Madelung Deformity

An important anatomic consideration in MD is the normal position of the distal radial articular surface. The following four features of this articular surface should be noticed radiographically:

• Radial inclination - This is the angle formed by a line from the distal radioulnar joint (DRUJ) to the radial styloid and a line perpendicular to the shaft of the radius through the lunate fossa; it is normally 21-23°
• Radial length - This is the difference in longitudinal level between the lunate fossa and the radial styloid; it is normally 12-15 mm
• Volar tilt - This is measured on a lateral radiograph and is the angle formed between a line perpendicular to the radial shaft and a line through the dorsal and volar rims of the radial joint surface; normally, it is 10-15°
• Ulnar variance - This is the relative difference in height between the radial and ulnar distal articular surfaces; these should be level with each other

One third of cases of MD are transmitted in an autosomal dominant fashion. The condition has a variable expression and 50% penetrance. MD is bilateral in 50% of cases and is primarily found in females. A number of affected kindreds of patients with MD have been described. Numerous cases of multiple patients with MD within families have been reported.[3, 4] One report detailed five generations of family members with bilateral MD without signs of dyschondrosteosis.

Finally, a primary chromosomal association with MD has been observed in patients with Turner syndrome (karyotype XO). MD was the presenting sign of Turner syndrome in one young girl, while Henry and Thorburn diagnosed Turner syndrome incidentally when studying a series of patients with MD who underwent cytogenetic evaluation.[5]

Molecular genetic studies clarified the association of the female predominant MD, dyschondrosteosis, and the missing X chromosome in Turner syndrome. The idea that an X chromosomal translocation caused dyschondrosteosis was first proposed in 1985. An X chromosomal translocation also was found to be associated with MD in 1997.

In 1998, groups from London, England, and Switzerland established a marker that was linked to the pseudoautosomal region (PAR1) of the X and Y chromosomes. Within families affected by a short stature dysplasia, the groups found deletions and a premature stop codon (exon 4) in the short stature homeobox-containing gene, SHOX, that segregated the marker in band Xp22.[6]

In 2000, another group reported that the SHOX gene mutation was found in patients with dyschondrosteosis and MD in multiple cases. Families with this mutation and individuals with Turner syndrome (both essentially hemizygous individuals for the SHOX gene) and families with a history of MD have been shown to exhibit a variable expression of MD and dyschondrosteosis. This variability indicates that a modifier gene on another area of the X chromosome or on an autosomal gene most likely is involved also.[7] The interaction of the SHOX locus with other genes and transcription factors is not known. The SHOX mutation continues to be the subject of many studies.[8, 9]

Henry and Thorburn classified MD into the following four etiologic groups[5] :

• Posttraumatic
• Dysplastic
• Chromosomal or genetic (Turner syndrome)
• Idiopathic or primary

Posttraumatic MD has been found following repetitive trauma or following a single event that disrupts growth of the distal radial ulnar-volar physis. Bone dysplasias associated with MD include the following:

• Multiple hereditary osteochondromatosis
• Ollier disease
• Achondroplasia
• Multiple epiphyseal dysplasias
• Mucopolysaccharidoses (eg, Hurler and Morquio syndromes)

Secondary causes of wrist deformity that may mimic MD include the following:

• Sickle-cell disease
• Infection
• Tumor
• Rickets

The most important dysplasia associated with MD, however, is dyschondrosteosis.[10, 11, 12, 13]

Dyschondrosteosis is a form of mesomelic dwarfism associated with MD that was first described by Leri and Weill in 1929. It is characterized by variable short stature, short forearms, and tibial/fibular shortening. Specifically, height is less than the 25th percentile, the radius is 75% of the length of the humerus, and the tibia is 85% of the length of the femur. Dyschondrosteosis becomes more clinically pronounced during adolescence. No other abnormalities are commonly associated with it.

