The evaluation of a tibial eminence fracture begins as an evaluation of a patient with knee pain and likely effusion after an injury, most commonly either hyperextension and femur rotation such as attempting to stop abruptly on a bicycle and sporting injuries or a direct blow to the femur with a flexed knee [9]. Radiographic evaluation should proceed first to ensure there are no other skeletal injuries and then distal neurovascular function should be assessed to ensure that the hyperextension did not injure the popliteal artery or nerve. Varus and valgus stress can be applied to ensure the collateral ligaments are intact but with radiographic diagnosis of a tibial spine fracture Lachman and anterior drawer testing should be avoided. In a low energy injury patient it is uncommon for there to be other ligamentous injuries although associated meniscal pathology is common [10–12]. In a higher energy injury such as a pedestrian struck by a motor vehicle multiple ligament injuries may be seen [1].

Plain radiography establishes the diagnosis of tibial eminence fracture and the Meyers and McKeever classification can be established (Fig. 20.3). In a recent study, measurement of fracture displacement was found to be consistent between plain films and CT scan with a difference of 1 mm between modalities [13]. Therefore, the authors concluded that advance imaging may not be necessary to evaluate displacement, but that there could be a need for MRI to establish an association of other injuries, such as meniscus tear or articular extension of the fracture.

Early clinical and MRI studies [14] noted meniscal pathology in adult injuries but considered associated pathology rare in children. More recently MRI studies [10–12] have found rates of 30–40% of meniscal tears as well as meniscal entrapment and bone contusion to be present in association with tibial eminence fractures advocating for MRI evaluation. In these studies there were no associated meniscal tears or entrapment associated with type 1 fractures; therefore it remains logical that more displaced injuries are more likely to have associated pathology.

Kocher used plain radiographs in 2004 [15] to compare notch width index to see if this was predictive of whether the patient would suffer an ACL tear or a tibial eminence fracture. They found that on average the ACL tear group had narrower notch width indices (0.230 vs. 0.253, p = 0.020). However, Samora in 2015 [16] also evaluated notch width index and found no statistical difference between ACL, tibial eminence fracture, and controls, but did find that the intercondylar roof inclination angle was less, indicating a steeper intercondylar roof, in ACL tear patients compared to controls. Tibial eminence fracture patients had a flatter intercondylar roof as evidenced by a larger intercondylar roof angle compared to controls.

Our current practice includes MRI imaging on borderline cases with displacement close to 5 mm on plain film that would not otherwise be treated operatively. If the displacement is more than 5 mm after any attempted reduction, then we proceed with operative reduction in order to minimize risks in outcomes. Associated injuries are then evaluated and treated at the time of surgery without the need for preoperative advanced imaging.

Noted above there is an association with entrapment of the meniscus causing residual displacement or blocking reduction. The typical culprit is the medial meniscus (Fig. 20.4) although inter-meniscal ligament and lateral meniscus have also been reported [17, 18]. While this is debated by some authors [19], others report rates of 26–30% entrapment of meniscus in type 2 fractures that do not reduce in extension and 48–65% of type 3 fractures [10, 17]. Edmonds and colleagues found a rate of 32% meniscal entrapment in their comparative study, and case reports and series have been published highlighting this entity including 9 out of 10 entrapped medial menisci in a series by Mah [20–23]. These numbers validate an operative strategy involving any fracture with residual displacement, whether classified as type 2 or 3.

Once the decision is made to operate the approach must be decided. Residual laxity and arthrofibrosis are common complications of operative intervention for tibial eminence fractures. Early reports involved open reduction with or without fixation. McLennan in 1982 demonstrated that arthroscopic reduction and pin fixation could be an effective treatment for type 3 fractures [24]. There have been 3 comparison studies of open and arthroscopic management of tibial eminence fractures. Melzer et al. found that 14 arthroscopic and 14 open reduction patients resulted in similar isokinetic evaluation of knee joint musculature with a slight decrease in extensor muscle strength [25]. Watts, Larson, and Milbrandt in a recent study found that the main factors contributing to arthrofibrosis were delayed time to surgery >7 days and prolonged operative time >120 min; and although their study was likely underpowered to perform this multivariable analysis, their conclusion that surgeons should approach these fractures in whatever manner they can achieve comfortably and efficiently is likely a good one [26].

