M ultiple L i g amen t K n e e

Injuries
Current State and Proposed Classification

Wesley Bradley Dosher, MDa, Garrett T. Maxwell, BSb,

Ryan J. Warth, MDa, Christopher D. Harner, MDa,*

KEYWORDS
 Multiple ligament knee  Knee dislocation  Complex knee  Classification

KEY POINTS
 An effective classification system will enhance communication between providers, facili-
tate accurate and consistent reporting in the literature, and guide management protocols
to improve patient outcomes.
 The current classification systems for multiligamentous knee injuries do not meet all of
these criteria.
 Traditional classification systems provide little information on timing of injury (acute,
chronic), grade of injury (partial, complete), details on the specific location of the anatomic
structures injured, meniscus and articular cartilage injuries, fracture types (avulsion vs
non-avulsion), and details regarding concomitant injuries (skin, tendons, meniscus, and
cartilage, among others).

INTRODUCTION

The overall prevalence of multiligamentous knee injuries (MLKIs) has seen a steady
rise in the United States over recent years.1 However, although MLKIs can be simply
defined as any injury involving more than one knee ligament, very few MLKIs are clin-
ically, functionally, or prognostically equivalent. Currently, the most frequently used
classification system is the anatomically based Schenck classification.2 Other less-
often used classification systems focus on either the direction of dislocation or the
level of energy involved in producing the injury.1,3 Although traditional classification
systems provide some details regarding a patient’s knee injury, they lack sufficient
detail to guide clinical management and thus have limited prognostic value.

Disclosure Statement: The authors have no disclosures.

a
Department of Orthopaedic Surgery, University of Texas Health Science Center at Houston,
6400 Fannin Street, Suite 1700, Houston, TX 77030, USA; b University of Texas Health Science
Center at Houston, McGovern Medical School, 6400 Fannin Street, Suite 1700, Houston, TX
77030, USA
* Corresponding author.
E-mail address: Christopher.Harner@uth.tmc.edu

Clin Sports Med 38 (2019) 183–192

https://doi.org/10.1016/j.csm.2018.11.006 sportsmed.theclinics.com
0278-5919/19/ª 2018 Elsevier Inc. All rights reserved.
184 Dosher et al

The purpose of this article is to give an overview of some existing musculoskeletal

classification systems in use, define what we believe are important additional compo-
nents that should be considered, and provide suggestions for what a “new” classifica-
tion system for MLKIs may look like.

WHAT MAKES AN EFFECTIVE CLASSIFICATION SYSTEM?

Before we introduce a new classification system, we must first define what compo-
nents go into an effective classification system. An effective classification system
will enhance communication between providers, facilitate accurate and consistent
reporting in the literature, and guide management protocols to improve patient out-
comes. We have chosen 2 clinically accepted classification systems currently in use
as examples.
The Burkhart Classification of Rotator Cuff Tears
The creators of this classification system believe a valuable classification system al-
lows for communication between clinicians and researchers, provides information
on treatment and prognosis, and allows for comparison of epidemiologic data and
treatment outcomes.4 The investigators realize the importance of using modern imag-
ing techniques in describing the geometric classification of rotator cuff tears. The sys-
tem not only describes the various tear patterns, but also suggests treatment options
and prognosis for each tear subtype.
The Vancouver Classification of Periprosthetic Femur Fractures
This classification system is based off of reproducible evaluation of plain radiographs
that were then validated with intraoperative findings.5,6 The important feature seen
with this scheme is that the classification of fractures was directly used to determine
operative intervention, which we believe to be an important aspect of any classifica-
tion system. The Vancouver system is also easily communicated between clinicians
and in the literature.
The list of classification systems in the orthopedic literature is extensive, but for the
purposes of this article, the 2 previously discussed classification systems will serve as
examples of effective, clinically meaningful classification systems.

