Is Peroneal Nerve Injury Associated With Worse Function After Knee Dislocation?

Abstract

Background

Peroneal nerve palsy is a frequent and potentially disabling complication of multiligament knee dislocation, but little information exists on the degree to which patients recover motor or sensory function after this injury, and whether having this nerve injury–with or without complete recovery–is a predictor of inferior patient-reported outcome scores.

Questions/purposes

The purposes of this study were to (1) report on motor and sensory recovery as well as patient-reported outcomes scores of patients with peroneal nerve injury from multiligament knee dislocation; (2) compare those endpoints between patients who had partial versus complete nerve injuries; and (3) compare patient-reported outcomes among patients who sustained peroneal nerve injuries after knee dislocation with a matched cohort of multiligament knee injuries without nerve injury.

Methods

Thirty-two patients were identified, but five did not have 2-year followup and are excluded (16% lost to followup). Twenty-seven patients (24 male, three female) with peroneal nerve injury underwent multiligament knee reconstruction and were followed for 6.3 years (range, 2-18 years). Motor grades were assessed by examination and outcomes by International Knee Documentation Committee (IKDC) and Lysholm scores. Retrospectively, patients were divided into complete (n = 9) and partial nerve palsy (n = 18). Treatment for complete nerve palsy included an ankle-foot orthosis for all patients, nonoperative (one), neurolysis (two), tendon transfer (three), nerve transfer (one), and combined nerve/tendon transfer (one). Treatment for partial nerve palsy included nonoperative (12), neurolysis (four), nerve transfer (one), and combined nerve/tendon transfer (one). Furthermore, patients without nerve injury were matched by Schenck classification, age, and sex. Data were analyzed using univariate and multivariate models.

Results

Overall, 18 patients (69%) regained antigravity ankle dorsiflexion after treatment (three complete nerve palsy [38%] versus 15 partial nerve palsy [83%]; p = 0.06). One patient with complete nerve palsy (13%) and 13 patients with partial nerve palsy (72%) regained antigravity extensor hallucis longus strength (p = 0.01). IKDC and Lysholm scores were similar between complete nerve palsy and partial nerve palsy groups. After controlling for confounding variables such as patient age, body mass index, injury interval to surgery, mechanism of injury, bicruciate injury, and popliteal artery injury status, there was no difference between patients with peroneal nerve injury and those without on Lysholm or IKDC scores.

Conclusions

With multiligament knee dislocation and associated peroneal nerve injury, patients with partial nerve injury are more likely to regain antigravity strength when compared with those with a complete nerve injury, but their overall function may not improve. After controlling for confounding variables in a multivariate model, there was no difference in Lysholm or IKDC scores between patients with peroneal nerve injury and those without.

Level of Evidence

Level III, therapeutic study. See Guidelines for Authors for a complete description of levels of evidence.

Introduction

Traumatic knee dislocations can be devastating because of injury to the ligaments, articular cartilage, menisci, and the neurovascular structures. Although multiligament knee injury is considered rare, earlier reports likely underestimate the true incidence because many knee dislocations spontaneously reduce before presentation [7]. Peroneal nerve palsy is a frequent complication of knee dislocation with a reported frequency ranging from 14% to 40% [14]. The peroneal nerve is at risk because it is held tightly against the fibular head [11] and injury often is associated with proximal fibular fractures and posterolateral corner injuries [5, 17]. Recovery of nerve injury is variable, reported as ranging from 31% to 75% [2, 11, 23, 24, 27]. However, morbidity associated with persistent nerve dysfunction and foot drop is considerable [3].

One previous study demonstrated that increased body mass index and male sex are associated with increased risk of peroneal nerve injury with multiligament knee injury, and young age is a favorable prognostic factor for recovery [23]. However, the authors in that report did not categorize nerve palsies into complete versus incomplete injuries, which limits the ability to assess injury severity and subsequent nerve-related outcomes. Several recent studies have reported on the subjective and objective outcome scores of patients who sustain multiligamentous knee injury [11, 12, 17], but there remains a paucity of literature regarding how concomitant peroneal nerve palsy, and its potential recovery, affects outcome [20]. To our knowledge, no previous studies have compared a matched cohort of patients with and without peroneal nerve palsy in the setting of multiligament knee injury. Improved understanding would allow the treating surgeon to better counsel patients about potential treatments and the prognosis for recovery.

