The PDF of the article you requested follows this cover page.  

This is an enhanced PDF from The Journal of Bone and Joint Surgery

1970;52:1090-1103. J Bone Joint Surg Am.CHARLES S. NEER, II    

THREE-PART AND FOUR-PART DISPLACEMENTDisplaced Proximal Humeral Fractures: PART II. TREATMENT OF

This information is current as of June 5, 2010

Reprints and Permissions

Permissions] link. and click on the [Reprints andjbjs.orgarticle, or locate the article citation on

to use material from thisorder reprints or request permissionClick here to

Publisher Information

www.jbjs.org20 Pickering Street, Needham, MA 02492-3157The Journal of Bone and Joint Surgery

Displaced Proximal Humeral Fractures

PAnT II. TREATMENT OF rfHIOEEI)A1OT ANI) 1�OUR-PART DISPLACEMENT*

BY CHARLES S. NEEIt, II, M.D.t, NEW YORK, N. Y.

Freiii the Dparlnient of Orthopaulic Surgery, College of Physicians and Surgeons-, Columbia

L.nivcrstty, (111(1 The New York ()rthopacdic IIo.s-pital, Columbia-Presbyterian iledical Center,

�Vew York

Iii the �)rece(ling article, a descriptiotu 1115(1 classification of fractures o)f tlse

proximal euid (.)f thue humerus was Presetste(i. Sitsce the select.iots of treatment. seemed

nu)st. (hifficl.llt its patients with three-part. and four-part displacen�ents, it was

tisouglit importatit. to compare tise results of the various methods used its recent

years. The j)urpose of this study was to analyZe the results of closed reduction, open

reduction, ausd l)rosthset.ic reconstruction atsd to cotssider tise teclsnical probleins

etscountered itt the treat.metst of these specific groups of fractures.

Material and Method

A consecutive series of 1 17 j)atietsts �vit.hs displaced three-part and four-part

fractures atid fracture-dislocations, followed for a minin�um period of one year, were

atiItlyze(I. All were treated by members of the permatsetit staff of the Xew York

Orthiopaedic Hospital-Columbia-Presbyterian Medical Center between the years

1953 lUi(l 1969. �fh age of tise patietits raisged from twenty-five to eighty-four

years 1115(1 averaged 53.3 years. The distribution of their ages by decades was: third,

two; fourth, teti; fifth, tweisty-otse; sixth, forty-three; seventh, tlsirt.y-otse; eighths,

nine; 1111(1 uiiusthi, one. rfls(� 1Ett.ieIstI4 1usd beets followed for from otse to sixteen years;

the ltverltge j)et’iO(l of follow-up after itsjury waS 4.5 years. Thirty-seveti pat.ieists

were followed for fi’om oi5e to two years; forty-six, two to five years; fourteen, five to

tets s-ears; Iuti(l twenty, ten to sixteets years. r1�l)(� results in these patieusts were rated

by the nutnet-ical systens (lescribe(l its the j)recednsg i)aper.

11’elttfllo’tst of the eustire series, according to classification, is show-n its Table I.

Closed reduction utsder atiesthesia was atten�j)ted in seveisty-seven patients and t.lse

1)Ositiotl \\�ItS accepted iii tlsirty-otse patietst.s ��-ho were theti treated by a Velpeau

bandage, a hanging cast., or overhead ultiar-piti tractiots. Opeti reduction was

performed iii forty-three patients iii whom the articular segmetst. was discarded itsfive atsd internal fixation was used iii thse remaitsing thirty-eight: ��-ire loops its

sixteeti, Rush nails in eight, silk or nylon in six, splnses in three, Kirschtser wires or

screw-s its fout’, 11usd staples in one. Prosthetic replacement of the Isead atsd recots-

structious of the tuberosit.ies were (lone �fl fort.v-t.hsree patients.

Findings and Results

‘I’/i ree-Part Fractures and Fracture-Dislocations

Group Jj/

There were twetity-three fractures of this type, in w-hichs there was a displace-

meist of over 1.0 centimeter of the greater tuberosity and the shaft but the articular

segmerst remaitsed its contitsuity w-it.h the lesser t.uberosit.y. The attached .subscapu-

* Head iii part at the Annual Meeting of The American Academy of Orthopaedic Surgeolss,

New York, N.Y., January 21, 1969.

t 161 Fort Washingtotu Aveisue, New York, N. Y. 10032.

