REVASCULARIZATION OF THE PROXIMAL POLE WITH IMPLANTATIONOF A VASCULAR BUNDLE AND BONE-GRAFTING*

BY DIEGO L. FERNANDEZ. M.D.L AND sTEPHAN EGGLI. M.D.-~. ’BERN. SWITZERLAND

Investigation performed at the Department of Orthopaedic Surgery, Lindenhof Hospital Bern

ABSTRACT: Eleven patients who had an ununitedfracture of the scaphoid associated with loss of theblood supply to the proximal fra~mnent were managedoperatively with a combination of an inlay cortico-cancellous bone graft from the lilac crest and implan-tation of the second dorsal intermetacarpal artery, itsaccompanying venae comitantes, and a thin cuff of:~erivascular tissue. The absence of the blood supply tothe proximal pole was evidenced both by radiographicchanges -- which included increased bone density,absence of normal trabecnlae, and cystic changes --and by failure to observe bleeding bone during theoperation.

There were ten men and one woman. The averageduration of non-union was fourteen months (range, sixto thirty-three months). Six patients had had previousunsuccessful operative attempts to obtain umon. E~ghtnon-unions ~ reLrfWiWthe proximal one-third and three,at the waist of the scaphoid. Union was achieved in tenpatients at an average of ten weeks postoperatively.

According to the wrist-scoring system of the MayoClinic, at an average of five years (range. 2.5 to elevenyears), three patients had a grade of excellent; three,goud; three, fair; and two, poor. Four patients had sub-sequent reconstructive procedures: radial styloidectomy,s~’loidectomy and resection of osteophytes, radioscapho-lunate arthrodesis, and total wrist arthrodesis were per-formed in one patient each.

The treatment df non-unions ot the scaphoid that autogenous graft from the iliac crest was employed in.are complicated by avascular necrosis of the proximal each patient .....fragment remains a challenge. The unfavorable re~- {,dc~;, ,,~,~sults reported by a number of investigators who have Materials and Methods

~]onal autogenous non-vascularized bone- From 1981 to 1992, ninety-six ununited fractures of

//grafting proceduresm~26 have led some authors to the scaphoid were treated operatively by the senior one

/recommend other operations.These have included exci- of us (D. L. F.) to obtain union. Eleven of these non-

|sion of the proximal fragment and replacement with a unions were thought to be associated with avascular

~prosthesis’~e3 or an allograft ~ as well as salvage proce- necrosis of the proximal segment, and they formed the

\ basis of the, current study.\ The presence of avascular necrosis was established\ *No benefits in any form have been received or will be received ............

w~ttl use ot specific radiographic and chmcal criteria~rom a commercial par’tv related directly or indirectly to the subject "o~this article.No fu~ads ~vere received in support of this study. - The radiographic criteria included an increase in the

\tDepartment of Orthopaedic Suroerv Lindenhof Hospital " v r nnn~n~-~. ,, , bone denstb, a loss of the normal trabecula,Br%aert~ea~Sttmr~SnS~~e~ai ol)~)rCthH£~0ale~icBejun;;e~Z~r~’vned~sit of Bern ance, collapse of the subchondral bone, cystic changes,

P P , Y , - .

selspit~CH-301.0 Bern, Switzerland. and deformity of the osseous segment. Clinically, a find-

Vf) L. 77-A,~. O. 6,]UN~Jg,95.¢1 ,. a ~ 0,,

"~ ~ L ~/ /0.-~ ’~ ’ ’ 883

dures such as resection arthroplasty or partial or totalarthrodesis ,of the wrists~~s.

Alternatively, several investigators have tried to ob- ..thin union in the presence of an avascular proximalsegment by providing additional vascularity to the siteof the non..union. This has been attempted throughthe implantation of a vascular pedicte~’t’~ or vascularizedbone~oraft-~-~.~:°:~-’~-~:. The former technique was first de-scribed by Hori et al. ’’~, who demonslrated, in experi-

mental animal models, active proliferation of bloodvessels and new-bone formation when a vascular pedi-cle or bundle consisting of peripheral vessels (an artery,venae comitantes, and perivascular .tissue) was im-planted into an isolated or necrotic bone. Success withimplantation of a vascular pedicle has been reported inthe treatment of Kienbdck disease~-’-m’:’~. To our knowl-edge, however, other than the original paper by Horiet hi. ’’~. which described the procedure in one patient,the English-language literature contains no reports ofthe use of this method for the treatment of a non-unionof the scaphoid with avascular necrosis of the proximalfragment.

In the current study, we report our experience withthe operative treatment of eleven ununited fracturesof the scaphoid in which the proximal segment was "demonstrated to be avascular. A combination of im-plantation of a vascular pedicle from the second dorsalintermetacarpal artery into the proximal segment of thescaphoid and use of an inlay corticocancellous strut of

884 _ . - ¯ D.L. FERNANDEZ AND STEPH,~N EGGLI¯ ..~_7.- ’ " " .._-Z~-:~-_.. __. .... . :: .... _ ,.

I Brick- M. 24 L Open Prox. 6 8 11 Mild 65/70 55/70 10115 2~/30 3~35layer reduct, third

+ int.fix. forcarpalfract.-disloc.

