- Email: [email protected]
Graft reconstruction of the interosseous membrane in conjunction with metallic radial head replacement: A cadaveric study
Graft Reconstruction of the Interosseous Membrane in Conjunction With Metallic Radial Head Replacement: A Cadaveric Study Samir G. Tejwani, MD, Keith L. Markolf, PhD, Prosper Benhaim, MD, Los Angeles, CA
Purpose: Longitudinal radioulnar dissociation (Essex-Lopresti injury) occurs when traumatic axial loading through the wrist disrupts the interosseous membrane (IOM) of the forearm and fractures the radial head. Proximal migration of the radius results in an ulnar-positive wrist, which can lead to painful ulnar-sided wrist degeneration and distal radioulnar joint instability. The purpose of this study was to measure the ability of an IOM reconstruction used in combination with a metal prosthetic radial head implant to reduce distal ulnar forces in a cadaveric model. The effects of varying the initial graft pretension on distal ulnar force were also studied. Methods: Twelve fresh frozen and thawed cadaveric forearms had a miniature load cell installed to record force in the distal ulna as the wrist was loaded axially to 134 N of compression force in neutral rotation. Intact forearms were tested first with the elbow in valgus and varus alignments. Loading tests were repeated after (1) insertion of a metal radial head implant that restored radius anatomic length, (2) excision of the IOM (with a radial head implant), and (3) reconstruction of the IOM using a palmaris longus tendon autograft (with a radial head implant). The implant then was removed and loading tests were repeated using 3 levels of initial graft pretension. Results: Mean distal ulnar forces with an intact forearm were 23% of applied wrist force in the varus alignment and 12% in the valgus alignment. Mean force levels after insertion of the implant were 18% (varus) and 13% (valgus); these were not significantly different from corresponding values for the intact forearm. Mean force levels after section of the IOM were 30% (varus) and 14% (valgus); these were not significantly different from corresponding values for the intact forearm (varus and valgus) but the mean for varus was significantly greater than the corresponding value with an implant. After IOM reconstruction with a palmaris longus tendon tensioned to 22 N mean distal ulnar forces were 8% (varus) and 7% (valgus); these means were significantly less than the corresponding values for all prior test conditions. With the radial head removed increasing the level of graft pretension reduced significantly mean distal ulnar force.
From the Department of Orthopaedic Surgery, Biomechanics Research Section, David Geffen School of Medicine, University of California Los Angeles, Los Angeles, CA. Received for publication February 3, 2004; accepted in revised form July 7, 2004. No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article. Reprint requests: Keith L. Markolf, PhD, UCLA Dept of Orthopaedic Surgery, Biomechanics Research Section, Rehabilitation Building, 1000 Veteran Ave, Rm 21-67, Los Angeles, CA 90095. Copyright © 2005 by the American Society for Surgery of the Hand 0363-5023/05/30A02-0016$30.00/0 doi:10.1016/j.jhsa.2004.07.022
The Journal of Hand Surgery
335
336
The Journal of Hand Surgery / Vol. 30A No. 2 March 2005
Conclusions: With the IOM resected insertion of a metal radial head implant alone did not reduce distal ulnar forces to intact forearm levels. When an IOM reconstruction was performed in combination with the implant mean distal ulnar force was reduced significantly to a level below that for the intact forearm. Applying pretension to the graft displaced the radius distally thereby making the wrist more ulnar negative and reducing distal ulnar force. Our results suggest that an IOM reconstruction used in combination with a metal radial head implant theoretically could help reduce distal ulnar impaction in an Essex-Lopresti injury. (J Hand Surg 2005;30A:335–342. Copyright © 2005 by the American Society for Surgery of the Hand.) Key words: Essex-Lopresti, graft reconstruction, interosseous membrane, radial head replacement.
In 1946 Curr and Coe1 described fracture-dislocation of the radial head in conjunction with distal radioulnar joint (DRUJ) dislocation. Today this pattern of injury is named after Essex-Lopresti, who in 1951 described traumatic axial loading through the wrist resulting in DRUJ instability, interosseous membrane (IOM) rupture, and fracture of the radial head.2 In 1992 Trousdale et al3 reported on proximal migration of the radius after comminuted radial head fracture and IOM injury, giving the condition the term longitudinal radioulnar dissociation (LRUD). During gripping activities proximal migration of the radius results in an ulnar-positive wrist with subsequent ulnocarpal impaction, painful ulnar-sided wrist degeneration, and DRUJ instability.4 –7 Interosseous membrane reconstruction has been studied in cadavers and performed in patients for the treatment of LRUD3,7–10 (Pfaeffle et al, presented at the 49th annual meeting of the Orthopaedic Research Society, 2003; Osterman et al, presented at the 67th annual meeting of the American Academy of Orthopaedic Surgeons, 2000). Radial head replacement with a metal prosthesis also has been recommended in cases of comminuted radial head fracture in which internal fixation was not feasible.6,8,11 The purpose of this study was to measure the ability of an IOM graft reconstruction to alter distal ulnar forces when used in combination with a metallic radial head implant. The experimental objectives were to measure directly distal ulnar forces at the wrist as the forearm was loaded axially to 134 N under the following test conditions: forearm with IOM intact, after removal of the radial head and insertion of a radial head implant that restored anatomic radial length (IOM sectioned), after IOM graft reconstruction (radial head implant in place), and after varying the initial graft pretension (radial head removed).
Methods Twelve fresh frozen and thawed cadaver forearms were used for this study. The donors ranged from 65 to 91 years of age (mean, 76 years); 8 were men and 4 were women. All cadaver forearm radiographs showed normal anatomy without significant prior wrist, forearm, or elbow pathology. All forearms were potted proximally at the distal humerus and distally using the central 3 metacarpals. The first and fifth metacarpals were removed. A specially designed miniature load cell was connected to prongs cemented into the distal part of the ulna for measurement of distal ulnar force. The potted humerus was mounted to a fixture attached to the crosshead of a materials testing system (MTS) machine (Model 812, MTS Systems Corp, Minneapolis, MN); and the elbow was flexed to 90° (Fig. 1). Load was applied to the potted central 3 metacarpals of the hand at a rate of approximately 1 mm/s; the maximum load applied was 134 N. All testing was performed with the wrist in neutral flexion– extension which was established by aligning the plane of the 3 potted metacarpals with the flexion– extension plane of the elbow. Prior studies from this laboratory have shown that load sharing at the wrist is affected by the varus– valgus position of the elbow.12–15 Valgus testing position was defined by applying a 2.5-Nm moment to the elbow that forced the radial head to contact the capitellum. Varus testing position was defined by applying a 2.5-Nm moment to the elbow that created a gap between the radial head and the capitellum. Further details on specimen mounting, load cell installation, and testing procedure are documented in our prior publications.12,14 The radial head implant used for this study (Evolve Modular Radial Head System, Wright Medical Technology, Arlington, TN) is a 2-part modular prosthesis with separate stem and head components of varying lengths and diameters. Using a hand-held micrometer with an accuracy of 0.05 mm we took a
Tejwani, Markolf, and Benhaim / IOM Reconstruction With Radial Head Replacement
Figure 1. A forearm specimen installed into the MTS machine.
