|
 |
|||||||||||||||||||
J.P. Meyrueis, A. Cazenave
Clinique Saint-Michel - Avenue d'Orient - F-83000 Toulon
Some surgeons claim to have virtually no dislocations after total hip replacement (THR). Others - including the authors of this article - are concerned about the frequency of this complication, which is distressing to the patients and their carers.
Two methods are in current use to reduce the rate of THR dislocations:-
- cups with a long posterior wall; and
- cup augments.
To the best of our knowledge, there is nothing in the literature to document the efficacy of either technique in statistical terms.
For the past 12 years, we have been particularly interested in the use of cup augments. We have inserted 294 of these devices. In 283 cases, this was done with a view to preventing dislocation, in patients who had been given 28 mm heads.
An earlier comparison had shown augments to be of little value with 22 mm heads.
By 1983, we had developed a first range of polyethylene (PE) augments to fit the chamfer on the cups used at our centre. It would appear that our development was contemporaneous with the work done by Grammont.
The first publication on augments was by Olerud and Karlstrom, in 1985. They cut a sector off a spare cup, and attached it with screws to the cup used for THR, in order to prevent recurrent dislocation.
When the patients were reviewed at between nine months and three years, it was found that the device had worked well in all of the six cases discussed in that paper.
In 1988, Mogensen et al reported two cases, who were successfully managed with this technique.
In the same year, Graham et al used that technique in three patients; however, in all the patients thus treated the hips continued to dislocate. The failure was due to impingement of the femoral head on the augmentation, causing the head to be levered out of the socket.
In 1989, Gie et al reported on ten augments. Of these, seven remained successful, while three had failed.
In 1990, Nicholas et al described a series of mechanical experiments, and proposed an augmentation made of PE with a metal backing plate. This implant was successfully used by them in three patients. The paper did not, however, give details of the length of follow-up.
In 1991, Watson et al, of the same team at Musgrave Park Hospital, Belfast, presented two cases, with details of follow-up. One device was found to be a success at 18 months; the other had failed, with recurrence of the dislocation with screw breakage at seven months.
The total number of cases described in the literature is only 26, with seven failures 26.90% and 19 successfully managed patients (73.10%).
DISLOCATIONS
INCIDENCE
The dislocation rate varies, with different authors giving figures that range from 0.11% to 9%.
In a control series of 150 28 mm heads inserted, via a posterior approach, between 1982 and 1983, we found our dislocation rate to be 5.11% at between four and six years.
Whilst there is some argument about the merits and demerits of the various approaches, it would appear, from the majority of studies, that the use of a posterior approach harbours a much greater risk of dislocation (5.80% at the Mayo Clinic, according to Woo & Morrey) than do the anterior approaches (2.30%, according to the same authors). We feel that the posterior approach has a number of advantages that tend to outweigh the dislocation risk. However, that is a different subject.
If one considers Charnley (or Charnley-type) devices inserted at different centres by senior surgeons trained in the same techniques, using the same approach, the dislocation rate will be seen to vary within a wide range, from less than 1% to very high rates that had made the surgeons concerned abandon the conventional socket.
CLASSIFICATION
There are three ways in which dislocations may be classified, with interesting results:-
(A) In terms of the event(s) precipitating the dislocation
- Spontaneous dislocation, or true dislocation, occurs following an ordinary activity of daily living, such as getting up from a low seat or out of a car, etc.
- Traumatic dislocation follows a violent blow to the hip. If the patients had not undergone hip replacement, the trauma sustained would have led to fracture of the neck of the femur.
In studies with long follow-up, there will always be cases of traumatic dislocation, unless they are deliberately excluded from the analysis (a not unreasonable step to take by the investigators). Traumatic dislocations must, however, be specified in any study of post-THR dislocation.
(B) In terms of aetiology
P. Fontes, J. Benoit, A. Lortat-Jacob and R. Didry have devised a system that may usefully be applied for the classification of dislocations under this heading:-
- Dislocation following faulty implant placement, most often because of socket malposition, less often because of faulty placement of the femoral component; sometimes, as a result of levering by impingement of the femoral neck on the edge of the cup.
