Pr Eric E. GabisonOphthalmology · Cornea & refractive · Paris
FR EN
HomePro areaKeratoconus › Keratoconus & ectasias (2026 course)
Course contents ▾
  1. Definitions & nosology
  2. Clinical background & risk
  3. Advanced genetics
  4. Pathophysiology
  5. Topography, tomography & epithelium
  6. Classification & staging
  7. PMD — specifics
  8. Corneal cross-linking (CXL)
  9. Topo-guided PTK ± CXL
  10. Illustrated clinical case
  11. Intracorneal rings (ICRS)
  12. CAIRS (allogeneic)
  13. DALK — deep lamellar
  14. Acute hydrops
  15. Decision algorithm
Corneal ectasias — from risk factors to reconstructive surgery

Keratoconus &
corneal ectasias

A reference course: understanding keratoconus and pellucid marginal degeneration as the failure of a matrix — detecting it at the earliest stage, stabilising it, regularising it, then reconstructing it once transparency is lost.

A single logic runs through the whole spectrum: a matrix that loses its integrity (collagen/LOX genetics + proteolysis + oxidative stress + defective cross-linking, triggered by eye rubbing) → a biomechanical deformation that we detect, stabilise (CXL), regularise (ICRS/CAIRS, PTK) and reconstruct (DALK) once the cornea becomes opaque.
≈ 1/375 to 1/500
Prevalence revised upward with tomography (vs the "classic" 1/2000) — subclinical forms are under-diagnosed. Screened populations.
#1
Eye rubbing is the major mechanical and pro-inflammatory factor, dose-dependent — and the first therapeutic lever: make it stop.
≈ 90 %
Stabilisation rate of progressive keratoconus after cross-linking (riboflavin + UVA) — the cornerstone of all management.
37 eyes
Consecutive double-docking DALK series (DD-DALK, Gabison et al.), Cornea 2023 — big bubble under intraoperative OCT control.
Keratoconus vs PMD — corneal mapsaxial topography · pachymetry · cross-sectionKERATOCONUSPMD (pellucid)inferior steepening“crab-claw”38445056Dthinnest point under apexthin inferior peripheral band400480560µmthin cornea + apical coneprotrusion above the band
Surgical videos · egabison.com

The techniques described in this course (cross-linking, intracorneal rings, double-docking DALK) are illustrated by the site's surgical videos:

▶ Keratoconus (CXL · rings · DALK) ▶ Corneal transplants
Chapter 01

Definitions & nosological framework

Keratoconus (KC) is a progressive corneal ectasia, bilateral but asymmetric, long described as non-inflammatory in the classic sense — a dogma now qualified. It combines stromal thinning and a paracentral or inferior conical protrusion, producing a progressive irregular myopic astigmatism.

Pellucid marginal degeneration (PMD) is an ectasia with an inferior peripheral band of thinning (typically 4 to 8 o'clock, 1-2 mm from the limbus), the corneal protrusion sitting above the thinnest zone, generating a strong against-the-rule astigmatism.

The ectasia spectrum

Around these two entities gravitate Terrien marginal degeneration (inflammatory, vascularised), keratoglobus (diffuse global thinning) and iatrogenic post-LASIK/PRK ectasia. All share a biomechanical failure of the stromal framework; their distinction rests on the topography of the thinning.

Key distinction — inferior KC vs PMD

The "crab-claw" topography is seen in both: it is the pachymetry map that decides. In keratoconus the thinnest point is para-central, under the apex; in PMD it is inferior peripheral, with the protrusion sitting above the thin band.

Chapter 02

Clinical background & risk factors

Epidemiology

The classic prevalence of 1/2000 has been revised upward (≈ 1/375 to 1/500 in some tomography-screened populations), because subclinical forms are under-diagnosed. Keratoconus begins at puberty, progresses into the 3rd-4th decade, then tends to stabilise — a slowing attributed to age-related natural cross-linking and glycation. Earlier, more severe forms predominate in populations with high consanguinity, high sun exposure (UV) and high atopy prevalence.

Risk factors — the true "background"

Risk factors and their clinical relevance
FactorMechanismPractical consequence
Eye rubbingMechanical + pro-inflammatory trauma, dose-dependent, often asymmetricLever #1 Explains asymmetry and the sleeping side; must be stopped systematically.
AtopyEczema, asthma, rhinoconjunctivitis, vernal keratoconjunctivitis (VKC)Surface inflammation + rubbing: a self-sustaining loop.
Genetic/familial backgroundFamily history, polygenic inheritanceQuestion relatives + tomographic screening.
SyndromicDown syndrome, Leber congenital amaurosis, Ehlers-Danlos, Marfan, mitral prolapse, Brittle Cornea Syndrome (ZNF469/PRDM5)Targeted screening; sometimes early and severe.
Floppy eyelid, sleep apnoeaSide-sleeping position, eyelid laxityLook for OSAS; correct sleep posture.
HormonalFlares at puberty, during pregnancyCloser monitoring during hormonally sensitive periods.

Clinical signs

Functional: progressive visual loss, unstable evolving astigmatism, halos, monocular diplopia, growing spectacle intolerance. Placido/retro-illumination signs: "scissors" reflex on retinoscopy, Charleux "oil-droplet" shadow on fundus examination.

