Reverse-geometry rigid gas-permeable lenses reshape the cornea overnight — non-surgical, reversible myopia correction and pediatric axial-elongation control
Orthokeratology (ortho-k) begins with a precision optical map of the cornea. Placido-disc or Scheimpflug topographers capture tens of thousands of elevation and curvature points across the corneal surface, feeding a fitting algorithm that must reshape a living, self-healing tissue with sub-micron accuracy — all without surgery.
Every ortho-k fit starts with corneal topography — typically a Placido-disc system (concentric illuminated rings reflected off the tear film) or a Scheimpflug rotating-camera system (Pentacam, Orbscan) that reconstructs true elevation data for both anterior and posterior corneal surfaces.
Key measurements extracted: • Flat-K and Steep-K: keratometric power (D) along the flattest and steepest corneal meridians, typically 42.00–46.00 D in normal eyes • Eccentricity (e-value): quantifies how quickly the cornea flattens from center to periphery (normal e≈0.4–0.6); a higher e-value indicates a cornea that will respond more predictably to reverse-geometry molding • Sagittal height (SAG): the depth of the corneal "bowl" measured at a given chord diameter (commonly 10–15 mm), which is the single most important parameter for reverse-geometry lens base-curve selection • Central corneal thickness (CCT): 540–560 µm typically; pachymetry is repeated throughout treatment to monitor epithelial thinning • Manifest refraction and cycloplegic refraction: confirms true refractive error, ruling out accommodative spasm
Candidacy screening also excludes active corneal infection, significant dry eye, keratoconus or irregular astigmatism (detected via topographic irregularity indices), and corneas that are too thin for safe epithelial redistribution.
Unlike a conventional rigid contact lens — which simply approximates corneal curvature — an orthokeratology lens is deliberately built with a base curve flatter than the cornea. This inverted (reverse) curvature relationship is the entire mechanism of action: it is what generates the fluid forces that reshape the epithelium overnight.
A modern overnight ortho-k lens (e.g., Paragon CRT, Euclid Emerald, Wave, Vervey/GOV designs) is defined by four concentric curves, each engineered for a distinct optical or mechanical function:
1. Base curve (BC) — the central optical zone, 5.5–6.5 mm in diameter. Deliberately fit FLATTER than the patient's flat-K by an amount proportional to the correction sought (roughly 0.50–0.75 D of curve flattening per diopter of myopia to be corrected, following the Jessen factor / modified Munnerlyn-type relationships used in fitting software). This flat central fit is what allows the lens to press against and compress the central epithelium.
2. Reverse curve (RC) — immediately peripheral to the base curve, this curve is engineered STEEPER than the base curve (and steeper than the natural corneal periphery) by roughly 4–7 D. Because it is steeper than the tissue beneath it, it does not touch the cornea — instead it vaults over it, creating an annular fluid reservoir 2–3 mm wide that fills with tear fluid trapped by capillary action and lid pressure. This reservoir is the "reverse geometry" that gives the technique its name and is the direct source of the mid-peripheral steepening effect.
3. Alignment curve (AC) — also called the "landing zone," this curve is fit to closely parallel the actual peripheral corneal curvature (essentially aligning with corneal topography 7–9 mm from center). It centers the lens, distributes bearing force evenly, and prevents decentration that would otherwise cause off-axis treatment and induced astigmatism/glare.
4. Peripheral curve (PC) — the outermost curve, flattened relative to the alignment curve to create edge lift (typically 0.6–1.0 mm of rise). This allows tear exchange under the lens with each blink and eyelid movement during sleep, preventing tear stagnation, debris trapping, and hypoxia.
Fitting software (e.g., Paragon CRT's "Corneal Refractive Therapy" algorithm, or Euclid's Emerald design suite) takes the topography map, target refractive correction, and corneal SAG value as inputs and outputs the precise curve radii and zone diameters — often iterated after an initial diagnostic lens trial and follow-up topography.
The base curve is typically flatter than flat-K by roughly (0.5 x diopters of correction) + 0.75D empirically — e.g., a −3.00 D correction target commonly uses a base curve about 2.25–3.00 D flatter than the patient's flat keratometry reading. Different proprietary algorithms (Jessen factor, tangent method) adjust this relationship for each certified lens design.
Once the reverse-geometry lens is inserted and the eyelid closes for sleep, a closed physiological system is created between the posterior lens surface and the corneal epithelium. Within this sealed compartment, tear fluid — normally free to drain — becomes a hydraulic medium that translates lens geometry directly into tissue-shaping mechanical force.