2 MechanicM, 20 L Below- Prox. 20 9.5 7 -- 65/75 50/60 10/20 35/35 4L’45the- thirdelbowcast

3 Truck M.28 R* Matti- Prox. 12 12 6.2 -- 5~/70 65/70 15/15 25125 40/40driver Russe third

inlaygraft

4 Manager M.30 R* Matti- Waist 10 I0 5.6 Se~: 3,9/7515/75 0/10 5i30 25"40 Arthro-Russe desis.inlay radio-graft scapho-

lunatejoints

16 I I 4.3 -- 7017065/70 20/20 25,’25 35;38 ¯ ~ 19 905 Student M, 22 L* Below- Prox.the- thirdelbowcast

6 Home- F, 27 R lntcrposit. Prox.maker grail third

fix.7 Carpenter M.35 L* Closed Prox.

redact, third~- below-the-elbowcast forfract.-disloc.

Rad. sty- 16 85loidect.

12 S0 (hood

18 80

32 45

8 12 3.9 -- 60;65 50170 10/15 30i35 20,L5 -- 15 75

Good

Poor

Exc.

Fair

8 Road M.31 R* None Prox. 33 Persist. 3 Sev. 0/70 0/75 0120 0/35 40145 Arthro- 24 35worker third non- desis.

union wrist9 Mason M,26 R* Interposit. Prox. 15 12 2.9 -- 7017565/70 10110 30/30 37/35 -- 17 100 Excl

graft, thirdscrc~wfix.

10 Roofer M. 19 L None Waist 9 8.6 2.7 -- 5017040/75 10120 20/30 42152 -- 14 7¢) FairII Welder M.29 R* Interposit.Waist 12 8 2.5 Mod. 4017560/70 15/15 15/30 40/55 Rad. sty- 24 65 Fair

graft, loidect..Herbert removalscrewy of osteo-fix. phytes

*Dominant hand.

ing of sclerotic bone without visible punctate bleeding fixation of an acute fracture associated with a perilunate

ja~o nts after debndement of the proximal segment was dislocation. In the remaining five patients, an estab-.~t./~_~ .__re~q~.ir.ed_to_..confirm.-thediagn-osi~~ lished non-union had been treated with a bone-grafting

~’,,~: ~...d~-i~Slqeither magnetic resonance imaging nor technetium- ] procedure (a Matti-Russe inlay bone graft in two [Cases~ .%9~,~ta~99m bone-scannino was used in t~’---~-- ~7-Z--’/

4]~-- ,~,~, ~ ............... ~. 3 rind and interpositional bone-grafting and screw.,>..Oz" - :leen-pa~ients-we-t-6-ifi-dff-~i-nd one was a woman; the fixation in three [Cases 6, 9, and 11]). I.J.0jkreeT~. tients.

average age was twenty-six years (range, nineteen to the non-union had developed after ~)treat-thirty-five years) (Table I). The dominant hand was fected in seven patients. The average duration of thenon-union of the scaphoid was fourteen months (range,six to thirty-three months). Six patients had had a pre-vious unsuccessful operative procedure. Of these pa-tients, one (Case 1) had had open reduction and internal

ment of the initial fracture. Of these three patients, two(Cases 2 and 5) had been managed with immobilizationin a below-the-elbow thumb-spica cast for three monthseach. and the third (Case 7) had had closed reductionof a transscaphoid perilunate dislocation and immobi-lization in a below-the-elbow cast for sixteen weeks. Two

THE JOURNAL OF BONE AND JOINT SURGERY

respecti+~-ly::: : :::-~~~:::~~-~: = :-pfindate ble~ding points are seen &e~--witgt~e~OUrni-~V:.:.L ....Eight non-unions were loca~d in the proximal one- quet inflated: however, if t~e surgeon is in doubt as to N ~ ~ ~~

third of the scaphoid and three, at the level of the scaph- whether or not a small red spot is in fact a vessel, the

old waist, tourniquet may be released and the surface may be" " " o o ........Preoperanvely, plato ra&o.raphs and tomo.rams observed during continuous ~rrx~anon w~th sahne solu-

sl:..~wed increased bone density and an absence of nor-mal osseous trabecutae in all patients. Additional cysticchanges and collapse of the subchondral bone of theproximal segment were observed in the seven patients(Cases 2.3, 5, 7, 8, 9, and ll) who had a longer durationof the non-union. ~Oi" ~

With the exception of one patient (Case 11). forwhom the preoperative radiographs exhibited a dorsallyrotated lunate with an increased scapholunate angle of7/? degre.es and signs of early periscaphoid osteoarthrosis.ad patients had normal carpal angles, with scapholunateangles ranging from 30 to 60 degrees and no radiographicevidence of associated carpal instability.

The most decisive criterion for inclusion in the studywas the clinical finding of a proximal fragment that wasdense, sclerotic, and avascular in association with a total

tion while the bleeding from the surrounding soft tissuesis blocked with a sponge.

After the absence of vascularitv of the fragment ofthe proximal pole in association with dense sclm:oticbone has been confirmed, the central portion of theproximal fragment is excavatedburrs for subsequent insertion of an~autooEnous peograft from the iliac crest. The diameter of this cavitvshould not exceed five millimeters. A rectangular troughis prepared on the dorsoradial aspect of the distal frag-ment with use of a small oscillating saw and small chis-els. The most distal transverse cut of the trough is madeslightly obliquely into the depth of the scaphoid in or-der to provide a trapezoidal surface to lock the graftinto place. After the trough in the distal fragment and acylindrical central cavity in the proximal fragment have

absence of bleeding points after d6bridement. Vascu!_cr- been prepared, the scaphoid is reduced and the length.

ity was assessed i,_._..~ntraoperatix’el~, accord" ,, ¯ " - width, and depth of the defect are measured in milli-

~,~s established by Green’~.~~s,i~ered ’~neters. A corticocancellous peg graft is cut off the iliac<oo~erou~.~!n~ ,t ,vas ][rest. with use ot’,power tools and small chisels, in ac-considered fair or poor if s~~ints w~re fordance with th~: measurements of the defect of the

"N~.bserved. The proximal polled avascular//scaphoid. ~dditional cancellous bone chips are obtained~e bleeding points Were seen. . ~ with a curet, and the wound over the iliac crest is closed.