- Dislocation as a result of loss of joint constraint, because of weakening of the periarticular muscles (mainly gluteus medius, but also the short external rotators) compounding the effect of the excision of the capsule. Weakening of the gluteus medius does not necessarily signify wasting of the muscle, but a new and different pattern of muscle function, permitting loss of joint constraint in flexion.
Faulty placement and loss of joint constraint will often be found to occur together.
(C) In terms of time to dislocation
Using the system proposed by J.P. Daly and B.F. Morrey, we may distinguish among three time-frames:-
- Early dislocation - occurring within three months following arthroplasty.
This is by far the most frequent form of dislocation, accounting for 90% of the events seen in Coventry's study.
In Williams' study, 70% of the dislocations occurred within the first post-arthroplasty month.
Ali Khan found a rate of 66% in the first five weeks.
These dislocations are generally due to faulty implant placement, and favoured by postoperative soft tissue relaxation.
Overall, 75% of all dislocations are early ones.
Unless the investigation of the patient yields evidence of major component malposition, these dislocations may be treated, with good prospects of success, by closed reduction and immobilization.
- Secondary dislocation occurring between four months and five years from arthroplasty (9.6% in Coventry's study).
These dislocations are often caused by malposition of the cup, or by abnormalities of the abductors.
Investigations should be directed towards detecting component malposition; management consists in the correction of implant position to prevent an otherwise very probable dislocation event.
- Late dislocation at five years and beyond.
Late dislocation rates vary greatly in different studies.
Thus, Coventry reports 0.40% late dislocations. This figure does not, however, include subluxations and traumatic dislocations.
Daly and Morrey give a rate of 12.60%, apparently excluding traumatic dislocations but including subluxations that cause recurrent disabling snapping during flexion-internal rotation movements. These authors account for late subluxation in terms of progressive stretching of the pseudocapsule, brought on by the inflammation caused by particulate debris.
Our control study of non-augmented cups did not have sufficiently long follow-up to cover late dislocations.
ESTABLISHED TREATMENTS
Dislocations may be managed conservatively, by simple closed reduction; or surgically.
(A) Closed reduction followed by a few weeks of immobilization is justified in early dislocations.
On average, the reduced hip will remain stable in 72% of the cases. The rate found by Ali Khan was 81%; Woo & Morrey observed 65%; while Courtois gives a figure of 70%.
A dislocation that recurs once will recur again in 77% of the cases. Usually, recurrence is due to major implant malposition, which will need to be corrected with surgery. External immobilization after reduction does not statistically reduce the likelihood of recurrence.
(B) Surgical reduction, if need be with a correction of component (usually cup) placement, and sometimes involving tensioning of the gluteal muscles, is a major procedure. There is a high risk of recurrence, since stability is obtained in only 68.60% of the cases, on average. The following figures have been found in the literature: Fraser and Wroblewski, 76%; Woo and Morrey (1982), 69%; Daly and Morrey (1992), 61%.
It should, however, be borne in mind that the statistical results of closed and of surgical reduction cannot be compared with each other, since the two techniques would be used to treat different types of patients, with surgical reduction more often resorted to for the management of recurrent dislocations.
Ali Khan found 21.80% of dislocated prostheses to remain unstable despite all attempts at surgical treatment. This instability may necessitate the removal of the implant. At the time of writing, the rate appears excessively high.
AUGMENTED CUPS
To date, we have inserted 294 augmented cups, using a posterior approach, in patients who had been given 28 mm femoral heads. All the augmented cups have, by now, been followed up for more than four months, i.e. they are all beyond the early dislocation time-frame.
RESULTS
A. Influence of the augment on the dislocation rate
A.1. Treatment of dislocations in patients managed with conventional cemented PE cups, inserted without an anti-dislocation augment.
Eleven augments were inserted after reduction of the dislocation, with a view to preventing recurrent dislocation. All these cases have been follow up for more than five years.