The mechanism combines hydrostatic (fluid pressure) and mechanical (direct compressive) forces acting on the highly deformable corneal epithelium — the only corneal layer thin and plastic enough to be reshaped without cutting:
Central zone — compressive "squeeze film" force: Because the lens base curve is flatter than the cornea, the two surfaces are forced toward parallel contact under lid pressure. The thin tear layer trapped between them (a "squeeze film," 10–15 µm) cannot escape instantaneously — as the eyelid presses the rigid lens down with every blink and by sustained closed-eye pressure, hydrostatic pressure rises in this thin central film and is transmitted directly onto the epithelial surface, encouraging outward (peripheral) redistribution of epithelial cells and thinning of the central epithelium.
Mid-periphery — negative pressure / reservoir effect: Under the reverse curve, the tear reservoir is thicker (15–25 µm) and forms a relative low-pressure pocket compared to the compressed central zone. This pressure differential, combined with the higher fluid volume available, favors epithelial cell accumulation and thickening in this annulus — cells that were pushed centrally migrate to and pool in this reservoir zone over successive nights.
Measured onset: optical coherence tomography (OCT) studies (e.g., Alharbi & Swarbrick, and subsequent epithelial-mapping studies) show detectable central epithelial thinning after a single night of wear (as little as 6–8 hours), with roughly 60–75% of the total treatment effect achieved within the first night and effect continuing to build and stabilize over 7–14 nights as epithelial remodeling reaches steady state.
The stroma and Bowman's layer are essentially unaffected — all measurable change occurs in the epithelium (normally 50–55 µm thick), which is the only refractive-surface layer thin, metabolically active, and mechanically compliant enough to redistribute overnight and regenerate during the day when the lens is removed.
The epithelium is a five-to-seven-cell-layer, self-renewing tissue that completely turns over roughly every 7 days. This constant renewal is what makes ortho-k both effective and fully reversible: cells respond plastically to mechanical cues each night, and the whole effect fades within days to weeks after lens wear stops.
Confocal microscopy and anterior-segment OCT epithelial-thickness mapping studies (Haque, Reinstein, and others) have directly visualized what reverse-geometry molding does at the tissue level:
• Central epithelial thinning: within the ~6 mm optical (treatment) zone, epithelial thickness drops by roughly 5–20 µm depending on correction magnitude — a −3.00 D correction typically requires on the order of 10–14 µm of central thinning, following an approximately linear dose-response relationship (roughly 3–4 µm of central thinning per diopter corrected, though this varies by lens design and individual corneal biomechanics).
• Mid-peripheral thickening (the annular "plus lenticular" zone): directly beneath the reverse-curve reservoir, epithelial thickness increases by 2–8 µm, forming a ring roughly 1–3 mm wide, typically centered 3–4 mm from the visual axis. This annulus is optically critical — see the myopia-control mechanism in Stage 6.
• Stromal thickness: essentially unchanged (within measurement noise of ±2–3 µm) — this is a defining distinction from LASIK/PRK, where the stroma itself is ablated by excimer laser. Ortho-k is purely an epithelial redistribution phenomenon; no tissue is removed, only relocated.
• Cellular mechanism: basal epithelial cells, and to a lesser extent the wing and superficial cell layers, undergo compression-induced migration. Because the epithelium fully renews roughly every 7 days via limbal stem cell-derived basal cell proliferation and centripetal/anterior migration, the mechanical "memory" of lens-induced reshaping must be continuously reinforced by nightly wear — this is precisely why ortho-k effects regress within 1–3 weeks (occasionally faster) once lens wear is discontinued, unlike the permanent tissue removal of refractive surgery.
Clinicians verify ortho-k success not by symptoms alone but by repeat corneal topography, specifically an axial-power difference map (post-treatment map minus baseline map). A well-centered, adequately powered treatment produces an instantly recognizable concentric "bullseye" pattern that has become the signature image of orthokeratology.
A difference (subtractive) topographic map overlays pre-treatment and post-treatment axial power data, color-coded typically with cool colors (blue/green) representing flattening (power loss) and warm colors (red/orange) representing steepening (power gain):
• Central blue zone: the treatment zone, corresponding almost exactly to the lens base curve diameter (~6 mm). Power reduction here directly corresponds to the myopic correction achieved — a well-fit −3.00 D treatment shows roughly 3.00 D of central flattening on axial power maps, consistent with epithelial thinning translating into reduced anterior corneal curvature (flatter cornea = less converging power = correction of myopia's excessive eye length/power mismatch).
