Osteoarthritis and Cartilage (1997) 5, 173-182 © 1997 Osteoarthritis Research Society 1063=4584/97/030173 + 10 $12.00/0

OSTEOARTHRITIS and

CARTILAGE Seria l k i n e m a t i c ana lys i s of the c a n i n e h i n d l i m b jo in t s a f ter d e a f f e r e n t a t i o n and a n t e r i o r c r u c i a t e l i g a m e n t t r a n s e c t i o n

BY JOEL A. VILENSKY*, BRIAN L. O'CoNNORt, KENNETH D. BRANDT:~§, ELIZABETH A. DUNN* AND PAMELA I. ROGERSt

*Department of Anatomy, Indiana University School of Medicine, Fort Wayne; ~-Department of Anatomy, Indiana University School of Medicine, Indianapolis; ¢Rheumatology Division, Indiana University School of Medicine and §Indiana University Multipurpose Arthritis and Musculoskeletal Diseases

Center, Indianapolis

S u m m a r y

Objective and Design: transection of the anterior cruciate ligament 2 weeks after ipsilateral hindlimb deafferentation leads to osteoarthritis of the knee joint within 3 weeks. We analyzed the gait of six dogs that underwent this procedure in order to identify kinematic changes that could account for this rapid joint degeneration. All animals were video taped 1, 3, 6, 9 and 13 weeks after surgery while they trotted on a treadmill.

Results: In each dog, extension of the hip, knee and ankle joints of the unstable limb was increased, and the yield phase of the unstable knee was delayed or attenuated. When killed, five of six dogs showed a large full-thickness cartilage ulcer on the distal and/or anterior surface of the medial femoral condyle of the unstable knee; in the sixth dog, a smaller ulcer was observed. However, the severity of pathology in each individual was not obviously related to differences among the dogs in postoperative joint kinematics.

Conclusions: These data, and results of prior studies in humans and dogs, suggest that knee hyperextension resulting from limb deafferentation, and knee instability resulting from anterior cruciate ligament transection, operate in concert to create a mechanical environment (i.e., increased tibiofemoral separation and changes in the loading of articular surfaces) that results in rapid joint breakdown.

Key words: Joint degeneration, Osteoarthritis, Gait, Cartilage, Joint instability.

Introduct ion

TRANSECTION of the can ine a n t e r i o r (crania l ) c ruc i a t e l i gamen t (ACLT) is a wide ly-accep ted model for s tudy ing the e t iopa thogenes i s of osteo- a r th r i t i s (OA) [1-4]. In th is model , jo in t b r e a k d o w n progresses slowly, and r emains mild for a t leas t 3 yea r s a f te r ACLT; the rea f t e r , however , t he disease p rogresses more rapid ly , even tua l ly r e su l t i ng in fu l l - th ickness 10ss of a r t i c u l a r ca r t i l age on the dis ta l sur face o f t h e m e d i a l femora l condy le [2]. D e v e l o p m e n t of end-s tage OA in this model is g r ea t ly a c c e l e r a t e d if sensory inpu t f rom the ips i la te ra l l imb is ex tens ive ly i n t e r r u p t e d by dorsa l roo t g a n g l i o n e c t o m y (DRG; L4-$1 seg- ments) 2-weeks before A C L T [5, 6].

In p rev ious s tudies, we showed t h a t sag i t ta l

Submitted: 18 April 1996; accepted 16 October 1996. Supported by USPHS Grant AR20582 Address correspondence to: Joel A. Vilensky; Ph.D., Indiana

University School of Medicine, 2101 Coliseum Blvd., E., Fort Wayne, 1N 46805, U.S.A.

p lane k inema t i c s a re changed from base l ine in dogs a f te r e i t he r DRG or ACLT. In the fo rmer animals , which do no t deve lop OA for at leas t 16 m o n t h s a f te r su rge ry [5], ex tens ion of the dea f f e r en t ed knee jo in t was inc reased t h r o u g h o u t the ga i t cycle. This i nc reased ex tens ion was p a r t i c u l a r l y p r o m i n e n t a t t o u c h d o w n ( app rox ima te ly 20 ° ) [7, 8]. In the l a t t e r an imals ( those t h a t had u n d e r g o n e A C L T and t h a t deve loped mild OA), knee f lexion (yield) dec rea sed du r ing we igh t accep tance , 'p ropuls ive ex t ens ion ' w~as a t t e n u a t e d du r ing s tance, and the d u r a t i o n of s t ance was in i t i a l ly dec reased [9, 10].

In the p resen t s tudy, we descr ibe the sag i t ta l p lane k inema t i c s of six DRG/ACLT dogs, and four dogs t h a t u n d e r w e n t sham-DRG/ACLT. In under- t ak ing th is s tudy, we p re sumed t ha t t he acce le ra- t ion of OA in this model is a t least p a r t l y due to inc reased m e c h a n i c a l damage to the insens i t ive (and, the re fore , unp ro t ec t ed ) jo in t r e su l t i ng f rom a l t e red pa t t e rn s of jo in t movement , and t ha t a carefu l analys is of sag i t ta l p lane k inema t i c s will r evea l one or more of these pa t te rns .