CURRENT CLASSIFICATION SYSTEMS

Energy and Velocity
A precise classification of what constitutes high-energy versus low-energy injury has
not been directly described in the knee dislocation literature. In general, the literature
considers victims of motor vehicle collisions, motorcycle collisions, auto versus
pedestrian injuries, and other “high-speed” injuries as “high energy.” This is in contrast
to the classic use of low energy, which classically has referred to patients who have
sporting injuries or sustain a fall from standing.7
It is worth taking time to briefly mention the new addition of the “ultra–low-velocity
knee dislocation” (ULVKD) described by Azar and colleagues.8 They use the term to
describe dislocations that occur “during activities of daily living, such as stepping
off a curb, stepping off a stair, or simply falling while walking.” With an ever-growing
obese population in the United States, the incidence of ULVKD could see a steady
rise. This is important because accurate diagnosis of these injuries can be difficult
given their mechanism and challenging physical examination. The consequences of
missing these injuries can be catastrophic, given their relatively high vascular and
neurologic injury rates.8
 Multiple Ligament Knee Injuries 185

Directional
The directional classification was described by Kennedy in 1963.3 The classification
system is simple, in that it describes dislocations as anterior, posterior, medial, lateral,
or rotatory. Rotatory dislocation can be subclassified as anterolateral, posterolateral,
anteromedial, and posteromedial. This system is limited because more half of knee dis-
locations spontaneously reduce before assessment in the emergency department.9
The directional classification of dislocations does allow for easy communication be-
tween clinicians but fails to provide a prognosis or guide clinical treatment of patients.
Anatomic
The most commonly used classification system is the anatomically based Schenck
classification (Table 1). The Schenck classification was initially described in 1994,
modified by Wascher and colleagues9 in 1997 to include vascular injuries as well as
specify medial versus lateral injuries, and then described in detail in Robert Schenck’s
2003 article with the conclusion “Classifying knee dislocation is best performed based
on what structures are torn, and use of the anatomic system allows for communication
and surgical planning.”2,9,10 Schenck does make the concession at the end of the
2003 article2 that recommends surgeons take into account the energy of injury,
even though his classification system does not directly address this component.
Looking at the anatomic classification system critically, it does fulfill some of the
necessary components of an effective classification system and certainly represents
a significant improvement of previous classification schemes. The system has in
particular allowed for significantly improved communication between providers over
the past several decades. However, the classification does still lack the necessary
specificity to facilitate accurate and consistent reporting. The most important issue
with this system is that it does not consistently guide clinical decision making. This
is further illustrated in our case examples at the end of the article.

IMPORTANCE OF STANDARDIZED LITERATURE REPORTING

Performing clinical studies of MLKI with a high level of evidence presents multiple ob-
stacles, including the relative infrequency of the injuries, the wide variety of injury
mechanisms and patterns, the lack of a clinically more detailed classification system,
and the varied treatment options available to surgeons who manage these injuries. To
create a new or modified classification system will require adding additional informa-
tion to help clinicians make informed decisions on treatment. Table 2 is a summary of
a systematic literature search of the PubMed and EMBASE databases (60 studies) that

Table 1
Modified Schenck classification that is currently used in clinical practice

Category Injury Patterns

KD-I ACL or PCL 1 collateral
KD-II ACL 1 PCL
KD-III-M ACL 1 PCL 1 medial
KD-III-L ACL 1 PCL 1 lateral
KD-IV ACL 1 PCL 1 medial 1 lateral
KD-V Any MLKI with periarticular fracture

Specifiers: M, medial; L, lateral; N, nerve; C, artery.

