Think outside the cone: Applying the Gulani Keratoconus Classification

Editor's note: This is the second article in a 2-part series by Arun C. Gulani, MD, MS, and Aaishwariya A. Gulani, MD. In part 1, the authors introduced a keratoconus classification organized around the dominant problem preventing vision (visual, structural, or reconstructive) rather than disease stage. Part 2 applies that framework procedure by procedure, from cross-linking and intracorneal rings to corneal transplantation, phakic lenses, and staged surgical sequencing.

Cross-linking: stabilization is not vision rehabilitation

Keratoconus is not one disease awaiting one procedure. In part 1 of this series, I described how I classified each eye not by stage of disease but by the dominant problem preventing it from seeing, whether visual, structural, or reconstructive (salvage), with the patient's Best Vision Potential as the target. Within that framework, cross-linking, INTACS intracorneal rings, CAIRS, CTAK, laser, lens-based surgery, and transplantation are ingredients rather than competing procedures, and the same question applies to each: does it restore vision, or does it only change anatomy?

Corneal collagen cross-linking is an important advance because its principal purpose is biomechanical stabilization. In appropriately selected progressive keratoconus, it can strengthen corneal collagen and reduce the likelihood of continued ectatic progression. But stabilizing an irregular cornea does not automatically make that cornea optically regular or visually capable. A patient can undergo entirely successful cross-linking and still suffer from less than usable vision including continuing to experience ghosting, glare, halos, and dependence on specialty lenses.

When a patient arrives having been told cross-linking “failed,” the first task is not to relitigate the original procedure. It is to determine, with precision, what remains abnormal: Is the cornea still progressing, or is it stable but optically irregular? Is the residual limitation irregular astigmatism, a scar, residual refractive error, or higher-order aberration? Only once that is separated out does the next step become obvious. (Figure 1)

The evidence supports cross-linking's role as a stabilizing, first-line intervention once progression is confirmed. A randomized controlled trial found that cross-linking halted progression in the substantial majority of young patients with documented progressive keratoconus, reinforcing its place as a first-line treatment rather than a rescue procedure.¹³ Refinements since the original epithelium-off Dresden protocol (accelerated irradiation, transepithelial delivery, and individualized protocols now extending treatment to even thinner corneas) have widened cross-linking's reach to corneas once considered too fragile to treat.¹³ None of these refinements changes the underlying point: cross-linking is a structural intervention, and its success should be measured in arrested progression, not in restored vision.

I describe this during my lectures worldwide as my Kerato-Scoliosis concept: cross-linking an irregular cornea without first correcting its shape is analogous to applying cement to a scoliotic spine and calling it treatment (ensuring that this scoliotic spine does not bend anymore). The cement may well prevent further progression of curvature, but it does not straighten what is already bent (so the spine remains scoliotic in this state). If an eye can be brought toward a better optical shape first through a structural technique, cross-linking becomes the technique that locks that improved shape and structure in, rather than the technique that makes a distorted one permanent.¹ I apply the same Kerato-Scoliosis thinking well beyond primary keratoconus to keratoglobus, pellucid marginal degeneration, and surgically induced ectasia after LASIK, photorefractive keratectomy (PRK), small incision lenticule extraction (SMILE), or radial keratotomy: anywhere a distorted corneal shape is being offered stabilization in place of correction.¹ (Figures 2 and 3)

INTACS and intracorneal rings: why customization matters

Intracorneal ring technologies can meaningfully influence corneal geometry in selected keratoconus and ectasia eyes. But inserting a segment is not, by itself, customized surgery. Keratoconus is frequently asymmetric; the cone may be central, inferior, or markedly displaced, and the astigmatism can carry both regular and irregular components that do not perfectly correspond to the topography map. The number of segments, their size, location, orientation, and relationship to the patient's actual optics all matter: the implant is one component of an individualized plan, not the plan itself. (Figure 4)

Long-term outcome data support the technology itself: 5-year follow-up of Ferrara-type intracorneal ring segments has shown sustained improvement in visual acuity, refractive error, and topography, and follow-up extending as far as 16 years has confirmed durable keratometry stabilization in most eyes.¹³ Those outcomes are real. They remain, however, outcomes of geometry rather than guarantees of optical perfection, which is exactly why the implant must be sized, positioned, and combined according to the individual eye rather than inserted as a standardized step.

