Facial Nerve Embryology

Updated: Jul 06, 2016
Author: Timothy Els, MBChB; Chief Editor: Arlen D Meyers, MD, MBA 


In order to appreciate the complex embryology of the facial nerve, one has to have a basic understanding of cranial embryology as a whole. Although it may seem daunting to the casual reader, revisiting cranial embryology allows for a more comprehensive understanding of the final 3-dimensional structure of the nerve, as well as the inherent logic in its development. To this end, this article briefly discusses numerous important processes in head and neck embryology, namely the implications of patterning in hindbrain development, the diverse roles of neural crest cells, migration of the neural crest cells into the branchial arches (particularly the hyoid arch), and the genetic control of these processes. This may help prepare the otolaryngologist to comprehend and anticipate variations encountered in clinical practice, such as anticipating facial nerve anomalies in congenital stapes fixation.[1]  However, the main objective of this article is to outline the embryology of the facial nerve and its common clinical implications. The reader is referred to Embryology and Anomalies of the Facial Nerve and Their Surgical Implications, 2nd Ed for a more comprehensive review of the development of the facial nerve and the associated development of the ear (see table 1).[14]

The surgical anatomy and landmarks of the facial n The surgical anatomy and landmarks of the facial nerve.
Schematic illustration shows the facial nerve and Schematic illustration shows the facial nerve and its peripheral connections. Note the interconnections of cranial nerve (CN) VII with CN V, CN IX, and CN X.

The Mature Facial Nerve

While studying the embryology of the facial nerve, keep in mind the mature course and structure that is the end result of developmental events. The motor nucleus of the facial nerve is located in the reticular formation of the caudal pons. Upon leaving the motor nucleus, axons extend dorsally and medially, cranially and superficially, to bend around the abducens (sixth cranial nerve) nucleus. The fibers then exit the central nervous system (CNS) between the olive and the inferior cerebellar peduncle.

The sensory root (nervus intermedius) consists of (1) central projections of neurons located in the geniculate ganglion (general somatic fibers that synapse in the spinal nucleus of the trigeminal nerve and special afferent fibers that synapse in the nucleus solitarius) and (2) axons of parasympathetic neurons from the superior salivatory (lacrimal) nucleus. The nervus intermedius enters the CNS lateral to the motor root at the pontocerebellar groove.

After exiting the internal auditory canal, the facial nerve enters the middle ear, where it bends posteriorly (first, or medial, genu) and courses horizontally through the middle ear. Just anterior to the lateral aspect of the horizontal semicircular canal, the facial nerve curves gently (the second genu) to form the vertical, or mastoid, segment that exits via the stylomastoid foramen.


Overview of Hindbrain Development

The development of the hindbrain (rhombencephalon) and the subsequent delamination of the neural crest cells are interrelated processes that need to be understood to appreciate the development of the branchiomotor cranial nerves in general and the facial nerve in particular.[2]  Primary neurulation is a process that leads to the development of the neural tube from the neural plate. During the first 4 weeks of embryogenesis, the notochord induces axial ectoderm to form the neural plate, which then folds along its long axis to form the neural tube. The tube subsequently develops vesicles at its rostral end, which give rise to the forebrain (prosencephalon), midbrain (mesencephalon), and hindbrain (rhombencephalon).[3]  The latter then divides into the metencephalon and caudal myelencephalon.[2]

The rostral division of the neural tube into its 3 main sections falls under the control of homeobox (Hox) family of genes.[4]  The Hox genes, well described as the master regulators of development, encode a set of transcription factors that specify the identity of particular segments during embryogenesis.[5]  The hindbrain is subsequently further segmented under the influence of Hox genes to give rise to rhombomeres, which are 8 transient neuroepithelial segments, denoted as r1 to r8. These sequential processes are referred to as anteroposterior (AP) patterning (this is distinct from ventral-dorsal patterning, which is discussed below).[4, 3] The rhombomeres r2 to r 6 are the most important for understanding facial nerve development, whereas r1 lacks branchiomotor nerves and develops into the cerebellum.[3]  The early segmental organization of the hindbrain has long-term effects on cell differentiation, in that each rhombomere gives rise to specific branchiomotor neurons, cranial ganglia, and pathways of neural crest cell migration.[4]  The rhombencephalon eventually divides into two main sections: the rostral metencephalon (pons and cerebellum) and the caudal myelencephalon (medulla). The facial nerve exists between these two structures at 40 days of development.

