Cosmetic Surgical Anatomy of the Midcheek and Malar Mounds · the level of the zygomatico-facial nerve, showing the prezy-gomatic space. The preperiosteal fat is covered by an adher-ent

Cosmetic

Surgical Anatomy of the Midcheek and MalarMoundsBryan C. Mendelson, F.R.C.S.Ed., F.R.A.C.S., F.A.C.S., Arshad R. Muzaffar, M.D., andWilliam P. Adams, Jr., M.D.Melbourne, Australia, and Dallas, Texas

The anatomy of the midcheek has not been satisfacto-rily described to adequately explain midcheek aging andmalar mounds, nor has it suggested a logical approach totheir correction or provided sufficient detail for safe sur-gery in this area. This cadaver study, which was comple-mented by many operative dissections, located a missinglink: a glide plane space overlying the body of the zygoma.The space functions to allow mobility of the orbicularisoculi, where it overlies the zygoma and the origins of theelevator muscles to the upper lip. The space is a cleftbetween the sub–orbicularis oculi fat and the preperios-teal fat and is lined by a fine membrane. The anatomicboundaries are clearly defined by retaining ligaments,which correlate with the triangularity of the space. Severalanatomic features provide the functional characteristics ofthe prezygomatic space, including the (1) absence of di-rect attachments between the orbicularis in the roof to thefloor, (2) more rigid inferior boundary formed by thezygomatic ligaments, and (3) more mobile upper liga-mentous boundary formed by the orbicularis retainingligament (separating from the preseptal space of the lowerlid). These components determine the characteristic ag-ing changes that occur in this region and explain muchabout malar mounds. An appreciation of this anatomy hasseveral surgical implications. The prezygomatic space is ajunction area that can be approached from the temple,lower lid, and cheek. The zygomatic branches of the facialnerve to the orbicularis do not cross the space; rather, theycourse in the walls and in the sub–orbicularis fat withinthe roof of the space. (Plast. Reconstr. Surg. 110: 885,2002.)

Malar mounds have acquired some notoriety fortheir persistence in the face of surgical efforts toremove them.

—D. Furnas1

Correction of the midcheek is currently rec-ognized to be of fundamental importance infacial rejuvenation surgery. To date, the de-scriptions of the intrinsic anatomy of this re-gion have been unable to explain its aging

changes and insufficiently detailed to providefor predictably safe, and possibly even logical,surgical correction of this area. This is a keyreason that the subperiosteal plane has beenthe preferred level of dissection in surgicalrejuvenation of the midcheek.2–6

In common usage, the term midcheek refersto that part of the cheek medial to a line ex-tending from the frontal process of the zygomato the oral commissure and from the lower lidabove to the nasolabial fold below. Althoughthe midcheek may appear externally as a uni-form structure, in fact, it is composed of twofunctionally distinct parts, the prezygomaticpart that overlies the skeleton of the midcheek(body of the zygoma and maxilla) and theinfrazygomatic part below that covers the ves-tibule of the oral cavity. The skeleton underly-ing the midcheek is the major determinant ofthe shape of the midcheek, because it bothseparates and connects the orbital and oralcavities. The sphincter muscles that enclosethese bony cavities (orbicularis oculi and orbic-ularis oris) share the bony platform of themidcheek for their limited skeletal attach-ments, as do the retaining ligaments associatedwith each muscle.

The aging changes that appear in the mid-cheek largely reflect the effect of laxity andptosis of the soft tissue mass of the cheek rela-tive to the underlying skeleton. Because thesoft tissue mass is thinner above, the ptosisappears as a volume displacement. This affectsthe upper prezygomatic part by uncovering theanatomy of the orbit, such that the lower lid fat

From the Department of Plastic Surgery, University of Texas Southwestern Medical Center, and from Melbourne, Australia. Received forpublication July 27, 2001; revised November 1, 2001.

Presented at the Scientific Meeting of the American Society for Aesthetic Plastic Surgery, New York, May 5, 2001.

DOI: 10.1097/01.PRS.0000019706.34235.E5

885

bulges become revealed (whether they are nor-mal or excessive in amount) and the bonyorbital rim becomes uncovered. The junctionbetween the lid and cheek, which formerly wasindistinct, becomes quite defined with the ap-pearance of a palpebromalar groove inferolat-erally and a nasojugal groove medially. Theseconfluent grooves become increasingly promi-nent with further descent of the lid-cheek junc-tion. The displaced volume accumulates in theinfrazygomatic part of the cheek, most evidentas increasing fullness of the nasolabial fold.

Changes in the shape of the cheek accom-pany the tissue descent. Whereas youthfulcheeks characteristically have a rounded full-ness maximal over the prominence of the zy-goma, the volume changes associated with ag-ing progressively flatten the fullness andeventually form an infrazygomatic concavity.This is the midcheek furrow, which is a contin-uation of the nasojugal groove downward andoutward. The furrow bisects the soft tissuesoverlying the prezygomatic part of the mid-cheek at an angle so that the lateral segmentoverlying the zygoma becomes triangular, be-ing bounded above by the palpebromalargroove. When visibly enlarged this part formsthe clinical entity described as malar mounds,1also termed malar bags and malar crescent.7This discrete entity should be distinguishedfrom the malar fat pad,8 a term used to de-scribe the subcutaneous fat that gathers medialto the midcheek furrow and exaggerates thenasolabial fold. The anatomic basis for malarmounds has not been satisfactorily defined,which may account for their unpredictable re-sponse to surgical correction.

