100

SECTION IIORIGINAL ARTICLES

Biomechanical Evaluation of the Elbow Using Roentgen

Stereophotogrammetric Analysis

Denise Eygendaal, MD, PhD*; Edward R. Valstar, MSc*; Jens O. Söjbjerg, MD**; and Piet M. Rozing, MD, PhD*

CLINICAL ORTHOPAEDICS AND RELATED RESEARCHNumber 396, pp. 100–105© 2002 Lippincott Williams & Wilkins, Inc.

The medial collateral ligament complex is theprimary constraint of the elbow to valgus forcesand is composed of the anterior bundle, the pos-terior bundle, and a transverse part. Total andpartial ruptures have been described. Clinicaland radiologic examinations of medial or valgusinstability of the elbow are difficult. The effect ofdifferent stages of medial collateral ligamentruptures on ulnohumeral movement in cadaverswas determined to rationalize the use of physi-cal and radiologic examinations in different

stages of valgus instability in vivo. Using roent-gen stereophotogrammetric analysis, motion isdetermined between the humerus and ulna un-der valgus load and between the humerus andradius during maximal pronation of the fore-arm after various dimensions of medial collat-eral ligament lesions. The increase in distancebetween the humerus and ulna under a 15 N val-gus load varied from 2.7 mm to 9.8 mm. The in-crease in distance between the humerus andproximal radius with the forearm in pronationin an intact specimen and after transsection ofthe anterior medial collateral ligament and pos-terior medial collateral ligament in the anteriordirection was 9.7 mm. These results suggest thatdetection of partial ruptures with clinical andradiologic examinations is difficult. Anteriormovement of the radial head can be used as anadditional parameter of valgus instability.

The medial collateral ligament complex, alsoreferred to as the ulnar collateral ligament, iscomposed of an anterior bundle, a posterior

From the *Department of Orthopaedics, Leiden Univer-sity Medical Center, Leiden, The Netherlands; and the**Department of Orthopaedics, University Hospital ofAarhus, Denmark.This study was financially supported by the Dutch AnnaFoundation.Reprint requests to Denise Eygendaal, MD, PhD, Mar-grietstraat 40, 5122 HV Rijen, The Netherlands.Received: June 5, 2000.Revised: March 7, 2001; June 1, 2001.Accepted: August 8, 2001.

bundle, and a transverse part. The transversepart does not span the elbow and cannot con-tribute to stability.4,5,20,26 The anterior bundleis the primary restraint to valgus and internalrotatory forces and can be subdivided into ananterior band and a posterior band. The poste-rior medial collateral ligament is a fan-shapedthickening of the capsule and was reported notto give any significant stability to the el-bow.1,3,24,26 Partial ruptures of the medial col-lateral ligament can cause subtle instability ofthe elbow and disabling pain in athletes en-gaged in overhead throwing. Posttraumatic to-tal rupture of the medial collateral ligamentcan result in frank subluxation.6–8,11,12,14

Distinction of tear size is considered impor-tant because complete and large tears requiresurgical correction, whereas small partial tearscan be treated conservatively.6,11,21,22 Clinicalexamination of medial or valgus instability ofthe elbow is difficult, and ancillary tests suchas dynamic radiography under valgus load ormagnetic resonance imaging are not reli-able.13,18,23,25 This makes valgus instability proneto delayed recognition and under-treatment.

Roentgen stereophotogrammetric analysishas been proven to be an accurate method formeasuring micromotion of prosthetic implantswith respect to the surrounding bone.9,23,27 Ac-curacy as much as 0.2 mm for translations and0.3� for rotations have been reported.9,23

In this study of medial instability of the el-bow, roentgen stereophotogrammetric analy-sis was used to detect motion after selectivetranssection of the medial collateral ligamentcomplex.

The aim of this cadaver study was to definethe kinematic changes under valgus load aftervarious transsections of the medial collateralligament to rationalize the use of a physicalexamination and valgus load radiography indifferent stages of medial collateral ligamentinjuries in vivo. Radial head motion was de-termined during maximal pronation aftertranssection of the medial collateral ligamentto detect an additional parameter, which mightfacilitate the diagnosis of medial collateral lig-ament insufficiency in vivo.