~~%ya.t.,,~ ~ The fraaments of the scaphoid are separated slightlyOperative Procedure with a spreader clamp, and the graft is inserted in the

With the patient under general anesthesia and thelilac crest draped free, the scaphoid is approachedthrough an extensite dorsoradial exposure (Fig. 1). Theproximal part of the incision begins at the Lister tu-bercle and extends distally for four centimeters, run-ning parallel to the extensor pollicis longus tendon andending at the dorso-ulnar aspect of the base of thethumb. After identification and protection of the sen-sory branches of the radial nerve, an interval is devel-oped between the extensor carpi radialis longus andbrevis and the extensor pollicis longus. The wrist capsuleis incised along the long axis of the scaphoid. The arte-rial nutrient branches arising from the radial artery areidentified as they enter the dorsal ridge of the distalfragment, and they are protected throughout the proce-dure. Next, the site of the non-union is visualized, andthe interposed fibrous tissue is debrided with a scalpel,with care being taken not to damage the articular carti-lage. A small spreader clamp may be used to distract thesite of the non-union in order to visualize the ununitedsurface of the proximal fragment.

A sharp curet is employed to debride the scleroticsurfaces at the site of the non-union. If the surface of

VOL. 77-A, NO. 6, JUN~5~1995

central cavity in the proximal fragment.The graft is thenreduced into the trough of the distal fragment, with carebeing taken to ensure that the cortical surface of thegraft lies clorsoradially and that the surface of the non-union is not distracted by an oversized graft. Additionalcystic cavities and the site of the non-union may bepacked with free cancellous bone chips before the peggraft is reduced into the distal fragment. Depending onthe size of the proximal fragment, one or two smoothKirschner wires are used to stabilize the non-union.If the proximal fragment is very small (less than one-third of the size of the scaphoid), a single 1.0 to 1.2-millimeter Kirschner wire is inserted parallel to thepeg graft in the most radial aspect of the scaphoid. TheKirschner wire is inserted with a power drill underimage intensification without entry into the scapho-trapezial joint: it is then directed to the proximal p01eof the scaphoid. Placement of the wire should notjeopardize the point of entry of the vascular bundle,which will be slightly ulnar to the center of the proxi-mal pole, close to the scapholunate junction. If the prox,imal fragment is large enough, a second Kirschner wiremay be inserted in the most ulnar aspect of the scaphoid,

886 ....... D.L. FERNANOEZ AND STEPHAN EGGLt .. .......

t)~n~l~-~,~~t~ ~ole across

:~;~.:::~..~=~-..~,.~,~-,~,~-,~.~-~~~" :~-:~-_.~mast,: - ta~d.:~re~utl~,: ust~fn~r.to the ihac-creSt~-~

:- ~juncuon--~8 a~onented shghtly obhquely ~n a radto--...... ~~:’" .... .~} ~ ) volar affection. ~e vascular bundle is passed through

....... ..../" x~,;].,.:....... ~ ] t.~e drill-hole ~na dorsal-to-palmar d,recnon (F,g. 3. c).~~]:.:-~] ,~ ]2e passage is facilitated by the attachment of a fine

""-"" r.~absorbable suture to the end of the pedicte and the

F,~ I use of a straight needle to ~uide the suture and the

~ ~ vascular ped~cle through the bone Care must be takenFios 1 and 3"Schematicdrawin,,s showin the operative technique " ~ ¯ "Fi~. l:~e extensile dorsoradial incision for grafting and revascu- tO avoid a focal constriction of the pedicte as it is ro-

lariz~tion of~oid wit h use of the second dorsal intermeta-carpal artery and vein~

"~’H~e location of the inlay bone ~raft and the vascular pedicle as wellas the ideal position o( the ~irschner wires for fixation of the

with care being taken not to violate the scaphocapitatejoint (Fig. 2).

~e skin incision is then extended distally to thesecond dorsal web space, and the extensor tendons" I

to the index finger are retracted toward the ulna. ~esecond dorsal intermetacarpal artery and its venae com-itantes are identified between the first and secondmetacarpals. The vessels lie underneath a thin layer ofaponeurosis that covers the interosseous muscles (Fig.3, a). This fascial layer is split longitudinally to the levelof the second web space, and the vascular bundle iscarefully dissected from proximal td distal, with carebeing taken to elevate the artery and its venae com-itantes together with a thin layer of perivascular areolartissue. Small collateral branches found during dissec-tion are identified and coagulated. The vascular bundleis dissected up to the level of the web space, and thevessels are ligated with a 5-0 monofilament suture andare transected. Freeing of the second intermetacarpalartery and its venae comitantes from the base of thesecond metacarpal to the level of the web space pro-duces a pedicle, five to six centimeters long. that willreadily reach the proximal pole of the non-union.