direct measurement of the distance between the articular surface of the radial head and a reference mark on the proximal radius at a fixed distance from the radial head; this was used to establish anatomic radial length. After a transverse cut was made in the radial neck with an oscillating saw an implant was selected that restored anatomic length and matched the diameter of the radial head articular surface. In the event that the trial implant thickness selected did not restore precisely full anatomic length ring-shaped disk spacers were placed over the stem beneath the
337
head of the implant to restore the reconstructed radial length to within approximately 0.5 mm of that for the intact specimen. Because of ease of assembly and implant insertion trial aluminum implants identical in geometry to the actual implants were used in this study. The intact forearms first were loaded axially to 134 N in both varus and valgus alignments. The radial head was exposed and excised through the posterolateral approach described by Kocher.16 The trial radial head implant then was inserted as described above. The volar aspect of the forearm was dissected through the approach described by Henry17 and the IOM was exposed in its entirety. The central band of the IOM was found consistently to be more robust than the surrounding thinner, membranous tissue that composed the proximal and distal portions of the IOM. The central portion of the central band insertion was identified on both the radius and ulna and 3.0-mm drill holes were made in both bones to allow ultimately for tendon graft passage. The drill holes were oriented transverse to the longitudinal axis of the radius and ulna and were placed in the midaxial line of each bone. The IOM was sectioned in its entirety from distal to proximal under direct visualization and the forearms were loaded for 1 cycle to 134 N in both varus and valgus alignments. Of note, the tunnel holes were made in the radius and ulna before IOM sectioning to ensure accuracy in the identification of the orientation of the central band. Through a separate incision the palmaris longus tendon was identified and harvested from its characteristic superficial location between the flexor carpi radialis (FCR) and flexor digitorum superficialis tendons. The attached palmaris longus muscle was dissected free and each end of the tendon was fixed with a Bunnell stitch using a no. 2 nonabsorbable suture (Ethicon, Somerville, NJ). As expected based on previous publications, of the 12 forearms used in this testing sequence the palmaris longus tendon was absent in 2 (17%); a replacement palmaris longus tendon was harvested from a similarly sized forearm in these cases.18,19 Soft tissue was dissected from both sides of the bone tunnels in the radius and ulna and the edges were rounded with a high-speed burr. The palmaris graft first was secured to the distal ulna by wrapping the suture around a screw-post and washer construct. The graft was then passed through the ulnar and radial tunnels using a wire-loop suture passer. After the palmaris graft was pulled through the radial tunnel it was pretensioned to 22 N as verified by a
338
The Journal of Hand Surgery / Vol. 30A No. 2 March 2005
Figure 2. The palmaris longus tendon IOM graft reconstruction performed with a tunnel method and screw-post and washer fixation to the radius and ulna.
hand-held calibrated spring scale with a resolution of 1.1 N. The tendon graft was then secured to a screwpost and washer construct located approximately 1 to 2 cm distal to the radial tunnel (Fig. 2). All palmaris grafts in this study were preconditioned with 10 cycles of axial loading to 45 N by using the MTS machine. Preliminary testing in our laboratory showed that without burring the tunnel holes at the site of tendon entry the palmaris tendon graft was unable to undergo 10 cycles of preconditioning as a result of premature graft rupture. No grafts ruptured after burring of the tunnel holes. After preconditioning palmaris graft pretension was reset to 22.3 N and the forearms were then loaded for 1 cycle in both varus and valgus alignments. Finally the radial head implant was removed and the palmaris tendon graft was pretensioned to a zero load (0 N), defined as a taut graft with no additional preload placed on it. Loading tests were repeated with 0 N, 22 N, and 45 N of pretension applied to the graft. At each pretension level the forearm was
loaded axially for 1 cycle; the elbow was in a position midway between the varus–valgus positions determined for the intact forearm. A 1-way repeated-measures analysis of variance model was used to determine the significance of differences in mean distal ulnar forces between the following conditions: intact specimen, implant with intact IOM, implant with sectioned IOM, and implant with IOM reconstruction. A similar model was used to compare mean distal ulnar forces between the following conditions (radial head removed): intact IOM and graft pretensioned to 0 N, 22 N, and 45 N. Multiple pairwise comparisons between means were made using the Student-Neuman-Keuls procedure. The level of significance was p ⬍ .05.
Results Mean distal ulnar forces with an intact forearm were 23% of applied wrist force in varus alignment and 12% in valgus alignment (Fig. 3). Mean force levels after insertion of the implant were 18%
Tejwani, Markolf, and Benhaim / IOM Reconstruction With Radial Head Replacement
339
Discussion
Figure 3. Distal ulnar force recorded at 134 N of applied wrist load (expressed as a percentage of load applied to the wrist) with varus and valgus elbow alignments for 4 test conditions: intact forearm, radial head implant, implant with IOM sectioned, and implant with IOM reconstruction (pretension set at 22.3 N). Mean values are shown with standard deviations indicated by error bars. All comparisons between test conditions are significantly different from each other (p ⬍ .05) except where indicated by “n.s.”.
(varus) and 13% (valgus); these were not significantly different from corresponding values for the intact forearm (Fig. 3). Mean force levels after section of the IOM (implant in place) were 30% (varus) and 14% (valgus); these were not significantly different from corresponding values for the intact forearm (varus and valgus) but the mean for varus was significantly greater than the corresponding value with an implant (Fig. 3). After IOM reconstruction with a palmaris longus tendon tensioned to 22 N mean distal ulnar forces were 8% (varus) and 7% (valgus); these means were significantly less than the corresponding values for all prior test conditions (Fig. 3). With the radial head resected (implant removed) mean distal ulnar force with an intact IOM was 24% of applied wrist force (Fig. 4); mean force was not significantly changed after IOM reconstruction with a palmaris tendon graft that was tightened slightly before fixation but not tensioned intentionally (0 N) (Fig. 4). Pretensioning the graft to 22 N reduced mean distal ulnar force to 15% of applied wrist force; this was significantly less than mean force for the intact-IOM and 0-N–pretension conditions (Fig. 4). Pretensioning the graft to 45 N further reduced mean distal ulnar force to 9% of applied wrist force; this was significantly less than means for intact-IOM, 0-N, and 22 N pretension conditions (Fig. 4).
The primary restraint to proximal migration of the radius is the radial head as it contacts the capitellum.20 –22 In 1930 Brockman23 described proximal translation of the radius 3 months after radial head fracture. In 1992 Trousdale et al3 reported on 20 cases of proximal migration of the radius after comminuted radial head fracture, naming the pattern of injury LRUD. Longitudinal radioulnar dissociation can occur acutely with an Essex-Lopresti injury or late after the IOM attenuates with time.3,6,24,25 Shepard et al14 showed that for each 1 mm of radial shortening after radial head excision the distal ulnar load increased by 10%. Drobner and Hausman26 concluded similarly that a positive ulnar variance from proximal radial migration of 2 mm increases distal ulnar load by 20% to 40%. Early on this can manifest during activities involving a strong grip when there is a dynamic increase in ulnar-positive variance.4,5 With more severe radial migration the new distal ulnar position can limit forearm supination and wrist extension as the carpus impinges on the ulnar head; wrist pain, DRUJ instability, and decreased wrist and forearm motion may result.6,7
Clinical Studies of IOM Reconstruction Hotchkiss7 attempted IOM reconstruction in 1 patient using a palmaris tendon weave coupled with cross-pinning of the radius and ulna. The clinical
Figure 4. Distal ulnar force recorded at 134 N of applied wrist load (expressed as a percentage of load applied to the wrist) for the following test conditions (radial head removed): native IOM and after reconstruction of the IOM with a palmaris longus tendon autograft pretensioned to 0 N, 22 N, and 45 N. All comparisons between test conditions are significantly different from each other (p ⬍ .05) except where indicated by “n.s.”.
