POSTGRAD. MED. J. (1964), 40, 521

QUADRICEPSPLASTYTHE TREATMENT OF STIFF KNEE FOLLOWING TRAUMA

E. A. NICOLL, C.B.E., M.D., F.R.C.S.

Chesterfield Royal Hospital and Mansfield General Hospital.

THERE are some joints, such as the wrist,ankle and tarsus, in which stiffness is com-patible with excellent function, and even theeffects of an arthrodesis may be masked almostbeyond recognition. In the knee joint, however,any serious limitation of movement is a veryconsiderable handicap, the effects of whichcan neither be disguised nor overcome. Thepatient is debarred from a considerable rangeof occupations and recreations; he may haveto change both his job and his hobbies, andhe is constantly being embarrassed by all thepeople who fall over his outstretched leg inbuses, theatres, cinemas and the like, many ofwhom in their ignorance do not hesitate toupbraid him for a clumsy and inconsideratefellow.

Such stiffness may follow either diseaseor trauma. When it is the result of infectionor arthritis, attempts at restoring movementare not only disappointing but dangerous, andif any surgery at all is called for because ofpain, instability or progressive deformity, it isusually better to do an arthrodesis. When it fol-lows trauma, however, the outlook is almostcompletely reversed and the most gratifyingresults can be obtained, even in the stiffestand most unpromising knees, by the opera-tion of quadricepsplasty. This is in fact oneof the most rewarding operations in theorthopaedic surgeon's repertoire, but one getsthe impression that it is also one of the mostneglected, simply because its possibilities arenot sufficiently appreciated. Since Bennett'soriginal paper in 1922 and Thompson's classiccontribution in 1944, the subject has receivedvery little attention in the literature, thoughsuch reports as have appeared have been con-sistently favourable (Judet, 1959; van Nes,1962; Nicoll, 1963; Hesketh, 1963).This neglect of a good operation is all the

more perplexing when one reflects on the

enormous amounts of time, energy andingenuity that have been spent, for the mostpart unprofitably, on the stiff hip. Moreover,the type of stiff knee most amenable to sur-

gery occurs predominantly in young, fit andotherwise active people, so it should notlightly be ,accepted either by surgeon or

patient. The problem is bound to increase inthe future as the mounting toll of road acci-dents pursues its apparently inexorable course.These high velocity injuries produce exactlythe kind of fracture that can result in astiff knee, and it is therefore important thatthe operation of quadricepsplasty, which hassuch rewarding possibilities, should be more

widely known. The object of the presentpaper is to present an analysis of 38 cases

operated on during the past 16 years, theresults of which have been most encouraging.

CausesThe causes of stiff knee requiring quad-

ricepsplasty in the present series are shown inTable 1.

TABLE 1

CAUSES OF KNEE STIFFNESS (38 PATIENTS)

Cause Number ofpatients

Fracture of femoral shaftNormal union ........ . 9Delayed union ...... ... 9 26Non-union requiring grafting 8

Fracture of femoral condyles ... 2Fracture-dislocation of knee ... 2Fracture of tibial plateau ...... 3Synovectomy ...... ..... 2Hip conditions ... ... ... ... 3

Total ... 38

by copyright. on M

arch 16, 2020 by guest. Protected

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.40.467.521 on 1 S

eptember 1964. D

ownloaded from

POSTGRADUATE MEDICAL JOURNAL

FIG. 1.-Superimposed radiographs showing the ex-cursion of the patella between full extension and130o of flexion.

By far the commonest cause is fracture ofthe femoral shaft, particularly if this is inthe lower half of the bone and is complicatedby severe soft tissue damage, comminution ordelayed union. In the present series, 26 cases(70%) followed fractures of the femur and17 of these (65%) were complicated by delayedunion or non-union. It would be wrong to inferfrom these figures, however, that stiffness ofthe knee is a frequent complication of femoralshaft fractures. Half the patients in the presentseries were referred from other centres, leavingonly 13 cases in 16 years of entirely endogenousorigin. These came from two hospitals servinga population of 400,000 people in an area ofheavy industry with a high accident rate. It istherefore a comparatively rare complication,as of course it should be in any properlyorganised accident service. Indeed, the sur-geon who could produce a large series entirelyof his own making would probably feel some-what reticent about publishing his experiences.