After the vascular bundle has been elevated, a 2.7-

FI(?. a: The non-union of the scaphoid is exposed through a limited

dorsoradial capsulotomy. Note the vascular bundle of the secondintermetacarpal artery, lying between the first and second metacar-pals. b: An inlay graft has been inserted and a 2.7-millimeter hole hasbeen drilled in the proximal fragment: the vascular bundle has beenraised and is held with a suture, c: The vascular bundle is passedthrough the proximal pole.

JOURNAL OF BONE AND JOINT SIJRGERY

Nf)N-t;NION I)F rile S(’AI’II()ID

IS

points or constriction. Hemostasis is achieved carefully the patient was encouraged t~ use the hand for activitiesbefore closure of the wound. The wrist capsule is closedwith interrupted sutures at the proximal aspect of thearthrotomy; however, the capsule is left open at thei3oint of entry of the vascular pedicle, to prevent con-striction of the pedicle.

In patients in whom a previous operation has beenperformed through a palmar approach, the non-unionof the scaphoid is approached through the original inci-sion. If a screw is in place across the non-union, thescrew channel and the adjacent osteolytic cavities aredebridcd with use of small curets. Under fluoroscopiccontrol, the screw channel is enlarged to a diameter

~f approximately four to 4.5 millimeters with use ofprogressively thicker drill-bits and burrs. A corticocan-cellous peg graft from the iliac crest is inserted in thischannel from distal to proximal with a rounded impac-tot. and internal fixation of the scaphoid is performedwith txvo 1.0 to 1.2-millimeter-diameter Kirschner wires.The vascular bundle is thdn implanted through a sepa-rate dorsal incision (Figs. 4-A. 4-B. and 4-C), as de-scribed before.

DH,’fgO[?t’I’(I[[I’C ,~[tllHlgUlllt-’ll. t lllld

Postoperatively, an above-the-elbow thumb-spicacast was worn for four weeks, followed bya below-the-elbow thumb-spica cast for another four weeks. At eightweeks, the cast was removed, and the Kirschner wires

of daily living: however, a removable wrist brace wasrecommended for strenuous activities for another fourweeks. If tomographic evidence of union was uncertain,the below-the-elbow thumb-spica cast was worn for an-other two weeks, and anteroposterior tomograms of thescaphoid were again made.

The final evaluation was based on both subjectiveand objective criteria, including the patient’s satisfac-tion. residual pain. active range of motion of the wrist.grip stren,,th= . results of a sensor\’ examination, andability to work. The patient’s satisfaction was asses.sedbv askin: wrist wasbetter than before the operation and by asking if hesl~,;-ree to have thEo~ agatn ~n-t~~st were to la~iar cOndifi6n. Allpatients were asked whether they had pain in the wristduring normal activities of daily living or during work.Pain was considered mild if it occurred at the extremes .~,--of the active ranae of motion of the wrist, but the patient

was netther phvstcallv nor i~).’cholo,,tcallvXdtsturbed¯moderate tl the pattent ~vas phys~-ezffl,~ iSr psycholog~-callv disturbed, or both. during strenuous manual la-

bor: and severe if it occurred during ~tivitic~ of dailyliving and at rest, .~,jfkAA,-- ’

The active tahoe of motion of i-he wrist was mea-sured with a goniometer and was compared with thatof the contralateral wrist. Sensory evaluation included

~~’-." ...,,.e;2~ ~, il~:~,,,,~.~ ~1~.~’..:,.:.~". ~1¢:; ~ ~

FIG. 4-A F~G, 4-B F~G. 4-C

Figs. 4-A, 4-B, and 4-C: Case 1 l. Radiographs of a twenty-nine-year-old man.Fig. 4-A: A persistent non-union of the scaphoid with avascular necrosis of the proximal fragment. ~e non-union had failed to heal after

interpositional bone-grafting and fixation with a Herbert screw.Fig. 4-B: Immediately after the index operation.Fig. 4-C: Eight months after the index operation, the non-union has healed.

VOL 77-A. NO. 6. JUNE 1995

Fir;. 5-A FI,:~. 5-B FI(;. 5-C

Figs. 5-A through 5-G: Case 2. A twenty-year-old man.Fig. 5-A: Radiographic appearance of the non-union of the scaphoid with avascularity of the proximal fragment, before the index operation.

’Fig. 5-B: Radiograph made immediately postopera~.i;’el.v.Fig. 5-C: Tomogram made at nine and one-half weeks, showing union.

F~c,. 5-D ...... FIG.

(~..~ ~~0 "~ Radiographs made one and four years postoperativeb:

l~h and pin-prick sensibility tests and a two- with residual pain, functional status, range of motion,point discrimination test of the median, radial, and ulnar and grip strength being granted a maximum of 25 pointsdermatomes. Particular attention was given to altera- each. A score of 90 to 100 points indicates an excellenttions of sensibility and to the presence of a neuroma of result: 80 to 89 points, a good result; 65 to 79 points, athe superficial radial nerve at the level of the operative fair result; and less than 65 points, a poor result.scar. Grip strength was measured with a Jamar dyna- For the radiographic assessment of the early results,mometer (Therapeutic Equipment. Clifton, New Jer- the criterion for union was the disappearance of a gapsey), was expressed as an absolute value of kilograms- at the site of the non-union, with evidence of bridgingforce, and was compared with that of the uninjured bone trabeculae on both plain radiographs and trispiralwrist. The ability to work was evaluated on the basis of tomograms. Restoration of normal bone density equalwhether or not the patient had returned to his or her to that of the neighboring carpal bones as well as re-original occupation and was able to work full-time (100 appearance of normal cancellous trabecular structureper cent) or was restricted (to 25 to 50 per cent of the on plain radiographs were considered signs of revas-normal working time), cularization. Irregular areas of patchy sclerosis in the

For the final combined assessment of the late results, proximal pole were interpreted as a sign of partial re-the wrist-scoring system of the Mayo Clinic: was used, vascularization. For the radiographic assessment of

TIlE .I()URN,.\I. ()[.’ P;()NE AND JOINT

t) indicating a normal appcarance:-~rade:t2slight nar-rowing of the joint space: grade 2. marked narroxvingof the joint space and formation of osteophytes: and_,-ad~ 3. complete loss of the joint space, formation oL,stcophytes, and subchondral bone cysts.

