Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
Current anatomical literature has widely adopted the terminology recommended by the and published in Terminologia Anatomica (1998). More than 100 years ago it was recognised that eponyms are not descriptive and that more uniform anatomical terms should be developed to promote better and clearer communications involving anatomical structures. The first text on anatomical nomenclature ( Nomina Anatomica ) was thus published in Latin in Basel, Switzerland in 1895 and has since been updated at regular intervals. Sections on embryology and histology have been added, and the seventh and latest volume changed its name to Terminologia Anatomica (TA) and was published in 1998. Anatomical nomenclature is the basis for medical terminology and it is paramount that scientists and clinicians throughout the world use the same terminology to facilitate communication and avoid confusion. Therefore the anatomical terminology used in this chapter is following that outlined by Terminologia Anatomica (1998). However, eponymous names still continue to be used in much of the literature.
The cornea occupies about 7% of the area of the outer coat of the eye and the anterior surface is a horizontal oval of approximately 11.7 × 10.6 mm. The posterior corneal surface is regarded as spherical ( ). When made an optical assessment of the corneal width using the corneoscleral sulcus as a landmark, they arrived at a corneal diameter of 12.9 × 11.6 mm.
Most corneal diameter assessments appear to use the visible iris diameter as a definition of corneal diameter but found the limbus to be at least 1 mm wide. However, sources have indicated that it is up to 2 mm in width ( , ) – see Section 9, Addendum, available at: https://expertconsult.inkling.com/ .
The cornea is formed by five distinct and very uniform layers:
epithelium
anterior limiting lamina (eponymous term: Bowman's layer)
substantia propria or stroma
internal limiting lamina (Descemet's layer)
endothelium.
The average human cornea is 535 µm thick with a range for normal between 473 µm and 596 µm. It thickens by 23%, or about 100 µm, from the centre to periphery, and is steeper centrally and flatter peripherally ( ) and has the shape of a minus meniscus lens (see also ‘Pachometry/Pachymetry’, Chapter 8 ). Corneal thickness is a factor that influences the intraocular pressure reading. A thicker cornea leads to an artificially higher reading and vice versa.
The outermost layer of the cornea is squamous stratified epithelium, from five to eight cells thick, and is probably the most regular of its type in the body ( Fig. 3.1a ).
The epithelium is thinnest centrally, between 48 µm and 51 µm thick ( , , , ), increasing from about 50–60 µm about 3 mm from the centre ( ) and about 70 µm near the limbus.
Peripherally, the epithelium is continuous with the conjunctiva (see ‘ Conjunctiva ’, p. 43 of this chapter), which initially becomes thicker over the limbus. As it moves into the conjunctiva, it becomes thinner, except for the palisades of Vogt, where the epithelium has areas of thickening projected inwards. The smooth external and internal surfaces are lost as the epithelium progresses into the conjunctiva.
The basal cells are columnar with a spherical or slightly oval nucleus containing dispersed chromatin (euchromatic).
Their cytoplasm contains:
many intermediate filaments, mostly grouped into bundles (tonofilaments)
free ribosomes
sparse mitochondria
a small amount of granular endoplasmic reticulum
glycogen granules
occasional Golgi complexes (centres for intracellular synthesis and metabolic activity).
The basal cells are covered by several layers of cells that become smaller, flatter and broader with distance from the base; the nuclei become more ovoid and finally disciform in the squamous superficial cells. The first two or three rows of suprabasal cells often overlap the apices of more than one underlying cell and are called umbrella or wing cells ( Fig. 3.1b ). The organelles are similar throughout the epithelium, except that filament aggregation is not observed in the squamous cells, and they often contain numerous small vesicles.
Hemidesmosomes attach the basal cells to the epithelial basement membrane. They are so numerous that they occupy at least one-third of the area of the membrane ( Fig. 3.2 ). Hemidesmosomes are macula junctions (or macula adherens – where two cells adhere to each other) continuous with basal cell tonofilaments and consist of local thickenings with increased density of the plasma membrane ( lamina densa ) opposite a thickened zone of the basement membrane. The lighter interval between the two thickened membranes ( lamina lucida ) contains numerous fine connecting filaments. On the opposite side of the junction, the basement membrane is attached to the underlying stroma with anchoring filaments, which in turn are attached to fine branched collagen structures called attachment plaques ( ) ( Figs 3.3 and 3.4 ).