Forearm shortening in dyschondrosteosis tends to be bilateral and appears almost identical to that in primary MD, except that the proximal radius is involved in patients with dyschondrosteosis (see the first image below), and the proximal radius is not involved in primary MD (see the second and third images below). Both dyschondrosteosis and MD are transmitted in an autosomal dominant fashion with a female predominance.[14]

With such a similarity between primary MD and Leri-Weill dyschondrosteosis, the reason why many authors have described them as a single entity is apparent. However, many children with unilateral and bilateral MD have normal stature and no other characteristic of dyschondrosteosis. Therefore, it is reasonable to describe primary MD as a separate, albeit related, condition.

To differentiate primary MD from dyschondrosteosis, Felman and Kirkpatrick suggested the following criteria[15] : Primary MD can be defined as occurring in a child above the 25th percentile in height with no family history of dyschondrosteosis and no history of other secondary causes. Conversely, a patient who is shorter than 5 ft (~1.5 m) at skeletal maturity, has proximal involvement of the radius, and has a relatively short tibia and fibula may have mesomelic dwarfism.

Several hundred cases of MD have been presented in the literature since its first description. However, no published reports exist on the actual frequency of MD in the population.[16]

Goals for surgical correction of MD consist primarily of pain relief and correction of the cosmetic deformity. A secondary goal is to increase ROM. Most patients who undergo biplane osteotomy with ulnar length adjustment or the crescentic radial osteotomy of Carter and Ezaki attain both primary goals.[17] Usually, ROM is, at best, only moderately improved. It should be noted that pain relief can be substantial and can be achieved quickly following release of the volar Vickers ligament.

In a retrospective review, Del Core et al evaluated long-term clinical and radiographic outcomes after Vickers ligament release and distal radial physiolysis in a small group of skeletally immature patients (N = 6; 8 wrists) with symptomatic MD.[18]  At 1 year of clinical follow-up, seven of the eight wrists were pain-free, and six were completely pain-free at final follow-up. Motion in flexion, extension, pronation, supination, radial, or ulnar deviation was similar between preoperative status and status on long-term follow-up. At the final follow-up, two patients had undergone a subsequent procedure (one radial dome osteotomy and one ulnar-shortening osteotomy).

Symptoms of Madelung deformity (MD) usually begin during adolescence in girls aged 10-14 years; patients rarely present when younger than 8-9 years. Because MD is observed so rarely in males, it has been proposed that a true severe case of MD never occurs in males.

Patients experience increasing deformity and pain in the wrist with decreased range of motion (ROM). On physical examination, the hand is translated volarly to the long axis of the forearm. The ulna, being relatively unaffected, abuts the carpus and becomes prominent dorsally relative to the carpus and hand. ROM is decreased, with a limitation of supination, dorsiflexion, and radial deviation. Pronation and flexion usually are normal. (See the images below.)

A few case reports have described extensor tendon rupture occurring in conjunction with chronic Madelung deformity.[19]

The diagnosis of Madelung deformity (MD) is confirmed radiographically with posteroanterior (PA) and lateral views of the forearm and wrist. (See the images below.)

Carter and Ezaki standardized a method of obtaining the PA forearm radiographs in patients with MD.[17]  Both right and left forearm images are placed on the same radiographic plate for comparison. The true PA radiograph is obtained with the forearm in neutral rotation, the elbow at 90° of flexion, and the shoulder abducted to 90°. These authors have found this method of analysis useful in preoperative planning, and it is highly reproducible by radiology staff.

A lateral radiograph is obtained by rotating the shoulder to the side of the patient and laying the ulnar border of the forearm on the plate.

