Veterinary Surgery

37:749755, 2008

Comparison of Arthroscopy and Arthrotomy for Diagnosis of

Medial Meniscal Pathology: An Ex Vivo Study
ANTONIO POZZI,

DMV, MS, Diplomate ACVS,

BLAKE EASON HILDRETH, III,

PAIVI J. RAJALA-SCHULTZ, DVM, PhD

DVM,

and

ObjectiveTo evaluate the sensitivity and specicity of arthroscopy and arthrotomy for diagnosis
of medial meniscal pathology and to evaluate the diagnostic value of medial meniscal probing.
Study DesignEx vivo study.
AnimalsCadaveric canine stie joints (n 30).
MethodsStie joints were assigned to either a cranial cruciate ligament (CrCL) decient or intact
group. Within each stie joint, no medial meniscal tear, a peripheral detachment, or 1 of 3 variants
of vertical longitudinal tears of the medial meniscus were created. Each stie joint had arthroscopy,
craniomedial (CrMed), and caudomedial (CdMed) arthrotomy. Diagnoses were made by both
observation and probing. Sensitivity, specicity, and correct classication rate (CCR) for diagnosing
the state of the medial meniscus using both observation and probing with all diagnostic methods
were calculated. Odds ratios were calculated to determine if probing increased diagnostic accuracy.
ResultsArthroscopy with probing was the most sensitive and specic diagnostic method and
had the highest CCR. For arthrotomy, CrMed was the most sensitive in CrCL-decient and CdMed
the most sensitive in stable, CrCL-intact stie joints. For all methods, probing increased their
diagnostic accuracy.
ConclusionsArthroscopy is the most accurate diagnostic method; however, probing the medial
meniscus enhances the diagnostic accuracy of all methods.
Clinical RelevanceAccurate diagnosis of medial meniscal pathology is ideally achieved by means
of arthroscopy; however, if arthrotomy is chosen, CrMed should be selected in unstable and CdMed
in stable stie joints. Regardless, medial meniscal probing should be performed to increase
diagnostic accuracy.
r Copyright 2008 by The American College of Veterinary Surgeons

accurate diagnosis of medial meniscal pathology is warranted to provide appropriate treatment and ultimately
maximize long-term outcome.
Vertical longitudinal tears are the most frequently
occurring medial meniscal tear, accounting for
57%
of medial meniscal tears.3 A vertical longitudinal tear is
referred to as a bucket handle tear if the axial border
is displaced, thereby becoming more apparent during
examination. The degree of displacement of the handle
and the location and complexity of the tear could potentially inuence the sensitivity (Sn) of the diagnostic

INTRODUCTION

EDIAL MENISCAL pathology, a commonly

reported sequela of cranial cruciate ligament (CrCL)
disease, has an incidence of 2077%.14 Medial meniscal
injury has been cited as a late complication of surgical
management of the CrCL-decient stie joint.5,6 Postliminary medial meniscal tears may be a result of residual
stie joint instability and latent tears represent a failure of
diagnosis at surgery and may result in persistent lameness
and require additional surgical treatment.5 Therefore, an

From the Departments of Veterinary Clinical Sciences and Veterinary Preventive Medicine, College of Veterinary Medicine, The Ohio
State University, Columbus, OH.
Address reprint requests to Antonio Pozzi, DMV, Department of Small Animal Clinical Sciences, University of Florida College of
Veterinary Medicine, PO Box 100126, HSC Gainesville, FL 32610-0126. E-mail: pozzia@vetmed.u.edu
Submitted April 2008; Accepted July 2008
r Copyright 2008 by The American College of Veterinary Surgeons
0161-3499/08
doi:10.1111/j.1532-950X.2008.00442.x

