Key points

Introduction

When considering the optimal reconstructive options for a particular cheek defect, familiarity with the basic anatomy of the face is critical, including both surface anatomy and contour arising from the underlying bony structures and soft tissues. The general borders of the cheek include the temple, infraorbital rim and lower eyelid, preauricular sulcus (including borders with the tragus/helix/lobule), mandible, lateral nasal wall, nasal ala, and melolabial fold (upper lip). It is also important to consider the physical properties of the skin and underlying tissues to be rearranged, as improper planning may lead to unsightly distortion of facial anatomy. This consideration is especially important for the facial creases of the cheek and the regions of the cheek adjacent to the lower eyelid, nasal ala, and lips. As with any facial defect, anatomic subunits and relaxed skin tension lines (RSTLs) play a key role in the selection and orientation of local flap repair for cheek defects. Based on anatomic and structural properties, the cheek may be divided into aesthetic subunits.

Over the years, different investigators have offered various patterns by which the aesthetic units of the cheek may be defined. For instance, Weerda has described division of the cheek into 6 anatomic units, including the upper, medial, and lateral divisions of the medial and lateral cheeks, whereas Bradley and Murakami have partitioned the cheek into medial, lateral (mandibular), zygomatic, and buccal divisions. Menick posits that aesthetic cheek units vary from individual to individual and are dynamic based on age, hairline, hairstyle, facial hair, and facial expression. Given the diversity of the cheek subunit, the primary reconstructive goals of cheek defects include restoration of skin color and texture, which are more conspicuous than variations in contour and subunit outline.

We have divided the cheek into the following regions with functional and aesthetic significance: medial cheek, perilabial (buccal) cheek, lateral cheek, and zygomatic cheek. Menick advocates that the subunit principle should be followed whenever possible in reconstruction of aesthetic regions of the face, although slight variations from “normal” are much more forgiving in cheek reconstruction when compared with the nose, lips, and eyelids. Moreover, subtle asymmetries in color, texture, and surface topography are generally less obvious in the cheek because direct visual comparison of one cheek to its paired contralateral unit is limited in most views aside from the direct anterior view. However, if possible, the surgeon should use the contralateral normal subunit as a template to recreate symmetry and the optimal cosmetic outcome. The benefit of en bloc reconstruction of entire aesthetic subunits and/or use of strategic scar placement at aesthetic borders is that resultant raised or depressed scars tend to be more subtle within the natural contours, shadows, and accents of the native facial structure that define subunit borders.

If placement along subunit borders is impossible, defects should be closed with incisions parallel to RSTLs, inherently minimizing skin tension closure, because RSTLs are determined by the elasticity of the underlying tissue and resultant tension on the skin. Although skin of the medial and buccal cheek is thicker and mobile, skin of the superolateral cheek is relatively affixed to fascia lying beneath. This fixation owes to various retaining ligaments anchoring the skin of the cheek to underlying bone. Overlying skin is very tightly fixed to the zygoma due to particularly robust retaining ligaments called the McGregor patch. The labiomandibular crease is another point of dense ligamentous attachment that is formed from attachments from the mandibular retaining ligament to the overlying skin. The topography of the cheek is determined from the aforementioned retaining ligaments and fibrous attachments to the superficial musculoaponeurotic system (SMAS), as well as the underlying facial skeleton, malar fat pad, and the muscles of facial expression. In general, larger reconstructions often yield better outcomes due to more robust subdermal pedicle and via en bloc subunit repair. Likewise, deliberate standing cutaneous deformity excision commonly yields optimal results.

The vascular supply to the cheek skin is derived primarily from branches of the external carotid artery and is of critical importance when considering flap design to optimize potential flap viability. The primary arterial supply to the cheek skin is the facial artery and its angular branch. Additional arterial contributions to the cheek include the infraorbital branch of internal maxillary artery and the transverse facial branch of superficial temporal artery. There are frequently anastomotic connections between these arteries as well. Branches of these named arteries ultimately supply the dermal and subdermal plexuses that perfuse the cheek skin. The venous drainage of the cheek largely mirrors its arterial supply, including facial vein, superficial temporal vein, and retromandibular vein with drainage into the internal and external jugular venous systems.

