Biomechanical Comparison of Modified Adams-Berger and DX technique in DRUJ Reconstruction

Abstract

Background  Adams-Berger ligamentoplasty is a widely accepted reconstruction for unrepairable triangular fibrocartilage complex (TFCC) injuries with instability. Failure of the reconstruction and recurrent instability is still a clinical problem. Internal brace augmentation of tendon grafts is gaining more popularity, but use in the distal radioulnar joint (DRUJ) is not yet published.

Questions/Purposes  The purpose of this study was to compare a novel anatomical DRUJ reconstruction with a modified Adams-Berger reconstruction to answer which technique stabilize better the DRUJ and which has enough stabilizing effect to allow early mobilization.

Methods  Nine matched pairs of cadaveric upper extremities were used. The dorsopalmar translations in the DRUJ that occurred with 50 N load were measured before and after detachment of the TFCC from the ulna and after ligament reconstruction with either modified Adams-Berger procedure or DX Swivelock technique. Internal brace augmented palmaris longus tendon grafts were used in all reconstruction.

Results  In the Adams-Berger group, the injured and the reconstructed displacements were significantly higher than the native, while in the DX group both the native and the reconstructed displacements were significantly lower than the injured. The mean (standard deviation) change of translations was 0.46 (1.94) mm after Adams-Berger and 2.51 (1.31) mm after DX reconstruction, implying significant better stabilizing effect of the latter.

Conclusions  DX Swivelock reconstruction showed better time zero stabilizing effect compared with Adams-Berger procedure, regaining almost normal stability of the DRUJ.

The distal radioulnar joint (DRUJ) has a complex motion pattern during forearm pronation–supination: rotation, dorsal–palmar translation, and proximal–distal pivoting of the radius. In addition to the geometry of the bones, the soft tissue structures stabilize the DRUJ, the triangular fibrocartilage complex (TFCC) being the main stabilizer. ^{1 2} After avulsion of the TFCC from the ulna, a surgical reattachment of the TFCC to the fovea can stabilize the DRUJ. In chronic and irreparable cases, reconstruction of the radioulnar ligaments could be necessary.

Brian Adams published an anatomical ligament reconstruction technique, aiming to recreate both the dorsal and the palmar limbs of the DRU ligaments in a triangular configuration. ³ A free palmaris longus (PL) tendon graft is passed through a tunnel in the radius, both limbs are entered into the ulnocarpal joint and passed through a tunnel in the ulnar head, and finally wrapped around the neck of the ulna, tightened, and fixed by suturing the limbs to each other. This method, described in detail in Adams' and Berger's paper, ⁴ became the “gold standard” for the treatment of chronic DRUJ instabilities, and satisfying long-term results have been published since. ⁵ However, failure of the reconstruction with recurrence of instability is still a problem, despite the use of new fixation methods, such as the use of an interference screw for fixation of the tendon graft.

Augmentation of the free tendon graft with nonresorbable suture material is gaining emerging popularity, aiming to withstand elongation or rupture of the graft. Promising results are published in different ligament reconstructions with this internal brace concept, both in animal models and in clinical studies. ⁶

The purpose of this study was to compare a novel anatomical DRUJ ligament reconstruction method, with a modified Adams-Berger reconstruction in a cadaver model, using internal brace augmented free tendon grafts. The main study question was, which technique stabilized the DRUJ better? The second question was, which technique had enough stabilizing effect to allow early mobilization?

Methods

Specimen Preparation

Both upper extremities of nine fresh-frozen cadavers (9 male) with a median age of 49 years (range: 22–64 years) were used. The limbs were amputated mid-humerus level and any skeletal disorders and deformities were radiologically ruled out. After thawing in room temperature, all the specimens were prepared in the same way. The shaft of the humerus was cleaned of soft tissues, keeping the elbow joint and forearm muscle insertions intact, and then embedded in a fiberglass resin block. From the middle of the forearm to the metacarpal base level, the skin was also removed. The radius metaphysis was exposed both on the volar and the dorsal side, without damaging the tendons and joint capsule. Just proximal to the Lister̀s tubercle, a 6 mm dorsopalmar tunnel was drilled through the distal radius. The exact placement and orientation of the tunnel were checked under fluoroscopy. A standard threaded commercial 5.75 mm bolt was fastened with washers and a self-locking nut on the dorsal side. The ulna shaft was also exposed, and 4 cm proximally from the tip of the ulnar styloid, a 1.5 mm olive wire was placed in a dorsopalmar direction while keeping the forearm in a neutral pronosupination position. The specimens were kept moist with intermittent saline application under preparation and testing.