• Mid-peripheral red annulus: the relative steepening ring directly beneath where the reverse curve reservoir sat. This ring is functionally a "plus lens" built into the peripheral cornea itself — it is not a side effect but an intentional and clinically important feature (see Stage 6).
• Green transition/alignment zone: corresponds to the lens alignment curve, ideally showing minimal power change, confirming the lens landed correctly on unaltered peripheral cornea.
Centration quality: an off-center bullseye (decentered >0.5 mm from the pupil center) predicts glare, halos, and reduced/irregular correction, and prompts refitting with adjusted lens diameter or a different alignment-curve design. Symmetric, well-centered bullseyes correlate strongly with best visual outcomes and predictable, stable diopter correction — topography becomes the primary objective endpoint, more reliable than subjective vision alone during the adaptation period.
Orthokeratology serves two distinct, clinically validated purposes: same-day, glasses-free unaided vision by directly correcting refractive error, and — increasingly its most important application — slowing the progression of myopia in children by exploiting the very same mid-peripheral steepening ring that appears on the bullseye map.
FDA clearance for overnight corneal reshaping lenses (Paragon CRT, 2002; various Euclid, Wave, and other designs subsequently) covers myopia up to −6.00 D, with best predictability and fastest visual stabilization in the −0.75 to −4.00 D range. Above roughly −4.00 to −5.00 D, treatment becomes progressively less predictable — larger amounts of epithelial redistribution are required, adaptation takes longer (sometimes 2–4 weeks instead of the typical several days to 2 weeks), and residual refractive error is more likely to remain, sometimes requiring adjunctive low-power spectacles for critical tasks.
Typical published outcomes (multiple peer-reviewed case series and trials, e.g. Cheung & Cho, Lipson et al.): approximately 65–75% of low-to-moderate myopes (up to about −4.00 D) achieve unaided visual acuity of 20/20 or better within the first few days, with the great majority achieving 20/40 or better (typical legal driving threshold in many jurisdictions) within the first night or two of wear. Full stabilization of both refraction and topography is typically reached by 7–14 nights of consistent wear, after which nightly (not necessarily consecutive-24-hour) wear maintains the effect.
The mid-peripheral steepening annulus is not merely a fitting byproduct — it is now understood to be the direct mechanism behind ortho-k's myopia-control effect in children, which has become its most important modern clinical application.
In untreated myopic eyes, images focus correctly on the central fovea but the peripheral retinal image typically falls BEHIND the peripheral retina (hyperopic peripheral defocus) because the myopic eye is axially elongated more than its optical periphery. This peripheral hyperopic defocus is believed to be a biological signal promoting further axial (front-to-back) eye growth — the retina "senses" it needs to grow to catch up to the focal plane, particularly implicated via retinal-scleral signaling pathways.
Ortho-k's mid-peripheral steepening ring adds extra plus (converging) power specifically in the peripheral cornea, shifting the peripheral image plane forward — onto or in front of the peripheral retina — converting peripheral hyperopic defocus into peripheral myopic defocus. This is believed to signal the eye to slow, rather than accelerate, axial elongation.
Controlled trial evidence: multiple randomized and case-controlled studies quantify the effect via axial length measurement (IOLMaster biometry), including ROMIO (Retardation of Myopia in Orthokeratology, Cho & Cheung 2012, 2-year RCT) showing roughly 43% less axial elongation in ortho-k wearers versus single-vision spectacle controls; HK-ROMIO and TO-SEE follow-up studies; and additional trials such as those informing the broader "CRAYON"-type pediatric ortho-k literature, collectively reporting axial elongation reductions in the range of roughly 30% to over 60% depending on baseline myopia, age, and treatment zone diameter — meta-analyses (e.g., Sun et al. 2015; Li et al.) place the pooled effect at approximately 45% mean reduction in axial elongation rate versus controls. Smaller treatment-zone lenses (creating a steeper, closer-in defocus ring) have shown even greater relative myopia-control effect in some studies, at some cost to optical quality/night vision.
A 2-year prospective RCT (ROMIO study, Cho & Cheung, Investigative Ophthalmology & Visual Science 2012) found axial elongation of only 0.36 mm in ortho-k-treated children versus 0.63 mm in spectacle-wearing controls — a 43% relative reduction — establishing orthokeratology as one of the most evidence-supported optical interventions for slowing pediatric myopia progression, alongside low-dose atropine and multifocal soft contact lenses.