If the reconstructive surgeon is also responsible for locoregional control, such as in management of certain cutaneous malignancies, familiarity with lymphatic drainage of the cheek is important for situations in which sentinel lymph node biopsy, neck dissection, or adjuvant therapy is considered. First-echelon lymphatic drainage from the cheek includes submandibular, preauricular, and submental lymph nodes. Furthermore, second-echelon lymphatic drainage from the cheek consists of superficial jugular lymph nodes.

Until margin status can be determined, avoidance of wound bed/margin distortion is strongly recommended. For lesions with high risk of local recurrence (eg, melanoma, Merkel cell carcinoma, dermatofibrosarcoma protuberans) or in need of immunohistochemistry to confirm the diagnosis, a 2-blade square technique for confirming margins can be used to avoid an open wound while awaiting margin status. Sophisticated tissue rearrangements should be reserved until confirmation of clear margins, particularly in melanomas managed with sentinel lymph node biopsy whose accuracy may be altered via disruption of dermal lymphatics and tissue rearrangement. The wound may be dressed with a bolster or closed via techniques that do not distort the wound margin, including primary closure, healing by secondary intention, or full-thickness or split-thickness skin graft. On the other hand, most nonmelanoma skin cancers, including squamous cell and basal cell carcinoma, have a favorable cure rate with microsurgery, and can be reliably reconstructed during a single-stage procedure. Simpler reconstructive options also may be optimal in patients at high risk for flap or graft failure (for example, heavy smokers, diabetic individuals, patients with severe vasculopathy, or patients at risk of hematoma, such as patients on high-dose antiplatelet or anticoagulant medications) or in patients at risk of developing additional primary lesions at or near the reconstruction (such as immunosuppressed patients or those with strong genetic predisposition to cutaneous malignancies, eg, xeroderma pigmentosa). However, in general, the optimal reconstructive option is often determined based on the location, size, and depth of the cheek defect.

A secondary function of the cheek and upper lip is to serve as a platform for the nose. In general, if cheek defects also involve the adjacent nasal and upper lip subunits, the cheek and upper lip should be reconstructed before and independent from the nasal defect to reestablish the base for nasal reconstruction. Flaps and grafts should be designed with their borders terminating at the nasofacial and alar-facial sulci. Subsequently, portions of the defect involving the nasal sidewall or ala are best reconstructed once the nasal platform is reestablished and healed, allowing any potential distortion from wound contraction to have already declared itself before nasal reconstruction. One exception to this general guideline is the shark island pedicle flap (SIPF), discussed later in this article, which allows single-stage reconstruction of the medial cheek and nasal ala with a single advancement flap without blunting critical sulci.

Introduction

When considering the optimal reconstructive options for a particular cheek defect, familiarity with the basic anatomy of the face is critical, including both surface anatomy and contour arising from the underlying bony structures and soft tissues. The general borders of the cheek include the temple, infraorbital rim and lower eyelid, preauricular sulcus (including borders with the tragus/helix/lobule), mandible, lateral nasal wall, nasal ala, and melolabial fold (upper lip). It is also important to consider the physical properties of the skin and underlying tissues to be rearranged, as improper planning may lead to unsightly distortion of facial anatomy. This consideration is especially important for the facial creases of the cheek and the regions of the cheek adjacent to the lower eyelid, nasal ala, and lips. As with any facial defect, anatomic subunits and relaxed skin tension lines (RSTLs) play a key role in the selection and orientation of local flap repair for cheek defects. Based on anatomic and structural properties, the cheek may be divided into aesthetic subunits.