MATERIALS AND METHODS

Five consecutive fresh cadavers were used within 12hours after death. The donors were three males andtwo females (mean age, 71 years). Ten elbows, withnormal ranges of motion and clinical stability, werefixed in a commercially available stress device (Te-los GA/HE stress device, Austin & Associates, Fall-ston, MD). This device was equipped with a screw-threaded shaft that permits valgus stress to be appliedgradually; the applied force is monitored on a light-emitting digital readout. The elbow was flexed 25�,and the upper arm was externally rotated and ab-ducted 65�. The wrist was fixed in neutral position be-tween two roller bars, and the elbow was fixed to thecorpus throughout the experiment (Figs 1, 2).

Dissection was done using a muscle-splittingapproach through the flexor carpi ulnaris andpronator teres muscle just distal to the medial epi-condyle to observe the medial collateral ligamentwithout penetrating the joint capsule. In all elbows,the individual bands of the medial collateral liga-ment could be assessed. The length and width of theanterior medial collateral ligament were deter-mined using a caliper.

Roentgen stereophotogrammetric analysis is ahighly accurate measurement technique, which isused to calculate the three-dimensional motions of

Number 396March, 2002 Biomechanical Evaluation of the Elbow 101

Fig 1. The Telos stress device consists of ascrew-threaded shaft that permits valgus stressto be applied. The force is applied at the lateralepicondyle of the elbow and monitored on a light-emitting digital readout. The elbow was flexed25�, and the upper arm was externally rotatedand abducted 65�. The wrist was fixed betweentwo roller bars, and the elbow was fixed to thecorpus throughout the experiment.

bony structures. To obtain high accuracy, artificiallandmarks have to be used. In this study, three, thin,transparent Plexiglas disks (Medis, Leiden, TheNetherlands) with metal screw-threaded pins con-taining eight tantalum beads each, developed by thetechnical department of Leiden University MedicalCentre, were used as artificial landmarks for thethree bony elements. The disks were rigidly fixedto the bone of the medial epicondyle, proximal ra-dial shaft, and base of the coronoid process of theulna by three separate percutaneous incisions.

The roentgen stereophotogrammetric analysisset-up consisted of two synchronized roentgentubes positioned at approximately 1.50 m above aPlexiglas calibration cage (Medis) with 1-mm tan-talum balls. Both roentgen tubes made an angle of20� with the vertical. An xray cassette holding afilm was inserted underneath the calibration cage.The elbow specimen from a cadaver was posi-tioned above the Plexiglas calibration cage (Fig 3).

In the first five specimens, the anterior band ofthe anterior medial collateral ligament was tran-sected, followed by transsection of the posterior

band of the anterior medial collateral ligament, andfinally the posterior medial collateral ligament. Inthe other five specimens, the posterior band of theanterior medial collateral ligament was transected,followed by transsection of the anterior band of theanterior medial collateral ligament, and finally theposterior medial collateral ligament, because theposterior medial collateral ligament has beenproven to be a secondary stabilizer.1–3

Synchronized radiographs were taken in an un-loaded condition with 15 N force applied to the lat-eral epicondyle of the humerus in the intact situationand after selective transsection of the medial collat-eral ligament. Radiographs with the forearm inmaximal internal rotation, without valgus stress,were taken in the intact situation and after transsec-tion of the anterior medial collateral ligament andthe posterior medial collateral ligament. Motion be-tween the humerus and ulna was determined in themediolateral direction, craniocaudal direction, andanteroposterior (AP) direction. Motion between thehumerus and proximal radius was determined in theprevious mentioned directions with the arm in full

Clinical Orthopaedics102 Eygendaal et al and Related Research

Fig 3. The experimental setup is shown with theelbow positioned above the calibration cage inwhich an xray cassette can be inserted. Two syn-chronized roentgen tubes are positioned 1.50 mabove the cage.

Fig 2. Screw-threaded shaft: 1 � turning grip; 2� threaded spindle; 3 � release button for freemotion; 4 � charging indicator lamp; 5 � socketfor charging equipment; 6 � digital readout; and7 � front cushion pad.

Number 396March, 2002 Biomechanical Evaluation of the Elbow 103

pronation in the intact situation and after total trans-section of the medial collateral ligament.

