|Year : 2007 | Volume
| Issue : 1 | Page : 75-101
Developmental field reassignment in unilateral cleft lip: Reconstruction of the premaxilla
Michael H Carstens
Division of Plastic Surgery, Saint Louis University, USA
Michael H Carstens
Division of Plastic Surgery, Saint Louis University
Source of Support: None, Conflict of Interest: None
|How to cite this article:|
Carstens MH. Developmental field reassignment in unilateral cleft lip: Reconstruction of the premaxilla. Indian J Plast Surg 2007;40:75-101
|How to cite this URL:|
Carstens MH. Developmental field reassignment in unilateral cleft lip: Reconstruction of the premaxilla. Indian J Plast Surg [serial online] 2007 [cited 2017 Mar 28];40:75-101. Available from: http://www.ijps.org/text.asp?2007/40/1/75/32670
In the 21 st century the greatest stimulus for progress in cleft surgery will come from more a more accurate model of facial development and how clefts originate. Victor Veau  accurately predicted this: "All cleft surgery is merely applied embryology." The revolution in developmental biology has not yet been incorporated into surgical practice. The drawings and terminology used in textbooks today are based on the work of Wilhelm His in the 1870's.  Cleft repairs are therefore designed based on the anatomy as it appears in the newborn. Measurements are taken and the tissues are geometrically manipulated. But the anatomy of a cleft as seen at the time of birth is far different from its original configuration in the embryonic face. From its onset at gastrulation, the clefting event unleashes pathologic processes that predictably alter the original embryonic anatomy over time to produce what we recognize at term as a cleft lip. Left uncorrected, these processes will remain operative throughout facial growth. This explains why geometric cleft repairs relapse over time, requiring revision.
Developmental field repair (DFR) is based upon the neuromeric model of craniofacial embryology. , The goals of DFR surgery are: (1) resolution of all pathologic processes of clefting (deficiency, displacement, division and distortion); (2) dissection along embryonic separation planes (subperiosteal release); (3) preservation of blood supply to the alveolar mucoperiosteum; (4) primary unification of the dental arch; and (5) reassignment of all developmental fields to their correct relationships. Before describing the surgical technique, certain basic concepts of field theory should be understood.
Craniofacial anatomy results from the assembly of specific, identifiable developmental fields . Fields are discrete units composed of ectoderm, mesoderm, endoderm and neural crest. At each level of the embryo all germ layers specific to that level can be referenced back to developmental zones of the neuraxis known as a neuromeres . These are named using the nomenclature of neuroembryology. The master plan of the entire embryo is determined by 6 prosomeres (prosencephalon-forebrain), 2 mesomeres (mesencephalon-midbrain), 12 rhombomeres (rhomboencephalon-hindbrain and 31 myelomeres (spinal cord). ,,
| » Rubenstein|| |
Applications of neuromeric anatomy provide a potential embryonic "map" of all craniofacial structures with important implications for diagnosis and surgery. Exclusive of the cranial base (basisphenoid and posterior) and parietal bone, the craniofacial skeleton is made from exclusively from neural crest. Thus the cell populations producing the ethmoid, presphenoid, premaxilla and vomer all originate in antero-posterior order from the neural folds in genetic register with the 1 st rhombomere (abbreviated r1). The inferior turbinate, palatine bone, maxilla, alisphenoid (greater wing) and zygoma arise from the neural crest of the 2 nd rhombomere (r2). The squamous temporal, mandible, malleus and incus are r3 neural crest bones.
Non-neural crest craniofacial bones come from paraxial mesoderm (PAM) lying immediately adjacent to the neural tube. PAM is divided into individual units shaped like popcorn balls called somitomeres , each one in register with its corresponding neuromere. The first seven somitomeres (Sm) contain the myoblasts for all muscles of the orbit and the first three pharyngeal arches. For example, the mandible comes from r3 neural crest and all muscles originating from it arise from Sm3. Beginning with Sm8 all somitomeres undergo a further transformation into somites , each having a dermatome, myotome and sclerotome. The first four somites (derived from Sm8-Sm11) produce the cranial base posterior to the sphenoid, the muscles of the tongue and part of the sternocleidomastoid and trapezius. These are called occipital somites .
Disturbances at a particular neuromeric level can affect individual or multiple fields to be deficient or absent. Thus, isolated cleft palate (unassociated with cleft lip) represents a deficiency state of the vomer. This occurs as a spectrum. As the vomer is progressively smaller it lifts away from the plane of the palatal shelves and the cleft extends forward toward the incisive foramen. In mild cases of cleft palate associated with Pierre Robin sequence , a reduction in the horizontal plate of the r2 palatine bone is seen. Soft palate muscles are consequently normal but divided. As the pathology worsens, reduction in the horizontal plate of the r2 maxilla creates the well-known "horse-shoe" palate cleft. Submucous cleft palate , on the other hand, involves pathology in the 3 rd pharyngeal arch. Somitomeres 6 and 7 contain the myoblasts of levator, uvulus, palatopharyngeus and superior constrictor. These can be globally affected. Frequently, persistent VPI follows a seemingly simple palatoplasty, requiring further surgery. Failure to stratify cleft palate by embryologic mechanism explains much of the confusion currently extent in the speech and surgical literature .
Finally. Treacher-Collins syndrom e provides an example in which all r2 developmental fields of the midface are affected: the maxillary, palatine and zygomatic bones are all small. The septum, vomer and premaxilla (being r1 structures) are unaffected. For this reason, the central midface projects normally while the dimensions of the palate, maxilla and zygoma are constricted.
Developmental fields form in a specific spatio-temporal sequence. Each one builds upon its predecessors. Making a face is much akin to assembling a house with magical pieces of Lego®, each one of which will grow over time. Imagine a Lego house made from 20 pieces (4 on the floor and 5 stories high). All pieces are growing independently. If a cornerstone piece is removed, the 19 remaining pieces undergo a deformation and the house tilts into the deficiency site. The missing Lego piece in cleft lip is the premaxilla . The physiologic basis of DFR is the reconstruction of the premaxillary field. ,, This chapter presents the reconstructive application of these principles to the surgery of labiomaxillary clefts.
The pathologic anatomy of cleft formation
The pathologic anatomy of unilateral and bilateral labiomaxillary clefts stems from a tissue deficiency state localized to the lower lateral piriform fossa. The tissue at fault is the mesenchyme of the ipsilateral premaxilla. Neural crest stem cells responsible for synthesis of the presphenoid and ethmoid arise from the mesencephalic neural folds and are genetically identified with the 1 st rhombomere (r1). Note that the basisphenoid is not neural crest in origin; it comes from paraxial mesoderm from somitomere 1 (Sm1). Sm1 lies just outside the neural tube at level r1 and is in register with it. Immediately caudal to this population are neural crest cells immediately above the rostral rhomboencephalon with the 2 nd rhombomere (r2). The most rostral zone of r2 (herein referred to as r2') is the likely source material for the vomer and PMx. The more caudal zone of r2 produces inferior turbinate (IT), palatine bone (P), maxilla (Mx), alisphenoid (AS) and zygoma (Z).
These cell populations migrate forward into the developing face in a strict temporo-spatial order. The sphenoid is laid down first, followed by the ethmoid. In like manner, formation of PMx is the prerequisite for the appearance of V. Formation of the PMx and V requires the pre-existence of the perpendicular plate of the ethmoid (PPE). The function of the PPE is to provide a cellular scaffold by which r2' neural crest cells can reach the midline.  [Figure - 1],[Figure - 2] In holoprosencephaly (HPE) the PPE can be absent. PMx and V cannot develop correctly; a wide bilateral cleft results. ,,
The piriform fossa of humans and some high primates is assembled as the fusion of the frontal process of the premaxilla (PMxF) and the frontal process of the maxilla (MxF). In all other vertebrate skulls PMxF and MxF are readily visible as two distinct entities.  Evolution foreshortened the human snout. The two fields became superimposed, PMxF becoming telescoped internal to MxF . This lamination is responsible for the strength of the piriform rim. Plating of the piriform "buttress" in fracture repairs takes advantage of this bicortical anatomy.
| » Piriform|| |
The developmental field in which the PMx resides consists of an epithelium and a mesenchyme. Formation of the PMx results from interactions between these tissues. The premaxilla has several anatomic subcomponents, these are assembled in a strict sequence. , In dental terminology the central incisor is called "A" and the lateral incisor is called "B." A erupts before B. Therefore the central incisor field (PMxA) is biologically "older" than the lateral incisor field (PMxB). Neural crest mesenchyme flows forward along the previously-established PPE. It first encounters the epithelium corresponding to PmxA and then "spills over" into zone PMxB. The time sequence of dental eruption (central incisor A > lateral incisor B) is a manifestation of the relative biologic "maturity" of the mesenchymal field within which each tooth develops. The frontal process field (PMxF) is a vertical offshoot of PMxB; this subfield is the biologically "newest" tissue.
Pathology affecting the PMx occurs as a spectrum based on this original developmental pattern. A deficiency state of the PMx will first occur in the most distal aspect of the frontal process (ie. at its most cranial extent). As the mesenchymal deficit worsens, frontal process will be reduced in a cranial-caudal gradient. "Scooping out" of the piriform rim results; the nasal lining is pulled down as well. This causes depression of the alar base and a downward-lateral displacement of the lateral crus. Biologic signals from PMxF regulate epithelial stability and therefore affect lip formation (vide infra). When the signal strength is minimally disturbed the lip is normal despite the piriform distortion. Therefore the forme fruste manifestation of premaxillary deficiency is a cleft lip nose with a perfectly normal lip. 