The purposes of this study were to (1) report on motor and sensory recovery as well as patient-reported outcomes scores of patients with peroneal nerve injury from multiligament knee dislocation; (2) compare those endpoints between patients who had partial versus complete nerve injuries; and (3) compare patient-reported outcomes among patients who sustained peroneal nerve injuries after knee dislocation with a matched cohort of multiligament knee injuries without nerve injury.

Patients and Methods

In 2007, our institution implemented a prospective registry dedicated to the tracking of patient outcomes after multiligament knee reconstruction. Patients before 2007 were retrospectively added to the registry. The institutional board approved the registry, and all patients provided informed consent before participation. Patients included in the registry were evaluated preoperatively and prospectively followed at 1, 2, 3, 5, 10, 15, and 20 years postoperatively. We retrospectively reviewed 160 patients who underwent surgical reconstruction of a multiligament knee injury from 1993 to 2010. Thirty-two (20%) of these patients sustained a peroneal nerve palsy in association with their initial knee injury. Five of these patients did not have 2-year followup and were excluded (16% lost to followup). The remaining 27 patients formed the peroneal nerve palsy cohort.

Patients in the peroneal nerve palsy cohort were subdivided into two groups based on their initial physical examination by the attending surgeon: complete peroneal nerve palsy and partial peroneal nerve palsy. Motor examination grading was standardized as previously described [4]. Complete nerve palsy was defined as 0/5 tibialis anterior (TA) strength and 0/5 extensor hallucis longus (EHL) strength and complete sensation loss as previously described [23]. Partial nerve palsy was defined as a deficit in at least one of the following measures: TA strength or EHL strength or peroneal nerve sensation (or any combination). Nine patients were in the complete nerve palsy group and 18 in the partial nerve palsy group. There were no significant differences in patient demographics or injury characteristics (Table 1).

Table 1.

Patient demographics and injury characteristics

Patients	Total (N = 27)	CNP (n = 9)	PNP (n = 18)	p value
Age (years; range)	30 (15-52)	26 (16-39)	33 (15-52)	0.6
Sex (male:female)	24:3	8:1	17:1	0.8
Mechanism
High velocity	9 (33.3%)	1 (11.1%)	8 (44.4%)	0.2
Low velocity	18 (66.7%)	8 (88.9%)	10 (55.6%)
Interval to surgery (days); mean (range)	111 (1-451)	99 (1-451)	117 (1-393)	0.7
Interval to surgery
Acute	10 (37%)	4 (44.4%)	6 (33.3%)	0.9
Delayed (> 3 weeks)	17 (63%)	5 (55.6%)	12 (66.7%)
KD class
KD-I	4 (14.8%)	2 (22.2%)	2 (11.1%)	0.1
KD-II	0	0	0
KD-III(L)	12 (44.4%)	6 (66.7%)	6 (33.3%)
KD-IV	7 (25.9%)	0	7 (38.9%)
KD-V	4 (14.8%)	1 (11.1%)	3 (16.7%)
Vascular injury
Yes	8 (29.6%)	2 (22.2%)	6 (33.3%)	0.9
No	19 (70.4%)	7 (77.8%)	12 (66.7%)

CNP = complete nerve palsy; PNP = partial nerve palsy; KD = knee dislocation classification.

All patients underwent surgical intervention for their multiligamentous knee injuries. Ligamentous reconstruction and/or repair was performed at the discretion of the operating surgeons (MJS, BAL). Soft tissue allografts and/or autografts were used when necessary to reconstruct combinations of anterior cruciate ligament (ACL), posterior cruciate ligament (PCL), fibular collateral ligament (FCL), medial collateral ligament (MCL), and posterolateral corner (PLC) injury. Standardized graft and tunnel preparation as well as graft fixation techniques were used based on surgeon preference with evolving surgical techniques during the study period. Injury patterns resulted in ACL reconstruction in 23 patients (85%), PCL reconstruction in 20 patients (74%), and bicruciate reconstruction in 18 (67%) of patients. The vast majority of patients (n = 26 [96%]) had lateral-sided injury and required surgical treatment of FCL or PLC injury. The majority of patients (58%) had injury to four of five ligamentous knee structures (ACL, PCL, MCL, FCL, and/or PLC). All five of these structures were injured in 12% of patients. The MCL was usually intact and only 17% of patients required MCL repair or reconstruction.