The hip remained stable following augmentation in six cases.
Five cases (45.45%) redislocated. Three recurrences were stabilized with a second augmentation. In two cases, the position of the cup had to be changed.
The rate of immediate success was 54.55%. This contrasts with a mean 72% success rate reported in the literature for simple closed reduction.
Thus, the overall results of augmentation after dislocation, without correction of the cup position in malposition cases, tend to be poor.
A.2. Preventive augmentation
Two hundred and eighty-three augmentations were performed with a view to preventing dislocation.
Initially, augments were used only in high-risk patients (the elderly, patients with neurological conditions, etc.).
Later, we adopted a policy of using augments almost routinely. Nowadays, the decision whether to use an augmentation or a long posterior wall cup is taken on a case-by-case basis.
A.2.1. Preventive augmentation of conventional cemented PE cups (Fig. 1)
Sixty-three augments were added to conventional cemented PE cups, with 28 mm heads. These cases have been followed up for between six and twelve years.
 Fig. 1 Augment on cemented PE cup |
- Early and secondary dislocations: 1.50%
Out to the fifth year, i.e. during the early and secondary dislocation time-frames, the results were excellent, with only one early dislocation, in a very elderly patient with wasted muscles. The femoral head had come out of the cup, deforming the free edge of the bearing piece.
- Traumatic late dislocations: 3.17%
There were two cases of late dislocation, following heavy falls. The events occurred at six years and at nine years, respectively, with breakage of the augment in one case, and pulling out of the screws in the other case. In both cases, the cup was properly placed, and was left in situ. A slightly different position was chosen for the new bearing piece.
Overall, in the cases with long follow-up, no subluxations were encountered. This is what would be expected, since the augment would make it impossible for the femoral head to partly come out of the socket.
A.2.2. Preventive augmentation of the PE insert of cementless cups
Two hundred and twenty augments were added to cementless Neptune sockets, which have a dual-coated (alumina and hydroxyapatite) shell (Fig. 2). The femoral head used was a 28 mm one. The surgical approach was through a posterior incision (Fig. 3).
 |
| Fig. 2 Augment on PE insert of cementless dual-coated Neptune cup |
 |
| Fig. 3 Augment on cementless cup, seen at surgery |
These patients have been followed up for between five months and five years.
The bearing piece and the PE socket are predrilled to accept the screws, with a bearing piece placement pattern that will ensure absence of friction between the head and the augment, under normal conditions.
Contact occurs only if the 28 mm head tries to come out of the socket.
In this series of 220 patients, there were six dislocations (2.72%):-
- Three spontaneous early dislocations (1.36%)
- The first occurred on Day 10, in a female patient suffering from advanced multiple sclerosis. A new, metal-reinforced augment was inserted, and the hip was stable when reviewed one year later.
- The second dislocation happened at two months, when the patient was getting out of a car. A new, reinforced augment was inserted. The hip has been stable for the past four years.
- The third event occurred at 80 days, in a non-compliant alcoholic weighing 100 kg, whose cup was malpositioned. The cup position was corrected, and a new augment was added. Since then, the hip has re-dislocated, for no apparent reason.
- Three secondary dislocations (1.36%)
- One occurred spontaneously, at 16 months, when the patient was getting up.
- Two were associated with overt trauma.
In the three cases, the augment was changed, and the hips have been stable since that revision.
Follow-up in this series has been too short to include late dislocations.
A.2.3. Overall results of preventive augmentation (Table)
| RESULTS OF PREVENTIVE CUP AUGMENTATION | |||
| Early dislocation | Secondary dislocation | Late dislocation | |
| Follow-up | < 3 mo | 3 mo - 5 yrs | > 5 yrs |
| Without augmentation % dislocation |
3.75% | 1.36% | 0.4% - 12.6% in different studies |
| With augmentation % dislocation |
1.14% | 1.06% | 3.17% |
| Causes | all spontaneous |
1/3 spontaneous 2/3 traumatic |
all traumatic |
The records of the 283 patients treated with preventive augmentation, with a follow-up of between five months and 12 years, were available for a statistical analysis of early dislocation, which is by far the most frequent type of dislocation in patients who have not been given a cup augment.