All the 26 cases following fracture of thefemoral shaft resulted from conservativetreatment. There was no instance of stiff knee

following intramedullary nailing, although thisprocedure had been adopted in nearly a thirdof all femoral shaft fractures treated duringthe 16-year period under review. Moreover,these operations were all done from theanterior approach, so the current fashion ofcondemning this approach on the grounds thatit causes stiff knees is not supported by thefacts. Stiffness is caused not by a clean sur-gical incision through muscle, but by the pro-longed immobilisation of damaged soft tissuesspanning both sides of a joint.

Mechanism of StiffnessThe mechanism of knee stiffness following

trauma can be explained in terms of thepathological anatomy. Blocking may occur inany or all of four different places:-1. Lower third of thigh and suprapatellar

pouch, due to fibrosis of vastus intermedius.2. Retropatellar, due to adhesions from patella

to femur.

September, 1964522by copyright.

on March 16, 2020 by guest. P

rotectedhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.40.467.521 on 1 Septem

ber 1964. Dow

nloaded from

NICOLL: Quadricepsplasty3. Paracondylar, due to adhesions of the

aponeurotic expansions of the vasti to thesides of the femoral condyles.

4. Inside the knee joint, due to adhesionsbetween the femoral and tibial articularsurfaces, or to contracture of the posteriorcruciate ligament.Adhesions in the first three places act by

limiting the distal excursions of the patelladuring flexion and it is not always realised justhow much this is. The superimposed radio-graphs in Fig. 1 show that in fact it travelsa distance of 3½inches (9 centimetres) through-out its full range and more than half of thisoccurs in the first 60°. This alsorepresents the excursion of the quadricepstendon and anything that anchors this tendonwill limit flexion. If, for example, there areadhesions which reach full stretch after thepatella has travelled 2 inches, movement will belimited to about 60 degrees. Thequadriceps tendon is a complex mechanism,consisting of two central components (rectusfemoris and vastus intermedius) which fuse to.form a single tendon above their insertioninto the patella, and two lateral expansions(vastus medialis and lateralis) which pass roundthe sides of the patella, to which they areattached before continuing distally to beinserted into the ligamentum patellae and thetibia. In the normal knee, these aponeuroticexpansions glide freely over the femoral con-dyles, from which they are separated by asulcus lined by synovial membrane. If thissulcus is obliterated by adhesions, theaponeurotic expansions get stuck down, and byvirtue of their attachment to the patella canlimit flexion just as effectively as when thepatella itself is stuck. This is, in fact, one ofthe commonest mechanisms of stiffness in allhinge joints and it can occur from prolongedimmobilisation alone, even when the joint itselfhas not been involved in the fracture.The component of the quadriceps most

frequently damaged is the vastus intermedius,which is wrapped round the bone and there-fore most liable to be torn or contused in thetype of comminuted fracture by direct violenceso frequently seen in the present day highvelocity injury. The muscle becomes fibroticand inextensible and through its deep attach-ment to the rectus femoris tendon it anchorsthe patella like a check rein (Fig 2). Thistethering action is clearly seen in idiopathiccontracture of the vastus intermedius, a rare

Fig. 2 (a)

Fig. 2 (b)

FIG. 2.-Diagrammatic representation of effect offibrosis of vastus intermedius and of adhesionsinside the knee joint.

(a) Knee extended : adhesions relaxed.(b) Knee flexed : adhesions tight.

condition occurring in young children. Flexioncan be restored quite easily in these patientsby dividing the fibrotic intermedius tendon anddisconnecting it from the rest of the quad-riceps apparatus.The patella is frequently bound down to

the front of the femur by adhesions (Fig. 2)and in several cases in the present series it wasstuck so firmly that it had to be dissected offwith a scalpel.