Slit-lamp: Fleischer ring (ferritin deposits at the base of the cone), Vogt striae (deep vertical stromal folds, disappearing on pressure), apical thinning, Munson sign (V-shaped deformation of the lower lid on downgaze), Rizzuti sign (nasal focusing of the limbal beam), apical scars (advanced forms), and acute hydrops from Descemet rupture (see dedicated chapter).

Clinical message

Any progressive astigmatism, any unusual against-the-rule or oblique astigmatism, any atopic patient or eye-rubber → tomography + epithelial mapping, even with preserved acuity. Subclinical forms are clinically silent.

Chapter 03

Advanced genetics

Mode of transmission

Keratoconus is mostly sporadic; 6 to 10 % of familial forms. Inheritance is complex, polygenic, with incomplete penetrance and variable expressivity, and a strong multifactorial component: genes × rubbing × atopy × UV.

Candidate genes and loci

ZNF469
The most robust signal; regulator of collagen synthesis; mutated in Brittle Cornea Syndrome; associated with corneal thinness (CCT).
VSX1
Historical; modest and debated contribution.
LOX
Lysyl oxidase, the enzyme of collagen cross-linking; its under-expression links genetics directly to pathophysiology (defective endogenous cross-linking).
COL4A3/4A4/5A1
Structural collagens; overlap with the loci of central corneal thickness.
Others
TGFBI, SOD1 (oxidative stress), MIR184 (microRNA, syndromic form with anterior cataract), DOCK9, HGF, RAB3GAP1, CAST, FNDC3B, WNT10A, IL-1 cluster.
Focus — PPIP5K2, a gene with substantiated causality

Identified by exome/genome sequencing in a 4-generation KC family (Khaled, Bykhovskaya… Rabinowitz & Liu, Sci Rep 2019): the missense variant rs35471301, p.Ser419Ala co-segregates with keratoconus, on a highly conserved residue. PPIP5K2 is a bifunctional kinase/phosphatase of inositol pyrophosphates (InsP7/InsP8).

Causal validation by mouse model: the gene-trap Ppip5k2 mouse develops irregular corneal surfaces and stromal thinning reproducing the KC phenotype (confirmed 2024, IOVS, Akoto et al., across three lines): localised "KC-like" thinning in the heterozygote → diffuse "keratoglobus-like" in the homozygote. One of the few KC genes whose causality is supported by a concordant animal model.

GWAS & practical relevance

Genome-wide association studies show a strong overlap with the genes determining CCT (FOXO1, FNDC3B, RXRA-COL5A1, MPDZ-NFIB…): keratoconus appears as the "extreme" of a continuum of corneal thinning in a predisposed subject. In practice, genetic counselling and screening of relatives today rest on tomography rather than routine molecular testing; the LOX/cross-linking, proteolytic (MMP/TIMP) and oxidative-stress axes outline future therapeutic targets.

Chapter 04

Pathophysiology — the mechanistic core

Three interdependent pillars converge toward loss of structural integrity of the stroma: proteolytic imbalance, oxidative stress, defective cross-linking — all mechanically amplified by rubbing.

1. Proteolytic imbalance (front line)

  • MMP over-expression: MMP-1 (collagenase), MMP-2 and MMP-9 (gelatinases), MMP-13 → degradation of collagen I/III and matrix.
  • Fall in inhibitors: ↓ TIMP-1 (even TIMP-3) → increased net gelatinase activity.
  • Cathepsins (B, G, V/L) and other serine proteases increased.
  • Epithelium→stroma loop: rubbing stimulates pro-inflammatory mediators (IL-1) and stromal MMP synthesis by keratocytes → localised matrix degradation at the cone apex.

2. Micro-inflammation (the "non-inflammatory" dogma revised)

↑ IL-6, TNF-α, IL-1, MMP-9 in tears and at the surface; IL-1-mediated keratocyte apoptosis, reduced keratocyte density. Keratoconus is not a keratitis, but a surface low-grade inflammation sustains the proteolytic cascade.

3. Oxidative stress & defective cross-linking

↑ ROS, ↓ antioxidant defences (SOD1, ALDH), mitochondrial DNA damage, accumulation of nitrotyrosine and lipid peroxidation products; ROS activate pro-MMPs (synergy). In parallel, ↓ LOX activity leads to insufficient collagen cross-linking, with proteoglycan abnormalities and disorganisation of fibril diameter and spacing.

4. Biomechanics — the measurable translation

The ORA shows ↓ corneal hysteresis (CH) and ↓ corneal resistance factor (CRF). The Corvis ST translates softening into dynamic indices: SSI (Stress-Strain Index), CBI (Corvis Biomechanical Index) and above all TBI (Tomographic-Biomechanical Index by Vinciguerra/Ambrósio), the best subclinical-detection tool by combining tomography and biomechanics.

Biomechanical vicious circle

thinning → ↑ local stress → protrusion → thinning: a self-aggravation that only cross-linking (CXL) durably interrupts.

Unified mental model

genetic predisposition (LOX/collagen/CCT) + rubbing/atopy → surface inflammation (IL-1) → ↑MMP/↓TIMP + ROS → proteolysis + defective cross-linking → focal thinning → biomechanical failure → ectasia.