173

174 Vilensky et al.: K i n e m a t i c ana lys i s o f jo in t s af ter dea f f eren ta t ion and ACLT

M e t h o d s

All aspects of this study were performed in accordance with inst i tut ional and Nat ional Insti- tute of Heal th regulat ions governing the t rea tment of ver tebra te animals. The study was ini t ia ted with the prior approval of the Purdue Universi ty Animal Care Committee, West Lafayet te , Indiana, the adminis t ra t ive uni t governing animal usage at the combined Indiana Univers i ty -Purdue Univer- sity campus in Fort Wayne.

TRAINING AND SURGERY

Using food rewards, 10 mixed-breed female dogs (approx. 20 kg) were t ra ined to t ro t at 2.3 m/s on a motorized treadmill (2.4x0.5 meter belt) [9]. After a t ra ining period tha t averaged 3 weeks (range 14-30 days), six dogs were subjected to left L4~$1 DRG followed by left ACLT (DRG/ACLT) 2 weeks later [5, 6].

The goal of our neurosurgical procedure was to in ter rupt L4-$1 sensory nerves as completely as possible. Briefly, under aseptic conditions, DRG (N=6) was accomplished by creat ing a midline skin incision from about L2-$2, stripping the muscles from the vertebrae, and drilling holes in the laminae of each vertebra. Then, for each segment, the in t radura t fat was removed by suction, and the dorsal root was identified and separated from the ventral root. When this was accomplished, the dorsal root was cut and re t rac ted laterally, and as much of the dorsal root ganglion as possible was removed wi thout jeopar- dizing the spinal nerve or ventra l root (generally, it was possible to remove approximately 0.5-1.0 cm of each dorsal root, and 50-75% of the mass of each ganglion). The muscles were re tu rned to their normal positions, and the wound closed by sutures. The sham-DRG (N= 4) was performed identically, except tha t the dorsal root was not cut, and the ganglion was not removed.

To t ransect the anter ior crucia te l igament (N=10), each dog was placed supine with its ipsilateral hip flexed approximately 45 ° and its ipsilateral knee flexed approximately 90 °. Under aseptic conditions, a midline incision was made superficial to the patel lar ligament, the skin re t rac ted medially, and the joint entered by dissecting th rough the infrapate l lar fat pad. Bleeding was control led by electrocautery. The ligament then was visualized and t ransected with either a number 15 or number 11 scalpel blade, the joint r insed free of any blood using saline and the wound closed.

We used a treadmill in this study, as in our pr ior kinemat ic studies [7-10], to ensure tha t the dogs ambulated at precisely the same speed pre- and postoperat ively. We killed the dogs 13 weeks after ACLT because we had previously shown [5] tha t knee joint pathology is well-developed in this model within this period.

VIDEO TAP I NG

The a n i m a l s were evaluated preoperat ively (before DRG or sham DRG) and 1, 3, 6, 9 and 13 weeks after ACLT ( 'after surgery'). Both sides of each animal were video taped against a back- ground grid by a high-speed video camera (120 fields/sec) located 6 m from the treadmill. Each video taping episode for each side of each dog lasted 10-20 s, and was considered a separate ' trial ' .

VIDEO ANALYSIS

Using the Peak Measurement Analysis System (Peak Performance Technologies, Boulder, CO, U.S.A.), the trials were divided into 'strides', each of which was defined as the interval between consecut ive touchdowns of the ~ntact (contralat- eral; right) hind paw. The r ight h~ndlimb was used to define strides because its movements were more consistent in the early postoperat ive sessions than those of the deafferented, unstable left hindlimb. In all but two (of 116) trials we obtained 10 acceptable consecutive strides per animal. An acceptable stride was defined as one in which the animal moved forward or backward no more than 3 cm on the moving belt during the course of the stride (i.e., the dog's velocity matched tha t of the treadmill). For each stride that was evaluated, we determined the dura t ion of stance (when the paw was in contact with the treadmill 's surface) and swing (when the paw was not in contact with the treadmill 's surface) of the left hindlimb for each of the 10 strides tha t const i tuted a trial. The mean durat ions were subsequently computed for each animal. We were unable to collect data from the left or r ight side of one DRG/ACLT dog 3 weeks after surgery; from the r ight side of the same dog at 9 weeks after surgery; and, from the left side of one sham-DRG/ACTL animal 1 week after surgery.