Translations were determined using roentgenstereophotogrammetric analysis software (Medis).This software automatically does fast and accuratemeasurements in digitized radiographs.27 Markersare automatically detected and labeled.

A paired Student’s t test was applied to the re-sults of the measurements of this study, in whichthe intact situation was compared with several lev-els of transsection using the SPSS software pro-gram (SSPS Inc, Chicago, IL). The results are ex-pressed as mean values and standard deviations.

RESULTS

The mean length of the anterior medial collat-eral ligament, measured with a caliper, was 26mm (range, 24–31 mm), and the mean widthof the midportion of the anterior medial col-lateral ligament was 5 mm (range, 4–7 mm).Distinction between the anterior and posteriorbands of the anterior medial collateral liga-ment could be made in all specimens but weredifficult to measure with a caliper. The poste-rior medial collateral ligament is more like a

thickening of the capsule and therefore couldnot be measured.

Motion Between the Humerus and UlnaThe quality of the roentgen stereophotogram-metric analysis radiographs in two of the spec-imens was insufficient to determine the in-crease of the distance between the humerusand ulna with an intact anterior medial collat-eral ligament. Therefore, this increase wasbased on the results in eight specimens. In onespecimen, the increase of the humeroulnar dis-tance after complete transsection could not bedetermined. The results of nine specimenswere used to determine the increase of dis-tance after complete transsection of the ante-rior medial collateral ligament.

Motion in the mediolateral direction wasless than 1.2 mm in all cases, and motion in theAP direction did not exceed 0.7 mm. Motion inthe craniocaudal direction, which representsmedial joint space widening, increased from2.7 mm in the intact situation to 9.8 mm aftertotal transsection of the medial collateral liga-ment (Table 1). Transsection of the anterior part

TABLE 1. Average Increase of Distance Between the Humerus and Ulna inMillimeters

Mean/Standard

Level of Transsection Direction N Minimum Maximum Deviation

Anterior medial collateral ligament ML 8 0.0 1.2 0.5/0.43intact, valgus load 15 N CC 8 1.6 3.7 2.7/0.96

AP 8 �0.7 �0.2 �0.3/0.27Partial (50%) transsection anterior part ML 5 0.0 1.1 �0.3/0.23

anterior medial collateral ligament first CC 5 5.1 5.8 5.5/0.31AP 5 �0.7 0.0 �0.3/0.23

Partial (50%) transsection posterior part ML 5 0.0 1.1 0.3/0.23anterior medial collateral ligament first CC 5 1.6 2.0 1.8/0.14

AP 5 �0.7 0.0 �0.3/0.23Complete (100%) transsection anterior ML 9 0.0 1.1 0.5/0.37

medial collateral ligament CC 9 6.3 8.6 7.4/0.77AP 9 �0.7 �0.1 �0.5/0.25

Complete (100%) transsection anterior ML 10 0.1 1.2 0.9/0.37medial collateral ligament and posterior CC 10 7.0 12.1 9.8/1.97medial collateral ligament AP 10 �0.7 0.7 �0.1/0.54

ML � mediolateral direction; CC � craniocaudal direction; AP � anteroposterior direction.

All p values were � 0.001. Negative values refer to motion in posterior direction.

of the anterior medial collateral ligament re-sulted in a 5.5-mm increase in distance in thecraniocaudal direction. Transsection of the pos-terior part of the anterior medial collateral liga-ment resulted in a much smaller increase of dis-tance between the humerus and ulna (0.5 mm).

Motion Between Humerus and RadialHeadIn one specimen, the quality of roentgen stereo-photogrammetric analysis radiographs was in-sufficient to determine the increase of distancebetween the humerus and radial head. The in-crease of distance between the humerus andradial head is based on the measurements ofnine specimens. The increase of distance be-tween the humerus and proximal radius in themediolateral and craniocaudal directions wasnot considered relevant: it was less than 1 mm.Transsection of the anterior medial collateralligament and the posterior medial collateralligament resulted in displacement in the AP di-rection of 9.7 mm (standard deviation, 0.61mm) compared with the intact situation undermaximal internal rotation (p � 0.001; Table 2).