Biomechanical Testing

The setup was similar to that of the study of Haugstvedt. ⁷ The specimens were mounted on an Instron 8871 Hydraulic Machine (Canton, MA), with 90 degrees of flexion in the elbow joint and neutral forearm rotation, with the palm facing down. The distal ulna shaft was solidly fixed by metal blocks around the olive wire, and the bolt in the radius was attached to a 5 KN load cell ( Fig. 1A ). Linear testing at a speed of 1 mm/s was performed with force control. From a position where the ulna was positioned approximately in the middle of the sigmoid notch, we started with palmar translation of the radius relative to the ulna ( Fig. 1B ). When the force reached the 50 N limit, the radius was pulled with the same speed in the opposite direction ( Fig. 1C ), until the force again reached the 50 N limit, then returned to the starting point. During the testing, the force and displacement were simultaneously measured and plotted.

Fig. 1.

( A ) The humerus and the ulna are firmly fixed with 90 degrees elbow flexion and neutral rotation; the radius is mounted to the testing machine with a bolt. ( B ) Palmar translation of the radius relative to the ulna. ( C ) Dorsal translation of the radius. The asterisk marks the ulnar head.

Simulating TFCC Injury

Immediately after testing the native specimens, we performed minimally invasive destabilizing of the DRUJ, keeping the specimens attached to the testing machine ( Fig. 2 ). We sharply cut both the styloid and the foveal insertions of the TFCC with a pointed scalpel through a stab incision of the capsule, dorsal to the ulnar styloid, as previously described, ⁷ then repeated the testing with the same setup.

Fig. 2.

Minimally invasive cutting of the triangular fibrocartilage complex insertions from both the ulnar styloid and the fovea with a scalpel. The contour of the ulna is outlined.

Surgical Technique

After the second testing, the right- and left-sided specimens of each pair were randomized to two different reconstructions, and performed randomly by four surgeons in two pairs. The standard dorsal approach to the DRUJ through the sheath of the extensor digiti minimi (EDM) tendon was used. The fovea was cleaned, and an oblique tunnel was drilled from the medial side of the ulna to the fovea, under fluoroscopic guidance with a cannulated 3.0 mm drill. A full length 2 mm thick strip of the PL tendon was harvested as a graft and both ends were whipstitched with 2–0 FiberWire (Arthrex, Naples, FL). As an internal brace augmentation, we used a 1.3 mm SutureTape (Arthrex, Naples, FL) in both reconstructions. The fixation to the radius was the only difference between the groups.

Adams-Berger Reconstruction

In the Adams-Berger group, a dorsopalmar oriented guidewire was placed under fluoroscopy in the distal–ulnar corner of the radius, parallel to and with a 3 mm distance from the joint surfaces of the lunate facet and the sigmoid notch. The guidewire was then drilled over with a cannulated 3.0 mm drill. The internal brace augmented tendon graft was pulled inside this tunnel using a wire loop.

DX Reconstruction

In the novel technique, we placed 1.35 mm guidewires in the dorsal–distal and palmar–distal corners of the sigmoid notch, facing the center of the distal radius ( Fig. 3 ). When the correct placement was fluoroscopically approved, 10 mm deep socket holes were drilled with a 3.5 mm cannulated drill bit. The graft was divided into two equally long limbs and both ends were whipstitched. 1.3 mm SutureTape was used for augmentation of both independent limbs ( Fig. 4A ), one was fixed in the palmar ( Fig. 4B ) and the other in the dorsal hole using 3.5 mm DX SwiveLock SL Anchors (Arthrex, Naples, FL).

Fig. 3.

Fluoroscopic control of the guidewire placement in the palmar–distal corner of the sigmoid notch. Note the drill hole in the radius metaphysis for attachment of the metal bolt.

Fig. 4.

( A ) The fork of the DX SwiveLock SL Anchor is loaded with the tendon graft and the SutureTape. ( B ) The anchor right before it is seated into the palmar hole in the radius. The contour of the radius is outlined. ( C ) The whipstitched graft and the SutureTape are pulled through the joint capsule with a grasper. ( D ) The palmar graft, marked with the asterisk, is laying on top of the capsule. The SutureTape is covered by the graft and is not visible.