The aforementioned study by Edmonds et al. found that the amount of reduction obtained open versus arthroscopic was similar (means: 9.1 mm and 8.6 mm, respectively) and both methods were able to obtain a statistically significant better reduction than closed treatment (mean: 2.3 mm). There was a similar rate of arthrofibrosis between the two surgical methods (ORIF 11.1%, Arthroscopic 12.5%) and there was no arthrofibrosis in the closed management group although the closed management group had a 16.7% risk of need for future operative intervention [13]. This study also highlighted the risk of arthrofibrosis as two patients suffered physeal fractures during manipulation of arthrofibrosis. From these studies it seems that the outcomes achieved are not dependent on open or arthroscopic approach to the fracture. Once again, it appears that the approach can be dependent on surgeon preference between open and arthroscopic techniques.

The overwhelming majority of literature regarding tibial eminence fractures consists of case series focusing on the type of fixation best utilized [27–64]. This includes suture anchors, suture-button suspensory devices, meniscal pin fixation devices, no fixation relying on tucking the fracture fragment under the inter-meniscal ligament to maintain reduction, screws, and series utilizing a combination of these fixation methods. All of these series demonstrate the technique of the authors and demonstrate good results validating the fixation choice.

The biomechanical testing of these fixation types has yielded a variety of results regarding fixation performance. Bong in 2005 compared three strands of #2 Fiberwire sutures vs. a 4.0 mm cannulated screw in 7 matched pairs of fresh frozen human cadaveric knees with a mean specimen age of 76.8 years [65]. Simulated fracture was performed with an osteotome and then fixed so that each cadaver received both fixation methods, one in each knee. These specimens were then loaded at 20 mm/min to failure. The mean ultimate failure load in the Fiberwire group at 319 N was significantly higher than the screw fixation at 125 N. The mean stiffness was not statistically different with Fiberwire 63.0 N/mm and screw 49.9 N/mm. The screws all failed by pullout of cancellous bone representing an issue with using older cadavers to test methods of fixation and translating to use in teenagers with improved bone mass.

In 2007 Eggers et al. evaluated 1 screw, 2 screws, 1 mm Ethibond and #5 (0.8 mm) Fiberwire in a pig model [66]. They found that the Fiberwire had the highest initial stiffness when loaded to failure, the Ethibond and both screw groups were similar indicating that a second screw did not add any stiffness to the construct. After 1000 loading cycles to simulate early postoperative ambulation only 3 of 8 specimens in both screw groups survived while all 16 of the specimens in the two suture groups were intact with 2.7 mm displacement in the Ethibond group and 1.3 mm displacement in the Fiberwire group. The specimens were then loaded to failure after the cyclic testing and found to have similar yield loads. Taking into account the failures during the cyclic loading this group favors suture fixation with Fiberwire based on the ability to withstand trauma during initial rehab as indicated by the initial stiffness as well as maintain reduction during cyclic loading such as during postoperative rehabilitation.

Mahar et al. in 2008 used bovine specimens to compare #2 Ethibond sutures, 3 resorbable nails, 1 resorbable screw, and 1 metallic screw and found no statistically significant difference in initial stiffness or ultimate failure force between cohorts [67]. They did find that after 200 cycles of loading there was 1 mm more displacement in the suture fixation and resorbable screw group when compared to the resorbable nails and metal screw group indicating possible clinical significance of increased laxity.

Aoki in 2011 found that at least 2 strands of #2 Fiberwire would place ultimate failure load above 500 N which they felt was a goal threshold for physiologic ACL stresses during walking [68]. Sawyer in 2012 noted statistically significant higher ultimate load (340 N) with a suture bridge construct using anchors compared to a 3.5 mm cannulated screw and a single #2 Fiberwire suture [69]. The displacement after cyclical loading was not found to be statistically significant. Anderson in 2013 evaluated physeal sparing methods of fixation and found that two strands of #2 Fiberwire tied over a suture button is biomechanically superior to 4 strands of #0 PDS suture button and a single 3.5 mm cannulated screw with higher median yield load, 100% survival after 1000 loading cycles, less median creep after 1000 loading cycles, and a higher median peak failure load after cyclic testing [70]. They also tested two suture anchors loaded with 2 strands each of #2 Fiberwire finding wider inter-quartile indexes indicating less consistent fixation although median ultimate load to failure was similar to the FiberWire suture button. They advocate Fiberwire tied over a suture button despite their initial hypothesis that suture anchors would provide the best fixation. These studies overall highlight that suture fixation provides stronger initial fixation and performs well under cyclical loading that will result from early rehabilitation.