The AP segmentation of the hindbrain and the development of the pharyngeal (branchial) arches are intimately related.[4, 3]  Each rhombomere and its linked pharyngeal arch share the same combination of Hox genes, creating a “Hox code.”[3]  Neuronal progenitors (mainly neural crest cells), originating from the rhombomeres, display programmed migratory behaviors, sending axons along defined trajectories to peripheral targets due to their segmental identity.[2]  The origin of the neural cell progenitors along the AP axis of the hindbrain determines segmental identity; sensory or motor function is determined by the position along the mediolateral axis.[2]  The branchiomotor neurons of the first 3 branchial arches (cranial nerves V, VII and IX) are derived from neural crest cells, originating from r2, r4, and r6, which migrate ventrolaterally in 3 streams toward arches I, II, and III, respectively. Considerably fewer neural crest cells derived from r3 and r5 migrate laterally; these cells typically migrate caudally or rostrally to join the even-numbered streams.[4]  However, the facial visceromotor (parasympathetic) nucleus develops from r5.[3]

Various additional genes participate in regulating the steps involved in AP patterning of the hindbrain into rhombomeres (alongside the aforementioned  Hox family). The relationship between segmentation genes (Kreisler and Hoxa) is complex, with genes directing growth (Krox20), genes directing rhombomere identity (Hoxa2, Hoxa3, Hoxb2, Hoxb3), and those controlling cellular restrictions (EphA4 and EphA7). However, the Hox gene family and the Ephrin/Eph receptor gene families appear to be the most important for establishing and maintaining the segmentation of the rhombencephalon.[4]

In the ventral-dorsal (or mediolateral) axis, the hindbrain divides into basal and alar plates under the influence of “sonic hedgehog” (Shh) and bone morphogenic protein (BMP) genes, respectively.[2, 3]  This process, known as mediolateral patterning, occurs by induction of the primordia of the longitudinal tracts of the developing central nervous system[3, 6] . The basal plates give rise to somatomotor, special visceral efferent (branchial) and general visceral efferent (autonomic) nuclei, whereas the alar plates give rise to sensory nuclei, respectively.[6]  The boundary between these two entities is in all species studied marked by a sulcus limitans.[6]


The Hindbrain Nuclei of the Facial Nerve

The motor nuclei of the facial nerve develop early during embryogenesis; the sensory components follow later with the formation of the geniculate ganglion and nervus intermedius. Cranial motor neurons can be classified into 3 subtypes: general visceral efferents (general visceromotor or parasympathetic), somatomotor, and branchiomotor (special visceral efferents or special visceromotor).[7]  Note that the term "visceromotor" denotes the group that includes both the general visceromotor (parasympathetic) and special visceromotor (branchiomotor). The basal plate of the hindbrain can be divided into a medial zone or column and an intermediomedial zone or column.[6]  The medial zone gives rise to somatomotor nuclei of (IV, VI, and XII), as well as the rostral end of the spinal motor column. The intermedioventral zone contains both general and special visceromotor cell masses, formed by the motor nuclei of V, VII, IX, and X.[6]

Human facial nerve motor nuclei develop around the fourth week of embryogenesis from r4 and migrate caudally through r5 into r6, subsequently forming a loop or genu around the abducens (CN VI) somatomotor nucleus in r5. This explains the course of the facial nerve in relation to the abducens, as well as the combined fallout seen in Moebius syndrome.[7, 3]  The general visceromotor neurons develop from r5 exclusively.[3]  Although the branchiomotor and visceromotor neurons of the facial nerve develop from the basal plate, they exit the brainstem dorsally from the alar plate.[7]  Both the visceromotor and branchiomotor cells bodies are located in their hindbrain nuclei; the sensory neuron (somatic and taste) cell bodies are located within the geniculate ganglion.