When surgery involving the superficial mus-culoaponeurotic system (SMAS) was intro-duced, it was not considered safe to extend theSMAS dissection into the area of the mid-cheek,9,10 notwithstanding the earlier excur-sion of Skoog11 into the infrazygomatic part.Although it provided a significant benefit tothe lateral cheek and jawline, this limitedSMAS surgery had minimal impact on the mid-cheek. Surgery of the midcheek developed sub-sequent to the descriptions of the retainingligaments of the cheek,12–14 when the focus inface lift surgery had extended to correction ofthe nasolabial fold. Release of the masseteric-cutaneous and the lateral zygomatico-cutane-ous ligaments enabled the SMAS to correct theinfrazygomatic part of the midcheek. However,when only that portion is corrected and the

prezygomatic part is not similarly corrected, adisharmony arises. Currently, there are severaldifferent approaches to the prezygomatic re-gion in clinical practice.

The prezygomatic SMAS can be bypassed bychanging the plane of dissection over the mid-cheek. The new plane can be on the outersurface of the orbicularis oculi with elevationof the subcutaneous fat (the malar fat pad),15

which is then repositioned independently ofthe underlying SMAS.8,16,17 The subperiostealplane is also used to bypass the SMAS of themidcheek. When the “deep plane” of dissec-tion18 is used for correction of the lateral andinfrazygomatic SMAS, it is logical to extend theplane so that the resultant SMAS flap now in-cludes the prezygomatic SMAS (which here islargely the orbicularis oculi and its fasciae).Previously, concerns about the risks of facialnerve injury had discouraged this approachuntil it was recognized that a “safe interval”exists between the temporal and zygomatic fa-cial nerve branches overlying the zygoma.12

This area has subsequently been described as asurgical “safe space.”19 Although the bound-aries of the space have not been specificallydefined, it is used in several face lift tech-niques.20–22 The safe-space approach to theprezygomatic SMAS can also be by enteringdirectly through the lower lid and temple in-dependent of the lateral cheek.23–26

MATERIALS AND METHODS

This is the second segment of a two-partanatomic study in this issue using the samecadaver material and methods described in thepreceding article.27

In addition, the detailed observations fromthe precise surgical dissections of hundreds offace lift, temporal lift, and midcheek proce-dures (performed by the senior author(B.C.M.) have provided supplementaryinformation.

RESULTS

The part of the orbicularis oculi that contin-ued into the cheek (pars orbitalis) was exten-sive and situated immediately deep to the sub-cutaneous fat. The thickness of the fat layernotably increased as the dissection proceededinferiorly from the lid margin. The muscle fi-bers were seen as loops that extended a vari-able distance down the cheek. Rather thanbeing semicircular and concentric to the lidmargin, as expected from textbook illustra-

886 PLASTIC AND RECONSTRUCTIVE SURGERY, September 1, 2002

tions, the lower fibers took a more ellipticalcourse and extended along the summit of thenasolabial fold, where for a distance, the mus-cle fibers were close and parallel to the naso-labial crease. Inferior to the zygoma the orbic-ularis was less clearly defined because the moreinferior loops of muscle were thinner andsomewhat dispersed without the muscle layerhaving a well-defined lower edge. Here themuscle layer was so thin as to be almost lost inthe thickness of the surrounding fat (i.e., bothsuperficial and deep to the muscle layer). Thefascia associated with the orbicularis musclelayer is less defined on the subcutaneous sur-face than on the deep surface and is grosslyminimal in the cheek.

In the prezygomatic region the fat deep tothe orbicularis oculi is arranged in two well-defined strata (Fig. 1), separated by a naturalcleavage plane. A thin, transparent membraneadheres tightly to the outer surface of thedeeper layer of fat. The seemingly complexanatomy of this area can be more simply de-

scribed in terms of an actual prezygomaticspace having a floor and a roof connected bywalls or borders above (cephalad) and below(caudal).

The origins of the three elevator muscles ofthe upper lip (i.e., zygomaticus major, zygo-maticus minor, and levator labii superioris)tend to follow the curvature of the lower bor-der of the underlying bone, which narrowsmedially toward the orbital rim. Accordingly,the line of origin is in the form of a crescent,concave above with the medial end higher thanthe lateral (Figs. 2 and 3) The medial end ofthe origin of the levator labii superioris ex-tends virtually to the orbital rim. From therethe broad line of the levator origin inclines atan angle of nearly 45 degrees away from theorbit. The narrow origin of the zygomaticusminor (4.0 to 5.0 mm wide) is the lowest pointof the crescent from the orbital rim. The originof the more significant zygomaticus major (8 to10 mm wide) is located higher on the zygoma(at least halfway up). Between this origin andthe zygomatico-facial foramen (or foramina),which is considerably closer to the orbital rim

FIG. 1. Concept of sectional anatomy of the midcheek, atthe level of the zygomatico-facial nerve, showing the prezy-gomatic space. The preperiosteal fat is covered by an adher-ent membrane (blue); both fat and membrane cover theorigins and the proximal part of the lip elevator muscles(shown here on the zygomaticus minor). The roof of theprezygomatic space is formed (from outside in) by skin andsubcutaneous fat (both removed), the orbicularis oculi andorbicularis muscle fascia with sub–orbicularis oculi fat ad-herent to the underside of the fascia. The membrane reflectsoff the floor to line the upper and lower boundaries. Theorbicularis retaining ligament (above) separates the prezygo-matic space from the preseptal space of the lower lid. Thezygomatico-cutaneous ligaments (below) separate the spacefrom the soft, yellow fat external to the buccinator and themucosa lining the vestibule of the oral cavity. The main motorbranch to the orbicularis courses within this fat immediatelyoutside the prezygomatic space.