Abbreviations: ACL, anterior cruciate ligament; MLKI, multiligamentous knee injury; PCL, poste-
rior cruciate ligament.
 186
Dosher et al
Table 2
Reporting frequency of Critical Factors (CFs) reported in the literature relevant to the management of multiligamentous knee injuries across the most
commonly declared subtopics

Most Commonly Reported Subtopics of Interest

±External ±Collateral Repair vs Graft
Variables Reported ±Operative Fixation ±Delayed ±Staging Surgery Reconstruction Selection
Number of studiesa 5 7 9 9 7 12 2
Age 5 2 7 9 6 11 2
BMI 0 0 1 1 1 1 0
Anticipated noncompliance 0 1 0 0 1 1 0
Neurovascular status 5 5 6 6 3 7 0
Degree of polytrauma 1 2 3 2 1 2 0
Reduction statusb 0 0 2 1 0 0 1
Injury chronicityc 5 7 9 9 7 12 2
Injury graded 2 3 7 6 6 7 2
Injury locatione 0 0 0 0 0 1 0
Injury typef 2 0 2 2 1 4 0
Status of collaterals 4 7 9 8 5 11 2
Presence of fractures 4 4 6 7 2 6 2
Nature of prior knee surgeries 0 0 0 1 0 0 0
Nature of concomitant injuries 3 2 4 5 3 7 1

Note: Shaded boxes represent CFs specific to the corresponding sub-topics of interest.
a
N 5 60 studies addressed a total of 70 subtopics.
b
Reported whether the knee had spontaneously reduced on presentation, or whether the knee was reduced on initial radiographs.
c
Reported breakdown of acute and chronic injuries.
d
Reported breakdown of partial and complete ligamentous injuries.
e
Reported location of injury along the length of each ligamentous structure (proximal, mid-substance, distal).
f
Reported breakdown of soft tissue injuries versus bony avulsion injuries.
 Multiple Ligament Knee Injuries 187

we performed to determine the reporting frequency of critical factors that would be

necessary for specific clinical decision (eg, operative vs nonoperative, repair vs recon-
struction, early vs delayed surgery). The gray boxes represent what we deem to be the
3 most important patient parameters for each question of interest; clearly, there exists
wide variability in the frequency of the most important variables. The lack of consistent
reporting standards for clinical studies involving MLKIs prevents between-study com-
parisons, even among groups of studies that attempt to answer the same clinical
question. Table 2 provides some important variables that should be included in future
MLKI studies, but can also serve as a guideline from which a new classification system
for MLKIs could be constructed.

WHAT SHOULD A NEW MULTILIGAMENTOUS KNEE INJURY CLASSIFICATION

SYSTEM INCLUDE?

Using retrospective data from 287 patients who presented to our institution with
MLKIs, we evaluated relationships between the current KD (knee dislocation) clas-
sification and subsequent management strategies to identify injury characteristics
that could predict surgical management (Warth RJ, unpublished data, 2018). We
found that the KD classification in isolation was not predictive of the type of surgery
performed or the need for staged procedures. We found that surgical management
strategies became much more predictable after considering the combinations of
structures injured, specifying the grades of ligament injuries, and the specifying
the location of medial-sided injuries (proximal, mid-substance, distal). For example,
posterior cruciate ligament (PCL) injuries were predictably treated surgically when
combined with an anterior cruciate ligament (ACL) injury, partial PCL tears were
much less likely to be treated surgically than complete PCL tears, and PCL sur-
geries were much more likely to be staged when a concomitant lateral-sided injury
was present. Medial-sided injuries were significantly less likely to be treated surgi-
cally, whereas lateral-sided injuries were predictive of surgical treatment; partial
tears involving either the medial or lateral side were more likely repaired primarily,
whereas complete tears were more often reconstructed with a graft. We also found
that, with respect to medial-sided injuries, distal tears were more likely to undergo
suture repair, mid-substance tears were much more likely to be reconstructed with
a graft, and proximal tears were more likely to be treated nonoperatively. Surgical
staging was predicted by the presence of concomitant fractures (after exclusion
of avulsion injuries), nerve injuries, and vascular injuries (Table 3). There were no
significant relationships between injury timing (acute, chronic) and any of the treat-
ment options analyzed.