The sensory nuclei of the facial nerve develop from the alar plate of the hindbrain. The alar plate can be divided into an intermediodorsal zone or column and a dorsal zone or column.[6]  The caudal part of the intermediolateral zone contains a centers known as lobus vagi or solitary tract nucleus, which receive viscerosensory (mostly gustatory) fibres from the VIIth, IXth, and Xth cranial nerves. The intermediolateral zone also contains several general somatosensory centers and special somatosensory centers. The general somatosensory centers include the diffuse nucleus tractus descendens of the trigeminal nerve and the nucleus princeps (main sensory nucleus) of the same nerve, which receive fibers from the nervus intermedius.[6]

Although the patterning of the developing facial nuclei have been described, examining the factors that initiate the development of these nuclei and regulate the growth and directionality of cranial nerve axon growth to their targets is important. The Shh family of genes play a pivotal role in inducing motor neurons throughout the length of the vertebrate neural tube; in this regard, motor neurons are completely absent in Shh knockout mice.[7]  Other components of the Shh signaling pathway have also been identified, such as transmembrane proteins (patched and smoothened), as well as transcription factors (Gli), all with various activating or inhibitory functions.[7]

Apart from the induction of motor neurons in the basal plate, numerous genes are necessary to specify branchiomotor/visceromotor neuron identities within their respective rhombomeres. As described for rhombomere patterning, these genes include krox20 (responsible for r3 and r5), kreisler (r5, r6), Hoxa2 (r1, r2), Hoxa1, and Hoxb1 (expressed from r3/r4 caudally).[7]  Hoxb1 is particularly important for CN VII development, as Hoxb1 knockout mice do not develop a facial motor nucleus at all. Further genes directly impacting the developing facial motor nucleus include Pbx4, Meis, Nkx6.1, and Gata2[7] . In this regard, mice deficient in Pbx4 display failure of r4 development, whereas deficiencies in Meis lead to maldevelopment of r3-r5. Similarly, in Gata2 knockout mice, CNVII motor nuclei fail to migrate caudally, which is partly due to the associated downregulation of Nkx6.1, which normally specifies a ventral branchiomotor neural progenitor domain associated with the developing facial motor nucleus.[7]

After induction of its motor nuclei in the brain stem, the motor neuron must migrate from its nucleus to ultimately innervate specific facial muscles. This occurs in a 2-stage process that involves axonal growth from the basal plate towards its exit point, subsequent extension into the appropriate branchial arch, and cessation of migration once the designated muscle is met. With regards to the former, genes encoding repellant molecules play an important role in directing facial motor neuron axons away from the basal plate. These genes include netrins and semaphorins and encode chemorupulsive molecules.[7]  The possibility of chemoattractant molecules toward hindbrain exit points has also been postulated but as yet have not been identified. The control of motor neuron axonal growth into the periphery similarly depends on chemoattractant and chemorepellant mechanisms. Chemorepellants belong to the semaphorin family, whereas chemoattractant genes include ephrins, hepatic yet growth factor, brain derived growth factor, ciliary derived neurotrophic factor, and cardiotrophin.[7]  However, further discussion of the development of the facial nerve outside of the hindbrain requires a review of the neural crest (NC).


The Neural Crest and Hyoid Arch Invasion

A basic understanding of cranial NC cell migration in the head and neck is important to appreciate facial nerve migration into the hyoid arch. This, of course, implies that the reader has a working knowledge of the NC in general. The NC is a pluripotent cell population that originates from the junction between the neuroepithelium of the developing neural tube and non-neural ectoderm, which later delaminates from the developing neural tube.[4, 8]  NC cells are critical for the development of the vertebrate head and neck because they give rise to most of the skull bones, the facial skeleton, the visceral skeleton, peripheral neurons, and glia.[4, 8]