Ten of the cloven non-ttnions healed at anof ten weeks (range, cigi~t to txveive weeks) after useo~catmcnt protocol. Four patients (Cases I. 2,and 1!) had evidence of union when the KirsclmcFxxircs were removed and the initial tonlo~i’alllS wcrcnladc: the rcmainill~ sexon ~)atlCnts nccdcd :Idditionai:nmobiiization in the bclow-ihc-clbow thumb-spica cast"cCtltthd" ill" ~IlCOIlC~LIsivc [OIllOEFLIIIlS. The cast was rc-

:1loved at tw{}-wcck inlcFvals, alld l’CpCat antcrop{}s-

.~, iOl tomt}Er:lltls xvcl-c llladc. Of these seven patients.

three had hcal~ne ut ten {}r c!cvcn weeks, fl]rcc had

’,llC ll~)ll-tllllt)ll. £I’:l~Illcll[:ltiOll. :lIld Ct)ll:~psC Of ’,[IC

:If {ilC wrist dcsDitc F:Idio~t’ttp~lic LHliOIU ..kt the tilllC

’ hc Fc~q~cr:ttio~l. ci~ht illOI1[11s a(tcr the rcco~lstructio~l

~hc Fadit~scaphoid ioint were found, and an arthrodcsis~l ~hc r:~dioscapho[unatc .ioints was pcrtornlcd. Two act-dititmul pnticnts had Fcsiduat pare in the wrist, wl~ichwas mild in one (Case 1) and moderate in the otherI Case t t): both had radiographic evidence of impinge-mcntot+ the scaphoid on the radial stvloid process xvithearly degenerative changes, evidenced bv a slight nar-rowing of the joint space. The pain in these two patientswas Fe~ieved by a radial styloidectomy in one {Caseand the removal of ostcophytes in the other {Case ll),at seven and ten months, respectively, after.the index,+potation.

No other early complications, such as pin-track in-tecfion or sensory disturbances in the area of the radial

nerve, were observed.

k,~e i~e,s’t~lts

The results were assessed at a mean of five veaFs(range. 2.5 to eleven years) postoperatively. With theexception of the two patients (Cases 4 and 8) who hadhad a secondary arthrodesis, all patients were satis-~icd with the late result. The two patients {Cases 1 and!1~ who had subsequently had a radial stvloidectomvwere pain-free at the latest tollow-up evaluatione

dominant hand than in the non-dominant hand at thelatest follow-up evaluation, one (Case 3) had equalstrength in both hands, and four had less strength (fif-teen kilograms-force less in two [Cases 4 and 11], fivekilograms-force less in one [Case 8], and three kilo-grams-force less in one [Case 5]) in the dominant hand.

Ten of the eleven patients returned to their originalwork at an average of 4.8 months after the operation.The remaining patient (Case 8), who had had a totalwrist arthrodesis and had been a road worker, had tochange his occupation as he was no longer able to use apick and shovel without difficulty.

Radiographic evidence of degenerative changes wasnoted in three patients at the latest follow-up eval-i/ation. In two of these patients (Cases 1 and 7), theoriginal injury had been a transscaphoid perilunate dis-location, which may account for the development ofsecondary osteoarthrotic changes at the level of themid-carpal joint. In both patients, the changes weremild, with slight,narrowing of the joint space and for-mation of osteophytes on the dorsal surface. The third

¯ patient (Case ll) had evidence of initial osteoarthroticchanges of the radioscaphoid joint, with formationof osteophytes at the level of the radial stvloid pro-cess. Comparison of the radiographs made soon afterhealing of the scaphoid with those made at the latestfollow-up evaluation failed to show any deteriorationof the scapholunate angle or of the carpal height in anypatient.

The over-all tunctional result, according to the wrist-scoring system of the Mayo Clinic:, was excellent inthree patients, good in three, fair in three, and poor intwo. Analysis of the three unsatisfactory results showedthat the treatment had failed to provide revasculariza-tion in one patient (Case 8), while two patients (Cases4 and 11) had residual, disabling pain in the wrist de-spite union. Retrospectively, the initial periscaphoid de-generative changes were underestimated, and perhaps asalvage procedure should have been performed insteadof revascularization. Relief of pain after an arthrodesisof the radioscapholunate joints (Case 4), a total arthro-desis of the wrist (Case 8), and a styloidectomy (Case11) was satisfactory in all three patients.

SCrA~tiz~SCA~Sweenl~v;Pea°rr-tol~rc an~su’stained a fractimal third of the scaphoid on the left, non-dominant side and wasmanaged with a below-the-elbow thumb-spica east for three months.The fracture failed to unite; however, the patient had no symptomsreferable to the non-union.