The rate found was 1.41%.
All the dislocations were spontaneous.
The rate given above will not be subject to change in the future, since all the THRs in the series were done more than four months ago.
The rate of secondary dislocations, at between four months and five years, is at present 1.06%.
Of these dislocations, 0.36% were spontaneous, while 0.70% were traumatic.
The rate given above may increase during the time it takes for all the THRs to pass the five-year follow-up mark.
The rate of late dislocations could be established in only 63 cemented cups, which were the only ones to have more than five years' follow-up.
The rate observed was 3.17%. All the late dislocations were traumatic in origin.
Because of this accident-related aetiology, the rate must be expected to rise as time goes by.
(B) Complications
B.1. Bearing piece screw breakage
Among the 294 augments, there were three cases of breakage of the lowermost screw; of these, two were spontaneous, and one followed a fall.
In a fourth case, the two lowermost screws broke when the patient fell out of bed, directly onto his hip (Fig. 4).
 Fig. 4 Traumatic breakage |
These breakages involving one or two screws did not interfere with function.
Apart from the overtly traumatic cases, screw breakage undoubtedly reflects the great stress under which the augments have to work when resisting femoral head dislocation from a malpositioned socket. We saw this in another two cases, where the lowermost screw broke, at 14 months and 20 months, respectively; these breakages led to the two secondary dislocations reported above.
Thus, the overall rate of screw breakage was 2.04%.
B.2. Loosening of screws
In four THR patients (1.36%), partial or total loosening of the lowermost bearing piece screw was seen. In one case, the screw had migrated towards the lower part of the joint.
Function was not affected.
In a fifth patient, however, the lowermost screw had migrated towards the lower part of the joint, six months after the insertion of the augment. One year later, the patient complained of a feeling of "something catching"; radiographs taken at that time showed that the other two screws had partially loosened.
The patient underwent revision three years after insertion. The middle and upper screws were found to be partially loosened, with the bearing piece standing 2-3 mm away from the cup. The bearing piece and the insert were removed, and replaced by a long posterior wall socket.
Overall, we found a 1.70% rate of bearing piece screw loosening.
C. Influence of augmentation on cup fixation
The most frequently asked question is whether cup augmentation affects the fixation of the socket.
In order to answer this question, we carried out a study, which was presented at the 1991 SOFCOT meeting. The study involved 235 cemented implants, with 50 preventive augmentations placed on conventional cemented PE cups. Follow-up was between five years and nine years.
In this study, there were
- fifty-seven stable radiolucent lines, less than 1 mm wide, around 185 non-augmented cups (30.81%);
- eleven stable radiolucent lines, less than 1 mm wide, around 50 augmented cups (22%).
Whilst this result was somewhat surprising, it was statistically significant, the clear message being that cup augmentation does not induce loosening of the socket.
D. Influence of augmentation on PE insert fixation in cementless sockets
In 220 cases, followed up for between five months and five years, no loosening of the PE insert in the Neptune titanium shell was observed.
E. Metal-backed augmentation
In common with Nicholas, we have performed metal-backed augmentation.
Bearing pieces with a metal backing were used in two cases of early dislocation that had occurred despite the use of augmentation.
In early dislocations, the screws have not been cyclically stressed for long enough to suffer fatigue failure. In such cases, the femoral head comes out of the socket deforming the edge of the bearing piece; this may be prevented by reinforcing the bearing piece with a metal backing plate (Figs. 5 and 6).
 |
| Fig. 5 THR for pertrochanteric fracture in OA, in an MS patient. Non-reinforced preventive augment did not prevent dislocation. New augmentation with metal-backing. |
 |
| Fig. 6 Metal backing of bearing piece, seen at surgery |
Both cases were successfully managed with this technique.