From among the 10 strides within a trial, we analyzed the joint angles of the stride with the dura t ion tha t best approximated the mean for tha t animal during tha t session (which, presumably, represented a ' typical ' stride). For every frame within the stride, one of the authors (E.A.D.) manipulated a computer 's mouse so that the cursor

Osteoarthri t i s and Cartilage Vol. 5 No. 3 175

on the video screen was placed over the estimated posit ion of the center-of-rotation of the fifth metatarsophalangeal , ankle (talocrural), knee (stifle) and hip (coxofemoral) joints. After smooth- ing the data, coordinates were used to calculate the angular values of the ankle, knee and hip [cf. Fig. i in 8].

To evaluate the rel iabil i ty of our technique for measuring knee joint angles, we digitized the same stride on 3 consecutive days, and subtracted the lowest value from the highest for each of the measured points. The average difference across the entire stride was 4.2 ± 1.7 °. For data smoothing, we used a But terworth filter algori thm (Peak Perform- ance Technologies, Boulder, CO, U.S.A.) with the cut-off frequency selected automat ical ly by the optimization software of the Peak system. For each animal, we determined changes in hip, knee and ankle joint angles within a stride (angular displacement) by plott ing the smoothed values from successive frames against the frame number (time), beginning at touchdown (stance) of the left paw. We then calculated the mean angular displacement pat terns for both experimental groups by dividing the displacement pat tern of each dog into 1% intervals (normalization) and calculat ing the mean and s tandard deviation for each group at each of the 100 time points (mean angle values) (Data Pac II software, Run Technol- ogies, Laguna Hill, CA, U.S.A.). Last, we plotted the mean patterns for each experimental group of dogs for each testing session.

E V A L U A T I O N O F J O I N T P A T H O L O G Y

At death, both knees of each dog were removed en-bloc. The joint space was opened widely by freeing the patella from its superior, medial and lateral at tachments, a n d reflecting it inferiorly. The joint was then immersed in 10% buffered formalin. After fixiation, the joint was disarticu- lated, photographed and inspected for evidence of OA. The location and extent of cart i lage fibrilla- t ion and the depth (full-thickness or part ial thickness) of carti lage u l c e r s w e r e noted. If full-thickness ulcers were present, the exposed bone was examined for the presence of eburnat ion or porosity. In addition, the thickest portion of the joint capsule was measured in millimeters, the menisci were examined for damage, and the location and size of osteophytes were recorded.

One of the authors (B:L.O.) graded the carti lage lesions 0-IV, where 0 = normal carti lage surface; I = fibrillated surface; II = partial-thickness ulcer; I I I= full-thickness ulcer with no dimension > 1 cm;

IV = full-thickness ulcer width at least one dimen- s ion > 1 cm.

Unfortunately, the knees of one dog in the sham-DRG/ACL group were inadver ten t ly dis- carded; therefore, pathologic f indings are pre- sented for only three of these dogs.

S T A T I S T I C A L A N A L Y S I S

The kinematics of the DRG/ACLT and sham- DRG/ACLT dogs were evaluated separate ly be- cause we were primarily interested in how each animal compared to its baseline (preoperative) condition. For each group, a repeated measure analysis of variance (ANOVA) was used to evaluate every fifth value of the 100 values used to construct the mean joint displacement patterns. This was done for the ankle, knee and hip joint values of both hindlimbs for all trials, i.e., a separate ANOVA was run for each of the six joints within each group of dogs. When a signif icant difference (P < 0.05) was detected among corre- sponding joint values for successive weeks (e.g., touchdown value for the left knee joint for the DRG/ACLT dogs), the preoperative values were compared with the respective postoperative values using Dunnett ' s procedure to determine those sessions during which significant differences exist compared with the preoperative, i.e., baseline, condition.

A repeated measure ANOVA was also used to compare the mean durat ion of preoperative versus postoperative stance, and preoperative versus postoperative swing of the DRG/ACLT dogs. Because no significant differences were found for the DRG/ACLT dogs, this evaluat ion was not under taken for the sham-operated group.

To summarize the amount of variabil i ty in knee joint angles from dog to dog, which is depicted in Fig. 2 and 3, the s tandard deviations for each of the 100 points in a mean gait cycle within each group were added and divided by 100 to obtain a s i n g l e measure of variabil i ty for the entire gait cycle.

R e s u l t s

E X T E N T O F D E A F F E R E N T A T I O N

The ipsilateral patel lar reflex was abolished after DRG and most regions of the ipsilateral hindlimb were unresponsive to pin prick. However, in no case was the limb completely insensitive, and most animals reacted to pin prick along the posterior aspect of the leg and thigh, and the plantar and dorsal surfaces of the paw. The

176 Vilensky e t al.: Kinematic analysis of joints after deafferentation and ACLT

sham-operated animals showed no loss of sen- sation.

G E N E R A L L O C O M O T I O N

Immediately after DRG, the dogs did not always use their deafferented limb during ambulat ion, but by 2 weeks after DRG (i.e., at the time of the ACLT) the limb was typical ly incorpora ted into the gait pat tern. DRG/ACLT dogs used their unstable limb immediately after ACLT, but one of the sham-DRG/ ACLT dogs in te rmi t ten t ly walked with the un- stable limb elevated for up to 3 weeks after surgery and another for up to 6 weeks. Al though DRG/ACLT dogs were not re luc tan t to use the unstable limb, the movements of tha t limb ~vere occasionally var iably abnormal, e.g., the left paw was not or iented parallel to the plane of progression or the paw slipped unpredic tably when placed on the substrate.