Twenty months after the initial injury, a second injury occurred,and the patient began to have pain with strenuous activity. Antero-posterior, lateral, and scaphoid radiographs revealed an ununited

its venae comitantes (F~ig. 5-B).Nine and one-half weeks postoperatively, anteroposterior tomo-

grams showed osseous union and incorporation of the inlay graft (Fig.5-C). Radiographs made one year later revealed that the scaphoidhad regained a normal-appearing osseous contour, and there wasevidence of bone trabeculae brid~in~ the site of the non-union (Fig.5-D). The vascular channel was still visible.

Four years postoperatively, the radiographs revealed normal can-cellous bone trabeculae in the tip of the proximal pole without evi-dence of increased sclerosis (Fig. 5-E). At the time of the latestfollow-up examination, seven years postoperatively, the patient wasfree of pain and the range of motion of the wrist was 65 degrees of .~!palrnar flexion. 50 degrees of extension, 10 degrees of radial deviation,and 35 degrees of ulnar deviation (Figs. 5-F and 5-G).

Discussion

The healing potential of an ununited scaphoid de-pends on two factors: vascularity and stability. Becauseof iits vulnerable blood supply and the loss of retamtml~tmentous support, unstable and proximal non-tinionsof-tfie-sn~phoicl nave tgeen assoctat.ecl w~tn decre,qsed(rates-g-6f-finion after conventional bone-7---’--

eeC~:7 With the advent of improved techniques ~::of~ternal fixation in combination with autogenous ~:

bone-grafting, a substantial improvement in the rateof union of unstable non-unions of the scaphoid hasbeen reported bv several authors’.m". However, theinlay-grafting procedure of Russe has resulted in a highrate of union of stable non-unions and has also beenrecommended for scaphoids that have failed to uniteafter an initial bone-grafting procedure-~*.’3.

The rate of success with conventional grafting islower when the proximal pole of the scaphoid is com-pletely avascularm"m-’~. GreenL~ analyzed the effect ofimpaired vascularity on the results of the Russe.pro-cedure and reported union in ten of fourteen patientsin ’whom the vascularity was spotty or diminished andfailure of the fracture to unite in five patients in whomthe proximal pole was totally avascular. Green sug-gested that the best indication of true avascularity ofthe proximal pole is the absence of punctate bleedingpoints on the cancellous surface.

More recently, it was suggested that magnetic reso-nance imaging of the scaphoid allows a more accuratedetermination of the state of the bone circulation:~.>.

Urban et al. 3° performed both a qualitative and a quan-titative histomorphometric analysis of six scaphoidspe, cimens. Four of these specimens had a non-union,with presumed avascular necrosis of the proximal poleand increased radiographic density; one was obtainedfrom a patient who had Preiser disease; and one wasan intact scaphoid, removed as part of a proxima!-rowcarpectomy, which served as a control. Two-millimeter-thick longitudinal and coronal sections from each ~spec-

THE JOURNAL OF BONE A INT SURGERY

central part 6ftfi~ro~m~fpble~as mor6 a fte~t~-~-t~hn :7 ~e6~:~ss crib~db~ B ehUn:~i Leung~nd Hfihg~ and K~wfiiwas the adjacent palmar or dorsal surface. In one pa- and Yamamoto~. Kuhlmann et alY reported the success-tient, the findings on preoperative magnetic resonanceimaging correlated with the histological findings. An in-teresting additional observation was that each specimenhad histological evidence of mild. moderate, or severedegenerative changes in the surrounding cartilage. Ur-ban et al. concluded that random intraoperative biopsyalone is not accurate for the diagnosis of avascular ne-crosis of the entire proximal pole of the scaphoid andthat the clinician must rely on other methods to evaluatethe vascular status.

Vascular impairment of the proximal segment of ascaphoid fracture can be separated into three catego-ries: transient ischemia: partial or reversible necrosis:and complete, irreversible necrosis~s. In transient isch-emia. the proximal pole is temporarily deprived of itsintraosseous circulation because of the fracture: thecirculation is restored when the fracture unites. Tran-sient ischemia is associated with increased radiograp~cde"h~sit~ withoht subc,hondral collapse.deformity, orabnormal bone trabeculae. Partial or reversible avas-cular necrosis, as seen in some instances of non-union,rel~eased bone densitv associated with a dis-appearance of normal cancellous bone trabeculation.deformity, and cystic changes. Complete, irreversible ne-crosis is characterized by the chfi.ffges of partial n~c~o~’isa~ll as by subchondral collapse and fragmentationsimilar to those seen in the late stages of Preiser orKienb6ck disease. At this stage, revascularization mavnot be possible since fragmentation is associated withpermanent structural changes of the bone matrix"-". Notonly is the vascular supply absent, but there is also in-congruity of the joint surfaces because of post-ischemiccoll~pse and fragmentation as well as degenerativechanges of the cartilage envelope. For this reason, an .operation that is done to promote union will compro.~_-.i~ise the chances--Td~pain r_e_lief and for revention~~~eq_ue_nt___o~e_o_a~t_~ro_ticchang-~6-g-6f the w~st. However, if the proximal fragment~s ~schem~c but not deformed and there ts no ewdenceof advanced periscaphoid osteoarthrosis or of estab-lished carpal collapse, it seems that a logical way topreserve the anatomical integrity of the scaphoid is toaccelerate union with a revascularization procedure.

