Think outside the cone: The Gulani keratoconus classification

For decades, keratoconus has been taught and, too often, treated as a single disease awaiting a single procedure. A patient receives the diagnosis, and the diagnosis is expected to generate the treatment: cross-linking for progression, an intracorneal ring for irregularity, a transplant when nothing else has worked, and specialty contact lenses in between. This sequence feels orderly. In my experience, it is also the single greatest obstacle to visual rehabilitation in keratoconus.

I teach a different mindset, one I have taught for over 3 decades: “Think Outside the Cone.” The mistake is not any one technology. The mistake is treating keratoconus as a diagnosis rather than as an evolving optical system, and treating the cone itself, rather than the eye's visual potential, as the thing to be managed.

Two corneas can carry an identical diagnosis of keratoconus while representing entirely different surgical problems: different cone locations, different depths of thinning, different degrees of scarring, different refractive error, different prior surgery, and, most importantly, different visual potential. A diagnosis describes a category of disease. It does not describe an eye.

Keratoconus is the diagnosis. Vision is the goal.

Keratoconus is a progressive corneal disorder in which the normally smooth cornea thins, steepens, and assumes an irregular, cone-like shape. Because the cornea contributes most of the eye's focusing power, this change in geometry produces ametropia (myopia and sometimes hyperopia), regular and irregular astigmatism, and higher-order aberrations in varying combinations. Patients describe ghosting, halos, glare, monocular diplopia, and a spectacle prescription that seems to change at every visit. As disease advances, glasses lose their ability to correct vision, and patients become dependent upon rigid gas-permeable or scleral lenses.

But keratoconus is not one disease of one severity. It exists along a spectrum extending from mild asymmetric astigmatism to corneas approaching even 90 D of steepness, with regions of stromal thickness under 200 µm and irregular astigmatism exceeding 20 D.¹ That same broader spectrum includes keratoglobus, pellucid marginal degeneration, and surgically induced ectasia following LASIK, SMILE, PRK, radial keratotomy, or hexagonal keratotomy: conditions that resemble keratoconus in appearance while behaving as distinct biomechanical problems.

Even our basic understanding of the disease keeps evolving. The word itself comes from the Greek keras (cornea) and conus (cone), a literal name for a condition first documented, though not yet named, in the eighteenth century: Benedict Duddell described a young patient's corneas as “very prominent, like obtused cones” in 1736, and it was not until 1854 that John Nottingham synthesized a century of scattered observations into the modern definition of keratoconus.²

Our biological understanding has moved just as much. Keratoconus was considered a purely non-inflammatory disease for most of the twentieth century, until research identified elevated interleukin-1, interleukin-6, tumor necrosis factor alpha, and matrix metalloproteinase-9 in keratoconic tears and corneas, alongside oxidative stress and progressive keratocyte apoptosis driving the stromal thinning.¹ Prevalence estimates have shifted too: a national US database study places prevalence at roughly 0.15%, highest in the western states, while population studies consistently find keratoconus several-fold more common in Middle Eastern and Indian populations than in White or East Asian populations.^1,3 None of this changes the disease. It changes how much of it we can now see, and therefore how early we can plan for it.

The question I ask, therefore, is never simply does this eye have keratoconus. It is: what, precisely, is preventing this eye from seeing, and is that problem visual, structural, or both?

Why I never begin with a procedure

“A doctor's inability should never become the patient's disability.” - #Gulanism

Early in my career it would have been easy to approach keratoconus the way our field so often teaches, by mastering one technology and applying it broadly. I chose a different path. Rather than asking which procedure I was best equipped to perform, I asked what an individual eye, with its unique situation, needed in order to see.