/

J O I N T P A T H O L O G Y

As expected, the stable knee joint in all dogs was grossly normal at death, whereas the unstable (left) knee exhibited prominent loss of a r t icu lar cartilage, osteophytes and capsular th ickening (Fig. 1). These changes were much more severe in dogs tha t underwent DRG before ACLT than in dogs tha t underwent sham-DRG before ACLT.

For example, in five of six DRG/ACLT dogs, a large full-thickness cart i lage ulcer (Grade IV) was present on the distal and/or anter ior surface of the medial femoral condyle of the unstable knee [Fig. l(a) and (b)]. In two of these five knees, the exposed bone on the femoral condyle and poster ior third of the medial t ibial condyle was eburnated. In one of these five DRG/ACLT dogs, full-thickness carti- lage ulcers were present also on the distal surface of the lateral femoral condyle and lateral t ibial condyle. In the sixth DRG/ACLT dog, a small full-thickness cart i lage ulcer (Grade III) was present at the distal aspect of the medial t rochlear ridge. This ulcer [Fig. 1(c)] was the least severe lesion identified in this group. The medial meniscus was frayed in all six DRG/ACLT dogs, and the lateral meniscus was frayed in two. In each dog, the medial capsule or medial port ion of the posterior capsule exceeded 0.75 cm in thickness (Table I). Last, in one dog, a small mass of loose, disorganized connect ive tissue was a t tached to the cut ends of the ligament. This tissue spanned an interval of approximately 0.5cm between the stumps of the l igament and was freely mobile when manipulated with a metal probe. In its consistency, it more closely resembled the fibro-fatty tissue of

!ii~!~ i~ ~ ~

FIG. 1. Distal femora from three DRG/ACLT dogs showing OA of varying severity. The femur from the deafferented and unstable knee is depicted on the right, and that from the contralateral knee, which appears normal, on the left. (a) A large cartilage ulcer is present on the distal aspect of the medial femoral condyle of the unstable knee. The upper-half of the ulcer consists of porous bone surrounding an area of fibrous tissue, whereas the lower-half consists of smoothly polished subchondral bone (Grade IV). (b) A full-thickness cartilage ulcer is present at the anterior aspect of the medial femoral condyle, adjacent to the intercondylar groove. The exposed bone consists of a central porous region surrounded on three sides by smoothly polished bone (Grade IV). (c) A small full-thickness ulcer is present on the distal aspect of the medial trochlear ridge (Grade III).

O s t e o a r t h r i t i s and Carti lage Vol. 5 No. 3 177

Table I Frequency distribution of cartilage lesion grades*

N 0 I II III IV V

Femoral condyles DRG/ACLT 6 - 1 3 2 Sham-DRG/ACLT 3 1 2 - - -

Tibial condyles DRG/ACLT 6 1 1 1 1 2 Sham-DRG/ACLT 3 1 2 . . . .

*0 = normal cartilage suface; I = fibrillated surface; II -- thinned or partial thickness ulcer; III=full-thick- ness ulcer, with no dimension > 1 cm; IV = full-thickness ulcer, with at least 1 dimension > 1 cm; V = eburnation.

the infrapatel lar fat pad than firm scar tissue tha t might have been expected to contribute to the passive restraint system of the joint.

In contrast, of the four dogs tha t underwent sham-DRG/ACLT, none of the three tha t were available for analysis developed cartilage lesions tha t exceeded Grade II in severity. These lesions developed at the distal aspect of the medial t rochlear ridge (thinning, 2/3; fibrillation, 1/3); the distal (weight bearing) surface of the lateral femoral condyle (thinning, 1/3); and the medial tibial condyle (fibrillation, 2/3). The menisci were grossly normal in each case. In one of the sham-DRG/ACLT dogs, the thickness of the medial and/or posteromedial joint capsule approached 0.5 cm, but in the others it did not exceed 0.25 cm (Table I).

D U R A T I O N S O F S T A N C E A N D S W I N G P H A S E S

Postoperatively, in the DRG/ACLT dogs the durat ion of both the swing phase and the stance phase of the unstable limb were unchanged from baseline values (P > 0.05).

A N G U L A R D I S P L A C E M E N T

Ipsilateral knee

The preoperative knee displacement curve (Fig. 2) il lustrates the typical quadrupedal pattern, with an initial 'yield' (flexion) phase of stance associated with weight acceptance, followed by a 'propulsive' (extension) phase that ends with lift-off about midway through the cycle. The swing phase is then characterized by an initial epoch of flexion, followed by extension as the limb prepares for the next touchdown.