F. Augmentation and Hardinge approach
A limited number of Hardinge approaches performed outside this study had shown us that this approach does not lend itself to the treatment and the prevention of dislocation by the use of cup augments.
If the augment is posterior, dislocation will be anterior, because of the weakness of the anterior muscles and the cam effect of the augment.
If the augment is anterior, it will interfere with flexion.
The only correct site for the augment, therefore, is on the superior portion of the socket.
DISCUSSION
From our analysis of 294 cup augments inserted through a posterior approach, in patients who had been given 28 mm heads and followed up for between five months and 12 years, the following conclusions may be drawn:-
The results of cup augmentation to prevent dislocation were poor if pre-existing malposition of the cup was allowed to persist uncorrected.
Thus, the object of the procedure is not simply the fitting of an augment onto a reduced implant. However, once the cup position has been corrected, a bearing piece may be fitted.
Augmentation as a means of primary prevention did, however, appear to be beneficial.
Overall, the rate of dislocation (all types) dropped from 5.11% to 2.47% over comparable periods of time, from insertion to five years' follow-up.
A. As a function of precipitating factors
The rate of spontaneous (or true) dislocations was 1.76%. The rate of traumatic dislocations was 1.41%.
The augments are not sufficiently soundly fixed to withstand violent trauma of a degree of severity that would cause fracture of the femoral neck in non-implanted hips.
B. As a function of aetiological factors
While minor degrees of implant malposition may be successfully managed with cup augmentation, this technique will not be a safeguard against dislocation where the placement is seriously wrong. In such cases, dislocation may occur spontaneously, during a forced movement or, often, in a context of trauma.
In patients with severe neurological conditions, such as Parkinson's disease or multiple sclerosis, metal-backed augmentation may be effective.
C. As a function of time
Cup augmentation was found to decrease substantially the rate of early dislocations, which, in our patients, always were true, spontaneous dislocations.
The rate in augmented cases was 1.41%, as against 3.75% in the non-augmented patients.
Augmentation did not affect the rate of secondary dislocations in any major way: The rates were 1.06% with, and 1.36% without cup augmentation. This finding is accounted for by the fact that two out of three secondary dislocations were traumatic in origin.
Augmentation is unlikely to have a major effect on the rate of late dislocations, which are always caused by trauma.
With the use of cup augmentation, traumatic dislocations become the most frequently observed dislocation pattern following THR.
CONCLUSION
Cup augmentation is not the answer to all our prayers.
Where augments are used, from a posterior approach and with 28 mm heads, they do, however, halve the risk of dislocation from all causes, without producing more cup loosening.
Cup augments or long posterior wall cups?
The weakest point of an augment is the screw sites. Bearing pieces are, however, better able to contain the femoral head in the early post-arthroplasty period, while the soft tissues are healing.
At present, it is our policy to use long posterior wall devices for younger subjects, and for patients whose muscles are well preserved; equally, the use of these cups is predicated upon proper placement of the implant components. Cup augmentation, on the other hand, is still used in the elderly and/or if the THR is considered likely to dislocate.
Further comparisons and statistical analyses will need to be performed, to establish more clearly the indications of either technique.
|
Bibliographie
Charnley J., Cupic Z.
Courtois B., Variel R., Lesaout J.
Coventry MB.
Daly PJ., Morrey BF.
Falker CD., Poss R.
Fontes D., Benoit J., Lortat-Jacob A., Didry R.
Fontes D., Benoit J., Lortat-Jacob A., Didry R.
Fraser GA., Wroblewski BM.
Gie A., Scott T., Ling RSM.
Graham GP., Jenkins AIR., Mintowt CZYZ W.
Khan MAA., Brakenbury PH., Reynolds ISR.
Mogensen B., Arnason H., Jonsson GT.
Nicholas RM., Mollan RA., Calderwood J., Nixon JR., Watson P.
Olerud S., Karlstrom G.
Watson P., Nixon JR., Mollan RAB.
Williams JF., Gottesman MJ., Mallory TH.
Woo RYG., Morrey BF.
|