Over the years, different investigators have offered various patterns by which the aesthetic units of the cheek may be defined. For instance, Weerda has described division of the cheek into 6 anatomic units, including the upper, medial, and lateral divisions of the medial and lateral cheeks, whereas Bradley and Murakami have partitioned the cheek into medial, lateral (mandibular), zygomatic, and buccal divisions. Menick posits that aesthetic cheek units vary from individual to individual and are dynamic based on age, hairline, hairstyle, facial hair, and facial expression. Given the diversity of the cheek subunit, the primary reconstructive goals of cheek defects include restoration of skin color and texture, which are more conspicuous than variations in contour and subunit outline.

We have divided the cheek into the following regions with functional and aesthetic significance: medial cheek, perilabial (buccal) cheek, lateral cheek, and zygomatic cheek. Menick advocates that the subunit principle should be followed whenever possible in reconstruction of aesthetic regions of the face, although slight variations from “normal” are much more forgiving in cheek reconstruction when compared with the nose, lips, and eyelids. Moreover, subtle asymmetries in color, texture, and surface topography are generally less obvious in the cheek because direct visual comparison of one cheek to its paired contralateral unit is limited in most views aside from the direct anterior view. However, if possible, the surgeon should use the contralateral normal subunit as a template to recreate symmetry and the optimal cosmetic outcome. The benefit of en bloc reconstruction of entire aesthetic subunits and/or use of strategic scar placement at aesthetic borders is that resultant raised or depressed scars tend to be more subtle within the natural contours, shadows, and accents of the native facial structure that define subunit borders.

If placement along subunit borders is impossible, defects should be closed with incisions parallel to RSTLs, inherently minimizing skin tension closure, because RSTLs are determined by the elasticity of the underlying tissue and resultant tension on the skin. Although skin of the medial and buccal cheek is thicker and mobile, skin of the superolateral cheek is relatively affixed to fascia lying beneath. This fixation owes to various retaining ligaments anchoring the skin of the cheek to underlying bone. Overlying skin is very tightly fixed to the zygoma due to particularly robust retaining ligaments called the McGregor patch. The labiomandibular crease is another point of dense ligamentous attachment that is formed from attachments from the mandibular retaining ligament to the overlying skin. The topography of the cheek is determined from the aforementioned retaining ligaments and fibrous attachments to the superficial musculoaponeurotic system (SMAS), as well as the underlying facial skeleton, malar fat pad, and the muscles of facial expression. In general, larger reconstructions often yield better outcomes due to more robust subdermal pedicle and via en bloc subunit repair. Likewise, deliberate standing cutaneous deformity excision commonly yields optimal results.

The vascular supply to the cheek skin is derived primarily from branches of the external carotid artery and is of critical importance when considering flap design to optimize potential flap viability. The primary arterial supply to the cheek skin is the facial artery and its angular branch. Additional arterial contributions to the cheek include the infraorbital branch of internal maxillary artery and the transverse facial branch of superficial temporal artery. There are frequently anastomotic connections between these arteries as well. Branches of these named arteries ultimately supply the dermal and subdermal plexuses that perfuse the cheek skin. The venous drainage of the cheek largely mirrors its arterial supply, including facial vein, superficial temporal vein, and retromandibular vein with drainage into the internal and external jugular venous systems.

If the reconstructive surgeon is also responsible for locoregional control, such as in management of certain cutaneous malignancies, familiarity with lymphatic drainage of the cheek is important for situations in which sentinel lymph node biopsy, neck dissection, or adjuvant therapy is considered. First-echelon lymphatic drainage from the cheek includes submandibular, preauricular, and submental lymph nodes. Furthermore, second-echelon lymphatic drainage from the cheek consists of superficial jugular lymph nodes.