When compared with baseline, the mean dis- placement pat tern of the unstable knee of the DRG/ACLT dogs was characterized b y increased extension throughout the entire 13 week postoper-

ative period (P < 0.05). There was no apparent trend toward either an increase or decrease in hyperextension throughout this period of time (Fig. 2). At touchdown, extension of the ipsilateral knee was, on average, 11 degrees greater than the baseline value ( P < 0 . 0 5 at 6 and 9 weeks postoperative). There was no indicat ion of in- creased extension in the sham animals (Fig. 3).

In addition, yield in the DRG/ACLT dogs was delayed postoperatively, i.e., it occurred at ap- proximately 33% through the cycle ra ther than 23% as it had preoperatively (Fig. 2). The delay in yield was not consistent in all dogs at all testing sessions (Fig. 4). The magnitude of yield was reduced in some trials regardless of whether or not the onset of yield was delayed (e.g., dog 'B' in Fig. 4 generally showed no yield postoperatively). In the sham animals, a distinct yield phase was not evident at week 1, but appeared to be mildly reduced (P > 0.05) in subsequent weeks (Fig. 3).

The variat ion in the pattern of displacement in the DRG/ACLT animals (Fig. 2) was not obviously related to the severity of knee joint pathology. For example, the DRG/ACLT dog with the mildest joint pathology [Fig. 1(c)] exhibited postoperative knee displacement patterns tha t had the most marked deviation from baseline [Fig. 4(c)].

In general, the mean standard deviation (S.D.) of the knee joint angles did not change from preoperative values in either group of dogs. Thus, in the six DRG/ACLT dogs the mean S.D. throughout the durat ion of the stride was 10 °, except 13 weeks after surgery when the mean S.D. reached 13 °. Similarly, in sham-DRG/ACLT dogs, the mean S.D. of the knee angle (also about 10 °) did not differ from baseline. Despite this overall lack of change, the error bars in Fig. 2 suggest that postoperative variabil i ty in the DRG/ACLT dogs decreased during stance and increased during swing. This t rend was not evident in the sham-operated animals (Fig. 3).

Other joints

Table II summarizes the major characterist ics of the postoperative changes in the displacement patterns of the ipsilateral and contra la tera l joints in the DRG/ACLT and sham-DRG/ACLT groups. In comparison with the baseline values, the most prominent difference between the two groups was the increased extension of the ipsilateral hip, knee and ankle joints of the DRG/ACLT dogs during stance and swing. In contrast, in sham-DRG/ACLT animals, knee extension was not increased during either phase of the cycle, and extension of the

178 V i l e n s k y e t al.: K i n e m a t i c a n a l y s i s o f j o i n t s a f t e r d e a f f e r e n t a t i o n a n d ACLT

ipsilateral hip and ankle joints was increased only during stance.

In comparison with baseline, the changes in the displacement pat terns of the contralateral joints that developed after ACLT were similar in the two experimental groups. Flexion was generally in- creased during stance, and extension was gener- ally increased during swing (Table II).

D i s c u s s i o n

In the present study, we documented the sagittal plane kinematics and pathology of dogs whose ipsilateral hindlimb had been deafferented before ACLT. We then at tempted to identify posto~era-

tive kinematic changes that might account for the especially severe OA that this s tudy and previous ones [7, 8] have shown develops in these dogs in comparison with dogs with intact sensory nerves that have undergone ACLT [5, 6].

Despite the variable severi ty of the joint pathology within the DRG/ACLT and sham-DRG/ ACLT groups, the most severe OA changes in the sham-DRG/ACLT group were milder than the mildest OA pathology in the DRG/ACLT group (Table I). These observations are consistent with our previous findings, viz., t ha t dogs that undergo DRG before ACLT develop very severe OA, whereas OA in dogs that undergo ACLT or sham-DRG/ACLT remains mild for as long as 72

(a)

170 Stance

160- ~

150

12° f ito ~

IO0

90 i l 0 25

140- © ©

~130

Swing

50 75 Percent of cycle

100

(b) (c) _

1~0 ° 1 ~ ~

0 25 50 75 100 0 25 50 75 100 Percent of cycle Percent of cycle

(d) 17C,

160

150

140

~130

~ i20

110

100

(0 170"

/

I I I 25 50 75 100 Percent of cycle

(e) 170 - v v v v v v v v v i

160 ~ : ~

I k

120 - al

110 -

1 0 0 -

90 I I 0 25 50 75

Percent of cycle

160

150 -

40

~130

120

90 L 0 100 0 25 50 75 100

110

100

90 I I

Percent of cycle

FIG. 2. Mean angular displacement patterns for the deafferented and unstable knee. The heavy line depicted on the graphs is the mean preoperative pattern; the dashed line represents the respective mean postoperative pattern; the errors bar indicate _+ 1 S.D. of the postoperative curves except for the preoperative graph for which baseline variability is depicted. The arrowheads indicate a significant difference (P < 0.05) from the preoperative value. The abscissa for each graph is percent of cycle (stride), beginning at touchdown (1% of stride) and ending immediately before the next touchdown of that limb (100% of stride). On the preoperative graph the cycle is divided into stance and swing, and the stance component is divided into yield and propulsive phases. (a) Preoperative; (b) 1 week; (c) 3 weeks, (d) 6 weeks; (e) 9 weeks; (f) 13 weeks.