Although there is a chance that ischemic proximalpoles will be revascularized with Russe inlay-grafting~-~’~,

the results with this method have been unpredictable,especially when the recipient bone has been completelyavascularm~’’-~. If there is clinical and radiographic evi-dence of avascutar necrosis or of recurrent pseud-arthrosis due to instability, revascularization may beaccelerated by the use of vascularized bone grafts or bythe implantation of a vascular bundle as well as a con-

ful use of a radial graft t~at was pedicled on the radialbranch of the volar carpal arch in the treatment of threechronic non-unions after a Matti-Russe operation hadfailed. Brunelli et al. ~ used. for the same indication, avascularized bone graft from the radial distal aspect ofthe second metacarpal: the graft was pedicled on thesuperficial dorsal interosseous artery.

Zaidemberg et alY reported success with the use ofa vascularized bone graft from the radial stvloid processin eleven patients who had had a long-standing non-union and a failed Matti-Russe procedure. The graft waspedicled on a constant ascending retrograde branch ofthe radial artery, that runs deep to the first dorsal exten-sor compartment at the level of the radial stvloid p, ro-~,cess. The average duration to union in their series wa~.2jweeks (range. five to eight weeks). Guimberteau"arN~" Panconi" used a vascularized bone graft, obtained fromthe ulna and based on the ulnar artery, to treat recurrent

non-unions of the scaphoid in eight patients, all of whomhad had at least two previous failed procedures. Healingoccurred at an average of 4.6 months in all eight patients.The major disadvantage of this approach is the need totransect the ulnar arterv to move the graft distalward.This necessitates restoration of arterial continuity with afree venous graft. Pechlaner et aid reported the success-ful treatment of twenty-five non-unions of the scaphoidwith a free vascularized iliac-crest graft based on a vas-cular pedicle from the deep iliac circumflex artery. Thisprocedure was indicated for small, avascular proximalpoles and for lone-standin~ non-unions, for which therate of union at tl~eir institution had been as lowper cent with use of conventional Russe procedures. ,.-v~’ ,j,~,~

Implantation of a vascular bundle is an alternative/method that provides direct capillary ingrowth to avas-cular bone. Its clinical use. with or without curettage of~the central necrotic areas of the lunate, has gained pop-ularity in the treatment of the early stages of Kienb6ckd~sease -’. The advantages of the procedure are its sim-

ne~tand the shorterplicity operative time: there is nofor complex vascular dissection of arterial pedicles orfor microvascular reconstruction. Furthermore, manip-ulation of a free peg graft from the iliac crest is easierthan that of a graft that is attached either to muscle orto a vascular pedicle. Insertion of a smaller free graftfrom the iliac crest requires a smaller trough, and moreof the cartilage surface of the scaphoid can be pre-served, thereby diminishing the possibility of late os-teoarthrotic changes.

The vascular pedicle that was used in our studyproved to be of sufficient dimensions and was present ineach patient. An alternative to use of the second dorsalintermetacarpal vessels is use of the dorsal branch of

’̄ ~ wrist:at-t~:legel of the rsdio:ulnaf joint7It pr0videg: ~finSl-~agm6nt Bf:a ~caph0ih- find-Union s~ow; radio-- --~nutrient branches to the lunate and capitate and anas- graphic evidence~f~va~cularity, appears sclerotic, andtomoses distally with the dorsal carpal arch. is deprived of ble~g points intraoperatively and in

We believe that the acceptable results in our smh~lwhich there is no~%22~’o¢ .... .....series were due to the fact that both factors that govern tiV4 changes, thethe healing of a non-union ~ stability and vascularity divan attempt to gain union with this approach.m were addressed simultaneously. Although we areaware of the lack of experimental models, biopsy mate-rial, and comparisons with similar series, we postulatethat ultimat6 revascularization and healing of the non-union is probably obtained through two sourCes of cap-illary ingrowth: a direct source, through the vascularpedicle, and an indirect source, through incorpora-

Given the small number of patients treated with re-vascularization procedures reported to date and thelack of randomized series, the choice of a particularmethod still remains based on the personal experience;of the surgeon. The combined technique of inlav bone-grafting, internal fixation, and implantation of a vascularbundle is an attractive alternative to vascularized bone--

tion and revascularization of the inlay graft from the grafting procedures.

I. Bochud. R. C. and Bfichlen U.: KienbOck’s disease, early stage 3 -- height reconstruction and core revascularization of the lunate.J. Hand Surg., 19-B: 466-478, 1994.

¯2. Bradway, J. K.~ Amadio, P. C.; and Cooney, W. P.: Open reductioa and internal fixation of displaced, comminuted intra-articularfractures of the distal end of the radius. J. Bone and Joint Sttrg.. 7 I-A: 839-847. July 1989.

3.Urbaniak.Braun" R. st.M’:Eouis.Viablec.pediclev. Mosbv.bone1987.grafting in the wrist. In Microsurgery for Major Lhnh Reconstruction. pp.--0-__9. Edited by. J. R.

41 Brnnelll, F.t Mathnnlin. C.: and Saffar, P.: Description d’un greffon osscux vascularisd prdlcvd au niveau de la t6te du deuxi6mem4tacarpien. Ann. ddr. main. 11: 40-45, 1992.

5. Carrnzzellao ~. C.; Stern, P. J.: and Mnrduek, P. A.: The fate of failed bone graft surgery for scaphoid nonunions. J. lhmd &irg.. 14A:8{}{}-806. 1989.

6. Carter. P. R.: Malinin. T. i.: Abbe); P. A.: and Sommerkamp. T. G.: The scaphoid allograft: a new operation for treatment of the very¯proximal scaphoid nonunion or for the necrotic, fragmented scaphoid proximal pole.J, lhmd Surg.. 14A: 1-12.1989.