Cell borders are characterised by shallow interlocking ridges covering most of their surfaces, with little space between cells. The ridges are least frequent in the apposed membranes of basal cells. Cells are joined together by many desmosomes which, together with the interlocking ridges, provide the exceptional strength of the epithelium against shearing forces such as eye rubbing.
Gap junctions ( Fig. 3.1c ) are noted infrequently between the cells and at the epithelial surface; adjacent surface cells are joined by tight junctions (zonulae occludentes). Unlike the other junctions, they girdle the cells, forming close contact, but do not completely fuse. This limits permeability, allowing access to the intercellular space from the tear film, or vice-versa, only through discrete ion-selective pores ( ). Desmosomes are attached to the cytoplasmic tonofibrils, forming a girder-like desmosomal-cytoskeletal network supporting the epithelium layer. Gap junctions are responsible for intercellular communication by permitting ionic exchange and reducing electrical resistance between cells.
Unlike surface skin cells, corneal cells are normally unkeratinised and retain their organelles, which suggests that metabolic processes are still functioning. Superficially the cells look smooth when examined at medium magnification through a light microscope, but with electron microscopy they resemble the pile of a carpet. The superficial squamous cells form tightly packed microvilli and/or microplicae ( Figs 3.1 and 3.5 ) up to 0.75 µm long in humans.
As cells slough from the surface, desmosomes split to achieve detachment, and cytoplasmic extrusion at these points may be responsible for the microvilli development ( Fig. 3.6 ) ( ). Differences in the appearance of surface cells can be observed by specular ( ) or scanning microscopy ( ). The latter technique reveals light, intermediate and dark cells, depending on the surface form and size, which may change with age ( , ). The degree of overlap between the flat surface cells may account for this variation, although cells with a light electron reflex are generally smaller, whereas dark reflex cells tend to be larger and have a greater number of sides ( ).
Generations of new cells occur mainly in the basal cell layer (Machemer 1966). Epithelial mitosis displays a circadian rhythm in rats, being most common at 2–7 a.m. and least common at 2–7 p.m. ( , , ). Basal cells make room for new ones by migrating to the next layer and subsequently to the surface, becoming squamosed and sloughed away by the action of the eyelids. These cells constitute the greater part of proteinaceous material found in tears.
Autoradiographical studies of epithelial cell nuclei, labelled with tritiated thymidine, indicate that the life cycle of cells is 3.5–7 days in a variety of young animals ( , ). By arresting the division of cells at metaphase with colchicine, determined that 14.5% of epithelial cells are renewed daily in rats. Assuming surface cells desquamate at the same rate and that all cells are resident for a similar period, the epithelium could be fully regenerated in 7 days. Hanna and O'Brien (1961) used results of isotope studies of enucleated eyes to estimate the life cycle of human corneal epithelial cells to be 7 days, but suggested that complete turnover of the corneal epithelium took 14 days. However, corneal epithelial renewal is not as simplistic as a posterior to anterior movement of cells. Since the cornea is dependent on a peripheral influx of epithelial cells, the determination of the lifespan of a cell arriving from the conjunctiva or limbus will be more complex, and the time requirement for this travel must be longer than for the voyage from the basement membrane to the ocular surface.
Cell renewal, a feature of all epithelia, is dependent on stem cells, and the cornea is no exception. The stem cell is a slowly cycling basal cell that divides to produce a replica of itself and a transient amplifying cell, or a daughter cell, with the capacity to proliferate before differentiation. The self-renewed stem cell remains basal, and the transient amplifying cell is responsible, by relatively rapid division, for cell turnover.
The cornea is exceptional in lacking stem cells, and the cell cycling described above is dependent on transient amplifying cells migrated from the limbal conjunctiva, the product of stem cells located in large numbers on the limbal basement membrane ( , , , ) but likely also in palisades of Vogt ( ). Stem cells may have a lifelong ability to slowly regenerate, whereas the faster-dividing transient amplifying cells can divide only a limited number of times ( ).
proposed the X,Y,Z theory, which suggests an orderly and continuous death and controlled regeneration of epithelial cells of the corneal surface ( Fig. 3.7 ).
Cells move centripetally from limbus to cornea (Y).
At the same time, basal cells move vertically towards the surface (X).
X plus Y equals the cells desquamating from the corneal surface (Z) ( Fig. 3.7 ).