1090 THE JOURNAL OF BONE AND JOINT SURGERY

DISPLACED PROXIMAL HUMERAL FRACTURES 1091

TABLE I

TREATMENT BY CLASSIFICATION

No. of Closed Open

Patients Treatment Reduction Prosthesis

Three-PartGroup IV 23 7 14 2GroupV 13 3 7 1

VI anterior 17 6 5 6

VI posterior 8 2 4 2

Four-Part4-part fracture 19 7 5 7

VI anterior 31 4 3 22VI posterior 6 0 3 3

117 31 43 43

TABLE IIRESULTS OF CLOSED REDUcTu0N

No. of ReductiotsPatients Accepted

Satisfactory

Result

Three-PartGroup IV 12 7 0

Group V S 3 0VI anterior 13 6 2VI posterior 6 2 1

Fonir-Part

4-part fracture 12 7 0

\.l anterior 23 4 0\_I posterior 3 0 0

3

hans tendots interrsallv rotated the head, exaggeratitsg the defect. in the rotato)r cuff

and causing the articular surface to face posteriorly (Fig. 3-A).

Closed reductiots was consistently unsuccessful its twelve patietsts ; how-ever,

closed treatment in overhead tractiots or hanging casts was contitsued in six patietsts.

In two patients, treated its Isanging casts, non-unioti developed at the surgical-neck

level (Fig. 1-C). In the other four patients disabling malunioni� occurred, with resorp-

tion of the head in one (Table II, Fig. 1-A). Open reduction was performe(l its four-

teen Patieists (Figs. 3-A, 3-B, and 3-C), two of whom had beets initially treated

in traction. Their results were: four excelletst, five satisfactory, atsd five failure. The

five poor results from opets reductiots followed technical errors consistitsg its failure

to approximate tise tuberosities and to repair the defect in the rotator cuff. The

vertical fixatiors (Rushs nails, sphine.s, or Kirschtser wires) failed to hold the tuberos-

ities in apposition (Fig. 6-B). Noti-utsions occurred itu three of these patients at the

level of t.he surgical neck (Table III). A prosthsesis ��-as used early its the series itu

two patients and their results w-ere rated as one excellent and one satisfactory

(Table III).

Group V

In thirteen fractures there w-as displacemerst of the lesser tuberosity and shaft

w-hile the greater tuberosit.y remaitsed its continuity withs the articular segment. Thse

external rotators caused the articular surface to face anteriorly (Fig. 4-A).

Eight were treated by closed reduction without success. Closed treatment was

continued in five patieists, four in overhead traction, and one in a Velpeau bandage,

VOL. 52-A, NO. 6. SEPTEMBER 1970

Fm. 1-13 Fun. 1-C

Fig. 1-B: Atutetoposterior roeuutgetl((gt-atli sh((witug tnalmutuioiu of a ( u-loup-\ t litee-paut fi:ict mite iuuwhich t he aut ici tInt si uuface is fm-i tug �iiuI eti( )il\.

Fig. 1-C: Non-utuiotu of a ( tomip-I V t htee-patt ft:ict mire treated itua haiugiiug (‘list-

1092 C. S. NEER, II

THE JOURNAL OF BONE AND JOINT SURGERY

Fig.. I-A, I-B, at uI I-C: 1 ((‘i ut gem gtams sh Wj tug t he utus:utisfact ((iv result S (If (-II )se(l t (1(1 tiwttt

fou t lutee-1I:(tt fuact uu-e�.Fig. 1-A: Axill sty oct utgem gram sb wit ug t he nsau-ked 0(11511 (iv (lisI)la(-etsselut it u a ul)almuuiit edi-(I(ttl-IV thuee-pait ftaet tie. 1111’ liii i(’Iulai surface faces backward atu(l i� latgelv c veted liv t lie

great ci tibet-sit .

followed by futictiouial exercises. TIse fractures utsited but with sigtiificatit rotational

deformity (Fig. 1-B), and much of’ thse head resorbed in otie pat ietst. One elderly

pat ietst was satisfied with hiet’ t’ecovet’y but all teceive(l poor tiunserical rat itigs(Table II). Opets reduction was performed its seveti patietits (Figs. 4-A, 4-B, atid

4-C). The results were: two excellent, two satisfactory, 11usd t hree failut’es. Each

failure was (hue to loss of fixatioti followitig thse use of a Rusls nttil or sphitse. Wire-loop

fixatioui auid cuff repair cotssistetstly gave satisfactory results (Table III). Ouse

VOL. 52-A, NO. 6, SEPTEMBER 1971)

I)ISPLACED PROXIMAL HUMERAL FRA(TUIIES 1093

patietit had tt torts axillary artety that required a vein graft 1usd, (luritsg this proce-

dure, the fracture was tteated w’ith a prosthesis. I’�our yeat’s later site was givets au

unsatisfactory futsctional rating, largely, hsow-ever, because all three cords of the

brachiial h)leXus 1Usd beets damaged by the itsjury (Table III).