In eyes requiring this level of customization, I use what I call my Titratable G-INTAC technique: single or paired micro-INTACS placed according to individualized planning of entry point, axis, alignment, optical zone, and segment thickness, at differential depths and circumferences, executed through femtosecond laser and, in select extreme cases, manual dissection. I refer to this custom management of corneas as Corneal Contour Modification (C²M).¹ The goal is never simply to reduce keratometry or improve a topography printout. The goal is to decrease irregular astigmatism, arrest progression, and build a structural platform with optical manipulation, stable enough to support the visual technique that follows (topographical improvement is always a welcome side effect). (Figures 5–7)

When the first procedure doesn't work: revision, compensatory, and salvage

For eyes that fall into the Reconstructive (Salvage) category above—after failed CAIRS, KeraRing complications, INTACS failure, or CTAK complications—I rely on 3 strategic pathways: revision, compensatory, and salvage.¹ (Table 1; Figures 8–11)

CAIRS and CTAK: new ingredients, not universal answers

CAIRS and CTAK use donor tissue to influence corneal architecture without relying exclusively on tissue removal, an appealing property given that keratoconus is fundamentally a disease of thinning. These are genuinely exciting additions to the armamentarium. But newer does not automatically mean better for a given eye, and no single technology, however promising, should become a default answer. Each remains an ingredient whose value depends entirely on the recipe it is placed into. (Figure 12)

Extreme keratoconus and corneal scarring: extraordinary measurements, extraordinary planning

Perhaps the greatest test of a keratoconus philosophy occurs when standard algorithms begin to run out of answers. I have managed eyes with keratometry of 89.9 D and irregular astigmatism of 23.50 D, reducing astigmatism to 1.40 D, regular enough to undergo precise toric lens cataract surgery to unaided 20/25 vision.¹ (Figure 13)

In another eye, differential depth and zone INTACS decreased irregular astigmatism from 49.00 D to 5.00 D in a scarred cornea that had been relegated to a corneal transplant.¹ (Figure 14)

In another, with 24.7 D of irregular astigmatism, scarring, and stromal thickness of barely 200 µm, I placed compression sutures across the thinnest, de-roofed area and used an INTACS ring in the opposite quadrant to counterbalance that compression, engineering structural stability out of an eye most algorithms would have sent straight to the transplant list.¹ These are not routine cases, and they should never be read as implying that comparable measurements guarantee a comparable outcome in another eye; every eye is unique, and no visual outcome can be promised.

What these cases demonstrate is a principle rather than a result: extraordinary measurements should trigger extraordinary attention to individualized planning, not an automatic end to the conversation. A severely compromised eye frequently requires staged treatment: structure addressed first, optics addressed later, treatment unfolding over months rather than a single encounter.

Corneal scarring adds a further variable, because a scar affects both transparency and geometry. My Corneoplastique philosophy^5,6 centers on determining how much of a patient's visual disability originates from the cone, how much from the scar, how much from refractive error, and how much from irregular astigmatism, before designing a rational sequence. Only after these components are separated can useful native tissue be rehabilitated rather than reflexively replaced.

Can laser vision surgery ever be used in keratoconus?

This question demands precision. Conventional elective LASIK is generally inappropriate in established keratoconus, since additional stromal weakening can worsen biomechanical instability. That does not mean every laser application to every irregular cornea is identical. Our LaZrPlastique concept^4,7,8 involves refraction-driven optical planning in carefully selected irregular and scarred corneas.

Candidacy depends entirely on individual anatomy, stability, remaining tissue, prior surgery, and optical potential.^7,8 A topography map is extraordinarily important. But patients do not walk around looking through their topography maps; they look through their eyes. Topography improvement is a welcome side effect. Vision should be the goal.