O s t e o a r t h r i t i s a n d C a r t i l a g e Vol. 5 No. 3 179

(a)

1601 I /

150 ttan%TTT T Swing

130

12ot- T/It 110~- __! .~ I ~ T F I 100~- ~ _~ ~I--

80

70 0 25 50 75 100

Percent of cycle

~b) 160

150

140

130-

~120

110

100

9O

8O

70 0

" " ~ W W W W V

- !

I I 25 50 75 100 Percent of cycle

(c) 160

150

140'

130.

120

~110

100

90

80

70 I I I 25 50 75 Percent of cycle

I00

(d) 160

150

140

130

~120"

110

100

80

70 0

L

I I 25 50 75 Percent of cycle

t 100

(e)

160

150

140

130

~ 120 v II0

100

9O

8O

7O 0

I I . . . . I 25 50 75 Percent of cycle

FIG. 3. Identical to Fig. 2 except tha t data are for the

100

(0 160i_ ~

15o ~ T ~

13o _~_2I~±~-- -±

120

~ II0 --

I00 -

90 -

80 -

70 I I 0 25 50 75 100

Percent of cycle

sham-DRG/ACLT animals. Key as for Fig. 2.

weeks after knee surgery [6]. This may be associated w i t h t h e increase in knee extension in the DRG/ACLT group, which was not evident in the sham animals.

In ACLT dogs, and presumably, in DRG/ACLT dogs, especially rapid compression between the femur and tibia occurs at touchdown [11]. Furthermore, in ar t iculated canine cadaver knees after ACLT, a posterior shift occurs in medial and lateral contact areas after ACLT [12], which is similar to changes identified after ACLT in the hyperextended knee of human cadavers [13]. We speculate that in dogs who have undergone DRG before ACLT, sensation-dependent protect ion otherwise afforded by muscular activity to the unstable and painful joint is reduced or abolished. This could increase the potential for abnormal separat ion of the tibial and femoral surfaces, change the areas of ar t iculat ion in the unstable and hyperextended knee, and produce excessive impact loading [14]. Presumably, these mechanical

factors induce rapid carti lage wear, and acceler- ated OA.

We have previously reported that 12 weeks after surgery the peak vertical force exerted by the unstable limb in DRG/ACLT dogs and sham-DRG/ ACLT dogs is lower than the baseline value [15]. This may explain the delayed and/or reduced yield in both groups, i.e., the magnitude and, pre- sumably, form of the "yield curve' reflects the amo.unt and manner in which load is applied to the limb at touchdown [16]. Also, a l though the peak vertical forces generated by the ipsilateral limb of dogs that have undergone ACLT, sham-DRG/ ACLT, and DRG/ACLT dogs are all reduced from baseline, the forces are reduced the least in the DRG/ACLT dogs [15]. Therefore, among these groups, all Of Which have knee instability, the DRG/ACLT knee is consistently subjected to greater loading during each gait cycl.e.

In dogs subjected only to ACLT, the variabi l i ty in knee joint angles during gait increases [9, 11].

180 V i l e n s k y e t a l . : K i n e m a t i c a n a l y s i s o f j o i n t s a f t e r d e a f f e r e n t a t i o n a n d ACLT

(a) 180

160

140 ©

120

100

8G

"",-~ : ' , ~ [

",. : "d4. '

I I I I

Percent of cycle

(b) (c)

. , I I I I I

Percent of cycle

+ . : '

_L..--J-..--"-X\ ..."i ' . . . . . . !,']i

V I I . . . . . I I .... I

Percent of cycle

- - Preop. - - - - 1 Week --~;-- 3 Weeks ........... 6 Weeks . . . . 9 Weeks : . . . . . 13 Weeks

FIG. 4. Angular displacement patterns at baseline (heavy line) and postoperatively for the unstable and deafferented knee of the three DRG/ACLT dogs depicted in Fig. 1, which exhibited OA of varying severity at death.

Accordingly, we expected that the variabi l i ty in joint angles among the sham-DRG/ACLT dogs in the present study would increase. However, this seemed not to have occurred (Fig. 3), perhaps because these four animals had an unusual ly uniform response to knee instabil i ty (compare Fig. 3 with Fig. 4 in ref. 9).

Also, a l though the overall variabi l i ty of knee joint angle did not increase among DRG/ACLT dogs postoperat ively (except at 13 weeks after surgery), the postoperat ive pat terns of individual dogs were different from one another , as well as from their respective preoperat ive pat terns (Fig. 4). Thus, a l though the mean postoperat ive pat tern (Fig. 2) character ized the population, it did not necessari ly represent the kinematic pat tern employed by any one dog. This is not surprising because individual dogs are known to differ in their kinematic response to knee instabil i ty [11].