Group VI, A oterior

There were seventeen subcoracoid fracture-dislocatiouss of this type, its �vhicls

the lesser tuberosity and its soft parts remaitsed its coustiuiuitv with the articular

segment.

Closed i’ecluctiots was attempted its thirteen pat ietsts ausd foutid to be (lifficult.

The subscapularis, attached to the head,

tended to prevetst its being reduced (Fig.

2-A). Eyeii when reduction was accom-

plislsed. the subscapularis often rotated

the Isead so tlsat thse articular surface faced

posteriorly, causing it to appear to be

its the upside down position (Fig. 2-B).

Closed reduction was accepted its six

patieist.s, two of whsons obtained satis-

factory ratings. Late rotatiotial displace-

ment by the subscapularis led to limited

range utah discomfort in the four otlser

patieists treated by closed methods (Table

Ii). Open reduction, wire-loop fixatiots,

and cuff repair resulted in good ratitsgs ins

three patietits. Spotty avascular changes

its the head were the rule, but. the cir-

I ________FIG. 2-B

Figs. 2-A, 2-B, auud 2-C: Auit eu’opost eu-iou- r )eiut geuuogu’ams sh wiuug t lie diffic illles etuc( tit u I eted iii

(‘lOse(l te_itt(’t i(uuof fu’act ut’e-dislocat iouus.Fig. 2-A: Auuteu’iou- t hu’ee-pau’t lesiouu showuu while I tact iouu was applied tutudet atuest hiesia. The

suthscaptulatis, at t ached to the head, pu’eveuuted its beiuug u’educed. -

Fig. 2-B: Same type of fum’t uu’e-dislocat iou as that showuu itt Fig. 2-A followitug (‘lose(lu’e(lmucttot.

The head appeat’s to be upside dowtu becamtse the stubscaptularis has (lIsted it 50 that I he alt icularo’au-tilage faces posteriou’lv.

Fig. 2-C: htucteased displacemeuut atud bu�u’hial plexus svmptotiis followitug t hitee attenipts tou’edtuce auu auuteiiou’ fommu-paut lesiouu by matuiptmlatiouu.

Three-partGroup I� 14 4 5 5 2 1 1GroupV 7 2 2 3 1 1VI anterior 5 2 1 2 Ii 4 1VI posterior 4 2 1 1 2 2

Four-partFour-part

fractureVI anteriorVI posterior

Ft,. 3-A

FIG. 3-B FIG. 3-C

THE JOURNAL OF BONE AND JOINT SURGERY

1094 c. s. NEER, II

TABLE III

O1’ERATJv: RESULTS OF EIGI(TY-SIX FRACTURES AND FRACTURE-DISLOCATIONS

Open Re( luction (Forty-three) Prost .hesis (Forty-thre--)

No. ofPatients

Excellent(90 units)

Satis-factory

(80 units)

Unsatis-factory

(70 lunuts)Failure

(Less)No. of

PatuentsExcellent(90 units)

Satis- Insatis-factory factory

(80 units) (To lunuts)Failure

(Less)

30 10 9 11 ii 1 7 1 2

5 5 7 1 65 5 22 3 18 13 :t 3 3

13 32 4 2713

Figs. 3-A, 3-B, atud 3-C: Ui’oup-I\ I 1st-ce-partfracture iuu which the greatet- ttiberositv is de-tached allowitig t.he subscaptularis to t-otate thehead inward, causitig the art ictular surf ace to facebackward.

Fig. 3-A: Pt-eopet’at ive atut etopost em It (emit -

genogram shosviuug the typical uotatorv displace-metit.

Fig. 3-B: Phsotogt’aph made itu the opet’atitsgloom showitig the biceps tetudotu ( forceps ) over-

lyitug the greater t uberosi lv fi-agnset it atud the culltear at the i’otatom itutervah, Ski ti hI oks lime Oil theedge of the stubscapularis teuudoui which is causitugthe articuilar sturface to face postel-i(It-l.