749

750

DIAGNOSIS OF MEDIAL MENISCAL PATHOLOGY

method. Additionally, a tear may remain undiagnosed

if the axial border (or handle) is not displaced during
examination and stie joint manipulation. Therefore,
determining the most reliable method for diagnosing
vertical longitudinal tears is of utmost importance.
Advanced imaging techniques (MRI, CT arthrogram,
ultrasonography) have been advocated in the diagnosis of
medial meniscal pathology.79 These minimally invasive
diagnostic methods are often of value, but their use may
be limited because of availability and cost. Arthroscopy
and arthrotomy are also used for diagnosis of
medial meniscal pathology.1,4,10,11 Despite recognition of
arthroscopy as the reference standard for diagnosing
medial meniscal tears,7 stie joint arthrotomy is still
commonly used. During stie joint arthrotomy, diagnosis
of medial meniscal tears are commonly made only by
observation and probing is not routinely performed. It
seems intuitive that (1) vertical longitudinal tears may be
difcult to diagnose without displacement of the axial
border of the medial meniscus and (2) adequate observation by arthrotomy would depend on the method of
arthrotomy, the inherent stability of the stie joint, and
the degree of stie degenerative joint disease.
To our knowledge, the Sn and specicity (Sp) of
arthroscopy and arthrotomy for diagnosing medial
meniscal pathology in dogs have not been reported. Furthermore, probing of the medial meniscus as part of
the diagnostic examination of the medial meniscus has
seemingly not been investigated. We hypothesized that
arthroscopy is more sensitive and specic than arthrotomy for diagnosis of medial meniscal pathology
and probing enhances the Sn and Sp of both methods.
Thus, our purposes were to (1) determine the Sn, Sp, and
correct classication rate (CCR) of arthroscopy, craniomedial (CrMed) arthrotomy, and caudomedial (CdMed)
arthrotomy with both observation alone and with
probing of the medial meniscus for diagnosis of medial
meniscal pathology in CrCL-decient and -intact stie
joints and (2) to determine if probing increased their
overall diagnostic accuracy, assessed by odds ratio (OR)
calculation.
MATERIALS AND METHODS
Specimens
Cadaveric pelvic limbs (n 30 pair) were collected within
12 hours of death from adult dogs euthanatized for reasons
unrelated to this study. Dogs were of various breeds, but were
similar in size and body weight (2030 kg). Palpation of each
limb and evaluation of the stie joint during limb preparation
were performed to ensure the absence of orthopedic disease,
notably that involving the CrCL. After inclusion, body
weights were recorded (kg) and paired pelvic limbs harvested
by coxofemoral disarticulation while preserving all soft tissues.

Specimens were wrapped in saline (0.9% NaCl) solutionsoaked towels and stored at 101C until study, at which time
they were thawed at room temperature.

In Vitro Limb Preparation and Testing

Limbs were randomly distributed into test groups and procedures were performed randomly based on randomization
tables established before starting the study. Each pair of pelvic
limbs was randomly assigned to the following 5 groups based
on the state of the medial meniscus: (1) no medial meniscal
tear, (2) peripheral detachment, (3) partial thickness vertical
longitudinal tear, (4) complete vertical longitudinal tear, and
(5) double vertical longitudinal tears. Within each pair of pelvic
limbs, stie joints were randomly assigned to either a CrCLdecient or -intact group, for a total of 30 limbs in each group.
Thus, there were 12 stie joints with each of the above medial
meniscal states, 6 in the CrCL-decient and 6 in the CrCLintact group. The CrCL-decient group represented a grossly
unstable stie joint with a complete CrCL tear. The CrCLintact group represented a stable stie joint, which is commonly observed clinically when there is a partial CrCL tear.
After initial thawing, limbs were clipped from the
mid-diaphysis of the femur to the middiaphysis of the tibia
and preserved in saline-soaked laparotomy sponges when not
being manipulated. Soft tissues were kept moist during the
study by spraying the tissues with isotonic saline solution. All
unshaven areas were wrapped with bandaging material to
decrease potential bias from recognizing limbs between
diagnostic methods. A 3 cm cranial to caudal skin incision
was performed on the medial aspect of the stie joint. The
incision was extended through the subcutaneous tissue and
fascia. The medial collateral ligament was identied, and a
medial collateral desmotomy and medial capsulotomy was
performed to expose the tibial surface of the medial meniscus.
The corresponding medial meniscal tear (or none for the no
medial meniscal tear group) was created with a No. 11 scalpel
blade (Bard-Parker Rib-Back carbon steel No. 11 surgical
blade, Becton, Dickinson & Co, Franklin Lakes, NJ) in the
caudal pole of the medial meniscus (Fig 1).
In the peripheral detachment group, the coronal ligament
attaching the caudal pole of the medial meniscus to the joint
capsule was incised (Fig 1B). The partial thickness vertical
longitudinal tear was created as a 3-mm-long incision on the
tibial surface of the medial meniscus (Fig 1C). The full thickness vertical longitudinal tear was created similarly; however,
it was full thickness (Fig 1D). The double vertical longitudinal
tear had an axial full thickness vertical longitudinal tear and
an adjacent abaxial partial thickness vertical longitudinal tear
(Fig 1E). Whereas partial thickness vertical longitudinal tears
originating on the tibial surface and double vertical longitudinal tears are rare, inclusion of these tears is important
because they are considered diagnostic challenges for any
surgeon. This is particularly so for double vertical longitudinal
tears, which can easily be misdiagnosed as a single full thickness vertical longitudinal tear.
By means of 2-0 polydioxanone, the medial collateral
desmotomy and medial capsulotomy were repaired with 2

