12. Segmental aponeurectomy: early results

The results of our first 213 segmental aponeurectomies have already been published (Moermans, 1991). An update of the results with some additional comments and remarks is shown below.

12.1 The patients

We have now performed 329 segmental aponeurectomies on 268 patients: 63 women (23.5%) and 205 men (76.5%). In 14 more cases (4.1%) a segmental aponeurectomy was planned but could not be carried out because there was too much skin shortage or because we considered that the dissection of the neurovascular pedicles would be too risky. A classical limited fasciectomy with Z-plasties or a dermofasciectomy was performed instead.

The average age was 61.9 years (range 24 to 87). A family history of the disease was reported by 28.4;% of the patients. The unreliability of this information has already been discussed. Among the patients, 8.3% were alcoholics, 3.1% were under treatment for diabetes and 5.1% had been or were under treatment for epilepsy. In 10.6 % of the cases there was a history of local trauma. These proportions are almost the same as those noted in 1991 except for local trauma which, for unknown reasons, was clearly lower (6.7%) then.

43 (13.1%) of the 329 hands had undergone a previous operation on the same or other rays (15 or 4.6% showed an extension of the disease, 13 or 4% a recurrence and 15 or 4.6% a combination of the two). Those numbers are noticeably different from those of 1991 and testify to a change in my indications for segmental aponeurectomy. Actually, during the last three years, I have been avoiding this procedure for hands that have already been operated on before for two reasons:

To evaluate the results, the pre-and post-operative mobility of the M.P. and I.P. joints was measured. We also used the assessment formula proposed by Tubiana et al. (1968) and the impairment-of-function assessment of Swanson (1983), which is useful to show the functional impact of some possible complications such as loss of flexion.

12.2 Preoperative evaluation

The contractures ranged from 5° to 95° at the M.P. joints and at the P.I.P. joints from 5° to 110°. There was a contracted aponeurotic band in the first web in 40 hands. Detailed data are presented in table 1.

Table 12-1: Flexion contracture before and after operation (in degrees)

The average stage using Tubiana's formula was 3.0 (S.D. 2.0, range 1 to 13). Figure 1 shows the details of the distribution. As can be seen, all operated hands had some degree of contracture. None was operated at a pure nodular stage.

Figure 12-1: Preoperative Tubiana's grade

The average impairment of function was 6.1 % (S.D. 5.6, range 1 to 35). The distribution is shown in figure 2. As can be seen, in about 50% of the cases, the impairment of function was 4% or less, in 80% it was 9% or less.
Figure 12-2: Preoperative impairment of function


12.3 Postoperative evaluation

All patients were seen regularly until they had recovered a complete range of motion or we felt that the mobility would not improve any further. Most patients (fig. 3) were under treatment for less than 6 weeks (range 1 to 30, mean 5.4, median 4).
Figure 12-3: Duration of treatment in weeks


12.3.1 Complications

Eighteen complications (5.6%) were observed in 16 cases (4.9%).

12.3.1.1 Haematomas

Six small haematomas (1.8%) developed. One gave rise to a limited skin necrosis. There was no adverse consequences in the other five since they engendered neither infection nor skin necrosis.

12.3.1.2 Skin necrosis

A small area of skin necrosis was observed in three cases (0.9%) which were under treatment respectively for 4, 7 and 8 weeks. The wounds healed with local treatments and did not require skin flaps or skin grafts.

12.3.1.3 Infection

One patient (0.3%), a heavy drinker who was lost for follow-up, developed an infection.

12.3.1.4 Nerve lesions

Two digital nerves were sectioned at the beginning of this series (cases 21 and 30) when I was still using scissors to divide the aponeurotic bands. Three more nerves were sectioned later. These injuries were identified during the operation and the nerves immediately sutured (1.5%).

12.3.1.5 Sympathetic dystrophy

Three patients (all men) developed a severe sympathetic dystrophy (0.9%). They were treated for many weeks (17, 20 and 30). One of them was the only patient in this series whose function was worse after the operation than before.