Once the frontal process is eliminated, the deficiency state shows up in the lateral incisor field. Progressive degrees of PMx deficiency in the lateral incisor field cause incremental loss of alveolar bone. Normal alveolar development follows a gradient. It begins at the incisive foramen and progresses forward. Mild deficiency causes notching on the labial surface. As the deficiency worsens the notch deepens backward toward the incisive foramen. A critical lack of alveolar bone mass results in outright failure of lateral incisor development.
BM P-4 signals from this field are directly implicated in the mechanism of fusion between the lateral lip element and the prolabium. , BMP-4 emanating from PmxL forms a cranio-caudal chemical gradient. The strength of this gradient depends upon the total amount of available BMP-4; this in turn is proportional to the overall mesenchymal mass of PMxB. Reduction in mass of the lateral incisor field results in a diminution of the total BMP-4 signal. Lip fusion follows this same gradient. Mild weakness of the BMP-4 gradient will result in notching of the vermilion. As the situation worsens the extent of the lip cleft worsens in a cranial direction. The clinical spectrum of the so-called minimal cleft lip deformity faithfully reproduces this biologic sequence. 
In summation, variations in clefts involving the primary palate and the lip can be understood as interactions between deep plane fields of the premaxilla, the maxilla and superficial plane field of the lateral lip element. The mesenchyme of the lip has a different embryologic origin. It is genetically identified with the 2 nd rhombomere. Neural crest cells from r2 provide the dermis and the subcutaneous tissue of the alar base, while paraxial mesoderm from the 2 nd somitomere forms the anterior half of the squamous temporal bone and the cranial half of the parotid gland. All derivatives from level r2 can be mapped out within the sensory distribution of V2, the nucleus of which resides within r2. ,,,,,
The developmental anatomy of the nose, prolabium and premaxilla has been previously described in terms of neuromeric theory by this author. ,,, In contrast to the rest of the body, all dermis and submucosa of the head originates from neural crest cells , not from a dermatome associated with a somite.  Pre-dermal neural crest arises from three distinct zones of the embryo. The dermis of the forehead, nose and vestibular lining come from the caudal prosencephalon (above prosomeres p4-p1). This prosencephalic neural crest (PNC) migrates forward like a gigantic glacier to occupy the neural folds above prosomeres p6 and p5. The alar half of p6 and p5 creates the telencephalon (cerebrum). The basal halves of p6 and p5 plus all remaining prosomeres (p4-p1) synthesize the diencephalon (epithalamus, thalamus and hypothalamus). The neural folds above p6 and p5 are "sterile." They contain the pituitary, olfactory and optic placodes, but no neural crest. When PNC flows forward into this zone the placodes are activated and dermis is formed. Nasal vestibular lining from the cribriform plate forward to the internal nasal valve comes from p6 PNC. All remaining frontonasal dermis from the internal nasal valve forward to the hairline comes from p5 PNC.
PNC skin shares sensory innervation with the dura of the underlying frontal lobe. V1, the sensory nerve for this common zone has its nucleus with the 1 st rhombomere (r1). The neuroanatomy is analogous for the rest of the face. Rhombomeres r2 and r3 contain neurons supplying all skin and dura innervated by V2 and V3. This is the embryologic basis for the treatment of migraine headaches with Botox® injected into peripheral trigger points.
Design of surgical incisions for cleft repair follows this embryology. The boundary between these two fields is sharply demarcated within the naris. The skin of the anterior columella and philtrum is thus a p5 derivative. This skin is supplied by terminal branches of the internal carotid artery, the anterior ethmoid arteries and innervated by V1.  The skin making up the floor of the nose has a different origin. It extends from the base of the columella laterally and makes contact with the alar base. The medial (terminal) branch of the sphenopalatine artery innervates this skin. The innervation is from V2 and is shared with the ipsilateral incisors. Continuity between the p5 skin of the lateral columella and the r2' skin of the nasal floor makes possible the elevation of a skin-cartilage flap containing the medial crura with long skin flaps. Many years ago Vissarianov described this flap as a means of secondary cleft reconstruction. -
Based on signals from the underlying premaxilla, the r2 alar base produces a tongue of tissue that makes contact with the r1 lateral prolabium. This skin bridge sets up the floor of the nose. It also provides a mechanical platform by which mesenchyme from the lateral lip element can make contact with the p5 mesenchyme of the prolabium. Lip closure thus occurs. This process involves mesenchymal "flow" from the lateral lip element toward the prolabium. For this to occur, the epithelium covering the lateral lip element, the r2-r1 skin bridge and prolabium must undergo a genetically-induced breakdown. BMP-4 produced by the premaxillary field causes de-repression of Sonic hedgehog ( SHH ), a gene localized to these overlying epithelia. The protein product of SHH causes epithelial breakdown. Thus, absence or deficiency of an appropriate BMP-4 signal will lead to restricted expression of SHH , abnormal persistence of epithelium and failure of mesenchymal fusion. ,
The volume of the PMx determines whether or not lip closure can occur. First, a small premaxillar field manufactures small amounts of BMP-4. The amount of BMP-4 produced is critical to produce the epithelial breakdown necessary to permit mesenchymal merger. Second, when the premaxilla is too small, the physical distance between it and maxilla will exceed a critical dimension. Epithelial bridge formation between the alar base and the prolabium cannot occur. If this critical distance exists at the level of the incisive foramen, a cleft of the secondary palate will form. This is because the horizontal repositioning of the palatal shelf from the maxilla must make contact with the vomer just posterior to the incisive foramen. The process is just like a zipper. If initial contact is not made, fusion of the palatal shelf to the vomer cannot take place. Even if initial contact is made, a secondary palatal cleft can still result due to displacement of the vomer away from the midline. The vomer can become warped by the inequality of growth forces on either side of the cleft. Thus, the zipper may get started anteriorly but, as the process proceeds posteriorly, when it encounters the deviated vomer, a palatal cleft will open up. 
Reconstruction of the premaxilla
Cleft surgery that does not reconstruct the missing PMx does not solve the biologic problem. The pediatric face is a set of Lego® pieces, all growing over time. If one piece of the set is missing, with subsequent growth the remaining pieces will collapse into the deficiency site. Only when all the Lego pieces are in place can harmonious facial development occur.
The premaxillary developmental field consists of lining (present in its totality) and mesenchyme (missing). Lining can be recreated by: (1) subperiosteal dissection of the primary palate (r2' premaxilla and r2 maxilla); (2) subperiochondrial dissection of the p6 septum (sufficient to reduce the septum into the midline; (3) subperiosteal elevation of r2' vomer to close the nasal floor (sufficient to reduce the septum into the midline); and (4) subperiosteal rescue of the r2' nasal floor skin from the lateral prolabium. When these flaps are elevated and sutured, the primary palate becomes a box lined with neural crest stem cells, all of which carry membrane-bound BMP receptors. ,,
Mesenchymal replacement can be undertaken using two basic mechanisms. Autogenous bone graft from rib or hip provides mesenchymal cells originating from paraxial mesoderm. Incorporation of the graft into the primary palate occurs by osteoconduction . Implantation of recombinant human bone morphogenetic protein-2 (rhBMP-2) in combination with an activated collagen sponge (ACS) results in morphogen-based recruitment of stem cells from the environment into the sponge. Stem cell concentration and differentiation into osteoblasts results in the formation of bone native to the site. This process is known as osteoinduction .  Extensive pre-clinical work by Boyne demonstrated the ability of rhBMP-2 to form membranous bone, including reconstitution of surgically-created cleft palate defects in primates. , Studies demonstrating efficacy in maxillary lift combined with absence of donor site morbidity resulted in FDA approval for this indication. 
The technique of facial bone reconstruction using rhBMP-2/ACS implantation is known as in situ osteogenesis (ISO) and has been previously reported by this author and co-workers.  Resynthesis of a 12 cm mandibular defect in a 9 year old demonstrated the ability of ISO function effectively outside the range of blood supply associated with a critical-size defect.  BMP-2 mediated osteoinduction is accompanied by extensive recruitment of blood vessels from local environment.  For this reason selected bone defects can be resynthesized using ISO alone, without recourse to microsurgical tissue transplantation.
Distraction techniques have been successfully applied to ISO-regenerated bone. In a number #7 lateral cleft with a foreshortened mandibular body and absent ramus, distraction of the recipient periosteal chamber in a posterior and superior direction permitted synthesis of 3.5 cm of mandibular ramus with eventual articulation with the skull base via a pseudoarthrosis.  Distraction-assisted in situ osteogenesis (DISO) will be an alternative treatment to rib grafts in the reconstruction of the Pruzansky III mandibular defects in craniofacial microsomia. The bone produced will be membranous. The dissection is less problematic. There is avoidance of unpredictable growth of the rib graft outside the natural periosteal environment. Finally, the chest wall donor site is obviated. Histology of ISO-produced membranous bone is indistinguishable from that of the recipient site in both mandible and maxilla.
Alveolar clefts are lined with mucoperiosteum containing neural crest stem cells. Blood supply is excellent and the dimensions modest. For this reason, this author and co-workers reported 50 alveolar clefts successfully treated with rhBMP-2.  Precise surgical technique of implant placement and of soft tissue closure was emphasized in the current series of 200 cleft sites.  Long-term outcome of 43 cleft sites was assessed at one year post grafting with 3-D CT. The series was comprised of 23 unilateral clefts (6 primary and 17 secondary) and 12 bilateral clefts (3 primary and 9 secondary). Complete transverse fill (unification of the dental arch) was achieved in all cases. Vertical fill was improved to 75-100% when an inert bulking agent (tricalcium phosphate) was included. For these reasons, primary cleft repair using developmental field reassignment technique can ideally be combined with ISO to achieve primary unification of the dental arch without donor site morbidity.