POZZI, HILDRETH, AND RAJALA-SCHULTZ

751

Fig 1. (A) Caudocranial and proximodistal view of a normal canine stie joint. (B) Caudocranial and proximodistal view of a
canine stie joint with a peripheral detachment of the medial meniscus. On the proximodistal view, the dotted line denotes the
original location of the entire caudal pole of the medial meniscus. (C) Caudocranial and proximodistal view of a canine stie joint
with a partial thickness vertical longitudinal tear of the medial meniscus. On the proximodistal view, the dotted line denotes the
location of the tear on the ventral aspect of the medial meniscus, which is seen on the corresponding caudocranial view. (D)
Caudocranial and proximodistal view of a canine stie joint with a full thickness vertical longitudinal tear of the medial meniscus.
(E) Caudocranial and proximodistal view of a canine stie joint with a double vertical longitudinal tear of the medial meniscus. On
the proximodistal view, the dotted line denotes the location of the most cranial tear on the ventral aspect of the medial meniscus,
which is seen to be an incomplete tear on the corresponding caudocranial view.

locking loop sutures and a simple continuous appositional

pattern, respectively. A craniolateral stie joint arthrotomy
was then performed in all stie joints.12 No evidence of gross
CrCL pathology was observed in any stie joint. In those
stie joints assigned to the CrCL-decient group, a midsubstance CrCL transection was performed with a No. 11 blade.
Stie joints were closed routinely with 2-0 polydioxanone in a
simple continuous appositional pattern in the joint capsule
and stainless steel staples in the skin (Apposet ULC 35 W

Stainless Steel Skin Stapler, United States Surgical, Tyco

Healthcare Group LP, Norwalk, CT). All limbs were then
placed in a separate holding area, away from the site of
experimentation.
Each stie joint was then evaluated by 3 separate diagnostic methods (arthroscopy, CrMed, and CdMed) in random
order using randomization table established before starting
the study. During each diagnostic method, 1 investigator
(A.P.) evaluated the medial meniscus initially by observation

752

DIAGNOSIS OF MEDIAL MENISCAL PATHOLOGY

Statistical Analysis
s

Fig 2. Arthrex Vet Systems small joint hook tip probe.

alone and then with probing using a meniscal probe (Fig 2;

s
Small Joint Hook Tip Probe, Arthrex Vet Systems, Bonita
Springs, FL). During each diagnostic method, a diagnosis was
made by 1 investigator (A.P.) initially without probing, and
then with probing, adhering to the guidelines of the 5 states of
the medial meniscus (Fig 1). To maintain blindness throughout experimentation, each stie joint had 1 diagnostic procedure at a time, after which the limb was returned to the
separate holding area by 1 investigator (B.E.H.) and the next
randomly assigned limb acquired for diagnosis. At the end of
the study, all stie joints were dissected to verify that at no
point during experimentation further medial meniscal injury
was induced. If the state of the medial meniscus was different
from that initially induced, the limb was excluded from the
study.