Fig. 3-C: Showing the atuat ((nut-al repait �f the

lesion showtu iiu Figs. 3-A auud 3-B LIV at)t)t’oxi-mat itug the t uuheu-osities ssit Is a wire loop mudrepair of the rotator-cuff defe(’t . The sttblmtxat iotu

disappeared as the mtuscles regai tied t I )t ue.

FIG. 4-B FIG. 4-C

Figs. 4-A, 4-B, auud 4-C: Group-V three-part fracture iii which the lesser Imuberositv is deta(’heclallowitug the exteruual rotators to rotate the head exteruially, causiuug the articuular sturface to face

anteri( Irly.Fig. 4-A: Preoperative auuteroposterior roent geuiogram showiuug rotatory dusplacemeuut.Fig. 4-B: Photograph made in the operatilug room to show the artictular surface preseiuting jiust

heuieat h the clavipectoral fa.scia. The skin hook is on the lesser tuherosity. This appearauu(’e hassuggest ed the term false fracture-dislocation.

Fig. 4-C: Excellent wire loop repair of the lesiots shown its Figs. 4-A mid 4-B l.y approximatingthe tuberosities 11usd shaft ausd repairiusg the rotator-cuff defect.

I)ISPLACED PROXIMAL HUM ERAL FRACTURES 1093

VOL. 52-A, NO. 6, SEPTEMBER 1970

culatiots, derived from the lesser tuberosity, was adequate to prevent FesO)1’1)tiOfl

of the head (Fig. 6-C). One of thse two failures of open reduction was due to a 1)ost-

operative w’ound infection in a patient who had been in traction for two weeks

prior to surgery and the other to loss of fixation in a patietit treated withs screw

fixation (Table Ill). Prosthetic reconstruction was performed in six l)sutients,

ti) of whsoni hiad gletsoid fractures and whose shoulders had beets dislocated for

over two months. Thse results i�’ere: four satisfactory, otse unsatisfactory, and one

failure (Table III). Thse failure was due to a large glenoid-rim fracture 1usd persistent

subluxation (Fig. S-B). Thse utssatisfactory result was causeol by a brokets wire that

permitted thie greater tuberosity to) redisplace (Fig. S-A).

Gioup VI, Posterior

In eighst posterior fracture-disloca-

tiouss the greater tuberosity remained

attached to the head atsd a severe rota-

tional detormity ��‘as present through the

surgical neck. Four of the patients had

been i tsj ured clut’itig electroshock therapy.

Closed reductioti was attempted in

six patients with success in only one

patient, whso obtained ats excellent result.

Iticomplete reduction was accepted in one

other patient ivhso had non-union at the

surgical neck (Table II). Open reduction

was performed its four patients. The

anterior deltopectoral approach was used

and the head w’as relocated by leverage

with a flat itsst.rumetst. The results were:

;�fl:;t’�’� J�.._,_�-1.,

FIG. 5

1096 C. S. NEER, II

THE JOi.RNAL OF BONE AND JOINT SURGERY

tw’o excellent, one satisfactory, and one failure (Table Ill). The failure was due

to non-utsion followitsg liush-tsail fixations (Fig. 6-A). A prosthesis was used iti two

lesiouss, otse of whsich hsad beets repeatedly manipulated and was several weeks old.

The other had beets utidiagnosed for fout’ months. Botls patients evetitually a-

chsieved satisfactory ratitigs (Table III).

Foni’-Part Fractures aial Fi’aet arc-Dislocations

Four-Pail Fractuie

Nineteets fractures witis displacement of both tuberosities were encountered. Its

each patietst the at-ticular segmetst was displaced laterally, out of contact with the

gletsoid, and wa.s detachsed from a source of blood supply.

Closed reductiots was attempted its twelve patietst.s, five of’ wisom later came to

surgery. Noise of the reductions were good because retraction of the tuberosity

persisted, atsd it was tiot possible to obtain good contact for tise articular fragment.Nevertheless, closed treatment was cotstinued its seven patients, resulting in four

Ph (I ogi’aphi made i mu the ((perat.itsg ioom (if a fottt’-paut.atutetiot’ fract uum’e-dislocatioiu, �. n up VI,showiiug the detacised head segnsetst which is a ‘‘shell fragmetut’’ withotut blood supply. Closedu-educt ion hsad failed aiim! at stut’gerv it was tuot possible to ((htaiui good approximat iotu (If the frag-metut with the tet macted I tuberosities.

nots-uniotss at the surgical neck, one complete resorption of the lsun4eral head, and

two disabhitig malutsions, each wit.ls sigtsificant humeral-head necrosis (Table II).