FIG. 2. Extent of the prezygomatic space. The space over-lies the body of the zygoma and the origins of the lip elevatormuscles. It extends to the posterior border of the body of thezygoma and so can be accessed from the lower temporalregion and from the lower lid. The space is bounded by theretaining ligaments of the lower lid and cheek. The upperborder is formed by the orbicularis retaining ligament, whichbecomes confluent at the inferolateral orbital rim with thebroadly based lateral orbital thickening that overlies the fron-tal process of the zygoma. The zygomatic-cutaneous neuro-vascular pedicle is the only structure crossing the space.

Vol. 110, No. 3 / SURGICAL ANATOMY OF MIDCHEEK 887

(5 to 10 mm), there is a significant space of“exposed” bone 15 mm or more in width.

Adhering strongly to this area of bone is thedeeper of two layers of fat. This preperiostealfat forms a carpet-like padding approximately21⁄2 to 3 mm thick and is distinguished by itslocation and its appearance; it has a pale colorand coarse lobulation. The preperiosteal fatlayer not only covers the exposed bone, but italso extends inferiorly between the musclesand covers the origins and bellies of the mus-cles for some distance, becoming progressivelythinner as it spreads onto the muscles. Thisarrangement has the effect of concealing theactual muscle origins so that the muscles ap-

pear to have emerged from a carpet of prepe-riosteal fat. The thin transparent membranethat adheres to the preperiosteal fat also ex-tends inferiorly over the muscles. As a result ofthis arrangement, the floor of the prezygo-matic space is broader than expected; it notonly overlies the bone but it also extends overthe proximal ends of the muscles. Because ofthe concealment by the preperiosteal fat, theextent of the muscles under the floor is morethan may be expected. When blunt dissectionis used in this area, as with a sweeping motionof a blunt instrument, the smooth outer sur-face of the membrane stays intact and the un-derlying preperiosteal fat remains adherent tothe bone. The membrane extends beyond thefloor of the space, reflecting outward to linethe walls.

The inferior wall of the prezygomatic spaceis lined by the continuation of this membraneas it leaves the surface of the muscles inferiorto their origins and then sweeps more superfi-cially into the soft tissues of the cheek (Fig. 4).The most cephalad of the zygomatico-cutane-ous retaining ligaments that take origin be-tween the muscles are located directly inferior,and provide reinforcement, to the membraneso that there is a strong inferior boundary ofthe space. For the dissection to extend inferi-orly from within the prezygomatic space intothe sub-SMAS space of the lower cheek, it isnecessary to transect this ligamentous barrier.Whereas these ligaments have the prezygo-matic membrane like a coating on their uppersurface, the other surfaces are surrounded bysoft, loose, yellow fat that is quite distinct fromthe firm, white preperiosteal fat. The zygo-matico-cutaneous retaining ligaments extendoutward through the subcutaneous layer to thedermis, whereas the muscles remain in theplane of their origins as they extend down tothe lip. The histologic examination under-taken to show this arrangement was not wellvisualized. However, the histologic results froma related study (Fig. 5) demonstrated a signif-icant difference in the architecture of the col-lagenous fibrous tissue superficial and deep tothe orbicularis strata. In contrast to the per-pendicular orientation of the retinacula cutisfibers in the subcutaneous layer, immediatelydeep to the orbicularis fascia a horizontal ori-entation suggests a cleftlike structure withoutdirect attachment to the periosteum. A solitaryor specific subcutaneous extension of the zygo-

FIG. 3. Dissection of the floor and upper border of theprezygomatic space. The thick carpet of preperiosteal fat andthe overlying membrane of the floor of the space also concealthe origins of the lip elevator muscles zygomaticus major(Zmaj.), zygomaticus minor (Zmin.), and levator labii supe-rioris (not shown). The upper blue spacer indicates how thetemporal approach to the prezygomatic space passes deep tothe lower temporal branches of the facial nerve (TFN) to theorbicularis. The zygomatico-facial nerve (ZFN) is the onlystructure crossing the space. For orientation purposes, thezygomatico-facial nerve is directly cephalad to the zygomat-icus minor. The space narrows medially to a V and is broaderlaterally above the zygomaticus major. LOT, lateral orbitalthickening; ORL, orbicularis retaining ligament; OO, orbic-ularis oculi in the roof of the prezygomatic space lined bysub–orbicularis oculi fat (SOOF ).


matic ligaments or the membrane was notseen.

The orbital boundary of the prezygomaticspace is formed where the membrane sweepsonto the underside of the septum-like orbicu-laris retaining ligament (described in detail inthe preceding article27). This ligament sepa-rates the prezygomatic space from the presep-tal space of the lower lid and courses immedi-ately on the orbital side of the zygomatico-facial foramen and nerve (Figs. 3 and 6). Theligament becomes significantly more substan-

tial and less distensible as it ascends the lateralorbital rim to become expanded out as itmerges into the lateral orbital thickening.

The medial end of the prezygomatic spacebecomes triangular before being closed off bythe approximation of the medial end of thelevator muscle origin to the orbicularis originon the inferomedial orbital rim. The spaceextends lateral to the zygomatico-facial nerveand is closed at the root of the frontal processof the zygoma, well below the level of the lat-eral canthus, by the lateral orbital thickening(also described in the preceding article27),which extends across the width of the frontalprocess of the zygoma onto the deep temporalfascia. The posterior border of the prezygo-matic space has not been as clearly delineatedin either its position or structure. It overlies thebody of the zygoma at or near its junction withthe arch of the zygoma. It seems that reinforce-ment of the posterior boundary is provided by

FIG. 5. Histologic view of the cheek overlying the zygomaand proximal zygomaticus major. Retinacula cutis fibers (RC)extend from the fascia enclosing the orbicularis muscle (OO)to the dermis but do not extend deep to the orbicularis. Notethe absence of a direct attachment of the orbicularis (the roofof the space) because of the prezygomatic glide plane (preZ)deep to the orbicularis, which here extends to the proximalpart of the zygomaticus major (ZM) (Fialkov, J., and Men-delson, B., unpublished study).