FUTURE MULTILIGAMENTOUS KNEE INJURY CLASSIFICATION SYSTEMS

Our retrospective data indicated that the KD classification was not predictive of the
treatment provided. Using our data, it becomes apparent that a new classification sys-
tem should include each specific structure injured (ACL, PCL, medial structures, lateral
structures), modifiers for fractures and extensor mechanism injuries, nerve injuries, or
vascular injuries, as well as possibly including the specific anatomic location of struc-
tures injured (proximal, mid-substance, distal), especially for medial-sided injuries.
One tool that could better illustrate knee ligament injuries is the Müller map.11 The
Müller map would allow for a visual representation of specific structures injured and
can be further modified to show specific location and severity of injuries. An example
of this can be seen in Fig. 1. There would also have to be a place for modifiers with
regard to ipsilateral fractures and neurovascular injuries. Although our retrospective
188 Dosher et al

Table 3
Summary of predicted management decisions for specific ligamentous injuries according to logistic
regression
Partial Tear Complete Tear Surgery Staging
Injury Feature Prediction Odds Ratioa Prediction Odds Ratioa Prediction Odds Ratioa
ACL injury Recon 2.8 (2.0–3.8) Recon 2.9 (2.1–4.0) NS —
PCL injury Recon or 1.4 (1.4–1.5) Recon 1.9 (1.7–2.1) Staged when 1.3 (1.1–1.5)
Repair combined
with
lateral-side
injury
Medial-side Repair 1.3 (1.1–1.5) Recon or 1.4 (1.4–1.5) NS —
injury Repair
Proximal Repair 5.2 (1.7–15.7) NS — NS —
Mid-Substance NS — Recon 8.9 (1.0–80.2) NS —
Distal Repair 7.1 (2.1–24.5) Repair 24.7 (5.4–114.2) NS —
Lateral-side Repair 1.2 (1.1–1.4) Recon or 1.4 (1.4–1.5) Staged when 1.3 (1.1–1.5)
injury Repair combined
with PCL
injury
Fracture — — — — Staged 1.2 (1.1–1.4)
Nerve Injury — — — — Staged 1.2 (1.1–1.4)
Vascular Injury — — — — Staged 1.4 (1.3–1.6)

Dashes indicate the data field was either not applicable or relevant.
Abbreviations: ACL, anterior cruciate ligament; NS, predictive variables were not statistically significant
(P<.05); PCL, posterior cruciate ligament; Recon, reconstruction.
a
All reported odds ratios are statistically significant according to logistic regression analyses; 95% con-
fidence intervals in parentheses.

data did not show a statistically significant difference in the eventual treatment of
acute versus chronic injuries, we still feel that this should be included in the classi-
fication system, as it represents an important landmark during clinical decision
making.

CASE EXAMPLES

Using Muller maps, we have chosen several case examples to better illustrate the
potential confusion in classifying MLKIs using the current Schenck classification
(Table 4).

Fig. 1. The Müller map. ALL, Anterolateral Ligament; BT, Biceps Femoris Tendon; ITB, Ilioti-
bial Band; LCL, Lateral Collateral Ligament; MCL, Medial Collateral Ligament; POL, Popliteal
Oblique Ligament; PT, Patellar Tendon. (Adapted from Müller W. The knee: form and func-
tion. Springer-Verlag Berlin Heidelberg: Springer; 1982; with permission.)
 Multiple Ligament Knee Injuries 189

Table 4
Common clinical scenarios in which MLKIs are classified into the same KD categories but are
treated differently

Injury Pattern Among

KD Categories that
KD Require Different
Classification Treatment Strategies Corresponding Müller Map
KD-I Acute complete ACL
tear with complete
LCL tear

Chronic complete PCL

tear with a partial
posterolateral corner
tear (intact LCL)