Suture configurations often involve tying the suture over a post in the tibia. There are concerns for physeal tethering if nonabsorbable suture is left tied across the physis. Schneppendahl in 2013 investigated absorbable suture material (#5 Vicryl and #2 PDS) compared to #5 Fiberwire and found that PDS failed to reach a steady state during cyclical testing with 4 of 6 specimens suffering elongation of more than 2 mm [71]. After 200 cycles at up to 150 N they tested the Fiberwire and Vicryl to failure and found a statistically significant performance of the Fiberwire 306 N compared to 220 N for the Vicryl. Their conclusion however is that with the performance during cyclical testing Vicryl could be used as a bioabsorbable alternative to Fiberwire for suture fixation but that PDS should be avoided.

Suture comparison has also been performed clinically: Brunner et al. in 2016 looked at absorbable suture versus nonabsorbable suture tied over a post in a retrospective comparison and found no significant difference in IKDC score, Lysholm scores, and rolimeter testing between the two groups [72] (Fig. 20.5). There were three cases of arthrofibrosis in the absorbable suture fixation group and only one in the nonabsorbable tied over a post; however, all three arthrofibrosis in the absorbable suture group had a delay of at least 4 weeks prior to surgery and preoperative loss of motion. Eight out of 10 patients required removal of the post in the nonabsorbable group highlighting a major benefit of the nonabsorbable suture group. Liao et al. also in 2016 compared nonabsorbable sutures to absorbable suture anchors and found no significant difference comparing Lysholm and IKDC post-op scores. There was no arthrofibrosis reported in either group and there were three total cases of residual laxity with no difference in Lachman testing between the two groups.

The results of the biomechanical studies as well as the clinical tests demonstrate that suture fixation of a tibial eminence fracture yields reliable healing of the fracture with no need for hardware removal. The biomechanics of screw fixation are less clear with some studies finding suture fixation superior and others finding less fragment displacement with screw fixation and advocating fixation method choice being left up to the treating surgeon.

Current narrative within published reports on this fracture type has tended to include more operative fixation than conservative management, to note objective laxity despite subjective stability, and to focus on arthrofibrosis as the least desirable outcome. Yet, the initial case series by Meyers and McKeever as well as Zaricznyj advocated operative treatment for only the most displaced type 3 fractures [6–8]. In 1981, Molander reported on 35 patients noting that only 3 underwent operative intervention and other than healing with a noticeable prominence on radiographs, the nonoperative patients did well [73].

The late 1970s and early 1980s began to change the discourse on tibial eminence fracture outcomes. Initial reports in the Italian trauma literature began to highlight residual instability. Smith in 1984 [74] identified a patient treated 2.5 years previously for tibial eminence fracture who subsequently presented with residual instability, serving as a sentinel event within his practice. Historical commentary suggested that this was not an issue with some authors reporting good outcomes with even ACL excision as part of the treatment: 2 patients in Neer’s series and 3 in Meyers and McKeevers. Smith subsequently reviewed 15 patients of which 12 had type 3 injuries and underwent open reduction and internal fixation. In his series, only seven patients were completely free of symptoms. The other eight reported mild activity related pain and some avoidance of certain activities such as direction changes or skiing. On examination all had normal range of motion, four had atrophy of the thigh measured by thigh circumference compared to the contralateral thigh. One patient had a positive pivot shift, 11 had a positive anterior drawer test and 13 had a positive Lachman test.

In 1988 Baxter and Wiley reported on 45 patients 3–10 years after tibial eminence fractures. In their cohort of 32 total patients, they demonstrated a mean difference of 3.5 mm on anterior drawer testing whether treated by closed reduction or open reduction. There was no varus or valgus instability except in patients who had been pedestrians struck by an automobile. There was loss of full extension in 27 patients and 23 patients had a positive anterior drawer test; none of the patients complained of knee instability although 29 patients were aware of a difference between knees [75]. Willis, in 1993, performed clinical exam, KT arthrometer measurements, and subjective evaluation at a mean follow-up of 4 years and found 64% clinical instability, 74% objective instability, 10% complaints of pain, but no complaints of instability in 50 patients. Moreover, they found no difference in rates of instability whether treated closed or open [76].