340
The Journal of Hand Surgery / Vol. 30A No. 2 March 2005
history of the patient and the precise surgical method were not described. Immobilization was performed for 8 weeks, after which significant proximal migration of the radius occurred indicating failure of the reconstruction. A radioulnar synostosis was performed subsequently as a salvage procedure. In 1 patient with chronic LRUD and dynamic ulnar-positive variance secondary to a failed silicone radial head replacement (Silastic). Ruch et al10 performed a triangular fibrocartilage complex repair, IOM reconstruction with bone–patellar tendon– bone autograft, and titanium radial head replacement with good results reported at 2 years. Osterman et al reconstructed prospectively the IOM in 14 patients with chronic LRUD using a central-third bone–patellar tendon– bone graft; no radial head replacement was performed (Osterman et al, presented at the 67th annual meeting of the American Academy of Orthopaedic Surgeons, 2000). An average 3 mm of ulnar-positive variance existed and an ulnar shortening osteotomy (mean, 4 mm) was performed on each patient. At follow-up examination (average time, 32 months) pain improved in all patients, elbow motion was preserved in 13 patients, forearm rotation loss was less than 20° in 11 patients, and grip strength improved from 62% to 85% of normal. No patient required additional surgery for increased ulnar-positive variance secondary to proximal radial migration. Postsurgical radiographs showed maintenance of ulnar variance in 11 of the 14 patients.
Biomechanical Studies of IOM Reconstruction Sellman et al8 reconstructed the central band of the IOM in 9 cadaver forearms using a 16-mm braided polyester cord with a triple helical weave in a tunnel method similar to the one used in this study; the cord pretension was not specified. Sectioning the central 10 cm of the IOM decreased gross forearm stiffness to 70% that of the intact forearm. With 1 cycle of loading to 75 N after IOM reconstruction and radial head excision gross axial stiffness was 94% that of the intact forearm. With the addition of a radial head replacement (Silastic) to the IOM reconstruction gross forearm stiffness remained at 94%. After the radial head implant was replaced with one made of titanium gross forearm stiffness increased to 145%. Distal ulnar force was not measured, nor were the effects of variable graft pretension. In a radial head– deficient cadaver forearm model Skahen et al9 reported that based on preliminary
testing IOM reconstruction with the palmaris longus tendon was unable to limit proximal radial migration with an applied wrist load of 89 N. After radial head excision the triangular fibrocartilage complex and IOM were sectioned incrementally from proximal to distal in 6 specimens. After 1 cycle of loading to 89 N in neutral forearm rotation the FCR IOM reconstruction decreased proximal radial migration from 12 mm to 6 mm. In another 6 specimens the sequence of sectioning of the IOM and triangular fibrocartilage complex was reversed to distal to proximal. In this scenario after FCR IOM reconstruction proximal radial head migration was decreased from 10 mm to 3 mm. The researchers concluded that a-FCR tendon autograft used in IOM reconstruction was capable of preventing complete migration of the proximal radius to the capitellum but was not capable of restoring completely the longitudinal stability of the forearm. Pfaeffle et al compared single- and double-bundle FCR tendon autograft IOM reconstructions in 12 cadaver forearms; the grafts were passed through tunnels and fixed to a screw-post construct (Pfaeffle et al, presented at the 49th annual meeting of the Orthopaedic Research Society, 2003). Fifteen newtons of cyclic preconditioning was applied and the grafts were pretensioned to 90 N. After 1 cycle of loading to 139 N the researchers concluded that the double-bundle FCR reconstruction restored successfully normal forearm mechanics in LRUD based on the load-sharing profile of the radius and ulna. The radial head was left intact throughout testing, however, and proximal migration of the radius did not change between the intact, IOM-sectioned, and IOMreconstructed conditions, varying from 0.1 to 0.3 mm. This suggests that load was applied in a valgus position with the radial head contacting the capitellum. In this elbow position axial load would not adequately test the ability of the IOM reconstruction to limit proximal migration of a mobile radius as is the case in the varus position with a gap between the radial head and capitellum. For the purposes of the present study the palmaris longus tendon autograft was selected to reconstruct the IOM because of its general expendability and frequent use in upper- extremity tendon reconstruction procedures. Skahen et al,9 however, reported that based on preliminary testing IOM reconstruction with the palmaris longus tendon was unable to limit proximal radial migration under 89 N of applied wrist load. In our tests the palmaris graft underwent only 1 cycle of loading with 134 N of applied wrist force and could be expected to attenuate under fur-
Tejwani, Markolf, and Benhaim / IOM Reconstruction With Radial Head Replacement
ther cycles of loading. Other acceptable graft substitutes with larger cross-sectional areas such as FCR tendon autograft and bone–patellar tendon– bone allograft may be better suited to reconstructing the IOM clinically. Nevertheless we found that when the relatively smaller palmaris tendon was used for IOM reconstruction in conjunction with a metal radial head implant, distal ulnar load was not only maintained at levels of the intact forearm but was decreased to levels significantly lower than those of the intact forearm as graft pretension was increased. We would expect similar or perhaps greater reductions in distal ulnar force with grafts of increased crosssectional area. It is of interest to understand how increasing graft pretension could lower distal ulnar force. We believe the explanation lies in our graft-tensioning method. As described in the Methods section we first secured the graft to the distal ulna and then passed it through the ulnar and radial tunnels. When pretensioning the graft we applied a distally directed force to the graft before fixing the sutures to the radius on a screw-post and washer construct distal to the tunnel hole. Because the DRUJ was not fixed and the radius was free to displace distally relative to the ulna as the pretension was applied we believe the wrist became more ulnar-negative before mechanical testing. Although we did not measure ulnar variance directly or radiographically before or after applying pretension our data show that distal ulnar load decreased significantly from 24% to 9% as graft pretension increased from 0 N to 45 N. To quantify the magnitude– effect relationship between graft pretension and ulnar variance in future studies radiographic measurements of ulnar variance before and after graft pretensioning would be valuable. Data were collected with the elbow in the neutral varus–valgus position and the radial head implant removed. This ensured that there was no contact between the proximal radius and the capitellum and thus no osseous restriction of proximal radial migration that could affect load sharing at the wrist. Although a metal radial head replacement can restore the longitudinal axis of the forearm in the valgus elbow position after an Essex-Lopresti injury the gap that is created between the capitellum and the radial head in the varus elbow position allows proximal radial displacement as the forearm is loaded axially through the wrist. In our study sectioning the IOM with an implant in place increased significantly the distal ulnar load from 23% to 30% as the radial head implant was free to migrate proximally; this
