Stabilise & regularise · Chapters 08 → 09

Cross-linking & topo-guided PTK

Cross-linking freezes progression; topo-guided PTK regularises the surface. Two complementary steps whose entire safety hinges on two numbers: 400 µm of residual stroma and a capped ablation.

Chapter 08

Corneal cross-linking (CXL)

8.1 Principle

Riboflavin (photosensitiser) + UVA 370 nm → ROS generation → covalent cross-links collagen-collagen and collagen-proteoglycan → ↑ biomechanical rigidity and arrest of progression (≈ 90 % stabilisation). CXL stabilises without correcting: it does not restore vision, it prevents worsening.

8.2 Protocols

Cross-linking protocols (reference total dose: 5.4 J/cm²)
Protocol	Parameters	Note
Dresden (standard epi-off)	Riboflavin 0.1 % (dextran) · 3 mW/cm² × 30 min	Reference Longest follow-up, maximum depth of effect.
Accelerated	9 mW × 10 min · 18 mW × ~5 min · 30 mW × 3 min	Reciprocity law (Bunsen-Roscoe) with limits: lower depth efficacy at high intensity (oxygen-limited).
Iontophoresis / epi-on	Facilitated penetration without epithelial removal	Less penetrating/effective, but less morbid.
PACE / pulsed	Customised CXL, pulsed UV cycles	Improved stromal oxygenation.
Theranostic (iontophoresis)	Transepithelial, riboflavin delivered by iontophoresis; real-time monitoring of stromal riboflavin	The UV-A fluence is adapted continuously ("riboflavin score") → personalised, algorithm-driven dose.

Innovation — theranostic CXL & AI-adapted fluence

Theranostic CXL combines iontophoresis (transepithelial riboflavin delivery, epithelium preserved) with a "theranostic" UV-A device that measures stromal riboflavin concentration in real time (hyperspectral analysis, "riboflavin score") and adapts the UV-A fluence continuously — a personalised, algorithm-driven dose. Goal: to offset the poorer riboflavin penetration of transepithelial CXL and estimate cross-linking efficacy during the session. Evaluated in a randomised trial (ARGO protocol) with early one-year real-world data in progressive keratoconus.

8.3 Limits — to know by heart

Minimum stromal pachymetry ≥ 400 µm after de-epithelialisation (endothelial protection against UVA). Thin corneas → hypo-osmolar riboflavin to swell to ≥ 400 µm, or epi-on/adapted.
Keratometry: no strict safety threshold, but decreasing efficacy/predictability in very advanced cones (very high K max).

Contraindications

Cornea < 400 µm not swellable · corneal herpes (reactivation risk) · dense central scar · active infection · pregnancy (relative).

Chapter 09

Topo-guided PTK ± CXL — combined protocols

9.1 Objective

Regularise the anterior surface (reduce irregular astigmatism and coma, improve acuity) — not correct the whole refractive error — while preserving tissue and combining with CXL so as not to destabilise.

9.2 Two ablation logics

t-PTK (trans-epithelial, guided by the epithelial map): exploits epithelial remodelling; the epithelium acts as a smoothing "mask" — trans-epithelial ablation regularises by following the doughnut profile and paradoxically normalises the underlying stromal surface.
Partial topo-guided PRK (Athens / Cretan-type protocols): myopic/topo-guided tissue-saving photoablation, capped.

9.3 Combined protocols

Athens (Kanellopoulos)

Partial topo-guided PRK (max ablation ≈ ≤ 50 µm) + high-fluence CXL in the same session ("same-day").

Cretan

Variant with adapted CXL parameters.

Logic

Regularise first, then cross-link to lock it in.

9.4 Stromal & keratometric limits — imperative

Capped ablation depth: aim for ≤ 50 µm (regularisation goal, not full correction).
Residual stromal bed ≥ 400 µm to allow safe CXL after ablation.
Very steep cones (very high K max) = poor predictability; prefer stabilisation/ICRS over ambitious regularisation.

Absolute rule

Never perform "full" topo-guided PRK on a keratoconus (risk of aggravated ectasia): always partial + CXL.

Clinical case

Illustrated clinical case — stepwise corneal regularization & IOL calculation

69-year-old woman, right eye. Cortico-nuclear cataract, keratoconus for 60 years, rigid contact lenses for 30 years. Work-up performed 4 months after lens discontinuation and 4 months after intracorneal ring segment (ICRS) implantation. The case illustrates a combined strategy: regularize the cornea (intracorneal rings, then topography-guided laser on higher-order aberrations), make IOL calculation reliable (total keratometry TK / EKR integrating the posterior surface), and manage an endothelial graft on an old keratoplasty.

Pentacam Holladay EKR Detail Report and IOLMaster 700 TK (right eye): equivalent keratometry and biometry before/after corneal regularization — **Plate 1 — Total keratometry & biometry for IOL calculation.** Pentacam *Holladay EKR Detail Report* (OD): the **EKR65** (Equivalent K-Reading, which integrates the posterior surface) gives, over the 4.5 mm zone, K1 45.49 D / K2 47.08 D, astigmatism 1.59 D — an axial map with irregular astigmatism and infero-central steepening. Slit-lamp photo (ring reflex) and OPD-Scan aberrometry (high RMS, higher-order aberrations). Below, two *IOLMaster 700* measurements with total keratometry (TK) compared: initially highly astigmatic (K1 45.89 @90 / K2 50.25 @180, ΔK −4.36 D), then after regularization near-zero keratometric astigmatism (TK1 45.19 @136 / TK2 47.92 @46) — IOL calculation becomes reliable. TK / EKR is the key to IOL calculation on an ectatic cornea *(patient data anonymized)*.

Pentacam follow-up (axial, pachymetry, posterior elevation maps): baseline, after intracorneal rings, then after topography-guided laser on HOA — **Plate 2 — Topographic follow-up of stepwise regularization.** Three rows of Pentacam maps (axial / pachymetry / posterior elevation). *Top row*: baseline — inferior cone, K max > 57 D, markedly pathological posterior elevation (+84 µm). *Middle row — intracorneal rings*: partial flattening, reduced steepening. *Bottom row — topography-guided laser centered on higher-order aberrations*: more regular surface, improved posterior elevation. On the right, anterior-segment OCT measuring corneal thickness and ring depth.

Plate 3 — Topography-guided excimer laser (T-CAT) & DMEK on an old graft. Top, DMEK on an old graft (KT): two axial Topolyzer maps (Wavelight Allegretto) before / after, major astigmatism (Sim K ≈ 9–10 D) treated by therapeutic / topography-guided photoablation targeting higher-order aberrations. Bottom, Topography-guided Customized Ablation Treatment (T-CAT): customized ablation profile (max 114.87 µm, central 62.74 µm), intraoperative imaging at 10 / 50 / 90 % and pupil-position / cyclotorsion tracking via eye-tracker.

Take-home — the logic of the case

In long-standing keratoconus that develops a cataract, the sequence rings → topography-guided laser (T-CAT / PTK) → phaco with calculation on TK / EKR regularizes the irregular astigmatism and higher-order aberrations, makes IOL calculation reliable (total keratometry), and treats endothelial decompensation with DMEK even on an old graft. Regularize first, measure second.