Arthroscopy
Stie joint arthroscopy was performed using a 2.7 mm 301
foreoblique arthroscope with video capture (Model 882TE,
Stryker Corp., Kalamazoo, MI). The camera had a built-in
digital enhancer with 1.5 Lux/800 LOR and a 62 dB signal
to noise ratio. A craniolateral camera portal and a CrMed
instrumental portal were made as previously described.7 The
camera was inserted proximal to the fat pad to avoid
obstruction within the viewing window instead of debriding
the fat pad. If necessary, the pad was retracted distally
by gently pulling the scope cannula in a proximal to distal
direction. To improve observation of the medial meniscus, the
proximal tibia was cranially subluxated in relation to the distal
femoral condyles in the CrCL-decient stie joints using the
tibial compression maneuver. Skin edges were apposed with
stainless steel staples.

CrMed Arthrotomy
A CrMed arthrotomy was performed as previously
described.12 A Hohman retractor was used to provide further
exposure by cranially subluxating the proximal tibia. The stie
joint capsule and fascia were closed in a simple continuous
appositional pattern with 0 and 2-0 polydioxanone, respectively. Skin edges were apposed with stainless steel staples.

CdMed Arthrotomy
A CdMed arthrotomy was performed caudal to the medial
collateral ligament as previously described.13 A small Gelpi
retractor was used for joint capsule retraction. The stie joint
capsule was closed in a simple interrupted cruciate mattress
pattern with 0 polydioxanone. Skin edges were apposed with
stainless steel staples.

Pooled body weights among the 5 medial meniscal groups

were deemed normally distributed by means of a Kolmogorov
Smirnov normality test. Body weights in the 5 experimental
groups were compared by means of a 1-way ANOVA, with
statistical signicance established at Po.05. In each of the
5 medial meniscal groups, diagnoses were classied as true
positive, false positive, false negative, and true negative, and
2  2 contingency tables constructed. When calculating the Sn
and Sp for correctly diagnosing the state of the medial meniscus, a diagnosis was considered negative if any other state
of the medial meniscus was diagnosed other than the exact
state that was present. Both Sn and Sp with respective 95%
condence intervals (CI) were calculated by means of a
Fishers exact test. The use of 95% CIs allows comparison of
the methods: no overlap between CI indicates a statistically
signicant difference at the level of a 0.05.
For each diagnostic method, with respect to observation
alone or by probing in each of the 5 medial meniscal groups,
diagnoses from both CrCL-decient and CrCL-intact stie
joints were pooled together. Sn and Sp calculations were made
for (1) all stie joints and (2) CrCL-decient and CrCL-intact
stie joints separately. Using Sn and Sp, the CCR and
respective 95% CI were calculated. The CCR ([true positives true negatives]/all diagnoses) is a better indicator of the
accuracy of a particular diagnostic method, because it takes
into account both true positives and true negatives, i.e. all
correct diagnoses.
ORs with respective 95% CI were calculated to determine
if probing increased the diagnostic accuracy. OR calculations
were made for (1) all 5 medial meniscal groups pooled for each
method in both CrCL-decient and -intact stie joints and (2)
all 5 medial meniscal groups pooled for each method with
pooling of both CrCL-decient and -intact stie joints to
determine if the odds of reaching a correct diagnosis was
higher if probing was used than observation alone. In addition, an OR calculation was made for the likelihood of
misdiagnosing a medial meniscal tear if a medial meniscal tear
was truly present in the CrCL-decient group using arthrotomy and observation alone when compared with arthrotomy
and probing. All analyses were performed with commercial
software (GraphPad Prism v4.03, GraphPad Software Inc.,
San Diego, CA).