DISCUSSION

The magnitude of motion in the craniocaudaldirection after transsection of two differentparts of the anterior medial collateral ligamentobserved in this study confirms the biome-chanical importance of different parts of theanterior medial collateral ligament in valgusstability, as described in previous reports.1–3

The anterior portion of the anterior medial col-lateral ligament is the primary constraint tovalgus loads, and the posterior part is a sec-ondary constraint.1–3 The anterior and poste-rior bands show consistent reciprocal tighten-ing and can be injured separately, dependingon the angle of flexion during injury. Lesionsof the posterior part are more difficult to detectthan lesions of the anterior part based on thisstudy, because the increase in craniocaudal di-rection is much smaller after transsection ofthe posterior part.

In this study, the posterior medial collateralligament was observed to be a tertiary con-straint to valgus load. This tertiary stabilizingfunction of the posterior medial collateral lig-ament is in contrast with some previous stud-ies where it was reported that this bundle hada negligible stabilizing effect.15–17,22,24

Based on the current experimental data,rupture of the posterior part of the anterior me-dial collateral ligament is difficult to detect us-ing the increase of medial joint widening atclinical examination or dynamic elbow radi-ography using 15 N valgus load, because thisincrease is only 0.5 mm. However, duringoverhead throwing much higher valgus forcesmay be generated about the medial side of theelbow, so the little increase in the craniocau-dal direction seen in this study might representclinically important instability in the throwingathlete. Rupture of the anterior part of the an-terior medial collateral ligament results in a5.5-mm widening, and complete rupture of theanterior medial collateral ligament results in

Clinical Orthopaedics104 Eygendaal et al and Related Research

TABLE 2. Average Increase of Distance Between Humerus and Radial Head AfterComplete Transsection of the Anterior Medial Collateral Ligament and PosteriorMedial Collateral Ligament in Various Directions

Maximum Minimum Mean Standard Level of Transsection Direction (mm) (mm) (mm) Deviation

Complete (100%) transsection ML �1.2 �0.1 �0.8 0.41anterior medial collateral CC 0.1 0.7 0.4 0.21ligament and posterior AP 8.9 10.9 9.7 0.61medial collateral ligament

ML � mediolateral direction; CC � craniocaudal direction; AP � anteroposterior direction.

Negative values refer to motion in lateral direction. All p values were � 0.001.

n � 9.

7.4-mm widening of the joint space. Ruptureof the anterior medial collateral ligament andposterior medial collateral ligament results inan increase of 9.8 mm. These lesions are morelikely to be detected at physical or radiologicexaminations (Table 1).

Increase of motion of the radial head duringexternal rotation in the posterior direction hasbeen described extensively in cases of postero-lateral rotatory instability attributable to lateralcollateral insufficiency.19,21 This increase in ex-ternal rotation of the ulna and radius producesposterior motion of the radial head, which is de-tectable at physical examination using the pivotshift elbow test as described by O’Driscoll etal.19 Total rupture of the medial collateral liga-ment results in a statistically significant in-crease in internal rotation of the ulna.2,3,26 Inthis study, the motion of the radial head was 9.7mm in the anterior direction.

The anterior band of the anterior medial col-lateral ligament is a primary constraint, theposterior band is a secondary constraint, andthe posterior medial collateral ligament is a ter-tiary constraint to valgus forces. Detection ofpartial ruptures of the anterior medial collateralligament seems to be unlikely on radiologic orphysical examinations, whereas complete rup-tures are more likely to be detected in this way.Anterior translation of the radial head can beused as an additional parameter in physical andradiologic examinations, which might facili-tate the diagnosis of medial collateral ligamentinsufficiency of the elbow.

References1. Callaway GH, Field LD, Deng XH, et al: Biomechan-

ical evaluation of the medial collateral ligament of theelbow. J Bone Joint Surg 79A:1223–1231, 1997.

2. Eygendaal D, Olsen BS, Jensen SL, Seki A, SöjbjergJO: Medial instability of the elbow joint: Kinematicsand clinical relevance. J Shoulder Elbow Surg8:612–616, 1999.

3. Floris S, Olsen BS, Dalstra M, Söjbjerg JO, SneppenO: The medial collateral ligament of the elbow jointanatomy and kinematics. J Shoulder Elbow Surg7:345–352, 1998.