Philosophy of Developmental Field Reassignment (DFR): the 4 D's
Faulty embryogenesis of the premaxillary field affects the position and shape of surrounding fields in four distinct ways: Division of force vectors causes unequal growth, Deficiency -related collapse of partner fields into the void, Displacement of partner fields away from the midline. Distortion over time of structures such as the septum. DFR repair addresses these "4 D's" in reverse order. All distortions should be corrected. The displaced soft tissues if the face should be centralized into the midline. The deficiency site should be restored, using neighboring soft tissues and bone graft or ISO; and (4) division should be closed via unification of soft tissues. 
The execution of DFR surgery is based upon the concepts of Sotereanos.  Its guiding principles are five. , (1) Correct all pathologic states in the first operation (as above). (2) Respect embryonic developmental planes during dissection, ie. subperiosteal release of the soft tissue envelope for a tension-free closure. (3) Conserve blood supply to the periosteum, safeguarding it for future membranous bone synthesis. Fourth, align and unite the dental arch into a normal relationship. Fifth, reassign all developmental fields into correct relationships with each other. When properly executed, these principles result in the restoration of the functional matrix. When all the bone-forming soft tissue fields are spatially correct, all force vectors exerted upon bone will be correctly aligned. Subsequent growth of the face is directed back toward the normal. ,,
A new algorithm: Is lateral nasal wall deficiency relative or absolute?
In keeping with Victor Veau, cleft repair is a constant exploration of nature's experiments. Developmental field reassignment surgery is deliberately designed to address a tissue deficit of the lateral nasal wall and alveolus, the product of a congenitally small premaxillary field. That the lateral crus be entrapped cannot be in doubt. Its release into a normal position occasions a triangular tissue gap that must be filled. Proper airway reconstruction is the name of the game. At the same time, the alveolar cleft (a six-sided box) must be reconstructed. This requires flap coverage for its nasal surface (the "top" of the box). Can these two goals be accomplished with the same tissue?
A skin graft alone (particularly anterior auricular skin and cartilage) will effectively support lateral alar crus advancement. It cannot provide vascular coverage for the alveolar cleft "roof." At first blush, the LLC-NPP flap would seem big enough to accomplish both goals. After 7 years of work with this technique this author has concluded that paring from the prolabium is not sufficient. Premaxillary field soft tissues are not only relatively deficient from the lateral nasal wall: an absolute tissue deficit exists. Use of the premaxillary tissue mismatched to the prolabium is not always sufficient to solve the problem. The new algorithm of DFR surgery now calls for optional composite grafting of the lateral wall, followed by elevation and transposition of the LCC-NPP flap across the nasal surface of the alveolar cleft. The decision to graft is based upon the relative size of the defect versus that of the flap.
| » Surgical Technique of DFR|| |
This operation consists of markings, a five-step dissection sequence and a five-step closure.
Preparation and marking
The DFR operation is comprehensive, providing simultaneous correction of the lip, nose and primary palate. A more extensive dissection is required. The operation takes longer to perform (about 3-4 hours for a unilateral cleft). For this reason, patients come to surgery at 4-6 months of age. Prior to operation the dental arch is prepared with splinting (a form of infant presurgical orthopedics); this is begun as early as two weeks after birth. The emphasis of presurgical splinting is: (1) promoting anterior growth of the retro-displaced cleft maxilla; and (2) maintenance of the space in the alveolar cleft. If satisfactory maxillary shelf repositioning does not result by age 3 months, a lip adhesion procedure is carried out. When satisfactory dental arch correction is achieved, DFR repair is performed. This usually occurs 3-4 months after the lip adhesion.
The patient receives antibiotics and corticosteroids for swelling. A central V2 block is performed at the pterygopalatine fossa using 0.25% Marcaine® (bupivicaine hydrochloride). Approximately 3-5 cc per side is sufficient (maximum dose for 0.25% bupivicaine being 1 cc/kg). At the end of the case, the central block is reinforced by blocking the infraorbital nerves with bupivicaine, 1-2 cc per side. The child returns to recovery pain-free. The initial block ensures that substance P (a critical mediator in the pain cascade released in response to surgical trauma) will not be produced. In the absence of substance P the entire postop pain response is altered.
The surgical fields of the unilateral cleft are defined as follows. FIGURE (In neuromeric terms: A =p6 (red) + p5 (turquoise), B = r2' (pale gold), C is r2 (yellow) and D is r2 + r4 (yellow + orange). The prolabium is divided into two zones. Zone A is the true philtrum; it measures the width of the columella. It contains paired anterior ethmoid arteries (the terminal branches of the internal carotid system) and paired terminal branches of V1. Thus zone A contains two neurovascular developmental fields. Zone B of the prolabium is all tissue lateral to the philtrum: the non-philtral probium (NPP). The nasopalatine (medial sphenopalatine) artery, the terminal branch from the internal maxillary system, supplies the NPP. Bilateral clefts have two NPP zones. The cleft-side alar base is zone C. The lateral lip element below the alar groove is zone D. It contains a sphincter layer, the deep orbicularis oris (DOO) and the dilator layer, the superficial orbicularis oris (SOO) muscles. DOO develops in conjunction with the oral mucosa while SOO develops with the skin. A layer of fat conveniently separates these two layers. ,
The true destination of NPP soft tissue is in the nasal floor and the lateral nasal wall. NPP represents the soft tissue envelope corresponding to the lateral zone of the premaxilla, PMxB + PMxF. Nasal floor soft tissue corresponds to the lateral incisor field, PMxB. Soft tissue of the lateral nasal wall between the inferior turbinate and the lateral crus corresponds to the frontal process field, PMxF. Lateral wall soft tissue corresponds to the frontal process field, PMxF. Nasal floor soft tissue corresponds to the lateral incisor field, PMxB. The rationale of DFR surgery is to reassign the misplaced fields of the premaxilla into correct position. When the NPP flap is combined with the reflected mucoperiosteum from the margins of the bony cleft a surgical "pocket" is created. This site is subsequently filled with bone graft or a bone-producing cytokine (rhBMP-2) such that missing premaxillary fields are re-synthesized.
The NPP flap functions just like a "boxtop" to cover the alveolar cleft. It is almost always adequate. I am convinced that the "secret identity" of NPP is the original PMxB Because of the cleft this tissue is "shipwrecked" on the premaxilla. Unfortunately, NPP is frequently insufficient to replace PMxF. Every time the lateral crus is released and advanced a soft tissue defect appears. The defect represents missing PMxF. In my clinical experience this is almost always the size of the auricular cymba. A composite graft of anterior auricular skin and cartilage from the cymba is the most reliable replacement for this tissue. The cymba graft, in combination with the NPP flap, provides more than enough tissue to reconstruct the missing premaxilla.
Markings are carried out using a modification of the Millard system. (Paired anterior ethmoid neurovascular bundles define the true philtrum; this equals the width of the columella.). First to be marked is point 2, the junction of the normal philtral column with the white roll. The width of the columella at its base is then measured. This distance (usually 6-10 mm) is subsequently marked along the white roll medial to the philtral column as point 3. Distance 2-3 is the width of the philtrum; this contains the two anterior ethmoid arteries (approximately 4-6 mm apart). The midpoint of the philtrum (point 1 in the Millard system or nadir of cpid's bow) is irrelevant. Point 13, defined by the bulge of the footplate of the medial crus on the non-cleft side, marks the "shoulder" of the columella. The corresponding point 12 on the cleft side will be found displaced caudally and internally with respect to point 13. Total philtral height 2-13 should equal 3-12. Points 4 and 10 mark the midpoint of the alar bases on non-cleft and cleft sides, respectively.
On the lateral side, points 9 and 11 mark the terminus of skin within the nasal cavity. This is located just anterior to the inferior turbinate. Point 9 is the tip of flap D while point 11 is the tip of the future nostril sill C', so-called because it is in continuity with the alar base C. The dimensions of C' can be roughly mapped out by measuring the nostril sill on the non-cleft side. This is the distance from the midpoint of the non-cleft alar base (point 4) to the ipsilateral footplate point 13. Point 8 marks the natural transition of the white roll. Distance 8-9 should match 3-12.
Step 1. Lateral dissection: rescuing the nostril sill
The lateral lip is tensed with a single hook and the skin-mucosa margin is incised proceeding upward from 8 to 9, located just below and anterior to the inferior turbinate. From here the incision swings around laterally to 10, the midpoint of the alar base, (but not beyond it at this point). In this manner, the lateral lip flap D is separated from the alar base C. This step separates orbicularis from the nasalis. From point 9, a second, more internal , incision defines the triangular flap C'. The base width of C' can be deduced by measuring the dimensions on the non-cleft side.
From the lip a lateral vermilion flap (L) is pared off and dissected down to the alveolar cleft margin. Proper paring of L includes a small strip of lip skin 2-3 mm wide because it is "rolled-in," with an abnormal relationship to the underlying muscle. If the surgeon does not do this, the skin will pucker inward at the final closure. The L flap is the most optional tissue of the entire DFR operation. It is best brought medially and interposed between the nasal vestibular lining and the B flap. It can also be used as a free graft to the lateral nasal wall. No tissue is sacrificed in DFR.
