The RheumatoidMetacarpophalangealJoint

Frank D. Burke, MBBS, FRCS

KEYWORDS

� Rheumatoid arthritis � Metacarpophalangeal joint� Arthroplasty � Arthrodesis

The need for surgical intervention in rheumatoidhand deformities has been, to an extent, modifiedin recent years. This is in response to improvedmedical management involving the use of powerfulnew therapeutic agents. Progression of diseasemay be halted or slowed and the foreshortenedfingers arising from premature epiphyseal closurein Still disease (juvenile idiopathic arthritis) is nowa rarity. Nevertheless, conservative managementwill not control some cases and metacarpophalan-geal (MP) joint reconstruction is still required forthese patients.

THE CAUSES OF RHEUMATOIDMETACARPOPHALANGEAL DEFORMITIES

Twenty years ago, I treated apatient with rheumatoidarthritis with hand deformities. In her late teens, shesustained injuries riding a motorcycle. There wassubstantial bone loss at the right elbow; the woundshealed, leaving a flail right elbow and, somewhatunusually, no associated distal neurologic deficit.Five years later, she developed rheumatoid arthritisand was referred to me 5 years after that diagnosishad been confirmed. She presentedwith rheumatoidhand deformities. The flail right upper limb was wornin a silk scarf sling and its use was limited to feedingobjects into the hand from the left hand. Use for theactivities of daily living was minimal, with the lefthand taking on almost all the functions of both upperlimbs. There were severe rheumatoid deformities tothe left upper limb in contrast to the right, which hadno deformities. The patient offered a very obviousindication of the role of the activities of daily living inthe development of rheumatoid hand deformities.

The author has nothing to disclose.Pulvertaft Hand Centre, Royal Derby Hospital, UttoxeterE-mail address: frank.burke@virgin.net

Hand Clin 27 (2011) 79–86doi:10.1016/j.hcl.2010.09.0050749-0712/11/$ e see front matter � 2011 Elsevier Inc. Al

VOLAR SUBLUXATION OF THE PROXIMALPHALANX BASE

Flatt and Ellison1 described the force vectorsarising when an object is held between the thumband index fingertips (Fig. 1): 1 kg force applied tothe object gives rise to 6 kg in the line of the longdigital flexors with a 3-kg volar vector appliedacross the MP joint. This persistent force on useattenuates the incompetent rheumatoid soft tissuerestraints to the MP joint, creating progressivevolar subluxation to the base of the proximalphalanx (Fig. 2) with respect to the metacarpalhead. Total dislocation may occur and the baseof the proximal phalanx will then drift proximally,creating shortening to the digits and makingreconstructive procedures more difficult (Fig. 3).

Ulnar Drift

Many distal rheumatoid hand deformities can beexplained by more proximal joint malalignment.Synovitis of the wrist and attenuation of the liga-ments may lead to the patient with rheumatoidarthritis maintaining the wrist in radial deviation.Radial tilt at the carpus creates a significant ulnardeviating force to the MP joints. This is evidencedby the effect of surgery to correct ulnar drift of thefingers in cases where a fusion of the wrist hasbeen performed in the past and bone union hasoccurred in radial deviation of the carpus. Thebenefits of ulnar drift correction are likely to beshort lived in these cases if the position of the wristis left unchanged.

Pinch and chuck grip create ulnar deviatingforces to the digits during activities of daily living.

Road 28 Midland Place, Derby DE22 3NE 5, UK

l rights reserved. hand.th

eclinics.com

Fig. 1. The biomechanics of thumb index pinch. (Reprinted from Flatt AE. The care of the rheumatoid hand kine-siology. St Louis (MO): C V Mosby Company; 1974. p. 29; with permission.)

Fig. 2. Volar subluxation of the proximal phalanx. (Reprinted from Flatt AE. The care of the rheumatoid handkinesiology. St Louis (MO): C V Mosby Company; 1974. p. 29; with permission.)

Fig. 3. Volar dislocation of the metacarpophalangeal joint.