Several distinct radiographic features of MD exist. Dannenberg et al described elements of radiographic diagnosis after their review of 172 cases.[20]  Criteria include the following:

• Lateral and dorsal curvature of the radius
• Widened interosseous space
• True shortening of the total length of the radius
• Premature fusion of the ulnar half of the distal radial physis
• Focal osteopenia in the area of the ulnar portion of the distal radius
• Exostosis at the distal ulnar border of the radius
• Triangularization of the distal radial epiphysis
• Ulnar- and palmar-facing distal radial articular surface
• Relative dorsal subluxation of the ulna
• Increased radiodensity of the ulnar head
• Carpal wedging with the lunate at the apex of the wedge
• An arched curvature of the carpal bones in direct continuation of the dorsal bowing of the radius on the lateral radiograph

Carter and Ezaki added the observations that there is an increased radial tilt on the PA radiograph and that the radial epiphysis becomes teardrop-shaped.[17]  On the lateral view, the radius tilts volarward until the ulna appears dislocated from its normal articulation with the radius. The ulna lies dorsal to the proximal carpal row. The distal ulna is not actually subluxated dorsally. Rather, the hand and radius are translated palmarward, resulting in incongruence of the distal radioulnar joint (DRUJ). The unaffected ulna continues to grow, ultimately becoming longer than the ulnar aspect of the radius.

Vickers and Nielsen described a physeal measurement for preoperative and postoperative assessment of severity and correction, respectively.[21]  PA radiographs are used for measurement; lateral views have too much variation to be of value. The longitudinal and transverse axes of the radius are difficult to determine because of the considerable bowing of the radius. The ulna usually is straight, which provides a reliable longitudinal axis. Therefore, a perpendicular line to the long axis of the ulna can be passed through the most radial extent of the radial physis.

A line joining the most radial and ulnar extents of the physis is drawn. The angle subtended by the transverse line and the physeal line is termed the physeal angle. With increasing severity, the lunate follows the ulnar-volar corner and eventually becomes interposed between the radius and ulna. The relative position of the proximal aspect of the lunate can be measured from the transverse line.

McCarroll et al described the following threshold values for four radiographic measurements as leading to a diagnosis of MD[22] :

• Ulnar tilt ≥33º
• Lunate subsidence ≥4 mm
• Lunar fossa angle ≥40º
• Palmar carpal displacement ≥20 mm

Because the deformity is three-dimensional (3D), a better understanding may be obtained with tomography or computed tomography (CT). These studies may be helpful in precisely visualizing physeal and articular morphology. A CT-derived three-dimensional (3D) reconstruction may assist in clarifying the articular orientation.[23] However, Carter and Ezaki stated that after the complexity of this lesion is appreciated, CT and 3D imaging are not necessary for routine treatment.[17]

Operative treatment of Madelung deformity (MD) is indicated for pain relief and cosmetic improvement. These indications are consistent among many authors. Range of motion (ROM), especially in pronation and supination, usually is only minimally improved and has varied in different operative series.

In the skeletally immature patient with clear evidence of MD, the most likely cause of pain is tension within the Vickers ligament. Release of the ligament either alone or in combination with an osteotomy is indicated.

In the skeletally mature patient, the congruency of the radiocarpal joint and that of the distal radioulnar joint (DRUJ) are assessed. If an osteotomy will result in a secondarily congruous joint, then an osteotomy is indicated. However, if radiocarpal congruity is not possible to obtain, a radioscaphocapitate arthrodesis is indicated. Similarly, if DRUJ congruity is not possible, then a Darrach or Sauve-Kapandji procedure is indicated.

No specific contraindications for surgery exist other than those associated with any elective surgical procedure.

Nonoperative management may be helpful in skeletally mature individuals with MD and mild-to-moderate short-term wrist pain. If the pain is predominantly related to the DRUJ, then a sugar tong–type splint may relieve joint irritation from overactivity. If radiocarpal pain is the primary problem, then a volar splint may decrease symptoms. Patients may eventually decrease their manual activity levels to a point where their symptoms are manageable without surgery.

In contrast, the younger and skeletally immature patient with clear evidence of MD has pain that is caused by the tension within the Vickers ligament, and splinting most likely will not have a satisfactory result.