341
effect was not seen when the elbow was in valgus alignment and the implant was in contact with the capitellum. It can be inferred that ulnocarpal impaction and its clinical sequelae are at least in part due, to the relative increase in ulnar positivity at the wrist and the corresponding increased distal ulnar load seen after IOM rupture.6,7,14,26 Our results show that a metal radial head replacement coupled with IOM graft reconstruction can reduce distal ulnar force better than metal radial head replacement alone. If proximal radial migration can be limited effectively— particularly in the varus elbow position—and the wrist does not become relatively more ulnar positive then the clinical sequelae from ulnocarpal impaction potentially can be avoided. There are numerous potential drawbacks to the clinical application of our IOM reconstruction method. A significant amount of volar soft tissue dissection is required for installing and pretensioning the graft. During surgery there is a risk of injury to the anterior and posterior interosseous neurovascular bundles. After surgery forearm stiffness and limitation of forearm rotation secondary to scarring at the graft site may occur. Additionally simultaneous pretensioning and fixation of the graft to the screw-post and washer construct was particularly challenging given the small soft tissue window within which we were working. With application of the highest pretension (44 N) we observed the radius being pulled toward the ulna distally, thereby disrupting the parallel relationship between the 2 bones and potentially altering forearm rotation. The implications of potentially overreducing the distal ulnar force with this IOM reconstruction method are unknown. In theory significant alterations in the ulnar variance could disrupt the congruent articulation of the ulnar head with the sigmoid fossa of the distal radius, leading to pain at the DRUJ and problems with forearm rotation. Regardless of the graft tissue used for the IOM reconstruction graft abrasion, fraying, attenuation, and even rupture could occur without adequate burring of the tunnel holes in the radius and ulna. Gaining access to the inner tunnel holes of both the radius and ulna for burring of the tunnel edges was particularly difficult through our surgical approach. A primary goal in the treatment of Essex-Lopresti injury is to preserve the longitudinal axis of the forearm, thereby preventing changes in load sharing at the wrist. Although metallic radial head replacement is effective in limiting proximal radial migration in the valgus elbow position, in the varus position the radius is free to migrate under axial loading
342
The Journal of Hand Surgery / Vol. 30A No. 2 March 2005
when the IOM is injured. The resultant increased ulnar-positive variance at the wrist, which may first manifest dynamically, can lead ultimately to painful ulnar-sided wrist degeneration and instability. There are little published data showing the ability of the IOM to heal after it is injured with or without immobilization of the forearm. Our study shows the potential ability of an IOM reconstruction with a graft substitute to not only maintain the load-sharing profile at the wrist but actually to unload the distal ulna. Although the palmaris longus tendon autograft was used in this study a more robust graft substitute implemented with our pretensioning method could provide potentially a more durable reconstruction. Further study is warranted.
11.
12.
13.
14.
15.
16.
References 1. Curr JF, Coe WA. Dislocation of the inferior radio-ulnar joint. Br J Surg 1946;34:74 –77. 2. Essex-Lopresti P. Fractures of the radial head with distal radio-ulnar dislocation. Report of two cases. J Bone Joint Surg 1951;33B:244 –247. 3. Trousdale RT, Amadio PC, Cooney WP, Morrey BF. Radioulnar dissociation. A review of twenty cases. J Bone Joint Surg 1992;74A:1486 –1497. 4. Friedman SL, Palmer AK, Short WH, Levinsohn EM, Halperin LS. The change in ulnar variance with grip. J Hand Surg 1993;18A:713–716. 5. Tomaino MM. The importance of the pronated grip x-ray view in evaluating ulnar variance. J Hand Surg 2000;25A: 352–357. 6. Rozental TD, Beredjiklian PK, Bozentka DJ. Longitudinal radioulnar dissociation. J Am Acad Orthop Surg 2003;11(1): 68 –73. 7. Hotchkiss RN. Injuries to the interosseous ligament of the forearm. Hand Clin 1994;10:391–398. 8. Sellman DC, Seitz WH Jr, Postak PD, Greenwald AS. Reconstructive strategies for radioulnar dissociation: a biomechanical study. J Orthop Trauma 1995;9:516 –522. 9. Skahen JR 3rd, Palmer AK, Werner FW, Fortino MD. Reconstruction of the interosseous membrane of the forearm in cadavers. J Hand Surg 1997;22A:986 –994. 10. Ruch DS, Change DS, Koman LA. Reconstruction of longitudinal stability of the forearm after disruption of in-
17. 18.
19.
20.
21.
22.
23.
24. 25.
26.
terosseous ligament and radial head excision (Essex-Lopresti lesion). J South Orthop Assoc 1999;8:47–52. Knight DJ, Rymaszewski LA, Amis AA, Miller JH. Primary replacement of the fractured radial head with a metal prosthesis. J Bone Joint Surg 1993;75B:572–576. Markolf KL, Lamey DL, Yang S, Meals R, Hotchkiss R. Radioulnar load-sharing in the forearm. A study in cadavera. J Bone Joint Surg 1998;80A:879 – 888. Markolf KL, Dunbar AM, Hannani K. Mechanisms of load transfer in the cadaver forearm: role of the interosseous membrane. J Hand Surg 2000;25A:674 – 682. Shepard MF, Markolf KL, Dunbar AM. Effects of radial head excision and distal radial shortening on load-sharing in cadaver forearms. J Bone Joint Surg 2001;83A:93–100. Shepard MF, Markolf KL, Dunbar AM. The effects of partial and total interosseous membrane transection on load sharing in the cadaver forearm. J Orthop Res 2001;19:587– 592. Kocher T. Textbook of operative surgery. 3rd ed [translated by Stiled HJ, Paul CB]. New York: Macmillan, 1911. Henry AK. Extensile exposure. 2nd ed. Baltimore: Williams and Wilkins, 1970:100. Thompson JW, McBatts J, Danforth CH. Hereditary and racial variation in the musculus palmaris longus. Am J Phys Anthropol 1921;4:205–218. Thompson NW, Mockford BJ, Cran GW. Absence of palmaris longus muscle: a population study. Ulster Med J 2001;70(1):22–24. Hotchkiss RN, An KN, Sowa DT, Basta S, Weiland AJ. An anatomic and mechanical study of the interosseous membrane of the forearm: pathomechanics of proximal migration of the radius. J Hand Surg 1989;14A:256 –261. Morrey BF, An KN, Stormont TJ. Force transmission through the radial head. J Bone Joint Surg 1988;70A:250 – 256. Goldberg I, Peylan J, Yosipovitch Z. Late results of excision of the radial head for an isolated closed fracture. J Bone Joint Surg 1986;68A:675– 679. Brockman EP. Two cases of disability at the wrist joint following excision of the head of the radius. Proc R Soc Med 1930;24:904 –905. McGinley JC, Kozin SH. Interosseous membrane anatomy and functional mechanics. Clin Orthop 2001;383:108 –122. Mikic ZD, Vukadinovic SM. Late results in fractures of the radial head treated by excision. Clin Orthop 1983;181:220 – 228. Drobner WS, Hausman MR. The distal radioulnar joint. Hand Clin 1992;8:631– 644.