That distinction, sustained across 3 decades of corneal, refractive, and lenticular surgery, eventually became a published algorithm rather than remaining a personal habit. In 2024, my coauthors and I published “Innovative Keratoconus Surgical Algorithm: A Refractive Approach to Restoring Vision,” in the Indian Journal of Cataract and Refractive Surgery.¹ The algorithm did not create the philosophy; it simply gave structure and language to an approach I had been practicing since before several of the technologies discussed in this article existed. At its heart is a single target I call a patient's Best Vision Potential (BVP): not the vision their disease allows, but the vision their eye is genuinely capable of, once every optical and structural variable has been accounted for.¹

I teach this approach to fellow surgeons as “Think Outside the Cone.” At its core sits what I call the LEGO Concept: rather than treating keratoconus as one monolithic disease, I deconstruct each eye into its individual optical and structural abnormalities (a thin cornea, a decentered apex, irregular astigmatism, an associated refractive error) and address each piece deliberately, in the sequence that eye requires.¹ A keratoconus diagnosis sounds enormous to a patient and, frankly, to many surgeons. Broken into its components, it becomes a solvable problem rather than an overwhelming one.

The Gulani keratoconus classification: visual, structural, and reconstructive (salvage) keratoconus

Having seen and treated a wide range of keratoconus and corneal ectasia cases from around the world for over 3 decades, I have created an actionable keratoconus classification and surgical algorithm, organized not by stage of disease but by the dominant problem preventing vision.⁴ This classification does not replace clinical staging by topography, tomography, or biomechanical testing. It organizes surgical strategy around what those tests mean for a given eye. (Table 1)

For decades I have maintained my mental framework for every eye: the 5S System (Sight, Shape, Scar, Strength, and Site).⁴ Does the eye have demonstrable visual potential (Sight)? What is its true refractive architecture (Shape)? Is opacity genuinely limiting vision (Scar)? Is the cornea biomechanically capable of supporting reconstruction (Strength)? And where, precisely, does the pathology sit within the optical system (Site)? Running a keratoconus eye through those 5 questions is what turns an intimidating diagnosis into a short, specific list of problems to solve, and it is what generates the classification below.

Visual keratoconus

When an eye retains adequate corneal structure and meaningful visual potential, the objective becomes determining whether its optical abnormalities can be safely rehabilitated. Depending on the individual eye, this may involve refractive strategies including LaZrPlastique, phakic lens technologies such as EVO ICL, or pseudophakic lens-based techniques, without necessarily disturbing corneal architecture that has been confirmed to be measurable and stable and does not need to be disturbed. (Figures 1–3)

Within this category, I assess every eye's candidacy through what I call the LaZrPlastique Candidacy Classification. Level 1 treats laser as the primary intervention: Class A for a clear cornea, Class B for a scarred cornea treated first through Corneoplastique principles. Level 2 treats laser as a staged secondary step: Class A follows prior corneal surgery (INTACS, CAIRS, CTAK, KeraRing, Ferrara Ring, lamellar or penetrating keratoplasty, cross-linking, or conductive keratoplasty), while Class B follows prior intraocular surgery (phakic implantation, or cataract surgery with monofocal, toric, or extended-range IOLs). The classification exists so that a patient with a straightforward clear cornea and a patient with a heavily pre-treated one are never offered the same conversation about what to expect. (Figure 4)

Structural keratoconus

When the cornea itself is the primary obstacle—through progressive thinning, excessive steepening, scarring, or biomechanical instability—anatomy must be addressed before optics can be reliably rehabilitated. The structural pathway can involve cross-linking, intracorneal technologies such as INTACS, KeraRings, or Ferrara Rings, tissue-addition approaches such as corneal allogenic intrastromal ring segments (CAIRS) or CTAK, and, when necessary, different levels of lamellar or penetrating keratoplasty. (Figure 5)

Reconstructive (salvage) keratoconus

The third category is, in my experience, the fastest growing and the most misunderstood. These are patients who have already undergone cross-linking, INTACS, KeraRings, Ferrara Rings, CAIRS, CTAK, topography-guided laser treatment, or corneal transplantation elsewhere, and who remain visually disabled. The surgeon is no longer beginning with untouched anatomy. Previous surgery has changed the cornea, and that altered cornea, not the original diagnosis, must become the new starting point. (Figure 6)

A field moving toward algorithmic, individualized care

A comprehensive 2026 review of keratoconus management surveying current literature traces a comparable evolution in how our field stages and treats this disease.⁵ Prof. Marc Amsler proposed the first keratoconus staging system in 1946, later refined into the Amsler-Krumeich classification still taught today. The 2015 Global Consensus on Keratoconus and Ectatic Diseases moved the field toward multi-parameter diagnostic criteria: abnormal posterior elevation, abnormal corneal thickness distribution, and clinical non-inflammatory thinning, paired with explicit criteria for defining progression.