Intrinsically orchestrated apoptosis is necessary for the development and healthy maintenance of human tissue. found that during normal homeostatic conditions, approximately 1% of cells along the epithelial surface are undergoing apoptosis. This cell shedding is genetically controlled and regulated by a complex signal transduction cascade, which causes alterations in mitochondrial permeability and the execution of downstream death effectors. Apoptotic cell death, unlike necrotic cell death, occurs without concurrent inflammation and is characterised by cell shrinkage, chromatin condensation, cytoplasmic blebbing and nuclear degradation. Receptors in the cell membrane finish the process through phagocytosis.
Studies on rabbits have suggested that the highest incidence of apoptotic cell death occurs at the centre of the cornea ( ). This increased desquamation may be due to the overall centripetal movement of cells.
This is suggested in dark irides where short, linear strands of pigment encroach into the cornea, especially at the lower limbus ( ).
Adrenaline (epinephrine) and sensory denervation by sympathectomy decrease the mitotic rate in rats ( ), presumably as a result of denervation and adrenaline not circulating. In humans, the balance of evidence favours the presence of adrenergic innervation ( , ).
Appropriate regulation of mitosis is important in maintaining corneal epithelial health and that of the cornea as a whole. Sensory denervation also decreases the mitotic rate ( ), delays wound healing and increases epithelial permeability ( ), and it may result in neurotrophic keratitis. showed that corneal lesions induced by neonatal sensory denervation were limited by sympathectomy. This surprising result suggests that a balance in neuronal activity between sympathetic neurones and trigeminal sensory neurones may be critical for maintaining the normal physiology of the cornea.
In a contact lens wearer, fewer cells are being shed from the ocular surface ( , , ). Nitrogen goggle experiments and eyelid suturing suggest that hypoxia is an important factor in inhibiting epithelial shedding ( ). This stagnation in epithelial shedding among contact lens wearers may help explain the increased incidence of microbial corneal infections in these patients. Cells inhibited from programmed desquamation remain longer on the corneal surface, allowing more time for bacterial adhesion or indeed plasmalemmal (cell membrane) changes promoting bacterial adhesion.
Epithelial renewal slows during contact lens wear depending on the oxygen permeability of the lens ( , ). In particular:
basal cells divide more slowly
epithelial cells migrate more slowly
apoptotic epithelial cell death along the corneal surface is inhibited.
Derived from bone marrow with a lifespan of weeks, Langerhans cells migrate from the bloodstream via the conjunctival epithelium to the corneal epithelium, where they are much less common than in skin ( ).
Langerhans cells:
are smaller than epithelial cells
have a limited perikaryon
have numerous long, thin processes
are generally found on or close to the basement membrane ( Fig. 3.1b )
are present at birth throughout the layer and reduce in number away from the limbus so that none is normally present in the centre of the adult cornea ( ).
Langerhans cells are an integral component of the local immune response to microbial and possibly other antigens. In response to various corneal insults their number increases substantially ( ). For instance, overnight wear of contact lenses elevates their density ( ), but exposure to ultraviolet B radiation leads to a loss ( ). They bear HLA-DR antigens and may provide the first signal in host sensitisation leading to corneal graft rejection ( , , ).
If these are smaller than a pinhead, they are covered in about 3 hours by the neighbouring basal cells sending pseudopodia to cover the excavated area ( ). Normal mitosis is inhibited and plays no part in healing. A series of animal studies showed that physical damage caused by contact lenses does not appear to extend beyond the epithelium, thus sparing the epithelial basement membrane and underlying anterior limiting lamina ( , , ).
If a larger area of the cornea is denuded, cells from all layers of the surrounding epithelium migrate and flatten to cover the Wound. Mitosis is at first inhibited but recommences after a few hours and takes an active part in repair. found that, in rabbits, an area 2–3 mm in diameter will become covered within 24 hours, and in three days the area will have a normal appearance as determined by fluorescein staining. The time course of repair is independent of the cause, except in the case of burns, when it is delayed. Other experiments on rabbits have shown that the establishment of a tight adhesion of newly regenerated epithelium takes only a few days if the basement membrane is largely intact, but initially the new epithelium is not firmly attached and is susceptible to damage ( ).
Tight junction reformation is swift ( ) and may reestablish the epithelial barrier well before the layer is firmly attached to the stroma. It is prudent, therefore, to discontinue contact lenses wear for a few days after incurring significant epithelial damage to allow for reestablishment of desmosomes and tight junctions after deep epithelial trauma.
If the lesion lies close to the limbus, conjunctival cells take part in the migration, as may be ascertained clinically by observing the movement into the cornea of pigment from the limbus. Such migration may represent an acceleration of a normal slow centripetal movement of epithelial cells ( ). After total denudation of rabbit corneas, 50% coverage occurs after 24 hours, 75% after 48 hours ( ) and total coverage takes from 4–12 days ( , , ).