Despite the individual differences in displace- ment pat terns, knee hyperextens ion was a consist-

ent and pronounced feature of each DRG/ACLT dog th roughout the postoperat iveg period. Fur- thermore, during the stance phase of the dogs tha t underwent DRG/ACLT, the variabi l i ty in the knee joint angle appeared to decrease (Fig. 2). Thus, hyperextension was especially uniform during weight bearing, presumably the time, during the gait cycle, tha t joint damage occurs.

Jo in t hyperexten t ion is a common kinematic feature after deafferentat ion [17++19], and may be due to an intr insic proper ty of the central nervous system when sensory input from a limb is substantial ly reduced. Alternatively, it is possible tha t by contact ing the surface with a hyperex- tended limb, the animal generates a 'jolt ' at touchdown, which is sensed th roughout the body, and therefore, can be in terpre ted by the central nervous system as a signal tha t it is safe to place weight on the limb [20]. Cont inued hyperextension th roughout the remainder of the gait cycle may then be a byproduct of the hyperextension at touchdown.

Table II General changes in mean displacement patterns of ipsilateral and contralateral joints compared to the preoperative

(baseline) condition

Joint DRG/ACLT group Sham-DRG/ACLT group

Ipsilatera] Hip Knee Ankle

Contralateral Hip Knee Ankle

T Extension stance* & swing T Extension stance and swing*; $ yield*, cf. text

Extension stance* & swing

T Flexion stance* T Flexion stance*; T extension swing T Flexion stance*; T extension swing

T Extension stance* Loss or yield T Extension stance*

T Flexion stance T Flexion stance*; T extension swing* T Flexion stance*; T extension swing

T Increased; $ decreased; *P < 0.05 based on t-factor ANOVA and Dunnett's test compared with baseline (cf. text).

O s t e o a r t h r i t i s and Cart i lage Vol. 5 No. 3 181

In the DRG/ACLT dogs, the kinematics of the cont ra la te ra l joints were general ly similar to those of dogs tha t _underwent sham-DRG/ACLT. This suggests tha t residual sensat ion from the operated limb and sensat ion from other parts of the body (e.g., stable limbs, ves t ibular system, visual system) informed the centra l nervous system tha t compen- satory adjustments in the movements of the in tac t hindlimb were required to ensure adequate propulsion. Such compensat ion was, however, insufficient to protec t the unstable knee from rapid joint breakdown.

The poor cor re la t ion between the severi ty of pathologic changes and postoperat ive kinematics in DRG/ACLT dogs may be explained by three factors: first, the severi ty of joint pa thology may depend on qual i ta t ive or quant i ta t ive aspects of limb use during times when dogs were not being analyzed. Second, the extent of deafferentat ion, as reflected b y the response to 'pin prick' , varied among dogs. Notably, of the DRG/ACLT group, the mildest OA occurred in the dog that re ta ined the most sensation. This dog was also the one in which connect ive tissue spanned the cut ends of the ACL, perhaps helping to stabilize the joint. Presumably, dogs tha t re ta ined more sensation could detect knee pain and/or instabil i ty bet ter than dogs tha t re ta ined less sensation, and were more likely to use muscles to stabilize the joint, in an effort to avoid these sensations.

Presumably, residual sensation, par t icu lar ly in the left hind paw, enabled DRG/ACLT dogs to perceive when the paw was in contac t with the surface, and to preserve the basic t iming of movements at the selected treadmill speed. Therefore, the rapid development of OA in DRG/ACLT dogs cannot be a t t r ibu ted to an increase in the dura t ion of applied forces due to a lengthening of s tance phase .

We suggest tha t especially-rapid knee degener- at ion occurs in DRG/ACLT dogs because of the combinat ion of ACLT and hyperextension, which presumably creates a mechanical envi ronment tha t subjects the jo int to excessive t rauma with each paw contact . Indeed, humans with anter ior cruciate l igament deficiency also show greater knee extension at foot-strike than controls [21], and abnormal shock-absorbing mechanisms and knee hyperextens ion at foot-strike have been described in humans with knee pain who were considered to be 'preos teoar throt ic ' [22]. Because knee extension at touchdown is also increased in humans with per ipheral sensory loss [20], it is possible tha t the increased knee extension in pat ients with a deficient an ter ior crucia te liga- ment occurs because of reduced propr iocept ion

[23]. Regardless of the cause, the possibi l i ty tha t reducing knee extension at foot-str ike in humans with anter ior crucia te l igament deficiency and/or knee pain might reduce the preva lence and/or severi ty of knee OA war ran t s considerat ion.

A c k n o w l e d g m e n t s

We are grateful to Ms Carla Barrett for assistance with the care of the animals and video taping.