12.3.1.6 Loss of flexion

Beside the three patients cited above whose stiffness is taken into account in the impairment of function evaluation, all patients regained at least the flexion they had before the operation.

12.3.2 Results

12.3.2.1 Contracture

Following operation, extension at the M.P. joints ranged from 0° to 35° and the P.I.P. joints from 0° to 90°. A contracted band was still present in the first web of one hand. The detailed data and the comparison with the pre-operative values are shown in table 1. They confirm the typical better prognosis of the M.P. contractures, even in the radial rays, and the less satisfactory results of the P.I.P. joint of the little finger (fig. 4, tables 2 and 3). The I.P. joint of the thumb was involved in only two cases. The results are thus not meaningful for this joint.
Figure 12-4: Percentage of improvement of the contracture by ray and by joint


Table 12-2: Improvement of contracture by ray and by joint level


Table 12-3: Improvement of flexion contracture


The mean improvement was 97.9% (S.D. 10.3, range 0-100) at M.P. level and 75.4% (S.D. 35.5, range -200-100) at the P.I.P. level.

No statistically significant difference was found between the M.P. joints of the five rays (p<0.3 between the first and third rays that show the greatest difference between the improvement means; statistical procedure: Mann-Whitney test.).

The prognosis of the P.I.P. joint of the little finger is significantly worse than in the other rays (p<0.01 between the fourth and fifth rays). The improvement in the P.I.P. of the index finger is clearly poorer than in the adjacent middle finger but the observed difference is not statistically significant (p<0.4).

12.3.2.2 Tubiana's grade

The mean post-operative Tubiana's grade was 0.5 (S.D. 0.9, range 0-6). The mean improvement was 84.2% (S.D. 24.5, range 0-100). As shown in figure 5, 62% of the hands had a complete re-extension (Tubiana's grade 0), and 90% of the cases had a Tubiana's grade inferior or equal to 1 (total extension contracture less than 45°).
Figure 12-5: Pre and postoperative Tubiana's grade


12.3.2.3 Impairment of function (fig.6)

The mean post-operative impairment of function was 0.9% (S.D. 1.9, range 0-14). The mean improvement was 88.4% (S.D. 20.6, range -50 to 100).
Figure 12-6: Pre and post-operative impairment of function


As stated before, one patient, who developed a severe sympathetic dystrophy, lost some function because of the operation and had a negative improvement.

This confirms the benefit of assessing the impairment of function since it is the only measure that reveals that the condition of this patient was aggravated by the operation.

12.4 Factors determining the outcome

I have tried to identify the factors that could play a role in the outcome of the operation.

12.4.1 General factors

I have studied the relations between the improvement of the hand condition, measured by the Tubiana's grade and the impairment of function, and several variables: age and sex of the patients, previous operation, ectopic sites of the disease, involvement of both hands and personal history.

12.4.1.1 Age of the patient

As can be seen (fig. 7), no relation exists between the age of the patient at the time of the operation and the improvement of the contracture.
Figure 12-7: Plot of age of the patient with percentage of improvement of Tubiana's grade


12.4.1.2 Sex, previous operation, ectopic sites of the disease and involvement of both hands

 The sex of the patient and the involvement of both hands do not play a role in the short term outcome of the operation (table 4). On the contrary, patients who had already been operated before had a significantly less satisfactory improvement of the contracture and of the function. This observation, the hazards of a blind dissection in presence of scar tissue and the likelihood of further recurrences, as indicated earlier, made me avoid segmental aponeurectomy in previously operated hands.

Table 12-4: Relations between the postoperative improvement of the contracture and of the flexion and sex, previous operation, ectopic sites of the disease and involvement of both hands (statistical procedure: Mann-Whitney test)


Patients who had ectopic sites of the disease had also a less satisfactory evolution. This is another indication of the value of this factor as an index of the severity of the disease.