Once the cornea is completely covered, the epithelium is one or two cells thick ( ); after 2 weeks it is two to three cells thick ( ), but it takes several weeks before the epithelium is of normal thickness and adhesion structures are adequately reformed ( ). found that normal thickness might be attained in one region of a previously denuded cornea, whereas another area remains uncovered.
In rabbits, polymorphonuclear leucocytes appear in the basal lamina and at the edge of an abrasion, possibly as soon as three hours after epithelial abrasions. These white blood cells reach the cornea via the tears after full-thickness epithelial injuries from trauma or postrefractive surgery ( , ). Epithelial cells bordering the abrasion are flattened and develop surface ruffles and long, fine filopodia at their free edges ( ). These extend to form attachments to the basal lamina, giving the impression of drawing the cells forward into the area of the defect. Epithelial cells bordering a defect appear to increase their water content and surface area, facilitating the production of cell extensions ( ). found that epithelial repair in rabbits slows with age, although only slightly.
showed that wearing low-Dk rigid or thick low-water-content 2-hydroxyethyl methacrylate monomer (HEMA) lenses can, over an extended period, lead to a whittling away of epithelial cells layer by layer until only the basement membrane remains. Another form of epithelial thinning occurs in Ortho-K patients from pressure of the lens onto the ocular surface, leading to flattening of cells ( , , ). The epithelial cells have gel-like properties, and their shape can easily be altered by pressure. In the Ortho-K cornea the epithelium will be thinner where the lens bears and thicker in noncontact areas (see Chapter 19 ). This is a result of cellular compression and the shifting of cells away from areas where the pressure is exerted. A blunt shearing injury, such as a thumb in the eye, will cause the basal cells to be torn apart just beneath their nuclei, leaving the overlying epithelium intact but detached from the underlying cornea. This demonstrates the strength of the epithelium.
The ability of the epithelial cells to remain intact is best explained by the extensive interdigitations between cells and by the vast number of desmosomal cell junctions. The tall columnar basal cells are the weakest. Excessive force causes them to rupture between the basal cell membrane and the nucleus, leaving the basal cell membrane with some internal cytoplasmic fragments attached by hemidesmosomes to the basement membrane. Because the untraumatised basement membrane remains intact, rapid and complete epithelial regeneration can take place. Repair after photorefractive keratectomy is significantly different because the laser ablates the basement membrane, which will have to be resynthesised by the epithelium and reattached to the underlying anterior limiting lamina.
Wound healing may also be slowed by some antibiotics and antimicrobials ( ). High concentrations of fluoroquinolones and peptides slowed cultured corneal epithelial migration, but aminoglycosides had little effect and penicillins none.
The epithelium in wearers of contact lenses made of the HEMA material, especially thicker lenses worn on an extended-wear schedule, is subject to the development of microcysts at the basal cell level ( ). This is an asymptomatic clinical manifestation of altered epithelial metabolism (see Chapter 12 and Fig 12.5 ). Although the exact morphological composition of microcysts is unknown, proposed that they represent interepithelial pockets of cellular breakdown products. In clinical practice, microcysts indicate the level of epithelial hypoxic stress but are not a serious threat to epithelial health. This clinical phenomenon is typically not manifested in silicone hydrogel lens wearers.
If the eye has been bandaged, or if pressure is exerted through the eyelids, the epithelium wrinkles ( ), giving rise to a quickly fading mosaic visible with fluorescein staining. Corneal wrinkling also occurs where the intraocular pressure is very low.
For effects of orthokeratology, see Section 9, Addendum, available at: https://expertconsult.inkling.com/ .
Current silicone hydrogel lenses, especially worn on a daily wear basis, typically do not affect the epithelium along the corneal surface. However, care solutions may challenge the cells, leading to corneal staining.
The anterior limiting lamina is:
a cell-free layer of uniform thickness of about 8.2–10.7 µm in transverse section ( , )
a fine randomly oriented mesh of Type I collagen fibrils seen with electron microscopy ( Figs 3.8 and 3.9 ) terminating at the peripheral extreme of corneal limbus
modified anteriorly where the anchoring filaments, collagen Type VII, bridge the basement membrane and the electron-dense anchoring plaques. This structural arrangement is believed to explain the firm attachment of the epithelial basement membrane to the underlying stromal tissue
penetrated over its whole area by fine unmyelinated sensory nerve fibres, which pass from the stroma to the epithelium, losing their Schwann cell sheaths as they leave the stroma; these are more frequent peripherally (see ‘ Corneal Innervation ’, p. 50 )
relatively tough, as evidenced by frequent epithelial damage without involvement of this layer; if it does become damaged, fibrous scar tissue is laid down resulting in a permanent opacity, although some reduction of the initial scar area usually occurs.