In a report from 1995, McLennan performed second look arthroscopy on 10 patients with type 3 tibial eminence fractures. He found a correlation with the best function (based on Lysholm ratings, Tegner levels, and IKDC grade) and improvement in displacement with treatment by open reduction and fixation. The cohort that had undergone open reduction alone did not fair as well, but those treated by closed methods only had the worst overall outcomes in all rating scales. He concluded that open reduction and internal fixation should be performed for type 3 tibial eminence fractures [77]. Also in 1995 Janarv looked at mean 16 year follow-up for 61 patients and found 87% excellent or good knee function based on Lysholm score and no reports of poor subjective knee function despite a rate of 38% laxity among the cohort. They do note in their discussion that only five subjects had a Tegner score greater or equal to seven and only five subjects stated they had a lower level than desired on Tegner. They note that whether conscious or unconscious the patients who suffered tibial eminence fractures performed at a lower activity level making high satisfaction scores easier to achieve for these patients [78].

More recently Kocher [79] examined six children after arthroscopic reduction and fixation with a 3.5 mm cannulated screw and found 5 of 6 patients had an abnormal Lachman examination, 4 of 6 demonstrated differences in KT-1000 testing greater than 3 mm side to side and 2 of 6 patients had an abnormal pivot shift. Despite these findings of laxity the mean Lysholm score was 99.5, mean Tegner score was 8.7, and the mean Marshall score was 49. All of these studies point to excellent clinical results despite instrumented or other objective signs of ACL laxity after fixation.

A more recent study by Mitchell and colleagues used surveys to locate 73 patients from a 20-year span who were treated for tibial eminence fracture and found a 19% rate of subsequent ACL reconstruction [80]. There were no statistically significant associations with fracture classification, sex, or mechanism of injury and need for future ACL reconstruction. They did find increasing age at time of initial injury to be predictive of future ACL reconstruction with odds ratio of 1.3 for each year of age. Interestingly in their study, they had three patients with Meyers-McKeever type 2 fractures who went on to ACL reconstruction due to residual laxity after closed treatment without any further knee injury. All other patients who went on to ACL reconstruction had another traumatic event.

This present literature demonstrates that the overall satisfaction and stability afforded the knee after healing of a tibial eminence fracture is subjectively excellent in those patients treated for tibial eminence fractures. There is residual laxity, whether due to stretch of the ligament prior to failure of the bone, or displacement of the healing fragment it is not known. Across various methods of fixation however residual laxity occurs without notice from the patient.

Residual laxity may be considered a result of the injury, but arthrofibrosis resides at the other end of the spectrum and is more truly a complication of treatment. In a report in 2010 Vander Have and colleagues reported data from four pediatric centers compiling 32 patients with arthrofibrosis after fixation of a tibial eminence fracture [81]. All patients underwent operative stabilization with 28 undergoing arthroscopic reduction and fixation with sutures or screws, three patients underwent open reduction and internal fixation with a screw and one patient underwent closed reduction and pin fixation. Twenty-four patients underwent arthroscopic lysis of adhesions with adhesions identified in all patients. Eight patients underwent manipulation under anesthesia and three of these patients suffered distal femoral physeal fractures and two went on to growth arrest subsequently becoming two of the three patients who never regained full motion. The other 29 patients at 1 year had motion within 5 degrees of the contralateral side.

Patel and colleagues, in 2012, found that mobilization within 4 weeks of definitive treatment limited the risk of developing arthrofibrosis [82]. This included patients treated nonoperatively although none of those patients went on to develop arthrofibrosis. A similar result was found in the comparative study by Edmonds et al. that found arthrofibrosis only in the surgically treated patients at a similar rate whether treated arthroscopic or open [13]. Of note, the period of immobilization was similar across both operative groups and the closed reduction group in this series and all three groups were immobilized for a mean duration of just over 4 weeks. Watts found delayed surgery more than 7 days from injury and prolonged operative time longer than 120 min to be predictive of the risk of arthrofibrosis [26].

Parikh et al., in 2014, detailed aggressive management of postoperative stiffness intervening at the time of screw removal (Fig. 20.6) for both intra-epiphyseal and trans-epiphyseal screws [83]. The epiphyseal screw group was indicated for screw removal based on loss of terminal extension or hardware prominence while all of the transphyseal screws in skeletally immature patients were removed. A third of the epiphyseal and transphyseal screw patients underwent removal of screw and debridement of scar tissue in the notch. At final follow-up there was no arthrofibrosis noted and there were no cases of growth disturbances from temporary transphyseal screws. Kocher recently detailed pediatric patients being treated with dynamic splinting including 21 patients after tibial spine fracture and noted 58% success in avoiding manipulation or lysis of adhesions in the total cohort [84].