RESULTS
Body weights ranged from 20.5 to 29.6 kg (mean 
SD, 24.6  3.0 kg) and were not signicantly different
among groups (P .525). No limbs were excluded because of altering the state of the medial meniscus.
Sn, Sp, and CCR were calculated separately for each
diagnostic method both with observation alone and
probing for each medial meniscus state (Tables 1 and 2).
Results were similar for all states of the medial meniscus
both in CrCL-decient and -intact stie joints, such that
arthroscopy performed better than arthrotomy and prob-

753

POZZI, HILDRETH, AND RAJALA-SCHULTZ

Table 1. Sensitivity, Specicity, Correct Classication Rate, and Odds Ratio when Observation is Combined with Probing Compared with Observation
Alone for the Overall Diagnosis of 5 Medial Meniscal States when Results for All Stie Joints are Pooled
Cranial Cruciate Ligament Decient and Intact Stie Joints
Surgery
Arthroscopy
Craniomedial (CrMed)
Caudomedial (CdMed)

Diagnostic
Approach

Sensitivity
(Sn) (95% CI)

Observation
Probing
Observation
Probing
Observation
Probing

33
80
22
37
22
42

Specicity
(Sp) (95% CI)

(2247)
(6889),w,z
(1234)
(2550)w
(1234)
(2955)z

83
95
80
84
80
85

Correct Classication
Rate (CCR) (95% CI)

(7888)
(9197),w,z
(7585)
(7986)w
(7585)
(8090)z

73
92
69
75
69
77

Odds Ratio
(OR) (95% CI)

(6878)
(8895),w,z
(6374)
(6979)w
(6374)
(7281)z

8.0 (3.518.3)
2.1 (0.94.7)
2.6 (1.25.8)

Figures in parentheses indicate the 95% CI. The OR indicates the odds of diagnosing the state of the medial meniscus correctly with probing compared
with observation alone, with 41 indicating increased odds of a correct diagnosis.
Indicates a statistically signicant difference between observation and probing for that particular diagnostic method,
wBetween arthroscopy and CrMed using either observation or probing and
zBetween arthroscopy and CdMed using either observation or probing.

ing increased the correct classication of all methods

when compared with observation alone. Overall, probing
increased Sn and Sp of each diagnostic method. Arthroscopy with probing had the highest Sn (80%), Sp (95%),
and CCR (92%) when all stie joints were pooled (Table
1). Sn for CrMed and CdMed arthrotomies with observation alone was 22% (95% CI: 1234%), with a CCR of
69% (95% CI: 6374%) when all stie joints were pooled
(Table 1). Considering arthrotomies with probing,
CrMed was the most sensitive (47%; 95% CI: 2866%)
and had the highest CCR (79%; 95% CI: 7184%) in
CrCL-decient stie joints and CdMed was the most
sensitive (53%; 95% CI: 3472%) and had the highest
CCR (81%; 95% CI: 7487%) in stable, CrCL-intact
stie joints (Table 2).
When all stie joints were pooled, probing increased
the likelihood of accurately diagnosing the state of the
medial meniscus with arthroscopy, CrMed arthrotomy,
and CdMed arthrotomy by 8.0 (95% CI: 3.518.3), 2.1
(95% CI: 0.94.7), and 2.6 (95% CI: 1.25.8) times,
respectively, when compared with observation alone

(Table 1). In addition, we were 21.4 (95% CI: 2.7

169.8) times more likely to misdiagnose a meniscal tear if
an arthrotomy was performed with observation alone
compared with probing when a medial meniscal tear was
truly present.
DISCUSSION
We found that arthroscopic examination of the medial
meniscus using both observation and probing had higher
Sn, Sp, and CCR than arthrotomy for diagnosing medial
meniscal tears. Probing enhanced the Sn, Sp, and CCR
of all diagnostic methods and an accurate diagnosis by
either arthrotomy or arthroscopy was 2.12.6 or 8.0 times
more likely, respectively, when compared with observation alone. Between arthrotomies, CrMed was the most
sensitive and had the highest CCR in CrCL-decient stie
joints, whereas CdMed was the most sensitive and had
the highest CCR in stable, CrCL-intact stie joints.
Our results support that arthroscopy is the best diagnostic method for assessing medial meniscus pathology in

Table 2. Sensitivity (Sn), Specicity (Sp), Correct Classication Rate (CCR), and Odds Ratio (OR) when Observation is Combined with
Probing Compared with Observation for the Overall Diagnosis of all 5 Medial Meniscal States in Cranial Cruciate Ligament (CrCL) Decient and
Intact Stie Joints
CrCL-Decient Stie Joints
Surgery
Arthroscopy

Diagnostic
Approach

Observation
Probing
Craniomedial (CrMed) Observation
Probing
Caudomedial (CdMed) Observation
Probing
See Table 1 for key.