This concept is further underscored by the fact that many “successful” topography-guided PRK keratoconus cases including those with corneal scars where topography looks improved but patients were frustrated with poor vision, referred to us have finally been corrected to their vision end zones with LaZrPlastique. (Figures 15 and 16)

Scleral contact lenses: the Superman suit

Rigid gas-permeable and scleral lenses provide vision for many patients with keratoconus by creating a regular optical surface over an irregular cornea, and for some patients that remains an entirely appropriate non-surgical solution. I describe the scleral contact lens, when relied upon indefinitely without addressing the underlying ocular status, as a Superman suit: the 6-pack abs are visible on the front surface while it is worn, but take the suit off at night, and the real 40-inch waist is still there. The lens can mask keratoconus and allow usable vision while the untreated cornea continues to progress underneath, sometimes to the point that transplantation may eventually become unavoidable.

I do not approach lenses or surgery dogmatically. A patient who sees well and is comfortable in a scleral lens should not undergo surgery merely because the technology exists. Conversely, a patient should not be told that lifelong specialty contact lens dependence is the only option without first determining whether the anatomy offers a genuine opportunity for surgical rehabilitation.

In extreme keratoconus cases in which the patient cannot land at an unaided vision endpoint despite surgery, scleral and specialty contact lenses become a valuable partner in helping that patient reach their best vision potential (but having undergone surgery, that cornea is now stable and structurally secure). (Figure 17)

Preserve before replacing: corneal transplantation as a last resort

Corneal transplantation remains an essential, sight-restoring operation, and some advanced eyes genuinely require it. My published algorithm positions transplantation as a last resort when appropriate alternatives remain, rather than an automatic destination for every severe cone.¹ This is not an anti-transplant philosophy. It is preserve before replace. Before replacing a patient's cornea, the operative question is whether the existing tissue retains meaningful potential for structural and optical rehabilitation. When transplantation is genuinely the most appropriate solution, it should be performed as needed. When useful native tissue can reasonably be rehabilitated, that possibility deserves consideration first.

When corneal surgery does become necessary, it should not be viewed as one universal operation. Depending on the depth and nature of disease, options range across different levels of lamellar surgery, deep anterior lamellar keratoplasty, or penetrating keratoplasty, with the appropriate depth dictated by the individual anatomy. Even a successful graft does not necessarily complete the visual journey: a clear graft can still carry substantial astigmatism or refractive error, which is why I treat eventual optical rehabilitation as part of the surgical strategy from the outset, not an afterthought.

Between a ring segment and a full-thickness graft sits an artistically least interventional option that I use more than most surgeons realize is available: a Corneoplastique-based hand lamellar keratoplasty under topical anesthesia, performed in an office-based surgery suite with minimal instrumentation. It removes only the thinned, scarred anterior layers, leaves the patient's own posterior stroma, Descemet membrane, and endothelium untouched, and grafts donor tissue onto that platform to restore thickness and reduce protrusion; the graft is sutured with a Barraquer anti-torque running suture to keep induced astigmatism to a minimum.¹ The eye is not yet visually rehabilitated at that point. It has simply been given a foundation stable enough to support a visual procedure later, rather than being committed straight to full-thickness replacement.

Literature underscores why this sequencing matters. Registry data on penetrating keratoplasty for keratoconus show first-graft survival declining from roughly 89% at 10 years to 49% at 20 years and 17% at 30 years, a reminder that even our most durable current solution carries a survival curve that erodes over a patient's lifetime.¹³ Deep anterior lamellar keratoplasty, which preserves the host's own Descemet membrane and endothelium, carries a meaningfully lower rejection rate than full-thickness transplantation, with multi-surgeon series reporting roughly three-quarters of eyes reaching 20/40 vision or better within a year.¹³ None of this argues against transplantation when it is genuinely needed. It argues for treating transplantation as a considered destination rather than a reflexive one, and for favoring lamellar, tissue-preserving techniques over full-thickness replacement whenever the anatomy allows.

I encourage eye surgeons to learn and apply my Corneoplastique-based anterior hand lamellar keratoplasty, a surgery that can be mastered as an art with the least intervention and be applied in extreme cases of keratoconus to rebuild the structure that then becomes amenable to visual correction.