The surgical decision is based on the following four factors:

• Patient's age and the growth remaining in the distal radius
• Severity of the deformity
• Severity of the symptoms
• Clinical and radiographic findings

Operative treatment can be divided into three broad categories as follows:

• Procedures that correct the primary deformity of the radius
• Procedures that attempt to decrease pain and increase ROM by making a compensatory change in the ulna
• Procedures that address both

Multiple procedures address the deformity in the radius. Conceptually, these can be broken into the following three types:

• Procedures that change the growth or anatomy at the physis
• Procedures that change the bony anatomy of the metaphysis
• Procedures that are considered salvage-type joint-sacrificing procedures

The goal of ulnar procedures is to change the relationship of the relatively long ulna to the radius.

When the deformity is noticed early and significant growth remains, changing the growth pattern of the distal radial physis to correct the deformity is possible. In 1992, Vickers and Nielsen described the lesion in the volar and ulnar distal radius as both bony and ligamentous, and they stated that it is an inherent failure of focal growth and structural tethering of further growth. They described an ulnar-volar release for MD of the physis, called physiolysis (see the images below).

This allows, then, normal and compensatory growth to correct the deformity, much like the bony bridge resection of Langenskiöld. Vickers and Nielsen used the procedure of resecting the bony and ligamentous lesion (the first to correct MD using remaining growth to correct the deformity). They were the first to describe a ligamentous lesion as part of the pathology and also were the first to use the volar approach to address it.

Through a volar flexor carpi radialis (FCR) approach, the pronator is mobilized, and the distal radius is exposed. An osteotome is used to make a longitudinal split 5 mm from the DRUJ on the radius, and parallel transverse 1-mm sections are made until normal physis is identified. Some physis is left to overhang the metaphysis. Fat harvested from the forearm is packed into the bone defect. Completely excising the tethering ligament (Vickers ligament) between the lunate and the radius is critical. (See the images below.)

The Vickers ligament originates on the radius in a fossa that is seen radiographically as a flame-shaped radiolucency distal to a bone spur on the ulnar aspect of the distal metaphysis (see the images below). Dannenberg noted this characteristic as an area of osteopenia in his original radiographic criteria in 1938. Carter and Ezaki noted it in 91% of their cases.

The Vickers ligament is fibrous and fibrocartilaginous and 5-7 mm thick. It inserts into the anterior surface of the lunate and the anterior radioulnar ligament portion of the triangular fibrocartilage complex (TFCC). Most likely, it is a secondary pathologic structure to a primary bony derangement and may even be a coalescence of normal structures. This ligament may be an etiologic factor in MD pain. Following operative release of this ligament, patients state soon after surgery that their pain is largely relieved.

If the deformity has progressed in an older child and remaining growth is insufficient, several procedures can be used to correct the position of the distal radiocarpal joint surface. They generally consist of a biplane osteotomy, either closing or opening wedge, which corrects the position of the joint surface and brings the radius and ulna into a more proper position. If a positive ulnar variance remains, an ulna-shortening procedure can be performed.[24, 25, 26]

A less commonly used method is distraction histogenesis with the Ilizarov technique. Several authors have described experience with the technique of radial osteotomy and subsequent angular and length correction over time.[27, 28]  The goal is slow correction of the distal radius to make a more congruous DRUJ. Children treated with this method are either skeletally mature or nearly mature.

Half pins are commonly selected over tensioned thin wires to allow better ROM during distraction. Either three or four rings may be used, with fewer complications and less pain noted when three are used. The first goal after fixation placement and radial osteotomy is angular correction over a 2-week period. The second goal is lengthening of approximately 1.5 cm over the next 2 weeks, followed by consolidation for 3 weeks. The entire process in the distractor takes 8 weeks.

In 2000, Carter and Ezaki reported a combined procedure using a Vickers ligament release and a dome-shaped osteotomy of the radius to correct all of the aspects of the radial deformity, including the radial and volar translation of the distal metaphysis.[17]  This procedure is performed with a volar approach, and the Vickers ligament is excised entirely.