The 27 patients underwent a variety of surgical procedures including: ACL, PCL, FCL, and PLC reconstructions (n = 9); ACL reconstruction with FCL and PLC repairs (n = 2); ACL, PCL, and MCL reconstructions (n = 2); ACL and PCL reconstructions with MCL repair (n = 1); ACL, PCL, MCL, FCL, and PLC reconstructions (n = 1); ACL, PCL, FCL, and PLC reconstructions with MCL repair (n = 1); ACL, FCL, and PLC reconstructions (n = 1); ACL and FCL reconstructions with a PLC repair (n = 1); ACL and PCL reconstruction (n = 1); PCL, FCL, and PLC reconstructions (n = 1); PCL reconstruction with FCL and PLC repairs (n = 1); PCL and PLC reconstructions (n = 1); MCL, FCL, and PLC repairs (n = 1); FCL and PLC reconstructions (n = 1); FCL and PLC repairs (n = 1); FCL repair (n = 1); and ACL reconstruction (n = 1).

Postoperative rehabilitation was tailored to the pattern of ligament repair and/or reconstruction. After surgery, no deterioration of peroneal nerve function occurred in any cases.

The treatment algorithm and decision-making for intervention of an injured peroneal nerve have been previously published [16]. After careful diagnosis, conservative management is appropriate in the early phase of treatment, but the risk versus benefit of surgical exploration at the index operation is controversial. Because late surgical treatment has relatively poor results, however, surgery for persistent nerve injury including neurolysis, primary nerve repair, nerve grafting, nerve transfer, and posterior tibialis tendon transfer are all viable options. Peroneal nerve injuries were treated surgically in seven of the complete injuries (78%) and six of the incomplete injuries (33%; Table 2).

Table 2.

Treatment of nerve injury

Treatment	CNP (n = 9)	PNP (n = 18)
AFO	All patients	12
Nonoperative	1	12
Neurolysis	2	4
Tendon transfer	3	0
Nerve transfer	1	1
Combined tendon and nerve transfer	1	1
Additional surgery	1 below-knee amputation	1 knee arthroplasty and 1 knee arthrodesis

CNP = complete nerve palsy; PNP = partial nerve palsy; AFO = ankle-foot orthosis.

Peroneal nerve motor grade was assessed at last clinical followup in 26 patients (one patient treated with below-knee amputation was excluded) and outcome scores were obtained using the International Knee Documentation Committee (IKDC) [13] and Lysholm forms [28] in all patients at a minimum clinical followup of 2 years (mean, 6 years; range, 2–18 years). Antigravity strength was defined as 3/5 or greater. A comparison was performed between patients with and without antigravity strength to determine if there was a difference in patient-reported outcome scores (IKDC and Lysholm).

A second group of patients with multiligament knee injury without associated peroneal nerve injury was identified from the 160-patient database, forming the matched group without nerve palsy. The peroneal nerve palsy and control groups were matched based on modified Schenck classification [26], age, and sex. The patients in both groups are similar, except there were more vascular injuries in the peroneal nerve palsy group (Table 3).

Table 3.

Matched cohort demographics and injury characteristics

Patients	PN injury (n = 20)	No injury (n = 20)	p value
Age (years; mean [range])	31 (15-52)	30 (19-51)	0.8
Sex (male:female)	18:2	18:2	1.0
Mechanism
High velocity	6 (30%)	11 (55%)	0.1
Low velocity	14 (70%)	7 (35%)
Unknown		2 (10%)
Interval to surgery (days); mean (range)	109 (1-451)	107 (9-493)	0.9
Interval to surgery
Acute	8 (40%)	5 (25%)	0.5
Delayed (> 3 weeks)	12 (60%)	15 (75%)
KD class
KD-I	2 (10%)	2 (10%)	1.0
KD-II	0	0
KD-III(L)	10 (50%)	7 (35%)
KD-IV	5 (25%)	8 (40%)
KD-V	3 (15%)	3 (15%)
Vascular injury
Yes	7 (35%)	1 (5%)	0.05
No	13 (65%)	19 (95%)

PN = partial nerve; KD = knee dislocation classification.

Statistical Analysis

Descriptive analyses of baseline patient characteristics were performed with use of means and SDs for continuous variables and frequencies and percentages for discrete variables. Comparisons of patient characteristics between groups were conducted with use of independent-sample t-tests for continuous variables and chi-square or Fisher’s exact tests for categorical variables. Final patient outcome scores (IKDC, Lysholm) in each group were assessed with use of paired t-tests. Two-tailed tests were used for all statistical analyses with p value < 0.05 considered significant. Data were analyzed using univariate and multivariate models to assess for predictive factors between the groups. All analyses were performed with use of SPSS software (Version 18.0; SPSS, Chicago, IL, USA).