NC cells undergo a process called epithelium to mesenchyme transition (EMT), thereafter differentiating into a myriad of tissues types, including nerves, smooth muscle, cartilage, bone, and melanocytes.[8]  The NC cells separate (or delaminate) from the neural plate border and then disperse from the length of neural axis and migrate throughout the developing embryo. Cranial NC cells from the diencephalon and anterior mesencephalon invade the frontonasal process, whereas those from the hindbrain populate the branchial arches.[9]  This is a highly regulated process that begins as a continuous delamination wave in the cranial region but quickly separates into 3 main “streams,” which appear adjacent to even numbered rhombomeres (r2, r4, r6) and are conserved across a wide number of vertebrate species.[9, 8]

NCs destined to develop into head and neck ganglia migrate to dorsal positions whereas those destined to become cranial bone and cartilage continue to ventral locations within the branchial arches.[8]  The maintenance of the segregation of the streams has important implications for facial patterning because the NC cells from particular rhombomeres develop into genetically predetermined structures within their associated pharyngeal arches. When NC cells destined for a particular arch are grafted to adjacent arches, they still form the structures they were programmed to form due to the segmental identities imposed by the Hox code during hindbrain segmentation/patterning (as described above). [4, 3]  

Three main streams are named in accordance with the branchial arches they feed into, namely the mandibular (1st arch), hyoid (2nd arch) and branchial (third arch), with migratory streams separated by NC cell-free regions adjacent to rhombomeres r3 and r5.[8]  However, the migrating NC cells may also serve a signalling role in cranial embryology, which is generally overlooked; for example, this is noted in NC-directed migration of placode derived neurons within the intermediate nerve.[10] . One possibility is that migratory NC cells are signaling to their surrounding tissues while on the move and that the patterning of their migratory routes is crucial to position the signals coming from the NC cells.[8]

Operational models describe the expression of several genetic markers during neural crest development, the so-called NC gene-regulating network (NCGRN).[9]  This model links the expression of transcription factors and signalling molecules that determine NC cell induction, migration, and differentiation. Within the model is a sequential or temporal expression of genes that coincides with NC development.[9]  These genes include signaling molecules (BMP, FGF, Notch, RA, Wnt) that induce border specific genes (transcription factors Msx1, Msx2, Pax3, Pax7, Zic1), which direct the expression of NC specifiers (AP2, FoxD3, Snail2, Sox9, Sox10), as well as effector genes (Sox9, Sox10, Cad7, ColIIa, Ngn-1, Mitf, Dct).[9]  Failure of cranial NC migration leads to significant craniofacial, as well as cardiac abnormalities. Mice with Sox 10 mutations display deformed ganglia, whereas AP2-mutated mice exhibit craniofacial abnormalities and malformed cranial nerves.[9]

According to Kulesa et al,[11]  the 3 phases of cranial neural crest migration are as follows:

  1. Acquisition of directed migration to the dorsolateral pathway: This occurs via ErbB4, Eph/ephrin interactions, chemokines, and neuropili/semaphoring interactions.

  2. Homing to the branchial arches via segmental streams along the dorsolateral pathway: This occurs via guidance from chemokines, neuropilins, and their ligands. CXCR4/SDF-1 signaling is important in condensing and patterning the NC cells into the pharyngeal arches. Furthermore, semaphore-neuropilin interactions are necessary for the initiation of NC streams.

  3. Invasion of the branchial arch mesenchyme: This is a complex process and requires highly regulated, multiple guidance cues. Once again, neuropilins are involved in the signalling leading to the invasion of the branchial arches. Furthermore, VEGF has been shown to be a strong chemoattractive cue for NC cells into the second arch. Multiple other chemoattractive molecules have also been identified.