Until margin status can be determined, avoidance of wound bed/margin distortion is strongly recommended. For lesions with high risk of local recurrence (eg, melanoma, Merkel cell carcinoma, dermatofibrosarcoma protuberans) or in need of immunohistochemistry to confirm the diagnosis, a 2-blade square technique for confirming margins can be used to avoid an open wound while awaiting margin status. Sophisticated tissue rearrangements should be reserved until confirmation of clear margins, particularly in melanomas managed with sentinel lymph node biopsy whose accuracy may be altered via disruption of dermal lymphatics and tissue rearrangement. The wound may be dressed with a bolster or closed via techniques that do not distort the wound margin, including primary closure, healing by secondary intention, or full-thickness or split-thickness skin graft. On the other hand, most nonmelanoma skin cancers, including squamous cell and basal cell carcinoma, have a favorable cure rate with microsurgery, and can be reliably reconstructed during a single-stage procedure. Simpler reconstructive options also may be optimal in patients at high risk for flap or graft failure (for example, heavy smokers, diabetic individuals, patients with severe vasculopathy, or patients at risk of hematoma, such as patients on high-dose antiplatelet or anticoagulant medications) or in patients at risk of developing additional primary lesions at or near the reconstruction (such as immunosuppressed patients or those with strong genetic predisposition to cutaneous malignancies, eg, xeroderma pigmentosa). However, in general, the optimal reconstructive option is often determined based on the location, size, and depth of the cheek defect.

A secondary function of the cheek and upper lip is to serve as a platform for the nose. In general, if cheek defects also involve the adjacent nasal and upper lip subunits, the cheek and upper lip should be reconstructed before and independent from the nasal defect to reestablish the base for nasal reconstruction. Flaps and grafts should be designed with their borders terminating at the nasofacial and alar-facial sulci. Subsequently, portions of the defect involving the nasal sidewall or ala are best reconstructed once the nasal platform is reestablished and healed, allowing any potential distortion from wound contraction to have already declared itself before nasal reconstruction. One exception to this general guideline is the shark island pedicle flap (SIPF), discussed later in this article, which allows single-stage reconstruction of the medial cheek and nasal ala with a single advancement flap without blunting critical sulci.

Wound preparation

For optimal tissue apposition and eversion of the incisional closure, wound edges are prepared by orienting the wound’s incisional axis perpendicular to the skin surface. Following Mohs micrographic surgery, the excisional bed is often saucerized and commonly leaves wound edges oriented at an obtuse angle to the skin surface. These beveled edges are excised at a 90° angle to the skin surface at a depth appropriate for the proposed recipient flap or graft. For defects comprising more than 50% of an aesthetic subunit, the remainder of the subunit may be excised for optimal scar placement along the subunit border although this is done infrequently in practice.

Depending on the defect’s size, shape, depth, and location (including the subunits involved with and adjacent to the defect), several aesthetically pleasing reconstructive options are available to the reconstructive surgeon for defects of the cheek. One advantage of skin and soft tissue defects of the cheek is the inherent elasticity and mobility of these tissues in this area, allowing adjacent tissue recruitment, which in general lends to a high degree of reconstructive flexibility.

Tissue handling, flap survival, dissection depth

The dissection plane of local flaps of the cheek is typically within the subcutaneous plane, with blood supply based on the subdermal and subcutaneous plexus. Alternatively, depending on the desired flap thickness or wound tension, elevation of the flap may be performed in a sub-SMAS plane based on transverse facial artery perforators. For example, Jacono and colleagues demonstrated superior distal flap survival in deep-plane (sub-SMAS) cervicofacial advancement flap elevation in Mohs repair, thereby reducing the risk of distal edge necrosis and potential distortion of adjacent structures from suboptimal scarring. The sub-SMAS technique may also be used to reduce superficial wound tension. The study by Jacono and colleagues implied improved flap survival with deep-plane elevation in repair of larger defects, as well as the possibility of improved flap survival in active smokers due to more robust vascular supply. The dissection plane is also critically important in terms of identification and protection of the facial nerve. If the dissection plane is maintained superficial to the parotidomasseteric fascia, the nerve is typically well-protected, although once the flap is elevated anteriorly to the anterior border of the parotid gland, the nerve emerges from the parotid tissue and necessitates careful dissection and identification. For interpolated flaps, V-to-Y flaps, and other circumferential flap designs, preservation of musculocutaneous perforators is also critical.