Burke80

The Rheumatoid Metacarpophalangeal Joint 81

In addition, there are mild ulnar vector forces to theindex and long fingers as the flexor tendonstraverse the palm from the carpal tunnel. Attenua-tion of the retinacular fibers maintaining theextensor tendons over the metacarpal headfrequently lead to the extensor tendons migratinginto the valleys to the ulnar side of the metacarpalhead. This creates an additional ulnar vector,which will play a role in the development andseverity of ulnar drift.

CARE PATHWAYS WHEN CONSIDERINGSURGICAL OPTIONS

It is beyond the remit of this article to discuss themedical management of MP rheumatoid defor-mities. Two pathways need to be considered:

1. Direct referral to a surgeon from primary care.2. Referral arising out of a combined clinic with

rheumatologists.

The latter is the preferred management option,where rheumatologists, hand surgeons, andhand therapists meet in a combined clinic settingto provide consistent informed advice to thepatient with rheumatoid arthritis and where thepatient’s often unique functional difficulties canbe analyzed. Informed advice is not readily avail-able for patients with rheumatoid arthritis.Alderman and colleagues2 noted major differ-ences in opinion between American handsurgeons and rheumatologists when a variety ofupper limb rheumatoid surgery procedures wereassessed for their efficacy. Surgeons, perhapspredictably, have greater confidence in the valueof the procedures. In a similar study in the UnitedKingdom, as yet unpublished, a similar, but lesspolarized trend was noted. In the UK study, handtherapists were also asked their views on the effi-cacy of the index procedures. Their assessmentusually lay between the views of surgeons andrheumatologists.

HAND THERAPY ASSESSMENT ANDMANAGEMENT

The subject is beyond the immediate remit of thisarticle and this issue of the journal, but it is anessential part of the optimized care pathway fora rheumatoid combined clinic. The patient benefitsfrom functional assessment in workshop andkitchen, with application of aids and appliancesappropriate to their needs to minimize functionaldifficulties and to maximize independence. Thera-pists also have a valuable role in advising onpossible surgical options and can often usefullyact as patient advocates when decisions are being

made about choice of surgery. Their involvementcan optimize the consenting process.

JOINT PROTECTION EXERCISES AND SPLINTS

Resting night splints to correct ulnar deviation ofthe MP joint can be applied. Their effectivenessat controlling deformity is uncertain and patientcompliance may also be a difficulty. Dynamicsplints can be worn during the day with slingsunder the proximal phalanges to draw the baseof the proximal phalanx dorsally and the digitsradially, thereby overcoming volar subluxationand ulnar deviation. Compliance with suchregimes and the enthusiasm for their use arevaried. The technique is widely used in northernEuropean countries but less frequently used inthe United Kingdom and North America.

MP Joint Synovectomy and Joint Realignment

The technique was used quite frequently 30 yearsago but is used less frequently now, although thissurgical option retains support in several countriesin northern Europe. There is a strong tendency inNorth America and the United Kingdom to delaysurgical intervention in these cases until jointreplacement is considered to be necessary. Jointsynovectomy involves exposing the joint through3 possible incisions. A transverse incision can beused over the metacarpal heads; alternatively, 4longitudinal incisions can be used directly overthe metacarpal head or 2 longitudinal incisionseach between a pair of metacarpal heads. Thecapsule is incised on the radial side and synovec-tomy performed, preserving the joint capsule. Theextensor tendons are released from the valleys tothe ulnar side of the metacarpal head and relo-cated onto the metacarpal head. The radial collat-eral ligament can be reefed or the radial portion ofthe volar plate sutured dorsally to overcome volarsubluxation of the base of the proximal phalanx.The hand is rested on a volar slab with an ulnargutter for a few days and then mobilized ina dynamic splint by day, maintaining correctionof the ulnar drift and volar subluxation, for 8 to10 weeks. A static night splint is worn for severalmonths, seeking to maintain the correction.

