arthrodesis of the interphalangeal joints with headless compression screws

INSTRUCTIONAL COURSE LECTURE

AActfjtsctitdt

sfitFraa

Arthrodesis of the Interphalangeal JointsWith Headless Compression Screws

Stephen J. Leibovic, MD

From the Virginia Hand Center, Medical College of Virginia, Virginia Commonwealth University, Richmond, VA.

Interphalangeal arthrodesis is a reliable method of pain relief for arthritic proximal and distalinterphalangeal joints in the fingers. Indications include osteoarthritis, acute trauma, chronicreconstruction for trauma, rheumatoid and other inflammatory arthritides, and at the distalinterphalangeal joint, chronic mallet finger deformity and unreconstructible flexor tendon de-fects. Solid arthrodesis imparts stability to the digital skeleton. Headless compression screws canbe reproducibly inserted and are a good method to provide fixation adequate to accomplishinterphalangeal arthrodesis. Surgical technique involves a dorsal incision and preparing theskeleton for good bony apposition. Exact technique for screw insertion depends on the specificscrew used. Union rates range from 85% to 100% in published studies, with time to union of 7to 10 weeks. (J Hand Surg 2007;32A:1113–1119. Copyright © 2007 by the American Society forSurgery of the Hand.)Key words: Arthrodesis, screw, proximal interphalangeal joint, distal interphalangeal joint,osteoarthritis.

grcfaVwmmahg

jwjaspptpape

rthrodesis of the interphalangeal joints of thefingers is a successful and time-tested treatmentfor a variety of painful or disabling conditions.

t the proximal interphalangeal joint, indications in-lude arthritides of all kinds (osteo-, psoriatic, rheuma-oid, post-traumatic). It is also vital in reconstructionrom trauma, both acute and chronic, and from unstableoints where soft tissue stabilizing procedures are con-raindicated or have failed. It is sometimes used inevere contractures from Dupuytren’s disease or otherauses where release is unlikely to restore useful mo-ion. At the distal interphalangeal (DIP) joint, similarndications exist, and in addition, arthrodesis is used forreatment of chronic mallet finger deformity and flexorigitorum profundus avulsion or laceration when softissue repairs are impossible or unlikely to succeed.

In chronic painful cases, nonsurgical measureshould normally be exhausted before surgery is per-ormed. These include anti-inflammatory medication,ntra-articular steroid injections, and splinting. In acuteraumatic situations, treatment need is more immediate.amiliarity with a variety of methods of surgical arth-odesis equips the surgeon with the means to treatcute as well as chronic cases with the techniquesnd hardware best suited for the particular case.

Motion-preserving procedures for the interphalan- o

eal joints exist, but they have limitations. Biologicesurfacing techniques include volar plate arthroplasty,hondral resurfacing, and vascularized free joint trans-er. Each has met with limited success; the volar platerthroplasty has certainly been the most widely used.olar plate arthroplasty is suited for some situationsith destruction limited to the volar portion of theiddle phalangeal base, particularly after dorsal proxi-al interphalangeal (PIP) joint dislocations.1 Chondral

rthroplasty2 and free joint transfer,3 however, bothave significant donor morbidity for limited functionalains.

Synthetic silicone interposition arthroplasty has en-oyed popularity at the metacarpophalangeal joints asell as the PIP joints and, to a limited extent, the DIP

oints.4 More recently, total joint arthroplasty with an-tomic replacement of both articulating surfaces haseen increasing popularity with the advent of metal onlastic or pyrocarbon implants. Enthusiasm for motion-reserving procedures at the interphalangeal joints isempered, however, due to problems with stability, es-ecially in the index finger during pinching activities,nd implant failure. In such cases, or to avoid theseroblems, arthrodesis remains a gold standard. Moberg5

loquently expressed in 1960 that “the prime requisite
f a good digital arthrodesis is a painless and stable
The Journal of Hand Surgery 1113

uost

htlbstpwtfistuoc

THatbi(FNtb

STdecsdstcbwus

bTpI

saccflabbprw

sasde

BFeabf

Fas

1114 The Journal of Hand Surgery / Vol. 32A No. 7 September 2007

nion in proper position occurring in a reasonable spacef time.” Proximal and distal interphalangeal arthrode-is remains today a fundamental and frequently usedechnique for relief of pain and disability in the fingers.