The interface between the ALL and the stroma shows shallow (<1 µm) overlapping in both directions. Collagen fibrils from the ALL project into the stroma, and fine stromal lamellae have been noted to embed and apparently terminate in the ALL ( ).
The stroma:
constitutes 90% of the corneal thickness
gives the cornea its biomechanical strength
is transparent because of its regular structure and absence of blood vessels.
The basic components of the stroma are the lamellae, consisting of parallel collagen fibres, separated by matrix. The lamellae cross each other at various angles while maintaining an overall orientation parallel to the corneal surface. A transverse section of the central human cornea has an average of 242 lamellae ( ), with a rather narrow range of 232–252 lamellae for different individuals ( ). Interestingly, in keratoconus the lamellae count increases due to lamellae breaking down into smaller units, which is called lamellae splitting ( ). Lamellae severed by trauma or surgery are unable to regenerate or reconnect and will remain severed for life. This is why a postradial keratotomy cornea is permanently weakened and a penetrating keratectomy cornea can be dislodged easily from the host eye in a blunt trauma.
The fibrils are buried in a matrix of proteoglycans and have a periodicity characteristic of collagen ( Fig. 3.10 ). Collagen fibrils are of remarkably uniform calibre, and this is important for meeting the corneal need for transparency. In the freshly fixed rabbit cornea, collagen fibrils measure 33 nm with a narrow range of less than 5 nm ( ).
Lamellae in the central cornea have an average thickness of 1.2–2.0 µm ( , ). In the anterior third of the cornea, they run obliquely and become interwoven with each other, whereas posteriorly they are laid down more plainly with one over the other ( Fig. 3.10 ). discovered a diagonal upward and outward bias of the lamellae using polarised light, whereas , using x-ray diffraction, revealed vertical and horizontal orientations. Towards the periphery, some lamellae may lie approximately concentric with the limbus ( , ).
The issue whether lamellae bridge the cornea from limbus to limbus and how much they extend into the sclera has yet to be determined. Most sources claim that stromal lamellae span the full width of the cornea ( , , , , ), but proposed that this may not be the case. explored the stromal anterior limiting lamina interface and provided evidence that at least some anterior lamellae appear to terminate halfway and that the dominant feature of the terminating lamellae was electron-dense formations. These appeared to be a proteinaceous material anchoring the terminating collagen fibres in the stroma. Little or no overlapping was apparent at the stromal–ALL interface.
Lamellar widths are difficult to measure; most are up to 250 µm, but some appear to be in excess of 1 mm. Adjacent lamellae appear to be discrete, but there are occasional slightly oblique branches in the posterior two-thirds of the cornea, connecting one lamella with another. This arrangement explains the ease with which the deep stroma may be split parallel to the surface, as in preparation for corneal lamellar grafts.
What anatomical structure makes the peripheral cornea thicker than its central portion? Recent morphometric analysis of the stromal lamellar architecture has shown that, perhaps surprisingly, it is not an increased number of lamellae ( ). Despite frequently observed lamellar branching, the total number of lamellae in a transverse section remains the same throughout the cornea. However, the posterior lamellae are noticeably thicker at 3.0–3.7 µm on average, with some achieving 10 µm in thickness, whereas anterior to mid stroma lamellae maintain a uniform thickness from peripheral to central cornea. This appears to provide another argument that lamellae do not cross the entire cornea from limbus to limbus.
At the corneoscleral margin, the stromal lamellae undulate, branch and probably interweave. The fibrils of single lamellae remain parallel to each other, which has a profound effect on how much the cornea swells ( ). Fibril diameters vary 10-fold or more ( Fig. 3.11 ), and there is some variation in diameter throughout the sclera. The sclera acts as a clamp to limit corneal swelling; with the sclera removed, a cornea will swell several times greater than its normal thickness. The canal of Schlemm and the corneoscleral meshwork are located more posteriorly; a description of their structure and relationships is beyond the scope of this chapter.
Become a Clinical Tree membership for Full access and enjoy Unlimited articles
If you are a member. Log in here