12.4.1.3 Personal history

Table 5 shows that patients who reported a local trauma had a slightly better improvement than the others and that epileptic patients have a worse prognosis. Those differences are nevertheless not significant.

Table 12-5: Relations between personal history and the post-operative improvement of the contracture and of the function (statistical procedure: Mann-Whitney test)


12.4.1.4 Complications

No statistically significant relation was found between the factors mentioned above and the existence of complications.

12.4.2 Local factors

As we have seen, there is great variation in the results obtained at operation on the P.I.P. joint, so it is likely that many variables affect the correction of contracture at this joint. Moreover, the post-operative improvement of the P.I.P joints of the ring and little fingers are clearly different. The frequency of their involvement by the disease makes a study of the local factors that could have a prognostic value particularly interesting.

12.4.2.1 Pre-operative contracture at the P.I.P. level

It seems obvious that the degree of contracture will affect the extent of the correction but in the proximal inter-phalangeal joints of the fourth and fifth fingers, the pre-operative angle was oddly lowest in the group with the worst outcome. In all joints in which the outcome was estimated as same or worse (improvement negative or equal to zero), the pre-operative angle was less than 30° (table 6).

Table 12-6: Relation between the preoperative P.I.P. contracture and its improvement.


Figure 8 summarises the clear difference in the pre-operative contractures of the little finger observed in the three improvement groups. The same graph could be shown for the ring finger. The observed differences are statistically highly significant (p<0.0001; statistical methodology: Kruskal-Wallis one-way analysis of variance).
Figure 12-8: Boxplot of outcome by pre-operative contracture of the P.I.P. joint of the fifth finger. The black square in the clear rectangle represents the median. The lower border of the rectangle represents the 25th percentile, the upper border the 75th percentile and so 50 % of the cases have values within the box. The lower and upper extremities of the vertical lines represent respectively the 5th and 95th percentiles.


These observations deserve some discussion. They suggest that joint contractures of less than 30° are best not operated upon. The reason for the poor results is not clear. It may be that with such contractures the diseased fascia was not clearly visible. A well developed cord may not have been present and one may not have removed sufficient tissue. As a result the contracture was not corrected and the residual disease, augmented by the post-operative scarring, caused further joint contracture.

On the other hand, most cases which were improved but not perfect had a pre-operative contracture greater than 40° whereas almost 50% of those with perfect results had contractures between 30° and 40°.

Hence we can conclude that the pre-operative contracture of the P.I.P. joints indeed plays a role in the quality of the final correction. Schematically we can say that the results are usually bad under 30°, suddenly improve to be perfect between 30° and 40° and then worsen again above this value.

12.4.2.2 Pre-operative contracture at the M.P. level

Using the same statistical tests, it is possible to show that the degree of contracture at these joints had no effect on the outcome at the proximal interphalangeal joint.

12.4.2.3 Total extension deficit

The P.I.P. contracture of respectively the ring and little finger was substracted from the total extension deficit to verify if the contracture of the other fingers played a role in the extent of correction. The relation for the little finger is summarised in figure 9. The graph would be approximately the same for the ring finger. The observed difference is not significant (p<0.2; statistical methodology: Kruskal-Wallis one-way analysis of variance.).
Figure 12-9: Relation between outcome of the P.I.P. joint and the total extension deficit exclusive of this joint.


12.4.2.4 Number of involved rays

The relation between the number of involved rays and the outcome is summarised in table 7 and figures 10 and 11. In both fingers, the cases with perfect results had less involved rays. This difference is statistically significant (p<0.003 for the little finger and p<0.02 for the ring finger; statistical methodology: Kruskal-Wallis one-way analysis of variance).