R e f e r e n c e s

1. Brandt KD, Braunstein EM, Visco DM, O'Connor B, Heck D, Albrecht M. Anterior (cranial) cruciate ligament transection in the dog: a bona fide model of osteoarthritis, not merely of cartilage injury and repair. J Rheumatol 1991;18:436-446.

2. Brandt KD, Myers SL, Burr D, Albrecht M. Osteoarthritic changes in canine articular carti- lage, subchondral bone, and synovium fifty-four months after transection of the anterior cruciate ligament. Arthritis Rheum 1991;34:1560-1570.

3. Marshal JL, Olsson S-E. Instability of the knee: a long-term experimental study in dogs. J Bone Joint Surg [Am] 1971;53:1561-1570.

4. McDevitt CA, Gilbertson E, Muir H. An experimen- tal model of osteoarthritis: early morphological and biochemical changes. J Bone Joint Surg [Br] 1977;59:24-35.

5. O'Connor BL, Palmoski MJ, Brandt KD. Neurogenic acceleration of degenerative joint lesions. J Bone Joint Surg [Am] 1985;67:562-572.

6. O'Connor BL, Visco DM, Brandt KD, Albrecht M, O'Connor AB. Sensory nerves only temporarily protect the unstable canine knee joint from osteoarthritis. Evidence that sensory nerves reprogram the central nervous system after cruciate ligament transection. Arthritis Rheum 1993;36:1154-1163.

7. Vilensky JA, O'Connor BL, Brandt KD, Dunn EA, Rogers PI. Serial kinematic analysis of the canine knee after L4-$1 dorsal root ganglionectomy: implications for the cruciate deficiency model of osteoarthritis. J Rheumatol 1994;21:2113-2117.

8. Vilensky JA, O'Connor BL, Brandt KD, Dunn EA, Rogers PI. Serial kinematic analysis of canine hind limb joints after unilateral L4-$1 dorsal root ganglionectomy: insights into locomotor control mechanisms. J Electromygr Kinesiol 1995;5:169- 176.

9. Vilensky JA, O'Connor BL, Brandt KD, Dunn EA, Rogers PI, DeLong CA. Serial kinematic analysis of the unstable knee after transectiobn of the anterior cruciate ligament: temporal and angular changes in a canine model of osteoarthritis. J Orthop Res 1994;12:229-237.

10. Vilensky JA, O'Connor BL, Brandt KD, Dunn EA, Rogers PI. Serial kinematic analysis of the trunk and limb joints after anterior cruciate ligament transection: temporal, spatial, and angular changes in a canine model of osteoarthritis. J Electromyogr Kinesiol 1994;4:181-192.

11. Korvick DL, Pijanowski GJ, Schaffler DJ. Three-di- mensional kinematics of the intact and cranial

182 Vilensky e t al. : Kinematic analysis of joints after deafferentation and ACLT

cruciate ligament-deficient stifle of dogs. J Biomech 1994;27:77-87.

12. Flechtenmacher J, Borodkin JL, Hol]ister SJ, Dedrick DK. Tibial contact pressure distribution changes in the canine cruciate ligament model of osteoarthrosis. Trans Orthop Res Soc 1993;18:721.

13. Fuss FK. The restraining function of the cruciate ligaments on hyperextension and hyperflexion of the human knee joint. Anat Rec 1991;230:283-289.

14. Radin E, Paul IL. Responses of joints to impact loading. Arthritis Rheum 1971;14:356-362.

15. O'Connor BL, Visco DM, Brandt KD. Gait alter- ations in dogs with unstable knee joints: evidence that the central nervous system is reprogrammed to protect joints. Trans Orthop Res Soc 1992;17:478.

16. @rillner S. Locomotion in vertebrates: central mechanisms and reflex interaction. Physiol Rev 1975;55:247-304.

17. Goldberger ME. Locomotor recovery after unilateral

hindlimb deafferentation in cats. Brain Res 1977;123:59-74.

18. Wiesendanger M. Rigidity produced by deafferenta- tion. Acta Physiol Scand 1964;62:160-168.

19. Yaknitsa IA, Bulgakov NV, Pilyavskii AI. Kin- ematics of different types of locomotor movements in the deafferented rat. Neurophysiol 1987;19:394- 398.

20. Perry, J. Gait Analysis. Thorofare, NJ: Slack Incorporated; 1992.

21. Berchuck M, Andriachhi TP, Bach BR, Reider B. Gait adaptations by patients who have a deficient anterior cruciate ligament. J Bone Joint Surg [Am] 1990;72:871-877.

22. Radin EL, Yang KH, Riegger C, Kish VL, O'Connor JJ. Relationship between lower limb dynamics and knee joint pain. J Orthop Res 1991;9:398-405.

23. Barrack RL, Skinner HB, and Buckley SL. Proprio- ception in the anterior cruciate deficient knee. Am J Sports Med 1989;17:1-6.