inadequate notchplasty, improper tensioning, graft fixation, and insufficient graft material.  Plain x-rays can be used to evaluate fracture of the bony portions of the grafts, metallic screw integrity and position, union of the bony portions of the graft, tunnel placement, and size of screw tunnels.  MR can also be used to evaluate tunnel positioning, graft integrity, the other soft tissues of the knee.  The femoral tunnel should be at the intersection of Blumensaat's line (intercondylar roof line) and a line extended from the posterior femoral cortex.  Tibial tunnel placement should be posterior and parallel to the tibial intersection of Blumensaat's line.  One of the more common problems with the tunnels is placement that is too anterior (15).  Tunnels may enlarge over time but this does not seem to affect outcome (1).  Screw placement does not necessarily correlate with the tunnel lines.  Osseous incorporation of the patellar tendon grafts occurs by one year in most cases (1, 11).  On MR, an intact unimpinged graft should have homogenous low signal.  It should be posterior to and not in contact with the intercondylar roof (14).  Grafts impinging on the roof or sidewall will have irregular increased signal (6).  Due to metal artifact on T2 images, T1 may be more useful for assessing technical failure (2).

Biologic causes for failure include:  failed ligamentization (for those grafts which were not originally ligaments), infection, arthrofibrosis, and infrapatellar contracture syndrome.  Arthrofibrosis is a loss of extension of more than 10 degrees or flexion of greater than or equal to 15 degrees.  It could be caused by a cyclops lesion.  This is a fibrous nodule anterolateral to the tibial tunnel that impinges on the intercondylar notch when the knee is extended (Figure 3).  It causes an abrupt endpoint of extension with a popping sensation.  The cyclops lesion could result from drilling debris, ACL stump remnants, broken graft fibers, or repeated impingment of the graft on the notch (4).  Adequate notchplasty (anterolateral debidement of the tibial tunnel) is required to reduce the rate of cyclops formation, which can occur in as many as 5-10% of patients (12).  A cyclops lesion is seen on MR as a low-signal intensity nodule outlined by fluid.

External causes for failure include:  traumatic rupture, secondary instability of the knee, and improper rehabilitation (5).  Graft rupture, similar to native ACL rupture, is best demonstrated on MRI as has been described (Figure 4).  Caution is advised when using secondary signs of ACL rupture since some of these are normal in a post-operative state, such as an effusion (1).

Figure 4.  Sagittal MR of knee demonstrating failed graft.  The fibers are disrupted and there is high signal intensity in the graft substance.

Osteoarthritis (OA) is a common complication of ACL reconstruction and can be imaged by plain x-rays.  Changes usually can be seen by 7 years post-operative (8-9, 12).  A recent study showed that patellofemoral OA is more common than tibiofemoral OA (47% vs. 14-18%) an average of 7 years post-operative (7).  In this study, the patellofemoral OA corresponded in severity to degree of shortening of the patellar tendon.  The OA is more advanced in those patients for which the ACL injury was chronic (> 12 weeks) prior to surgery and when meniscectomy was required (8-9, 12).