MP JOINT ARTHROPLASTY FOR PATIENTSWITH RHEUMATOID ARTHRITISThe History of Joint Arthroplasty

Colonel Brannon and Klein3 designed a metal-hinged prosthesis for use at the MP and proximalinterphalangeal (PIP) joint level. In 1959, they re-ported on a 14-patient series (all but 2 involvingthe PIP joint). All were young service personnel

Burke82

and most retained a satisfactory result witha moderate range of motion. I had the opportunityto see one of the implants removed in Louisville in1976 (Fig. 4). The half-threaded rivet screw hadfractured but the patient had maintained goodfunction for 20 years. In 1961, Flatt4 reported ona stainless steel implant for MP or PIP joint use(Fig. 5) and reviewed the results of 242 prosthesesin 1972. MP joint range of motion was limited witha tendency to bone erosion. The implants wereconsidered to offer stability with limited move-ment, and the use in 1976, when I worked inIowa City, was restricted to the rheumatoid MPjoint of the thumb.In the years that followed the development of

Flatt’s implant, research into alternative designssplit into 2 broad streams: the use of the Swansonsilicone joint spacer, and the development of trueimplants constrained or unconstrained, seekingto emulate the now excellent results of total hipreplacement. Many of the designs were underbiomechanical evaluation in Iowa during my tenure(Gillespie and colleagues5). The sophisticateddesign of almost all these implants failed to deliverthe intended benefit and the Steffee, the St GeorgeBuchholz, the Schultz, and the Strickland werewithdrawn from use.The Swanson joint replacement, considered by

many to be a stop-gap implant until a “proper”joint was available, continued to offer benefits forpatients with rheumatoid hand deformities andwas recognized as an option that was safe andbeneficial to patients, particularly if their require-ments were low demand. More durable high-performance silicone was introduced to reduce

Fig. 4. (AeC). Brannon prosthesis.

the risk of implant fracture and, in more recentyears, alternative designs to the Swanson modelhave also gained acceptance (the Sutter, theAvanta, and the Neuflex). I continued to use theSwanson implant for patients with rheumatoidarthritis requiring MP joint arthroplasty.A further generation of MP joint arthroplasties

has been developed in recent years but assess-ment of benefits is clouded in most publicationsby the investigators offering relatively short-termreviews and tending to include PIP joint arthro-plasties with those applied to the MP joint. Thestudies often also include rheumatoid and osteo-arthritic cases. The longer-term benefit specificallyto patients with rheumatoid arthritis at the MP jointremains uncertain. Kujula and colleagues6 notedgratifying results but the study was limited to 7implants in 2 patients reviewed at 10 months.Parker and colleagues7 indicated short-term painrelief and a modest gain in motion, but with therisk of axial subsidence and erosions, particularlyin patients with rheumatoid arthritis. Parker andcolleagues7 observed that nonconstrained pyro-lytic carbon implants rely on stable soft tissuerestraints around the joint. These usually are avail-able in cases involving osteoarthritis but arecommonly deficient in cases presenting with rheu-matoid arthritis. Cook and colleagues8 offereda large pyrocarbon implant series of cases (26)reviewed at an average of 11.5 years. Twelvepercent of a larger series had required revision.Eighty percent of cases suffered from rheumatoidarthritis. The survival rate was 82% at 5 years and81% at 10 years. Of the 71 implants available forreview, 34 were noted to have exhibited mild to

Fig. 5. Flatt prothesis.

The Rheumatoid Metacarpophalangeal Joint 83

moderate subsidence. Recurrent volar subluxationdid not seem to be a significant problemwith thesepatients but ulnar deviation was noted to recur toprevious levels (average 20� preoperatively and19� at review). Surgery reduced the extensor lagand this benefit was noted to be retained tolong-term review. The postoperative range ofmotion increased modestly with further additionalgains at long-term review.

The long-term outcome for Swanson jointreplacements has been documented reasonablywell by Chung and colleagues9 in a systematicreview and by Goldfarb and Stern10 and Trail andcolleagues.11 The procedure offers good pain re-lief and a modest increase in range of motion,which tends to decrease over time. The arc ofmotion shifts to a position of greater extension.Implant fractures did not necessarily createa need for revision and rates of revision surgerywere low.

A limited number of articles have investigated thebenefits of alternative silicone implants. Parkkilaand colleagues12 did not find any appreciabledifference between Sutter implants and the Swan-son joint replacement. Delaney and colleagues13

compared theNeuflex implant against the Swansonand found a greater range of motion with the Neu-flex, but the study was small and limited to 2-yearreview. Moller and colleagues14 compared theAvanta silicone implant with the Swanson. Amodest increase in rangeat 2-year reviewoccurredwhen the Avanta had been used, but the implantwas associated with a higher fracture rate.