Finally, there are often adhesions in theknee joint itself (Fig. 2). These may extendfrom the femoral to the tibial condyles orthere may be a contracture of the posteriorcruciate ligament. This ligament is normallyslack in extension and taut in flexion. If itcontracts, as it may do if it is immobilised fora long time in its shortened position, it willlimit flexion.

September, 1964 523

by copyright. on M

arch 16, 2020 by guest. Protected

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.40.467.521 on 1 S

eptember 1964. D

ownloaded from

POSTGRADUATE MEDICAL JOURNAL

There is one other way in which flexion canbe blocked after all the above causes have beendealt with, and that is by contracture of therectus femoris itself. In this condition themuscle becomes so fibrous that it completelyloses its extensibility. To use an analogy, thepatella is then, as it were, attached to thepelvis by a piece of string instead of a pieceof elastic. Fortunately this is rare-it only hap-pened four times in 38 cases in the presentseries-but when it does occur it is comparableto contracture of the tendo Achillis causingfixed equinus of the ankle and must be dealtwith along similar lines by lengthening thetendon. This was the basis of the originalBennett procedure, but interference with thetendon should be avoided if at all possiblebecause it always leads to loss of activeextension. This is related to the long excursionof the tendon, which may have to belengthened as much as 3 inches to gain 130degrees of flexion. When the knee isstraightened again after such a lengtheningprocedure, the elongated tendon curls up in aseries of folds and it is obvious that thefibrotic rectus muscle is quite incapable oftaking up so much slack. In a normal kneethis would not matter because full extensioncan be achieved by the vasti alone, as can easilybe demonstrated in any case of patellectomyfor arthritis in which no attempt has beenmade to restore continuity between the rectustendon and the ligamentum patelle. However,in the circumstances in which the rectus femorishas to be lengthened in quadricepsplasty theexpansions of the vasti have already beendisconnected from the patella and theirextensor action seriously interfered with.However, if loss of voluntary extension has tobe accepted in order to gain a reasonableamount of flexion, it is a penalty well worthincurring, provided there is no loss of passiveextension.The disability in these cases of "extension

lag" is surprisingly slight, and in explainingthis to patients or their doctors there is nothingso convincing as visual demonstration. Oneex-patient who is kind enough to appear fromtime to time in this role has a total range of 120degrees with an extension lag of 20 degrees,having started with a total pre-operative rangeof 30 degrees and no lag. He drives heavylorries, walks with a perfect gait, goes up anddown stairs normally, cycles, skips, climbs anddances. He cannot run fast and he would not

be selected to take a penalty in a Cup Final,but otherwise it is difficult to demonstrate his"disability".

The OperationThe operation is based on an appreciation

of the morbid anatomy and must be planned togive access to all the sites at which adhesionsmay be encountered. It is impossible to tell inadvance where these will be, although thenature of the causative injury may give a clue.In injuries around the knee, for example, theyare most likely to involve the lateral expan-sions and the inside of the joint, but thesesites are also commonly affected followingfractures in the middle third of the femur. Itis rare to find only one or two sites affectedand common to find all four. Sometimes,having divided all visible adhesions at all sites,the knee still refuses to flex until manipulated.It then yields with an obvious giving way ofsomething in the back of the joint, eitheradhesions out of sight or a contractedposterior cruciate ligament.A vertical mid-line incision is used,

extending from three inches above thepatella to the level of the knee joint. Thismay have to be extended upwards if thereis extensive scarring of the vastus intermedius,but usually the rectus femoris is free in itsupper half. The flaps are dissected laterally andthe junctions of vastus medialis and vastuslateralis with rectus femoris are identified anddivided. These incisions are carried downthrough the expansions of the vasti on eachside of the patella and about half an inchfrom it, opening up the knee joint freely.Any fibrotic vastus intermedius muscle bindingdown the deep surface of the rectus femoristendon can now be seen and removed;adhesions from the deep surface of the patellato the femur are divided and the lateralexpansions of the vasti are mobilised down tothe level of the tibia. Finally, adhesions insidethe joint itself are divided. The knee is manipu-lated at each stage of the dissection untilflexion of about 120 degrees is obtained, atwhich stage it is not necessary to go anyfurther. In most cases it is impossible tosuture the vasti back into place and no attemptto do so should be made unless it can beachieved with the knee in 90 degrees of flexion.The skin must be sutured carefully, how-ever, because sound healing is essential toallow early exercise.