Purpose: Longitudinal radioulnar dissociation (Essex-Lopresti injury) occurs when traumatic axial loading through the wrist disrupts the interosseous membrane (IOM) of the forearm and fractures the radial head. Proximal migration of the radius results in an ulnar-positive wrist, which can lead to painful ulnar-sided wrist degeneration and distal radioulnar joint instability. The purpose of this study was to measure the ability of an IOM reconstruction used in combination with a metal prosthetic radial head implant to reduce distal ulnar forces in a cadaveric model. The effects of varying the initial graft pretension on distal ulnar force were also studied. Methods: Twelve fresh frozen and thawed cadaveric forearms had a miniature load cell installed to record force in the distal ulna as the wrist was loaded axially to 134 N of compression force in neutral rotation. Intact forearms were tested first with the elbow in valgus and varus alignments. Loading tests were repeated after (1) insertion of a metal radial head implant that restored radius anatomic length, (2) excision of the IOM (with a radial head implant), and (3) reconstruction of the IOM using a palmaris longus tendon autograft (with a radial head implant). The implant then was removed and loading tests were repeated using 3 levels of initial graft pretension. Results: Mean distal ulnar forces with an intact forearm were 23% of applied wrist force in the varus alignment and 12% in the valgus alignment. Mean force levels after insertion of the implant were 18% (varus) and 13% (valgus); these were not significantly different from corresponding values for the intact forearm. Mean force levels after section of the IOM were 30% (varus) and 14% (valgus); these were not significantly different from corresponding values for the intact forearm (varus and valgus) but the mean for varus was significantly greater than the corresponding value with an implant. After IOM reconstruction with a palmaris longus tendon tensioned to 22 N mean distal ulnar forces were 8% (varus) and 7% (valgus); these means were significantly less than the corresponding values for all prior test conditions. With the radial head removed increasing the level of graft pretension reduced significantly mean distal ulnar force.
From the Department of Orthopaedic Surgery, Biomechanics Research Section, David Geffen School of Medicine, University of California Los Angeles, Los Angeles, CA. Received for publication February 3, 2004; accepted in revised form July 7, 2004. No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article. Reprint requests: Keith L. Markolf, PhD, UCLA Dept of Orthopaedic Surgery, Biomechanics Research Section, Rehabilitation Building, 1000 Veteran Ave, Rm 21-67, Los Angeles, CA 90095. Copyright © 2005 by the American Society for Surgery of the Hand 0363-5023/05/30A02-0016$30.00/0 doi:10.1016/j.jhsa.2004.07.022
The Journal of Hand Surgery
335
336
The Journal of Hand Surgery / Vol. 30A No. 2 March 2005
Conclusions: With the IOM resected insertion of a metal radial head implant alone did not reduce distal ulnar forces to intact forearm levels. When an IOM reconstruction was performed in combination with the implant mean distal ulnar force was reduced significantly to a level below that for the intact forearm. Applying pretension to the graft displaced the radius distally thereby making the wrist more ulnar negative and reducing distal ulnar force. Our results suggest that an IOM reconstruction used in combination with a metal radial head implant theoretically could help reduce distal ulnar impaction in an Essex-Lopresti injury. (J Hand Surg 2005;30A:335–342. Copyright © 2005 by the American Society for Surgery of the Hand.) Key words: Essex-Lopresti, graft reconstruction, interosseous membrane, radial head replacement.
In 1946 Curr and Coe1 described fracture-dislocation of the radial head in conjunction with distal radioulnar joint (DRUJ) dislocation. Today this pattern of injury is named after Essex-Lopresti, who in 1951 described traumatic axial loading through the wrist resulting in DRUJ instability, interosseous membrane (IOM) rupture, and fracture of the radial head.2 In 1992 Trousdale et al3 reported on proximal migration of the radius after comminuted radial head fracture and IOM injury, giving the condition the term longitudinal radioulnar dissociation (LRUD). During gripping activities proximal migration of the radius results in an ulnar-positive wrist with subsequent ulnocarpal impaction, painful ulnar-sided wrist degeneration, and DRUJ instability.4 –7 Interosseous membrane reconstruction has been studied in cadavers and performed in patients for the treatment of LRUD3,7–10 (Pfaeffle et al, presented at the 49th annual meeting of the Orthopaedic Research Society, 2003; Osterman et al, presented at the 67th annual meeting of the American Academy of Orthopaedic Surgeons, 2000). Radial head replacement with a metal prosthesis also has been recommended in cases of comminuted radial head fracture in which internal fixation was not feasible.6,8,11 The purpose of this study was to measure the ability of an IOM graft reconstruction to alter distal ulnar forces when used in combination with a metallic radial head implant. The experimental objectives were to measure directly distal ulnar forces at the wrist as the forearm was loaded axially to 134 N under the following test conditions: forearm with IOM intact, after removal of the radial head and insertion of a radial head implant that restored anatomic radial length (IOM sectioned), after IOM graft reconstruction (radial head implant in place), and after varying the initial graft pretension (radial head removed).
Methods Twelve fresh frozen and thawed cadaver forearms were used for this study. The donors ranged from 65 to 91 years of age (mean, 76 years); 8 were men and 4 were women. All cadaver forearm radiographs showed normal anatomy without significant prior wrist, forearm, or elbow pathology. All forearms were potted proximally at the distal humerus and distally using the central 3 metacarpals. The first and fifth metacarpals were removed. A specially designed miniature load cell was connected to prongs cemented into the distal part of the ulna for measurement of distal ulnar force. The potted humerus was mounted to a fixture attached to the crosshead of a materials testing system (MTS) machine (Model 812, MTS Systems Corp, Minneapolis, MN); and the elbow was flexed to 90° (Fig. 1). Load was applied to the potted central 3 metacarpals of the hand at a rate of approximately 1 mm/s; the maximum load applied was 134 N. All testing was performed with the wrist in neutral flexion– extension which was established by aligning the plane of the 3 potted metacarpals with the flexion– extension plane of the elbow. Prior studies from this laboratory have shown that load sharing at the wrist is affected by the varus– valgus position of the elbow.12–15 Valgus testing position was defined by applying a 2.5-Nm moment to the elbow that forced the radial head to contact the capitellum. Varus testing position was defined by applying a 2.5-Nm moment to the elbow that created a gap between the radial head and the capitellum. Further details on specimen mounting, load cell installation, and testing procedure are documented in our prior publications.12,14 The radial head implant used for this study (Evolve Modular Radial Head System, Wright Medical Technology, Arlington, TN) is a 2-part modular prosthesis with separate stem and head components of varying lengths and diameters. Using a hand-held micrometer with an accuracy of 0.05 mm we took a
Tejwani, Markolf, and Benhaim / IOM Reconstruction With Radial Head Replacement
Figure 1. A forearm specimen installed into the MTS machine.