Methods for achieving interphalangeal arthrodesisave evolved. It is generally accepted that techniqueshat provide structural rigidity and compression areikely to have the highest success rate. K-wires haveeen used for years, although they suffer from lack oftructural rigidity. The tenon method6 and the bone pegechnique7 are time-consuming methods of bone car-entry that impart considerable rigidity to an arthrodesishen used alone or with just K-wires. More recently,

ension band wiring and headless compression screwxation of interphalangeal joint fusions have demon-trated consistent success with relatively low complica-ion rates. At the present time, they represent the prod-ct of 3 decades of investigations to increase the rigidityf interphalangeal arthrodesis while simplifying surgi-al techniques.

echniqueeadless compression screws used for interphalangeal

rthrodesis are either screws with leading and trailinghreaded portions with different thread pitch separatedy a smooth shaft, such as the TwinFix (Stryker-Leib-nger, Kalamazoo, MI), Herbert or Herbert/WhippleZimmer, Warsaw, IN), HBS (Orthosurgical, Miami,L), or Millennium (Millennium Medical, Santa Fe,M), or with a fully threaded, continuously variable

apered thread, such as the Acutrak (Acumed, Hills-oro, OR). Surgical techniques are similar for all.

urgical Approachhe surgical approach to the PIP joint is through aorsal, curvilinear incision around the joint (Fig. 1). Thextensor tendon is encountered directly under the skin,overed with thin filmy epitenon. There is no macro-copically defined separation between the extensor ten-on and the joint capsule at the PIP level. The combinedtructure of tendon and capsule is divided in the midlineo enter the joint. Care should be taken to ensure that theentral slip is not detached from the dorsal base of P2,ut that it is rather subperiosteally dissected so that itill re-adhere to this region without proximal retractionpon closure. The lateral bands are left undisturbed andhould continue to function to extend the DIP joint.

To fully expose the joint surfaces for appropriateone carpentry, both collateral ligaments are sectioned.his can be simply done from inside the joint, eithereeling them off the condyles of P1 or by direct section.
n either case, any redundant tissue after sectioning i
hould be excised so that it does not interfere with bonypposition. Care must be taken to protect the neurovas-ular bundles on both sides of the finger. Once theollateral ligaments are excised, the joint can be hyper-exed for further exposure. The volar plate will remainnd will prevent complete separation of P1 from P2. Asoth P1 and P2 are slightly shortened by subsequentone carpentry, the volar plate may act as a spacerreventing apposition of P1 to P2. It should either beemoved, or a central portion of it excised, at this stageith joint hyperflexion.Approach to the DIP joint is through a Y- or H-

haped incision over the joint. Care must be taken tovoid damage to the germinal matrix, close by theurgical field. The extensor tendon is sharply dividedirectly over the joint, and the collateral ligaments arexcised as in the PIP joint. The volar plate is excised.

one Preparationixation techniques currently available impart consid-rable stiffness and strength to the arthrodesis immedi-tely, unlike in years past when fixation consisted of allone carpentry with no metal, or K-wires only. There-ore, elaborate bone carpentry with tenons or bone pegs

igure 1. Curved incision on the dorsal side of the jointvoids the midline. This minimizes the potential for adhe-ions that might impede tendon function (Leibovic16).

s unnecessary. Typically, straight cuts are made with

aup

agmtmatffhfdcpdfl

mtse2

phfb

atmcaecbc

mjabhotgss

SSseaplpltadbu

Fojtaahta(

Fird

Stephen J. Leibovic / Arthrodesis of the Interphalangeal Joints 1115

n oscillating saw on both phalanges, or rongeurs aresed to fashion cup and cone or convex-concave op-osing surfaces for arthrodesis.