With its vermilion stripped off, the lateral lip margin is now entered and the orbicularis is split into its deep (constrictor) and superficial (dilator) components. The deep orbicularis oris (DOO) is shaped like the letter J and separated from the superficial orbicularis oris (SOO) by a layer of fat. The caudal margin of the two muscles defines the white roll and contains the labial artery. 
Step 2. Medial dissection: the NPP-lateral columellar advancement flap
The medial margin of the cleft (zone B in our diagram) consists of prolabial tissue lateral to the true philtrum. This is in continuity with lateral columellar skin and the medial crus of the alar cartilage. Conveniently, the blood supply of these two units is a watershed permitting disection of a very long flap. The B flap has two parts: (1) a skin/mucosa flap of the non-philtral philtrum (NPP) and (2) a chondrocutaneous flap containing the lateral columellar skin and medial crus (LCC). B = NPP + LCC
B has a rich blood supply. The NPP component is irrigated by the nasopalatine artery. The LCC component gets supply from 2-3 lateral branches of the ipsilateral anterior ethmoid. In addition, vessels in continuity with the lateral nasal system from the facial artery run along the surface of the alar cartilage itself. Once elevated, B is surprisingly long, reaching all the way across the cleft to the lateral nasal rim. (Figure B) is inset into the gap created by advancement of the p5 lateral vestibular lining (and the lateral crus). This achieves two important surgical goals: (1) cephalic advancement of the medial crus with increased tip projection; and (2) soft tissue augmentation of the lateral nasal wall. Inset of B permits release of the tethered lateral vestibular lining. Repositioning the lateral crus now occurs in what would otherwise be a Y-V pattern without the Y-V closure . Natural alar cartilage anatomy is accomplished without pinching together an already-deficient lining.
The future cleft-sided philtral column is determined by understanding the embryology of the prolabium. The true philtrum lies between points 2 and 3 and consists of two fields (each supplied by a separate branch of the anterior ethmoid artery). The philtral dermis comes from p5 prosencephalic neural crest and is innervated by V1. The width of the philtrum is roughly equal to that of the columella, generally measuring 6-10 mm. Because the philtrum develops on top of the r2' premaxilla it receives additional blood supply from the terminal branches of the sphenopalatine artery (SPA) emerging at the level of the septopremaxillary junction. Thus, the philtrum has a dual blood supply . All remaining prolabial skin and mucosa (the NPP) is an r2' derivative, supplied by the nasopalatine artery and innervated by V2. The prolabium of a bilateral cleft thus consists of four developmental units based on embryology, blood supply and innervation: two central philtral A fields and two lateral B fields.
The embryology of the columella is as follows. The anterior centrally-located skin comes from p5 and contains the paired anterior ethmoid arteries. On either side of that central swatch, the columellar skin extends backwards toward the septum. Just beneath the lateral columellar pillars lie the medial crura of the lower lateral cartilages (LLC). Because these pillars serve as the biologic "template" for the cartilage to form, dissection of skin away from the crura is extremely difficult. The LLCs are thus p5 neural crest derivatives. The upper lateral cartilages lie above the p6 vestibular lining and are therefore of p6 neural crest origin.
Correction of the cleft lip nose requires releasing the alar cartilage from two points of entrapment.  The deficiency state of the premaxilla causes a mechanical deformation of perfectly normal p5 skin on both sides of the nasal introitus. In the lateral nasal wall, the r2' skin overlying the premaxillary frontal process (PMxF) is reduced or absent. Consequently, p5 skin containing the lateral crus of the alar cartilage is dragged down into this "sinkhole" located just in front of the inferior turbinate. The lateral crus is therefore flattened. On the medial side, the deficiency of the r2' premaxillary lateral incisor field (PmxB) creates a "sinkhole" into which the p5 lateral columellar skin is displaced. This displaces the medial crus of the alar cartilage downward and inward compared with the non-cleft side.
The LCC-NPP flap is elevated as follows. Under tension, the NPP skin is elevated in continuity with the lateral columellar wall. The lateral (internal) incision is made first, separating the skin from the mucosa all the way to the junction between the premaxilla and vomer. Although this incision will eventually be carried up in front of the septum, it is advisable to stop here. Tension can thus be maintained on the NPP while it is separated from the philtrum proper.
Next, a second incision, parallel to the first, ascends from point 3 to the base of the columella. It then extends into the nose along the lateral sidewall of the columella. Previously I would continue this incision along the side of the columella (just anterior to the edge of the medial crus) and thence bring it directly into the nose as an infracartilagenous incision. This design made me concerned about possible scar contracture. At the suggestion of Dr. John Reinisch, I began to break up the incision by elevating a medially-based rectangular skin flap occupying the caudal half of the columella. Just beneath the Reinisch flap lies the footplate of the lateral nasal cartilage. At its cephalic margin, the incision is resumed along lateral columella. At the level of the intermediate crus the incision transitions beneath the soft triangle into a standard infracartilagenous incision and is then continued all the way to the piriform margin.
The anterior incision of the NPP-LCC flap is not sufficient to advance it. A second parallel incision is required. FIGURE It starts in the membranous septum about halfway up. It is then carried downward to the vomer and then re-directed forward to meet the lateral margin of the B flap. Note that the membranous septal incision provides immediate exposure of the cartilaginous septum.
The NPP-LCC flap now resembles a long boot, shaped like Italy. The "toe" of the boot extends along the prolabial margin. Beneath the "heel" of the boot lies the footplate of the alar cartilage. This landmark corresponds to point 12. Grasping the heel of B provides instant access to the medial aspect of the medial crus. This is a safe plane permitting dissection of the alar cartilage right into the nasal tip, with the following caveat. The B flap gets blood supply from the nasopalatine artery via 2-3 branches emerging at the junction of the premaxilla and vomer. Gentle spreading along the medial border of the cartilage will reveal these branches and preserve them. Additional blood supply descends along the skin. Formerly I would elevate these NPP-LCC flaps completely, never having an issue with ischemia. Now however, I think it prudent to preserve the nasopalatine branches when possible. This type of blunt dissection is more than sufficient to advance the nasal tip.
The septum is now dissected out and freed from the maxillary crest until it sits passively in the midline. As growth proceeds the centralized septum will no longer constitute an abnormal force vector tethering the nasal tip.
Step 3. Nasal dissection: "open-closed" rhinoplasty
At this juncture, a standard infracartilagenous incision is made. This incision is brought all the way to the piriform rim, following the natural fold between the nasal skin and the vestibular skin. The caudal extent of this incision terminates at the internal border of the triangular nostril sill flap B'. The exposure gained via the lateral columella-infracartilaginous incision allows a complete dissection of the dorsal nasal skin envelope as described by McComb.  The success of the McComb dissection is really an embryonic field separation. The deep layer comprises the p6 vestibular epithelium and p6 neural crest upper lateral cartilages, the blood supply to which comes from below via the ICA. The superficial layer is p5 nasal skin and the lower lateral nasal cartilages, the blood supply to which is also of ICA derivation. Interposed between these two layers, like a sandwich, is the SMAS layer of facial muscles derived from the 2 nd pharyngeal arch (r4-r5). The myoblasts come from somitomere 4. The nasal muscles are compartmentalized by r4 neural crest fascia (the SMAS). The blood supply to this intermediate 2 nd arch layer is from the facial artery (ECA). This mesenchyme provides an additional source of blood supply to the alar cartilages. Careful dissection of the lateral columellar walls from the columella discloses vessels running along the medial surface of the medial crura. These anastomose the lateral nasal vessels with the nasopalatine vessels.
Along the lateral nasal rim the dissection is carried right down to the piriform margin. The proper plane for separation is achieved by hugging the surface of the cartilage. The overlying r4 SMAS layer and the p5 skin layer are left in anatomic continuity. Very little bleeding is occasioned by this approach. As one reaches the piriform rim the periosteum is incised and stripped vertically. Subperiosteal elevation of the soft tissues lateral to the piriform rim preserves the facial artery arcade. The nasal skin envelope is then liberated cephalically all the way to the nasal bones.
Despite these maneuvers, the vestibular lining is still tight! The media crus has been completely released and advanced into the nasal tip but the lateral crus being remains splayed out and tethered. Releasing the lateral crus from the vestibular lining has been advocated in the past. This is technically difficult because it violates the embryology (recall that the alar cartilage arises as a neural crest response to a pattern embedded in the vestibular epithilium). The size and shape of the cleft-side alar cartilage has been demonstrated to be normal compared with the non-cleft side.  Consequently, the p5 vestibular lining "program" must be normal as well. Studies at UNC demonstrate that the overall surface area of the repaired cleft nostril is reduced by 30%.  The site and dimensions of the deficit correspond to the soft tissues of PMxF.
The soft tissues of the alveolar cleft PMxB can be found on the prolabium. The LCC-NPP flap effectively provides a vascularized "roof" for the bony cleft. Lateral nasal wall release "uncovers" a defect that frequently exceed the reach of LCC-NPP. The release begins at the inferior turbinate and continues along the junction of vestibular and nasal skin. At the apex of the lateral crus it becomes V-shaped. Due to the prior McComb dissection the lateral crus is advanced cephalically into symmetry with the normal side. The internal nasal valves will appear equal. Left behind is a raw spot roughly the size of the ear cymba. A composite graft using anterior skin and cartilage from the cymba provides a sturdy reconstruction for the airway. The LCC-NPP flap, inset without tension across the alveolus, comes to rest at the foot of the cymba graft.