Five open reductiotss were perfo)rmed, and the results its all were rated failure. Each

effort to obtaiui coustact bet.weets isead atsd tuberosities so that thse articular frag-

metst mighst become revascularized failed (Fig. 7-B). The head was removed its one

patietst atsd resorbed its three. Otse patient had a postoperative woutsd itsfection

(Table III). A prosthesis was used with recotsstructiots of the tuberosity ttuid cuff in

seveti patietits (Fig. 11). The results were otse excehletst 1usd six satisfactory (Table

III).

Group VI, A ntei’ior

There were t hsirtv-ouse subcoracoid fracture-dislocations its w-hsich bot Is t ubet’os-

ities were detached from the articular segmeist (Fig. 2-C). The t.uberosities were

usually retracted, but its twelve Patients tisey were held in approximatiots by the

rotator cuff.

Fi� 6-C

Figs. 6-A, 6-B, and 6-C: I )epicl ing errors of fixatioti at open redmuct iou of Ihsree-part fract tures.Fig. 6-A: Axillary roeuutgeuuograns made two years after liush-usail fixatiouu of a thtee-part

posterior fract ure-dislocat ion, showiuig usonu-ttuuious.Fig. 6-B: Auuteroposterior roeuut.geuuogram of a u�oti-uusious a Groiup-I V I hree-fragnseuut lesuous

that had beets fixed with Kirschuier wires autd without repair of the rotator cull.Fig. 6-C: Auut eropost erior roeuitgenogram of a three-part. auit erior fract ture-dislocat iou niade

two years after opeut rediictiouu. Good approximation of the tuberosities loud a good (luff repair was

obtaitsed bttt the wire loop (lid uuot secture thse shaft., resultiusg its delayed unious.

DISPLACED PROXIMAL HUMERAL FRACTURES 1097

VOL. 52-A, NO. 6, SEPTEMBER 1970

Closed reductiots was attempted in twenty-three atsd accepted its four patietsts.

Whsen the tuberosities were iseld together by the itstact cuff, thsey seemed to prevetst

relocatioti of the hsead. When they were retracted, it was tsot possible to obtain

adequate boise cotstact for the hsead. Manipulation appeared to cotitribute to

tseurological symptoms its five pat.ietsts (Fig. 2-C). All four lesioiss treate(l by closed

methods resulted its failure, two due to resorptiots of the head atid two due to tson-

union (Table II). Open reduction was performed its five patients (Fig. 5). It was

difficult to maiustain cotst.act. between the fragments, susd the hsead was discarded

during two operatioiss. All five results are classified failures because of discar(led or

ava.scular heads (Table III, Fig. 7-A). Tw’enty-two patietits were treated by

prosthetic replacement. Tise results were three excelleist, eight eets satisfactory, and

one failure (Table III). Tlse failure was due to a postoperative woutsd itsfect ion.

Group VI, Posterior

In six patients thse head was dislocated posteriorly and is-as detached; both

tuberosities were retracted.

FIG. 7-A FIG. 7-B

Figs. 7-A and 7-B: Complicatiouss of open reduction its four-part lesions.Fig. 7-A: Atsteroposterior roent.geusogram made two years after open reduction of a four-part

fractuure-dislocation which failed because of resorptioms of the head.Fig. 7-B: Anteroposterior roent.genogram made ouse year after failure of Rush-nail fixation for a

four-part. fractutre showiuug avascular usecrosis, Isous-union of the head, and tuberosity retractions,iusdicative of a large rotator-cuff defect.

1098 C. S. NEER, II

THE JOURNAL OF BONE AND JOINT SURGERY

TABLE IV

Loe.st. Coimphmc.sTuoxs: ONE HUNDRED .SND SEVENTEEN FRACTURES ANI) FR.scTuaE-DtsLoc;tTIoxs

T hree-part (61) F our-part (56)

ClosedTteatment

(20)

OpenReduction

(30)Prosthesis

(11)

Closed

Tueatnsetit(11)

OpenReduction

(13)Prosthesis

(32)

Infections

Nous-uumsiomi

Avascula.m mue(’tosis

�\1alutiiots

Neum’ova.scular

Head discarded

Urssat isfactot-’s- Result

32

12

1

62

2 2

1

63

2

1

16

5

1

17 11 * 3 ii 13 1

* 1’e(’lsfli(’al eu’rou’s in seven.