Keratoconus after transplant: when the graft is clear but vision is not

Some of the most instructive consultations involve patients told their graft looks excellent while their functional vision remains disappointing. These cases underscore the difference between anatomical success and optical success. Residual astigmatism, refractive error, and other optical abnormalities can persist despite a clear, healthy graft. The patient's question is unchanged from before surgery: how well can I see? That question still deserves an optical answer, not simply an anatomical one. (Figures 18 and 19)

Beyond the cornea: ICL and LenzOplastique in the keratoconus eye

For selected patients with keratoconus whose anatomy and refraction are appropriate, a phakic implantable Collamer lens can become part of a visual strategy. An ICL does not treat the underlying keratoconus; it addresses selected refractive components without removing further corneal tissue. As with every technique discussed here, the diagnosis alone does not determine candidacy; the entire optical system must be evaluated. I have placed the full range of ICL technology, from the Visian ICL to the EVO and EVO Toric ICL, in patients with keratoconus in the Visual category, at times combined with LaZrPlastique, INTACS, or cross-linking, and have on occasion corrected ICL complications in keratoconus eyes by elegant removal of a failed implant followed by surgical rehabilitation to a genuine visual endpoint.³

This mirrors the broader literature on phakic intraocular lenses in keratoconus, which has found both iris-fixated and posterior-chamber phakic lenses to offer good safety and predictable refractive outcomes in appropriately selected eyes, with eyes whose ectasia resembles pellucid marginal degeneration reported to benefit particularly well.¹³ Sometimes the most sophisticated plan is the one that touches the cornea least: in a keratoconus eye with high myopia, regular astigmatism, and a cornea too thin to safely manipulate further, I have gone straight to a phakic ICL and left the keratoconic cornea alone entirely, using the eye's own natural lens chamber as a permanent contact lens rather than adding risk to tissue that had none to spare.¹

Patients with keratoconus eventually develop age-related cataracts like everyone else, and cataract surgery in a keratoconus eye presents unique challenges: irregular curvature complicates keratometry and intraocular lens calculation, astigmatism carries both regular and irregular components, and previous cross-linking, rings, CAIRS, CTAK, laser surgeries, or transplantation further influences measurement and planning. My LenzOplastique philosophy⁹ extends the keratoconus strategy beyond the cornea: the lens implant becomes another ingredient in the recipe, intelligently manipulated to work holistically with the corneal component rather than selected in isolation. (Figure 20)

This approach is applicable in stable keratoconus cases: preoperative and post-cataract surgery corneal topography can show similar astigmatism even after that astigmatism has been internally neutralized with a toric lens implant. In another case, in a failed premium IOL cataract surgery, optical manipulation with a piggyback IOL induced a planned refractive error that then allowed surface LaZrPlastique to bring that case to 20/20+ vision without the need to exchange the previous surgeon’s lens implant.

In an extreme Salvage case referred from overseas, a patient was religiously opposed to having a lamellar corneal graft and was ready to wear a scleral lens implant. His demanding eye, with keratometry over 70 D, 9.00 D of irregular astigmatism, a fully scarred 200-µm cornea, and a traumatic, terminal white cataract, was addressed with LenzOplastique-directed cataract surgery to restore functional vision.¹

Staged vision engineering: inside-out and outside-in sequencing

Many complex keratoconus eyes require more than 1 procedure because more than 1 optical factor or structure limits vision. Rather than allowing tradition to dictate which procedure comes first, I individualize the sequence through what I call Outside-In and Inside-Out reconstruction,⁴ organized under my broader kerato-lenticulo-extended refractive (KLEAR) philosophy.¹⁰ Which direction I move depends entirely on which part of the optical system is holding the eye back.

Outside-In: structure before optics

I have described Outside-In and Inside-Out sequencing in detail in a previous Ophthalmology Times article on refractive lens exchange; the same principle applies to keratoconus. An Outside-In sequence is chosen when the cornea itself is the primary obstacle: unstable, scarred, or too irregular to provide reliable measurement. Structure and surface are addressed first, through Corneoplastique-based techniques, cross-linking, intracorneal stromal implant rings, or tissue addition, until the cornea becomes what I call measurable and stable. Only then does lens or refractive surgery become accurate.