Exit the turbinate flap
As the DFR evolved, I employed various solutions to patch the defect created by the vestibular release. One of these solutions made use of an anteriorly-based inferior turbinate flap as described by Noordhoff. , Although this tissue worked adequately, I had several reservations about it: (1) the dissection is subtle and difficult to teach; (2) the tissue type is distinct and not native to the nasal rim; and (3) healing of the donor site can be accompanied by crusting and bleeding; and (4) it did not make embryologic sense. Proper dissection and inset of flap NPP combined with use of a composite for FTSG made the turbinate flap unnecessary. The key to getting the most out of NPP is to include the prolabial vermilion with the skin.
Step 4. Intraoral dissection: sliding sulcus S flap
The rationale and design of the sliding sulcus mucoperiosteal flap stem directly from pioneering work at the University of Pittsburgh by Sotereanos.  This technique involves a gingival release on the cleft side carried out from the alveolar cleft to the buttress.  This permits wide subperiosteal dissection over the entire face of the maxilla below the infraorbital foramen as described by Delaire. , A 45 degree backcut up the buttress is performed. The attached gingival is released. The flap is covered on its undersurface by a sheet of periosteum, rendering it rather stiff. The Sotereanos maneuver is a mobilization of the S flap using a counter-incision in the periosteal sheet itself parallel to the gum line. The counter-incision, located half-way from the gingival margin to the infraorbital foramen, is made by just scoring through periosteum. It extends from the piriform margin straight lateral to the buttress. At this point it joins up with the previous backcut. These two incisions make a right angle: the periosteal counter-incision is transverse across the maxilla and the buttress incision is vertical. The combination of these two incisions releases the lower mucoperiosteal flap S from the upper mucoperiosteum attached to the orbit.
When S is released, it advances mesially about two tooth units. The margin of S that was freed from the lateral border of the cleft now extends across the alveolar cleft without tension and is sutured to the mucoperiosteum of the premaxilla. In this way, like a sliding door, the S flap seals up the anterior aspect of the alveolar cleft.
On the noncleft side, a similar subperiosteal dissection is done without recourse to a gingival release. A large bipedicle flap is created, permitting centralization of the previously lateralized soft tissue envelope. Note that when cleft lip occurs without an alveolar cleft. bilateral S flaps, elevated without gingival release , permit proper tension-free centralization of midface soft tissues.
The S flap then comes in two varieties, each of which has a specific benefit for the patient. The simple S flap is elevated in the subperiosteal plane across the entire face of the maxilla between two points of vertical release: the alveolar cleft and lateral wall of the buttress. It remains attached to the teeth. A gingival releasing incision is not required. By virtue of its ability to centralize the soft tissues of the entire midface, the simple S flap fulfills two important functions: (1) the creation of an esthetic soft tissue "drape"; and (2) the biologically active osteogeneic potential of the mucoperiosteal stem cells is transferred forward, thus enabling bone deposition to take place in centrically for better facial skeletal projection. The drawback of the simple S flap is that it cannot produce a mucoperiosteal flap to cover over the anterior aspect of the alveolar cleft. In such cases, a rectangular mucoperiosteal flap can be harvested immediately lateral to the alveolar cleft and brought over for closure. This design works for narrow clefts.
The compound S flap (as described by Sotereanos) implies a full gingival releasing incision all the way from the alveolar cleft back to the tuberosity. This involves additional operating time. Gingival release from an alveolus containing erupted primary or secondary dentition is technically simple. The bone is solid and the surgical plane easy to follow. Infants are a different matter. Prior to the 4 th month of life maxillary and alveolar bone is too soft to work with. DFR should be performed between ages 4-6 months. Dissection is facilitated using loupe magnification, a #15C or Beaver blade, an amalgam packer and Molt (#9) or Louisville periosteal elevators. A compound S flap can be mobilized two full tooth widths. The main advantage of the compound S flap is the dramatic increase in mucoperiosteal flap length it provides. This is generally the width of two dental units. The anterior aspect of the alveolar cleft is thus covered with stem cell-containing mucoperiosteum. Disadvantages of S flap relate to the temporary anatomic disruption of periocoronal attachments.
The S flap has distinct indications for use in primary and secondary cleft repair. These are of vital importance, particularly when one must operate under conditions that are less than ideal. In a primary cleft, whenever possible, pre-surgical orthodontic control of the arch should be accomplished. This may also involve a lip adhesion. The entire effort is directed toward normalizing arch dimensions. In these cases, compound S flaps are not required. Closing the anterior wall of the alveolar cleft can be accomplished with a rectangular superiorly-based mucoperiosteal flap elevated directly off the lateral incisor. The space will fill in nicely. If one desires, the flap donor site can be grafted with Alloderm® .
Limited resources and difficult logistics in economically developing countries force surgeons treating cleft patients to adopt differentand creative strategies, particularly when the alveolar cleft is wide. These children often have no access to pre-surgical orthopedics. They may never receive orthodontics. They do possess a ready source of bone graft from the rib. In such cases, achievement of a consolidated arch and elimination of oronasal fistula are reasonable goals for primary cleft surgery. The two key factors for success in these patients are: (1) ability to make a simple acrylic splint and secure it--2 mm screws work perfectly well; and (2) meticulous dissection of the alveolar cleft with attention to its vascular anatomy.
If a dental splint can be fabricated, lip adhesion can narrow the cleft down to dimensions permitting simultaneous DFR repair and alveolar grafting with a simple S flap. When this is not possible, a compound S flap will close the gap. In some very wide primary unilateral clefts I have used bilateral compound S flaps to successfully close gaps of 14-16 mm. Of course, this maneuver disconnects the frenulum from the midline. Postoperative remodeling of the alveolus reestablishes symmetry. In such cases, arch stability justifies the extensive intraoral dissection.
Secondary cleft reconstruction follows the same rules but with the proviso that gingival release incisions are fast and easy to accomplish. The floor of the alveolar reconstruction has already been provided by a pre-existent palatal repair. When orthodontic capability exists, such clefts should first be expanded to the original width prior to grafting. But when the patient has no access to ortho, S flap alveolar reconstruction can be combined with DFR to produce a permanent, aesthetically pleasing and physiologic result in one stage.
Step 5. Dissection of the primary palate
Beginning with the lateral nasal wall, the mucoperiosteum is elevated off the maxillary alveolar bone. The palatal margin is undermined as well. This results in a large mucoperiosteal envelope. This must be separated into nasal and oral components. This is done at the level of the "shoulder" of the alveolus. The incision can be extended backward a few mm along the edge of the palatal shelf. On the medial side, the septal mucoperichondrium and the vomerine mucoperiosteum are elevated in continuity. The surgeon then proceeds forward to elevate the mucoperiosteum off the premaxilla. Separation of the envelope occurs at the "shoulder" of the premaxilla. A vomer mucoperiosteal flap is elevated and sutured to the nasal mucosa of the cleft maxilla.
A six-sided "box" is created. The medial and lateral walls are raw bone. The floor is oral mucoperiosteum. The anterior wall is the sliding sulcus mucoperiosteum. The roof is the nasal mucoperiosteum + the NPP flap. The posterior wall is the vomer flap. These flaps are loaded with undifferentiated mesenchymal stem cells (MSC). The cambium layer of the periosteum is especially rich in stem cells. All neural crest MSCs contain membrane-bound receptors for BMP. This pocket is now ready to be filled with rib graft (iliac crest graft in older children) or with an rhBMP-2/ACS implant.
Alveolar reconstruction, both primary and secondary, demands precise knowledge regarding the vascular anatomy of mucoperiosteum bordering the alveolar cleft. This is a neglected area of cleft surgery. The blood supply to the premaxilla is the most complex (and potentially treacherous). It is derived from three sources. (1) Premaxilla is an r2' derivative. Like all mesenchymal structures caudal to r1, premaxilla is supplied from the external carotid. The vascular axis of the vomer and premaxilla is the medial sphenopalatine (nasopalatine) artery. PMx mucoperiosteum supplied by the SPA envelops the bone from above-downward, like a cloak thrown over a chair. (2) From below, PMx mucoperiosteum is continuous with that of the r2 maxillary palatal shelf (MxP), the arterial axis of which is the external carotid-based greater palatine artery. (3) a second anastomosis exists within the mucoperiosteum covering the anterior superior aspect of the premaxilla. This zone is in contact with the soft tissues of the prolabium, a p5 derivative supplied by the internal carotid-based anterior ethmoid arteries.
Premaxillary survival depends upon maintenance of sufficient contact between the bone and its enveloping mucoperiosteum. Alveolar cleft reconstruction cannot be accomplished without stripping away the vascular coverage of PMx, to a greater or lesser degree. Thus, premaxillary vascular anatomy dictates the direction in which the mucoperiosteum must be elevated. Surgical strategy for this maneuver is a direct consequence of the cleft type.
In unilateral palate clefts, the medial alveolar cleft margin is a continuous zone of external carotid from the SPA to the GPA. This permits downward dissection of the mucoperiosteum from the shoulder of the premaxilla or upward dissection from the oral margin. The premaxilla remains alive, based on its other sources. In bilateral palate clefts, no such palatal anastomosis exists. The mucoperiosteum can only be reflected upward. If this is done on both sides and if the prolabium is simultaneously elevated, premaxillary necrosis can ensue.
