4. Anatomy of the pathological aponeurosis
Dupuytren's disease induces the development of nodules, dermal pits, deformities of the palmar creases, cords and joint stiffness. In spite of the multiple associations and permutations of lesions that can be observed, their distribution does not happen haphazardly but on the contrary can be explained on the basis of pathological shortening of the normal fascial structures (McGrouther, 1982, 1985, 1990a; McFarlane, 1974, 1985b).
4.1 The palmar localisation
The affected structures are the palmar aponeurosis, the transverse fibres, the natatory ligaments and the aponeurosis of the hypothenar muscles.
4.1.1The palmar aponeurosis
The palmar aponeurosis is almost always involved in Dupuytren's disease.
Figure 4-1: The retraction of the superficial fibres of the pretendinous band is responsible for the development of nodules and dermal pits (from McGrouther 1990)
The superficial pretendinous bands that insert in the deep dermis approximately at a point midway between the distal palmar crease and the proximal digital crease are responsible for the development of nodules and dermal pits as well as for the contracture of the metacarpophalangeal joints (fig. 1). As we have seen previously, the pretendinous band for the thumb is not well individualised and that for the index finger is essentially inserted on the radial border of the palm. This explains why a MP contracture in those two rays is seldom seen and usually not significant.
For the surgeon some considerations derived from the anatomy are important:
- the palmar nodules are frequent; they have a defined positioning, proximal to the transverse fibres or distal to the distal palmar crease on the dermal insertion of the pretendinous bands;
- the distal nodule is thus always intimately adherent to the dermis and difficult to dissect; the risk of devascularisation of the small skin flaps that would be elevated at that level is high;
- the proximal nodule is not adherent to the skin and its dissection is easy;
- the dissection of the pretendinous bands is easy since they do not displace the neurovascular bundles when they retract;
- the prognosis after correction of a metacarpophalangeal joint contracture is thus good; this will be confirmed by the analysis of the operative results that will be presented.
The involvement of the intermediate and deep layers of the pretendinous fibres will be studied with the digital localizations.
4.1.2 The transverse fibres
As has been shown by Skoog (1967, 1985), these fibres are never involved and they should be preserved during the surgical correction. The extension of the transverse fibres to the thumb and the first web (proximal commissural ligament) is sometimes involved and its excision can be indicated.
4.1.3 The natatory ligament
They are often involved. Their retraction closes the web space and restricts the spreading of the fingers. This can lead to a significant impairment of function in certain occupations. For example, I had to operate a pianist, an organist and a harpsichord player for a restriction of the spreading out of the fourth and fifth fingers at an early stage of the disease.
The fibres of the natatory ligaments contribute to the formation of the lateral digital sheet. Through this, their retraction can lead to a contracture of the proximal interphalangeal joint.
The radial extension of the natatory ligament to the thumb and the first web (distal commissural ligament) can also be involved.
The complex interactions between the ulnar fibres of the natatory ligament and the abductor digiti minimi have been described previously. Their meeting point is often the seat of a voluminous nodule.
4.1.4 The hypothenar fascia
This fascia can sometimes be palpated particularly in presence of a nodule on the basis of the little finger.
4.2 The digital localisations.
Four cords play a role, alone or in association, in the contracture of the proximal interphalangeal joint. Each one originates in fascia that is normally present (McFarlane, 1985b). These normal fascial structures are normally present on both sides of the finger but usually only one side is affected by the disease. The retracted cords displace the vasculonervous bundles and a good comprehension of the anatomy is thus essential to avoid injuries during surgery.
4.2.1 The central cord
It is a direct extension of the pretendinous cord in the palm that remains directly subcutaneous. It lies between the neurovascular bundles. Over the proximal phalanx the cord is intimately attached to the skin but not to the tendon sheath. It attaches to the tendon sheath and periosteum of the middle phalanx, usually on one or other side but a symmetrical attachment is seen occasionally.
Some fibres of Grayson's ligament are almost always involved by the retraction of this cord. They attract the neurovascular bundle in a more central position.
The central cord is the commonest cause of proximal interphalangeal joint contracture. Its presence can usually be detected before a surgical correction.
 |
Figure 4-2: The schema on the left shows the four normal components of a spiral cord. 1 = pretendinous band; 2 = spiral band; 3 = lateral digital sheet; 4 = Grayson's ligament. On the right, as these structures shorten and form a cord, the neurovascular bundle is displaced to the midline (from McFarlane, 1985) |
4.2.2 The spiral cord
The spiral cord arises either from the intermediate layer of the pre-tendinous fibres of the palmar fascia or from the tendon of abductor digiti minimi (fig. 2). In both cases, it runs behind the neurovascular bundles just distally to the metacarpophalangeal joint and joins the lateral digital sheet. From there it extends in front of the neurovascular bundle and inserts on the flexor sheath of the middle phalanx through Grayson's ligament. The spiral cord thus winds around the neurovascular bundle, hence its name. It usually causes severe proximal interphalangeal joint contracture.