Currently the guest editor of this edition of HandClinics has been leading a multicenter studyreviewing the middle-term outcome (7 years) ofSwanson MP joint arthroplasty contrasting theiroutcome with a similar group of patients who elec-ted at the outset to continue nonoperative care byrheumatologists. Enrollment is complete but thestudy has 2 more years to run. Preliminary resultsreveal those seeking surgery were generally moredisabled by the rheumatoid process and the oper-ative intervention lifted their function (1 year aftersurgery) to the functional level of those who hadelected to enter the nonoperative limb of the study.Grip and pinch were not improved by surgery. Themedical wing of the study maintained functionalstatus at the first-year review with no observabledeterioration in function. Swanson arthroplastyoutcome was not undermined by more severeulnar drift preoperatively; they gained a greatercorrection. One study center has investigatedthe relationship between early postoperativecomplications and concomitant chemotherapy.Only one type of drug was stopped before thesurgery (Etanercept, discontinued 2e3 weekspreoperatively). All other medications (nonste-roidal anti-inflammatories biologics, cytotoxics,noncytotoxics, and steroids) were continued overthe operative period. There were no serious prob-lems to the 140 wounds studied. Four minorproblems occurred (slightly delayed wound heal-ing, suture granuloma, a rheumatoid flare, anda possible superficial wound infection) and allsettled satisfactorily.

Burke84

Assessment of Patients for Surgery

The combined clinic model with rheumatologistsand hand therapists is preferred. The principlesSouter15 offered in 1979 are still relevant. Themere existence of deformity is not necessarily anindication for surgery. Surgeon and therapistneed to look beyond the deformity and assessthe disability and whether surgical options willlikely improve matters. Decisions should be pacedwith a patient empowered by knowledge ofsurgery, rehabilitation, and likely outcome. A sepa-rate assessment by therapists in a workshop/kitchen environment is valuable.Particular care has to be taken to assess the

viability of the skin to the dorsum of the hand.Splinter hemorrhages to the nail beds may ques-tion the skin’s ability to heal with compromisedrehabilitation postoperatively or implant exposureand infection. A full assessment of all upper andlower limbs is required to ensure the anticipatedbenefit at MP joint level will not be underminedby other problems. Patients with rheumatoidarthritis are often candidates for several surgicaloptions at any given time. These options need tobe prioritized to the patients’ needs. Patientsneed to “gear-up” to elective surgery and mostpatients with rheumatoid arthritis are reluctant tosign up for more than one operative procedure ina year.

Operative Technique

Patients are preclerked and attend on the day ofsurgery. Anesthesia is usually by brachial blockor general anesthesia if preferred. The limb isprepped and draped with an upper arm tourni-quet in place. Care is taken to ensure the prepa-ration agent does not seep under the tourniquet(I prefer exsanguination using a sterile crepebandage that will be used at the end of theprocedure as part of the dressing). The smallamount of blood that is left in the vessels facili-tates identification. I prefer 4 longitudinal inci-sions over the MP joints with a similar incisionto the retinacular fibers, which are attenuatedon top of the joint. The retinacular fibers shouldthen be dissected off the joint capsule. Thecapsule is then incised and mobilized sufficientlyfor a fulsome view of the metacarpal head andneck. Synovectomy is performed; the extent ofhead and neck resection is variable. Preoperativeradiographs will help give some information as tothe degree of proximal phalanx base subluxationand proximal migration under the metacarpalhead. Minor migration may permit resection ofthe metacarpal head, preserving the collateralligament origins, but this will not be possible

with more severe deformity. Resect the headand complete the synovectomy if required. Drilla hole in the proximal phalanx base to allow theentry of a Swanson reamer to square off thecortical defect to the articular surface to accom-modate the base of the stem. The medulla of theproximal phalanx can then be reamed.I prefer to compact the metacarpal medullary