Preoperative planning must include a decision on theppropriate angle for arthrodesis. Proximal interphalan-eal joints of the index and middle fingers are typicallyore functional when fused at 15° to 30° of flexion. As

hey are used more in fine pinch and manipulation, aore extended position improves dexterity. The ring

nd small fingers are more functional when fused at 30°o 45°, as they are more involved in grip. As one movesrom the radial to the ulnar side of the hand, optimalusion angle increases. Individual patient function,owever, must be considered; a guitar player, in theretting hand, may want the joints more flexed than aata entry operator. Encouraging the patient to wearustom fabricated orthoplast splints at different anglesrior to operative planning will help them be part of theecision making process. At the DIP joint, 0° to 50° ofexion is generally most useful.When the desired fusion angle is determined, imple-entation must be planned depending upon the carpen-

ry method chosen. If saw cuts are to be used, both cutshould be angled. The sum of the angles on each bonequals the total amount of joint flexion obtained (Fig.

igure 2. Straight saw cuts at the condyles of P1 and the basef P2 can be fashioned to produce the appropriate amount ofoint flexion after fusion. If the angles of each cut with respecto the perpendicular axis of each bone is represented by �nd �, the final angle of fusion will equal the sum of thengles � and �, as shown here. The technique is unforgiving,owever, as the final angle of flexion as well as angulation in

he coronal plane are completely determined by the saw cutsnd they cannot be modified without recutting the boneLeibovic16).

). This produces the largest surface area of bone ap- m

osition for the arthrodesis. Coronal plane alignment,owever, is completely determined by the saw cuts; nourther adjustment can be made without recutting theone, which further shortens the digital skeleton.

An alternative is the use of rongeurs to fashion cupnd cone or convex-concave opposing surfaces on thewo bone ends (Fig. 3). Final positioning can be deter-ined after machining of the bone surfaces. Small

hanges in the angle of flexion as well as rotation can bedjusted without having to recut or recontour the bonends. Bone mills are available that can produce accurateup and cone configurations on the opposing surfaces,ut their use is cumbersome in the fingers; I preferareful use of a rongeur.

The digit is necessarily shortened somewhat by re-oval of cartilage and subchondral bone from opposing

oint surfaces. Especially in osteoarthritis with preoper-tive angulation in the coronal plane, the shortening cane considerable. If a sufficient amount of the volar plateas not been removed, it may prevent apposition of thepposing surfaces. A frequent complaint in osteoarthri-is is enlargement of the proximal or distal interphalan-eal joints by osteophytes, which may be deemed un-ightly or prevent the patient from wearing rings. Thesehould be removed at this time.

crew Insertionuccessful arthrodesis with a headless compressioncrew requires attention to detail in planning andxecution. At the PIP joint, the screw is insertedntegrade from the dorsal surface of the proximalhalanx. It is critical that the dorsal entry site is ateast 6 or 7 mm proximal to the joint surface torevent fragmentation of the dorsal cortex upon dril-ing. A K-wire is used to start a hole obliquelyhrough the dorsal proximal phalangeal cortex. Thengle of the hole in the sagittal plane is equal to theesired angle of arthrodesis. The K-wire is followedy the drill sequence required for the particular screwsed. In the case of the Herbert and mini-Herbert

igure 3. Rounded cup and cone configurations can be eas-ly fashioned at the proximal interphalangeal joint with aongeur. Then, the final angle of fusion can be adjusteduring insertion of the fixation device, and any rotational
alalignment can be easily corrected (Leibovic16).

stcbpcoahKpt

dagdtsSonada

twbc

Dnlsoteattrfltodpsiai

TTtpsTht

ieamstt8dcd

tgI

FifUtatd


crews, the “long, skinny” drill is used first to drillhrough the proximal phalanx into the medullaryanal of the middle phalanx (Fig. 4). This is followedy the “short, fat” drill for the trailing threads in theroximal phalanx only. The only two-piece headlessompression screw, the TwinFix, which has the the-retical advantage of providing more compressioncross the arthrodesis site, and the Acutrak screwave a single cannulated drill that will follow the-wire through the proximal phalanx into the middlehalanx. The drill must pass the cortical isthmus ofhe middle phalanx.