direct measurement of the distance between the articular surface of the radial head and a reference mark on the proximal radius at a fixed distance from the radial head; this was used to establish anatomic radial length. After a transverse cut was made in the radial neck with an oscillating saw an implant was selected that restored anatomic length and matched the diameter of the radial head articular surface. In the event that the trial implant thickness selected did not restore precisely full anatomic length ring-shaped disk spacers were placed over the stem beneath the
337
head of the implant to restore the reconstructed radial length to within approximately 0.5 mm of that for the intact specimen. Because of ease of assembly and implant insertion trial aluminum implants identical in geometry to the actual implants were used in this study. The intact forearms first were loaded axially to 134 N in both varus and valgus alignments. The radial head was exposed and excised through the posterolateral approach described by Kocher.16 The trial radial head implant then was inserted as described above. The volar aspect of the forearm was dissected through the approach described by Henry17 and the IOM was exposed in its entirety. The central band of the IOM was found consistently to be more robust than the surrounding thinner, membranous tissue that composed the proximal and distal portions of the IOM. The central portion of the central band insertion was identified on both the radius and ulna and 3.0-mm drill holes were made in both bones to allow ultimately for tendon graft passage. The drill holes were oriented transverse to the longitudinal axis of the radius and ulna and were placed in the midaxial line of each bone. The IOM was sectioned in its entirety from distal to proximal under direct visualization and the forearms were loaded for 1 cycle to 134 N in both varus and valgus alignments. Of note, the tunnel holes were made in the radius and ulna before IOM sectioning to ensure accuracy in the identification of the orientation of the central band. Through a separate incision the palmaris longus tendon was identified and harvested from its characteristic superficial location between the flexor carpi radialis (FCR) and flexor digitorum superficialis tendons. The attached palmaris longus muscle was dissected free and each end of the tendon was fixed with a Bunnell stitch using a no. 2 nonabsorbable suture (Ethicon, Somerville, NJ). As expected based on previous publications, of the 12 forearms used in this testing sequence the palmaris longus tendon was absent in 2 (17%); a replacement palmaris longus tendon was harvested from a similarly sized forearm in these cases.18,19 Soft tissue was dissected from both sides of the bone tunnels in the radius and ulna and the edges were rounded with a high-speed burr. The palmaris graft first was secured to the distal ulna by wrapping the suture around a screw-post and washer construct. The graft was then passed through the ulnar and radial tunnels using a wire-loop suture passer. After the palmaris graft was pulled through the radial tunnel it was pretensioned to 22 N as verified by a
338
The Journal of Hand Surgery / Vol. 30A No. 2 March 2005
Figure 2. The palmaris longus tendon IOM graft reconstruction performed with a tunnel method and screw-post and washer fixation to the radius and ulna.
hand-held calibrated spring scale with a resolution of 1.1 N. The tendon graft was then secured to a screwpost and washer construct located approximately 1 to 2 cm distal to the radial tunnel (Fig. 2). All palmaris grafts in this study were preconditioned with 10 cycles of axial loading to 45 N by using the MTS machine. Preliminary testing in our laboratory showed that without burring the tunnel holes at the site of tendon entry the palmaris tendon graft was unable to undergo 10 cycles of preconditioning as a result of premature graft rupture. No grafts ruptured after burring of the tunnel holes. After preconditioning palmaris graft pretension was reset to 22.3 N and the forearms were then loaded for 1 cycle in both varus and valgus alignments. Finally the radial head implant was removed and the palmaris tendon graft was pretensioned to a zero load (0 N), defined as a taut graft with no additional preload placed on it. Loading tests were repeated with 0 N, 22 N, and 45 N of pretension applied to the graft. At each pretension level the forearm was
loaded axially for 1 cycle; the elbow was in a position midway between the varus–valgus positions determined for the intact forearm. A 1-way repeated-measures analysis of variance model was used to determine the significance of differences in mean distal ulnar forces between the following conditions: intact specimen, implant with intact IOM, implant with sectioned IOM, and implant with IOM reconstruction. A similar model was used to compare mean distal ulnar forces between the following conditions (radial head removed): intact IOM and graft pretensioned to 0 N, 22 N, and 45 N. Multiple pairwise comparisons between means were made using the Student-Neuman-Keuls procedure. The level of significance was p ⬍ .05.
Results Mean distal ulnar forces with an intact forearm were 23% of applied wrist force in varus alignment and 12% in valgus alignment (Fig. 3). Mean force levels after insertion of the implant were 18%
Tejwani, Markolf, and Benhaim / IOM Reconstruction With Radial Head Replacement
339
Discussion
Figure 3. Distal ulnar force recorded at 134 N of applied wrist load (expressed as a percentage of load applied to the wrist) with varus and valgus elbow alignments for 4 test conditions: intact forearm, radial head implant, implant with IOM sectioned, and implant with IOM reconstruction (pretension set at 22.3 N). Mean values are shown with standard deviations indicated by error bars. All comparisons between test conditions are significantly different from each other (p ⬍ .05) except where indicated by “n.s.”.
(varus) and 13% (valgus); these were not significantly different from corresponding values for the intact forearm (Fig. 3). Mean force levels after section of the IOM (implant in place) were 30% (varus) and 14% (valgus); these were not significantly different from corresponding values for the intact forearm (varus and valgus) but the mean for varus was significantly greater than the corresponding value with an implant (Fig. 3). After IOM reconstruction with a palmaris longus tendon tensioned to 22 N mean distal ulnar forces were 8% (varus) and 7% (valgus); these means were significantly less than the corresponding values for all prior test conditions (Fig. 3). With the radial head resected (implant removed) mean distal ulnar force with an intact IOM was 24% of applied wrist force (Fig. 4); mean force was not significantly changed after IOM reconstruction with a palmaris tendon graft that was tightened slightly before fixation but not tensioned intentionally (0 N) (Fig. 4). Pretensioning the graft to 22 N reduced mean distal ulnar force to 15% of applied wrist force; this was significantly less than mean force for the intact-IOM and 0-N–pretension conditions (Fig. 4). Pretensioning the graft to 45 N further reduced mean distal ulnar force to 9% of applied wrist force; this was significantly less than means for intact-IOM, 0-N, and 22 N pretension conditions (Fig. 4).
The primary restraint to proximal migration of the radius is the radial head as it contacts the capitellum.20 –22 In 1930 Brockman23 described proximal translation of the radius 3 months after radial head fracture. In 1992 Trousdale et al3 reported on 20 cases of proximal migration of the radius after comminuted radial head fracture, naming the pattern of injury LRUD. Longitudinal radioulnar dissociation can occur acutely with an Essex-Lopresti injury or late after the IOM attenuates with time.3,6,24,25 Shepard et al14 showed that for each 1 mm of radial shortening after radial head excision the distal ulnar load increased by 10%. Drobner and Hausman26 concluded similarly that a positive ulnar variance from proximal radial migration of 2 mm increases distal ulnar load by 20% to 40%. Early on this can manifest during activities involving a strong grip when there is a dynamic increase in ulnar-positive variance.4,5 With more severe radial migration the new distal ulnar position can limit forearm supination and wrist extension as the carpus impinges on the ulnar head; wrist pain, DRUJ instability, and decreased wrist and forearm motion may result.6,7
Clinical Studies of IOM Reconstruction Hotchkiss7 attempted IOM reconstruction in 1 patient using a palmaris tendon weave coupled with cross-pinning of the radius and ulna. The clinical
Figure 4. Distal ulnar force recorded at 134 N of applied wrist load (expressed as a percentage of load applied to the wrist) for the following test conditions (radial head removed): native IOM and after reconstruction of the IOM with a palmaris longus tendon autograft pretensioned to 0 N, 22 N, and 45 N. All comparisons between test conditions are significantly different from each other (p ⬍ .05) except where indicated by “n.s.”.