This principle is exactly the same as in prolabial death reported by Millard, a combination of bilateral rotation backcuts (destroying the anterior ethmoid supply to the philtrum) and elevation away from the underlying premaxilla.  In bilateral cleft lip and palate, the anatomy of both prolabium and premaxilla consists of paired angiosomes. One can disrupt one or the other with impunity, but not both.
Fortunately, the robust blood supply of the maxilla comes to our rescue. Injection studies demonstrate the maxilla to be supplied by the pharyngeal branch of the facial artery and facial branch of the ascending pharyngeal artery.  It is independent of its mucoperiosteal cover. Meticulous dissection of the lateral alveolar cleft creates a continuous mucoperiosteal sheet uniting the lateral nasal wall with that of the hard palate. This contains a mirror-image external carotid anastomosis, this time between the medial sphenopalatine and greater palatine arteries. This can be stripped downward with impunity. It is a long flap, extending from just in front of the inferior turbinate all the way to the hard palate. The lateral alveolar mucoperiosteal flap is of sufficient size to close the entire floor of the alveolar cleft.
What then should be done for children BCL(P)? Must one sacrifice arch reconstruction to facial aesthetics or vice versa? The answer (no!) lies in a staged approach with careful dissection. Primary lip reconstruction must not violate the tissue plane separating the prolabium and mucoperiosteum. Even in a bilateral dissection, 50% of the premaxilla will remain perfused by both SPA and AEA vessels. Lateral nasal wall reconstruction with the cymba composite graft is appropriate, Liberation of the medial crura must be done without entering the territory of the medial sphenopalatine artery. If this appears difficult, nasal elevation and lip revision may have to be staged later. Alveolar cleft closure should be done by elevating the premaxillary mucoperiosteal flap upward and the maxillary mucoperiosteal flap downward.
Wide bilateral clefts require alternative strategies. Two principles to avoid trouble in these patients are: (1) meticulous primary closure of the nasal floor; and (2) delayed grafting of the alveolar cleft. A wide alveolar "roof" defect requires big nasal floor flaps. These are best harvested at the primary surgery, before other tissues get in the way. The key point here is to not cut the lateral mucoperiosteal flap until the medial flap is harvested from the premaxilla. Once the dimensions of medial flap are known, a superiorly-based lateral flap of sufficient size is elevated to complete the anterior nasal closure. At this point we are still left with a wide alveolar "floor" defect and not enough tissue to close it. What does one do? One waits it out, using a palatal acrylic splint (secured with pins or 2 mm screws) is maintain the transverse arch dimensions and prevent collapse.
6 months later, at palatoplasty, the long mucoperiosteal flaps elevated from the hard palate provided ideal coverage for the alveolar cleft "floor." Bilateral compound S flaps are raised. These will adequately seal off the anterior face of the alveolar cleft. The graft (using rhBMP-2 or rib) is placed. Once again, care must be taken to prevent transverse collapse until the bone graft has hardened. Intraoral splinting is thus continued for another 2-3 months.
Secondary grafting in bilateral cases follows the same principles. The vascular anatomy may have been altered (primary surgery may have elevated the prolabium away from the premaxilla). In such cases some degree of revascularization to the premaxilla occurs over time. Nonetheless, dissection of the premaxillary mucoperiosteum must be sparing. All that is required is elevation of sufficient tissue to close the soft tissue defect and place the graft. Nasal tip elevation can be safely accomplished using DFR technique and cymba grafts. The key point is to preserve soft tissue continuity between the vomer, the septal mucoperichondrium and the premaxilla: this is the site of entry of blood supply into the premaxilla.
Step 1. Elevation of the nasal tip
The lateral crus is stabilized into symmetry with the normal side using a transcutaneous 4-0 chromic mattress stitch. The V-shaped limbs of the releasing incision are closed. The donor site is reconstructed using a full-thickness retroauricular graft or a composite graft of anterior ear skin and cartilage. By trial and error I prefer the cymba as the donor site.
The tip is positioned anatomically using the Cronin nasal retractor (Padgett Instruments). The medial crura are battened together with 5-0 PDS. Suture suspension and modification of the alar cartilage can be readily executed by means of the open-closed approach as per the surgeon's preference. The author approach is predicated on the establishment of normal field relationships. Wide dissection of the nasal soft tissues combined with release from their "piriform prison" allows all fields to be passively held in position by a nasal stent inserted at the conclusion of the procedure. The author finds the Koken-type silicon stents (Porex Corp, Newman, Georgia) easy to use. The stent is placed at the end of the surgery. Closure of the nostril incision starts at the intermediate crus working medially down to the Reinisch flap. One then proceeds laterally to the margin of the lateral nasal wall. This involves placing 2-3 sutures over 5 mm.
The remainder of the lateral nasal incision will be filled by inset of the B flap. Reconstruction of the frontal process of the premaxilla adds the missing tissue to the lateral nasal wall (see below). This frequently involves placement of a composite for full thickness graft into the defect. The medial crural complex the then elevated with respect to the septum with 4-0 vicryl. This also closes the membranous septum counter-incision.
Step 2. Soft tissue reconstruction of the premaxilla
The roof of the missing premaxillary field is reconstructed based on meticulous closure of the nasal floor. This is carried out using a mouth gag, starting anteriorly at the incisive foramen. Because the space is tight using 5-0 Vicryl® on a small P-2 needle is helpful. The medial vomerine and lateral nasal mucoperiosteal flaps are closed all the way posterior to the end of the vomer flap. This provides correct orientation for inset of the B flap. The posterior margin of B is sutured from medial to lateral along the newly-united nostril floor. The tip of B is eventually inset into the donor site of the lateral crural advancement flap. Next, the alar nasal skin flap C is sutured to the anterior margin of B. The surface area of the lateral nasal wall is now restored.
The floor of the missing premaxillary field is reconstructed when the medial and lateral mucoperiosteal flaps harvested from the walls of the alveolar cleft are turned down into the mouth using 4-0 vicryl. The cleft side sliding sulcus flap S is advanced and secured using 4-0 Vicryl PS-2 around the dental units. Three or four such sutures will suffice. Optional suspension of the sulcus to maintain height can be done by passing a 3-0 vicryl up to the nasal floor and then back down into the sulcus as a mattress suture.
Step 3. Osseous reconstruction of the premaxilla: In situ osteogenesis (ISO)
The primary palate is reconstructed with Infuse® bone graft (Memphis, Tennessee Sofamor-Danek). The implant is prepared by soaking a Helistat® activated collagen sponge (ACS) (Plainsfield, N.J. Integra Life Siences) of preselected size with reconstituted rhBMP-2 applied uniformly over the ACS. The minimum time required for binding is 15 minutes; however, I usually wait 30 minutes. The Infuse® comes in three kit sizes 4.2 cc, 5.6 cc and 8.4 cc. These are used as follows: primary unilateral (small), primary bilateral (medium), secondary unilateral (medium) and secondary bilateral (large). The volume of the implant should match that of the defect. Mastergraft® tricalcium phosphate is useful for "bulking up" the implantation site. It acts passively, a space-occupying agent. One could use freeze-dried bone as well (the ideal bulking agent does not yet exist.) The Mastergraft is wrapped with the ACS collagen sponge like a fajita or sushi roll and placed into the defect. I like to place a vertical component into the alveolar walls and a horizontal component below the nasal floor. The pocket created by DFR dissection can be filled with iliac crest graft as well.
Step 4. Closure of the lip: Delaire concepts modified
Many years ago pioneering work by Delaire demonstrated that wide subperiosteal dissection yield significant aesthetic benefits of tissue draping and, at the same time, did not impair maxillary growth. , Indeed, when compared with supraperiosteal release, this approach is developmentally correct because it separates the osteogenic functional matrix (the periosteum and overlying soft tissues) from the product (the premaxilla and maxilla).
The suture sequence is as follows:
Delaire 1 : Centralization of the bilaterally displaced soft tissues is accomplished with 4-0 nylon sutures placed from below the flap to the paranasalis (levator) insertion into the superficial orbicularis oris (SOO). This insertion does not involve the alar base. It is generally located one fingerbreadth lateral to the alar base. The non-cleft side is sutured first to below the anterior nasal spine. This suture serves as a reference point for a similar suture from the cleft side.
Suspension of the DOO : This suture sest the depth of the sulcus on the cleft side. At the most cephalic margin the DOO is suspended from the septum with a 4-0 PDS.
Delaire 2 : The purpose of this suture is to control the height and curvature of the cleft side nostril. This is accomplished by connection of the nasalis to the anterior nasal spine However this step is optional at this point. If performed, Care should be taken to not tighten this and inadvertently narrow the nostril. When the remaining sutures are placed a decision can be made if a Delaire 2 is warranted. I usually decide upon a Delaire 2 at the end of the case to accentuate curvature
Orbicularis closure : Three or 4 sutures of 4-0 PDS are required for the DOO layer. The SOO layer is closed with 5-0 PDS making sure the loop is at the level of the dermal-epidermal junction.
Delaire 3 : The oblique head of SOO sets the aesthetic drape of the lip. The 5-0 PDS is obliquely passed upward from the cephalic edge through the base of the columella and then back down to the SOO as a mattress suture.
Step 5 Final adjustments: finessing the nostril floor
After closure of the lip, the alar base C is at time compressed medially by the movement of the neighboring D flap (lateral lip element). If so, the alar base must be translocated laterally. This can be accomplished by excision of a crescent of skin from the lateral lip element. C is then elevated and secured to the lateral via a buried 5-0 PDS suture. At times it is necessary to elevate the ala completely in order obtain sufficiently lateralization. p The tip of the nostril sill flap C' is now sutured just posterior to the columellar shoulder. Continuity between the columellar shoulder and the nostril sill is now reestablished. Perialar suturing with inverted 5-0 PDS sutures restores the alar crease very nicely.