Sn
(95% CI)
37
83
23
47
20
30

(2056)
(6594)
(1042)
(2866)
(839)
(1549)

Sp
(95% CI)
84
96
81
87
80
83

(7690)
(9199),z
(7387)
(7992)
(7287)
(7589)z

CCR
(95% CI)
75
93
69
79
68
72

CrCL-Intact Stie Joints

OR
(95% CI)

(6781)
8.6 (2.629.1)
(8896),w,z
(6276)
2.9 (1.08.7)
(7184)w
(6075)
1.7 (0.55.6)
(6479)z

Sn
(95% CI)
30
77
20
27
23
53

(1549)
(5890)
(839)
(1246)
(1042)
(3472)

Sp
(95% CI)
83
94
80
82
81
88

(7589)
(8898)w
(7287)
(7488)w
(7387)
(8193)

CCR
(95% CI)
72
91
68
71
69
81

OR
(95% CI)

(6479) 7.7 (2.424.3)

(8594),w
(6075)
1.5 (0.44.9)
(6377)w
(6276)
3.8 (1.211.4)
(7487)

754

DIAGNOSIS OF MEDIAL MENISCAL PATHOLOGY

dogs, because 490% of the medial meniscal states were

correctly diagnosed with arthroscopy when combined with
probing.3 The magnication and illumination provided by
arthroscopy allows the surgeon to evaluate regions of the
medial meniscus that are otherwise difcult to observe by
conventional arthrotomy. In addition, the joint can be
evaluated in a more anatomically normal position during
range of motion, translation, and rotation, which would
be more of a challenge with arthrotomy. We found that
arthrotomy had a lower Sn, Sp, and CCR than arthroscopy, suggesting that medial meniscal tears may be missed
at the time of joint evaluation and stabilization for CrCL
insufciency. These tears may become associated with
clinical signs soon after surgery or subsequent to limb use
in the early postoperative period.5,6
When rates of medial meniscal tears after tibial plateau
leveling osteotomy (TPLO) among groups of dogs undergoing CrMed with medial meniscal release, CrMed without medial meniscal release, and arthroscopy without
medial meniscal release were compared,5 medial meniscal
tears after TPLO were 4 times more likely after arthrotomy
with no medial meniscal release than after arthroscopy
without medial meniscal release. The authors proposed
that medial meniscal tears were more likely missed with
CrMed than arthroscopy, and that medial meniscal release
may prevent undiagnosed medial meniscal tears from
becoming clinical after TPLO.5 Our results are in agreement. Despite the fact that having high Sn is advantageous because one is more likely to accurately identify if
a medial meniscal tear is present, we believe that having
high Sp may be of equal importance, particularly depending on how the medial meniscus is addressed if a tear
is identied. Having low Sp will make no difference if the
treatment selected for any medial meniscal tear identied
is meniscectomy, as long as no medial meniscal tears are
missed (high Sn); however, high Sp is of utmost importance if the treatment selected for some medial meniscal
tears may include primary repair, which may ultimately
preserve the long-term function of the medial meniscus.
In our study, all methods had a higher Sp than Sn.
Similar to Sn, arthroscopy had a higher Sp than arthrotomy, with probing increasing the Sp of all methods when
compared with observation alone. We missed about 80%
of the meniscal tears using arthrotomy without probing.
Therefore, as previously suggested, a low rate of subsequent medial meniscal tears without performing medial
meniscal release can be achieved through an accurate
initial diagnosis of medial meniscal pathology and a joint
stabilization method that will protect the medial meniscus
from further injury. Further support for the importance
of improving the overall Sp of diagnosing medial meniscal tears is the fact that even medial meniscal release
is not completely protective against the development of
subsequent tears.5