4.2.3 The lateral cord
It is an involvement of the lateral digital sheet. The natatory ligament is almost always involved except on the ulnar border of the little finger. This cord attaches primarily to the skin and generally does not cause severe proximal interphalangeal joint contracture. The cord does not disturb the neurovascular bundle except that its bulk can push the bundle towards the midline.
A distal extension of the lateral cord can cause distal interphalangeal joint contracture.
4.2.4 The retrovascular cord
This cord was described by Thomine (1965, 1985). Its existence as an element distinct from Cleland's ligament is debated. It is a condensation of longitudinally oriented fibres behind the neurovascular bundle running between that bundle and Cleland's ligament. It would play no role in a proximal interphalangeal contracture but could be responsible of a distal interphalangeal joint deficit of extension.
4.3 Peculiar characteristics of each finger
The general description which has just been given applies well to the middle and ring fingers. The index and little fingers show some specific features.
4.3.1 The index finger
It is not very often involved and its contracture is somewhat different. A contracture of the MP joint is rarely present. The contracture of the PIP joint is usually not very important and is rarely due to a central cord. Generally the cord develops in continuity with the distal commissural ligament that joins with the radial digital sheet.
4.3.2 The little finger
Variations are most frequent in this finger. Whereas the MP contracture is only due to the pretendinous fibres, all the other cords can play a role in the retraction of the PIP joint. This is probably why the correction of the contracture is, as a rule, less satisfactory in this finger: the multiplicity of the involved fibres prevents to treat them all.
4.4 The dorsal localisations and knuckle pads
Swellings over the finger joints or knuckle pads are often found in association with Dupuytren's disease. They have been associated with a more aggressive course of the disease as well as a greater tendency for recurrence after operation (Hueston 1985c, 1990a). Others (Mikkelsen 1977) do not share this view.
Other, more subtle, knuckle changes can be found in patients with Dupuytren's disease. McGrouther (1990b) demonstrated that there is a spectrum of changes from normal through skin thickening to a palpable nodule. This may explain the wide differences in the reported incidence of pads.
The appearance of dorsal lesions could possibly be explained by the fascial continuity between palmar structures and the dorsal wrinkle skin over the proximal interphalangeal joint described by Lawet al. (1984) as the lateral peritendinous cutaneous fibres.
Knuckle changes can be observed even in non-contracted fingers and Dupuytren's disease seems therefore to be a more widespread affliction of the connective tissue than is immediately apparent from the study of the palm of the hand (McGrouther 1990b).
4.5 Conclusions
As we have seen, all the finger contractures that occur in Dupuytren's disease can be explained on the basis of the anatomy of the palmar fascia and of its extensions. It seems that no cord appears outside the normal pre-existing anatomical structures. For example, the pathognomonic nodule is firmly adherent to the deep dermis only if the fascia in which it develops has cutaneous insertions. However, all the parts of the palmar and digital aponeurosis are not equally involved by the pathological process. For instance, the aponeurosis of the thenar muscles is never involved whereas that of the hypothenar muscles is often invaded by the disease.
All the affected tissues have the same embryological origin namely the superficial layer. The pretendinous fibres, the natatory ligaments and Grayson's ligaments are inserted both on the skin and on the tendons. The deeper layer made up of the transverse fibres, the flexor retinaculum and the tendon sheaths has no cutaneous insertion and is never involved. The exception to this rule is the most radial part of the transverse fibres in the first web that has some cutaneous insertions and that often develops a retraction. The deepest layer made up of the intermetacarpal transverse ligaments and of Cleland's ligaments is never involved.
The fact that only the aponeurotic bands that have attachments in the deep dermis are affected by the disease is in agreement with the observation that recurrences are frequent after fasciectomy if the skin over the lesions is preserved whereas they are exceptional if the skin is excised and replaced by a skin graft (Hueston 1962, 1985d). This had led Hueston to suggest that the starting point of the disease lays outside the aponeurosis in the subcutaneous fibro-fatty lobules.
Does one have to look for the cause of Dupuytren's disease in the embryological origin of some layers of the palmar fascia as the current anatomical approach of the distribution of contractures suggests or is it a pathology of the deep dermis as the excellent results on recurrences by Hueston's skin excision and grafting imply? Could it be more simply a response to mechanical stimuli which, understandably, are more frequent and severe in the fibres attached to the skin?
The first hypothesis stresses the role of heredity, the second one sees Dupuytren's contracture as an acquired disease. The last one suggests a multifactorial origin: an abnormally intense response could be triggered by normal mechanical stimuli in predisposed individuals. The predisposition could be hereditary or drug induced for example.
To make a synthesis between these different approaches we need a better knowledge of the patho-physiological mechanisms involved in the appearance and development of the disease.