A final caveat concerns the lining of the lateral nasal defect. Experience with this technique shows that prevention of post operative contraction is paramount. This requires placement of a Porex® silicon conformer sized to the patient. The stent should be sutured in place with 4-0 nylon for 8 weeks. It is important that the B flap not be sutured to the sidewall of C with any tension. If tension exists I have found it prudent to use a small pinch graft of retroauricular skin is used.
| » Discussion|| |
Whenever a surgical procedure for a congenital condition in a growing child leads to a predictable pattern of relapse over time two inescapable conclusions must be drawn: (1) the biologic rationale of the procedure is incorrect; and (2) the anatomic pattern of relapse points an accusatory finger at the pathology. DFR surgery represents a new form of thinking about clefts by identification and rearrangement of specific developmental fields. The primary pathology of cleft lip is hypoplasia or absence of the distal premaxilla. DFR is designed to reconstruct the missing premaxilla using osteoinductive technologies for stem cell concentration and differentiation such as rhBMP-2 or by conventional osteoconductive technique (bone grafting). DFR incisions are made on the basis of vascular supply and embryology, not geometric manipulation. For this reason the design of DFR presents a series of specific solutions that speak to problems in cleft surgery hitherto inadequately addressed: distortion of the nasal envelope and septum, displacement of soft tissues away from the midline, the entrapped position of the medial crus and the "hidden" nostril sill. DFR treats mechanisms, not appearance.
The extraoral design of DFR is invariable; it is the same for both unilateral and bilateral clefts. Developmental fields are recognized, separated and rearranged into correct anatomic relationships. The intraoral design varies with cleft type. If no alveolar cleft is present, subperiosteal release and zygomaticomaxillary buttress backcut are sufficient to achieve centralization of the soft tissue envelope. Gingival release is not necessary. If a primary palate cleft exists, a complete sliding sulcus dissection with gingival release will allow for mesial translocation and coverage of the alveolar cleft.
DFR includes a straight-line repair, resulting in an anatomically correct philtral column. The distorted and seemingly foreshortened philtrum in clefts is familiar to all plastic surgeons. For many years, the rotation-advancement technique has been used to reorient the philtrum. So why is philtral rotation not required in DFR? The appearance of the philtrum in clefts stems from the overall malposition of the soft tissue in a falsely lateralized state. Thus, the philtrum and columella appear to be short, but are actually displaced into the nasal tip. DFR achieves de-rotation of the philtrum and columella by the following: (1) field separation; and (2) subperiosteal mobilization. The entire soft tissue complex in a unilateral cleft is laterally displaced "around the corner" of the premaxilla. When mobilized correctly off the bone it is brought forward around the curve of the bone and drop right into place.
The membranous bones of the maxilla and premaxilla are just a product of the soft tissue functional matrix that surrounds them. As the soft tissue grows, new bone is deposited and old bone resorbed according to the mechanical forces placed on the bone. The purpose of subperiosteal centralization in DFR is to change the biologic relationship of the bone product to the functional matrix. Over time, the former will adapt to the latter in its new, centralized position. Appreciation of the subperiosteal plane as the correct approach to surgically separate the functional matrix from its product (bone) ensures tension-free release and accuracy of muscle repair (individual muscle units can be identified by tugging on them from below). Thirty years of work by Delaire and others confirm the clinical accuracy and safety of this approach. Readers are also advised to study the pioneering work on primary bone grafting done by Rosenstein. , This 30 year follow-up of dental development in patients treated with careful presurgical orthodontics and rib grafting demonstrates clearly the safety and advantages of gaining control over the alveolar cleft at primary surgery.
The concepts and techniques of DFR are applicable to all forms of clefts. Because it separates out osseous pathology from soft tissue pathology, DFR is a true "cut as you go" technique. In secondary cleft surgery, DFR is capable of rescuing previously violated fields and reuniting them into their correct functional relationships. Sliding sulcus flaps function as elastic flaps as defined by Goldstein , and can readily be brought into the midline. Abbe flaps are virtually unnecessary, even in salvage cases involving total loss of the philtrum.
DFR is a practical application of developmental anatomy. It provides a means to analyze membranous bone formation and periosteal physiology. It embodies concepts of facial growth central to orthodontics. It makes use of neuromeric theory to map out and manipulate developmental fields. Proper implementation and study of cleft repair using DFR will provide a forum for dialog along craniofacial surgeons orthodontists and developmental biologists.
| » References|| |
|1.||Veau V, Recamier J. Bec-de Lievre: Formes Cliniques-Chirugie. Masson et Cie: Paris; 1938. |
|2.||His W. Beobachtungen zur Geschichte und Gamenbildung beim menschlichen Embryo. Kgl. Akad Wiss: 1901. p. 27. |
|3.||Carstens MH. Development of the facial midline. J Craniofac Surg 2002;13:129-90. [PUBMED] [FULLTEXT]|
|4.||Carstens MH. Neural tube programming and craniofacial cleft formation. I. The neuromeric organization of the head and neck. Eur J Paediatr Neurol 2004;8:181-210. |
|5.||Puelles L, Rubenstein JL. Development of the central nervous system. In : Rossant J, Tam PP. Mouse development: Patterning, Morphogenesis and organogenesis . Academic Press: San Diego; 2002. p. 37-54. |
|6.||Puelles L, Rubenstein JL. Forebrain gene expression domains and the evolving prosomeric model. Trends Neurosci 2003;26:469-76. [PUBMED] [FULLTEXT]|
|7.||Rubenstein JL, Puelles L. Development of the nervous system. In : Epstein CJ, Erickson RP, Wynshaw-Boris A, editors. Inborn errors of metabolism: The molecular basis of clinical disorders of morphogenesis. Oxford University Press: Oxford, England; 2004. p. 75-88. |
|8.||Carstens MG. Functional matrix cleft repair: A common strategy for unilateral and bilateral clefts. J Craniofac Surg 2000;11:437-69. |
|9.||Carstens MH. Function matrix cleft repair: Principles and techniques. Clin Plast Surg 2004;31:159-89. [PUBMED] [FULLTEXT]|
|10.||Lemire RJ, Cohen MM Jr, Beckwith JB, Kokich VG, Siebert JR. The facial features of holoprosencephaly in anencephalic human specimens. I. Historical review and associated malformations. Teratology 1981;23:297-303. |
|11.||Siebert JR, Kokich VG, Beckwith JB, Cohen MM Jr, Lemire RJ. The facial features of holoprosencephaly in anencephalic human specimens. II. Craniofacial anatomy. Teratology 1981;23:305-15. |
|12.||Liem KF Jr, Tremml G, Roelink H, Jessell TM. Dorsal differentiation of neural plate cells induced by BMP-mediated signals from epidermal ectoderm. Cell 1995;82:969-79. [PUBMED] [FULLTEXT]|
|13.||Liem, KF, Bemis WE, Walker WF, Grande L. Functional anatomy of the vertebrates: An evolutionary perspective, 4 rd ed. Thompson: Belmont, California; 2006. |
|14.||Barteczko K, Jacob M. A re-evaluation of the premaxillary bone in humans. Anat Embryol (Berlin) 2004;207:417-37. |
|15.||Mooney MP, Siegel IP, Kimes KR, Todhunter J. Premaxillary development in normal and cleft lip and palate human fetuses using three-dimensional computer reconstruction. Cleft Lip Palate J 1991;28:49-54. |
|16.||Carstens MH. The spectrum of minimal clefting: Process-oriented cleft management in the presence of an intact alveolus. J Craniofac Surg 2000;11:270-94. [PUBMED] |
|17.||Gong SG, Guo C. BMP4 gene is expressed at the putative site of fusion in the midfacial region. Differentiation 2003;71:228-36. [PUBMED] [FULLTEXT]|
|18.||Ashique AM, Fu K, Richman JM. Endogenous bone morphogenetic proteins regulate outgrowth and epithelial survival during avian lip fusion. Development 2002;129:4647-60. [PUBMED] [FULLTEXT]|
|19.||Lumsden A, Krumflauf R. Patterning the vertebrate neuraxis. Science 1996;274:1109-15 . |
|20.||Noden DM. Patterning of avian craniofacial muscles. Dev Biol 1986;116:347-56. [PUBMED] [FULLTEXT]|