The remaining steps of the reconstructions were similar. The volar limb of the graft and the internal brace were drawn into the joint using a grasper through a small incision of the volar capsule ( Fig. 4C, D ). Both limbs together with their respective internal brace ( Fig. 5A ) were pulled through the ulnar tunnel from inside out, recreating the volar and the dorsal DRU ligaments ( Fig. 5C ). This step could be challenging. We solved this with the use of a thin nitinol loop (SutureLasso SD wire loop, Arthrex, Naples, FL). With this loop, we retrieved the stitch of the palmar graft and the matching SutureTape through the ulnar tunnel, then placed the loop inside the tunnel again, before pulling the palmar graft limb through the tunnel. Then, using the loop again, we easily retrieved the stitch of the graft and the SutureTape of the dorsal limb, and by pulling the stitch, the dorsal graft could also be drawn through the ulnar tunnel. Alternatively, a small QuickPass Tendon Shuttle (Arthrex, Naples, FL) could be used to pass both grafts and SutureTapes in one step ( Fig. 5B ). While holding the forearm in neutral position, both limbs of the reconstruction could be tensioned, and fixed in the ulnar tunnel with a 3 × 8 mm PEEK Tenodesis Screw (Arthrex, Naples, FL) placed from the medial side through the hole in the ulna ( Fig. 6A ). The redundant tendon graft was then cut flush. A 3.0 mm diameter blind socket was drilled 1 cm proximal to the ulnar tunnel into the ulna. Both SutureTapes were fixed to the ulna with a 3.5 mm DX SwiveLock SL anchor ( Fig. 6B ), and the tails were cut flush. To finish off, the capsule-retinaculum flap was sutured with running 2–0 FiberWire, letting the EDM tendon lay on top. After completing the reconstruction, the specimens were tested again with the same setup, as described earlier.

Fig. 5.

( A ) The palmar (*) and the dorsal (#) limbs are seen through the dorsal arthrotomy. ( B ) Both limbs together with their respective internal bracing are pulled through the ulnar tunnel from inside out, using a Tendon Shuttle. ( C ) The triangular configuration of the distal radioulnar joint ligaments is recreated.

Fig. 6.

( A ) Under tensioning, both limbs are fixed in the ulnar tunnel with a Tenodesis screw. ( B ) Both SutureTapes are fixed to the ulna again with a 3.5 mm DX SwiveLock SL Anchor. The contour of the ulna is outlined.

Analysis

We collected three displacement curves for every specimen, the values from the native, the injured, and the reconstructed conditions. These values indicated how much dorsopalmar translation in the respective DRUJ could occur with a 50 N force. The data were statistically analyzed with a paired t -test, comparing the displacement between the different measurements within the same specimen. Using Bonferroni correction, the significance level was set to 0.017. The difference of translations after reconstruction was also calculated for all specimens, by subtracting the reconstructed displacement values from the injured values. These were compared between the groups with an independent sample t -test with a significance level of 0.05.

Results

The mean (standard deviation) native, injured, and reconstructed dorsopalmar displacements caused by 50 N were 7.29 (2.35), 11.11 (2.31), and 10.65 (1.99) mm in the Adams-Berger and 8.23 (2.96), 11.73 (3.6) and 9.22 (3.52) mm in the DX group, respectively ( Fig. 7 ).

Fig. 7.

The results of the biomechanical tests. The rectangular bars represent the mean, and the error bars the standard deviation of the displacements of the native, injured, and reconstructed specimens. The horizontal bars represent the p -values of the paired t -tests comparing the displacements between the different measurement values within the same specimen. Using Bonferroni correction, the significance level was set to p ≤ 0.017. In the Adams-Berger group, the native displacement values differed significantly from the injured and the reconstructed values, while in the DX group, the injured values differed significantly from the native and the reconstructed values.

The difference of translations was 0.46 (1.94) mm in the Adams-Berger and 2.51 (1.31) mm in the DX group, implying a significantly better stabilizing effect of the DX reconstruction on the DRUJ ( p  = 0.019).

In the Adams-Berger group, the native translation values were significantly lower than the injured and the reconstructed values; however, the two latter did not differ. In three specimens, the displacement after the Adams-Berger reconstruction was actually larger than after the injury. On the contrary, in the DX group, the injured values were significantly higher compared with both the native and the reconstructed values, with no significant difference between the two latter. In other words, the DX reconstruction was able to regain the stability in the DRUJ, while the Adams-Berger procedure failed.

Discussion

Chronic instability necessitates ligament reconstruction for stabilizing the DRUJ. Unfortunately, the failure of a free tendon graft reconstruction is not an uncommon complication, causing a clinically significant recurrence. In this study, we compared the Adams-Berger reconstruction with a novel technique, the DX reconstruction.

The mean limitation of our study is that we tested the reconstructions only with one cycle of a 50 N load in a linear fashion, in one position. This force limit was chosen from previously performed biomechanical studies on the DRUJ. ⁷ Probably cyclic loading in different positions or rotational cycles could better mimic the normal load in the DRUJ under mobilization. However, this study in this form was also time consuming, and the elongated time for further tests could lead to the natural breakdown of the cadaveric tissue.