bone and only ream the cavity when the bone isvery narrow (most commonly found in the ringmetacarpal). The largest implant is preferredbased on what will fit in the metacarpal; it wouldbe unusual to find sizing limited by the dimensionsof the proximal phalanx. Trial reductions are help-ful to check that enough metacarpal head hasbeen excised. Undue tightness will limit flexionpostoperatively; be aware that silicone trialimplants shrink slightly as a result of repetitiveautoclaving and a very tight trial fit probably meansthe implant chosen will be slightly too big. Theimplant is inserted into the metacarpal and thenfed into the proximal phalanx base with non-toothed forceps with the joint held in flexion. I donot use grommets (Trail and colleagues11 did notfind they improved outcome). If the implant isseated satisfactorily, I may reef the radial collateralligament with a nonabsorbable suture. I am not inthe habit of mobilizing part of the volar platedorsally to overcome volar subluxation. Theextensor tendon is then released, on its ulnarborder, and relocated over the implant and the ret-inacular fibers reefed to maintain the position. Theskin is closed without drains. I prefer to operate onthe fingers in twos, first the index and middlefingers with skin closure to both before startingon the ring and little fingers. Paraffin gauze anddressings are applied and the hand rested ona volar slab with an ulnar gutter to avoid ulnar devi-ation. The hand is then elevated with firm pressureover the wound and the tourniquet released andremoved immediately, to avoid the risk of venouscongestion during the phase of reactive hyper-emia. Elevation with a near straight elbow andfirm pressure to the operative area is continuedfor 3 to 4 minutes, when the reactive hyperemiais at its maximum. There were no hematomas inthe 140 wounds treated by this method in ourrecent case series nor do I recall hematoma prob-lems in the past. The patient is discharged thatevening with dressing changes at 3 to 5 daysand the creation of a static night splint to maintainMP joint extension and the avoidance of ulnardeviation. An extensor outrigger is applied fordaytime use, holding the MP joints in extensionwith a radial force applied to mold the scar tissuein the early weeks from surgery. The dynamicsplint is used for 6 to 8 weeks, depending on

Fig. 6. (A, B) Extensor pollicis longus to extensor pollicis brevis transfer.

The Rheumatoid Metacarpophalangeal Joint 85

progress and the static splint 10 to 12 weeks orlonger if patients prefer.

THE MP JOINT OF THE THUMB

Boutonniere and swan neck deformities have theeffect of shortening the effective length of thethumb, thereby reducing precision skills. Stabilitytakes precedence over mobility at the MP jointand arthrodesis presents a popular option, offeringpredictable lasting benefit. A longitudinal dorsalincision over the MP joint affords a good view ofthe joint; the incision is deepened to the joint andthe sides of the proximal phalanx base and meta-carpal head are exposed. A limited excision of thehead is performed with the plane of the excision in10� of flexion with respect to the long axis of thefirst metacarpal. A minimum of bone is thenexcised from the proximal phalanx base, cutperpendicular to the long axis of the bone. Ifbone stock is judged to be reasonable, fixationwith an oblique Kirschner (K)-wire and a cerclagewire is adequate and involves minimal hardware,which does not lie too close to the skin. Thedisparity in width between metacarpal head andproximal phalanx base allow for the twistedportion of the cerclage wire to lie in an unobtrusivearea down the side of the fusion site. The K-wire ispassed distally until it bites into the proximalphalanx cortex and the proximal end is then bentover, cut short, and rotated to lie as flat aspossible. The internal fixation can then be ignoredwith no need for removal after bone union. Patientswith rheumatoid arthritis suffer so many difficultiesin life that any surgical technique that offers themsimple postoperative care is to be recommended.If bone stock is considered to be poor, tensionband wiring offers improved stability with minimaladditional hardware.

Extensor Pollicis Longus to Extensor PollicisBrevis Transfer

A minority of patients present for assessment witha boutonniere deformity to the MP joint of the

thumb, where radiographs reveal a fairly satisfac-tory joint surface and the joint is stable. Thumblength can be regained and joint mobilitypreserved with an underused surgical option(Fig. 6). The extensor pollicis longus is mobilized,leaving the soft tissues on its immediate marginsin longitudinal continuity. It is released from thedistal phalanx. A mallet deformity will not occur ifthe soft tissues on the immediate margins of thetendon are preserved. A drill hole is then madenear the base of the proximal phalanx and thetendon drawn through and sutured under tensionto itself. Active extension of the MP joint is re-gained with improved thumb length and dexterity.The new tendon insertion is usually robust, whichallows early movement postoperatively.