A clean example: an eye with 9 associated complexities, where paired G-INTAC placement alone brought keratometry down from 89.9 D and reduced irregular astigmatism from 23.50 D to 1.40 D, followed by a premium toric cataract procedure once the cornea had finally become measurable.¹ The cornea had to move first. Nothing else could be trusted to work until it did. (Figure 21)

Inside-Out: optics before structure

An Inside-Out sequence is chosen when the cornea is stable enough to measure reliably but the lens or refractive error is the dominant limitation: an internal optical platform is established first, and the corneal surface is subsequently refined to work in symbiosis with it. One of many such illustrations is what I call the Orlando-to-Paris-via-Iceland case, which involved a patient left hyperopic, scarred, and irregularly astigmatic after cataract surgery elsewhere, with a further complexity: his previous surgeon had opened the capsular bag, making a straightforward lens exchange less attractive to him than it should have been.

Using my 5S system, I found a single technique that could resolve all 5 problems at once—a vector-derived, myopic LaZrPlastique approach—but the patient's eye was hyperopic, the opposite of what that technique corrects. So, I first added a piggyback lens implant to his existing one, deliberately pushing the eye into myopia (taking him, metaphorically, to Iceland, in the wrong direction, on purpose) so that the myopic LaZrPlastique procedure could then bring him the rest of the way to Paris: 20/20+, unaided, within minutes, without needles or pain.¹ The lens moved first. The cornea finished the job, and the patient regained his pilot’s license. (Figure 22)

When the sequence itself must be reconstructed

The most demanding cases are neither simple Outside-In nor simple Inside-Out; they require reconstructing a sequence that other surgeons already started and left incomplete. I use my Gulani Planning System (GPS)¹⁴ to map these, because the task is not to add one more procedure but to decide which previous interventions to leave alone, which to reverse, and in what order the remaining stages should unfold. In what I call my Orlando-to-California-to-Paris case, involving an eye referred after 12 surgical failures, among them multiple cross-linking, laser PTKs, PRK, INTACS, and ICL insertion and now with a cataract, I kept its intracorneal rings in place, removed his phakic lens elegantly, and the eye was then brought to 20/20− through a staged cataract procedure.¹⁴ I did not undo everything the referring surgeons had done. I picked up the baton where their work left off and carried it, deliberately, the rest of the way. (Figure 23)

The objective across every one of these sequences is the same: to intelligently move a disturbed optical system, stage by stage, toward a visual landing zone; each stage is designed to give the next its greatest opportunity to succeed.

Encouraging surgeons to have more than one answer

When colleagues ask how to evaluate a complex keratoconus referral, or how patients might evaluate them as a keratoconus surgeon, I return to these questions:

1. Does the surgeon have more than one answer?
2. Can that surgeon move fluently between cross-linking, INTACS, KeraRings, Ferrara Rings, CAIRS, and CTAK rather than depending on a single technology?
3. Manage corneal scarring and severe irregular astigmatism?
4. Evaluate a patient who remains visually disabled after cross-linking, or after a previous ring procedure, without reflexively repeating or abandoning it?
5. Perform advanced corneal reconstruction when it is genuinely required, and evaluate real alternatives before recommending transplantation?
6. Rehabilitate vision after a clear graft that still leaves a patient unhappy?
7. And understand the lens and cataract challenges that the same patient with keratoconus will eventually face?

As keratoconus care becomes more technologically rich, the limitations of a one-procedure philosophy become more apparent, not less.

Emerging frontiers: regenerating rather than replacing

The direction of the broader field increasingly validates the preserve-before-replace instinct at a biological level. Early clinical work with a cell-free, double-cross-linked bioengineered corneal implant has shown that native corneal stroma can be reshaped from within without removing existing tissue or placing a single suture, with reported improvements in corneal thickness, keratometry, and vision, including previously contact-lens-intolerant eyes regaining lens tolerance.¹³ Parallel work implanting a patient's own adipose-derived stem cells into the stromal layer of advanced keratoconic corneas has shown 3-year safety, with recovered corneal transparency and modest visual improvement, and no eye losing lines of vision.¹³

Neither approach belongs to routine practice yet, and both need far more follow-up before broad adoption. But their premise, generating and reinforcing the eye's own tissue rather than defaulting to replacement, is precisely the premise this algorithm has been built on for 3 decades. The tools are finally beginning to catch up to our time-tested philosophy.