The testing was initiated from a position where the ulna was positioned approximately in the middle of the sigmoid notch. It is difficult to place the ulna in the same position repeatedly relative to the radius. However, we believe that the different positioning would not affect the results, as we performed testing in both directions until reaching the load limit. Some more dorsal or palmar positioning would tension the soft tissues and the reconstruction on one side, which would cause less displacement one way, but more displacement in the other way, resulting the same total translation.

The neutral prosupination position was chosen because other static soft tissue stabilizers are less tight allowing more laxity in the DRUJ. In pronation the dorsal capsule, while in supination the pronator quadratus muscle, the volar capsule and the intraosseous membrane are more tensioned; this could theoretically cause less displacement. On the contrary, the limbs of the reconstruction are more tensioned in neutral position, so the stabilizing effect of the reconstruction is theoretically best investigated in the neutral position. Testing in more pronated or supinated positions would probably show less difference between the native, injured, and reconstructed displacements, decreasing the power of our analysis. We did not apply any load to the pronator quadratus muscle and the extensor carpi ulnaris tendon, which are important dynamic stabilizers of the DRUJ. This could be considered as further limitation of our study; however, our aim was to test the static stability.

In our study, we analyzed only the time zero stability of the reconstructions. The open surgical approach with detaching the dorsal capsule and the remnants of the foveal attachment after the mini-invasive cut has a further destabilizing effect on the DRUJ, affecting both reconstructions. Probably the reconstruction was not able to compensate for this effect in every specimen, and is most likely the reason why we found a greater displacement in several specimens after the Adams-Berger reconstructions, but not in the DX group.

In the original Adams-Berger reconstruction, the graft was originally sutured to the periosteum of the ulna and wrapped around the ulna shaft, suturing to itself. However, there was no specific fixation between the graft and the radius. This construct would most likely lead to even larger displacement with the same test setup. With healing of the joint capsule and tissue ingrowth to the free tendon grafts, the joints gradually gain better stability, though protection during this period with immobilization for 6 to 8 weeks is necessary. Some modifications of this procedure have been described, such as fixation of the graft to the ulna with interference screws ⁵ and arthroscopic assistance instead of open procedure. ⁸ This aims to increase the strength of the ulnar fixation and decrease the destabilizing effect of the arthrotomy, and therefore reduce the length of immobilization. The immobilization of the forearm often leads to decreased rotation, and the patient needs a long period of intensive physiotherapy to regain functional mobility. However, the forearm rotation is often decreased after the reconstruction, and passive exercises and stretching can lead to attenuation or rupture of the graft.

Several studies were published on internal brace augmentation for different joint stabilizing and tendon enhancement procedures: anterior cruciate ligament, ^{9 10} knee medial collateral ligament, ¹¹ patellar tendon, ¹² rotator cuff, ^{13 14} talofibular ligament, ^{15 16} thumb metacarpophalangeal joint, ulnar collateral ligament, ^{17 18} and scapholunate ligament. ¹⁹ The internal brace augmentation principle could be applied for DRUJ reconstructions as well. It could probably decrease the risk of elongation or failure of the graft causing recurrent instability. However, long-term results with internal brace augmentation of ligament reconstruction are as of yet not available.

Our intention was to carry out the procedures as similar as possible, to decrease the differences between the groups. We used solid fixation of the graft ensured by correct tunnel and interference screw placement. To increase time zero stability, we chose to add internal brace augmentation, which could also eliminate the mechanic failure due to the weak tendon graft. The real difference was in the radial fixation of the graft. With the DX technique, we could fix the grafts to more anatomically correct positions than with the Adams-Berger procedure. In addition, in the latter, there was no specific fixation between the graft and the radius, meaning that the graft could slide inside the radial tunnel, while the DX SwiveLock anchors could attach the graft to the bone in a very stable way.

For the Adams-Berger reconstruction, a surgical approach to both the palmar and dorsal side of the radius is necessary. The DX technique could be performed all inside the joint with arthroscopic assistance, with only a small open approach to the distal ulna. We chose to carry out the reconstructions open, to decrease the time frame of the study and for better reproducibility. Carratalá Baixauli et al described a similar technique, where they fixed the free tendon grafts all-arthroscopically with DX anchors to the radius and achieved good clinical results in four cases. ²⁰ However, they did not use internal brace augmentation and immobilized the wrists for 6 weeks.

In this cadaver model, we found a greater stabilizing effect of the posttraumatically lax DRUJ with the DX SwiveLock fixation compared with the modified Adams-Berger procedure. Due to the higher time zero strength of the DX reconstruction, it could possibly allow early mobilization of the wrist without an increased risk for failure.