524 September, 1964by copyright.

on March 16, 2020 by guest. P

rotectedhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.40.467.521 on 1 Septem

ber 1964. Dow

nloaded from

NICOLL: QuadricepsplastyIf everything at all the usual sites has been

dealt with and there is still insufficient flexionwith a very tight rectus femoris at the limitof movement, then there must be an actualcontracture of the rectus femoris. In thesecircumstances the surgeon must be preparedto lengthen the tendon, but it is undesirableto exceed 2½ inches. The same technique isused as in lengthening the Achilles tendon,by splitting it in the coronal plane.The secret of success in this operation is

that once having started, the surgeon must beprepared to go on until he has achieved hisobjective of 120 degrees of flexion. As thedissection proceeds, things begin to look moreand more alarming, and the inexperiencedoperator may well get the impression half waythrough that he is about to ruin the patient'sknee for life. If he steels himself against thetemptation to pull out in time, the kneelooks such a shambles at the end of theoperation that he may well panic and startsewing things up again. This is a fatal mistake,for he is merely recreating the conditions thathe has been at such pains to disturb. Nothingbut the skin should be sutured, but this mustbe done very carefully indeed to ensure primaryhealing and allow early movement.

In the present series the operation has beenperformed under tourniquet control and itrarely takes more than half-an-hour. Noattempt has been made to secure bleedingpoints by removing the tourniquet beforeclosing the skin. No vessels of any size arecut and haemostasis is easily secured by apressure bandage with the knee in flexion.If 120 degrees of movement has been achievedthe knee is fixed in plaster at 90 degrees forthe first three days. This flexed position hastwo advantages-it helps to secure haemostasisby keeping everything in moderate tension,and it minimises the risk of adhesions re-forming in extension during the early dayswhen the knee is too painful to move. Beforefixing the knee in this position, however, thesurgeon must be completely satisfied that thereis no undue tension on the skin wound thatmight interfere with primary healing.

After three days the plaster is removedand exercises on slings commenced. Betweenexercise periods the knee is kept in flexionon a plaster back slab, but at night it issupported in extension.

Selection and After-CareThe after-care is so important and so pro-

FIG. 3.-"Do-it-yourself" exercises. Extension of thenormal knee is used to assist flexion of the stiffknee.

longed that no patient should be selectedfor this operation who is not manifestlycapable of co-operating to the full and driv-ing himself to the limit. Most patients willalready have been assessed from this aspectbecause they will have attended forexercises and physiotherapy over a period ofmany months. The essential pre-requisites area determined patient, a determined surgeon,and a first-class rehabilitation unit in thatorder. The patient should have beenthoroughly indoctrinated during his preliminarytreatment and should suffer from no illusions asto the relative merits of active exercises byhimself and the laying on of hands by aphysiotherapist, however charming. All thepatients in the present series had failed toyield to manipulation under anaesthesia afterthey had been at a standstill for somemonths under treatment in a well organisedrehabilitation unit.

After operation, treatment must be con-tinued for upwards of six months. Pooltherapy, active exercises and quadriceps drill,form the basis of the after-care and the "do-it-yourself" type of assisted flexion illustratedin Fig. 3 is particularly useful. It is a simplegadget that can easily be fixed up in thepatient's home. One patient who came fromGreece went home a few weeks after opera-tion equipped only with this device and agood deal of initiative and determination. Hedid his pool therapy in the Aegean Sea,which he reached by exercising his quadricepson a non-stationary bicycle. He achieved anexcellent result.