A small 1.1 or 1.5 mm rongeur is used to enlarge theorsal hole in the proximal phalanx to prevent fracturend fragmentation upon screw insertion. With a ron-eur, small bites can be taken around the hole in theorsal cortex. Cortical purchase of the trailing thread ofhe screw is not necessary for adequate fixation. If thecrew being used requires a tap, it is used at this point.crew length is then estimated by eye after experiencer using fluoroscopy and a K-wire. Measurement needot be precise, as the leading end of the screw may restnywhere at or beyond the cortical isthmus of the mid-le phalanx. Usually, screws between 16 and 22 mmre appropriate for PIP joint arthrodesis. The assistant

igure 4. The “long, skinny” drill from the Herbert screw sets inserted first. The hole must be made sufficiently far backrom the joint to avoid fragmentation of the dorsal cortex.sually 6–7 mm is enough. The “short, fat” drill is then used

o enlarge the trailing hole in the dorsal cortex of P1. Finally,rongeur is used to enlarge the hole sufficiently that the

railing head of the Herbert screw can easily pass deep to the

morsal cortex (Leibovic16).

hen aligns the joint and maintains firm compressionhile the screw is advanced until it is flush with oreneath the dorsal cortex. Final rotational alignmentan be adjusted after screw insertion (Fig. 5).

Insertion of a headless compression screw into theIP joint is done retrograde. A preliminary K-wire isecessary only if the hardware being used has a cannu-ated drill. It is drilled first antegrade from the jointurface distal. The drill will find its way down the canalf the distal phalanx and exit through the tuft. As it tentshe skin at the pulp, a small incision is made largenough for exit of the drill. The drill is then withdrawnnd reinserted retrograde through the pulp incision, intohe tuft hole and into the joint. Under direct vision, it ishen directed into the middle of the denuded and deco-ticated head of the middle phalanx. Up to 10° ofexion at the arthrodesis site is usually possible if at this

ime the drill is directed from slightly volar to the centerf the head of the middle phalanx dorsally toward theorsal cortex until it reaches just beyond the middlehalangeal isthmus, seen on fluoroscopy (Fig. 6). If thecrew requires a larger diameter trailing thread hole, its made in the distal phalanx only. Then, while thessistant reduces and compresses the joint, the screw isnserted (Fig. 7).

echnical Pitfallshe major complication of PIP joint arthrodesis with

hese screws is fragmentation of the dorsal cortex of theroximal phalanx. If this occurs, all purchase is lost andcrew fixation must be abandoned or supplemented.his is best avoided by enlarging the dorsal corticalole with a rongeur until it is almost as large as thehread diameter of the trailing threads.

At the DIP joint, a significant complication can occurf the thread diameter exceeds the anteroposterior diam-ter of the distal phalanx, which it may. The averagenteroposterior diameter of the distal phalanx is 3.55m,8 while the trailing thread diameter of the Herbert

crew is 3.9 mm and the TwinFix screw, 4.0 mm. Inheir cadaver study, Wyrsch et al8 found penetration ofhe dorsal distal phalangeal cortex by screw threads in3% of their specimens and in all female specimensuring joint instrumentation with a Herbert screw. Thisan lead to stretching or disruption of the nail bed witheformity of the nail.

Occasionally, the diameter of the metaphysis ofhe middle phalanx is so large that the screw will notain purchase. This occurs most often in the thumb.f preoperative radiographs suggest this, alternate
eans of fixation should be used.

PAdtirtsbsmtaipjsjisu

rDug

Ft

Fgtrh


ostoperative Carewell-padded dressing on the digit or a small bulky

ressing on the hand incorporating a splint will protecthe digit from stress and allow for postoperative swell-ng. After 7 to 10 days, the dressing and sutures can beemoved, and a thermoplastic splint can be fitted to justhe involved joint. A scar pad may be included in theplint, which is applied with Velcro straps. Early mo-ilization of the adjacent joints prevents stiffness. Thecrew imparts substantial stability to the joint, allowinginimal external protection. The splint can be discon-

inued after 3 weeks during light activity, although forny stressful activities, it may be left in place. Six weekss usually sufficient for clinical union, judged by aainless joint on stress, although radiographic union,udged by presence of trabeculae crossing the arthrode-is site, may take 3 months or more. If the arthrodesedoint is pain free, lack of radiographic union should notmpede functional use of the hand. Resistive exerciseshould be begun, however, only after radiographicnion is evident.

Nonunion is rare, but should be declared if thereemains pain and instability 6 months after surgery.elayed unions may be splinted for longer thansual, though splinting for more than 3 months is

igure 5. (A) Posteroanterior and (B) lateral radiographs of ahe dorsal cortex of P1 to prevent cortical fragmentation.