Table 12-7: Relation between the number of involved rays and the outcome at the P.I.P. joint


Figure 12-10: Relation between the outcome of the P.I.P. joint of the ring finger and the number of involved rays


Figure 12-11: Relation between the outcome of the P.I.P. joint of the little finger and the number of involved rays

12.4.2.5 Presence of an ulnar cord in the little finger

In the little finger, a cord coming off the adductor digiti minimi is often found on the ulnar border of the finger. Its dissection is usually more difficult than that of the pre-tendinous cord. It could be the explanation for the worse prognosis of the P.I.P. joint of this finger.
Table 12-8: Relation between the presence of an ulnar cord and the outcome at the P.I.P joint of the fifth finger


The relation between the improvement and the presence of an ulnar cord as seen at the time of operation (table 8) was tested but no statistically significant difference was found (statistical methodology: Pearson's Chi square).

12.5 Discussion

Several aspects will be discussed separately. First we will discuss the operative results and compare them to those published in other studies and then we will examine the factors that determine the outcome of the operation.

12.5.1 Operative results: comparison with other studies

Segmental aponeurectomy was developed as a possible way to achieve complete correction of the contracture while trying to limit the post-operative discomfort and the sequelae of wide dissection. The results of surgery for Dupuytren's contracture have been assessed by a number of different methods and genuine comparison between procedures is extremely difficult. It is all the more so as many factors influence the results: the great variety in the severity of the disease, the operation performed and the post-operative complications, especially haematoma, skin necrosis nerve lesions and sympathetic dystrophy. The most complete study on this subject is the multicentric review published by McFarlane et al. (1990b). This review will be used as a comparison basis because it is the most extensive published to date, because they splitted the study by geographical area so that we can compare identical populations and because they studied almost the same parameters as we did.

The results achieved in these 329 segmental aponeurectomies are equal or even better than those published by McFarlane for Northern European countries and for other techniques even though he only considered patients operated for the first time.

From table 9, we can see that the patient profiles in McFarlane's review and in this study are almost the same except for the sex ratios which were already discussed in the previous chapter and for the fact that, in this study, 13.1 % of the cases had already been operated before. As we have already established before, we are studying the same population of patients.

Table 12-9: Comparison of patient profiles


Table 10 summarises the pre and post-operative contractures in both studies. Once again the differences in the pre-operative values are insignificant.

Table 12-10: Comparison of hand profiles. The table shows the mean contractures by rays and by joints (SD)


If we compare the outcome of the operations (table 11) we see big differences: the proportion of perfect results is much lower in McFarlane's review, mostly for the PIP joints, and there are clearly more cases ranked as same or worse.

Table 12-11: Comparison of outcome of operations


McFarlane reports an overall complication rate of 17 % (McFarlane et al., 1990c). This must be compared with 4.9 % in this series as above-mentioned. The comparison of the types of complications (table 12) is interesting. We must note that one complication often leads to another so there are fewer patients involved than complications listed. The percentage of nerve injuries is the same in both series even though the exposure of the neurovascular bundles is much smaller in segmental aponeurectomies. The percentages of infections, haematomas, skin losses, sympathetic dystrophy and loss of flexion are all greater in McFarlane's review. Infection may occur after skin loss and is likely after haematoma. All these complications are related to the skin undermining and the creation of dead spaces in the hand palm that I tried to avoid by limited incisions.

Table 12-12: Comparison of complications


In a further analysis of the results of treatment according to the type of operation performed, McFarlane (1990) concludes that "In patients having a local operation, the pre-operative contracture at the fifth metacarpophalangeal joint and the postoperative correction were greatest.
Also, a perfect outcome at the fifth proximal interphalangeal joint was more frequent after a local operation both in the palm and in the finger".

12.5.2 Factors determining the outcome

12.5.2.1 Localisation of the disease

The localisation of the disease is by far the most important of the factors that play a role in the final outcome after operation. At the metacarpophalangeal joint, as we have seen, the results are uniformly good whatever the digital ray. The same observation was made by McFarlane et al. for all types of operations in their multicentric survey. This joint cannot be used to evaluate methods of treatment.