The individual eye determines the recipe

The future of keratoconus care should not be framed as cross-linking vs INTACS vs CAIRS vs CTAK vs transplantation. These are not competitors. They are tools, and sophisticated keratoconus surgery lies in knowing which tool belongs in which eye, and when.

“Just stabilizing the keratoconus, improving topography, or delaying corneal transplant should certainly not be the only goal. Improving vision while allowing stability and topography improvement to be a welcome side effect is acceptable, while unaided vision should be the relentless goal. Even if impossible, the attempt for that goal should still be there.” — Arun C. Gulani, MD, MS

Whether a patient presents with newly diagnosed keratoconus, progressive disease, an extreme or scarred cornea, keratoglobus, pellucid marginal degeneration, post-LASIK ectasia, failed cross-linking, failed INTACS, previous CAIRS or CTAK, lifelong contact-lens dependence, a prior transplant, or poor vision despite multiple previous treatments elsewhere, the diagnosis should never automatically dictate the procedure.

Keratoconus is the diagnosis. The individual eye determines the recipe. Vision remains the goal.

“Think Outside the Cone” - #Gulanism

This is the second article in a 2-part series by Arun C. Gulani, MD, MS, and Aaishwariya A. Gulani, MD. Read part 1 here.

References
1. Gulani AC, Pandya RP, Gulani AA, Gulani YA. Innovative keratoconus surgical algorithm: a refractive approach to restoring vision. Indian J Cataract Refract Surg. 2024;1(2):101-111. doi:10.4103/ICRS.ICRS_35_24
2. Gulani AC, Probst L, Cox I, Vieth R. Wavefront in LASIK: the Zyoptix platform. Ophthalmol Clin North Am. 2004;17:173-181.
3. Gulani AC, Neumann AC, Fyodorov S. Phakic implants for high myopia: six years follow-up. Invest Ophthalmol Vis Sci. 1997;38:S539. Abstract.
4. Gulani AC. LaZrPlastique: refractive surgery beyond the procedure. The Ophthalmologist. June 2026.
5. Gulani AC. Corneoplastique: art of vision surgery. Indian J Ophthalmol. 2014;62(1):3-4.
6. Gulani A. Corneoplastique. Tech Ophthalmol. 2007;5(1):11-20.
7. Gulani AC, Gulani AA, Gulani YA. LaZrPlastique: pioneering a new era in laser vision surgery for universal application and democratization of refractive surgery. Indian J Cataract Refract Surg. 2025.
8. Gulani AC. LaZrPlastique: the non-cutting edge over LASIK. Ophthalmology Times. 2025.
9. Gulani AC. LenzOplastique: raising cataract surgery to an art, an experience. Ophthalmology Times. 2024;49(10):11-13.
10. Gulani AC. KLEAR: a new concept in refractive surgery. Ocular Surgery News. April 2018.
11. Gulani A. Principles of surgical treatment of irregular astigmatism in unstable corneas. In: Textbook of Irregular Astigmatism Diagnosis and Treatment. 2007.
12. Gulani AC. Shaping the future and reshaping the past: the art of vision surgery. In: Copeland and Afshari’s Principles and Practice of Cornea. 2013.
13. Ozir MA, Hashim SE, Nordin MH, Norizan S, Kairuddin MF. Management of keratoconus – current perspective in this era: a review article. Indian J Clin Exp Ophthalmol. 2026;12(1):31-41. doi:10.18231/j.ijceo.5388.1769843921
14. Gulani AC. Gulani Planning System (GPS): showing it like it is. EyeWorld. 2017.
15. Grzybowski A, McGhee CN. The early history of keratoconus prior to Nottingham’s landmark 1854 treatise on conical cornea: a review. Clin Exp Optom. 2013;96(2):140-145.
16. 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.