A dome-shaped osteotomy is made in the metaphysis so that the articular surface not only decreases in its radial inclination but also translates as it rotates. This translation reestablishes articular support to the lunate. A second maneuver translates the articular surface dorsally. The prominent volar bone is then removed, and the osteotomy is pinned into position with two Steinman pins. This not only corrects the deformity but also decreases pain and increases ROM. (See the images below.)

Volar ligament release with distal radial dome osteotomy has been shown to yield lasting correction of Madelung deformity.[29, 30]

In MD, the ulna grows normally and becomes longer than the radius. Because the radius is volar, the ulna appears to be subluxated dorsally. The incongruence at the DRUJ and the impingement of the radius on the ulna in supination may cause pain and contribute to decreased ROM in supination. To allow unrestricted rotation, several ulnar procedures have been described; these include ulnar shortening, ulnar head resection and a Sauve-Kapandji (Lauenstein) DRUJ arthrodesis,[31] and ulnar pseudoarthrosis. Several authors have advocated both radial and ulnar procedures.

The Darrach procedure has long been a treatment option for MD. In isolation, this construct may leave the carpus unstable, especially in light of the increased ulnar and volar slope of the radial articular surface, as a result of which the carpus tends to slide off of the ulnar side of the wrist. Several authors have devised procedures to solve this problem. The Sauve-Kapandji (Lauenstein) procedure may be a viable option for MD in that when the ulnar head is preserved, less chance exists for ulnar migration of carpus.

In 1975, Ranawat studied 13 wrists in eight patients with MD. His operative indications were pain and limitation of motion; deformity alone was not an indication for operation. Mild deformity was treated with a Darrach procedure, and severe deformity was treated with Darrach plus a biplane radial osteotomy. In patients treated with a Darrach procedure, the carpus tended to translate ulnarward. A radial osteotomy was theorized to improve the muscle balance about the wrist and to provide better axial bone support for the carpus on the radius in the absence of the ulnar head.

In 1993, Watson et al described another combined procedure in which the radial osteotomy was performed with both a closing wedge technique on the radial aspect of the metaphysis and an opening wedge on the ulnar aspect with the radial bone wedge. This technique preserved radial length but also required an ulnar head resection. Pain was significantly relieved in all 15 wrists.

White and Weiland described a combination procedure in a series of wrists with MD of posttraumatic etiology.[32]  The procedure included a volar closing wedge osteotomy along with a distal radioulnar arthrodesis and ulnar pseudoarthrosis proximal to the arthrodesis. The osteotomy of the radius improved the mechanics of the radiocarpal joint and prevented radiocarpal arthritis.

This procedure alone may not improve pain on supination/pronation due to the incongruent DRUJ. Resection of the distal ulna, as in the Darrach procedure, may solve this problem, but ulnar translation of the carpus may result. Fusing the DRUJ maintained carpal stability. Increase in the patient's ROM was minimal in any plane, but pain was significantly relieved, the carpus was stable, and the cosmetic result was satisfactory.

In the skeletally mature patient who presents with pain and in whom physical examination demonstrates limitation of motion and severe radiocarpal joint incongruity, a salvage procedure may be indicated. When incongruity is severe, any surgical manipulation of the joint surfaces would be difficult, and the joint would not be stable for smooth articulation. For long-term pain relief and a stable joint, a radiocarpal arthrodesis and distal ulna resection are indicated.

Postoperative management depends on what is done operatively. Cast immobilization is necessary for 6-8 weeks after an osteotomy of the radius or ulna to allow for bony healing. Hand therapy is necessary in children who are not able to regain ROM on their own after 2-3 months without restrictions.

In the author's experience, no specific complications after operative correction have arisen as problematic. However, isolated cases of postoperative wound infections, reflex sympathetic dystrophy, and recurrent deformity after continued growth are reported in the literature.

Children treated for MD should be monitored until surgical issues are resolved and then yearly until skeletal maturity is reached. After a physiolysis procedure, yearly radiographs can be used to document improvement in the position of the distal radial articular surface with growth.