340
The Journal of Hand Surgery / Vol. 30A No. 2 March 2005
history of the patient and the precise surgical method were not described. Immobilization was performed for 8 weeks, after which significant proximal migration of the radius occurred indicating failure of the reconstruction. A radioulnar synostosis was performed subsequently as a salvage procedure. In 1 patient with chronic LRUD and dynamic ulnar-positive variance secondary to a failed silicone radial head replacement (Silastic). Ruch et al10 performed a triangular fibrocartilage complex repair, IOM reconstruction with bone–patellar tendon– bone autograft, and titanium radial head replacement with good results reported at 2 years. Osterman et al reconstructed prospectively the IOM in 14 patients with chronic LRUD using a central-third bone–patellar tendon– bone graft; no radial head replacement was performed (Osterman et al, presented at the 67th annual meeting of the American Academy of Orthopaedic Surgeons, 2000). An average 3 mm of ulnar-positive variance existed and an ulnar shortening osteotomy (mean, 4 mm) was performed on each patient. At follow-up examination (average time, 32 months) pain improved in all patients, elbow motion was preserved in 13 patients, forearm rotation loss was less than 20° in 11 patients, and grip strength improved from 62% to 85% of normal. No patient required additional surgery for increased ulnar-positive variance secondary to proximal radial migration. Postsurgical radiographs showed maintenance of ulnar variance in 11 of the 14 patients.
Biomechanical Studies of IOM Reconstruction Sellman et al8 reconstructed the central band of the IOM in 9 cadaver forearms using a 16-mm braided polyester cord with a triple helical weave in a tunnel method similar to the one used in this study; the cord pretension was not specified. Sectioning the central 10 cm of the IOM decreased gross forearm stiffness to 70% that of the intact forearm. With 1 cycle of loading to 75 N after IOM reconstruction and radial head excision gross axial stiffness was 94% that of the intact forearm. With the addition of a radial head replacement (Silastic) to the IOM reconstruction gross forearm stiffness remained at 94%. After the radial head implant was replaced with one made of titanium gross forearm stiffness increased to 145%. Distal ulnar force was not measured, nor were the effects of variable graft pretension. In a radial head– deficient cadaver forearm model Skahen et al9 reported that based on preliminary
testing IOM reconstruction with the palmaris longus tendon was unable to limit proximal radial migration with an applied wrist load of 89 N. After radial head excision the triangular fibrocartilage complex and IOM were sectioned incrementally from proximal to distal in 6 specimens. After 1 cycle of loading to 89 N in neutral forearm rotation the FCR IOM reconstruction decreased proximal radial migration from 12 mm to 6 mm. In another 6 specimens the sequence of sectioning of the IOM and triangular fibrocartilage complex was reversed to distal to proximal. In this scenario after FCR IOM reconstruction proximal radial head migration was decreased from 10 mm to 3 mm. The researchers concluded that a-FCR tendon autograft used in IOM reconstruction was capable of preventing complete migration of the proximal radius to the capitellum but was not capable of restoring completely the longitudinal stability of the forearm. Pfaeffle et al compared single- and double-bundle FCR tendon autograft IOM reconstructions in 12 cadaver forearms; the grafts were passed through tunnels and fixed to a screw-post construct (Pfaeffle et al, presented at the 49th annual meeting of the Orthopaedic Research Society, 2003). Fifteen newtons of cyclic preconditioning was applied and the grafts were pretensioned to 90 N. After 1 cycle of loading to 139 N the researchers concluded that the double-bundle FCR reconstruction restored successfully normal forearm mechanics in LRUD based on the load-sharing profile of the radius and ulna. The radial head was left intact throughout testing, however, and proximal migration of the radius did not change between the intact, IOM-sectioned, and IOMreconstructed conditions, varying from 0.1 to 0.3 mm. This suggests that load was applied in a valgus position with the radial head contacting the capitellum. In this elbow position axial load would not adequately test the ability of the IOM reconstruction to limit proximal migration of a mobile radius as is the case in the varus position with a gap between the radial head and capitellum. For the purposes of the present study the palmaris longus tendon autograft was selected to reconstruct the IOM because of its general expendability and frequent use in upper- extremity tendon reconstruction procedures. Skahen et al,9 however, reported that based on preliminary testing IOM reconstruction with the palmaris longus tendon was unable to limit proximal radial migration under 89 N of applied wrist load. In our tests the palmaris graft underwent only 1 cycle of loading with 134 N of applied wrist force and could be expected to attenuate under fur-
Tejwani, Markolf, and Benhaim / IOM Reconstruction With Radial Head Replacement
ther cycles of loading. Other acceptable graft substitutes with larger cross-sectional areas such as FCR tendon autograft and bone–patellar tendon– bone allograft may be better suited to reconstructing the IOM clinically. Nevertheless we found that when the relatively smaller palmaris tendon was used for IOM reconstruction in conjunction with a metal radial head implant, distal ulnar load was not only maintained at levels of the intact forearm but was decreased to levels significantly lower than those of the intact forearm as graft pretension was increased. We would expect similar or perhaps greater reductions in distal ulnar force with grafts of increased crosssectional area. It is of interest to understand how increasing graft pretension could lower distal ulnar force. We believe the explanation lies in our graft-tensioning method. As described in the Methods section we first secured the graft to the distal ulna and then passed it through the ulnar and radial tunnels. When pretensioning the graft we applied a distally directed force to the graft before fixing the sutures to the radius on a screw-post and washer construct distal to the tunnel hole. Because the DRUJ was not fixed and the radius was free to displace distally relative to the ulna as the pretension was applied we believe the wrist became more ulnar-negative before mechanical testing. Although we did not measure ulnar variance directly or radiographically before or after applying pretension our data show that distal ulnar load decreased significantly from 24% to 9% as graft pretension increased from 0 N to 45 N. To quantify the magnitude– effect relationship between graft pretension and ulnar variance in future studies radiographic measurements of ulnar variance before and after graft pretensioning would be valuable. Data were collected with the elbow in the neutral varus–valgus position and the radial head implant removed. This ensured that there was no contact between the proximal radius and the capitellum and thus no osseous restriction of proximal radial migration that could affect load sharing at the wrist. Although a metal radial head replacement can restore the longitudinal axis of the forearm in the valgus elbow position after an Essex-Lopresti injury the gap that is created between the capitellum and the radial head in the varus elbow position allows proximal radial displacement as the forearm is loaded axially through the wrist. In our study sectioning the IOM with an implant in place increased significantly the distal ulnar load from 23% to 30% as the radial head implant was free to migrate proximally; this