Results

A total of 18 (69%) patients with peroneal nerve injuries in the overall cohort recovered antigravity or better TA strength. Of those 18, 14 patients recovered antigravity or greater EHL strength at latest followup. Eleven patients had intact sensation in the peroneal nerve distribution (all partial nerve palsy group), 12 had partial sensation (five complete, seven partial injury group), and three had no sensation (all complete injury group). At latest followup, nine patients (35%) were using an ankle-foot orthosis (AFO). Ten of 13 patients treated nonoperatively (77%) and eight of 13 patients treated operatively (62%) for persistent peroneal nerve palsy had antigravity or greater TA muscle strength. Full return of peroneal sensation was noted in eight of 13 patients with nonoperative nerve treatment (62%) and three of 13 patients who underwent operative treatment of persistent nerve palsy (23%). Analysis of TA and EHL motor strength at last clinical followup for each surgical intervention independent of any others performed (isolated tendon transfer, isolated nerve transfer, combined tendon and nerve transfer, or any surgical intervention) for patients with complete nerve palsy and partial nerve palsy revealed no significant difference in regaining antigravity TA or EHL motor strength.

There was no difference in IKDC (61.6 ± 17.7 versus 71.6 ± 23.5; p = 0.29) or Lysholm (63.0 ± 23.6 versus 77.0 ± 20.5; p = 0.34) scores in patients who did or did not regain antigravity (≥ 3/5) TA motor strength.

Three patients with complete nerve palsy (38%) and 15 patients with partial nerve palsy (83%) regained antigravity ankle dorsiflexion strength (p = 0.06). In comparison, one patient with complete nerve palsy (13%) and 13 patients with partial nerve palsy (72%) regained antigravity EHL strength (p = 0.01). IKDC scores were 59.4 ± 27.2 versus 66.7 ± 16.1 (p = 0.64) and Lysholm scores were 71.2 ± 20.1 versus 65.4 ± 24.8 (p = 0.71) with no difference between the complete nerve palsy and partial nerve palsy groups, respectively, with the numbers available.

After controlling for confounding variables such as patient age, body mass index, injury interval to surgery, mechanism of injury, bicruciate injury, and popliteal artery injury status, with the numbers available, there was no difference between patients with peroneal nerve injury and those without peroneal nerve injury on Lysholm (p = 0.72, 0.64, 0.87, 0.33, 0.06, 0.09, respectively) or IKDC scores (p = 0.65, 0.42, 0.24, 0.33, 0.09, 0.06). Of note, the average impact of a popliteal artery injury on Lysholm and IKDC scores was 16.6 (p = 0.06) and 14.1 (p = 0.09) points lower, respectively.

Discussion

Multiligament knee injuries can be devastating and concomitant peroneal nerve injury may cause chronic pain and persistent motor weakness. Existing knee dislocation studies focus on surgical treatment of the ligamentous injuries [1, 8, 17–19, 30], but the impact of associated injuries such as those to the popliteal artery or peroneal nerve can be just as vital. In this study, we evaluated motor and sensory recovery as well as patient-reported outcomes scores in patients with peroneal nerve injury after knee dislocation, compared recovery between partial and complete palsies, and established the impact of peroneal nerve injury by contrasting results with a cohort of patients without nerve injury.

There are several weaknesses of our study. Although this study represents the largest reported cohort of patients with surgically treated multiple ligamentous knee injury and associated peroneal nerve palsy of which we are aware, this remains a rare injury and relatively small patient numbers make statistical evaluations difficult. Second, the spectrum of partial versus complete nerve injury as well the variety of surgical treatment methods makes it challenging to demonstrate superiority of one treatment approach. Analysis was performed to determine if one surgical treatment of nerve injury had a positive effect on return of antigravity strength or outcome score compared with other treatments, but no differences were found. Third, there are many variables in each patient’s clinical history that can affect the clinical outcome scores. Therefore, a multivariate analysis was performed to control for some of these factors such as surgical timing, popliteal artery injury, and mechanism of injury. However, activity level was not used to match patients, so patients with demanding preoperative function may do worse because they are not able to get back to full preoperative activities compared with more sedentary patients who may be less disabled with a lower postoperative activity level. Fourth, the outcome measures used are commonly used in the literature, but they have not been validated to evaluate health-related quality of life in multiligament knee injury, and clearly there remains a need for a validated measure by which to assess these complicated injuries. Lastly, the retrospective nature of the study is a limitation and introduces potential biases related to treatments chosen. For example, surgical technique and timing were not randomized but were rather determined by injury type and surgeon preference. In addition, more severe peroneal nerve injuries were selected for intervention such as nerve or tendon transfer, and surgical techniques and treatment approaches evolved during the inclusion period. Like with most retrospective studies, there was loss of 2-year followup–in this study 16%. Patients lost to followup may have had inferior scores compared with those with complete followup, so these outcome scores may reflect more positively.