FIG. 4. The roof of the prezygomatic space (orbicularisoculi [OO] lined by sub–orbicularis oculi fat [SOOF]) hasbeen opened by a horizontal incision; then the upper andlower flaps are grasped by forceps and reflected to show thefloor and the upper and lower boundaries. The floor of theprezygomatic space is defined between the two continuous bluelines, which indicate where the membrane (PreZyg. Memb.) isreflected off the floor onto the upper border, formed by theorbicularis retaining ligament (ORL), and onto the lowerborder, where it directly overlies the upper surface of thezygomatic ligaments (ZL), nearest the space. The fenestra-tions through the membrane between these ligaments aredissection artifacts. The instrument is located immediately out-side the space in the soft, yellow fat on the underside of thesame retaining ligaments that form the inferior border. Thezygomatic ligaments farther outside the space are not linedby the membrane. The motor nerve to the orbicularis coursesoutside the space in the interval between the ligaments, asindicated by the instrument.


the most anterior of the smaller zygomatic lig-aments attached to the arch of the zygoma.

The roof of the prezygomatic space isformed by the orbicularis oculi (pars orbitalis),the underside of which is covered by the orbic-ularis muscle fascia, deep to which is a thinlayer of fat quite distinct from the preperiostealfat. This is the previously described sub–orbicularis oculi fat.28

The sub–orbicularis oculi fat is finely lobu-lated and distinctly yellow, in contrast to thepreperiosteal fat, and is less strongly adherentto the orbicularis than is the preperiosteal fatand membrane to the periosteum of the zy-goma. Coursing within the thin layer of sub–orbicularis oculi fat are multiple radiatingbranches of the zygomatico-facial nerve andvessels from above, and the motor branches tothe orbicularis oculi. No structures cross theprezygomatic space, although the zygomatico-facial nerve and vessels are more at the upperboundary than within the space.

Several significant observations were maderegarding the innervation of the orbicularis

oculi. The zygomatic branches of the facialnerve that innervate the orbicularis oculi enterthe pars orbitalis at or near its periphery andseem to do so in four distinct locations (Fig. 7).They originate off the main zygomatic branch,which remains at the “deep level” within thesoft, yellow fat at the base of the zygomaticretaining ligaments. Ultimately, the severalbranches become more superficial as they en-ter the sub–orbicularis oculi fat on the under-side of the orbicularis.

The nerve, previously considered the “main”innervation of the orbicularis oculi,29 separatesfrom the zygomatic nerve well before the latterpasses deep to the zygomaticus major. Thenerve continues alongside the parent branchfor a short distance before it diverges to be-come slightly more superficial and continue itsdirectly horizontal course over the outer sur-face of the zygomaticus major, within a centi-meter of the upper edge of the muscle origin.This orbicularis branch courses immediatelyoutside the prezygomatic space as it passes onthe inferior surface of the same zygomatic lig-ament forming part of the inferior boundary ofthe prezygomatic space, immediately lateral tothe zygomaticus major. This orbicularis branchcontinues medially onto the surface of the zy-gomaticus minor (in some cases passingthrough the outermost fibers) to the medialborder of the zygomaticus minor, where itabruptly changes course, from transverse to

FIG. 6. Dissection through the roof overlying the lateralextent of the prezygomatic space. The superficial fascia of theroof consists of the orbicularis oculi (OO), which is contin-uous with the temporoparietal fascia (TPF) over the temple.The orbicularis is lined by sub–orbicularis oculi fat (SOOF).The continuous blue line indicates the outward reflection of thesmooth membrane on the preperiosteal fat onto the under-surface of the orbicularis retaining ligament (ORL). Thisligament is confluent with the lateral orbital thickening(LOT) near the lateral canthus. The zygomatico-facial nerve(ZFN) is just inside the upper boundary.

FIG. 7. The pattern of innervation of the orbicularis oculiof the lower lid. The branches at 6 o’clock and 7 o’clockcourse from deep to the parotid–masseteric fascia to becomesuperficial into the sub–orbicularis oculi fat in immediateproximity to the retaining ligaments, where indicated by theyellow arrows. Red arrows indicate the three common surgicalapproaches to the prezygomatic space.


vertical, to enter the sub–orbicularis oculi fatoverlying the zygomaticus minor. It remains inthis layer, in the ceiling of the space, as itascends toward the lid margin.

An adjacent motor branch to the orbicularisis located in relation to the inferolateral cornerof the prezygomatic space. This branch sepa-rates from the main zygomatic nerve moreproximally, either within or just outside theparotid, and initially courses directly vertical tocross the lower border of the arch of the zy-goma before coursing transversely in the depthof the fat immediately cephalad to the lowerborder of the zygomatic arch. This brings itimmediately caudal to the most anterior of thesmaller zygomatic arch ligaments, only 10 mmor so lateral to the main zygomatic ligament(Fig. 7). If the previously described mainbranch to the orbicularis is considered to be at6 o’clock to the prezygomatic space, thisbranch is at 7 o’clock. Both branches to theorbicularis have similarities regarding their re-lationship to the prezygomatic space, being inintimate relationship to the ligament immedi-ately cephalad and separated from the prezy-gomatic space by ligament. Another similarityis their change of course immediately aftercrossing the base of the ligament, where theypass outward into the inferior-lateral edge ofthe orbicularis.