In contrast, the results at the proximal interphalangeal joint varied with affected fingers. Globally, the prognosis at the PIP level is worse than in any MP joint and the PIP of the little finger has the worst prognosis of all. Those differences are statistically significant as we have seen earlier. The same differences are observed with all types of operations at various degrees (McFarlane et al., 1990b).

12.5.2.2 Sex of the patient and involvement of both hands

These variables do not contribute to the final result as has been previously shown. McFarlane drew the same conclusion.

12.5.2.3 Previous operation

This factor plays, as we have seen, an important and statistically significant role in the correction of the contracture. The improvement of the contracture was 86.4 % in patients who were never operated before and only 69.8 % in the others. The abundance of scar tissue is certainly the underlying factor which cannot be corrected by a limited approach. This would explain why McFarlane, who reviewed more extensive operations, did not isolate a previous operation as a variable contributing to the outcome.

12.5.2.4 Ectopic sites

This factor plays also a significant role in the correction of the contracture. As we have seen in the chapter devoted to the epidemiology of the surgical patients, the presence of ectopic sites of the disease is certainly a good index of a more aggressive disease as it is positively related to a greater extension deficit, more involved rays and more involvement of the radial rays. This is another example of the value of this factor as an index of a stronger diathesis.

12.5.2.5 Personal history

McFarlane found that alcoholism had a negative effect on the correction of the contracture. In this series, this factor does not seem to play a role but epilepsy was associated with a worse prognosis and a history of local trauma with a better correction of the contracture. The observed differences are nevertheless not significant probably because of the small number of cases.

12.5.2.6 Timing of operation

Metacarpophalangeal joint contracture can almost always be corrected, but because PIP joint contracture is so difficult to correct, it has been the common view that the patient should be advised to have an operation as soon as the proximal joint begins to flex. We should modify this view in the light of what we have seen about the role of the pre-operative contracture on the outcome: the results are usually bad under 30°, suddenly improve to be perfect between 30° and 40° and then worsen again above this value. This has practical consequences. First, one is cautioned against operating upon early disease. It would seem best to wait until 30-40° of contracture is present. This was already known for the MP joints but is new for the PIP joints. Second, for the same reasons, minimal disease elsewhere in a hand that requires an operation is best left alone. This means that if a little or ring finger is operated upon, minimal disease often present in the other fingers should be left. The disease may not progress in those areas or the contracture can be made worse by removing it.

12.5.2.7 Metacarpophalangeal joint contracture

Legge and McFarlane (1980) have suggested that the degree of metacarpophalangeal joint contracture influenced the proximal interphalangeal joint result. This hypothesis has been tested with our data but no relation was found. McFarlane et al. (1990b) have tested his observation with the data collected in their multicentric survey and, with a much larger group of cases, found no relation either. We can conclude that the degree of contracture at the MP level has no bearing upon the outcome at the proximal interphalangeal joint.

12.5.2.8 Number of involved rays

As we have seen earlier, the number of involved rays plays a negative role on the outcome at the PIP joints of the ring and little fingers. There is no obvious explanation for this observation. The most likely one is that if there are more operated rays the postoperative swelling and pain are increased and active mobilisation is delayed.

12.6 Conclusion

The comparison of the results achieved with segmental aponeurectomies with those published for other procedures supports our reliance on this less extensive and less aggressive surgical technique. Indeed, the correction of the contracture is equal and often better than that obtained with classical operations and the overall complication rate is much lower.

The analysis of the variables contributing to the outcome of the operation allowed us to isolate some important factors: localisation of the disease, previous operation, ectopic sites and number of involved rays. It also allowed us to determine the best timing of the operation.

Those factors are the same as those identified for other procedures and we are there approaching the reality of the situation that Hueston (1982) summarised by saying: "Fundamentally the patient produces the disease. The surgeon attempts to control it".