PIP joint fusion using a Herbert screw. Note the enlarged hole in

enerally ill advised. 1

igure 6. For DIP joint arthrodesis, the drill is passed ante-rade from the joint surface distally through the end of theuft and the pulp of the finger. It is then withdrawn andeinserted retrograde, from where it enters the remains of theead of P2. This can sometimes be done in approximately
0° of flexion at the DIP joint.

RIpP2cH4dphadtsoptd

nww

chsedsi

Fsi s can


esultsnterphalangeal joint arthrodesis using headless com-ression screws has enjoyed high success rates. At theIP joint, reported nonunion rates vary between 0% and% (Table 1). Leibovic and Strickland,9 in a studyomparing PIP fixation with K-wires, tension band,erbert screw, and plate found nonunion rates of 21%,.5%, 0%, and 50%, respectively. There was a clearecrease in nonunion rate with the Herbert screw com-ared to other means of fixation. Stern and Fulton,10

owever, in a study of complications of DIP jointrthrodesis found nonunion rates unchanged betweenifferent surgical techniques of K-wire, interfragmen-ary loop wire plus K-wire, and Herbert screw. In theirtudy, the nonunion rate was 11% to 12%, independentf fixation used. Both studies agreed that patients withsoriatic arthritis had the highest nonunion rates at thewo joints, although the rates of nonunion with otheriagnoses differed (Table 2).

Complications are not common. Brutus et al11 had 4onunions, 2 from infection in the DIP joint. Thereere 4 postoperative infections, easily treated. There

igure 7. Radiographs of a DIP joint fusion using an Acutraktability. The large diameter (3.5 mm) of the trailing threads mn loss of purchase by the screw in the bone. Sometimes thi

as a 7% incidence of nail deformity, but no significant

omplaints of fingertip tenderness. Stern and Fulton10

ad no osteomyelitis in DIP arthrodeses with Herbertcrews, although there was a 2% incidence of osteomy-litis across all fixation methods. There was, however,eep wound infection in 4% of fusions with Herbertcrews. There was an incidence of dorsal skin necrosisn 15% of joints fused with a Herbert screw. Leibovic

ed screw. An additional K-wire has been inserted for addedak through the cortex of the distal phalanx, which will result

be salvaged with the addition of a K-wire.

Table 1. Nonunion Rate and Time to Union inInterphalangeal Arthrodesis With HeadlessCompression Screws

Author JointNonunionRate (%)

Time toUnion

(weeks)

Ayres et al13 PIP 2 —Katzman et al14 PIP and DIP 0 8.1Leibovic and

Strickland9 PIP 0 7Stern and

Fulton10 DIP 11 10Gomez et al15 DIP 5 8

11

taperay bre

Brutus et al DIP 15 8–10

as

wagpara

Dacsrt

itctdbtwtfhiicsMaw

R

aa

tv

R

1

1

1

1

1

1

1


nd Strickland9 had 1 case of superficial infection aftercrew fixation, or 3%, and no osteomyelitis.

Established nonunion, while rare, should be treatedith reoperation using cancellous bone graft. Good

pposition of vascular cancellous surfaces should be theoal of treatment, and fixation should be as secure asossible. Malunion can occur, but it is best avoided byttention to detail. Angular alignment in all planes andotational alignment must be assessed intraoperativelynd corrected when needed.

Dorsal skin necrosis is more likely to occur over theIP joint than the PIP joint. Blood supply is less dense

nd the skin is tighter. Tight skin closure and overlyompressive dressings must be avoided. As the digit ishortened somewhat by the resection required for arth-odesis, skin flaps should not normally be needed at theime of operation.

Interphalangeal joint arthrodesis has proven effectiven relieving pain and instability in cases of osteoarthri-is, inflammatory arthritides, and trauma. The PIP jointontributes 85% of intrinsic digital flexion and 30% ofhe combined overall flexion of the finger,12 and painfulysfunction at the joint can be debilitating. The mor-idity associated with a fused joint is a worthwhile priceo pay for the increased function and comfort associatedith a painless joint. At the DIP joint, arthrodesis is also

olerated very well. As it is accepted that rigid fixationacilitates healing of an arthrodesis, fixation techniquesave evolved over the last 4 decades with a goal ofncreasing the rigidity of the fixation, while simplifyingts application and insertion. Fixation with headlessompression screws accomplishes rigid fixation with atraightforward surgical technique. It is consistent withoberg’s requirements for a good digital arthrodesis

nd is my preferred technique for digital arthrodesis

Table 2. Radiographic Nonunion Rate byDiagnosis in Interphalangeal Arthrodesis(Various Fixation Methods)