KD-II Chronic complete ACL

tear with partial
PCL tear

Acute complete ACL

tear with complete
PCL tear

(continued on next page)

190 Dosher et al

Table 4
(continued )
Injury Pattern Among
KD Categories that
KD Require Different
Classification Treatment Strategies Corresponding Müller Map
KD-III-M Acute and complete
tear of ACL/PCL/MCL

Acute complete ACL tear,

complete PCL tear,
partial MCL tear

KD-III-L Chronic and complete

tear of ACL/PCL/LCL

Acute complete ACL tear,

complete PCL tear,
popliteus tear
(intact LCL/ITB)

(continued on next page)

 Multiple Ligament Knee Injuries 191

Table 4
(continued )
Injury Pattern Among
KD Categories that
KD Require Different
Classification Treatment Strategies Corresponding Müller Map
KD-IV Chronic complete ACL,
complete PCL,
complete distal MCL,
complete LCL
(intact popliteus)

Acute complete ACL,

partial PCL, complete
MCL, complete LCL
distal avulsion

KD-V Acute partial ACL,

complete PCL
mid-substance,
Schatzker 4 medial
plateau

Acute complete ACL,

complete PCL
mid-substance,
complete LCL
with fibular
head avulsion

Abbreviations: ACL, anterior cruciate ligament; LCL, lateral collateral ligament; MCL, medial collat-
eral ligament; MLKI, multiligamentous knee injury; PCL, posterior cruciate ligament.
Adapted from Müller W. The knee: form and function. Springer-Verlag Berlin Heidelberg:
Springer; 1982; with permission.
192 Dosher et al

SUMMARY

The use of classification systems to enhance communication between providers,

facilitate accurate and consistent reporting in the literature, and guide management
protocols to improve patient outcomes remains critical in modern orthopedics. The
current MLKI classification system is now more than 20 years old and certainly rep-
resents an improvement over previous classification systems; however, it is our hope
that by creating a classification system that includes each specific structure injured
(ACL, PCL, medial structures, lateral structures), modifiers for fractures, extensor
mechanism injuries, nerve injuries, or vascular injuries, as well as specific anatomic
location of structures injured meniscus and articular cartilage injuries, and injury
timing (acute, chronic) that we can use this information to guide further research in
this field and to ultimately improve patient care.

REFERENCES

1. Arom GA, Yeranosian MG, Petrigliano FA, et al. The changing demographics of
knee dislocation: a retrospective database review. Clin Orthop Relat Res 2014;
472:2609–14.
2. Schenck R. Classification of knee dislocations. Oper Tech Sports Med 2003;11:
193–8.
3. Kennedy JC. Complete dislocation of the knee joint. J Bone Joint Surg Am 1963;
45:889–904.
4. Davidson J, Burkhart SS. The geometric classification of rotator cuff tears: a sys-
tem linking tear pattern to treatment and prognosis. Arthroscopy 2010;26:417–24.
5. Duncan CP, Masri BA. Fractures of the femur after hip replacement. Instr Course
Lect 1995;44:293–304.
6. Brady OH, Garbuz DS, Masri BA, et al. The reliability and validity of the Vancou-
ver classification of femoral fractures after hip replacement. J Arthroplasty 2000;
15:59–62.
7. Shelbourne KD, Porter DA, Clingman JA, et al. Low-velocity knee dislocation. Or-
thop Rev 1991;20:995–1004.
8. Azar FM, Brandt JC, Miller RH 3rd, et al. Ultra-low-velocity knee dislocations. Am
J Sports Med 2011;39:2170–4.
9. Wascher DC, Dvirnak PC, DeCoster TA. Knee dislocation: initial assessment and
implications for treatment. J Orthop Trauma 1997;11:525–9.
10. Schenck RC Jr. The dislocated knee. Instr Course Lect 1994;43:127–36.
11. Müller W. The knee: form and function. Springer-Verlag Berlin Heidelberg:
Springer; 1982.