SUMMARY

SwansonMP joint arthroplasty remains a beneficialoperative intervention in carefully selected cases.The combined clinic with rheumatologists is thepreferred environment in which these casesshould be considered. Hand therapists offer valu-able additional benefits in terms of optimized care(advice on splints and appliances) and patientadvocacy during the consenting process. Therole of the latest generation of pyrocarbonimplants applied to the rheumatoid MP jointremains uncertain. The vulnerability of the patientwith rheumatoid arthritis to attenuation of liga-ments may undermine any lasting benefit with py-rocarbon implants through joint subluxation,recurrent ulnar drift, and bone erosion around theimplant stem.

REFERENCES

1. Flatt AE, Ellison MR. Restoration of rheumatoid

finger function III: a follow-up note after 14 years of

experience with a metallic hinge prosthesis.

J Bone Joint Surg Am 1972;54:1317e22.

2. Alderman AK, Chung KC, Kim HM, et al. Effective-

ness of rheumatoid hand surgery: contrasting

Burke86

perceptions of hand surgeons and rheumatologists.

J Hand Surg Am 2003;28:3e11.

3. Brannon EW, Klein G. Experiences with a finger

joint prosthesis. J Bone Joint Surg Am 1959;41:

87e102.

4. Flatt AE. Restoration of rheumatoid finger joint func-

tion: interim report on trial of prosthetic replacement.

J Bone Joint Surg Am 1961;42:753e74.

5. Gillespie TE, Flatt AE, Youm Y, et al. Biomechanical

evaluation of metacarpophalangeal joint prosthesis

designs. J Hand Surg Am 1979;4:508e21.

6. Kujala S, Bongiorno V, Leppilahti J, et al. Pyrolytic

carbon MCP arthroplasty: a report of seven pros-

theses implanted into two patients. J Orthop Trau-

matol 2006;7(1):12e5.

7. Parker WL, Rizzo M, Moran SL, et al. Preliminary

results of non constrained pyrolytic carbon arthro-

plasty for metacarpophalangeal joint arthritis.

J Hand Surg Am 2007;32:1496e505.

8. Cook SD, Beckenbaugh RD, Redondo J, et al.

Long-term follow-up of pyrolyctic carbon

metacarpophalangeal implants. J Bone Joint Surg

Am 1999;81:635e48.

9. Chung KC, Kowalski CP, Kim HM, et al.

Patient outcomes following Swanson silastic

metacarpophalangeal joint arthroplasty in the rheu-

matoid hand: a systematic overview. J Rheumatol

2000;27(6):1395e402.

10. Goldfarb CA, Stern PJ. Metacarpophalangeal joint

arthroplasty in rheumatoid arthritis. J Bone Joint

Surg Am 2003;85:1869e78.

11. Trail IA, Martin JA, Nuttall D, et al. Seventeen years

survivorship analysis of silastic metacarpophalan-

geal joint replacement. J Bone Joint Surg Br 2004;

86:1002e6.

12. Parkkila T, Belt EA, Hakaka M, et al. Comparison of

Swanson and Sutter metacarpophalangeal joint ar-

throplasties in patients with rheumatoid arthritis:

a prospective and randomised trial. J Hand Surg

Am 2005;30:1276e81.

13. Delaney R, Trail IA, Nuttall D. A comparative study of

outcome between the neuflex and Swanson meta-

carpophalangeal joint replacements. J Hand Surg

Br 2005;30:3e7.

14. Moller K, Sollerman C, Geijer M, et al. Avanta versus

Swanson silicon implants in the MCP joint. Rando-

mised comparison of 30 patients followed for 2

years. J Hand Surg Br 2005;30:8e13.

15. Souter WE. Planning treatment of the rheumatoid

hand. Hand 1979;11:3e16.