Diagnosis

Stern and Fulton10

(DIP Joint)Nonunion Rate (%)

Leibovic andStrickland9

(PIP Joint)Nonunion (%)

Osteoarthritis 13 0Post trauma

reconstructionchronic 7 9

Acute trauma 22 12Rheumatoid

arthritis 13 15Psoriatic arthritis 22 29

hen there is adequate bone stock.

eceived for publication February 14, 2007; accepted June 13, 2007.No benefits in any form have been received or will be received fromcommercial party related directly or indirectly to the subject of this

rticle.Corresponding author: Stephen J. Leibovic, MD, Virginia Hand Cen-

er, 2819 N. Parham Rd., Richmond, VA 23294; e-mail: [email protected] © 2007 by the American Society for Surgery of the Hand0363-5023/07/32A07-0024$32.00/0doi:10.1016/j.jhsa.2007.06.010

eferences1. Eaton RG, Malerich MM. Volar plate arthroplasty for the

proximal interphalangeal joint: A ten year review. J HandSurg 1980;5A:260.

2. Hasegawa T, Yamano Y. Arthroplasty of the proximal in-terphalangeal joint using costal cartilage grafts. J Hand Surg1992;17B:583–585.

3. Foucher G, Lenoble E, Smith D. Free and island vascularizedjoint transfer for proximal interphalangeal reconstruction: aseries of 27 cases. J Hand Surg 1994;19A:8–16.

4. Zimmerman NB, Suhey PV, Clark GL, Wilgis EF. Siliconeinterpositional arthroplasty of the distal interphalangeal joint.J Hand Surg 1989;14A:882–887.

5. Moberg E. Arthrodesis of finger joints. Surg Clin N. Am1960;40:465–470.

6. Lewis RC, Nordyke MD, Tenny JR. The tenon method ofsmall joint arthrodesis in the hand. J Hand Surg 1986;11A:567–569.

7. Potenza AD. A technique for arthrodesis of finger joints.J Bone Joint Surg 1973;55A:1534–1536.

8. Wyrsch B, Dawson J, Aufranc S, Weikert D, Milek M.Distal interphalangeal joint arthrodesis comparing tension-band wire and Herbert screw: A biomechanical and dimen-sional analysis. J Hand Surg 1996;21A:438–443.

9. Leibovic SJ, Strickland JW. Arthrodesis of the proximalinterphalangeal joint of the finger: Comparison of the use ofthe Herbert screw with other fixation methods. J Hand Surg1994;19A:181–188.

0. Stern PJ, Fulton DB. Distal interphalangeal joint arthrodesis:An analysis of complications. J Hand Surg 1992;17A:1139–1145.

1. Brutus JP, Palmer AK, Mosher JF, Harley BJ, Loftus JB.Use of a headless compressive screw for distal interphalan-geal joint arthrodesis in digits: Clinical outcome and reviewof complications. J Hand Surg 2006;31A:85–89.

2. Little JW, Herndon JH, Thompson JS. Examination of thehand. In: Reconstructive Plastic Surgery. Vol 6.Philadelphia: WB Saunders Inc., 1977:2973.

3. Ayres JR, Goldstrohm GL, Miller GJ, Dell PC. Proximalinterphalangeal joint arthrodesis with the Herbert screw.J Hand Surg 1998;13A:600–603.

4. Katzman SS, Gibeault D, Dickson K, Thompson JD. Use ofa Herbert screw for interphalangeal joint arthrodesis. ClinOrtho 1993;296:127–132.

5. Gomez CL, Proubasta I, Escriba I, Itarte J, Caceres E. Distalinterphalangeal joint arthrodesis: Treatment with Herbertscrew. J Southern Orth Assoc 2003;12:154–159.

6. Leibovic SJ. Arthrodesis of the proximal interphalangealjoint of the finger using tension band wiring or Herbert
screws. Atlas of the Hand Clinics 1998;3:17–30.

arthrodesis of the interphalangeal joints with headless compression screws

Documents