Just below the level of the lateral canthus,several transversely oriented (lower temporalor upper zygomatic) branches continue theirpassage through the temporoparietal fasciainto the orbicularis at about the level of theposterior border of the frontal process of thezygoma (i.e., 15 to 17 mm lateral to the lateralcanthus). A medial branch to the orbicularis islocated at the medial extent of the prezygo-matic space near the medial side of the levatorlabii superioris. The origin of this small branchwas not satisfactorily defined, but it seems toarise from the main zygomatic branch, whichhas been described as entering the levator labiisuperioris alaeque nasi,30 the angular nerve.

DISCUSSION

The three-dimensional arrangement of thesoft tissues in the prezygomatic part of the cheekis unique in that it accommodates the merging ofthe dynamic and mobile tissues of the lower eye-lid with those of the infrazygomatic cheek (peri-oral). Accordingly, the prezygomatic space canbe considered to be a transition zone. The skel-etal attachments of the lower lid (orbicularis oc-

uli and orbicularis retaining ligament) have arestricted attachment to the upper border of thebone, whereas the continuation of the orbicu-laris in the cheek (pars orbitalis) remains super-ficial as it extends well down in the roof of theprezygomatic space and covers the proximal partof the muscles to the upper lip.

The elevator muscles of the upper lip, beingmore powerful, not unexpectedly have a moreextensive attachment on the zygoma and max-illa. This attachment is not located near thelower border (which it would be if it mirroredthe lid structure); rather, it extends more thanhalfway up the bony surface. The periostealorigins of the zygomatic ligaments are in closeproximity to the origins of these muscles. Fromthis common origin, the ligaments course in adifferent direction to the muscles. Whereas themuscles remain “deep,” on the skeletal planethe ligaments radiate outward to the dermis ofthe cheek, and in so doing, the ligaments de-fine the lower border of the prezygomatic partof the cheek, separating it from the infrazygo-matic part.

The orbicularis retaining ligament is the an-atomic structure responsible for defining thepalpebro-malar groove (the visible junction be-tween the preseptal part of the lower lid andthe cheek). Downward displacement of the lid-cheek junction is associated with a lengtheningand attenuation of the preseptal part of the lidas it extends into the area formerly occupied bythe upper cheek. For bulging lower lid fat toextend beyond the orbital rim, the retainingligament must also undergo distension alongwith the preseptal segment of the lid. Accord-ingly, the lid structures encroach into theprezygomatic space.

The prezygomatic space is an entity, eventhough it may require dissection to be re-vealed, with characteristics similar to those ofthe other well-known spaces within the subsu-perficial fascia areola cleavage plane of the face(e.g., subgaleal, temporal, and sub-SMAS ofthe outer cheek). That is, there is a spatialseparation between the superficial and deepfacial fascias, the boundaries of the space aredefined by retaining ligaments (from the peri-osteum to the superficial fascia), and facialnerve branches do not course through butrather around the space within the boundaries,such as within a ligamentous wall. Similar tothe other spaces, the prezygomatic space seemstight in younger patients and becomes moreapparent and easier to surgically expand in


older patients. Presumably, this is because ofthe development of laxity of the fibrous con-nective tissue components of the walls andwithin the superficial fascia forming the roof.

The prezygomatic space provides for inde-pendent mobility of the orbicularis oculi in theroof of the space from the lip elevator musclesunderlying the floor of the space and viceversa. The considerable movement of the skinof the lower lid and upper cheek that occurson contraction of the orbicularis must be asso-ciated with an equivalent degree of displace-ment of the underside of the muscle. For theprezygomatic space to function to allow a glid-ing movement of the overlying orbicularis, theroof itself must be mobile and not directlyattached to the underlying deep fascia.

Accordingly, it is only the peripheral attach-ments at the boundaries that stabilize the roof.The septum-like orbicularis retaining ligament ofthe upper boundary is considerably less strongthan the zygomatic ligaments of the lowerboundary. The limited rigidity of the ligament ofthe upper boundary allows for the significantmovement that occurs across this boundary onorbicularis contraction, compared with the “fixa-tion” of the lower boundary. In some ways, theprezygomatic space has similarities with the glideplane space that exists above the superior orbitalrim deep to the lower part of the frontalis andorbicularis oculi.31

The shape of the malar mounds, triangularwith the apex medially, mirrors that of theunderlying prezygomatic space, being definedby the same ligamentous boundaries. Laxity oftissue tone within the roof of the prezygomaticspace seems to be the key to the presence anddegree of malar mounds (Fig. 8) inasmuch assquinting causes the mounds to efface whenorbicularis contraction tightens this layer andmoves it superomedially.

The presence of a malar septum was de-duced by Pessa and Garza32 and Pessa et al.33 toexplain the several clinical entities associatedwith malar mounds. This delicate structure tothe underside of the dermis of the lowerboundary of the mound is difficult to visualizeon histologic examination32 and was not specif-ically seen in our study, although there areconsiderable subcutaneous retinacula cutis at-tachments between the orbicularis (in the roofof the space) and the overlying dermis. Thepreperiosteal membrane can be seen quiteclearly during some surgical dissections when itreflects off the outer surface of the preperios-

teal fat and onto the upper surface of thezygomatic ligaments. It is presumed that thismembrane extends through the overlying or-bicularis and subcutaneous layer into the der-mis. The location of the membrane and liga-ment directly underlie the well-defined lowerborder of the malar mounds. Presumably, thestructure of the prezygomatic space, definedinferiorly by this membranous reflection, is anintegral factor in the localization of periorbitalecchymosis in addition to the subcutaneouspartitioning.

Malar mounds are most often seen as a com-ponent of the aging changes of the midcheek.Laxity of the tissues forming the roof of theprezygomatic space could be a phenomenonlocalized to the roof, or it could be secondaryto a weakening of the upper peripheral attach-ments of the roof. This laxity can be demon-strated by the application of finger pressureonto the mound with movement in a side-to-side or superior direction. This considerablemobility reduces when the laxity is surgicallycorrected. Although the basis for malarmounds seems to be the downward sliding ofthe tissues of a poorly supported roof of theprezygomatic space against the resistance ofthe more strongly supported lower boundary,secondary changes are observed clinically inthe overlying tissue layers, namely, recurring orchronic edema involving the subcutaneous fat,with or without skin pleats.