Apart from the chemoattractive cues alluded to above, numerous mechanisms have been suggested to explain the directed migration of NC cells from the hindbrain to the branchial arches. These include cell nudging, population pressure, and contact inhibition of movement and polarized cell movement.[11]  Another way of conceptualizing the control of NC cell migration is to consider regulation as the interplay between positive and negative external regulators.[8]  Negative regulators include the ephrins and their Eph receptors, as well as the class 3 semaphorins and their neuropilin/plexin receptors. These genes prevent the entry of NC cells into specific zones. A similar role is performed by the EGF-like receptor ErbB4. Cephalic NC cells also encounter physical barriers; the otic placode obstructs the flow of early migrating cells. In terms of positive regulators, two main groups of molecules can be identified: chemoattractants and permissive factors. CXCR4 belongs to the former category, whilst VEGF, FGF, and PGDF belong to the latter. For a comprehensive overview of cranial NC migration, the reader is directed to the work of Kulesa et al and Theveneau.[8, 11]

The preceding description of the NC serves as a precursor to discussion of the facioacoustic primordium, which develops during the third week of embryogenesis from neural crest cells destined for the hyoid arch. It is attached to the metencephalon just rostral to the otic vesicle and gives rise to the sensory portions of the facial nerve as well as the geniculate ganglion in conjunction with the epibranchial placode of the hyoid arch.[10]  The primordium develops ventrally to the deep surface of the epibranchial placode, a thickened area of ectoderm located at the dorsal and caudal aspect of the first branchial groove. Placodes are transient thickenings of embryonic head ectoderm and give rise to the distal sensory ganglia of cranial nerves VII, IX, and X.[10]  The epibranchial placode of the second arch gives rise to gustatory fibres of the facial nerve as well as the geniculate ganglion, whereas the NC cells give rise to the general somatic sensory fibers.[10]  However, the targeting of the central axons of placodal derived gustatory fibres to the rhombencephalon, as well as the migration of these neurons, depends on neural crest signalling.[10]

The developing facial nerve becomes more superficial and rostral as it courses towards the placode.[6]


The Development of the Fallopian Canal

The fallopian (facial nerve) canal is an important structure for the otologic surgeon because it protects the nerve from damage during middle ear surgery. In-depth knowledge of its development and typical mature structure is vital for performing safe ear surgery. The mature facial canal is roughly Z-shaped and courses from the internal acoustic meatus to the stylomastoid foramen and encloses the labyrinthine, tympanic, and mastoid segments of the facial nerve.[12]  The development of the fallopian canal can be divided into 3 stages: before 16 weeks' gestation, from 16-21 weeks' gestation, and 22-25 weeks' gestation.[12]

Before 16 weeks' gestation, the otic capsule is cartilaginous and the developing perichondrium splits to envelop the facial nerve.[12]  This results in a cartilage base medially and a condensed mesenchymal covering on the lateral aspect of the nerve. The otic capsule develops from the otic placode, which invaginates to form the otic pit and, subsequently, the otic vesicle during the fourth week of gestation.[1]  At around the same time, the facial nerve canal develops as a sulcus running along the lateral aspect of the developing otic capsule. The common temporal origin of the stapes and facial nerve from the hyoid arch may explain the association between malformed stapes and facial nerve aberrations.[1]  The canal passes dorsally and caudally from the geniculate ganglion above the stapes and around the developing round window. The canal is, however, dehiscent from the labyrinthine to the end of the tympanic segment at this point.

The development of the tympanic portion of the facial canal depends on a direct apposition between the developing facial nerve and the otic capsule, ultimately resulting in a mature position between the lateral semicircular canal and stapes foot plate.[13]  This dependence on direct contact with the optic capsule may explain why the tympanic segment is most prone to dehiscence and anomalous positioning. The initial path of the nerve distal to the geniculate ganglion is straight in early embryonic life but by the end of 8 weeks' gestation, the orientation of the facial nerve within the developing temporal bone has been established. The ultimate position of the canal and completeness of bony covering is determined by the concomitant development of the stapes and bony labyrinth.