7. Cooney, W. P., llh Dobyns, J. H.; and Linscheid. R. L.: Nonunion of the scaphoid: analysis of the results from bone ~2raftina. J. thmdSttrg.. 5: 343-354. 1980. " ......

8. Cooney, W. P.; Linscheid, R. L.; and Dobyns, J. H.: Scaphoid fractures. Problems associated with nonunion and avascular necrosis.Orthop. Clin. North Anterica. 15: 381-391. 1984.

9. Coone); W. P.: Linscheid. R. L.: Dobyns, J. H.; and Wnud. M. B.: Scaphoid nonunion: role of anterior interpositional bone grafts. J. ~haulSurg., 13A: 635-650. 1988.

I0. Fernandez, D. L.: Anterior bone grafting and conventional lag screw fixation to treat scaphoid nonunions. J. Hand Sttrg.. 15A: 140-147, 1990.

11. Fernandez, D. L.: Scaphoid non-union: current approach to management. In ~Vrist Disorders, pp. 156-164. Edited by R. Nakamura. R. L.Linscheid, and T. Miura. New York, Springer, 1992.

12. Foucher, G., and Saffar, P.: Revascularization of the necrosed lunate. Staze I and Ii With a dorsal intermetacarpal arteriovenous pedicle.J. chir. main, I: 259,1982.

13. Green, D. P.: The effect of avascular necrosis on Russe bone grafting for scaphoid nonunion. J. Hand Surg.. 10A: 597-605, 1985.14. Guimberteau, J. C, and Panconi, B.: Recalcitrant non-union of the scaphoid treated with a vascularized bone graft based on the ulnar

artery. J. Bone and Joint Sttrg., 72-A: 88-97, Jan. 1990.15. Herbert, T. J.: The Fractured Scaphoid, pp. 121-138. St. Louis. Quality Medical, 1990.16. Herbert, T. J., and Fischer, W. E.: Management of the fractured scaphoid using a new bone screw. J. Bone and Joint Surg., 66-B(1):

114-123. 1984.17. Herness, D., and Posner, M. A.: Some aspects of bone grafting for non-union of the carpal navicular. Analysis of 41 cases. Acta Orthop.

Scandinavica, 48: 373-378, 1977.18. Hori, Y.; Tamai, S.; Okuda, H.; Sakamoto, H.; Takita, T.; and Masuhara, K.: Blood vessel transplantation to bone. J. Hand Sttrg.. 4:

23-33, 1979.19. Hull, W. J.; House, J. H.; Gustiilo, R. B.; Kleven, L.; and Thompson, W.: The surgical approach and source of bone graft for symptomatic

nonunion of the scaphoid. Clin. Orthop., 115: 241-247, 1976.20. Kawai, H., and Yamamoto, K.: Pronator quadratus pedicled bone graft for old scaphoid fractures. J. Bone and Joint Surg., 70-B(5):

829-831, 1988.21. Kuhlmann, J. N.; Mimoun, M.; Boabighi, A.; and Baux, S.: Vascularized bone graft pedicled on the volar carpal artery for non-union of

the scaphoid. J. Hand Sttrg., 12-B: 203-210, 1987.22. Leung, P. C., and Hung, L. K.: Use of pronator quadratus bone flap in bony reconstruction around the wrist. J. Hamt Sarg., 15A: 637-

640, 1990.23. Mulder, J. D.: The results of 100 cases of pseudarthrosis in the scaphoid bone treated by the Matti-Russe operation. J. Bone a;ut Joint

Surg., 50-B(I): 110-I 15, 1968.

THE JOt/RNAL OF BONE AND JOINT SUP, GERY

NON-UNION OF TIlE $CAPHO1D 893

.... ~. ¯ "4. ~Pechlanei’, $4_ Hussl. H.I and-K~nzeI, K. H.: Al.ter~tive ~perationsmethode bei Kahn~einpseudarthrosen Prospektive Studie. Hand-

7- " 28. ~al, ~2 Yaji~’H~fMiZUmrtd~/~=~H~]~ 7~e~t~n~f~v~i~b~~ ~ith qascular bundle-:implantation. In Book o/ "

Abstracts of the 45th Annual Meeting of the American Society for Surgery of the Hand. Toronto. Ontario. Canada. 1990.29. Trumble, T. E.: Avascular necrosis after scaphoid fracture: a correlation of magnetic resonance i~aging and histology. J. Hand Surg.,

15A: 557-5~, 1990.30. Urban, M. A;; Green, D. E; and Aufdemoae, T. B.: ~e patchy configuration of scaphoid avascular necrosis. J. Hand Surg., 18A: 669-

674, 1993..; 1. Verdan. C., and Narakas, A.: Fractures and pseudarthrosis of the scaphoid. Surg. Clin. North America. 48: 1083-i095. 1968.32. Zaidemberg, C,; Siebe~, J. W.; and Angrigiani, Cz A new vasculafized bone graft for scaphoid nonunion. J. Hand Surg.. 16A: 474-

478, 1991.33. Zemel, N. P.; Stark, H. H.; Ashwo~h, C. R.; Rickard, T. A.: and Anderson, D. R.: Treatment of selected patients with an ununited

fracture of the proximal part of the scaphoid by excision of the fragment and insertion of a ca~ed silicone-rubber spacer. Z Bone andJoint Surg., 66-A: 510-517, April 19~.

VOL. 77-A. NO. 6. JUNE 1995