Why malar mounds are present in some

FIG. 8. Concept of the pathogenesis of malar mounds.Laxity within the roof of the prezygomatic space may beprimary or secondary to laxity of the upper boundary (ORL).Movement is resisted below because the zygomatico-cutane-ous ligaments (ZL), which form the structure of the lowerboundary, are stronger and more resistant to laxity.


young people can only be postulated. Conceiv-ably, there may be variations among individu-als regarding the tone within the roof of theprezygomatic space either primarily or second-ary to relatively poor support of the superiorboundary. There is a recognized associationbetween baggy, allergic lower lids and the pres-ence of malar mounds, the common denomi-nator being laxity of the orbicularis retainingligament. The strength of the lower border(zygomatico-cutaneous ligaments) would beundiminished (because of the absence of aginglaxity), leaving the mounds high and well de-fined. Another possible variation may be in thevolume of fat contained within the space rela-tive to the tightness of fixation, such that bulg-ing of the roof is due to increased pressurewithin the space. The concept of natural vari-ations is not without precedent, being ob-served in other situations in younger persons,such as constitutionally heavy nasolabial foldsor low-set brows, with a tendency toward pre-mature aging of the feature.

Swelling of malar mounds is often associatedwith swelling of herniated lower lid fat.1 Thereis an anatomic explanation for patients withthis combination of bulges typically having arelatively worsened appearance of the malarmound after undergoing correction of thelower lid bulge alone. To accommodate thevolume of herniated lid fat, there must be con-siderable distension of the orbicularis retain-ing ligament, and both of these factors wouldcontribute to laxity of the roof of the prezygo-matic space. Traditionally, a blepharoplastycorrects the displaced lid fat and tightens theanterior lamella of the lid (skin muscle flap),which may then contrast with (and therebyexaggerate) the uncorrected laxity of the roofof the malar mound immediately below (Fig.8).

Numerous and diverse treatments have beenadvocated for the correction of malar moundsover the years, each directed to one of thecomponent tissue layers. Restoration of skintone has been approached by direct skin exci-sion in selected cases.34 Laser resurfacing of theoverlying skin has more recently been de-scribed as being notably successful in the treat-ment of mounds.35 Treatment directed to thefat layers has been advocated either by super-ficial liposuction of subcutaneous fat36 or directexcision of sub–orbicularis oculi fat.28 Excisionof preperiosteal fat has also been performed.All fat excision techniques are associated with

prolonged postoperative swelling. Excision oforbicularis muscle has also been suggested,1,20

as has direct excision of the full thickness ofthe lax tissue of the roof.37 Reducing the visiblelower edge of the mound has also been sug-gested, using either soft-tissue filling (lipoinfil-tration) or skeletal augmentation with a teartrough silicone implant.34 Transection of themembrane of the lower boundary at a superfi-cial subcutaneous level has not been specifi-cally described in the literature. However, tran-section at the deep subcutaneous level(junction of orbicularis oculi and subcutane-ous fat) has been widely performed in the con-text of midcheek rejuvenation, but, disappoint-ingly, it is associated with exaggeration ofmalar mound swelling for a prolonged periodafter surgery.16,17

The predisposition to prolonged postopera-tive swelling is a characteristic of malar moundsand suggests a lymphatic origin, the basis ofwhich is not resolved by this study. In the ab-sence of a proven effective treatment for malarmounds, treatment should be directed to thecorrection of laxity in the roof of the prezygo-matic space when this is present. The treat-ment of other components (e.g., fat excision ordermal tightening with laser resurfacing)should be complementary to the surgical res-toration of tone within the roof. In the past,the extended lower lid blepharoplasty proce-dure has been described to correct thislaxity.38,39

Tightening of laxity of the orbicularis and itsfascia logically should be performed accordingto the established principles of SMAS surgery.40

This first requires a complete release of therestraining effect of those ligamentous attach-ments that restrict the beneficial effect ofSMAS redraping. In this context, the ligamentsrequired to be released are those of the supe-rior and superolateral boundaries of the prezy-gomatic space, namely, the lateral orbital thick-ening attachment to the frontal process of thezygoma and the confluent stronger lateral partof the orbicularis retaining ligament (where itstill provides resistance.) The laxity of the roofof the space is then “taken up” by redrapingthe SMAS. The vectors for this correction re-quire a predominantly vertical lift. Because thisredraping of the upper cheek is beyond theeffective range of a face lift, a temporal ap-proach is required, not for the usual indicationof elevating the position of the outer brow butrather so the vectors can be applied to the


upper lateral boundary of the prezygomaticspace.

The absence of sufficient vertical correctionis the reason malar mounds worsen when aninfrazygomatic face lift only is performed inthe presence of laxity over the zygoma. Correc-tion of the laxity of the prezygomatic regioncan be accomplished by inclusion of the SMASof this area (i.e., the roof of the prezygomaticspace) in the face lift flap and incorporation ofa temporal lift. This is the anatomic basis forthe composite SMAS face lift described byHamra20 about 10 years ago, but extended toinclude the temporal SMAS (the completecomposite face lift). Because the prezygomaticspace is a junction area linking the lower lid,the lower temporal space, and the lateral andinfrazygomatic parts of the cheek, in surgerythe prezygomatic space can be approachedand defined from all but the last direction. Itprovides the common anatomic pathway forthe midcheek correction achieved by a deeptemporal lift and face lift and for the preperi-osteal midcheek lift performed through thelower lid approach. These can be performed assolitary procedures or in combination.