Closed reductiots \V1t5 attempted witisout success in three patients (Table II).

Three lesiotss treated by open reduction resulted iii failure because of avascular

necrosis (Tstble III). The results w-ere rated satisfactory in three pat.ietst.s in whom a

h)rosthsesis was used (Table III).

Coin plicat bus

There were tio significant systemic complicatiouss. Five patients had sustained

their injuries duritsg electroshock therapy for psychoses, sevets were alcoholic, one

hladI epileptic seizures, two had multiple injuries, atsd seven had petsdiisg litigation.

Xotse of these problems proved an overwhelming deterrent to recovery. Local

complications are summarized in Table IV.

I nfection: Three postoperative wound infect.ioiss occurre(l. Two followed open

re(luctiolss that had beets (lelayed. One followed prostisetic replacement and was

promptly cotstrohled by removal of thse prosthesis and the administration of local

an(1 systemic antibiotics. All resulted in failure.

Fig S-C: Roemstgenogram nsade with the arm at the side fottr years after delayed surgery for afract.ure-dislocatious complicated by myositis ossificauis.

Fig. 8-I): Same patieust a.s shown iui Fig. 8-C with the arm overhead, demouustmatiuig that adequate

range had beeui attained four years after the complicatiots.

DISPLACED PROXIMAL HUMERAL FRACTURES 1099

VOL. 52-A. NO. 6, SEPTEMBER 1970

Figs. 8-A thsrough S-D: Complicat iotis of prosthetic replacemeuut, illtustrated by auutetoposteriorroeustgenograms.

Fig. 8-A: Result rated failure because a brokets wiu’e allowed tlse I (uberosity I( met uact omu theseveuuth day aft.er operat.ioui.

Fig. 8-B: Failure to recouust rtuct a fouur-nsoust h-old Isuslerior fiact mure-disl( (‘at l( itu with a large

gleuioid-rim fract tire.

Non-union: There i�’ere sixteets established non-unions at the surgical tseck of

whsich time followed closed treatment, six a hangitsg clLst (Fig. 1-C), and two over-

head t.ractiols. Distractiots appeared to be a causal factor. Six non-utsiotss followed

opens reduction of three-part fractures tisat were inadequately itsternally fixed (Figs.

1100 C. S. NEER, II

THE JOURNAL OF BONE AND JOINT SUtIGEtOY

. Ascending branch� anterior circumflex

Ful. �)

Illuuslmatitug hue tsseth(((I ((f explou’iuug a(’(tte fu’act.uuu’es. The defect iuu the totatou’ iuuteuval is (Ic-vel( (I)Cd Pt( )XitiulullV, av( imliuug i i uj uu’v I o I lie blo d su�)plv If t he 1ie�sd svhicls emuI (‘is b (t u(’ i uu I lie

vit’iuuitv ((f 1Ise liicipital gmo�(ve �uud tubeu’osities.

Fig. I I ): ‘Fhie usset Isoti f uepaiti i ug a I luuee-patt lesi (t u aft eu ( pet u ted tu(’t i( Ii u- ‘l’lie I mil iet� si I l(’s ute

appu-oximated �suud secuuued to the shaft with stuouig buuied wiue sutmutes, Two wit-c loops ate tuow

puefeu’ued. flue rotator-t-uuff (Iefe(’t is repan.Fig. 11: TIse method (f m’epaim-itug a foui’-paut fmact mute with a puost lsesis.All I mihetosit v fuagnseuuts

ate appu’oxiusated uuudeu t Ise head of t he pu’ost hesis wit ls two hturied wiu’e loops TIst’ defect iii I hueu((tat((t (‘tiffis suOti’ed It is tt((t aiwa s iue(’essau’y 1(1 divide the louug head of the biceps

6-A and 6-B). Nouu-uiiioii of the head segment was pt’so’tit iii most patietits witls

avascular tiect’osis (F’ig. 7-B).

A u’a.sculai’ necrosis’: Cohlapse of the at’t icular suu’face occurred its only two of

t I uirt �‘ t hsree-part lesions treated by opeti reductiots (Table IV). (.)f t I sirt eeti pat.ietits

witls four-part lesions treatetl by open reduction, however, the head was (lisCau’(led

iii five 1111(1 resO)rbe(.l in six (Fig. 7-A).