The labyrinthine portion of CN VII is positioned in a shallow sulcus on the superior portion of the otic capsule between the developing cochlea and anterior semicircular canal. This position is stable and constant, and dehiscence of this portion of the mature canal is rare.[13]  The nerve furthermore develops a tight connection with the cochlea and runs parallel to its first turn. This explains the course of the nerve between the internal acoustic meatus and geniculate ganglion.[13]

The mastoid segment of the facial canal develops postnatally as mastoid growth ensues, although the mastoid begins to develop from the surrounding mesenchyme from around 15 weeks' gestation.[13]  The stylomastoid foramen is, however, much more superficially located at birth than in the adult. During 11 weeks' gestation, branches develop from the facial nerve between the stapedius and the developing chorda. These branches join with branches of CN IX and X to supply sensation to the external auditory canal and auricle.

From 16-21 weeks' gestation, ossification of the otic capsule occurs by multiple endochondral ossification centers along the facial sulcus, at the same time superior and inferior rims (or clasps) develop from the facial sulcus.[12]  At 22-25 weeks' gestation, the superior and inferior rims ossify by dismal ossification of the mesenchymal covering of the canal. The rims develop from anterior to posterior, and the superior rim contributes more than half of the circumference of the canal.[12, 13]  The inferior rim may be significantly shallower; if fusion is incomplete, dehiscence of the canal persists in the adult. Given the direction of development and ossification of the rims, the most common area of dehiscence is the inferior portion of the segment of canal adjacent to the stapes, which is the posterior aspect of the tympanic segment of the fallopian canal.[13]


Development of the Facial Nerve in Weeks

Weeks 0-4 (0-6 mm)

The rhombencephalon (or hindbrain) is divided into the myelencephalon (caudal), which becomes the medulla oblongata, and the metencephalon (cranial), which becomes the pons and cerebellum. The facioacoustic (acousticofacial) primordium appears during the third week (4.2 mm crown-rump length [CRL]). It is attached to the metencephalon just cranial to the otic vesicle. The facial part of the acousticofacial primordium migrates cranial and ventral to end adjacent to the epibranchial placode, which is located on the dorsal and caudal aspect of the first branchial cleft.

By the end of the fourth week of gestation (4.8-6.5 mm CRL), the facial nerve splits into 2 parts: the caudal and rostral trunks. The chorda tympani nerve exits rostrally and courses ventrally to the first pharyngeal pouch to enter the mandibular arch. Shortly thereafter, the nerve approaches the epibranchial placode, inducing the appearance of the large, dark nuclei of neuroblasts that represent the future geniculate ganglion.

Weeks 5-6 (7-17mm)

Mesenchymal concentrations that form the cephalic muscles are seen in association with their nerves, while the epibranchial placode disappears and the geniculate ganglion is identifiable. The greater superficial petrosal nerve (GSPN) is present. The chorda tympani nerve enters the mandibular arch and terminates just proximal to the submandibular ganglion, near a branch of the trigeminal nerve that will become the lingual nerve. The posterior auricular nerve appears near the chorda tympani.

Complete separation of the facial and acoustic nerves is apparent, and a discrete nervus intermedius develops, making this an important temporal reference point for gestational disorders that affect both systems. The GSPN courses to the lateral aspect of the developing internal carotid artery (ICA), where it joins the deep petrosal nerve and continues as the nerve of the pterygoid canal. It terminates in a group of cells that will become the pterygopalatine ganglion. At this point, the most distal branches of the facial nerve are a loose network or interconnecting twigs.

Week 7 (18-31 mm)

The nervus intermedius is now smaller than the motor root and enters the brainstem between the vestibulocochlear nerve and the motor root of the facial nerve. The chorda tympani and lingual nerve unite proximal to the submandibular gland. The posterior auricular nerve now divides into cranial and caudal branches.

Several branches are visible in the peripheral portion of the seventh nerve. All of the peripheral branches lie deep to the myoblastic laminae that will form the facial muscles. At the end of the seventh week, the separations between the terminal branches continue to increase to the extent that all peripheral divisions can be identified.

The parotid gland is beginning to develop from the parotid bud at this stage. The temporal, zygomatic, and upper buccal branches are superficial to the parotid primordium, while the lower buccal, mandibular, and cervical branches are deeper. Multiple facial muscles appear at this time as well, including the zygomaticus major and minor, depressor anguli oris, buccinators, and frontalis.