In CrCL-decient stie joints, probing of the medial

meniscus increased the odds of correctly diagnosing medial meniscal tears with arthrotomy and arthroscopy
by
23 and 9 times, respectively. Furthermore, arthrotomy with observation alone increased the likelihood of
incorrectly diagnosing a vertical longitudinal or peripheral detachment tear by a factor of 21 times when compared with probing when only stie joints with medial
meniscal tears were considered. It is evident from these
results that probing the medial meniscus should be part
of standard evaluation during both arthroscopy and
arthrotomy. The tip of the probe can be inserted in areas
that are not visible and can be used to detect previously
unidentied vertical longitudinal tears by feeling for
ssures, gaps, or irregularities. An accurate diagnosis of
partial or full thickness vertical longitudinal tears would
allow the surgeon to choose either partial meniscectomy
or meniscal repair, depending on the overall integrity of
the medial meniscus and the location of the pathology
with respect to meniscal blood supply and technical
feasibility for repair.14
Study Limitations
The experimental design of our study had some limitations. First, we used normal stie joints to evaluate the
Sn and Sp of diagnostic methods routinely performed
in pathologic stie joints. Both observation and manipulation of the medial meniscus may be more difcult in a
stie joint with chronic CrCL insufciency because of
periarticular brosis and synovial proliferation. Despite
using normal stie joints, we found lower Sn than anticipated for both arthroscopy and arthrotomy. This nding
may be a result of the type of medial meniscal tears
selected for study. We included only peripheral detachments and vertical longitudinal tears because these tears
are considered challenging to diagnose. Additionally, we
created medial meniscal tears in normal menisci, whereas
clinically, pathologic menisci are typically associated with
gross abnormalities.15 Despite this limitation, we believe
that our model is relevant for early vertical longitudinal
tears, which may be present in grossly normal menisci.
This particular medial meniscal tear is commonly minimally displaced or partial thickness, providing a diagnostic challenge even for the most-experienced surgeons.16
Our overall diagnostic accuracy may have been lower
because of our method of creating medial meniscal tears.
The surgical approach itself may have induced artifactual
changes, resulting in the high incidence of false positive
diagnoses in the group in which there was no medial
meniscal tear. Also, we created our vertical longitudinal
tears from the tibial surface of the medial meniscus;
however, clinically, partial thickness vertical longitudinal
tears may originate on either the femoral or tibial surface.

POZZI, HILDRETH, AND RAJALA-SCHULTZ

As mentioned, medial meniscal tears originating on the

tibial surface may be more of a diagnostic challenge than
those originating on the femoral surface. However, our
approach was ultimately chosen to minimize interference
with the subsequent approaches for our diagnostic methods and to provide adequate observation of the tibial
surface of the medial meniscus.
Although potential bias existed, because we knew
5 different states of the medial meniscus existed, prior
knowledge was necessary to prevent diagnosing particular tear types that were not included in the study. We
included double vertical longitudinal tears as one type of
tear of the medial meniscus. These tears are commonly
treated with partial meniscectomy by removing the axial
ap and debriding the remaining medial meniscus as
appropriate.16 We did not debride the axial portion of the
tear using any diagnostic method because it would have
changed the state of the medial meniscus; however, it is
likely that debridement of the axial portion of the medial
meniscus would have increased Sn for all diagnostic
methods, and is an additional limitation of our study.
A single investigator performed all of the diagnostic
methods, in an attempt to reduce variability that may
have been introduced by using multiple examiners. For
this reason, the results for each diagnostic method should
be interpreted carefully, because they depend on the
relative experience and skill of the examiner. In addition,
no information on repeatability or widespread usability
of the diagnostic method used can be extrapolated from
our results.
Arthroscopy and CrMed were less sensitive in stable
compared with unstable stie joints. In a stable stie
joint, the caudal pole of the medial meniscus is difcult to
evaluate because of an inability to cranially subluxate the
proximal tibia. This lower Sn conrms that manipulation
of the tibia is important for observation of the medial
meniscus. In a stable stie joint, it may be useful to
evaluate the medial meniscus at different exion angles
because a partially torn CrCL may not be taut in both
exion and extension; however, if the medial meniscus
cannot be fully observed, CdMed arthrotomy may allow
better exposure for probing the medial meniscus.
Summarily, we found that arthroscopic evaluation of
the medial meniscus should be performed for its high Sn,
Sp, and CCR in diagnosing peripheral detachments and
vertical longitudinal tears. Meniscal probing should be
part of a routine stie joint evaluation, regardless of the
diagnostic method used.
ACKNOWLEDGMENTS
Arthrex Vet Systems provided the meniscal probe that
was used during the study. The assistance of Tim Vojt in
producing the gures is gratefully acknowledged.