_llalunion: Rot ato)ry displacemetst in thsree-part fractures could not be cots-

tt’olled by closed met Isods (Figs. 1-A, 1-B, 2-A, mtuid 2-B). The result usually itscluded

DISPLACED PROXIMAL HUMERAL FRACTURES 1101

significant discomfort. and loss of external rotation and abduction. One patient with

fixed retractiots of the greater tuberosit.y and malunion of the glenoid (Fig. S-B) and

another with loss of fixation of the tuberosities (Fig. 8-A) accounted for two of the

failures of prosthetic replacement (Table IV).

Neurom’ascular: Tets patients were first seen with neurological deficits, involvitsg

the axillarv 1Usd median iserves in four, axilla.ry and radial nerves in two, ulisar userve

ins two, and all three in two. Repeated efforts at closed reductiots was thought to have

ilsCreased t.hsese symptoms in several patients (Fig. 2-C). Deficits persisted its four

patients, one of whom received an unsatisfactory rating (Table IV).

Transitory subluxation: This w’as common after all types of treatmeust (Fig.

3-C). It disappeared as tise muscles regained tone.

Myositis ossificans: Pericapsular bone formation occurred after all three

methods of treatment : three after closed reduction, five after opets reduction, atsd

six after prosthetic replacement. (Figs. S-C and S-D). Predispositsg factors seemed to

be soft-tissue injury as in fracture-dislocations, repeated manipulatiots, and delayed

reduction beyond sevets days.

Failure of metal: There was no instance of dislocation or loosening of the stem

of a prosthesis. The ��‘ire loops fragmented in one patieist, between two and tsine

years after surgery, but caused no discomfort..

Over-all Results

The results of closed reduction, opets reduction, and prosthetic replacenienst are

summarized in Tables II and III. A minimum of one year was required for reason-

able recovery aisd, in general, the results improved with time.

Closed treatment appeared inadequate for active patietsts in eithser group. It

produced satisfactory ratings in only three of thirty-nilse patients with three-part

lesions arid consistently failed in thirty-eight patients with four-part displacements.

Open reduction was performed in thirty patients with three-part displacements

and appeared effective. There were eleven failures but thsese w’ere largely due to

technical errors (Figs. 6-A and 6-B). Suture fixation of the tuberosities atsd cuff

repair yielded satisfactory or excellent results in 86 per cent. (Fig. 10). Severe

necrosis w’iths resorption of the head rarely occurred. Patients with four-part lesions,

however, had uniformly poor results because the head had beets discarded or

resorbed.

Prosthetic replacement was used in thirty-tw-o patieists w’ith four-part. displace-

meists (Fig. 11). The typical result was satisfactory but imperfect. atsd t.he recovery

period was prolonged. The average numerical rating was 82 units. There w’ere few

complicatiotss. These results in patients with four-part lesiotis were quite superior

to those of either closed or open reduction.

Operative Technique

During the course of this study, some techniques were found to be of value and

some problems and errors in technique became apparent.

1. Anteroposterior and lateral roentgenograms of the upper end of the humerus

are essential preoperat.ively in planning the procedure and avoiditig unnecessary

tissue damage.

2. Reconstruction is difficult and delay beyond two w’eeks renders it. more

difficult, because of fixed retraction of the tuberosities, formation of cicatrix,

deposition of pericapsular bone, and softening of the fragmeists.

3. The preoperative pain in the sisoulder reduces the efficacy of skits prepara-

t.ion, and the injured tissue invites infection. It is important that the skins be cleansed

scrupulously under anesthesia and that. adherent plastic drapes be used.

VOL. 52-A. NO 6, SEPTEMBER 1970

1102 c. S. NEER, II

4. The deltopectoral approach with detachment of tise ansterior 7.6 centimeters

of the (leltOid from clavicle is preferred. Acromiotsectomy or inadequate reattach-

meust of tise deltoid is especially disabling.

5. After divisiots of the clavipectoral fascia, the w’outsd should be irrigated free

of clots �U5d the teisdon of the long head of the biceps shsould be used aS a guide to

the interval between the greater and lesser tuberosities (Fig. 3-B).

(1. Itsjury to the blood supply of the head and the anterolateral artery � cats be

avoi(le(l by developiisg the itsterval defect in the rotator cuff (Fig. 9). Its doubtful

cases the itsterval shsould be explored before a final decision to discard the head is

made.