Week 8 (32-49 mm)

A sulcus develops around the facial nerve that is the beginning of the fallopian canal. The orbicularis oris, levator anguli oris, and orbicularis oculi muscles appear.

Week 9 (50-60 mm)

Auricularis anterior, corrugator supercilii, occipital and mandibular platysma, and levator labii superioris alaeque nasi muscles appear. All the cranial nerves more closely resemble their adult relationships.

Weeks 10-15 (61-80 mm)

Extensive branching of the peripheral portions of the facial nerve occurs at this stage. Communication with the trigeminal nerve (via infraorbital, buccal, auriculotemporal, and mental branches) occurs in the perioral and infraorbital regions. The vertical portion of the facial nerve begins in the middle ear, and its overall relationship to external and middle ear structures is far more anterior than in the adult. Branches that will supply sensation to the external auditory canal arise between the stapedius and chorda tympani nerves.

Intricate connections between the superficial and deep lobes of the parotid and their relation to the facial nerve develop. By the fifteenth week, the geniculate ganglion is fully developed, and the facial nerve's relationship to middle ear structures is more fully developed.

Week 16 to birth (146 mm)

All definitive communications of the facial nerve are established by the 16th week. At 26 weeks' gestation, ossification has progressed to partial closure of the previously formed sulcus into the fallopian canal.

In late fetal life, the fallopian canal is closed by bone in most areas, except in the anterior cranial portion, where it remains open to form the facial hiatus along the floor of the middle cranial fossa. At least 25%, and as many as 55%, of fallopian canals are dehiscent, with the most common location adjacent to the oval window. At birth, the anatomy of the facial nerve approximates that of the adult, except for its exit through the more superficially located stylomastoid foramen. Adult anatomy will form in this region as the mastoid tip develops after birth. (See Table 1, below.)

Table 1. Summary of the Derivatives of the Second Branchial Arch (Open Table in a new window)

Pharyngeal Arch





II = Hyoid (Reichert cartilage)

Cranial nerve VII (Facial nerve)


  • Muscles of facial expression

  • Buccinator

  • Posterior belly of digastric

  • Stylohyoid muscle

  • Stapedius muscle

  • Manubrium of malleus

  • Long process incus

  • Stapes (except for footplate)

  • Facial canal

  • Styloid process

  • Stylohyoid ligament

  • Lesser cornu of hyoid

  • Upper body of hyoid


Congenital Facial Paralysis

Abnormalities of the facial nerve may occur in conjunction with malformations of the ear, in isolation without associated anomalies, or in conjunction with a variety of syndromes that include abnormalities elsewhere in the body.

In the newborn, the otolaryngologist evaluating a facial paresis or facial palsy must decide whether it is congenital or acquired. One in 2000 live births has a unilateral facial palsy, with a 90% spontaneous recovery rate. Approximately 75-80% of palsies in newborns are related to birth trauma. A history of forceps delivery, prolonged labor, ecchymosis over the mastoid, or hemotympanum raises suspicion for birth trauma.

The presence of bilateral facial paralysis, other cranial nerve deficits, or other anomalies suggests a developmental etiology. Early, accurate diagnosis is important if the etiology is traumatic. In rare cases, surgery and facial nerve repair may be required in the newborn if the etiology is determined to be traumatic.

The evaluation of facial nerve paralysis includes the use of electromyograms (EMGs), evoked electromyograms (EEMGs), and computed tomography (CT) scans. If the etiology is traumatic, the nerve can be stimulated for 3-5 days postnatal; fibrillation potentials on EMG develop 14-21 days after birth. If the cause is not traumatic, treatment generally is delayed. Eye protection is rarely required in congenital facial paralysis.

In patients with congenital malformations, eliciting the fetal age at which development was arrested is usually possible. This allows for elucidation of the anatomy of the malformed structure based on its normal course of embryologic development. Furthermore, if anomalies are present in other organ systems (in particular, the kidney), they often reflect arrested development at the same time during development. In this way, the surgeon should be able to predict the location of the facial nerve, particularly in the case of middle ear malformation.