Damage to the innervation of the orbicularisoculi is the only significant risk when operatingin this region. The multiple motor branches tothe orbicularis are “at risk” where they crossfrom the deeper facial plane to the more su-perficial plane of the orbicularis in the lowerand lateral boundaries of the prezygomaticspace. This is where they are under the “pro-tection” of retaining ligaments. From there thebranches continue in the plane of the sub–orbicularis oculi fat to the preseptal part of theorbicularis.41 The lower lid and temporal ap-proaches are relatively safe as they enter theprezygomatic space from above, on either sideof the zygomatico-facial nerve, and do not en-counter the zygomatic branches. However, thetemporal approach has the potential for a trac-tion neurapraxia of the lowest temporalbranch to the lateral canthus, because itcourses immediately posterior to a zygomaticarch ligament (Fig. 7).

The SMAS face lift approaches the prezygo-matic space through the inferolateral bound-ary, where it is necessary to dissect in the nar-row interval between two ligaments, the majorzygomatic ligament immediately lateral to thezygomaticus major and the most anterior ofthe zygomatic arch ligaments. Because of theproximity of the two nerves of the orbicularis

to these ligaments, releasing these ligamentsfrom the facial skeleton has the potential riskof damage to the innervation of the orbicu-laris. This risk is present not only when theSMAS face lift approach is used to access theprezygomatic space but also when dissection isextended through the lower boundary of theprezygomatic space to correct the infrazygo-matic region of the cheek; i.e., release of thezygomatic ligaments at the inferolateral cornerof the prezygomatic space is associated withthis risk regardless of the surgical approach tothe prezygomatic space.

The zygorbicular face lift approach42 was de-veloped in recognition that the lower lid pro-vides a simpler and safer approach to the mid-face, through the area now known to be theprezygomatic space, and possibly a safer ap-proach to release of the zygomatic ligaments.

CONCLUSIONS

• The triangular prezygomatic space is a junc-tion area between the lower eyelid (above),the lower temple (superolaterally), the lat-eral cheek, and the infrazygomatic part of thecheek (below).

• The roof of the prezygomatic space containsthe orbicularis oculi (pars orbitalis).

• Two distinct layers of fat exist deep to theorbicularis muscle fascia separated by aprezygomatic space.

• The sub–orbicularis oculi fat is superficial tothe prezygomatic space, and the preperios-teal fat is deep in the floor of the space. Adistinct membrane is firmly attached to thepreperiosteal fat.

• The only anatomic structure crossing theprezygomatic space is the zygomatico-facialnerve, crossing immediately adjacent to thethicker part of the orbicularis retainingligament.

• Malar mounds, defined by the ligamentousboundaries of the underlying prezygomaticspace, reflect the laxity of the roof of thespace.

• The zygomatic motor branches to the orbic-ularis oculi course from deep to superficialimmediately outside the prezygomaticspace.

• The zygomatic motor branches to the orbic-ularis enter the sub–orbicularis oculi fatnear the periphery of the orbicularis andtraverse within the sub–orbicularis oculi fatin the roof of the space.


• Of the three surgical approaches to theprezygomatic space, the lateral approach(face lift) is the most at risk of injury to theorbicularis motor branches.

Dr. Bryan C. Mendelson109 Mathoura RoadToorak, Victoria 3142Melbourne, [email protected]

ACKNOWLEDGMENTS

The authors thank Dr. Fritz Barton (Emeritus Professor),Dr. Rod Rohrich (Professor), and Dr. Steven Byrd (AssociateProfessor), The University of Texas Southwestern MedicalCenter at Dallas, for their encouragement, support, and as-sistance with this project.

REFERENCES

1. Furnas, D. W. Festoons, mounds, and bags of the eyelidsand cheek. Clin. Plast. Surg. 20: 367, 1993.

2. Psillakis, J. M., Rumley, T. O., and Camargos, A. Sub-periosteal approach as an improved concept for cor-rection of the aging face. Plast. Reconstr. Surg. 82: 383,1988.

3. Ramirez, O. M., Maillard, G. F., and Musolas, A. Theextended subperiosteal face lift: A definitive soft-tissueremodeling for facial rejuvenation. Plast. Reconstr.Surg. 88: 227, 1991.

4. Hester, T. R., Codner, M. A., McCord, C. D., and Nahai,F. Transorbital lower-lid and midface rejuvenation.Oper. Techn. Plast Reconstr. Surg. 5: 163, 1998.

5. Gunter, J. P., and Hackney, F. L. A simplified trans-blepharoplasty subperiosteal cheek lift. Plast. Reconstr.Surg. 103: 2029, 1999.

6. Hobar, P. G., and Flood, J. Subperiosteal rejuvenationof the midface and periorbital area: A simplified ap-proach. Plast. Reconstr. Surg. 104: 842, 1999.

7. Hamra, S. T. Repositioning the orbicularis oculi musclein the composite rhytidectomy. Plast. Reconstr. Surg. 90:14, 1992.

8. Owsley, J. Q. Lifting the malar pad for correction ofprominent nasolabial folds. Plast. Reconstr. Surg. 91:463, 1993.

9. Mitz, V., and Peyronie, M. The superficial musculo-apo-neurotic system (SMAS) in the parotid and cheekarea. Plast. Reconstr. Surg. 58: 80, 1976.

10. Jost, G., and Levet, Y. Parotid fascia and face lifting: Acritical evaluation of the SMAS concept. Plast. Reconstr.Surg. 74: 42, 1984.