The Belin ABCD staging system, introduced in 2016 and updated in 2020, went further still, incorporating anterior and posterior surface data, thinnest pachymetry, and visual acuity across 5 stages, and is now regarded by many as superseding the systems that preceded it.⁵ Deshmukh and colleagues built a 2023 treatment algorithm directly on the ABCD progression display, dividing management into progressive and non-progressive pathways, while Sinjab proposed a simpler functional classification the same year, grading disease as mild, moderate, or severe by symptoms and visual acuity rather than imaging alone.⁵

Each of these systems answers an important question: how advanced this disease is. None of them, by itself, answers the question I believe determines the actionable operation: what, specifically, is preventing this eye from seeing, and does the answer point toward optical rehabilitation, structural rehabilitation, or reconstruction of a previously treated eye. My Visual / Structural / Reconstructive framework does not compete with ABCD staging or the Global Consensus criteria; it is the surgical translation from “staging” into “strategy,” to thereby deliver the only ask of every patient: vision.

The ingredient vs recipe principle

Cross-linking, INTACS, CAIRS, CTAK, ICL (STAAR Surgical), LaZrPlastique, and corneal transplantation are not competing procedures. They are ingredients. Having an ingredient does not create a recipe. The expertise lies in deciding which ingredient belongs in which eye, in what magnitude, at what location, and, critically, in what sequence.

This reframes a question I hear constantly from both patients and referring colleagues: which technology is best? CAIRS vs INTACS. CTAK vs CAIRS. Cross-linking vs rings. These comparisons assume one technology must defeat another. Each technology addresses a different component of the same problem. Cross-linking addresses biomechanical stability. Intracorneal implants and tissue-addition techniques influence corneal geometry. Optical rehabilitation addresses residual refractive abnormality. Corneal surgery addresses anatomy no longer suitable for lesser intervention. The useful question is never which procedure is universally best; it is which procedure, or sequence of procedures, this eye requires.

In practice, this has grown into a genuinely large toolkit: across single techniques and their combinations, the algorithm now spans numerous individual keratoconus procedures and unlimited combination strategies.¹ That breadth is not the point in itself; it exists because every eye referred to me has been addressed individually while aspiring to its best visual capacity. (Figures 7 and 8)

Part 2 of this series examines how this approach applies to specific procedures and will be available on October 15, 2026.

References
1. Gulani AC, Pandya RP, Gulani AA, Gulani YA. Innovative keratoconus surgical algorithm: A refractive approach to restoring vision. Indian Journal of Cataract and Refractive Surgery. 2024;1(2):101–111. doi:10.4103/ICRS.ICRS_35_24
2. Grzybowski A, McGhee CN. The early history of keratoconus prior to Nottingham's landmark 1854 treatise on conical cornea: A review. Clinical and Experimental Optometry. 2013;96(2):140–145.
3. Munir SZ, Munir WM, Albrecht J. Estimated prevalence of keratoconus in the United States from a large vision insurance database. Eye & Contact Lens. 2021;47(9):505–510.
4. Gulani AC. LaZrPlastique: Refractive surgery beyond the procedure. The Ophthalmologist. June 2026.
5. Ozir MA, Hashim SE, Nordin MH, Norizan S, Kairuddin MF. Management of keratoconus – current perspective in this era: A review article. Indian Journal of Clinical and Experimental Ophthalmology. 2026;12(1):31–41. doi:10.18231/j.ijceo.5388.1769843921