If flexion is being lost, the surgeon should

September, 1964 525by copyright.

on March 16, 2020 by guest. P

rotectedhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.40.467.521 on 1 Septem

ber 1964. Dow

nloaded from

526 POSTGRADUATE MEDICAL JOURNAL September, 1964

not hesitate to manipulate the knee underanaesthesia during the post-operative months.This was done in about half the cases in thepresent series, but since adopting the flexedpost-operative position, it has become lessnecessary.

TABLE 2SUMMARISED RESULTS IN 38 CASES OF

QUADRICEPSPLASTY.There were 2 failures due to avoidable technical faultsand 2 exceptional cases with 90 and 120 degrees ofmovement before operation (see text), both of whomachieved full movement.

34 REMAINING CASES

Mini- Maxi- Aver-mum mum age

Interval (months) betweeninjury and .quadricepsplasty 4 36 14(excluding 1 case of 15 years) _____

DegreesPre-operative range nil 60 40Final range 60 140 106Flexion gain 30 100 70Extension lag (4 cases) 15 40 32

ResultsQuadricepsplasty is undoubtedly one of the

most rewarding operations in orthopaedicsurgery, and why it has failed to achieve thepopularity it deserves is a complete mystery.The results in the present series are sum-marised in Table 2.The average pre-operative range of move-

ment was 40 degrees and all but two patientshad 60 degrees or less. These two patientswere exceptional in that they had ranges of90 degrees and 120 degrees respectively,which is normally considered adequate. Butboth required full flexion to enable them toresume pre-accident work. The patient with120 degrees was a young miner on light workwho asked to have the operation because heneeded 150 degrees in order to work at thecoal face, which to him meant a 50%increase in earning power. This seemed areasonable enough indication, and in fact the

crucial extra movement was obtained veryeasily.There were two failures, due to technical

faults which prevented the normal after-carefrom being carried out. The average gain ofmovement in the remainder was 70 degrees,and only 4 patients failed to reach 90degrees of flexion or more. There was anextension lag in 4 cases. This varied from 15to 40 degrees, but in none of these was thereany loss of passive extension, so function wasexcellent and all four patients were completelyconvinced that the operation had been worth-while.

Summary1. Thirty-eight cases of stiff knee treated by

the operation of quadricepsplasty arereviewed.

2. The causes and pathological anatomy areanalysed. The operative technique isdescribed and this is based on anappreciation of the pathological anatomy.

3. Selection and after-care are of great im-portance and the latter must be persistentand prolonged. Manipulation underanaesthesia during the after-care period wasused in about half the cases, but theneed for this has diminished sinceadopting fixation in the flexed kneeposition for the first three post-operativedays.

4. The results were most rewarding. Therewere two technical failures-both avoid-able-but in the remaining 36 cases theaverage gain of flexion was 70 degrees.Lengthening of the rectus femoris tendonhad to be carried out in 4 cases only.This should be avoided if possible since itleads to loss of active extension. This didnot occur in the other cases, except as atemporary phenomenon lasting six totwelve months. Even if active extensionhas to be sacrificed it is a penalty wellworth incurring, provided there is no lossof passive extension.

REFERENCESBENNETT, G. E. (1922): Lengthening of the Quadriceps Tendon, J. Bone Jt. Surg., 4, 279.HESKETH, K. T. (1963): Experienoes with the Thompson Quadricepsplasty, Ibid., 45-B, 491.JUDET, R. (1959): Mobilisation of the Stiff Knee, Ibid., 41-B, 856.NES, C. P. VAN, SEN. (1962): Quadricepsplasty, Ibid., 44-B, 954.NICOLL, E. A. (1963): Quadricepsplasty, Ibid., 45-B, 483.rHOMPSON, T. C. (1944): Quadricepsplasty to Improve Knee Function, Ibid., 26, 366.

by copyright. on M

arch 16, 2020 by guest. Protected

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.40.467.521 on 1 S

eptember 1964. D

ownloaded from