Flap survival and cosmesis are improved with a variety of basic soft tissue techniques, which include atraumatic tissue handling, such as using sharp skin hooks and sharp retractors for gentle tissue retraction, tissue handling with toothed forceps to avoid crushing, preparation of the wound bed, and maintaining flap moisture with periodic application of moist gauze. Furthermore, the risk of distal wound necrosis also can be limited with avoidance of wound closure under tension, closure of the incisions in layers, and use of pulley sutures when appropriate. Undue tension at the distal point of a large flap may be avoided by anchoring the flap with more proximal tacking sutures before inset of the distal flap. This technique is also particularly helpful in avoiding necrosis. Finally, care should be taken during repair such that wound edges are everted (for which interrupted vertical mattress sutures can be particularly useful), and superficial nonabsorbable sutures should be promptly removed (ideally within 7–10 days) to avoid crosshatching.

Healing of cheek defects by secondary intention, primary closure, and scar placement along aesthetic boundaries

In general, healing by secondary intention has a limited role, as it tends to lead to suboptimal aesthetic outcomes in cheek defect repair. This approach is generally reserved almost exclusively for small defects of the nasofacial, melolabial, preauricular, or alar-facial sulci. Resultant scars from secondary intention tend to contract and lack requisite convexity to reconstruct defects of the remainder of the cheek, thereby making contracture and discoloration more conspicuous. Conversely, the relative concavity of these prominent sulci often conceals scars well. Primary closure provides an excellent option for small (<1 cm) cheek defects, those nearing aesthetic subunit borders (especially within the nasofacial, melolabial, preauricular, or alar-facial sulci), or those able to be closed parallel to RSTLs.

The amenability of a lesion to primary closure is largely dependent on tissue laxity, but in general is a viable option for defects consisting of 30% or less of the total aesthetic subunit. On the other hand, larger defects often rely on recruitment of tissue from the adjacent cheek or from the neck. Elliptical closure can be used to accomplish ideal scar orientation in circular or ovoid defects that commonly arise from micrographic resection. Because ample undermining is required to close increasingly large defects, differential or asymmetric undermining is one technique that is often useful to avoid disturbing the natural positions of facial anatomy and for preferential incisional placement within sulci or aesthetic borders. Furthermore, extending incisions to obscure standing cutaneous cones within aesthetic borders is also useful, as is incising intervening tissue between a defect and the nearest anatomic boundary. Primary closure is a good option for defects bordering the nasal sidewall that obscures the closure within the nasofacial sulcus. Likewise, the preauricular crease is useful to conceal lateral cheek defects, and the lateral rhytids offer additional disguise for small superolateral cheek defects with proximity to the lateral canthus. To better plan for incisional placement in the zygomatic region, the patient can exaggerate the lateral rhytids by clenching the eyes preoperatively, which accentuates the wrinkles along which the closure’s linear axis will lie. Additionally, for defects of the medial zygomatic subunit threatening to impinge on the lids or lateral canthus, an M-plasty may be used to shorten the length of the incision, thereby preventing it from distorting the eyelid. In general, flaps transposed, advanced, or rotated toward the eyelids and lateral canthus transmit lateral tension on the lateral canthus and eyelids. This tension and the tissue distortion that results can be lessened by anchoring the flap to the static underlying periosteum surrounding the orbit.

The melolabial fold, the prominent sulcus serving as the boundary between the medial cheek and upper lip, represents another vastly important aesthetic landmark. Although in certain circumstances the melolabial fold provides a reconstructive challenge, it also allows for camouflage of primary closure of small defects. Differential undermining also may be used to preferentially position scars in aesthetically optimal locations. These techniques may be useful for closure of medium-sized perilabial and perinasal cheek defects adjacent to aesthetic boundaries and facial creases or oriented parallel to RSTLs.