|21.||Noden DM. Interactions and fates of avian craniofacial mesenchyme. Development 1988;103:121-40. [PUBMED] |
|22.||Noden DM. Origins and patterns of craniofacial mesenchymal tissues. J Craniofac Genet Dev Biol 1991;11:192-213. [PUBMED] |
|23.||Noden DM, Trainor PA. Relations and interactions between cranial mesoderm and neural crest populations. J Anat 2005;207:575-601. [PUBMED] [FULLTEXT]|
|24.||Helms JA, Cordero D, Tapadia MD. New insights into craniofacial morphogenesis. Development 2005;132:851-61. [PUBMED] [FULLTEXT]|
|25.||Carlson BR. Human embryology and developmental biology, 3 rd ed. Mosby: St. Louis; 2004. |
|26.||Vissarionov VA. Analysis of the results of reconstructive plastic surgery of unilateral clefts of the upper lip based on data of the surgical section 1970-1977. In : Pathogenesis, treatment and prophylaxis of cosmetic deformities and deficiencies. Institute of Cosmetology: Moscow; 1982. p. 118-21. |
|27.||Vissarionov VA. Correction of the nasal tip deformity following unilateral repair of clefts of the upper lip. Plast Reconst Surg 1989;83:341-7. [PUBMED] |
|28.||Zhang Z, Song Y, Zhao X, Zhang X, Fermin C, Chen Y. Rescue of cleft palate in Msx-1 deficient mice by transgenic Bmp4 reveals a network of BMP and Shh signaling in the regulation of mammalian palatogenesis. Development 2002;129:4135-40. [PUBMED] [FULLTEXT]|
|29.||Gilbert S. Developmental Biology, 8 th ed. Mass, Sinauer: Sunderland; 2006. |
|30.||Wozney J. Overview of bone morphogenetic proteins. Spine 2002;27:52-8. |
|31.||Ebara S, Nakayama K. Mechanisms for the action of bone morphogenetic proteins and regulation of their activity. Spine 2002;27:S10-5. [PUBMED] [FULLTEXT]|
|32.||Valentin-Opran A, Wozney J, Csiima C, Lilly L, Reidel GE. Clinical evaluation of recombinant human bone morphogenetic protein-2. Clin Orthop Related Res 2002;395:110-20. |
|33.||Boyne PJ. Animal studies of application of rhBMP-2 in maxillofacial reconstruction. Bone 1996;19:83S-92S. [PUBMED] |
|34.||Boyne PJ, Nath R, Nakamura A. Human recombinant BMP-2 in osseous reconstruction of simulated cleft palate defects. Br J Oral Maxillofac Surg 1998;36:84-90. [PUBMED] |
|35.||Boyne PJ, Marx RE, Nevins M, Triplett G, Lazaro E, Lilly LC, et al . A feasibility study evaluating rhBMP-2/absorbable collagen sponge for maxillary sinus floor augmentation. Int J Periodontic Restorative Dent 1997;17:11-25. |
|36.||FDA Dental Advisory Board hearing on Infuse® bone graft for maxillary sinus and alveolar extraction socket defects. Gaithersburg: Maryland; November 9, 2006. |
|37.||Carstens MH, Chin M, Li J. In situ osteogenesis (ISO): Regeneration of 10-cm defect in porcine model using recombinant human bone morphogenetic protein-2 (rhBMP-2) and Helistat ® absorbable collagen sponge. J Craniofac Surg 2005;16:1033-42. |
|38.||Chao M, Donovan T, Sotelo C, Carstens MH. In situ osteogenesis of hemimandible in a 9-year-old boy: Osteoinduction via stem cell concentration. J Craniofac Surg 2006;17:405-12. [PUBMED] [FULLTEXT]|
|39.||Ruberte J, Carretero A, Navarro M, Marcucio RS, Noden D. Morphogenesis of blood vessels in the head muscles of avian embryo: spatial, temporal, and VEGF expression analyses. Dev Dynam 2003;227:470-83. |
|40.||Carstens M, Chin M, Ng T, Tom WK. Reconstruction of #7 facial cleft with distraction assisted in situ osteogenesis (DISO): Role of recombinant human bone morphogenetic protein-2 with Helistat activated collagen sponge. J Craniofac Surg 2005;16:1023-32. |
|41.||Chin M, Ng T, Tom WK, Carstens MH. Repair of alveolar clefts with recombinant human bone morphogenetic protein. J Craniofac Surg 2005;16:778-89. |
|42.||Chin M, Ng T, Tom WK, Carstens MH. 5 th International Conference on Distraction Osteogenesis in Craniofacial Surgery. Paris, France; June, 2006. |
|43.||Carstens MH. The sliding sulcus procedure: Simultaneous repair of unilateral clefts of the lip and primary palate-a new technique. J Craniofac Surg 1999;10:415-29. [PUBMED] |
|44.||Demas PN, Sotereanos GC. Closure of alveolar clefts with Corticocancellous block grafts and marrow: A retrospective study. J Oral Maxillofac Surg 1988;46:682-7. [PUBMED] |
|45.||Delaire J. The potential role of facial muscles in monitoring maxillary growth and morphogenesis. In : Carlson DS, McNamara JA Jr, editors. Muscle Adaptation and Craniofacial Growth. (Center for Human Growth and Development, Craniofacial Growth Monograph No. 8). University of Michigan: Ann Arbor, MI; 1978. p. 157-80. |
|46.||Markus AF, Delaire J, Smith WP. Facial balance in cleft lip and palate. I. Normal development and cleft palate. Br J Oral Maxillofac Surg 1992;30:287-95. |
|47.||Moss ML, Vilmann H, Das Gupta G, Skalak R. Craniofacial growth in space-time. In : Carlson DS, McNamara Jr JA, editors. Craniofacial biology. Center for Human Growth and Development, University of Michigan: Ann Arbor, MI; 1981. |
|48.||Park GC. The importance of accurate orbicularis repair of the orbicularis oris muscle in unilateral cleft lip. Plast Reconstr Surg 1995;96:780-8. |
|49.||Carstens MH. Correction of the bilateral cleft using the sliding sulcus technique. J Craniofac Surg 2000;11:137-67. [PUBMED] |
|50.||Carstens MH. Correction of the unilateral cleft lip nasal deformity using the sliding sulcus procedure. J Craniofac Surg 1999;10:346-64. [PUBMED] |
|51.||McComb H. Treatment of the unilateral cleft lip nose. Plast Reconstr Surg 1975;55:596-601. [PUBMED] |
|52.||Stenstrom SJ. The alar cartilage and the nasal deformity in unilateral cleft lip. Plast Reconstr Surg 1966;38:223-31. [PUBMED] |
|53.||Drake AF, Davis JU, Warren DW. Nasal airway size in cleft and noncleft children. Laryngoscope 1993;103:915-7. [PUBMED] |
|54.||Noordhoff MS, Chen YR, Chen KT. The surgical technique for the complete unilateral cleft lip-nasal deformity. Op Tech Plast Surg 1995;2:164-74. |
|55.||Noordhoff MS. The surgical technique for the unilateral cleft lip-nasal deformity. Noordhoff Craniofacial Foundation: Taipei, Taiwan; 1997. |
|56.||Millard DR. Cleft Craft II. Bilateral and rare deformities. Little, Brown: Boston; 1977. p. 259-65. |
|57.||Siebert JW, Angrigiani C, McCarthy JG, Longaker MT. Blood supply of the Lefort I maxillary segment: an anatomic study. Plast Reconstr Surg 1997;100:843-51. [PUBMED] [FULLTEXT]|
|58.||Delaire J, Precious D, Gordeef A. The advantage of wide subperiosteal exposure in primary surgical correction of labial maxillary clefts. Scand J Plast Reconstr Surg Hand Surg 1989;22:147-51. |
|59.||Precious DA, Delaire J. Surgical considerations in patients with cleft deformities. In : Bell WH, editor. Modern practice in orthognathic and reconstructive surgery. WB Saunders: Philadelphia; 1992. p. 390-425. |
|60.||Rosenstein SW, Grassechi M, Dado DV. A long-term retrospective assessment of facial growth, secondary surgery need and maxillary lateral incisor status in a surgical-orthodontic protocol for complete clefts. Plast Reconstr Surg 2003;111:1-13. |
|61.||Rosenstein SW. Early bone grafting of alveolar cleft deformities. J Oral Maxillofac Surg 2003;61:1078-81. [PUBMED] [FULLTEXT]|
|62.||Goldstein MH. A tissue expanding vermilion myocutaneous flap for lip repair. Plast Reconstr Surg 1984;73:768-70. [PUBMED] |
|63.||Goldstein MH. The elastic flap for lip repair. Plast Reconstr Surg 1990;85:446-52. [PUBMED] |
[Figure - 1], [Figure - 2], [Figure - 3], [Figure - 4], [Figure - 5], [Figure - 6], [Figure - 7], [Figure - 8], [Figure - 9], [Figure - 10], [Figure - 11], [Figure - 12], [Figure - 13], [Figure - 14]
|This article has been cited by|
||Rhinoplastie de fente primaire et secondaire
| ||Jean-Claude Talmant,Jean-Christian Talmant |
| ||Annales de Chirurgie Plastique Esthétique. 2014; |
|[Pubmed] | [DOI]|
||Cleft lip, nose, and palate: The nasal septum as the pacemaker for midfacial growth
| ||Hall, B.K. and Precious, D.S. |
| ||Oral Surgery, Oral Medicine, Oral Pathology and Oral Radiology. 2013; 115(4): 442-447 |
||Cleft lip, nose, and palate: the nasal septum as the pacemaker for midfacial growth
| ||Brian K. Hall,David S. Precious |
| ||Oral Surgery, Oral Medicine, Oral Pathology and Oral Radiology. 2013; 115(4): 442 |
|[Pubmed] | [DOI]|
||Primary Reconstruction of Alveolar Clefts Using Recombinant Human Bone Morphogenic Protein-2
| ||Michael A. Fallucco,Michael H. Carstens |
| ||Journal of Craniofacial Surgery. 2009; 20(Suppl 2): 1759 |
|[Pubmed] | [DOI]|
||Primary reconstruction of alveolar clefts using recombinant human bone morphogenic protein-2: Clinical and radiographic outcomes
| ||Fallucco, M.A., Carstens, M.H. |
| ||Journal of Craniofacial Surgery. 2009; 20(8 SUPPL 2): 1759-1764 |