341
effect was not seen when the elbow was in valgus alignment and the implant was in contact with the capitellum. It can be inferred that ulnocarpal impaction and its clinical sequelae are at least in part due, to the relative increase in ulnar positivity at the wrist and the corresponding increased distal ulnar load seen after IOM rupture.6,7,14,26 Our results show that a metal radial head replacement coupled with IOM graft reconstruction can reduce distal ulnar force better than metal radial head replacement alone. If proximal radial migration can be limited effectively— particularly in the varus elbow position—and the wrist does not become relatively more ulnar positive then the clinical sequelae from ulnocarpal impaction potentially can be avoided. There are numerous potential drawbacks to the clinical application of our IOM reconstruction method. A significant amount of volar soft tissue dissection is required for installing and pretensioning the graft. During surgery there is a risk of injury to the anterior and posterior interosseous neurovascular bundles. After surgery forearm stiffness and limitation of forearm rotation secondary to scarring at the graft site may occur. Additionally simultaneous pretensioning and fixation of the graft to the screw-post and washer construct was particularly challenging given the small soft tissue window within which we were working. With application of the highest pretension (44 N) we observed the radius being pulled toward the ulna distally, thereby disrupting the parallel relationship between the 2 bones and potentially altering forearm rotation. The implications of potentially overreducing the distal ulnar force with this IOM reconstruction method are unknown. In theory significant alterations in the ulnar variance could disrupt the congruent articulation of the ulnar head with the sigmoid fossa of the distal radius, leading to pain at the DRUJ and problems with forearm rotation. Regardless of the graft tissue used for the IOM reconstruction graft abrasion, fraying, attenuation, and even rupture could occur without adequate burring of the tunnel holes in the radius and ulna. Gaining access to the inner tunnel holes of both the radius and ulna for burring of the tunnel edges was particularly difficult through our surgical approach. A primary goal in the treatment of Essex-Lopresti injury is to preserve the longitudinal axis of the forearm, thereby preventing changes in load sharing at the wrist. Although metallic radial head replacement is effective in limiting proximal radial migration in the valgus elbow position, in the varus position the radius is free to migrate under axial loading
342
The Journal of Hand Surgery / Vol. 30A No. 2 March 2005
when the IOM is injured. The resultant increased ulnar-positive variance at the wrist, which may first manifest dynamically, can lead ultimately to painful ulnar-sided wrist degeneration and instability. There are little published data showing the ability of the IOM to heal after it is injured with or without immobilization of the forearm. Our study shows the potential ability of an IOM reconstruction with a graft substitute to not only maintain the load-sharing profile at the wrist but actually to unload the distal ulna. Although the palmaris longus tendon autograft was used in this study a more robust graft substitute implemented with our pretensioning method could provide potentially a more durable reconstruction. Further study is warranted.
11.
12.
13.
14.
15.
16.
References 1. Curr JF, Coe WA. Dislocation of the inferior radio-ulnar joint. Br J Surg 1946;34:74 –77. 2. Essex-Lopresti P. Fractures of the radial head with distal radio-ulnar dislocation. Report of two cases. J Bone Joint Surg 1951;33B:244 –247. 3. Trousdale RT, Amadio PC, Cooney WP, Morrey BF. Radioulnar dissociation. A review of twenty cases. J Bone Joint Surg 1992;74A:1486 –1497. 4. Friedman SL, Palmer AK, Short WH, Levinsohn EM, Halperin LS. The change in ulnar variance with grip. J Hand Surg 1993;18A:713–716. 5. Tomaino MM. The importance of the pronated grip x-ray view in evaluating ulnar variance. J Hand Surg 2000;25A: 352–357. 6. Rozental TD, Beredjiklian PK, Bozentka DJ. Longitudinal radioulnar dissociation. J Am Acad Orthop Surg 2003;11(1): 68 –73. 7. Hotchkiss RN. Injuries to the interosseous ligament of the forearm. Hand Clin 1994;10:391–398. 8. Sellman DC, Seitz WH Jr, Postak PD, Greenwald AS. Reconstructive strategies for radioulnar dissociation: a biomechanical study. J Orthop Trauma 1995;9:516 –522. 9. Skahen JR 3rd, Palmer AK, Werner FW, Fortino MD. Reconstruction of the interosseous membrane of the forearm in cadavers. J Hand Surg 1997;22A:986 –994. 10. Ruch DS, Change DS, Koman LA. Reconstruction of longitudinal stability of the forearm after disruption of in-
17. 18.
19.
20.
21.
22.
23.
24. 25.
26.
terosseous ligament and radial head excision (Essex-Lopresti lesion). J South Orthop Assoc 1999;8:47–52. Knight DJ, Rymaszewski LA, Amis AA, Miller JH. Primary replacement of the fractured radial head with a metal prosthesis. J Bone Joint Surg 1993;75B:572–576. Markolf KL, Lamey DL, Yang S, Meals R, Hotchkiss R. Radioulnar load-sharing in the forearm. A study in cadavera. J Bone Joint Surg 1998;80A:879 – 888. Markolf KL, Dunbar AM, Hannani K. Mechanisms of load transfer in the cadaver forearm: role of the interosseous membrane. J Hand Surg 2000;25A:674 – 682. Shepard MF, Markolf KL, Dunbar AM. Effects of radial head excision and distal radial shortening on load-sharing in cadaver forearms. J Bone Joint Surg 2001;83A:93–100. Shepard MF, Markolf KL, Dunbar AM. The effects of partial and total interosseous membrane transection on load sharing in the cadaver forearm. J Orthop Res 2001;19:587– 592. Kocher T. Textbook of operative surgery. 3rd ed [translated by Stiled HJ, Paul CB]. New York: Macmillan, 1911. Henry AK. Extensile exposure. 2nd ed. Baltimore: Williams and Wilkins, 1970:100. Thompson JW, McBatts J, Danforth CH. Hereditary and racial variation in the musculus palmaris longus. Am J Phys Anthropol 1921;4:205–218. Thompson NW, Mockford BJ, Cran GW. Absence of palmaris longus muscle: a population study. Ulster Med J 2001;70(1):22–24. Hotchkiss RN, An KN, Sowa DT, Basta S, Weiland AJ. An anatomic and mechanical study of the interosseous membrane of the forearm: pathomechanics of proximal migration of the radius. J Hand Surg 1989;14A:256 –261. Morrey BF, An KN, Stormont TJ. Force transmission through the radial head. J Bone Joint Surg 1988;70A:250 – 256. Goldberg I, Peylan J, Yosipovitch Z. Late results of excision of the radial head for an isolated closed fracture. J Bone Joint Surg 1986;68A:675– 679. Brockman EP. Two cases of disability at the wrist joint following excision of the head of the radius. Proc R Soc Med 1930;24:904 –905. McGinley JC, Kozin SH. Interosseous membrane anatomy and functional mechanics. Clin Orthop 2001;383:108 –122. Mikic ZD, Vukadinovic SM. Late results in fractures of the radial head treated by excision. Clin Orthop 1983;181:220 – 228. Drobner WS, Hausman MR. The distal radioulnar joint. Hand Clin 1992;8:631– 644.