4. Fuss FK: The ulnar collateral ligament of the humanelbow joint: Anatomy, function and biomechanics. JAnat 175:203–212, 1991.

5. Hotchkiss RN, Weiland AJ: Valgus stability of theelbow: J Orthop Res 5:372–377, 1987.

6. Jobe FW, Stark H, Lombardo SJ: Reconstruction of

the ulnar collateral ligament in athletes. J Bone JointSurg 68A:1158–1163, 1986.

7. Josefsson PO, Gentz CF, Johnell O, Wendeberg B:Surgical versus non-surgical treatment of ligamen-tous injuries following dislocation of the elbow joint.J Bone Joint Surg 69A:605–608, 1987.

8. Josefsson PO, Johnell O, Gentz CF: Long-term se-quelae of simple dislocation of the elbow. J BoneJoint Surg 66A:927–930, 1984.

9. Kärrholm J: Roentgen stereophotogrammetry: Re-view of orthopaedic applications. Acta Orthop Scand60: 491–503, 1989.

10. King GJW, Morrey BF, An KN: Stabilizers of the el-bow. J Shoulder Elbow Surg 2:165–174, 1993.

11. Kuroda S, Sakamaki K: Ulnar collateral ligamenttears of the elbow joint. Clin Orthop 208:266–271,1984.

12. Lesin BE, Balfour GW: Acute rupture of the medialcollateral ligament of the elbow requiring recon-struction. J Bone Joint Surg 86A:1278–1280, 1986.

13. Mirowitz SA, London JT: Ulnar collateral ligamentinjury in baseball pitchers: MR imaging evaluation.Radiology 185:573–576, 1992.

14. Morrey BF: Fundamentals and General Considera-tions. In Morrey BF (ed). The Elbow and Its Disor-ders. Ed 2. Philadelphia, WB Saunders 28–33, 1993.

15. Morrey BF, An KN: Functional anatomy of the liga-ments of the elbow. Clin Orthop 201:84–90, 1985.

16. Morrey BF, Chao EY: Passive motion of the elbowjoint. J Bone Joint Surg 58A:501–509, 1976.

17. Morrey BF, Tanaka S, An KN: Valgus stability ofthe elbow. Clin Orthop 265:187–195, 1991.

18. Nakanishi K, Masatomi T, Ochi T, et al: MR arthrog-raphy of the elbow: Evaluation of the UCL of the el-bow. Skeletal Radiol 25:629–634, 1996.

19. O’Driscoll SW, Bell DF, Morrey BF: Posterolateralrotatory instability of the elbow. J Bone Joint Surg73A:440–446, 1991.

20. O’Driscoll SW, Morrey BF, An KN: Origin of the medial ulnar collateral ligament. J Hand Surg17A:164–167, 1992.

21. Olssen BS, Söjbjerg JO, Dalstra M, Sneppen O: Kine-matics of the lateral ligamentous constraints of the el-bow joint. J Shoulder Elbow Surg 5:333–341, 1996.

23. Rijke AM, Goitz HT, McCue FC, Andrews JR, BerrSS: Stress radiography of the medial elbow liga-ments. Radiology 191:213–216, 1994.

22. Ryd L, Albrektsson BEJ, Carlsson L, et al: Roentgenstereophotogrammetric analysis as a predictor of me-chanical loosening of knee prosthesis. J Bone JointSurg 77B:377–383, 1995.

24. Schwab GH, Bennet JB, Woods W, Tullos HS: Bio-mechanics of elbow instability: The role of the medialcollateral ligament. Clin Orthop 146:42–52, 1980.

25. Schwartz ML, Salem AZ, Morwessel RM, AndrewsJR: Ulnar collateral ligament injury in the throwingathlete: Evaluation with saline enhanced MR-arthrography. Radiology 197:297–299, 1995.

26. Söjbjerg JO, Ovesen J, Nielsen S: Experimental el-bow instability after transection of the medial collat-eral ligament. Clin Orthop 287:186–190, 1987.

27. Vrooman HA, Valstar ER, Brand GJ, et al: Fast andaccurate automated measurements in digitized stereo-photogrammetric radiographs. J Biomech 12:491–498,1998.

Number 396March, 2002 Biomechanical Evaluation of the Elbow 105