Most hereditary conditions that include facial paralysis are manifest at the time of birth. However, a few hereditary syndromes are associated with the development of facial paralysis later in life. (See Table 2, below.)

In addition, many hereditary and congenital malformations are associated with abnormal facial nerve anatomy in the presence of normal nerve function. The otolaryngologist must be familiar with these conditions, because abnormal development may place the nerve at increased risk of injury during otologic surgery.

Table 2. Developmental Syndromes Associated With Facial Nerve Abnormalities* (Open Table in a new window)


Facial Nerve Abnormality


Bulbopontine paralysis with progressive sensorineural hearing loss

  • Facial paresis common

  • Complete facial paralysis uncommon


  • Autosomal recessive inheritance

  • More females affected than males

  • Involvement of bulbar musculature

  • Slowly progressive, bilateral sensorineural hearing loss

DiGeorge syndrome

Facial paralysis reported

  • Multiple anomalies of craniofacial, cardiovascular, and visceral structures

  • Absent and/or hypoplastic thymus and parathyroid glands

  • Abnormal aural development

Dominant craniometaphyseal dysplasia

Unilateral and bilateral facial paralysis reported

  • Autosomal dominant

  • Metaphyseal widening of limbs and bony overgrowth of facial bones and skull

  • Obliteration of mastoid air cells

  • Conductive and sensorineural hearing loss

  • Manifest in early infancy or childhood

Hemifacial microsomia

Facial paralysis

Facial asymmetry, including unilateral microtia, macrostomia, and failure of mandibular ramus and condyle to form

Hereditary acoustic neuromas

Facial paresis and/or palsy

  • Autosomal dominant

  • Manifests in third decade

  • Symptoms secondary to tumor encroachment on nerves

Melkersson-Rosenthal syndrome

Recurrent alternating facial paralysis

  • Recurrent alternating facial paralysis

  • Edema of lips, face, and eyelids

  • Cheilitis and fissured tongue

  • Positive family history

Möbius syndrome

  • Bilateral facial paralysis

  • Masklike facies

  • Unilateral or bilateral abducens paralysis

  • Deformities of extremities

  • Aplasia of brachial and thoracic muscles



Facial paralysis, which may be acute and recurring

  • "Marble bone disease"

  • Uncommon genetic disorder characterized by increasing skeletal density

  • Two forms, benign and malignant

  • Benign form is autosomal dominant and not commonly associated with facial paralysis

  • Malignant form is autosomal recessive (known as Albers-Schönberg disease) and is associated with facial paralysis, deafness, blindness, and abnormal ear and mastoid; death often results at a young age due to infection

Recessive craniometaphyseal dysplasia

Unilateral facial paralysis

  • Autosomal recessive bone disease

  • Major features - Glabella and paranasal prominence with severe mandibular prognathism

  • Nasal obstruction (complete)

  • Ocular hypertelorism

  • Progressive visual disturbance and blindness

  • Severe hearing loss


  • Facial paralysis possibly apparent at birth

  • Often appears early in childhood

  • Generally unilateral, eventually becoming bilateral


  • Autosomal recessive

  • Generalized osteosclerosis, hyperostosis of mandible, calvaria, pelvis, and clavicles

  • Hearing loss usually bilateral, may be conductive, sensorineural, or mixed

Sickle cell disease

Facial paralysis observed to occur during a crisis

  • Hemoglobinopathy observed primarily in African Americans

  • Recurrent attacks of fatigue, weakness, abdominal pain, anorexia, and jaundice

  • May have sensorineural hearing loss during a crisis


von Recklinghausen neurofibromatosis

Facial paralysis possible from a neurofibroma of the facial nerve or secondary to encroachment by an acoustic schwannoma

  • Autosomal dominant

  • Multiple skin tumors, cutaneous pigmentation (cafe-au-lait spots)

  • Incidence of 1 in 2000

  • Malignant degeneration in 3-12%


*Modified from Sataloff, 1991