Much like primary closure, the leading border of advancement, transposition, or rotational flap closures can be strategically placed within the nasofacial, melolabial, preauricular, and alar-facial sulci by means of excision of intervening tissues to position standing cones and incisions along these borders. The inherent elasticity and mobility of medial cheek tissue adjacent to the defect allows closure of the defect without appreciable tension or distortion of surrounding anatomy while conserving tissue and positioning incisions along aesthetic boundaries. The nasofacial sulcus, alar-facial sulcus, and melolabial crease often provide ideal opportunities to hide defects using this technique.

Periorbital cheek defects and considerations for local flaps bordering the orbit and lower lid

When it comes to maintaining both functional and aesthetic outcomes in cheek reconstruction, eyelid considerations are paramount. Therefore, when planning repair of cheek defects adjacent to the inferior orbital rim, lower eyelid, medial canthus, and/or the lateral canthus, one should anticipate risk for distortion of the eyelid or canthi, as appropriate planning to avoid major functional and aesthetic implications. As a general rule, a vertical downward vector is avoided in proximity of the lower lid. When a local flap is transferred near the eyelids or orbital rim, the border of the flap should be anchored to the underlying periosteum at the orbital rim, medial canthus, and/or lateral canthus to alleviate unwanted tension and reduce the risk of ectropion and lid traction.

For large defects adjacent to the nasal ala, lateral nasal wall, and inferior orbital rim, large cervicofacial flaps and rotational/advancement flaps are often required, and scar concealment presents a unique challenge. In young patients with adequate orbicularis oculi tone, the subciliary line often offers optimal scar camouflage for defects of the perinasal and periocular cheek. In such cases, both the resultant tension vectors imposed by flap transfer as well as the physical properties exhibited by the transferred tissue are primary concerns. Care must be taken in such cheek advancements to avoid ectropion, inferior lid retraction, and chronic lower lid edema. Elderly individuals lacking lower lid tone and robust lymphatic drainage are particularly susceptible to persistent edema. Therefore, patient selection to determine preoperative lid laxity is crucial to avoid functional and aesthetic abnormalities. Neglect of such considerations carries risk of corneal abrasion, keratitis, and even blindness.

Several techniques have been described to assess lid laxity, most of which were initially developed for use during preoperative assessment before blepharoplasty. Two such tests include the eyelid distraction test and the eyelid snap test, and these also may be used to screen patients who may be suited for subciliary advancement flaps. In the eyelid distraction test, the lower lid is pulled away from the eyeball and the distraction of the lid from the eyeball should not exceed 6 mm. Alternatively, the eyelid snap test estimates eyelid muscular tone and consists of pulling the lower lid toward the lower orbital rim. A lower eyelid with adequate orbicularis tone should return to the resting position spontaneously. A lower lid with significant laxity may require one or more blinks to return to resting position. An eyelid with normal eyelid snap test should tolerate postoperative scar healing without significant eyelid retraction following advancement flap closure involving the subciliary line. For patients thought to be poor candidates for subciliary advancement flap closure, an alternative option for incisional placement includes the inferior bony orbital rim. This option also maintains aesthetic boundaries (medial cheek and lower eyelid borders) while still placing the incision in an aesthetically inconspicuous location. Moreover, routine concurrent lateral canthopexy or temporary tarsorrhaphy have also been advocated for advancement flap closures nearing the lower lid to further prevent retraction. Irrespective of the ultimate incisional placement for superomedial cheek defects, the use of suspension sutures to anchor the flap to the infraorbital rim periosteum is recommended to alleviate inferior traction on the eyelid. For further medial support, the superomedial tip of the advancement flap may be deepithelialized and secured to the periosteum of the medial canthal region with dermis-to-periosteum anchoring sutures.