In the present study, two-thirds of patients who sustained a peroneal nerve injury at the time of knee dislocation were able to regain antigravity ankle dorsiflexion. This compares favorably to the previously reported nerve recovery rates (31%-75%) [2, 11, 23, 24, 27]. The variability of the existing literature likely results from injury heterogeneity and the definition of “recovery.” For example, Peskun et al. [23] report a 31% recovery rate in their series of 26 peroneal nerve palsies with knee dislocation but defined recovery as full strength and sensation. Similarly, Niall et al. [22] demonstrated complete motor recovery in 21% but partial recovery in only 29%. In the current study, we used antigravity muscle strength as a clinically meaningful recovery, and this likely resulted in higher percentages of patients defined as having recovered. In addition, patients in our series were treated with a variety of interventions, including neurolysis and nerve and tendon transfers, which have been previously demonstrated to improve recovery and function [9, 16, 21]. Despite surgical treatment, the peroneal nerve has a poor intrinsic ability for recovery, perhaps as a result of its variable and diminished intraneural blood supply [15]. Previous studies have demonstrated that most peroneal nerve injuries occur in the setting of posterolateral corner injuries as a result of poor tolerance of the nerve to accommodate stretch with changes in limb position during trauma [22, 25]. This observation is confirmed in our study because 26 of the 27 peroneal neurapraxias occurred with posterolateral corner injuries. At 6-year average followup, one-third of our patients continued to wear an AFO. In a series of 27 patients with isolated peroneal nerve injury, de Bruijn et al. [6] report 11% AFO use at 5-year followup. However, patients in their series did not have concomitant knee dislocation and perhaps were better able to tolerate slight gait abnormalities with an otherwise normal knee.

In agreement with previous studies, we found that partial nerve injuries had greater recovery potential than complete injuries. Engebretsen and colleagues [7] reported on motor recovery 2 years after injury in 85 patients who underwent treatment for knee dislocation with 18 patients sustaining nerve palsy. None of the patients with complete nerve palsy recovered sensation or useful motor function, which was in contrast to those with partial palsy who regained some function. Goitz and Tomaino [10] demonstrated an improved prognosis for patients with incomplete nerve palsy after nonoperative management. Peskun et al. [23] did not stratify partial versus complete peroneal nerve injuries in their study, but identified young age (younger than 30 years) as a favorable prognostic factor for recovery. In general, axonal regeneration and recovery in peripheral nerve injuries tend to decrease with age, and worse recovery has been shown in older patients [29]. Sex has been previously demonstrated to be associated with peroneal nerve injury [23], and our study also has a preponderance of males included, but no clear explanations have been elucidated.

In this matched, controlled series of patients with multiligament knee dislocation, we demonstrate similar functional outcome scores in those who sustained concomitant peroneal nerve injury. This may be secondary to the small sample size of the current study. However, it may be a limitation in the outcome measures currently used for assessing patients with multiligamentous knee dislocation. Perhaps IKDC and Lysholm do not accurately reflect peroneal nerve dysfunction in which walking and stairclimbing are limited in two-thirds of patients [6]. These same patients are also restricted in their leisure activities and work duties [6]. A patient with foot drop walks with a vaulting or circumduction gait, which can negatively affect lower extremity function. Clearly there is a need for development of a validated outcome measure in this patient population to reflect the contribution of vascular and nerve injuries to the patient’s overall function. On the other hand, overall quality of life, function, and pain are all adversely affected by posttraumatic peroneal neuropathy [3, 23], but this may not add much further disability to an already devastating injury of knee dislocation. The treatment of knee dislocation with associated peroneal nerve injury is challenging.

Our series demonstrated that patients with a partial nerve injury are more likely to regain antigravity strength than those with complete nerve injury. After matching patients with knee dislocation with peroneal nerve injury with those without injury, there was no difference in Lysholm or IKDC scores when we controlled for confounding variables. This may reflect that current outcome scores inadequately capture the influence of concomitant peroneal nerve dysfunction with multiligament knee injury. At this time, we cannot make definitive recommendations for treatment of peroneal nerve injury with knee dislocation. Future multicenter trials are needed to develop better outcome measures, both to assess the effectiveness of any intervention and to optimize decision-making for this devastating knee injury.