11. Skoog, T. Plastic Surgery: New Methods and Refinements.Philadelphia: Saunders, 1974.

12. Bosse, J. P., and Papillon, J. Surgical anatomy of theSMAS at the malar region. In Transactions of the 9thInternational Congress of Plastic and Reconstructive Sur-gery. New York: McGraw-Hill, 1987. P. 348.

13. Furnas, D. W. The retaining ligaments of the cheek.Plast. Reconstr. Surg. 83: 11, 1989.

14. Stuzin, J. M., Baker, T. J., and Gordon, H. L. The re-lationship of the superficial and deep facial fascias:Relevance to rhytidectomy and aging. Plast. Reconstr.Surg. 89: 441, 1992.

15. Barton, F. E., Jr. Rhytidectomy and the nasolabial fold.Plast. Reconstr. Surg. 90: 601, 1992.

16. Owsley, J. Q., and Fiala, T. G. Update lifting the malarfat pad for correction of prominent nasolabial folds.Plast. Reconstr. Surg. 100: 715, 1997.

17. Owsley, J. Lifting the malar fat pad for correction ofprominent nasolabial folds. Aesthetic Surg. J. 19: 153,1999.

18. Hamra, S. T. The deep-plane rhytidectomy. Plast. Re-constr. Surg. 86: 53, 1990.

19. Mendelson, B. C. Extended sub-SMAS dissection andcheek elevation. Clin. Plast. Surg. 22: 325, 1995.

20. Hamra, S. T. Composite rhytidectomy. Plast. Reconstr.Surg. 90: 1, 1992.

21. Aston, S. J. Correcting mid-facial laxity. Presented at theCutting Edge Symposium, New York, October 1–5,1996.

22. Mendelson, B. C. SMAS fixation to the facial skeleton:Rationale and results. Plast. Reconstr. Surg. 100: 1834,1997.

23. Hinderer, U. T., Urriolagoitia, F., and Vildosola, R. Theblepharo-periorbitoplasty: Anatomical basis. Ann.Plast. Surg. 18: 437, 1987.

24. Hinderer, U. T. The forehead and vertical transpalpe-bral pretemporal-preperiosteal midface rhytidec-tomy. Presented at the Maliniac Lecture, Dallas, No-vember 1996.

25. Byrd, H. S., and Andochick, S. E. The deep temporallift: A multiplanar, lateral brow, temporal, and upperface lift. Plast. Reconstr. Surg. 97: 928, 1996.

26. Byrd, H. S. The extended browlift. Clin. Plast. Surg. 24:233, 1997.

27. Muzaffar, A. R., Mendelson, B. C., and Adams, W. P., Jr.Surgical anatomy of the ligamentous attachments ofthe lower lid and lateral canthus. Plast. Reconstr. Surg.110: 873, 2002.

28. Aiache, A. E., and Ramirez, O. H. The sub–orbicularisoculi fat pads: An anatomic and clinical study. Plast.Reconstr. Surg. 95: 37, 1995.

29. Freilinger, G., Gruber, H., Happak, W., and Pechmann,U. Surgical anatomy of the mimic muscle system andthe facial nerve: Importance for reconstructive andaesthetic surgery. Plast. Reconstr. Surg. 80: 686, 1987.

30. Caminer, D., and Boyd, B. The innervation of the pro-cerus and corrugator muscles and its significance infacial surgery. Presented at the Annual Meeting of theAmerican Society of Plastic and Reconstructive Sur-geons, San Francisco, September 1997.

31. Knize, D. M. An anatomically based study of the mech-anism of eyebrow ptosis. Plast. Reconstr. Surg. 97: 1321,1996.

32. Pessa, J. E., and Garza, J. R. The malar septum: Theanatomic basis of malar mounds and malar edema.Aesthetic Surg. J. 17: 1, 1997.

33. Pessa, J. E., Zadoo, V. P., Adrian, E. K., Woodwards, R.,and Garza, J. R. Anatomy of a “black eye”: A newlydescribed fascial system of the lower eyelid. Clin. Anat.11: 157, 1998.

34. Flowers, R. Cosmetic surgery, state of the art. Advancesin Plast. Reconstr. Surg. 8: 319, 1992.

35. Roberts, T., and Yokoo, K. M. In pursuit of optimalperiorbital rejuvenation, laser resurfacing with orwithout blepharoplasty and browlift. Aesthetic Surg. J.18: 321, 1998.

36. Rosenberg, G. J. Correction of saddlebag deformity ofthe lower eyelids by superficial suction lipectomy.Plast. Reconstr. Surg. 96: 1061, 1995.

37. Castanares, S. A comparison of blepharoplasty tech-


niques. In F. Masters and J. R. Lewis (Eds.), AestheticSurgery of the Eyelids and Face. St. Louis: Mosby, 1972.

38. Small, R. G. Extended lower lid blepharoplasty. Arch.Ophthalmol. 99: 1402, 1981.

39. Adamson, P. A., Tropper, G. J., and McGraw, B. L. Ex-tended blepharoplasty. Arch. Otolaryngol. Head NeckSurg. 117: 606, 1991.

40. Mendelson, B. C. Surgery of the superficial muscu-

loaponeurotic system: Principles of release, vectors,and fixation. Plast. Reconstr. Surg. 107: 1545, 2001.

41. Ramirez, O. A., and Santomarina, R. Spatial orientationof motor innervation to the lower orbicularis oculimuscle. Aesthetic Surg. J. 20: 107, 2000.

42. Hamra, S. T. The zygorbicular dissection in compositerhytidectomy: An ideal midface plane. Plast. Reconstr.Surg. 102: 1646, 1998.


Documents

Cosmetic Surgical Anatomy of the Midcheek and Malar Mounds · the level of the zygomatico-facial nerve, showing the prezy-gomatic space. The preperiosteal fat is covered by an adher-ent