7. At open reduction, fixation by two strong buried wire loops is much more

efficietst than by screws, nails, or Kirschner wires. After the tuberosities are secured

together atsd to the shaft., the rotator cuff is repaired (Fig. 10).

S. if a prosthesis is us-ed (Fig. 11), a tight fit of the s-tern within the medullary

canal is essential. Four stem sizes are curretitly available, each with appropriate drill

specifications. The articular surface must be positioned so as to face in 30 degrees ofretroversion to provide stability against dislocation. The fragrnetsts of thse t.uberosi-

ties are approximated betseaths the prosthesis w’iths two wire loops and nhe rotator

cuff is tisen repaired.

9. Adhesions cats be mitsimized by the use of assisted extertial rotat.iots exercises.

These are started at about four days after surgery atsd progress as rapidly as the

repair j)ernsits, working itsitially for range and much later for strength.

10. Patients sisould be warned preoperatively of thse protracted course of re-

isabilitlttiots, and later they shsould be etscouraged by their surgeots to have con-

fidetice Itildi take pride in tiseir accomplishments.

Discussion

Diuritsg the period required to collect. this series, more than 1300 proximal humer-

111 fractures were seen its our clinic. This figure does not, however, accurately portray

tlse prevalence of the severe multifragment itsjuries ansotsg injuries to the proximal

end of the Isumerus sitsce matsy in the series w’ere referred from other hospitals. The

majority of patients who sustaitsed these disablitsg fractures were healthy in-

dividluals its their active �‘ears. Therefore, although these injuries are uticommots,

the�’ pose su significatst problem.

Tlse classification used may seem cumbersome, but. it. seems muds more repre-

sentative of the actual type of lesions encountered than thse oversimplified classifica-

tiotss of the past. For thse purpose of evaluating therapy, at least, our classification

hsa.s l)rOved useful. For example, if thse over-all results after opens reduction are com-

1)ared with those of Ktsight. and Mayne, there were 44 per cent. satisfactory result�sin this series of forty-tlsree procedures as compared withs 46 per cent satisfactory

results its twenty-six procedures ins the series of Knight and Mayne. Yet ��‘hen t.hse

techtsically good opens reductiotss in our three-part fractures are considered, tise

results were excellent or satisfactory in 86 per cent. It. is not possible to separate

thsis group its Knight. 1usd Maytse’s series. Thus, it is. possible isow to recognize the

value of open reductions ins the various types of fracture and to detect. its deficietscies

in the four-part. fracture.

Summary

A stu(ly of 117 thsree-part tusd four-part displaced proximal humeral fractures,

followed for from otse to sixteen years, is presented. The ages of the patients averaged

55.3 years. Treatment began w’iths closed reduction in sevetsty-seven patients, the re-

sult of whsich were accepted ins thirty-one. Open reduction was doise in forty-thsree

THE JOURNAL OF BONE AND JOINT SURGERY

DISPLACED PROXIMAL HUMERAL FRACTURES 1103

patietst.s, and prosthetic replacement in forty-three patient��. Their results were

rated by a numerical system. Closed reduction was found inadequate for active,

healthy patients in either group. This was because of uncontrollable rotatory (us-

placement in three-part fractures and avascular necrosis of the detached lie-sd in

four-part fractures. \Iost of the poor results of open reduction ins three-part (lis-

placements w’ere due to errors in reduction or fixation while those in four-part

displacements w’ere due to avascular necrosis of the head. It. w’as concluded that thse

preferable method for three-part fractures was open reduction tusd that for four-part

fractures w’as prosthetic replacement. Using these indications, the typical result w’as

satisfactory but imperfect atid many months were required for maximum recovery.

Surgical errors and technique are discussed.

References

1. KNIGHT, II. A., ausd MAYNE, J. A.: Commiusuted Fractures amsd Fracture 1)islocatiouss Ituvolvinigthe Artictular Surface of the Humeral Head. J. Boise ausd Joint. Surg., 39-A: 1343-1355, l)ec.1957.

2. LAtNG, P. G.: The Arterial Supply of the Adult humerus. J. Bouse and Joint Surg., 38-A: 1105-1116, Oct. 1956.

3. NEEuS, C. S., II: Displaced Proximal hlumeral Fractures. Part I. Classificatiouu and Evalualious.J. Boise ausd Joiust Surg., 52-A: 1077-1089, Sept. 1970.

VOL 52-A, NO. 6, SEPTEMBER 1970