755

REFERENCES
1. Bennett D, May C: Meniscal damage associated with cruciate
disease in the dog. J Small Anim Pract 32:111117, 1991
2. Flo GL, DeYoung D: Meniscal injuries and medial meniscectomy. J Am Anim Hosp Assoc 14:683689, 1978
3. Ralphs SC, Whitney WO: Arthroscopic evaluation of menisci
in dogs with cranial cruciate ligament injuries: 100 cases
(19992000). J Am Vet Med Assoc 221:16011604,
2002
4. Flo GL: Meniscal injuries. Vet Clin North Am Small Anim
Pract 23:831843, 1993
5. Thieman KM, Tomlinson JL, Fox DB, et al: Effect of meniscal
release on rate of subsequent meniscal tears and ownerassessed outcome in dogs with cruciate disease treated
with tibial plateau leveling osteotomy. Vet Surg 35:705710,
2006
6. Metelman LA, Schwarz PD, Salman M, et al: An evaluation
of three different cranial cruciate ligament surgical stabilization procedures as they relate to postoperative meniscal
injuries. Vet Comp Orthop Traumatol 8:118123, 1995
7. Mahn MM, Cook JL, Cook CR, et al: Arthroscopic verication of ultrasonographic diagnosis of meniscal pathology
in dogs. Vet Surg 34:318323, 2005
8. Samii VF, Dyce J: Computed tomographic arthrography
of the normal canine stie. Vet Radiol Ultrasound 45:
402406, 2004
9. Martig S, Konar M, Schmokel HG, et al: Low-eld MRI and
arthroscopy of meniscal lesions in ten dogs with experimentally induced cranial cruciate ligament insufciency.
Vet Radiol Ultrasound 47:515522, 2006
10. Hoelzler MG, Millis DL, Francis DA, et al: Results of
arthroscopic versus open arthrotomy for surgical management of cranial cruciate ligament deciency in dogs. Vet
Surg 33:146153, 2004
11. Kivumbi CW, Bennett D: Arhroscopy of the canine stie
joint. Vet Rec 109:241249, 1981
12. Piermattei DL, Johnson KA: Approach to the stie joint
through medial incision, in Piermattei DL, Johnson KA
(eds): An Atlas of Surgical Approaches to the Bones and
Joints of the Dog and Cat. Philadelphia, PA, Saunders, 2004,
pp 346349
13. Piermattei DL, Johnson KA: Approach to the medial
collateral ligament and caudo-medial part of the stie
joint, in Piermattei DL, Johnson KA (eds): An Atlas of
Surgical Approaches to the Bones and Joints of the Dog
and Cat. Philadelphia, PA, Saunders, 2004, pp 360363
14. Cook J: Meniscal repair and regeneration, in 2007 American
College of Veterinary Surgeons Symposium, October 18
21. Chicago, IL 2007, pp 326327
15. Jackson J, Vasseur PB, Griffey S, et al: Pathologic changes
in grossly normal menisci in dogs with rupture of the
cranial cruciate ligament. J Am Vet Med Assoc 218:1281
1284, 2001
16. Whitney WO: Arthroscopically assisted surgery of the stie
joint, in Beale BS, Hulse DA, Schulz KS, et al. (eds): Small
Animal Arthroscopy. Philadelphia, PA, Saunders, 2003, pp
116157